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 Alpha options:}
256 [@b{-mdebug} | @b{-no-mdebug}]
257 [@b{-replace} | @b{-noreplace}]
258 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
259 [@b{-F}] [@b{-32addr}]
263 @emph{Target ARC options:}
269 @emph{Target ARM options:}
270 @c Don't document the deprecated options
271 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
272 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
273 [@b{-mfpu}=@var{floating-point-format}]
274 [@b{-mfloat-abi}=@var{abi}]
275 [@b{-meabi}=@var{ver}]
278 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
279 @b{-mapcs-reentrant}]
280 [@b{-mthumb-interwork}] [@b{-k}]
284 @emph{Target Blackfin options:}
285 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
292 @emph{Target CRIS options:}
293 [@b{--underscore} | @b{--no-underscore}]
295 [@b{--emulation=criself} | @b{--emulation=crisaout}]
296 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
297 @c Deprecated -- deliberately not documented.
302 @emph{Target D10V options:}
307 @emph{Target D30V options:}
308 [@b{-O}|@b{-n}|@b{-N}]
312 @emph{Target EPIPHANY options:}
313 [@b{-mepiphany}|@b{-mepiphany16}]
317 @emph{Target H8/300 options:}
321 @c HPPA has no machine-dependent assembler options (yet).
325 @emph{Target i386 options:}
326 [@b{--32}|@b{--n32}|@b{--64}] [@b{-n}]
327 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
331 @emph{Target i960 options:}
332 @c see md_parse_option in tc-i960.c
333 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
335 [@b{-b}] [@b{-no-relax}]
339 @emph{Target IA-64 options:}
340 [@b{-mconstant-gp}|@b{-mauto-pic}]
341 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
343 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
344 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
345 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
346 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
350 @emph{Target IP2K options:}
351 [@b{-mip2022}|@b{-mip2022ext}]
355 @emph{Target M32C options:}
356 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
360 @emph{Target M32R options:}
361 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
366 @emph{Target M680X0 options:}
367 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
371 @emph{Target M68HC11 options:}
372 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}]
373 [@b{-mshort}|@b{-mlong}]
374 [@b{-mshort-double}|@b{-mlong-double}]
375 [@b{--force-long-branches}] [@b{--short-branches}]
376 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
377 [@b{--print-opcodes}] [@b{--generate-example}]
381 @emph{Target MCORE options:}
382 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
383 [@b{-mcpu=[210|340]}]
386 @emph{Target MICROBLAZE options:}
387 @c MicroBlaze has no machine-dependent assembler options.
391 @emph{Target MIPS options:}
392 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
393 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
394 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
395 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
396 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
397 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
398 [@b{-mips64}] [@b{-mips64r2}]
399 [@b{-construct-floats}] [@b{-no-construct-floats}]
400 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
401 [@b{-mips16}] [@b{-no-mips16}]
402 [@b{-mmicromips}] [@b{-mno-micromips}]
403 [@b{-msmartmips}] [@b{-mno-smartmips}]
404 [@b{-mips3d}] [@b{-no-mips3d}]
405 [@b{-mdmx}] [@b{-no-mdmx}]
406 [@b{-mdsp}] [@b{-mno-dsp}]
407 [@b{-mdspr2}] [@b{-mno-dspr2}]
408 [@b{-mmt}] [@b{-mno-mt}]
409 [@b{-mmcu}] [@b{-mno-mcu}]
410 [@b{-mfix7000}] [@b{-mno-fix7000}]
411 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
412 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
413 [@b{-mdebug}] [@b{-no-mdebug}]
414 [@b{-mpdr}] [@b{-mno-pdr}]
418 @emph{Target MMIX options:}
419 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
420 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
421 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
422 [@b{--linker-allocated-gregs}]
426 @emph{Target PDP11 options:}
427 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
428 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
429 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
433 @emph{Target picoJava options:}
438 @emph{Target PowerPC options:}
440 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
441 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
442 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
443 @b{-mbooke}|@b{-mpower4}|@b{-mpr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
444 @b{-mpower7}|@b{-mpw7}|@b{-ma2}|@b{-mcell}|@b{-mspe}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
445 [@b{-many}] [@b{-maltivec}|@b{-mvsx}]
446 [@b{-mregnames}|@b{-mno-regnames}]
447 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
448 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
449 [@b{-msolaris}|@b{-mno-solaris}]
450 [@b{-nops=@var{count}}]
454 @emph{Target RX options:}
455 [@b{-mlittle-endian}|@b{-mbig-endian}]
456 [@b{-m32bit-ints}|@b{-m16bit-ints}]
457 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
461 @emph{Target s390 options:}
462 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
463 [@b{-mregnames}|@b{-mno-regnames}]
464 [@b{-mwarn-areg-zero}]
468 @emph{Target SCORE options:}
469 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
470 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
471 [@b{-march=score7}][@b{-march=score3}]
472 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
476 @emph{Target SPARC options:}
477 @c The order here is important. See c-sparc.texi.
478 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
479 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
480 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
485 @emph{Target TIC54X options:}
486 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
487 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
492 @emph{Target TIC6X options:}
493 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
494 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
495 [@b{-mpic}|@b{-mno-pic}]
499 @emph{Target TILE-Gx options:}
500 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
503 @c TILEPro has no machine-dependent assembler options
508 @emph{Target Xtensa options:}
509 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
510 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
511 [@b{--[no-]transform}]
512 [@b{--rename-section} @var{oldname}=@var{newname}]
517 @emph{Target Z80 options:}
518 [@b{-z80}] [@b{-r800}]
519 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
520 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
521 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
522 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
523 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
524 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
528 @c Z8000 has no machine-dependent assembler options
537 @include at-file.texi
540 Turn on listings, in any of a variety of ways:
544 omit false conditionals
547 omit debugging directives
550 include general information, like @value{AS} version and options passed
553 include high-level source
559 include macro expansions
562 omit forms processing
568 set the name of the listing file
571 You may combine these options; for example, use @samp{-aln} for assembly
572 listing without forms processing. The @samp{=file} option, if used, must be
573 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
576 Begin in alternate macro mode.
578 @xref{Altmacro,,@code{.altmacro}}.
581 @item --compress-debug-sections
582 Compress DWARF debug sections using zlib. The debug sections are renamed
583 to begin with @samp{.zdebug}, and the resulting object file may not be
584 compatible with older linkers and object file utilities.
586 @item --nocompress-debug-sections
587 Do not compress DWARF debug sections. This is the default.
590 Ignored. This option is accepted for script compatibility with calls to
593 @item --debug-prefix-map @var{old}=@var{new}
594 When assembling files in directory @file{@var{old}}, record debugging
595 information describing them as in @file{@var{new}} instead.
597 @item --defsym @var{sym}=@var{value}
598 Define the symbol @var{sym} to be @var{value} before assembling the input file.
599 @var{value} must be an integer constant. As in C, a leading @samp{0x}
600 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
601 value. The value of the symbol can be overridden inside a source file via the
602 use of a @code{.set} pseudo-op.
605 ``fast''---skip whitespace and comment preprocessing (assume source is
610 Generate debugging information for each assembler source line using whichever
611 debug format is preferred by the target. This currently means either STABS,
615 Generate stabs debugging information for each assembler line. This
616 may help debugging assembler code, if the debugger can handle it.
619 Generate stabs debugging information for each assembler line, with GNU
620 extensions that probably only gdb can handle, and that could make other
621 debuggers crash or refuse to read your program. This
622 may help debugging assembler code. Currently the only GNU extension is
623 the location of the current working directory at assembling time.
626 Generate DWARF2 debugging information for each assembler line. This
627 may help debugging assembler code, if the debugger can handle it. Note---this
628 option is only supported by some targets, not all of them.
630 @item --size-check=error
631 @itemx --size-check=warning
632 Issue an error or warning for invalid ELF .size directive.
635 Print a summary of the command line options and exit.
638 Print a summary of all target specific options and exit.
641 Add directory @var{dir} to the search list for @code{.include} directives.
644 Don't warn about signed overflow.
647 @ifclear DIFF-TBL-KLUGE
648 This option is accepted but has no effect on the @value{TARGET} family.
650 @ifset DIFF-TBL-KLUGE
651 Issue warnings when difference tables altered for long displacements.
656 Keep (in the symbol table) local symbols. These symbols start with
657 system-specific local label prefixes, typically @samp{.L} for ELF systems
658 or @samp{L} for traditional a.out systems.
663 @item --listing-lhs-width=@var{number}
664 Set the maximum width, in words, of the output data column for an assembler
665 listing to @var{number}.
667 @item --listing-lhs-width2=@var{number}
668 Set the maximum width, in words, of the output data column for continuation
669 lines in an assembler listing to @var{number}.
671 @item --listing-rhs-width=@var{number}
672 Set the maximum width of an input source line, as displayed in a listing, to
675 @item --listing-cont-lines=@var{number}
676 Set the maximum number of lines printed in a listing for a single line of input
679 @item -o @var{objfile}
680 Name the object-file output from @command{@value{AS}} @var{objfile}.
683 Fold the data section into the text section.
685 @kindex --hash-size=@var{number}
686 Set the default size of GAS's hash tables to a prime number close to
687 @var{number}. Increasing this value can reduce the length of time it takes the
688 assembler to perform its tasks, at the expense of increasing the assembler's
689 memory requirements. Similarly reducing this value can reduce the memory
690 requirements at the expense of speed.
692 @item --reduce-memory-overheads
693 This option reduces GAS's memory requirements, at the expense of making the
694 assembly processes slower. Currently this switch is a synonym for
695 @samp{--hash-size=4051}, but in the future it may have other effects as well.
698 Print the maximum space (in bytes) and total time (in seconds) used by
701 @item --strip-local-absolute
702 Remove local absolute symbols from the outgoing symbol table.
706 Print the @command{as} version.
709 Print the @command{as} version and exit.
713 Suppress warning messages.
715 @item --fatal-warnings
716 Treat warnings as errors.
719 Don't suppress warning messages or treat them as errors.
728 Generate an object file even after errors.
730 @item -- | @var{files} @dots{}
731 Standard input, or source files to assemble.
739 @xref{Alpha Options}, for the options available when @value{AS} is configured
740 for an Alpha processor.
745 The following options are available when @value{AS} is configured for an Alpha
749 @include c-alpha.texi
750 @c ended inside the included file
757 The following options are available when @value{AS} is configured for
762 This option selects the core processor variant.
764 Select either big-endian (-EB) or little-endian (-EL) output.
769 The following options are available when @value{AS} is configured for the ARM
773 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
774 Specify which ARM processor variant is the target.
775 @item -march=@var{architecture}[+@var{extension}@dots{}]
776 Specify which ARM architecture variant is used by the target.
777 @item -mfpu=@var{floating-point-format}
778 Select which Floating Point architecture is the target.
779 @item -mfloat-abi=@var{abi}
780 Select which floating point ABI is in use.
782 Enable Thumb only instruction decoding.
783 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
784 Select which procedure calling convention is in use.
786 Select either big-endian (-EB) or little-endian (-EL) output.
787 @item -mthumb-interwork
788 Specify that the code has been generated with interworking between Thumb and
791 Specify that PIC code has been generated.
799 @xref{Blackfin Options}, for the options available when @value{AS} is
800 configured for the Blackfin processor family.
805 The following options are available when @value{AS} is configured for
806 the Blackfin processor family.
810 @c ended inside the included file
817 See the info pages for documentation of the CRIS-specific options.
821 The following options are available when @value{AS} is configured for
824 @cindex D10V optimization
825 @cindex optimization, D10V
827 Optimize output by parallelizing instructions.
832 The following options are available when @value{AS} is configured for a D30V
835 @cindex D30V optimization
836 @cindex optimization, D30V
838 Optimize output by parallelizing instructions.
842 Warn when nops are generated.
844 @cindex D30V nops after 32-bit multiply
846 Warn when a nop after a 32-bit multiply instruction is generated.
852 The following options are available when @value{AS} is configured for the
853 Adapteva EPIPHANY series.
856 @xref{Epiphany Options}, for the options available when @value{AS} is
857 configured for an Epiphany processor.
862 The following options are available when @value{AS} is configured for
863 an Epiphany processor.
866 @include c-epiphany.texi
867 @c ended inside the included file
875 @xref{i386-Options}, for the options available when @value{AS} is
876 configured for an i386 processor.
881 The following options are available when @value{AS} is configured for
886 @c ended inside the included file
893 The following options are available when @value{AS} is configured for the
894 Intel 80960 processor.
897 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
898 Specify which variant of the 960 architecture is the target.
901 Add code to collect statistics about branches taken.
904 Do not alter compare-and-branch instructions for long displacements;
911 The following options are available when @value{AS} is configured for the
917 Specifies that the extended IP2022 instructions are allowed.
920 Restores the default behaviour, which restricts the permitted instructions to
921 just the basic IP2022 ones.
927 The following options are available when @value{AS} is configured for the
928 Renesas M32C and M16C processors.
933 Assemble M32C instructions.
936 Assemble M16C instructions (the default).
939 Enable support for link-time relaxations.
942 Support H'00 style hex constants in addition to 0x00 style.
948 The following options are available when @value{AS} is configured for the
949 Renesas M32R (formerly Mitsubishi M32R) series.
954 Specify which processor in the M32R family is the target. The default
955 is normally the M32R, but this option changes it to the M32RX.
957 @item --warn-explicit-parallel-conflicts or --Wp
958 Produce warning messages when questionable parallel constructs are
961 @item --no-warn-explicit-parallel-conflicts or --Wnp
962 Do not produce warning messages when questionable parallel constructs are
969 The following options are available when @value{AS} is configured for the
970 Motorola 68000 series.
975 Shorten references to undefined symbols, to one word instead of two.
977 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
978 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
979 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
980 Specify what processor in the 68000 family is the target. The default
981 is normally the 68020, but this can be changed at configuration time.
983 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
984 The target machine does (or does not) have a floating-point coprocessor.
985 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
986 the basic 68000 is not compatible with the 68881, a combination of the
987 two can be specified, since it's possible to do emulation of the
988 coprocessor instructions with the main processor.
990 @item -m68851 | -mno-68851
991 The target machine does (or does not) have a memory-management
992 unit coprocessor. The default is to assume an MMU for 68020 and up.
999 For details about the PDP-11 machine dependent features options,
1000 see @ref{PDP-11-Options}.
1003 @item -mpic | -mno-pic
1004 Generate position-independent (or position-dependent) code. The
1005 default is @option{-mpic}.
1008 @itemx -mall-extensions
1009 Enable all instruction set extensions. This is the default.
1011 @item -mno-extensions
1012 Disable all instruction set extensions.
1014 @item -m@var{extension} | -mno-@var{extension}
1015 Enable (or disable) a particular instruction set extension.
1018 Enable the instruction set extensions supported by a particular CPU, and
1019 disable all other extensions.
1021 @item -m@var{machine}
1022 Enable the instruction set extensions supported by a particular machine
1023 model, and disable all other extensions.
1029 The following options are available when @value{AS} is configured for
1030 a picoJava processor.
1034 @cindex PJ endianness
1035 @cindex endianness, PJ
1036 @cindex big endian output, PJ
1038 Generate ``big endian'' format output.
1040 @cindex little endian output, PJ
1042 Generate ``little endian'' format output.
1048 The following options are available when @value{AS} is configured for the
1049 Motorola 68HC11 or 68HC12 series.
1053 @item -m68hc11 | -m68hc12 | -m68hcs12
1054 Specify what processor is the target. The default is
1055 defined by the configuration option when building the assembler.
1058 Specify to use the 16-bit integer ABI.
1061 Specify to use the 32-bit integer ABI.
1063 @item -mshort-double
1064 Specify to use the 32-bit double ABI.
1067 Specify to use the 64-bit double ABI.
1069 @item --force-long-branches
1070 Relative branches are turned into absolute ones. This concerns
1071 conditional branches, unconditional branches and branches to a
1074 @item -S | --short-branches
1075 Do not turn relative branches into absolute ones
1076 when the offset is out of range.
1078 @item --strict-direct-mode
1079 Do not turn the direct addressing mode into extended addressing mode
1080 when the instruction does not support direct addressing mode.
1082 @item --print-insn-syntax
1083 Print the syntax of instruction in case of error.
1085 @item --print-opcodes
1086 print the list of instructions with syntax and then exit.
1088 @item --generate-example
1089 print an example of instruction for each possible instruction and then exit.
1090 This option is only useful for testing @command{@value{AS}}.
1096 The following options are available when @command{@value{AS}} is configured
1097 for the SPARC architecture:
1100 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1101 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1102 Explicitly select a variant of the SPARC architecture.
1104 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1105 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1107 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1108 UltraSPARC extensions.
1110 @item -xarch=v8plus | -xarch=v8plusa
1111 For compatibility with the Solaris v9 assembler. These options are
1112 equivalent to -Av8plus and -Av8plusa, respectively.
1115 Warn when the assembler switches to another architecture.
1120 The following options are available when @value{AS} is configured for the 'c54x
1125 Enable extended addressing mode. All addresses and relocations will assume
1126 extended addressing (usually 23 bits).
1127 @item -mcpu=@var{CPU_VERSION}
1128 Sets the CPU version being compiled for.
1129 @item -merrors-to-file @var{FILENAME}
1130 Redirect error output to a file, for broken systems which don't support such
1131 behaviour in the shell.
1136 The following options are available when @value{AS} is configured for
1137 a @sc{mips} processor.
1141 This option sets the largest size of an object that can be referenced
1142 implicitly with the @code{gp} register. It is only accepted for targets that
1143 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1145 @cindex MIPS endianness
1146 @cindex endianness, MIPS
1147 @cindex big endian output, MIPS
1149 Generate ``big endian'' format output.
1151 @cindex little endian output, MIPS
1153 Generate ``little endian'' format output.
1165 Generate code for a particular @sc{mips} Instruction Set Architecture level.
1166 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1167 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1168 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1169 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips64}, and
1171 correspond to generic
1172 @samp{MIPS V}, @samp{MIPS32}, @samp{MIPS32 Release 2}, @samp{MIPS64},
1173 and @samp{MIPS64 Release 2}
1174 ISA processors, respectively.
1176 @item -march=@var{CPU}
1177 Generate code for a particular @sc{mips} cpu.
1179 @item -mtune=@var{cpu}
1180 Schedule and tune for a particular @sc{mips} cpu.
1184 Cause nops to be inserted if the read of the destination register
1185 of an mfhi or mflo instruction occurs in the following two instructions.
1189 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1190 section instead of the standard ELF .stabs sections.
1194 Control generation of @code{.pdr} sections.
1198 The register sizes are normally inferred from the ISA and ABI, but these
1199 flags force a certain group of registers to be treated as 32 bits wide at
1200 all times. @samp{-mgp32} controls the size of general-purpose registers
1201 and @samp{-mfp32} controls the size of floating-point registers.
1205 Generate code for the MIPS 16 processor. This is equivalent to putting
1206 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1207 turns off this option.
1210 @itemx -mno-micromips
1211 Generate code for the microMIPS processor. This is equivalent to putting
1212 @code{.set micromips} at the start of the assembly file. @samp{-mno-micromips}
1213 turns off this option. This is equivalent to putting @code{.set nomicromips}
1214 at the start of the assembly file.
1217 @itemx -mno-smartmips
1218 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1219 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1220 @samp{-mno-smartmips} turns off this option.
1224 Generate code for the MIPS-3D Application Specific Extension.
1225 This tells the assembler to accept MIPS-3D instructions.
1226 @samp{-no-mips3d} turns off this option.
1230 Generate code for the MDMX Application Specific Extension.
1231 This tells the assembler to accept MDMX instructions.
1232 @samp{-no-mdmx} turns off this option.
1236 Generate code for the DSP Release 1 Application Specific Extension.
1237 This tells the assembler to accept DSP Release 1 instructions.
1238 @samp{-mno-dsp} turns off this option.
1242 Generate code for the DSP Release 2 Application Specific Extension.
1243 This option implies -mdsp.
1244 This tells the assembler to accept DSP Release 2 instructions.
1245 @samp{-mno-dspr2} turns off this option.
1249 Generate code for the MT Application Specific Extension.
1250 This tells the assembler to accept MT instructions.
1251 @samp{-mno-mt} turns off this option.
1255 Generate code for the MCU Application Specific Extension.
1256 This tells the assembler to accept MCU instructions.
1257 @samp{-mno-mcu} turns off this option.
1259 @item --construct-floats
1260 @itemx --no-construct-floats
1261 The @samp{--no-construct-floats} option disables the construction of
1262 double width floating point constants by loading the two halves of the
1263 value into the two single width floating point registers that make up
1264 the double width register. By default @samp{--construct-floats} is
1265 selected, allowing construction of these floating point constants.
1268 @item --emulation=@var{name}
1269 This option causes @command{@value{AS}} to emulate @command{@value{AS}} configured
1270 for some other target, in all respects, including output format (choosing
1271 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
1272 debugging information or store symbol table information, and default
1273 endianness. The available configuration names are: @samp{mipsecoff},
1274 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
1275 @samp{mipsbelf}. The first two do not alter the default endianness from that
1276 of the primary target for which the assembler was configured; the others change
1277 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
1278 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
1279 selection in any case.
1281 This option is currently supported only when the primary target
1282 @command{@value{AS}} is configured for is a @sc{mips} ELF or ECOFF target.
1283 Furthermore, the primary target or others specified with
1284 @samp{--enable-targets=@dots{}} at configuration time must include support for
1285 the other format, if both are to be available. For example, the Irix 5
1286 configuration includes support for both.
1288 Eventually, this option will support more configurations, with more
1289 fine-grained control over the assembler's behavior, and will be supported for
1293 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1300 Control how to deal with multiplication overflow and division by zero.
1301 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1302 (and only work for Instruction Set Architecture level 2 and higher);
1303 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1307 When this option is used, @command{@value{AS}} will issue a warning every
1308 time it generates a nop instruction from a macro.
1313 The following options are available when @value{AS} is configured for
1319 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1320 The command line option @samp{-nojsri2bsr} can be used to disable it.
1324 Enable or disable the silicon filter behaviour. By default this is disabled.
1325 The default can be overridden by the @samp{-sifilter} command line option.
1328 Alter jump instructions for long displacements.
1330 @item -mcpu=[210|340]
1331 Select the cpu type on the target hardware. This controls which instructions
1335 Assemble for a big endian target.
1338 Assemble for a little endian target.
1344 See the info pages for documentation of the MMIX-specific options.
1351 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1352 for a PowerPC processor.
1356 @c man begin OPTIONS
1357 The following options are available when @value{AS} is configured for a
1360 @c man begin INCLUDE
1362 @c ended inside the included file
1367 @c man begin OPTIONS
1369 See the info pages for documentation of the RX-specific options.
1373 The following options are available when @value{AS} is configured for the s390
1379 Select the word size, either 31/32 bits or 64 bits.
1382 Select the architecture mode, either the Enterprise System
1383 Architecture (esa) or the z/Architecture mode (zarch).
1384 @item -march=@var{processor}
1385 Specify which s390 processor variant is the target, @samp{g6}, @samp{g6},
1386 @samp{z900}, @samp{z990}, @samp{z9-109}, @samp{z9-ec}, or @samp{z10}.
1388 @itemx -mno-regnames
1389 Allow or disallow symbolic names for registers.
1390 @item -mwarn-areg-zero
1391 Warn whenever the operand for a base or index register has been specified
1392 but evaluates to zero.
1400 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1401 for a TMS320C6000 processor.
1405 @c man begin OPTIONS
1406 The following options are available when @value{AS} is configured for a
1407 TMS320C6000 processor.
1409 @c man begin INCLUDE
1410 @include c-tic6x.texi
1411 @c ended inside the included file
1419 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1420 for a TILE-Gx processor.
1424 @c man begin OPTIONS
1425 The following options are available when @value{AS} is configured for a TILE-Gx
1428 @c man begin INCLUDE
1429 @include c-tilegx.texi
1430 @c ended inside the included file
1438 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1439 for an Xtensa processor.
1443 @c man begin OPTIONS
1444 The following options are available when @value{AS} is configured for an
1447 @c man begin INCLUDE
1448 @include c-xtensa.texi
1449 @c ended inside the included file
1454 @c man begin OPTIONS
1457 The following options are available when @value{AS} is configured for
1458 a Z80 family processor.
1461 Assemble for Z80 processor.
1463 Assemble for R800 processor.
1464 @item -ignore-undocumented-instructions
1466 Assemble undocumented Z80 instructions that also work on R800 without warning.
1467 @item -ignore-unportable-instructions
1469 Assemble all undocumented Z80 instructions without warning.
1470 @item -warn-undocumented-instructions
1472 Issue a warning for undocumented Z80 instructions that also work on R800.
1473 @item -warn-unportable-instructions
1475 Issue a warning for undocumented Z80 instructions that do not work on R800.
1476 @item -forbid-undocumented-instructions
1478 Treat all undocumented instructions as errors.
1479 @item -forbid-unportable-instructions
1481 Treat undocumented Z80 instructions that do not work on R800 as errors.
1488 * Manual:: Structure of this Manual
1489 * GNU Assembler:: The GNU Assembler
1490 * Object Formats:: Object File Formats
1491 * Command Line:: Command Line
1492 * Input Files:: Input Files
1493 * Object:: Output (Object) File
1494 * Errors:: Error and Warning Messages
1498 @section Structure of this Manual
1500 @cindex manual, structure and purpose
1501 This manual is intended to describe what you need to know to use
1502 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1503 notation for symbols, constants, and expressions; the directives that
1504 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1507 We also cover special features in the @value{TARGET}
1508 configuration of @command{@value{AS}}, including assembler directives.
1511 This manual also describes some of the machine-dependent features of
1512 various flavors of the assembler.
1515 @cindex machine instructions (not covered)
1516 On the other hand, this manual is @emph{not} intended as an introduction
1517 to programming in assembly language---let alone programming in general!
1518 In a similar vein, we make no attempt to introduce the machine
1519 architecture; we do @emph{not} describe the instruction set, standard
1520 mnemonics, registers or addressing modes that are standard to a
1521 particular architecture.
1523 You may want to consult the manufacturer's
1524 machine architecture manual for this information.
1528 For information on the H8/300 machine instruction set, see @cite{H8/300
1529 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1530 Programming Manual} (Renesas).
1533 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1534 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1535 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1536 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1539 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1543 @c I think this is premature---doc@cygnus.com, 17jan1991
1545 Throughout this manual, we assume that you are running @dfn{GNU},
1546 the portable operating system from the @dfn{Free Software
1547 Foundation, Inc.}. This restricts our attention to certain kinds of
1548 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1549 once this assumption is granted examples and definitions need less
1552 @command{@value{AS}} is part of a team of programs that turn a high-level
1553 human-readable series of instructions into a low-level
1554 computer-readable series of instructions. Different versions of
1555 @command{@value{AS}} are used for different kinds of computer.
1558 @c There used to be a section "Terminology" here, which defined
1559 @c "contents", "byte", "word", and "long". Defining "word" to any
1560 @c particular size is confusing when the .word directive may generate 16
1561 @c bits on one machine and 32 bits on another; in general, for the user
1562 @c version of this manual, none of these terms seem essential to define.
1563 @c They were used very little even in the former draft of the manual;
1564 @c this draft makes an effort to avoid them (except in names of
1568 @section The GNU Assembler
1570 @c man begin DESCRIPTION
1572 @sc{gnu} @command{as} is really a family of assemblers.
1574 This manual describes @command{@value{AS}}, a member of that family which is
1575 configured for the @value{TARGET} architectures.
1577 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1578 should find a fairly similar environment when you use it on another
1579 architecture. Each version has much in common with the others,
1580 including object file formats, most assembler directives (often called
1581 @dfn{pseudo-ops}) and assembler syntax.@refill
1583 @cindex purpose of @sc{gnu} assembler
1584 @command{@value{AS}} is primarily intended to assemble the output of the
1585 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1586 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1587 assemble correctly everything that other assemblers for the same
1588 machine would assemble.
1590 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1593 @c This remark should appear in generic version of manual; assumption
1594 @c here is that generic version sets M680x0.
1595 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1596 assembler for the same architecture; for example, we know of several
1597 incompatible versions of 680x0 assembly language syntax.
1602 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1603 program in one pass of the source file. This has a subtle impact on the
1604 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1606 @node Object Formats
1607 @section Object File Formats
1609 @cindex object file format
1610 The @sc{gnu} assembler can be configured to produce several alternative
1611 object file formats. For the most part, this does not affect how you
1612 write assembly language programs; but directives for debugging symbols
1613 are typically different in different file formats. @xref{Symbol
1614 Attributes,,Symbol Attributes}.
1617 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1618 @value{OBJ-NAME} format object files.
1620 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1622 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1623 @code{b.out} or COFF format object files.
1626 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1627 SOM or ELF format object files.
1632 @section Command Line
1634 @cindex command line conventions
1636 After the program name @command{@value{AS}}, the command line may contain
1637 options and file names. Options may appear in any order, and may be
1638 before, after, or between file names. The order of file names is
1641 @cindex standard input, as input file
1643 @file{--} (two hyphens) by itself names the standard input file
1644 explicitly, as one of the files for @command{@value{AS}} to assemble.
1646 @cindex options, command line
1647 Except for @samp{--} any command line argument that begins with a
1648 hyphen (@samp{-}) is an option. Each option changes the behavior of
1649 @command{@value{AS}}. No option changes the way another option works. An
1650 option is a @samp{-} followed by one or more letters; the case of
1651 the letter is important. All options are optional.
1653 Some options expect exactly one file name to follow them. The file
1654 name may either immediately follow the option's letter (compatible
1655 with older assemblers) or it may be the next command argument (@sc{gnu}
1656 standard). These two command lines are equivalent:
1659 @value{AS} -o my-object-file.o mumble.s
1660 @value{AS} -omy-object-file.o mumble.s
1664 @section Input Files
1667 @cindex source program
1668 @cindex files, input
1669 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1670 describe the program input to one run of @command{@value{AS}}. The program may
1671 be in one or more files; how the source is partitioned into files
1672 doesn't change the meaning of the source.
1674 @c I added "con" prefix to "catenation" just to prove I can overcome my
1675 @c APL training... doc@cygnus.com
1676 The source program is a concatenation of the text in all the files, in the
1679 @c man begin DESCRIPTION
1680 Each time you run @command{@value{AS}} it assembles exactly one source
1681 program. The source program is made up of one or more files.
1682 (The standard input is also a file.)
1684 You give @command{@value{AS}} a command line that has zero or more input file
1685 names. The input files are read (from left file name to right). A
1686 command line argument (in any position) that has no special meaning
1687 is taken to be an input file name.
1689 If you give @command{@value{AS}} no file names it attempts to read one input file
1690 from the @command{@value{AS}} standard input, which is normally your terminal. You
1691 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1694 Use @samp{--} if you need to explicitly name the standard input file
1695 in your command line.
1697 If the source is empty, @command{@value{AS}} produces a small, empty object
1702 @subheading Filenames and Line-numbers
1704 @cindex input file linenumbers
1705 @cindex line numbers, in input files
1706 There are two ways of locating a line in the input file (or files) and
1707 either may be used in reporting error messages. One way refers to a line
1708 number in a physical file; the other refers to a line number in a
1709 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1711 @dfn{Physical files} are those files named in the command line given
1712 to @command{@value{AS}}.
1714 @dfn{Logical files} are simply names declared explicitly by assembler
1715 directives; they bear no relation to physical files. Logical file names help
1716 error messages reflect the original source file, when @command{@value{AS}} source
1717 is itself synthesized from other files. @command{@value{AS}} understands the
1718 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1719 @ref{File,,@code{.file}}.
1722 @section Output (Object) File
1728 Every time you run @command{@value{AS}} it produces an output file, which is
1729 your assembly language program translated into numbers. This file
1730 is the object file. Its default name is
1738 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1740 You can give it another name by using the @option{-o} option. Conventionally,
1741 object file names end with @file{.o}. The default name is used for historical
1742 reasons: older assemblers were capable of assembling self-contained programs
1743 directly into a runnable program. (For some formats, this isn't currently
1744 possible, but it can be done for the @code{a.out} format.)
1748 The object file is meant for input to the linker @code{@value{LD}}. It contains
1749 assembled program code, information to help @code{@value{LD}} integrate
1750 the assembled program into a runnable file, and (optionally) symbolic
1751 information for the debugger.
1753 @c link above to some info file(s) like the description of a.out.
1754 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1757 @section Error and Warning Messages
1759 @c man begin DESCRIPTION
1761 @cindex error messages
1762 @cindex warning messages
1763 @cindex messages from assembler
1764 @command{@value{AS}} may write warnings and error messages to the standard error
1765 file (usually your terminal). This should not happen when a compiler
1766 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1767 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1768 grave problem that stops the assembly.
1772 @cindex format of warning messages
1773 Warning messages have the format
1776 file_name:@b{NNN}:Warning Message Text
1780 @cindex line numbers, in warnings/errors
1781 (where @b{NNN} is a line number). If a logical file name has been given
1782 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1783 the current input file is used. If a logical line number was given
1785 (@pxref{Line,,@code{.line}})
1787 then it is used to calculate the number printed,
1788 otherwise the actual line in the current source file is printed. The
1789 message text is intended to be self explanatory (in the grand Unix
1792 @cindex format of error messages
1793 Error messages have the format
1795 file_name:@b{NNN}:FATAL:Error Message Text
1797 The file name and line number are derived as for warning
1798 messages. The actual message text may be rather less explanatory
1799 because many of them aren't supposed to happen.
1802 @chapter Command-Line Options
1804 @cindex options, all versions of assembler
1805 This chapter describes command-line options available in @emph{all}
1806 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
1807 for options specific
1809 to the @value{TARGET} target.
1812 to particular machine architectures.
1815 @c man begin DESCRIPTION
1817 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1818 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1819 The assembler arguments must be separated from each other (and the @samp{-Wa})
1820 by commas. For example:
1823 gcc -c -g -O -Wa,-alh,-L file.c
1827 This passes two options to the assembler: @samp{-alh} (emit a listing to
1828 standard output with high-level and assembly source) and @samp{-L} (retain
1829 local symbols in the symbol table).
1831 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1832 command-line options are automatically passed to the assembler by the compiler.
1833 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1834 precisely what options it passes to each compilation pass, including the
1840 * a:: -a[cdghlns] enable listings
1841 * alternate:: --alternate enable alternate macro syntax
1842 * D:: -D for compatibility
1843 * f:: -f to work faster
1844 * I:: -I for .include search path
1845 @ifclear DIFF-TBL-KLUGE
1846 * K:: -K for compatibility
1848 @ifset DIFF-TBL-KLUGE
1849 * K:: -K for difference tables
1852 * L:: -L to retain local symbols
1853 * listing:: --listing-XXX to configure listing output
1854 * M:: -M or --mri to assemble in MRI compatibility mode
1855 * MD:: --MD for dependency tracking
1856 * o:: -o to name the object file
1857 * R:: -R to join data and text sections
1858 * statistics:: --statistics to see statistics about assembly
1859 * traditional-format:: --traditional-format for compatible output
1860 * v:: -v to announce version
1861 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1862 * Z:: -Z to make object file even after errors
1866 @section Enable Listings: @option{-a[cdghlns]}
1876 @cindex listings, enabling
1877 @cindex assembly listings, enabling
1879 These options enable listing output from the assembler. By itself,
1880 @samp{-a} requests high-level, assembly, and symbols listing.
1881 You can use other letters to select specific options for the list:
1882 @samp{-ah} requests a high-level language listing,
1883 @samp{-al} requests an output-program assembly listing, and
1884 @samp{-as} requests a symbol table listing.
1885 High-level listings require that a compiler debugging option like
1886 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1889 Use the @samp{-ag} option to print a first section with general assembly
1890 information, like @value{AS} version, switches passed, or time stamp.
1892 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1893 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1894 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1895 omitted from the listing.
1897 Use the @samp{-ad} option to omit debugging directives from the
1900 Once you have specified one of these options, you can further control
1901 listing output and its appearance using the directives @code{.list},
1902 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1904 The @samp{-an} option turns off all forms processing.
1905 If you do not request listing output with one of the @samp{-a} options, the
1906 listing-control directives have no effect.
1908 The letters after @samp{-a} may be combined into one option,
1909 @emph{e.g.}, @samp{-aln}.
1911 Note if the assembler source is coming from the standard input (e.g.,
1913 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1914 is being used) then the listing will not contain any comments or preprocessor
1915 directives. This is because the listing code buffers input source lines from
1916 stdin only after they have been preprocessed by the assembler. This reduces
1917 memory usage and makes the code more efficient.
1920 @section @option{--alternate}
1923 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
1926 @section @option{-D}
1929 This option has no effect whatsoever, but it is accepted to make it more
1930 likely that scripts written for other assemblers also work with
1931 @command{@value{AS}}.
1934 @section Work Faster: @option{-f}
1937 @cindex trusted compiler
1938 @cindex faster processing (@option{-f})
1939 @samp{-f} should only be used when assembling programs written by a
1940 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1941 and comment preprocessing on
1942 the input file(s) before assembling them. @xref{Preprocessing,
1946 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1947 preprocessed (if they contain comments, for example), @command{@value{AS}} does
1952 @section @code{.include} Search Path: @option{-I} @var{path}
1954 @kindex -I @var{path}
1955 @cindex paths for @code{.include}
1956 @cindex search path for @code{.include}
1957 @cindex @code{include} directive search path
1958 Use this option to add a @var{path} to the list of directories
1959 @command{@value{AS}} searches for files specified in @code{.include}
1960 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
1961 many times as necessary to include a variety of paths. The current
1962 working directory is always searched first; after that, @command{@value{AS}}
1963 searches any @samp{-I} directories in the same order as they were
1964 specified (left to right) on the command line.
1967 @section Difference Tables: @option{-K}
1970 @ifclear DIFF-TBL-KLUGE
1971 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1972 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1973 where it can be used to warn when the assembler alters the machine code
1974 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1975 family does not have the addressing limitations that sometimes lead to this
1976 alteration on other platforms.
1979 @ifset DIFF-TBL-KLUGE
1980 @cindex difference tables, warning
1981 @cindex warning for altered difference tables
1982 @command{@value{AS}} sometimes alters the code emitted for directives of the
1983 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
1984 You can use the @samp{-K} option if you want a warning issued when this
1989 @section Include Local Symbols: @option{-L}
1992 @cindex local symbols, retaining in output
1993 Symbols beginning with system-specific local label prefixes, typically
1994 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
1995 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
1996 such symbols when debugging, because they are intended for the use of
1997 programs (like compilers) that compose assembler programs, not for your
1998 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
1999 such symbols, so you do not normally debug with them.
2001 This option tells @command{@value{AS}} to retain those local symbols
2002 in the object file. Usually if you do this you also tell the linker
2003 @code{@value{LD}} to preserve those symbols.
2006 @section Configuring listing output: @option{--listing}
2008 The listing feature of the assembler can be enabled via the command line switch
2009 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2010 hex dump of the corresponding locations in the output object file, and displays
2011 them as a listing file. The format of this listing can be controlled by
2012 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2013 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2014 @code{.psize} (@pxref{Psize}), and
2015 @code{.eject} (@pxref{Eject}) and also by the following switches:
2018 @item --listing-lhs-width=@samp{number}
2019 @kindex --listing-lhs-width
2020 @cindex Width of first line disassembly output
2021 Sets the maximum width, in words, of the first line of the hex byte dump. This
2022 dump appears on the left hand side of the listing output.
2024 @item --listing-lhs-width2=@samp{number}
2025 @kindex --listing-lhs-width2
2026 @cindex Width of continuation lines of disassembly output
2027 Sets the maximum width, in words, of any further lines of the hex byte dump for
2028 a given input source line. If this value is not specified, it defaults to being
2029 the same as the value specified for @samp{--listing-lhs-width}. If neither
2030 switch is used the default is to one.
2032 @item --listing-rhs-width=@samp{number}
2033 @kindex --listing-rhs-width
2034 @cindex Width of source line output
2035 Sets the maximum width, in characters, of the source line that is displayed
2036 alongside the hex dump. The default value for this parameter is 100. The
2037 source line is displayed on the right hand side of the listing output.
2039 @item --listing-cont-lines=@samp{number}
2040 @kindex --listing-cont-lines
2041 @cindex Maximum number of continuation lines
2042 Sets the maximum number of continuation lines of hex dump that will be
2043 displayed for a given single line of source input. The default value is 4.
2047 @section Assemble in MRI Compatibility Mode: @option{-M}
2050 @cindex MRI compatibility mode
2051 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2052 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2053 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
2054 configured target) assembler from Microtec Research. The exact nature of the
2055 MRI syntax will not be documented here; see the MRI manuals for more
2056 information. Note in particular that the handling of macros and macro
2057 arguments is somewhat different. The purpose of this option is to permit
2058 assembling existing MRI assembler code using @command{@value{AS}}.
2060 The MRI compatibility is not complete. Certain operations of the MRI assembler
2061 depend upon its object file format, and can not be supported using other object
2062 file formats. Supporting these would require enhancing each object file format
2063 individually. These are:
2066 @item global symbols in common section
2068 The m68k MRI assembler supports common sections which are merged by the linker.
2069 Other object file formats do not support this. @command{@value{AS}} handles
2070 common sections by treating them as a single common symbol. It permits local
2071 symbols to be defined within a common section, but it can not support global
2072 symbols, since it has no way to describe them.
2074 @item complex relocations
2076 The MRI assemblers support relocations against a negated section address, and
2077 relocations which combine the start addresses of two or more sections. These
2078 are not support by other object file formats.
2080 @item @code{END} pseudo-op specifying start address
2082 The MRI @code{END} pseudo-op permits the specification of a start address.
2083 This is not supported by other object file formats. The start address may
2084 instead be specified using the @option{-e} option to the linker, or in a linker
2087 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2089 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2090 name to the output file. This is not supported by other object file formats.
2092 @item @code{ORG} pseudo-op
2094 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2095 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2096 which changes the location within the current section. Absolute sections are
2097 not supported by other object file formats. The address of a section may be
2098 assigned within a linker script.
2101 There are some other features of the MRI assembler which are not supported by
2102 @command{@value{AS}}, typically either because they are difficult or because they
2103 seem of little consequence. Some of these may be supported in future releases.
2107 @item EBCDIC strings
2109 EBCDIC strings are not supported.
2111 @item packed binary coded decimal
2113 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2114 and @code{DCB.P} pseudo-ops are not supported.
2116 @item @code{FEQU} pseudo-op
2118 The m68k @code{FEQU} pseudo-op is not supported.
2120 @item @code{NOOBJ} pseudo-op
2122 The m68k @code{NOOBJ} pseudo-op is not supported.
2124 @item @code{OPT} branch control options
2126 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2127 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2128 relaxes all branches, whether forward or backward, to an appropriate size, so
2129 these options serve no purpose.
2131 @item @code{OPT} list control options
2133 The following m68k @code{OPT} list control options are ignored: @code{C},
2134 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2135 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2137 @item other @code{OPT} options
2139 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2140 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2142 @item @code{OPT} @code{D} option is default
2144 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2145 @code{OPT NOD} may be used to turn it off.
2147 @item @code{XREF} pseudo-op.
2149 The m68k @code{XREF} pseudo-op is ignored.
2151 @item @code{.debug} pseudo-op
2153 The i960 @code{.debug} pseudo-op is not supported.
2155 @item @code{.extended} pseudo-op
2157 The i960 @code{.extended} pseudo-op is not supported.
2159 @item @code{.list} pseudo-op.
2161 The various options of the i960 @code{.list} pseudo-op are not supported.
2163 @item @code{.optimize} pseudo-op
2165 The i960 @code{.optimize} pseudo-op is not supported.
2167 @item @code{.output} pseudo-op
2169 The i960 @code{.output} pseudo-op is not supported.
2171 @item @code{.setreal} pseudo-op
2173 The i960 @code{.setreal} pseudo-op is not supported.
2178 @section Dependency Tracking: @option{--MD}
2181 @cindex dependency tracking
2184 @command{@value{AS}} can generate a dependency file for the file it creates. This
2185 file consists of a single rule suitable for @code{make} describing the
2186 dependencies of the main source file.
2188 The rule is written to the file named in its argument.
2190 This feature is used in the automatic updating of makefiles.
2193 @section Name the Object File: @option{-o}
2196 @cindex naming object file
2197 @cindex object file name
2198 There is always one object file output when you run @command{@value{AS}}. By
2199 default it has the name
2202 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
2216 You use this option (which takes exactly one filename) to give the
2217 object file a different name.
2219 Whatever the object file is called, @command{@value{AS}} overwrites any
2220 existing file of the same name.
2223 @section Join Data and Text Sections: @option{-R}
2226 @cindex data and text sections, joining
2227 @cindex text and data sections, joining
2228 @cindex joining text and data sections
2229 @cindex merging text and data sections
2230 @option{-R} tells @command{@value{AS}} to write the object file as if all
2231 data-section data lives in the text section. This is only done at
2232 the very last moment: your binary data are the same, but data
2233 section parts are relocated differently. The data section part of
2234 your object file is zero bytes long because all its bytes are
2235 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2237 When you specify @option{-R} it would be possible to generate shorter
2238 address displacements (because we do not have to cross between text and
2239 data section). We refrain from doing this simply for compatibility with
2240 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2243 When @command{@value{AS}} is configured for COFF or ELF output,
2244 this option is only useful if you use sections named @samp{.text} and
2249 @option{-R} is not supported for any of the HPPA targets. Using
2250 @option{-R} generates a warning from @command{@value{AS}}.
2254 @section Display Assembly Statistics: @option{--statistics}
2256 @kindex --statistics
2257 @cindex statistics, about assembly
2258 @cindex time, total for assembly
2259 @cindex space used, maximum for assembly
2260 Use @samp{--statistics} to display two statistics about the resources used by
2261 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2262 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2265 @node traditional-format
2266 @section Compatible Output: @option{--traditional-format}
2268 @kindex --traditional-format
2269 For some targets, the output of @command{@value{AS}} is different in some ways
2270 from the output of some existing assembler. This switch requests
2271 @command{@value{AS}} to use the traditional format instead.
2273 For example, it disables the exception frame optimizations which
2274 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2277 @section Announce Version: @option{-v}
2281 @cindex assembler version
2282 @cindex version of assembler
2283 You can find out what version of as is running by including the
2284 option @samp{-v} (which you can also spell as @samp{-version}) on the
2288 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2290 @command{@value{AS}} should never give a warning or error message when
2291 assembling compiler output. But programs written by people often
2292 cause @command{@value{AS}} to give a warning that a particular assumption was
2293 made. All such warnings are directed to the standard error file.
2297 @cindex suppressing warnings
2298 @cindex warnings, suppressing
2299 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2300 This only affects the warning messages: it does not change any particular of
2301 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2304 @kindex --fatal-warnings
2305 @cindex errors, caused by warnings
2306 @cindex warnings, causing error
2307 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2308 files that generate warnings to be in error.
2311 @cindex warnings, switching on
2312 You can switch these options off again by specifying @option{--warn}, which
2313 causes warnings to be output as usual.
2316 @section Generate Object File in Spite of Errors: @option{-Z}
2317 @cindex object file, after errors
2318 @cindex errors, continuing after
2319 After an error message, @command{@value{AS}} normally produces no output. If for
2320 some reason you are interested in object file output even after
2321 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2322 option. If there are any errors, @command{@value{AS}} continues anyways, and
2323 writes an object file after a final warning message of the form @samp{@var{n}
2324 errors, @var{m} warnings, generating bad object file.}
2329 @cindex machine-independent syntax
2330 @cindex syntax, machine-independent
2331 This chapter describes the machine-independent syntax allowed in a
2332 source file. @command{@value{AS}} syntax is similar to what many other
2333 assemblers use; it is inspired by the BSD 4.2
2338 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2342 * Preprocessing:: Preprocessing
2343 * Whitespace:: Whitespace
2344 * Comments:: Comments
2345 * Symbol Intro:: Symbols
2346 * Statements:: Statements
2347 * Constants:: Constants
2351 @section Preprocessing
2353 @cindex preprocessing
2354 The @command{@value{AS}} internal preprocessor:
2356 @cindex whitespace, removed by preprocessor
2358 adjusts and removes extra whitespace. It leaves one space or tab before
2359 the keywords on a line, and turns any other whitespace on the line into
2362 @cindex comments, removed by preprocessor
2364 removes all comments, replacing them with a single space, or an
2365 appropriate number of newlines.
2367 @cindex constants, converted by preprocessor
2369 converts character constants into the appropriate numeric values.
2372 It does not do macro processing, include file handling, or
2373 anything else you may get from your C compiler's preprocessor. You can
2374 do include file processing with the @code{.include} directive
2375 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2376 to get other ``CPP'' style preprocessing by giving the input file a
2377 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2378 Output, gcc.info, Using GNU CC}.
2380 Excess whitespace, comments, and character constants
2381 cannot be used in the portions of the input text that are not
2384 @cindex turning preprocessing on and off
2385 @cindex preprocessing, turning on and off
2388 If the first line of an input file is @code{#NO_APP} or if you use the
2389 @samp{-f} option, whitespace and comments are not removed from the input file.
2390 Within an input file, you can ask for whitespace and comment removal in
2391 specific portions of the by putting a line that says @code{#APP} before the
2392 text that may contain whitespace or comments, and putting a line that says
2393 @code{#NO_APP} after this text. This feature is mainly intend to support
2394 @code{asm} statements in compilers whose output is otherwise free of comments
2401 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2402 Whitespace is used to separate symbols, and to make programs neater for
2403 people to read. Unless within character constants
2404 (@pxref{Characters,,Character Constants}), any whitespace means the same
2405 as exactly one space.
2411 There are two ways of rendering comments to @command{@value{AS}}. In both
2412 cases the comment is equivalent to one space.
2414 Anything from @samp{/*} through the next @samp{*/} is a comment.
2415 This means you may not nest these comments.
2419 The only way to include a newline ('\n') in a comment
2420 is to use this sort of comment.
2423 /* This sort of comment does not nest. */
2426 @cindex line comment character
2427 Anything from a @dfn{line comment} character up to the next newline is
2428 considered a comment and is ignored. The line comment character is target
2429 specific, and some targets multiple comment characters. Some targets also have
2430 line comment characters that only work if they are the first character on a
2431 line. Some targets use a sequence of two characters to introduce a line
2432 comment. Some targets can also change their line comment characters depending
2433 upon command line options that have been used. For more details see the
2434 @emph{Syntax} section in the documentation for individual targets.
2436 If the line comment character is the hash sign (@samp{#}) then it still has the
2437 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2438 to specify logical line numbers:
2441 @cindex lines starting with @code{#}
2442 @cindex logical line numbers
2443 To be compatible with past assemblers, lines that begin with @samp{#} have a
2444 special interpretation. Following the @samp{#} should be an absolute
2445 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2446 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2447 new logical file name. The rest of the line, if any, should be whitespace.
2449 If the first non-whitespace characters on the line are not numeric,
2450 the line is ignored. (Just like a comment.)
2453 # This is an ordinary comment.
2454 # 42-6 "new_file_name" # New logical file name
2455 # This is logical line # 36.
2457 This feature is deprecated, and may disappear from future versions
2458 of @command{@value{AS}}.
2463 @cindex characters used in symbols
2464 @ifclear SPECIAL-SYMS
2465 A @dfn{symbol} is one or more characters chosen from the set of all
2466 letters (both upper and lower case), digits and the three characters
2472 A @dfn{symbol} is one or more characters chosen from the set of all
2473 letters (both upper and lower case), digits and the three characters
2474 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2480 On most machines, you can also use @code{$} in symbol names; exceptions
2481 are noted in @ref{Machine Dependencies}.
2483 No symbol may begin with a digit. Case is significant.
2484 There is no length limit: all characters are significant. Symbols are
2485 delimited by characters not in that set, or by the beginning of a file
2486 (since the source program must end with a newline, the end of a file is
2487 not a possible symbol delimiter). @xref{Symbols}.
2488 @cindex length of symbols
2493 @cindex statements, structure of
2494 @cindex line separator character
2495 @cindex statement separator character
2497 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2498 @dfn{line separator character}. The line separator character is target
2499 specific and described in the @emph{Syntax} section of each
2500 target's documentation. Not all targets support a line separator character.
2501 The newline or line separator character is considered to be part of the
2502 preceding statement. Newlines and separators within character constants are an
2503 exception: they do not end statements.
2505 @cindex newline, required at file end
2506 @cindex EOF, newline must precede
2507 It is an error to end any statement with end-of-file: the last
2508 character of any input file should be a newline.@refill
2510 An empty statement is allowed, and may include whitespace. It is ignored.
2512 @cindex instructions and directives
2513 @cindex directives and instructions
2514 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2515 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2517 A statement begins with zero or more labels, optionally followed by a
2518 key symbol which determines what kind of statement it is. The key
2519 symbol determines the syntax of the rest of the statement. If the
2520 symbol begins with a dot @samp{.} then the statement is an assembler
2521 directive: typically valid for any computer. If the symbol begins with
2522 a letter the statement is an assembly language @dfn{instruction}: it
2523 assembles into a machine language instruction.
2525 Different versions of @command{@value{AS}} for different computers
2526 recognize different instructions. In fact, the same symbol may
2527 represent a different instruction in a different computer's assembly
2531 @cindex @code{:} (label)
2532 @cindex label (@code{:})
2533 A label is a symbol immediately followed by a colon (@code{:}).
2534 Whitespace before a label or after a colon is permitted, but you may not
2535 have whitespace between a label's symbol and its colon. @xref{Labels}.
2538 For HPPA targets, labels need not be immediately followed by a colon, but
2539 the definition of a label must begin in column zero. This also implies that
2540 only one label may be defined on each line.
2544 label: .directive followed by something
2545 another_label: # This is an empty statement.
2546 instruction operand_1, operand_2, @dots{}
2553 A constant is a number, written so that its value is known by
2554 inspection, without knowing any context. Like this:
2557 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2558 .ascii "Ring the bell\7" # A string constant.
2559 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2560 .float 0f-314159265358979323846264338327\
2561 95028841971.693993751E-40 # - pi, a flonum.
2566 * Characters:: Character Constants
2567 * Numbers:: Number Constants
2571 @subsection Character Constants
2573 @cindex character constants
2574 @cindex constants, character
2575 There are two kinds of character constants. A @dfn{character} stands
2576 for one character in one byte and its value may be used in
2577 numeric expressions. String constants (properly called string
2578 @emph{literals}) are potentially many bytes and their values may not be
2579 used in arithmetic expressions.
2583 * Chars:: Characters
2587 @subsubsection Strings
2589 @cindex string constants
2590 @cindex constants, string
2591 A @dfn{string} is written between double-quotes. It may contain
2592 double-quotes or null characters. The way to get special characters
2593 into a string is to @dfn{escape} these characters: precede them with
2594 a backslash @samp{\} character. For example @samp{\\} represents
2595 one backslash: the first @code{\} is an escape which tells
2596 @command{@value{AS}} to interpret the second character literally as a backslash
2597 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2598 escape character). The complete list of escapes follows.
2600 @cindex escape codes, character
2601 @cindex character escape codes
2604 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2606 @cindex @code{\b} (backspace character)
2607 @cindex backspace (@code{\b})
2609 Mnemonic for backspace; for ASCII this is octal code 010.
2612 @c Mnemonic for EOText; for ASCII this is octal code 004.
2614 @cindex @code{\f} (formfeed character)
2615 @cindex formfeed (@code{\f})
2617 Mnemonic for FormFeed; for ASCII this is octal code 014.
2619 @cindex @code{\n} (newline character)
2620 @cindex newline (@code{\n})
2622 Mnemonic for newline; for ASCII this is octal code 012.
2625 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2627 @cindex @code{\r} (carriage return character)
2628 @cindex carriage return (@code{\r})
2630 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2633 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2634 @c other assemblers.
2636 @cindex @code{\t} (tab)
2637 @cindex tab (@code{\t})
2639 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2642 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2643 @c @item \x @var{digit} @var{digit} @var{digit}
2644 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2646 @cindex @code{\@var{ddd}} (octal character code)
2647 @cindex octal character code (@code{\@var{ddd}})
2648 @item \ @var{digit} @var{digit} @var{digit}
2649 An octal character code. The numeric code is 3 octal digits.
2650 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2651 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2653 @cindex @code{\@var{xd...}} (hex character code)
2654 @cindex hex character code (@code{\@var{xd...}})
2655 @item \@code{x} @var{hex-digits...}
2656 A hex character code. All trailing hex digits are combined. Either upper or
2657 lower case @code{x} works.
2659 @cindex @code{\\} (@samp{\} character)
2660 @cindex backslash (@code{\\})
2662 Represents one @samp{\} character.
2665 @c Represents one @samp{'} (accent acute) character.
2666 @c This is needed in single character literals
2667 @c (@xref{Characters,,Character Constants}.) to represent
2670 @cindex @code{\"} (doublequote character)
2671 @cindex doublequote (@code{\"})
2673 Represents one @samp{"} character. Needed in strings to represent
2674 this character, because an unescaped @samp{"} would end the string.
2676 @item \ @var{anything-else}
2677 Any other character when escaped by @kbd{\} gives a warning, but
2678 assembles as if the @samp{\} was not present. The idea is that if
2679 you used an escape sequence you clearly didn't want the literal
2680 interpretation of the following character. However @command{@value{AS}} has no
2681 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2682 code and warns you of the fact.
2685 Which characters are escapable, and what those escapes represent,
2686 varies widely among assemblers. The current set is what we think
2687 the BSD 4.2 assembler recognizes, and is a subset of what most C
2688 compilers recognize. If you are in doubt, do not use an escape
2692 @subsubsection Characters
2694 @cindex single character constant
2695 @cindex character, single
2696 @cindex constant, single character
2697 A single character may be written as a single quote immediately
2698 followed by that character. The same escapes apply to characters as
2699 to strings. So if you want to write the character backslash, you
2700 must write @kbd{'\\} where the first @code{\} escapes the second
2701 @code{\}. As you can see, the quote is an acute accent, not a
2702 grave accent. A newline
2704 @ifclear abnormal-separator
2705 (or semicolon @samp{;})
2707 @ifset abnormal-separator
2709 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2714 immediately following an acute accent is taken as a literal character
2715 and does not count as the end of a statement. The value of a character
2716 constant in a numeric expression is the machine's byte-wide code for
2717 that character. @command{@value{AS}} assumes your character code is ASCII:
2718 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2721 @subsection Number Constants
2723 @cindex constants, number
2724 @cindex number constants
2725 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2726 are stored in the target machine. @emph{Integers} are numbers that
2727 would fit into an @code{int} in the C language. @emph{Bignums} are
2728 integers, but they are stored in more than 32 bits. @emph{Flonums}
2729 are floating point numbers, described below.
2732 * Integers:: Integers
2737 * Bit Fields:: Bit Fields
2743 @subsubsection Integers
2745 @cindex constants, integer
2747 @cindex binary integers
2748 @cindex integers, binary
2749 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2750 the binary digits @samp{01}.
2752 @cindex octal integers
2753 @cindex integers, octal
2754 An octal integer is @samp{0} followed by zero or more of the octal
2755 digits (@samp{01234567}).
2757 @cindex decimal integers
2758 @cindex integers, decimal
2759 A decimal integer starts with a non-zero digit followed by zero or
2760 more digits (@samp{0123456789}).
2762 @cindex hexadecimal integers
2763 @cindex integers, hexadecimal
2764 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2765 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2767 Integers have the usual values. To denote a negative integer, use
2768 the prefix operator @samp{-} discussed under expressions
2769 (@pxref{Prefix Ops,,Prefix Operators}).
2772 @subsubsection Bignums
2775 @cindex constants, bignum
2776 A @dfn{bignum} has the same syntax and semantics as an integer
2777 except that the number (or its negative) takes more than 32 bits to
2778 represent in binary. The distinction is made because in some places
2779 integers are permitted while bignums are not.
2782 @subsubsection Flonums
2784 @cindex floating point numbers
2785 @cindex constants, floating point
2787 @cindex precision, floating point
2788 A @dfn{flonum} represents a floating point number. The translation is
2789 indirect: a decimal floating point number from the text is converted by
2790 @command{@value{AS}} to a generic binary floating point number of more than
2791 sufficient precision. This generic floating point number is converted
2792 to a particular computer's floating point format (or formats) by a
2793 portion of @command{@value{AS}} specialized to that computer.
2795 A flonum is written by writing (in order)
2800 (@samp{0} is optional on the HPPA.)
2804 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2806 @kbd{e} is recommended. Case is not important.
2808 @c FIXME: verify if flonum syntax really this vague for most cases
2809 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2810 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2813 On the H8/300, Renesas / SuperH SH,
2814 and AMD 29K architectures, the letter must be
2815 one of the letters @samp{DFPRSX} (in upper or lower case).
2817 On the ARC, the letter must be one of the letters @samp{DFRS}
2818 (in upper or lower case).
2820 On the Intel 960 architecture, the letter must be
2821 one of the letters @samp{DFT} (in upper or lower case).
2823 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2827 One of the letters @samp{DFRS} (in upper or lower case).
2830 One of the letters @samp{DFPRSX} (in upper or lower case).
2833 The letter @samp{E} (upper case only).
2836 One of the letters @samp{DFT} (in upper or lower case).
2841 An optional sign: either @samp{+} or @samp{-}.
2844 An optional @dfn{integer part}: zero or more decimal digits.
2847 An optional @dfn{fractional part}: @samp{.} followed by zero
2848 or more decimal digits.
2851 An optional exponent, consisting of:
2855 An @samp{E} or @samp{e}.
2856 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2857 @c principle this can perfectly well be different on different targets.
2859 Optional sign: either @samp{+} or @samp{-}.
2861 One or more decimal digits.
2866 At least one of the integer part or the fractional part must be
2867 present. The floating point number has the usual base-10 value.
2869 @command{@value{AS}} does all processing using integers. Flonums are computed
2870 independently of any floating point hardware in the computer running
2871 @command{@value{AS}}.
2875 @c Bit fields are written as a general facility but are also controlled
2876 @c by a conditional-compilation flag---which is as of now (21mar91)
2877 @c turned on only by the i960 config of GAS.
2879 @subsubsection Bit Fields
2882 @cindex constants, bit field
2883 You can also define numeric constants as @dfn{bit fields}.
2884 Specify two numbers separated by a colon---
2886 @var{mask}:@var{value}
2889 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2892 The resulting number is then packed
2894 @c this conditional paren in case bit fields turned on elsewhere than 960
2895 (in host-dependent byte order)
2897 into a field whose width depends on which assembler directive has the
2898 bit-field as its argument. Overflow (a result from the bitwise and
2899 requiring more binary digits to represent) is not an error; instead,
2900 more constants are generated, of the specified width, beginning with the
2901 least significant digits.@refill
2903 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2904 @code{.short}, and @code{.word} accept bit-field arguments.
2909 @chapter Sections and Relocation
2914 * Secs Background:: Background
2915 * Ld Sections:: Linker Sections
2916 * As Sections:: Assembler Internal Sections
2917 * Sub-Sections:: Sub-Sections
2921 @node Secs Background
2924 Roughly, a section is a range of addresses, with no gaps; all data
2925 ``in'' those addresses is treated the same for some particular purpose.
2926 For example there may be a ``read only'' section.
2928 @cindex linker, and assembler
2929 @cindex assembler, and linker
2930 The linker @code{@value{LD}} reads many object files (partial programs) and
2931 combines their contents to form a runnable program. When @command{@value{AS}}
2932 emits an object file, the partial program is assumed to start at address 0.
2933 @code{@value{LD}} assigns the final addresses for the partial program, so that
2934 different partial programs do not overlap. This is actually an
2935 oversimplification, but it suffices to explain how @command{@value{AS}} uses
2938 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2939 addresses. These blocks slide to their run-time addresses as rigid
2940 units; their length does not change and neither does the order of bytes
2941 within them. Such a rigid unit is called a @emph{section}. Assigning
2942 run-time addresses to sections is called @dfn{relocation}. It includes
2943 the task of adjusting mentions of object-file addresses so they refer to
2944 the proper run-time addresses.
2946 For the H8/300, and for the Renesas / SuperH SH,
2947 @command{@value{AS}} pads sections if needed to
2948 ensure they end on a word (sixteen bit) boundary.
2951 @cindex standard assembler sections
2952 An object file written by @command{@value{AS}} has at least three sections, any
2953 of which may be empty. These are named @dfn{text}, @dfn{data} and
2958 When it generates COFF or ELF output,
2960 @command{@value{AS}} can also generate whatever other named sections you specify
2961 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2962 If you do not use any directives that place output in the @samp{.text}
2963 or @samp{.data} sections, these sections still exist, but are empty.
2968 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
2970 @command{@value{AS}} can also generate whatever other named sections you
2971 specify using the @samp{.space} and @samp{.subspace} directives. See
2972 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2973 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2974 assembler directives.
2977 Additionally, @command{@value{AS}} uses different names for the standard
2978 text, data, and bss sections when generating SOM output. Program text
2979 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2980 BSS into @samp{$BSS$}.
2984 Within the object file, the text section starts at address @code{0}, the
2985 data section follows, and the bss section follows the data section.
2988 When generating either SOM or ELF output files on the HPPA, the text
2989 section starts at address @code{0}, the data section at address
2990 @code{0x4000000}, and the bss section follows the data section.
2993 To let @code{@value{LD}} know which data changes when the sections are
2994 relocated, and how to change that data, @command{@value{AS}} also writes to the
2995 object file details of the relocation needed. To perform relocation
2996 @code{@value{LD}} must know, each time an address in the object
3000 Where in the object file is the beginning of this reference to
3003 How long (in bytes) is this reference?
3005 Which section does the address refer to? What is the numeric value of
3007 (@var{address}) @minus{} (@var{start-address of section})?
3010 Is the reference to an address ``Program-Counter relative''?
3013 @cindex addresses, format of
3014 @cindex section-relative addressing
3015 In fact, every address @command{@value{AS}} ever uses is expressed as
3017 (@var{section}) + (@var{offset into section})
3020 Further, most expressions @command{@value{AS}} computes have this section-relative
3023 (For some object formats, such as SOM for the HPPA, some expressions are
3024 symbol-relative instead.)
3027 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3028 @var{N} into section @var{secname}.''
3030 Apart from text, data and bss sections you need to know about the
3031 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3032 addresses in the absolute section remain unchanged. For example, address
3033 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3034 @code{@value{LD}}. Although the linker never arranges two partial programs'
3035 data sections with overlapping addresses after linking, @emph{by definition}
3036 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3037 part of a program is always the same address when the program is running as
3038 address @code{@{absolute@ 239@}} in any other part of the program.
3040 The idea of sections is extended to the @dfn{undefined} section. Any
3041 address whose section is unknown at assembly time is by definition
3042 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3043 Since numbers are always defined, the only way to generate an undefined
3044 address is to mention an undefined symbol. A reference to a named
3045 common block would be such a symbol: its value is unknown at assembly
3046 time so it has section @emph{undefined}.
3048 By analogy the word @emph{section} is used to describe groups of sections in
3049 the linked program. @code{@value{LD}} puts all partial programs' text
3050 sections in contiguous addresses in the linked program. It is
3051 customary to refer to the @emph{text section} of a program, meaning all
3052 the addresses of all partial programs' text sections. Likewise for
3053 data and bss sections.
3055 Some sections are manipulated by @code{@value{LD}}; others are invented for
3056 use of @command{@value{AS}} and have no meaning except during assembly.
3059 @section Linker Sections
3060 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3065 @cindex named sections
3066 @cindex sections, named
3067 @item named sections
3070 @cindex text section
3071 @cindex data section
3075 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3076 separate but equal sections. Anything you can say of one section is
3079 When the program is running, however, it is
3080 customary for the text section to be unalterable. The
3081 text section is often shared among processes: it contains
3082 instructions, constants and the like. The data section of a running
3083 program is usually alterable: for example, C variables would be stored
3084 in the data section.
3089 This section contains zeroed bytes when your program begins running. It
3090 is used to hold uninitialized variables or common storage. The length of
3091 each partial program's bss section is important, but because it starts
3092 out containing zeroed bytes there is no need to store explicit zero
3093 bytes in the object file. The bss section was invented to eliminate
3094 those explicit zeros from object files.
3096 @cindex absolute section
3097 @item absolute section
3098 Address 0 of this section is always ``relocated'' to runtime address 0.
3099 This is useful if you want to refer to an address that @code{@value{LD}} must
3100 not change when relocating. In this sense we speak of absolute
3101 addresses being ``unrelocatable'': they do not change during relocation.
3103 @cindex undefined section
3104 @item undefined section
3105 This ``section'' is a catch-all for address references to objects not in
3106 the preceding sections.
3107 @c FIXME: ref to some other doc on obj-file formats could go here.
3110 @cindex relocation example
3111 An idealized example of three relocatable sections follows.
3113 The example uses the traditional section names @samp{.text} and @samp{.data}.
3115 Memory addresses are on the horizontal axis.
3119 @c END TEXI2ROFF-KILL
3122 partial program # 1: |ttttt|dddd|00|
3129 partial program # 2: |TTT|DDD|000|
3132 +--+---+-----+--+----+---+-----+~~
3133 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3134 +--+---+-----+--+----+---+-----+~~
3136 addresses: 0 @dots{}
3143 \line{\it Partial program \#1: \hfil}
3144 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3145 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3147 \line{\it Partial program \#2: \hfil}
3148 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3149 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3151 \line{\it linked program: \hfil}
3152 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3153 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3154 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3155 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3157 \line{\it addresses: \hfil}
3161 @c END TEXI2ROFF-KILL
3164 @section Assembler Internal Sections
3166 @cindex internal assembler sections
3167 @cindex sections in messages, internal
3168 These sections are meant only for the internal use of @command{@value{AS}}. They
3169 have no meaning at run-time. You do not really need to know about these
3170 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3171 warning messages, so it might be helpful to have an idea of their
3172 meanings to @command{@value{AS}}. These sections are used to permit the
3173 value of every expression in your assembly language program to be a
3174 section-relative address.
3177 @cindex assembler internal logic error
3178 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3179 An internal assembler logic error has been found. This means there is a
3180 bug in the assembler.
3182 @cindex expr (internal section)
3184 The assembler stores complex expression internally as combinations of
3185 symbols. When it needs to represent an expression as a symbol, it puts
3186 it in the expr section.
3188 @c FIXME item transfer[t] vector preload
3189 @c FIXME item transfer[t] vector postload
3190 @c FIXME item register
3194 @section Sub-Sections
3196 @cindex numbered subsections
3197 @cindex grouping data
3203 fall into two sections: text and data.
3205 You may have separate groups of
3207 data in named sections
3211 data in named sections
3217 that you want to end up near to each other in the object file, even though they
3218 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3219 use @dfn{subsections} for this purpose. Within each section, there can be
3220 numbered subsections with values from 0 to 8192. Objects assembled into the
3221 same subsection go into the object file together with other objects in the same
3222 subsection. For example, a compiler might want to store constants in the text
3223 section, but might not want to have them interspersed with the program being
3224 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3225 section of code being output, and a @samp{.text 1} before each group of
3226 constants being output.
3228 Subsections are optional. If you do not use subsections, everything
3229 goes in subsection number zero.
3232 Each subsection is zero-padded up to a multiple of four bytes.
3233 (Subsections may be padded a different amount on different flavors
3234 of @command{@value{AS}}.)
3238 On the H8/300 platform, each subsection is zero-padded to a word
3239 boundary (two bytes).
3240 The same is true on the Renesas SH.
3243 @c FIXME section padding (alignment)?
3244 @c Rich Pixley says padding here depends on target obj code format; that
3245 @c doesn't seem particularly useful to say without further elaboration,
3246 @c so for now I say nothing about it. If this is a generic BFD issue,
3247 @c these paragraphs might need to vanish from this manual, and be
3248 @c discussed in BFD chapter of binutils (or some such).
3252 Subsections appear in your object file in numeric order, lowest numbered
3253 to highest. (All this to be compatible with other people's assemblers.)
3254 The object file contains no representation of subsections; @code{@value{LD}} and
3255 other programs that manipulate object files see no trace of them.
3256 They just see all your text subsections as a text section, and all your
3257 data subsections as a data section.
3259 To specify which subsection you want subsequent statements assembled
3260 into, use a numeric argument to specify it, in a @samp{.text
3261 @var{expression}} or a @samp{.data @var{expression}} statement.
3264 When generating COFF output, you
3269 can also use an extra subsection
3270 argument with arbitrary named sections: @samp{.section @var{name},
3275 When generating ELF output, you
3280 can also use the @code{.subsection} directive (@pxref{SubSection})
3281 to specify a subsection: @samp{.subsection @var{expression}}.
3283 @var{Expression} should be an absolute expression
3284 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3285 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3286 begins in @code{text 0}. For instance:
3288 .text 0 # The default subsection is text 0 anyway.
3289 .ascii "This lives in the first text subsection. *"
3291 .ascii "But this lives in the second text subsection."
3293 .ascii "This lives in the data section,"
3294 .ascii "in the first data subsection."
3296 .ascii "This lives in the first text section,"
3297 .ascii "immediately following the asterisk (*)."
3300 Each section has a @dfn{location counter} incremented by one for every byte
3301 assembled into that section. Because subsections are merely a convenience
3302 restricted to @command{@value{AS}} there is no concept of a subsection location
3303 counter. There is no way to directly manipulate a location counter---but the
3304 @code{.align} directive changes it, and any label definition captures its
3305 current value. The location counter of the section where statements are being
3306 assembled is said to be the @dfn{active} location counter.
3309 @section bss Section
3312 @cindex common variable storage
3313 The bss section is used for local common variable storage.
3314 You may allocate address space in the bss section, but you may
3315 not dictate data to load into it before your program executes. When
3316 your program starts running, all the contents of the bss
3317 section are zeroed bytes.
3319 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3320 @ref{Lcomm,,@code{.lcomm}}.
3322 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3323 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3326 When assembling for a target which supports multiple sections, such as ELF or
3327 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3328 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3329 section. Typically the section will only contain symbol definitions and
3330 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3337 Symbols are a central concept: the programmer uses symbols to name
3338 things, the linker uses symbols to link, and the debugger uses symbols
3342 @cindex debuggers, and symbol order
3343 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3344 the same order they were declared. This may break some debuggers.
3349 * Setting Symbols:: Giving Symbols Other Values
3350 * Symbol Names:: Symbol Names
3351 * Dot:: The Special Dot Symbol
3352 * Symbol Attributes:: Symbol Attributes
3359 A @dfn{label} is written as a symbol immediately followed by a colon
3360 @samp{:}. The symbol then represents the current value of the
3361 active location counter, and is, for example, a suitable instruction
3362 operand. You are warned if you use the same symbol to represent two
3363 different locations: the first definition overrides any other
3367 On the HPPA, the usual form for a label need not be immediately followed by a
3368 colon, but instead must start in column zero. Only one label may be defined on
3369 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3370 provides a special directive @code{.label} for defining labels more flexibly.
3373 @node Setting Symbols
3374 @section Giving Symbols Other Values
3376 @cindex assigning values to symbols
3377 @cindex symbol values, assigning
3378 A symbol can be given an arbitrary value by writing a symbol, followed
3379 by an equals sign @samp{=}, followed by an expression
3380 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3381 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3382 equals sign @samp{=}@samp{=} here represents an equivalent of the
3383 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3386 Blackfin does not support symbol assignment with @samp{=}.
3390 @section Symbol Names
3392 @cindex symbol names
3393 @cindex names, symbol
3394 @ifclear SPECIAL-SYMS
3395 Symbol names begin with a letter or with one of @samp{._}. On most
3396 machines, you can also use @code{$} in symbol names; exceptions are
3397 noted in @ref{Machine Dependencies}. That character may be followed by any
3398 string of digits, letters, dollar signs (unless otherwise noted for a
3399 particular target machine), and underscores.
3403 Symbol names begin with a letter or with one of @samp{._}. On the
3404 Renesas SH you can also use @code{$} in symbol names. That
3405 character may be followed by any string of digits, letters, dollar signs (save
3406 on the H8/300), and underscores.
3410 Case of letters is significant: @code{foo} is a different symbol name
3413 Each symbol has exactly one name. Each name in an assembly language program
3414 refers to exactly one symbol. You may use that symbol name any number of times
3417 @subheading Local Symbol Names
3419 @cindex local symbol names
3420 @cindex symbol names, local
3421 A local symbol is any symbol beginning with certain local label prefixes.
3422 By default, the local label prefix is @samp{.L} for ELF systems or
3423 @samp{L} for traditional a.out systems, but each target may have its own
3424 set of local label prefixes.
3426 On the HPPA local symbols begin with @samp{L$}.
3429 Local symbols are defined and used within the assembler, but they are
3430 normally not saved in object files. Thus, they are not visible when debugging.
3431 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols:
3432 @option{-L}}) to retain the local symbols in the object files.
3434 @subheading Local Labels
3436 @cindex local labels
3437 @cindex temporary symbol names
3438 @cindex symbol names, temporary
3439 Local labels help compilers and programmers use names temporarily.
3440 They create symbols which are guaranteed to be unique over the entire scope of
3441 the input source code and which can be referred to by a simple notation.
3442 To define a local label, write a label of the form @samp{@b{N}:} (where @b{N}
3443 represents any positive integer). To refer to the most recent previous
3444 definition of that label write @samp{@b{N}b}, using the same number as when
3445 you defined the label. To refer to the next definition of a local label, write
3446 @samp{@b{N}f}---the @samp{b} stands for ``backwards'' and the @samp{f} stands
3449 There is no restriction on how you can use these labels, and you can reuse them
3450 too. So that it is possible to repeatedly define the same local label (using
3451 the same number @samp{@b{N}}), although you can only refer to the most recently
3452 defined local label of that number (for a backwards reference) or the next
3453 definition of a specific local label for a forward reference. It is also worth
3454 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3455 implemented in a slightly more efficient manner than the others.
3466 Which is the equivalent of:
3469 label_1: branch label_3
3470 label_2: branch label_1
3471 label_3: branch label_4
3472 label_4: branch label_3
3475 Local label names are only a notational device. They are immediately
3476 transformed into more conventional symbol names before the assembler uses them.
3477 The symbol names are stored in the symbol table, appear in error messages, and
3478 are optionally emitted to the object file. The names are constructed using
3482 @item @emph{local label prefix}
3483 All local symbols begin with the system-specific local label prefix.
3484 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3485 that start with the local label prefix. These labels are
3486 used for symbols you are never intended to see. If you use the
3487 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3488 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3489 you may use them in debugging.
3492 This is the number that was used in the local label definition. So if the
3493 label is written @samp{55:} then the number is @samp{55}.
3496 This unusual character is included so you do not accidentally invent a symbol
3497 of the same name. The character has ASCII value of @samp{\002} (control-B).
3499 @item @emph{ordinal number}
3500 This is a serial number to keep the labels distinct. The first definition of
3501 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3502 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3503 the number @samp{1} and its 15th definition gets @samp{15} as well.
3506 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3507 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3509 @subheading Dollar Local Labels
3510 @cindex dollar local symbols
3512 @code{@value{AS}} also supports an even more local form of local labels called
3513 dollar labels. These labels go out of scope (i.e., they become undefined) as
3514 soon as a non-local label is defined. Thus they remain valid for only a small
3515 region of the input source code. Normal local labels, by contrast, remain in
3516 scope for the entire file, or until they are redefined by another occurrence of
3517 the same local label.
3519 Dollar labels are defined in exactly the same way as ordinary local labels,
3520 except that they have a dollar sign suffix to their numeric value, e.g.,
3523 They can also be distinguished from ordinary local labels by their transformed
3524 names which use ASCII character @samp{\001} (control-A) as the magic character
3525 to distinguish them from ordinary labels. For example, the fifth definition of
3526 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3529 @section The Special Dot Symbol
3531 @cindex dot (symbol)
3532 @cindex @code{.} (symbol)
3533 @cindex current address
3534 @cindex location counter
3535 The special symbol @samp{.} refers to the current address that
3536 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3537 .long .} defines @code{melvin} to contain its own address.
3538 Assigning a value to @code{.} is treated the same as a @code{.org}
3540 @ifclear no-space-dir
3541 Thus, the expression @samp{.=.+4} is the same as saying
3545 @node Symbol Attributes
3546 @section Symbol Attributes
3548 @cindex symbol attributes
3549 @cindex attributes, symbol
3550 Every symbol has, as well as its name, the attributes ``Value'' and
3551 ``Type''. Depending on output format, symbols can also have auxiliary
3554 The detailed definitions are in @file{a.out.h}.
3557 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3558 all these attributes, and probably won't warn you. This makes the
3559 symbol an externally defined symbol, which is generally what you
3563 * Symbol Value:: Value
3564 * Symbol Type:: Type
3567 * a.out Symbols:: Symbol Attributes: @code{a.out}
3571 * a.out Symbols:: Symbol Attributes: @code{a.out}
3574 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3579 * COFF Symbols:: Symbol Attributes for COFF
3582 * SOM Symbols:: Symbol Attributes for SOM
3589 @cindex value of a symbol
3590 @cindex symbol value
3591 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3592 location in the text, data, bss or absolute sections the value is the
3593 number of addresses from the start of that section to the label.
3594 Naturally for text, data and bss sections the value of a symbol changes
3595 as @code{@value{LD}} changes section base addresses during linking. Absolute
3596 symbols' values do not change during linking: that is why they are
3599 The value of an undefined symbol is treated in a special way. If it is
3600 0 then the symbol is not defined in this assembler source file, and
3601 @code{@value{LD}} tries to determine its value from other files linked into the
3602 same program. You make this kind of symbol simply by mentioning a symbol
3603 name without defining it. A non-zero value represents a @code{.comm}
3604 common declaration. The value is how much common storage to reserve, in
3605 bytes (addresses). The symbol refers to the first address of the
3611 @cindex type of a symbol
3613 The type attribute of a symbol contains relocation (section)
3614 information, any flag settings indicating that a symbol is external, and
3615 (optionally), other information for linkers and debuggers. The exact
3616 format depends on the object-code output format in use.
3621 @c The following avoids a "widow" subsection title. @group would be
3622 @c better if it were available outside examples.
3625 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3627 @cindex @code{b.out} symbol attributes
3628 @cindex symbol attributes, @code{b.out}
3629 These symbol attributes appear only when @command{@value{AS}} is configured for
3630 one of the Berkeley-descended object output formats---@code{a.out} or
3636 @subsection Symbol Attributes: @code{a.out}
3638 @cindex @code{a.out} symbol attributes
3639 @cindex symbol attributes, @code{a.out}
3645 @subsection Symbol Attributes: @code{a.out}
3647 @cindex @code{a.out} symbol attributes
3648 @cindex symbol attributes, @code{a.out}
3652 * Symbol Desc:: Descriptor
3653 * Symbol Other:: Other
3657 @subsubsection Descriptor
3659 @cindex descriptor, of @code{a.out} symbol
3660 This is an arbitrary 16-bit value. You may establish a symbol's
3661 descriptor value by using a @code{.desc} statement
3662 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3663 @command{@value{AS}}.
3666 @subsubsection Other
3668 @cindex other attribute, of @code{a.out} symbol
3669 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3674 @subsection Symbol Attributes for COFF
3676 @cindex COFF symbol attributes
3677 @cindex symbol attributes, COFF
3679 The COFF format supports a multitude of auxiliary symbol attributes;
3680 like the primary symbol attributes, they are set between @code{.def} and
3681 @code{.endef} directives.
3683 @subsubsection Primary Attributes
3685 @cindex primary attributes, COFF symbols
3686 The symbol name is set with @code{.def}; the value and type,
3687 respectively, with @code{.val} and @code{.type}.
3689 @subsubsection Auxiliary Attributes
3691 @cindex auxiliary attributes, COFF symbols
3692 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3693 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3694 table information for COFF.
3699 @subsection Symbol Attributes for SOM
3701 @cindex SOM symbol attributes
3702 @cindex symbol attributes, SOM
3704 The SOM format for the HPPA supports a multitude of symbol attributes set with
3705 the @code{.EXPORT} and @code{.IMPORT} directives.
3707 The attributes are described in @cite{HP9000 Series 800 Assembly
3708 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3709 @code{EXPORT} assembler directive documentation.
3713 @chapter Expressions
3717 @cindex numeric values
3718 An @dfn{expression} specifies an address or numeric value.
3719 Whitespace may precede and/or follow an expression.
3721 The result of an expression must be an absolute number, or else an offset into
3722 a particular section. If an expression is not absolute, and there is not
3723 enough information when @command{@value{AS}} sees the expression to know its
3724 section, a second pass over the source program might be necessary to interpret
3725 the expression---but the second pass is currently not implemented.
3726 @command{@value{AS}} aborts with an error message in this situation.
3729 * Empty Exprs:: Empty Expressions
3730 * Integer Exprs:: Integer Expressions
3734 @section Empty Expressions
3736 @cindex empty expressions
3737 @cindex expressions, empty
3738 An empty expression has no value: it is just whitespace or null.
3739 Wherever an absolute expression is required, you may omit the
3740 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3741 is compatible with other assemblers.
3744 @section Integer Expressions
3746 @cindex integer expressions
3747 @cindex expressions, integer
3748 An @dfn{integer expression} is one or more @emph{arguments} delimited
3749 by @emph{operators}.
3752 * Arguments:: Arguments
3753 * Operators:: Operators
3754 * Prefix Ops:: Prefix Operators
3755 * Infix Ops:: Infix Operators
3759 @subsection Arguments
3761 @cindex expression arguments
3762 @cindex arguments in expressions
3763 @cindex operands in expressions
3764 @cindex arithmetic operands
3765 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3766 contexts arguments are sometimes called ``arithmetic operands''. In
3767 this manual, to avoid confusing them with the ``instruction operands'' of
3768 the machine language, we use the term ``argument'' to refer to parts of
3769 expressions only, reserving the word ``operand'' to refer only to machine
3770 instruction operands.
3772 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3773 @var{section} is one of text, data, bss, absolute,
3774 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3777 Numbers are usually integers.
3779 A number can be a flonum or bignum. In this case, you are warned
3780 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3781 these 32 bits are an integer. You may write integer-manipulating
3782 instructions that act on exotic constants, compatible with other
3785 @cindex subexpressions
3786 Subexpressions are a left parenthesis @samp{(} followed by an integer
3787 expression, followed by a right parenthesis @samp{)}; or a prefix
3788 operator followed by an argument.
3791 @subsection Operators
3793 @cindex operators, in expressions
3794 @cindex arithmetic functions
3795 @cindex functions, in expressions
3796 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3797 operators are followed by an argument. Infix operators appear
3798 between their arguments. Operators may be preceded and/or followed by
3802 @subsection Prefix Operator
3804 @cindex prefix operators
3805 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3806 one argument, which must be absolute.
3808 @c the tex/end tex stuff surrounding this small table is meant to make
3809 @c it align, on the printed page, with the similar table in the next
3810 @c section (which is inside an enumerate).
3812 \global\advance\leftskip by \itemindent
3817 @dfn{Negation}. Two's complement negation.
3819 @dfn{Complementation}. Bitwise not.
3823 \global\advance\leftskip by -\itemindent
3827 @subsection Infix Operators
3829 @cindex infix operators
3830 @cindex operators, permitted arguments
3831 @dfn{Infix operators} take two arguments, one on either side. Operators
3832 have precedence, but operations with equal precedence are performed left
3833 to right. Apart from @code{+} or @option{-}, both arguments must be
3834 absolute, and the result is absolute.
3837 @cindex operator precedence
3838 @cindex precedence of operators
3845 @dfn{Multiplication}.
3848 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3854 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3857 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3861 Intermediate precedence
3866 @dfn{Bitwise Inclusive Or}.
3872 @dfn{Bitwise Exclusive Or}.
3875 @dfn{Bitwise Or Not}.
3882 @cindex addition, permitted arguments
3883 @cindex plus, permitted arguments
3884 @cindex arguments for addition
3886 @dfn{Addition}. If either argument is absolute, the result has the section of
3887 the other argument. You may not add together arguments from different
3890 @cindex subtraction, permitted arguments
3891 @cindex minus, permitted arguments
3892 @cindex arguments for subtraction
3894 @dfn{Subtraction}. If the right argument is absolute, the
3895 result has the section of the left argument.
3896 If both arguments are in the same section, the result is absolute.
3897 You may not subtract arguments from different sections.
3898 @c FIXME is there still something useful to say about undefined - undefined ?
3900 @cindex comparison expressions
3901 @cindex expressions, comparison
3906 @dfn{Is Not Equal To}
3910 @dfn{Is Greater Than}
3912 @dfn{Is Greater Than Or Equal To}
3914 @dfn{Is Less Than Or Equal To}
3916 The comparison operators can be used as infix operators. A true results has a
3917 value of -1 whereas a false result has a value of 0. Note, these operators
3918 perform signed comparisons.
3921 @item Lowest Precedence
3930 These two logical operations can be used to combine the results of sub
3931 expressions. Note, unlike the comparison operators a true result returns a
3932 value of 1 but a false results does still return 0. Also note that the logical
3933 or operator has a slightly lower precedence than logical and.
3938 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3939 address; you can only have a defined section in one of the two arguments.
3942 @chapter Assembler Directives
3944 @cindex directives, machine independent
3945 @cindex pseudo-ops, machine independent
3946 @cindex machine independent directives
3947 All assembler directives have names that begin with a period (@samp{.}).
3948 The rest of the name is letters, usually in lower case.
3950 This chapter discusses directives that are available regardless of the
3951 target machine configuration for the @sc{gnu} assembler.
3953 Some machine configurations provide additional directives.
3954 @xref{Machine Dependencies}.
3957 @ifset machine-directives
3958 @xref{Machine Dependencies}, for additional directives.
3963 * Abort:: @code{.abort}
3965 * ABORT (COFF):: @code{.ABORT}
3968 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3969 * Altmacro:: @code{.altmacro}
3970 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3971 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3972 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3973 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, @code{.bundle_lock}, @code{.bundle_unlock}
3974 * Byte:: @code{.byte @var{expressions}}
3975 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
3976 * Comm:: @code{.comm @var{symbol} , @var{length} }
3977 * Data:: @code{.data @var{subsection}}
3979 * Def:: @code{.def @var{name}}
3982 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3988 * Double:: @code{.double @var{flonums}}
3989 * Eject:: @code{.eject}
3990 * Else:: @code{.else}
3991 * Elseif:: @code{.elseif}
3994 * Endef:: @code{.endef}
3997 * Endfunc:: @code{.endfunc}
3998 * Endif:: @code{.endif}
3999 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4000 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4001 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4003 * Error:: @code{.error @var{string}}
4004 * Exitm:: @code{.exitm}
4005 * Extern:: @code{.extern}
4006 * Fail:: @code{.fail}
4007 * File:: @code{.file}
4008 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4009 * Float:: @code{.float @var{flonums}}
4010 * Func:: @code{.func}
4011 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4013 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4014 * Hidden:: @code{.hidden @var{names}}
4017 * hword:: @code{.hword @var{expressions}}
4018 * Ident:: @code{.ident}
4019 * If:: @code{.if @var{absolute expression}}
4020 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4021 * Include:: @code{.include "@var{file}"}
4022 * Int:: @code{.int @var{expressions}}
4024 * Internal:: @code{.internal @var{names}}
4027 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4028 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4029 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4030 * Lflags:: @code{.lflags}
4031 @ifclear no-line-dir
4032 * Line:: @code{.line @var{line-number}}
4035 * Linkonce:: @code{.linkonce [@var{type}]}
4036 * List:: @code{.list}
4037 * Ln:: @code{.ln @var{line-number}}
4038 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4039 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4041 * Local:: @code{.local @var{names}}
4044 * Long:: @code{.long @var{expressions}}
4046 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4049 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4050 * MRI:: @code{.mri @var{val}}
4051 * Noaltmacro:: @code{.noaltmacro}
4052 * Nolist:: @code{.nolist}
4053 * Octa:: @code{.octa @var{bignums}}
4054 * Offset:: @code{.offset @var{loc}}
4055 * Org:: @code{.org @var{new-lc}, @var{fill}}
4056 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4058 * PopSection:: @code{.popsection}
4059 * Previous:: @code{.previous}
4062 * Print:: @code{.print @var{string}}
4064 * Protected:: @code{.protected @var{names}}
4067 * Psize:: @code{.psize @var{lines}, @var{columns}}
4068 * Purgem:: @code{.purgem @var{name}}
4070 * PushSection:: @code{.pushsection @var{name}}
4073 * Quad:: @code{.quad @var{bignums}}
4074 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4075 * Rept:: @code{.rept @var{count}}
4076 * Sbttl:: @code{.sbttl "@var{subheading}"}
4078 * Scl:: @code{.scl @var{class}}
4081 * Section:: @code{.section @var{name}[, @var{flags}]}
4084 * Set:: @code{.set @var{symbol}, @var{expression}}
4085 * Short:: @code{.short @var{expressions}}
4086 * Single:: @code{.single @var{flonums}}
4088 * Size:: @code{.size [@var{name} , @var{expression}]}
4090 @ifclear no-space-dir
4091 * Skip:: @code{.skip @var{size} , @var{fill}}
4094 * Sleb128:: @code{.sleb128 @var{expressions}}
4095 @ifclear no-space-dir
4096 * Space:: @code{.space @var{size} , @var{fill}}
4099 * Stab:: @code{.stabd, .stabn, .stabs}
4102 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4103 * Struct:: @code{.struct @var{expression}}
4105 * SubSection:: @code{.subsection}
4106 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4110 * Tag:: @code{.tag @var{structname}}
4113 * Text:: @code{.text @var{subsection}}
4114 * Title:: @code{.title "@var{heading}"}
4116 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4119 * Uleb128:: @code{.uleb128 @var{expressions}}
4121 * Val:: @code{.val @var{addr}}
4125 * Version:: @code{.version "@var{string}"}
4126 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4127 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4130 * Warning:: @code{.warning @var{string}}
4131 * Weak:: @code{.weak @var{names}}
4132 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4133 * Word:: @code{.word @var{expressions}}
4134 * Deprecated:: Deprecated Directives
4138 @section @code{.abort}
4140 @cindex @code{abort} directive
4141 @cindex stopping the assembly
4142 This directive stops the assembly immediately. It is for
4143 compatibility with other assemblers. The original idea was that the
4144 assembly language source would be piped into the assembler. If the sender
4145 of the source quit, it could use this directive tells @command{@value{AS}} to
4146 quit also. One day @code{.abort} will not be supported.
4150 @section @code{.ABORT} (COFF)
4152 @cindex @code{ABORT} directive
4153 When producing COFF output, @command{@value{AS}} accepts this directive as a
4154 synonym for @samp{.abort}.
4157 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
4163 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4165 @cindex padding the location counter
4166 @cindex @code{align} directive
4167 Pad the location counter (in the current subsection) to a particular storage
4168 boundary. The first expression (which must be absolute) is the alignment
4169 required, as described below.
4171 The second expression (also absolute) gives the fill value to be stored in the
4172 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4173 padding bytes are normally zero. However, on some systems, if the section is
4174 marked as containing code and the fill value is omitted, the space is filled
4175 with no-op instructions.
4177 The third expression is also absolute, and is also optional. If it is present,
4178 it is the maximum number of bytes that should be skipped by this alignment
4179 directive. If doing the alignment would require skipping more bytes than the
4180 specified maximum, then the alignment is not done at all. You can omit the
4181 fill value (the second argument) entirely by simply using two commas after the
4182 required alignment; this can be useful if you want the alignment to be filled
4183 with no-op instructions when appropriate.
4185 The way the required alignment is specified varies from system to system.
4186 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or32,
4187 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4188 alignment request in bytes. For example @samp{.align 8} advances
4189 the location counter until it is a multiple of 8. If the location counter
4190 is already a multiple of 8, no change is needed. For the tic54x, the
4191 first expression is the alignment request in words.
4193 For other systems, including ppc, i386 using a.out format, arm and
4194 strongarm, it is the
4195 number of low-order zero bits the location counter must have after
4196 advancement. For example @samp{.align 3} advances the location
4197 counter until it a multiple of 8. If the location counter is already a
4198 multiple of 8, no change is needed.
4200 This inconsistency is due to the different behaviors of the various
4201 native assemblers for these systems which GAS must emulate.
4202 GAS also provides @code{.balign} and @code{.p2align} directives,
4203 described later, which have a consistent behavior across all
4204 architectures (but are specific to GAS).
4207 @section @code{.altmacro}
4208 Enable alternate macro mode, enabling:
4211 @item LOCAL @var{name} [ , @dots{} ]
4212 One additional directive, @code{LOCAL}, is available. It is used to
4213 generate a string replacement for each of the @var{name} arguments, and
4214 replace any instances of @var{name} in each macro expansion. The
4215 replacement string is unique in the assembly, and different for each
4216 separate macro expansion. @code{LOCAL} allows you to write macros that
4217 define symbols, without fear of conflict between separate macro expansions.
4219 @item String delimiters
4220 You can write strings delimited in these other ways besides
4221 @code{"@var{string}"}:
4224 @item '@var{string}'
4225 You can delimit strings with single-quote characters.
4227 @item <@var{string}>
4228 You can delimit strings with matching angle brackets.
4231 @item single-character string escape
4232 To include any single character literally in a string (even if the
4233 character would otherwise have some special meaning), you can prefix the
4234 character with @samp{!} (an exclamation mark). For example, you can
4235 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4237 @item Expression results as strings
4238 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4239 and use the result as a string.
4243 @section @code{.ascii "@var{string}"}@dots{}
4245 @cindex @code{ascii} directive
4246 @cindex string literals
4247 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4248 separated by commas. It assembles each string (with no automatic
4249 trailing zero byte) into consecutive addresses.
4252 @section @code{.asciz "@var{string}"}@dots{}
4254 @cindex @code{asciz} directive
4255 @cindex zero-terminated strings
4256 @cindex null-terminated strings
4257 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4258 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4261 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4263 @cindex padding the location counter given number of bytes
4264 @cindex @code{balign} directive
4265 Pad the location counter (in the current subsection) to a particular
4266 storage boundary. The first expression (which must be absolute) is the
4267 alignment request in bytes. For example @samp{.balign 8} advances
4268 the location counter until it is a multiple of 8. If the location counter
4269 is already a multiple of 8, no change is needed.
4271 The second expression (also absolute) gives the fill value to be stored in the
4272 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4273 padding bytes are normally zero. However, on some systems, if the section is
4274 marked as containing code and the fill value is omitted, the space is filled
4275 with no-op instructions.
4277 The third expression is also absolute, and is also optional. If it is present,
4278 it is the maximum number of bytes that should be skipped by this alignment
4279 directive. If doing the alignment would require skipping more bytes than the
4280 specified maximum, then the alignment is not done at all. You can omit the
4281 fill value (the second argument) entirely by simply using two commas after the
4282 required alignment; this can be useful if you want the alignment to be filled
4283 with no-op instructions when appropriate.
4285 @cindex @code{balignw} directive
4286 @cindex @code{balignl} directive
4287 The @code{.balignw} and @code{.balignl} directives are variants of the
4288 @code{.balign} directive. The @code{.balignw} directive treats the fill
4289 pattern as a two byte word value. The @code{.balignl} directives treats the
4290 fill pattern as a four byte longword value. For example, @code{.balignw
4291 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4292 filled in with the value 0x368d (the exact placement of the bytes depends upon
4293 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4296 @node Bundle directives
4297 @section @code{.bundle_align_mode @var{abs-expr}}
4298 @cindex @code{bundle_align_mode} directive
4300 @cindex instruction bundle
4301 @cindex aligned instruction bundle
4302 @code{.bundle_align_mode} enables or disables @defn{aligned instruction
4303 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4304 into fixed-sized @defn{bundles}. If the argument is zero, this mode is
4305 disabled (which is the default state). If the argument it not zero, it
4306 gives the size of an instruction bundle as a power of two (as for the
4307 @code{.p2align} directive, @pxref{P2align}).
4309 For some targets, it's an ABI requirement that no instruction may span a
4310 certain aligned boundary. A @defn{bundle} is simply a sequence of
4311 instructions that starts on an aligned boundary. For example, if
4312 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4313 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4314 effect, no single instruction may span a boundary between bundles. If an
4315 instruction would start too close to the end of a bundle for the length of
4316 that particular instruction to fit within the bundle, then the space at the
4317 end of that bundle is filled with no-op instructions so the instruction
4318 starts in the next bundle. As a corollary, it's an error if any single
4319 instruction's encoding is longer than the bundle size.
4321 @section @code{.bundle_lock} and @code{.bundle_unlock}
4322 @cindex @code{bundle_lock} directive
4323 @cindex @code{bundle_unlock} directive
4324 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4325 allow explicit control over instruction bundle padding. These directives
4326 are only valid when @code{.bundle_align_mode} has been used to enable
4327 aligned instruction bundle mode. It's an error if they appear when
4328 @code{.bundle_align_mode} has not been used at all, or when the last
4329 directive was @w{@code{.bundle_align_mode 0}}.
4331 @cindex bundle-locked
4332 For some targets, it's an ABI requirement that certain instructions may
4333 appear only as part of specified permissible sequences of multiple
4334 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4335 and @code{.bundle_unlock} directives define a @defn{bundle-locked}
4336 instruction sequence. For purposes of aligned instruction bundle mode, a
4337 sequence starting with @code{.bundle_lock} and ending with
4338 @code{.bundle_unlock} is treated as a single instruction. That is, the
4339 entire sequence must fit into a single bundle and may not span a bundle
4340 boundary. If necessary, no-op instructions will be inserted before the
4341 first instruction of the sequence so that the whole sequence starts on an
4342 aligned bundle boundary. It's an error if the sequence is longer than the
4345 Bundle-locked sequences do not nest. It's an error if two
4346 @code{.bundle_lock} directives appear without an intervening
4347 @code{.bundle_unlock} directive.
4350 @section @code{.byte @var{expressions}}
4352 @cindex @code{byte} directive
4353 @cindex integers, one byte
4354 @code{.byte} expects zero or more expressions, separated by commas.
4355 Each expression is assembled into the next byte.
4357 @node CFI directives
4358 @section @code{.cfi_sections @var{section_list}}
4359 @cindex @code{cfi_sections} directive
4360 @code{.cfi_sections} may be used to specify whether CFI directives
4361 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4362 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4363 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4364 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4365 directive is not used is @code{.cfi_sections .eh_frame}.
4367 @section @code{.cfi_startproc [simple]}
4368 @cindex @code{cfi_startproc} directive
4369 @code{.cfi_startproc} is used at the beginning of each function that
4370 should have an entry in @code{.eh_frame}. It initializes some internal
4371 data structures. Don't forget to close the function by
4372 @code{.cfi_endproc}.
4374 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4375 it also emits some architecture dependent initial CFI instructions.
4377 @section @code{.cfi_endproc}
4378 @cindex @code{cfi_endproc} directive
4379 @code{.cfi_endproc} is used at the end of a function where it closes its
4380 unwind entry previously opened by
4381 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4383 @section @code{.cfi_personality @var{encoding} [, @var{exp}]}
4384 @code{.cfi_personality} defines personality routine and its encoding.
4385 @var{encoding} must be a constant determining how the personality
4386 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4387 argument is not present, otherwise second argument should be
4388 a constant or a symbol name. When using indirect encodings,
4389 the symbol provided should be the location where personality
4390 can be loaded from, not the personality routine itself.
4391 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4392 no personality routine.
4394 @section @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4395 @code{.cfi_lsda} defines LSDA and its encoding.
4396 @var{encoding} must be a constant determining how the LSDA
4397 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4398 argument is not present, otherwise second argument should be a constant
4399 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4402 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4403 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4404 address from @var{register} and add @var{offset} to it}.
4406 @section @code{.cfi_def_cfa_register @var{register}}
4407 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4408 now on @var{register} will be used instead of the old one. Offset
4411 @section @code{.cfi_def_cfa_offset @var{offset}}
4412 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4413 remains the same, but @var{offset} is new. Note that it is the
4414 absolute offset that will be added to a defined register to compute
4417 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4418 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4419 value that is added/substracted from the previous offset.
4421 @section @code{.cfi_offset @var{register}, @var{offset}}
4422 Previous value of @var{register} is saved at offset @var{offset} from
4425 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4426 Previous value of @var{register} is saved at offset @var{offset} from
4427 the current CFA register. This is transformed to @code{.cfi_offset}
4428 using the known displacement of the CFA register from the CFA.
4429 This is often easier to use, because the number will match the
4430 code it's annotating.
4432 @section @code{.cfi_register @var{register1}, @var{register2}}
4433 Previous value of @var{register1} is saved in register @var{register2}.
4435 @section @code{.cfi_restore @var{register}}
4436 @code{.cfi_restore} says that the rule for @var{register} is now the
4437 same as it was at the beginning of the function, after all initial
4438 instruction added by @code{.cfi_startproc} were executed.
4440 @section @code{.cfi_undefined @var{register}}
4441 From now on the previous value of @var{register} can't be restored anymore.
4443 @section @code{.cfi_same_value @var{register}}
4444 Current value of @var{register} is the same like in the previous frame,
4445 i.e. no restoration needed.
4447 @section @code{.cfi_remember_state},
4448 First save all current rules for all registers by @code{.cfi_remember_state},
4449 then totally screw them up by subsequent @code{.cfi_*} directives and when
4450 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4451 the previous saved state.
4453 @section @code{.cfi_return_column @var{register}}
4454 Change return column @var{register}, i.e. the return address is either
4455 directly in @var{register} or can be accessed by rules for @var{register}.
4457 @section @code{.cfi_signal_frame}
4458 Mark current function as signal trampoline.
4460 @section @code{.cfi_window_save}
4461 SPARC register window has been saved.
4463 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4464 Allows the user to add arbitrary bytes to the unwind info. One
4465 might use this to add OS-specific CFI opcodes, or generic CFI
4466 opcodes that GAS does not yet support.
4468 @section @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4469 The current value of @var{register} is @var{label}. The value of @var{label}
4470 will be encoded in the output file according to @var{encoding}; see the
4471 description of @code{.cfi_personality} for details on this encoding.
4473 The usefulness of equating a register to a fixed label is probably
4474 limited to the return address register. Here, it can be useful to
4475 mark a code segment that has only one return address which is reached
4476 by a direct branch and no copy of the return address exists in memory
4477 or another register.
4480 @section @code{.comm @var{symbol} , @var{length} }
4482 @cindex @code{comm} directive
4483 @cindex symbol, common
4484 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4485 common symbol in one object file may be merged with a defined or common symbol
4486 of the same name in another object file. If @code{@value{LD}} does not see a
4487 definition for the symbol--just one or more common symbols--then it will
4488 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4489 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4490 the same name, and they do not all have the same size, it will allocate space
4491 using the largest size.
4494 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4495 an optional third argument. This is the desired alignment of the symbol,
4496 specified for ELF as a byte boundary (for example, an alignment of 16 means
4497 that the least significant 4 bits of the address should be zero), and for PE
4498 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4499 boundary). The alignment must be an absolute expression, and it must be a
4500 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4501 common symbol, it will use the alignment when placing the symbol. If no
4502 alignment is specified, @command{@value{AS}} will set the alignment to the
4503 largest power of two less than or equal to the size of the symbol, up to a
4504 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4505 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4506 @samp{--section-alignment} option; image file sections in PE are aligned to
4507 multiples of 4096, which is far too large an alignment for ordinary variables.
4508 It is rather the default alignment for (non-debug) sections within object
4509 (@samp{*.o}) files, which are less strictly aligned.}.
4513 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4514 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4518 @section @code{.data @var{subsection}}
4520 @cindex @code{data} directive
4521 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4522 end of the data subsection numbered @var{subsection} (which is an
4523 absolute expression). If @var{subsection} is omitted, it defaults
4528 @section @code{.def @var{name}}
4530 @cindex @code{def} directive
4531 @cindex COFF symbols, debugging
4532 @cindex debugging COFF symbols
4533 Begin defining debugging information for a symbol @var{name}; the
4534 definition extends until the @code{.endef} directive is encountered.
4537 This directive is only observed when @command{@value{AS}} is configured for COFF
4538 format output; when producing @code{b.out}, @samp{.def} is recognized,
4545 @section @code{.desc @var{symbol}, @var{abs-expression}}
4547 @cindex @code{desc} directive
4548 @cindex COFF symbol descriptor
4549 @cindex symbol descriptor, COFF
4550 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4551 to the low 16 bits of an absolute expression.
4554 The @samp{.desc} directive is not available when @command{@value{AS}} is
4555 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4556 object format. For the sake of compatibility, @command{@value{AS}} accepts
4557 it, but produces no output, when configured for COFF.
4563 @section @code{.dim}
4565 @cindex @code{dim} directive
4566 @cindex COFF auxiliary symbol information
4567 @cindex auxiliary symbol information, COFF
4568 This directive is generated by compilers to include auxiliary debugging
4569 information in the symbol table. It is only permitted inside
4570 @code{.def}/@code{.endef} pairs.
4573 @samp{.dim} is only meaningful when generating COFF format output; when
4574 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4580 @section @code{.double @var{flonums}}
4582 @cindex @code{double} directive
4583 @cindex floating point numbers (double)
4584 @code{.double} expects zero or more flonums, separated by commas. It
4585 assembles floating point numbers.
4587 The exact kind of floating point numbers emitted depends on how
4588 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4592 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4593 in @sc{ieee} format.
4598 @section @code{.eject}
4600 @cindex @code{eject} directive
4601 @cindex new page, in listings
4602 @cindex page, in listings
4603 @cindex listing control: new page
4604 Force a page break at this point, when generating assembly listings.
4607 @section @code{.else}
4609 @cindex @code{else} directive
4610 @code{.else} is part of the @command{@value{AS}} support for conditional
4611 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4612 of code to be assembled if the condition for the preceding @code{.if}
4616 @section @code{.elseif}
4618 @cindex @code{elseif} directive
4619 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4620 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4621 @code{.if} block that would otherwise fill the entire @code{.else} section.
4624 @section @code{.end}
4626 @cindex @code{end} directive
4627 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4628 process anything in the file past the @code{.end} directive.
4632 @section @code{.endef}
4634 @cindex @code{endef} directive
4635 This directive flags the end of a symbol definition begun with
4639 @samp{.endef} is only meaningful when generating COFF format output; if
4640 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4641 directive but ignores it.
4646 @section @code{.endfunc}
4647 @cindex @code{endfunc} directive
4648 @code{.endfunc} marks the end of a function specified with @code{.func}.
4651 @section @code{.endif}
4653 @cindex @code{endif} directive
4654 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4655 it marks the end of a block of code that is only assembled
4656 conditionally. @xref{If,,@code{.if}}.
4659 @section @code{.equ @var{symbol}, @var{expression}}
4661 @cindex @code{equ} directive
4662 @cindex assigning values to symbols
4663 @cindex symbols, assigning values to
4664 This directive sets the value of @var{symbol} to @var{expression}.
4665 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
4668 The syntax for @code{equ} on the HPPA is
4669 @samp{@var{symbol} .equ @var{expression}}.
4673 The syntax for @code{equ} on the Z80 is
4674 @samp{@var{symbol} equ @var{expression}}.
4675 On the Z80 it is an eror if @var{symbol} is already defined,
4676 but the symbol is not protected from later redefinition.
4677 Compare @ref{Equiv}.
4681 @section @code{.equiv @var{symbol}, @var{expression}}
4682 @cindex @code{equiv} directive
4683 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4684 the assembler will signal an error if @var{symbol} is already defined. Note a
4685 symbol which has been referenced but not actually defined is considered to be
4688 Except for the contents of the error message, this is roughly equivalent to
4695 plus it protects the symbol from later redefinition.
4698 @section @code{.eqv @var{symbol}, @var{expression}}
4699 @cindex @code{eqv} directive
4700 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4701 evaluate the expression or any part of it immediately. Instead each time
4702 the resulting symbol is used in an expression, a snapshot of its current
4706 @section @code{.err}
4707 @cindex @code{err} directive
4708 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4709 message and, unless the @option{-Z} option was used, it will not generate an
4710 object file. This can be used to signal an error in conditionally compiled code.
4713 @section @code{.error "@var{string}"}
4714 @cindex error directive
4716 Similarly to @code{.err}, this directive emits an error, but you can specify a
4717 string that will be emitted as the error message. If you don't specify the
4718 message, it defaults to @code{".error directive invoked in source file"}.
4719 @xref{Errors, ,Error and Warning Messages}.
4722 .error "This code has not been assembled and tested."
4726 @section @code{.exitm}
4727 Exit early from the current macro definition. @xref{Macro}.
4730 @section @code{.extern}
4732 @cindex @code{extern} directive
4733 @code{.extern} is accepted in the source program---for compatibility
4734 with other assemblers---but it is ignored. @command{@value{AS}} treats
4735 all undefined symbols as external.
4738 @section @code{.fail @var{expression}}
4740 @cindex @code{fail} directive
4741 Generates an error or a warning. If the value of the @var{expression} is 500
4742 or more, @command{@value{AS}} will print a warning message. If the value is less
4743 than 500, @command{@value{AS}} will print an error message. The message will
4744 include the value of @var{expression}. This can occasionally be useful inside
4745 complex nested macros or conditional assembly.
4748 @section @code{.file}
4749 @cindex @code{file} directive
4751 @ifclear no-file-dir
4752 There are two different versions of the @code{.file} directive. Targets
4753 that support DWARF2 line number information use the DWARF2 version of
4754 @code{.file}. Other targets use the default version.
4756 @subheading Default Version
4758 @cindex logical file name
4759 @cindex file name, logical
4760 This version of the @code{.file} directive tells @command{@value{AS}} that we
4761 are about to start a new logical file. The syntax is:
4767 @var{string} is the new file name. In general, the filename is
4768 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4769 to specify an empty file name, you must give the quotes--@code{""}. This
4770 statement may go away in future: it is only recognized to be compatible with
4771 old @command{@value{AS}} programs.
4773 @subheading DWARF2 Version
4776 When emitting DWARF2 line number information, @code{.file} assigns filenames
4777 to the @code{.debug_line} file name table. The syntax is:
4780 .file @var{fileno} @var{filename}
4783 The @var{fileno} operand should be a unique positive integer to use as the
4784 index of the entry in the table. The @var{filename} operand is a C string
4787 The detail of filename indices is exposed to the user because the filename
4788 table is shared with the @code{.debug_info} section of the DWARF2 debugging
4789 information, and thus the user must know the exact indices that table
4793 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4795 @cindex @code{fill} directive
4796 @cindex writing patterns in memory
4797 @cindex patterns, writing in memory
4798 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4799 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4800 may be zero or more. @var{Size} may be zero or more, but if it is
4801 more than 8, then it is deemed to have the value 8, compatible with
4802 other people's assemblers. The contents of each @var{repeat} bytes
4803 is taken from an 8-byte number. The highest order 4 bytes are
4804 zero. The lowest order 4 bytes are @var{value} rendered in the
4805 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4806 Each @var{size} bytes in a repetition is taken from the lowest order
4807 @var{size} bytes of this number. Again, this bizarre behavior is
4808 compatible with other people's assemblers.
4810 @var{size} and @var{value} are optional.
4811 If the second comma and @var{value} are absent, @var{value} is
4812 assumed zero. If the first comma and following tokens are absent,
4813 @var{size} is assumed to be 1.
4816 @section @code{.float @var{flonums}}
4818 @cindex floating point numbers (single)
4819 @cindex @code{float} directive
4820 This directive assembles zero or more flonums, separated by commas. It
4821 has the same effect as @code{.single}.
4823 The exact kind of floating point numbers emitted depends on how
4824 @command{@value{AS}} is configured.
4825 @xref{Machine Dependencies}.
4829 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
4830 in @sc{ieee} format.
4835 @section @code{.func @var{name}[,@var{label}]}
4836 @cindex @code{func} directive
4837 @code{.func} emits debugging information to denote function @var{name}, and
4838 is ignored unless the file is assembled with debugging enabled.
4839 Only @samp{--gstabs[+]} is currently supported.
4840 @var{label} is the entry point of the function and if omitted @var{name}
4841 prepended with the @samp{leading char} is used.
4842 @samp{leading char} is usually @code{_} or nothing, depending on the target.
4843 All functions are currently defined to have @code{void} return type.
4844 The function must be terminated with @code{.endfunc}.
4847 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4849 @cindex @code{global} directive
4850 @cindex symbol, making visible to linker
4851 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
4852 @var{symbol} in your partial program, its value is made available to
4853 other partial programs that are linked with it. Otherwise,
4854 @var{symbol} takes its attributes from a symbol of the same name
4855 from another file linked into the same program.
4857 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
4858 compatibility with other assemblers.
4861 On the HPPA, @code{.global} is not always enough to make it accessible to other
4862 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
4863 @xref{HPPA Directives, ,HPPA Assembler Directives}.
4868 @section @code{.gnu_attribute @var{tag},@var{value}}
4869 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
4872 @section @code{.hidden @var{names}}
4874 @cindex @code{hidden} directive
4876 This is one of the ELF visibility directives. The other two are
4877 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
4878 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4880 This directive overrides the named symbols default visibility (which is set by
4881 their binding: local, global or weak). The directive sets the visibility to
4882 @code{hidden} which means that the symbols are not visible to other components.
4883 Such symbols are always considered to be @code{protected} as well.
4887 @section @code{.hword @var{expressions}}
4889 @cindex @code{hword} directive
4890 @cindex integers, 16-bit
4891 @cindex numbers, 16-bit
4892 @cindex sixteen bit integers
4893 This expects zero or more @var{expressions}, and emits
4894 a 16 bit number for each.
4897 This directive is a synonym for @samp{.short}; depending on the target
4898 architecture, it may also be a synonym for @samp{.word}.
4902 This directive is a synonym for @samp{.short}.
4905 This directive is a synonym for both @samp{.short} and @samp{.word}.
4910 @section @code{.ident}
4912 @cindex @code{ident} directive
4914 This directive is used by some assemblers to place tags in object files. The
4915 behavior of this directive varies depending on the target. When using the
4916 a.out object file format, @command{@value{AS}} simply accepts the directive for
4917 source-file compatibility with existing assemblers, but does not emit anything
4918 for it. When using COFF, comments are emitted to the @code{.comment} or
4919 @code{.rdata} section, depending on the target. When using ELF, comments are
4920 emitted to the @code{.comment} section.
4923 @section @code{.if @var{absolute expression}}
4925 @cindex conditional assembly
4926 @cindex @code{if} directive
4927 @code{.if} marks the beginning of a section of code which is only
4928 considered part of the source program being assembled if the argument
4929 (which must be an @var{absolute expression}) is non-zero. The end of
4930 the conditional section of code must be marked by @code{.endif}
4931 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4932 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4933 If you have several conditions to check, @code{.elseif} may be used to avoid
4934 nesting blocks if/else within each subsequent @code{.else} block.
4936 The following variants of @code{.if} are also supported:
4938 @cindex @code{ifdef} directive
4939 @item .ifdef @var{symbol}
4940 Assembles the following section of code if the specified @var{symbol}
4941 has been defined. Note a symbol which has been referenced but not yet defined
4942 is considered to be undefined.
4944 @cindex @code{ifb} directive
4945 @item .ifb @var{text}
4946 Assembles the following section of code if the operand is blank (empty).
4948 @cindex @code{ifc} directive
4949 @item .ifc @var{string1},@var{string2}
4950 Assembles the following section of code if the two strings are the same. The
4951 strings may be optionally quoted with single quotes. If they are not quoted,
4952 the first string stops at the first comma, and the second string stops at the
4953 end of the line. Strings which contain whitespace should be quoted. The
4954 string comparison is case sensitive.
4956 @cindex @code{ifeq} directive
4957 @item .ifeq @var{absolute expression}
4958 Assembles the following section of code if the argument is zero.
4960 @cindex @code{ifeqs} directive
4961 @item .ifeqs @var{string1},@var{string2}
4962 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4964 @cindex @code{ifge} directive
4965 @item .ifge @var{absolute expression}
4966 Assembles the following section of code if the argument is greater than or
4969 @cindex @code{ifgt} directive
4970 @item .ifgt @var{absolute expression}
4971 Assembles the following section of code if the argument is greater than zero.
4973 @cindex @code{ifle} directive
4974 @item .ifle @var{absolute expression}
4975 Assembles the following section of code if the argument is less than or equal
4978 @cindex @code{iflt} directive
4979 @item .iflt @var{absolute expression}
4980 Assembles the following section of code if the argument is less than zero.
4982 @cindex @code{ifnb} directive
4983 @item .ifnb @var{text}
4984 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
4985 following section of code if the operand is non-blank (non-empty).
4987 @cindex @code{ifnc} directive
4988 @item .ifnc @var{string1},@var{string2}.
4989 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4990 following section of code if the two strings are not the same.
4992 @cindex @code{ifndef} directive
4993 @cindex @code{ifnotdef} directive
4994 @item .ifndef @var{symbol}
4995 @itemx .ifnotdef @var{symbol}
4996 Assembles the following section of code if the specified @var{symbol}
4997 has not been defined. Both spelling variants are equivalent. Note a symbol
4998 which has been referenced but not yet defined is considered to be undefined.
5000 @cindex @code{ifne} directive
5001 @item .ifne @var{absolute expression}
5002 Assembles the following section of code if the argument is not equal to zero
5003 (in other words, this is equivalent to @code{.if}).
5005 @cindex @code{ifnes} directive
5006 @item .ifnes @var{string1},@var{string2}
5007 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5008 following section of code if the two strings are not the same.
5012 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5014 @cindex @code{incbin} directive
5015 @cindex binary files, including
5016 The @code{incbin} directive includes @var{file} verbatim at the current
5017 location. You can control the search paths used with the @samp{-I} command-line
5018 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5021 The @var{skip} argument skips a number of bytes from the start of the
5022 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5023 read. Note that the data is not aligned in any way, so it is the user's
5024 responsibility to make sure that proper alignment is provided both before and
5025 after the @code{incbin} directive.
5028 @section @code{.include "@var{file}"}
5030 @cindex @code{include} directive
5031 @cindex supporting files, including
5032 @cindex files, including
5033 This directive provides a way to include supporting files at specified
5034 points in your source program. The code from @var{file} is assembled as
5035 if it followed the point of the @code{.include}; when the end of the
5036 included file is reached, assembly of the original file continues. You
5037 can control the search paths used with the @samp{-I} command-line option
5038 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5042 @section @code{.int @var{expressions}}
5044 @cindex @code{int} directive
5045 @cindex integers, 32-bit
5046 Expect zero or more @var{expressions}, of any section, separated by commas.
5047 For each expression, emit a number that, at run time, is the value of that
5048 expression. The byte order and bit size of the number depends on what kind
5049 of target the assembly is for.
5053 On most forms of the H8/300, @code{.int} emits 16-bit
5054 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5061 @section @code{.internal @var{names}}
5063 @cindex @code{internal} directive
5065 This is one of the ELF visibility directives. The other two are
5066 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5067 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5069 This directive overrides the named symbols default visibility (which is set by
5070 their binding: local, global or weak). The directive sets the visibility to
5071 @code{internal} which means that the symbols are considered to be @code{hidden}
5072 (i.e., not visible to other components), and that some extra, processor specific
5073 processing must also be performed upon the symbols as well.
5077 @section @code{.irp @var{symbol},@var{values}}@dots{}
5079 @cindex @code{irp} directive
5080 Evaluate a sequence of statements assigning different values to @var{symbol}.
5081 The sequence of statements starts at the @code{.irp} directive, and is
5082 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5083 set to @var{value}, and the sequence of statements is assembled. If no
5084 @var{value} is listed, the sequence of statements is assembled once, with
5085 @var{symbol} set to the null string. To refer to @var{symbol} within the
5086 sequence of statements, use @var{\symbol}.
5088 For example, assembling
5096 is equivalent to assembling
5104 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5107 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5109 @cindex @code{irpc} directive
5110 Evaluate a sequence of statements assigning different values to @var{symbol}.
5111 The sequence of statements starts at the @code{.irpc} directive, and is
5112 terminated by an @code{.endr} directive. For each character in @var{value},
5113 @var{symbol} is set to the character, and the sequence of statements is
5114 assembled. If no @var{value} is listed, the sequence of statements is
5115 assembled once, with @var{symbol} set to the null string. To refer to
5116 @var{symbol} within the sequence of statements, use @var{\symbol}.
5118 For example, assembling
5126 is equivalent to assembling
5134 For some caveats with the spelling of @var{symbol}, see also the discussion
5138 @section @code{.lcomm @var{symbol} , @var{length}}
5140 @cindex @code{lcomm} directive
5141 @cindex local common symbols
5142 @cindex symbols, local common
5143 Reserve @var{length} (an absolute expression) bytes for a local common
5144 denoted by @var{symbol}. The section and value of @var{symbol} are
5145 those of the new local common. The addresses are allocated in the bss
5146 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5147 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5148 not visible to @code{@value{LD}}.
5151 Some targets permit a third argument to be used with @code{.lcomm}. This
5152 argument specifies the desired alignment of the symbol in the bss section.
5156 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5157 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5161 @section @code{.lflags}
5163 @cindex @code{lflags} directive (ignored)
5164 @command{@value{AS}} accepts this directive, for compatibility with other
5165 assemblers, but ignores it.
5167 @ifclear no-line-dir
5169 @section @code{.line @var{line-number}}
5171 @cindex @code{line} directive
5172 @cindex logical line number
5174 Change the logical line number. @var{line-number} must be an absolute
5175 expression. The next line has that logical line number. Therefore any other
5176 statements on the current line (after a statement separator character) are
5177 reported as on logical line number @var{line-number} @minus{} 1. One day
5178 @command{@value{AS}} will no longer support this directive: it is recognized only
5179 for compatibility with existing assembler programs.
5182 Even though this is a directive associated with the @code{a.out} or
5183 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5184 when producing COFF output, and treats @samp{.line} as though it
5185 were the COFF @samp{.ln} @emph{if} it is found outside a
5186 @code{.def}/@code{.endef} pair.
5188 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5189 used by compilers to generate auxiliary symbol information for
5194 @section @code{.linkonce [@var{type}]}
5196 @cindex @code{linkonce} directive
5197 @cindex common sections
5198 Mark the current section so that the linker only includes a single copy of it.
5199 This may be used to include the same section in several different object files,
5200 but ensure that the linker will only include it once in the final output file.
5201 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5202 Duplicate sections are detected based on the section name, so it should be
5205 This directive is only supported by a few object file formats; as of this
5206 writing, the only object file format which supports it is the Portable
5207 Executable format used on Windows NT.
5209 The @var{type} argument is optional. If specified, it must be one of the
5210 following strings. For example:
5214 Not all types may be supported on all object file formats.
5218 Silently discard duplicate sections. This is the default.
5221 Warn if there are duplicate sections, but still keep only one copy.
5224 Warn if any of the duplicates have different sizes.
5227 Warn if any of the duplicates do not have exactly the same contents.
5231 @section @code{.list}
5233 @cindex @code{list} directive
5234 @cindex listing control, turning on
5235 Control (in conjunction with the @code{.nolist} directive) whether or
5236 not assembly listings are generated. These two directives maintain an
5237 internal counter (which is zero initially). @code{.list} increments the
5238 counter, and @code{.nolist} decrements it. Assembly listings are
5239 generated whenever the counter is greater than zero.
5241 By default, listings are disabled. When you enable them (with the
5242 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5243 the initial value of the listing counter is one.
5246 @section @code{.ln @var{line-number}}
5248 @cindex @code{ln} directive
5249 @ifclear no-line-dir
5250 @samp{.ln} is a synonym for @samp{.line}.
5253 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5254 must be an absolute expression. The next line has that logical
5255 line number, so any other statements on the current line (after a
5256 statement separator character @code{;}) are reported as on logical
5257 line number @var{line-number} @minus{} 1.
5260 This directive is accepted, but ignored, when @command{@value{AS}} is
5261 configured for @code{b.out}; its effect is only associated with COFF
5267 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5268 @cindex @code{loc} directive
5269 When emitting DWARF2 line number information,
5270 the @code{.loc} directive will add a row to the @code{.debug_line} line
5271 number matrix corresponding to the immediately following assembly
5272 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5273 arguments will be applied to the @code{.debug_line} state machine before
5276 The @var{options} are a sequence of the following tokens in any order:
5280 This option will set the @code{basic_block} register in the
5281 @code{.debug_line} state machine to @code{true}.
5284 This option will set the @code{prologue_end} register in the
5285 @code{.debug_line} state machine to @code{true}.
5287 @item epilogue_begin
5288 This option will set the @code{epilogue_begin} register in the
5289 @code{.debug_line} state machine to @code{true}.
5291 @item is_stmt @var{value}
5292 This option will set the @code{is_stmt} register in the
5293 @code{.debug_line} state machine to @code{value}, which must be
5296 @item isa @var{value}
5297 This directive will set the @code{isa} register in the @code{.debug_line}
5298 state machine to @var{value}, which must be an unsigned integer.
5300 @item discriminator @var{value}
5301 This directive will set the @code{discriminator} register in the @code{.debug_line}
5302 state machine to @var{value}, which must be an unsigned integer.
5306 @node Loc_mark_labels
5307 @section @code{.loc_mark_labels @var{enable}}
5308 @cindex @code{loc_mark_labels} directive
5309 When emitting DWARF2 line number information,
5310 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5311 to the @code{.debug_line} line number matrix with the @code{basic_block}
5312 register in the state machine set whenever a code label is seen.
5313 The @var{enable} argument should be either 1 or 0, to enable or disable
5314 this function respectively.
5318 @section @code{.local @var{names}}
5320 @cindex @code{local} directive
5321 This directive, which is available for ELF targets, marks each symbol in
5322 the comma-separated list of @code{names} as a local symbol so that it
5323 will not be externally visible. If the symbols do not already exist,
5324 they will be created.
5326 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5327 accept an alignment argument, which is the case for most ELF targets,
5328 the @code{.local} directive can be used in combination with @code{.comm}
5329 (@pxref{Comm}) to define aligned local common data.
5333 @section @code{.long @var{expressions}}
5335 @cindex @code{long} directive
5336 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5339 @c no one seems to know what this is for or whether this description is
5340 @c what it really ought to do
5342 @section @code{.lsym @var{symbol}, @var{expression}}
5344 @cindex @code{lsym} directive
5345 @cindex symbol, not referenced in assembly
5346 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5347 the hash table, ensuring it cannot be referenced by name during the
5348 rest of the assembly. This sets the attributes of the symbol to be
5349 the same as the expression value:
5351 @var{other} = @var{descriptor} = 0
5352 @var{type} = @r{(section of @var{expression})}
5353 @var{value} = @var{expression}
5356 The new symbol is not flagged as external.
5360 @section @code{.macro}
5363 The commands @code{.macro} and @code{.endm} allow you to define macros that
5364 generate assembly output. For example, this definition specifies a macro
5365 @code{sum} that puts a sequence of numbers into memory:
5368 .macro sum from=0, to=5
5377 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5389 @item .macro @var{macname}
5390 @itemx .macro @var{macname} @var{macargs} @dots{}
5391 @cindex @code{macro} directive
5392 Begin the definition of a macro called @var{macname}. If your macro
5393 definition requires arguments, specify their names after the macro name,
5394 separated by commas or spaces. You can qualify the macro argument to
5395 indicate whether all invocations must specify a non-blank value (through
5396 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5397 (through @samp{:@code{vararg}}). You can supply a default value for any
5398 macro argument by following the name with @samp{=@var{deflt}}. You
5399 cannot define two macros with the same @var{macname} unless it has been
5400 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5401 definitions. For example, these are all valid @code{.macro} statements:
5405 Begin the definition of a macro called @code{comm}, which takes no
5408 @item .macro plus1 p, p1
5409 @itemx .macro plus1 p p1
5410 Either statement begins the definition of a macro called @code{plus1},
5411 which takes two arguments; within the macro definition, write
5412 @samp{\p} or @samp{\p1} to evaluate the arguments.
5414 @item .macro reserve_str p1=0 p2
5415 Begin the definition of a macro called @code{reserve_str}, with two
5416 arguments. The first argument has a default value, but not the second.
5417 After the definition is complete, you can call the macro either as
5418 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5419 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5420 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5421 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5423 @item .macro m p1:req, p2=0, p3:vararg
5424 Begin the definition of a macro called @code{m}, with at least three
5425 arguments. The first argument must always have a value specified, but
5426 not the second, which instead has a default value. The third formal
5427 will get assigned all remaining arguments specified at invocation time.
5429 When you call a macro, you can specify the argument values either by
5430 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5431 @samp{sum to=17, from=9}.
5435 Note that since each of the @var{macargs} can be an identifier exactly
5436 as any other one permitted by the target architecture, there may be
5437 occasional problems if the target hand-crafts special meanings to certain
5438 characters when they occur in a special position. For example, if the colon
5439 (@code{:}) is generally permitted to be part of a symbol name, but the
5440 architecture specific code special-cases it when occurring as the final
5441 character of a symbol (to denote a label), then the macro parameter
5442 replacement code will have no way of knowing that and consider the whole
5443 construct (including the colon) an identifier, and check only this
5444 identifier for being the subject to parameter substitution. So for example
5445 this macro definition:
5453 might not work as expected. Invoking @samp{label foo} might not create a label
5454 called @samp{foo} but instead just insert the text @samp{\l:} into the
5455 assembler source, probably generating an error about an unrecognised
5458 Similarly problems might occur with the period character (@samp{.})
5459 which is often allowed inside opcode names (and hence identifier names). So
5460 for example constructing a macro to build an opcode from a base name and a
5461 length specifier like this:
5464 .macro opcode base length
5469 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5470 instruction but instead generate some kind of error as the assembler tries to
5471 interpret the text @samp{\base.\length}.
5473 There are several possible ways around this problem:
5476 @item Insert white space
5477 If it is possible to use white space characters then this is the simplest
5486 @item Use @samp{\()}
5487 The string @samp{\()} can be used to separate the end of a macro argument from
5488 the following text. eg:
5491 .macro opcode base length
5496 @item Use the alternate macro syntax mode
5497 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5498 used as a separator. eg:
5508 Note: this problem of correctly identifying string parameters to pseudo ops
5509 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5510 and @code{.irpc} (@pxref{Irpc}) as well.
5513 @cindex @code{endm} directive
5514 Mark the end of a macro definition.
5517 @cindex @code{exitm} directive
5518 Exit early from the current macro definition.
5520 @cindex number of macros executed
5521 @cindex macros, count executed
5523 @command{@value{AS}} maintains a counter of how many macros it has
5524 executed in this pseudo-variable; you can copy that number to your
5525 output with @samp{\@@}, but @emph{only within a macro definition}.
5527 @item LOCAL @var{name} [ , @dots{} ]
5528 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5529 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5530 @xref{Altmacro,,@code{.altmacro}}.
5534 @section @code{.mri @var{val}}
5536 @cindex @code{mri} directive
5537 @cindex MRI mode, temporarily
5538 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5539 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5540 affects code assembled until the next @code{.mri} directive, or until the end
5541 of the file. @xref{M, MRI mode, MRI mode}.
5544 @section @code{.noaltmacro}
5545 Disable alternate macro mode. @xref{Altmacro}.
5548 @section @code{.nolist}
5550 @cindex @code{nolist} directive
5551 @cindex listing control, turning off
5552 Control (in conjunction with the @code{.list} directive) whether or
5553 not assembly listings are generated. These two directives maintain an
5554 internal counter (which is zero initially). @code{.list} increments the
5555 counter, and @code{.nolist} decrements it. Assembly listings are
5556 generated whenever the counter is greater than zero.
5559 @section @code{.octa @var{bignums}}
5561 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5562 @cindex @code{octa} directive
5563 @cindex integer, 16-byte
5564 @cindex sixteen byte integer
5565 This directive expects zero or more bignums, separated by commas. For each
5566 bignum, it emits a 16-byte integer.
5568 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5569 hence @emph{octa}-word for 16 bytes.
5572 @section @code{.offset @var{loc}}
5574 @cindex @code{offset} directive
5575 Set the location counter to @var{loc} in the absolute section. @var{loc} must
5576 be an absolute expression. This directive may be useful for defining
5577 symbols with absolute values. Do not confuse it with the @code{.org}
5581 @section @code{.org @var{new-lc} , @var{fill}}
5583 @cindex @code{org} directive
5584 @cindex location counter, advancing
5585 @cindex advancing location counter
5586 @cindex current address, advancing
5587 Advance the location counter of the current section to
5588 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5589 expression with the same section as the current subsection. That is,
5590 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5591 wrong section, the @code{.org} directive is ignored. To be compatible
5592 with former assemblers, if the section of @var{new-lc} is absolute,
5593 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5594 is the same as the current subsection.
5596 @code{.org} may only increase the location counter, or leave it
5597 unchanged; you cannot use @code{.org} to move the location counter
5600 @c double negative used below "not undefined" because this is a specific
5601 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5602 @c section. doc@cygnus.com 18feb91
5603 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5604 may not be undefined. If you really detest this restriction we eagerly await
5605 a chance to share your improved assembler.
5607 Beware that the origin is relative to the start of the section, not
5608 to the start of the subsection. This is compatible with other
5609 people's assemblers.
5611 When the location counter (of the current subsection) is advanced, the
5612 intervening bytes are filled with @var{fill} which should be an
5613 absolute expression. If the comma and @var{fill} are omitted,
5614 @var{fill} defaults to zero.
5617 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5619 @cindex padding the location counter given a power of two
5620 @cindex @code{p2align} directive
5621 Pad the location counter (in the current subsection) to a particular
5622 storage boundary. The first expression (which must be absolute) is the
5623 number of low-order zero bits the location counter must have after
5624 advancement. For example @samp{.p2align 3} advances the location
5625 counter until it a multiple of 8. If the location counter is already a
5626 multiple of 8, no change is needed.
5628 The second expression (also absolute) gives the fill value to be stored in the
5629 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5630 padding bytes are normally zero. However, on some systems, if the section is
5631 marked as containing code and the fill value is omitted, the space is filled
5632 with no-op instructions.
5634 The third expression is also absolute, and is also optional. If it is present,
5635 it is the maximum number of bytes that should be skipped by this alignment
5636 directive. If doing the alignment would require skipping more bytes than the
5637 specified maximum, then the alignment is not done at all. You can omit the
5638 fill value (the second argument) entirely by simply using two commas after the
5639 required alignment; this can be useful if you want the alignment to be filled
5640 with no-op instructions when appropriate.
5642 @cindex @code{p2alignw} directive
5643 @cindex @code{p2alignl} directive
5644 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5645 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5646 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5647 fill pattern as a four byte longword value. For example, @code{.p2alignw
5648 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5649 filled in with the value 0x368d (the exact placement of the bytes depends upon
5650 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5655 @section @code{.popsection}
5657 @cindex @code{popsection} directive
5658 @cindex Section Stack
5659 This is one of the ELF section stack manipulation directives. The others are
5660 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5661 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5664 This directive replaces the current section (and subsection) with the top
5665 section (and subsection) on the section stack. This section is popped off the
5671 @section @code{.previous}
5673 @cindex @code{previous} directive
5674 @cindex Section Stack
5675 This is one of the ELF section stack manipulation directives. The others are
5676 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5677 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5678 (@pxref{PopSection}).
5680 This directive swaps the current section (and subsection) with most recently
5681 referenced section/subsection pair prior to this one. Multiple
5682 @code{.previous} directives in a row will flip between two sections (and their
5683 subsections). For example:
5695 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
5701 # Now in section A subsection 1
5705 # Now in section B subsection 0
5708 # Now in section B subsection 1
5711 # Now in section B subsection 0
5715 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
5716 section B and 0x9abc into subsection 1 of section B.
5718 In terms of the section stack, this directive swaps the current section with
5719 the top section on the section stack.
5723 @section @code{.print @var{string}}
5725 @cindex @code{print} directive
5726 @command{@value{AS}} will print @var{string} on the standard output during
5727 assembly. You must put @var{string} in double quotes.
5731 @section @code{.protected @var{names}}
5733 @cindex @code{protected} directive
5735 This is one of the ELF visibility directives. The other two are
5736 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5738 This directive overrides the named symbols default visibility (which is set by
5739 their binding: local, global or weak). The directive sets the visibility to
5740 @code{protected} which means that any references to the symbols from within the
5741 components that defines them must be resolved to the definition in that
5742 component, even if a definition in another component would normally preempt
5747 @section @code{.psize @var{lines} , @var{columns}}
5749 @cindex @code{psize} directive
5750 @cindex listing control: paper size
5751 @cindex paper size, for listings
5752 Use this directive to declare the number of lines---and, optionally, the
5753 number of columns---to use for each page, when generating listings.
5755 If you do not use @code{.psize}, listings use a default line-count
5756 of 60. You may omit the comma and @var{columns} specification; the
5757 default width is 200 columns.
5759 @command{@value{AS}} generates formfeeds whenever the specified number of
5760 lines is exceeded (or whenever you explicitly request one, using
5763 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5764 those explicitly specified with @code{.eject}.
5767 @section @code{.purgem @var{name}}
5769 @cindex @code{purgem} directive
5770 Undefine the macro @var{name}, so that later uses of the string will not be
5771 expanded. @xref{Macro}.
5775 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
5777 @cindex @code{pushsection} directive
5778 @cindex Section Stack
5779 This is one of the ELF section stack manipulation directives. The others are
5780 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5781 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5784 This directive pushes the current section (and subsection) onto the
5785 top of the section stack, and then replaces the current section and
5786 subsection with @code{name} and @code{subsection}. The optional
5787 @code{flags}, @code{type} and @code{arguments} are treated the same
5788 as in the @code{.section} (@pxref{Section}) directive.
5792 @section @code{.quad @var{bignums}}
5794 @cindex @code{quad} directive
5795 @code{.quad} expects zero or more bignums, separated by commas. For
5796 each bignum, it emits
5798 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5799 warning message; and just takes the lowest order 8 bytes of the bignum.
5800 @cindex eight-byte integer
5801 @cindex integer, 8-byte
5803 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5804 hence @emph{quad}-word for 8 bytes.
5807 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5808 warning message; and just takes the lowest order 16 bytes of the bignum.
5809 @cindex sixteen-byte integer
5810 @cindex integer, 16-byte
5814 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
5816 @cindex @code{reloc} directive
5817 Generate a relocation at @var{offset} of type @var{reloc_name} with value
5818 @var{expression}. If @var{offset} is a number, the relocation is generated in
5819 the current section. If @var{offset} is an expression that resolves to a
5820 symbol plus offset, the relocation is generated in the given symbol's section.
5821 @var{expression}, if present, must resolve to a symbol plus addend or to an
5822 absolute value, but note that not all targets support an addend. e.g. ELF REL
5823 targets such as i386 store an addend in the section contents rather than in the
5824 relocation. This low level interface does not support addends stored in the
5828 @section @code{.rept @var{count}}
5830 @cindex @code{rept} directive
5831 Repeat the sequence of lines between the @code{.rept} directive and the next
5832 @code{.endr} directive @var{count} times.
5834 For example, assembling
5842 is equivalent to assembling
5851 @section @code{.sbttl "@var{subheading}"}
5853 @cindex @code{sbttl} directive
5854 @cindex subtitles for listings
5855 @cindex listing control: subtitle
5856 Use @var{subheading} as the title (third line, immediately after the
5857 title line) when generating assembly listings.
5859 This directive affects subsequent pages, as well as the current page if
5860 it appears within ten lines of the top of a page.
5864 @section @code{.scl @var{class}}
5866 @cindex @code{scl} directive
5867 @cindex symbol storage class (COFF)
5868 @cindex COFF symbol storage class
5869 Set the storage-class value for a symbol. This directive may only be
5870 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
5871 whether a symbol is static or external, or it may record further
5872 symbolic debugging information.
5875 The @samp{.scl} directive is primarily associated with COFF output; when
5876 configured to generate @code{b.out} output format, @command{@value{AS}}
5877 accepts this directive but ignores it.
5883 @section @code{.section @var{name}}
5885 @cindex named section
5886 Use the @code{.section} directive to assemble the following code into a section
5889 This directive is only supported for targets that actually support arbitrarily
5890 named sections; on @code{a.out} targets, for example, it is not accepted, even
5891 with a standard @code{a.out} section name.
5895 @c only print the extra heading if both COFF and ELF are set
5896 @subheading COFF Version
5899 @cindex @code{section} directive (COFF version)
5900 For COFF targets, the @code{.section} directive is used in one of the following
5904 .section @var{name}[, "@var{flags}"]
5905 .section @var{name}[, @var{subsection}]
5908 If the optional argument is quoted, it is taken as flags to use for the
5909 section. Each flag is a single character. The following flags are recognized:
5912 bss section (uninitialized data)
5914 section is not loaded
5924 shared section (meaningful for PE targets)
5926 ignored. (For compatibility with the ELF version)
5928 section is not readable (meaningful for PE targets)
5930 single-digit power-of-two section alignment (GNU extension)
5933 If no flags are specified, the default flags depend upon the section name. If
5934 the section name is not recognized, the default will be for the section to be
5935 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
5936 from the section, rather than adding them, so if they are used on their own it
5937 will be as if no flags had been specified at all.
5939 If the optional argument to the @code{.section} directive is not quoted, it is
5940 taken as a subsection number (@pxref{Sub-Sections}).
5945 @c only print the extra heading if both COFF and ELF are set
5946 @subheading ELF Version
5949 @cindex Section Stack
5950 This is one of the ELF section stack manipulation directives. The others are
5951 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
5952 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
5953 @code{.previous} (@pxref{Previous}).
5955 @cindex @code{section} directive (ELF version)
5956 For ELF targets, the @code{.section} directive is used like this:
5959 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
5962 The optional @var{flags} argument is a quoted string which may contain any
5963 combination of the following characters:
5966 section is allocatable
5968 section is excluded from executable and shared library.
5972 section is executable
5974 section is mergeable
5976 section contains zero terminated strings
5978 section is a member of a section group
5980 section is used for thread-local-storage
5982 section is a member of the previously-current section's group, if any
5985 The optional @var{type} argument may contain one of the following constants:
5988 section contains data
5990 section does not contain data (i.e., section only occupies space)
5992 section contains data which is used by things other than the program
5994 section contains an array of pointers to init functions
5996 section contains an array of pointers to finish functions
5997 @item @@preinit_array
5998 section contains an array of pointers to pre-init functions
6001 Many targets only support the first three section types.
6003 Note on targets where the @code{@@} character is the start of a comment (eg
6004 ARM) then another character is used instead. For example the ARM port uses the
6007 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6008 be specified as well as an extra argument---@var{entsize}---like this:
6011 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6014 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6015 constants, each @var{entsize} octets long. Sections with both @code{M} and
6016 @code{S} must contain zero terminated strings where each character is
6017 @var{entsize} bytes long. The linker may remove duplicates within sections with
6018 the same name, same entity size and same flags. @var{entsize} must be an
6019 absolute expression. For sections with both @code{M} and @code{S}, a string
6020 which is a suffix of a larger string is considered a duplicate. Thus
6021 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6022 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6024 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6025 be present along with an additional field like this:
6028 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6031 The @var{GroupName} field specifies the name of the section group to which this
6032 particular section belongs. The optional linkage field can contain:
6035 indicates that only one copy of this section should be retained
6040 Note: if both the @var{M} and @var{G} flags are present then the fields for
6041 the Merge flag should come first, like this:
6044 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6047 If @var{flags} contains the @code{?} symbol then it may not also contain the
6048 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6049 present. Instead, @code{?} says to consider the section that's current before
6050 this directive. If that section used @code{G}, then the new section will use
6051 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6052 If not, then the @code{?} symbol has no effect.
6054 If no flags are specified, the default flags depend upon the section name. If
6055 the section name is not recognized, the default will be for the section to have
6056 none of the above flags: it will not be allocated in memory, nor writable, nor
6057 executable. The section will contain data.
6059 For ELF targets, the assembler supports another type of @code{.section}
6060 directive for compatibility with the Solaris assembler:
6063 .section "@var{name}"[, @var{flags}...]
6066 Note that the section name is quoted. There may be a sequence of comma
6070 section is allocatable
6074 section is executable
6076 section is excluded from executable and shared library.
6078 section is used for thread local storage
6081 This directive replaces the current section and subsection. See the
6082 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6083 some examples of how this directive and the other section stack directives
6089 @section @code{.set @var{symbol}, @var{expression}}
6091 @cindex @code{set} directive
6092 @cindex symbol value, setting
6093 Set the value of @var{symbol} to @var{expression}. This
6094 changes @var{symbol}'s value and type to conform to
6095 @var{expression}. If @var{symbol} was flagged as external, it remains
6096 flagged (@pxref{Symbol Attributes}).
6098 You may @code{.set} a symbol many times in the same assembly.
6100 If you @code{.set} a global symbol, the value stored in the object
6101 file is the last value stored into it.
6104 On Z80 @code{set} is a real instruction, use
6105 @samp{@var{symbol} defl @var{expression}} instead.
6109 @section @code{.short @var{expressions}}
6111 @cindex @code{short} directive
6113 @code{.short} is normally the same as @samp{.word}.
6114 @xref{Word,,@code{.word}}.
6116 In some configurations, however, @code{.short} and @code{.word} generate
6117 numbers of different lengths. @xref{Machine Dependencies}.
6121 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6124 This expects zero or more @var{expressions}, and emits
6125 a 16 bit number for each.
6130 @section @code{.single @var{flonums}}
6132 @cindex @code{single} directive
6133 @cindex floating point numbers (single)
6134 This directive assembles zero or more flonums, separated by commas. It
6135 has the same effect as @code{.float}.
6137 The exact kind of floating point numbers emitted depends on how
6138 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6142 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6143 numbers in @sc{ieee} format.
6149 @section @code{.size}
6151 This directive is used to set the size associated with a symbol.
6155 @c only print the extra heading if both COFF and ELF are set
6156 @subheading COFF Version
6159 @cindex @code{size} directive (COFF version)
6160 For COFF targets, the @code{.size} directive is only permitted inside
6161 @code{.def}/@code{.endef} pairs. It is used like this:
6164 .size @var{expression}
6168 @samp{.size} is only meaningful when generating COFF format output; when
6169 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6176 @c only print the extra heading if both COFF and ELF are set
6177 @subheading ELF Version
6180 @cindex @code{size} directive (ELF version)
6181 For ELF targets, the @code{.size} directive is used like this:
6184 .size @var{name} , @var{expression}
6187 This directive sets the size associated with a symbol @var{name}.
6188 The size in bytes is computed from @var{expression} which can make use of label
6189 arithmetic. This directive is typically used to set the size of function
6194 @ifclear no-space-dir
6196 @section @code{.skip @var{size} , @var{fill}}
6198 @cindex @code{skip} directive
6199 @cindex filling memory
6200 This directive emits @var{size} bytes, each of value @var{fill}. Both
6201 @var{size} and @var{fill} are absolute expressions. If the comma and
6202 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6207 @section @code{.sleb128 @var{expressions}}
6209 @cindex @code{sleb128} directive
6210 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6211 compact, variable length representation of numbers used by the DWARF
6212 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6214 @ifclear no-space-dir
6216 @section @code{.space @var{size} , @var{fill}}
6218 @cindex @code{space} directive
6219 @cindex filling memory
6220 This directive emits @var{size} bytes, each of value @var{fill}. Both
6221 @var{size} and @var{fill} are absolute expressions. If the comma
6222 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6227 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6228 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6229 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6230 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6238 @section @code{.stabd, .stabn, .stabs}
6240 @cindex symbolic debuggers, information for
6241 @cindex @code{stab@var{x}} directives
6242 There are three directives that begin @samp{.stab}.
6243 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6244 The symbols are not entered in the @command{@value{AS}} hash table: they
6245 cannot be referenced elsewhere in the source file.
6246 Up to five fields are required:
6250 This is the symbol's name. It may contain any character except
6251 @samp{\000}, so is more general than ordinary symbol names. Some
6252 debuggers used to code arbitrarily complex structures into symbol names
6256 An absolute expression. The symbol's type is set to the low 8 bits of
6257 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6258 and debuggers choke on silly bit patterns.
6261 An absolute expression. The symbol's ``other'' attribute is set to the
6262 low 8 bits of this expression.
6265 An absolute expression. The symbol's descriptor is set to the low 16
6266 bits of this expression.
6269 An absolute expression which becomes the symbol's value.
6272 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6273 or @code{.stabs} statement, the symbol has probably already been created;
6274 you get a half-formed symbol in your object file. This is
6275 compatible with earlier assemblers!
6278 @cindex @code{stabd} directive
6279 @item .stabd @var{type} , @var{other} , @var{desc}
6281 The ``name'' of the symbol generated is not even an empty string.
6282 It is a null pointer, for compatibility. Older assemblers used a
6283 null pointer so they didn't waste space in object files with empty
6286 The symbol's value is set to the location counter,
6287 relocatably. When your program is linked, the value of this symbol
6288 is the address of the location counter when the @code{.stabd} was
6291 @cindex @code{stabn} directive
6292 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6293 The name of the symbol is set to the empty string @code{""}.
6295 @cindex @code{stabs} directive
6296 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6297 All five fields are specified.
6303 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6304 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6306 @cindex string, copying to object file
6307 @cindex string8, copying to object file
6308 @cindex string16, copying to object file
6309 @cindex string32, copying to object file
6310 @cindex string64, copying to object file
6311 @cindex @code{string} directive
6312 @cindex @code{string8} directive
6313 @cindex @code{string16} directive
6314 @cindex @code{string32} directive
6315 @cindex @code{string64} directive
6317 Copy the characters in @var{str} to the object file. You may specify more than
6318 one string to copy, separated by commas. Unless otherwise specified for a
6319 particular machine, the assembler marks the end of each string with a 0 byte.
6320 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6322 The variants @code{string16}, @code{string32} and @code{string64} differ from
6323 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6324 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6325 are stored in target endianness byte order.
6331 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6332 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6337 @section @code{.struct @var{expression}}
6339 @cindex @code{struct} directive
6340 Switch to the absolute section, and set the section offset to @var{expression},
6341 which must be an absolute expression. You might use this as follows:
6350 This would define the symbol @code{field1} to have the value 0, the symbol
6351 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6352 value 8. Assembly would be left in the absolute section, and you would need to
6353 use a @code{.section} directive of some sort to change to some other section
6354 before further assembly.
6358 @section @code{.subsection @var{name}}
6360 @cindex @code{subsection} directive
6361 @cindex Section Stack
6362 This is one of the ELF section stack manipulation directives. The others are
6363 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6364 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6367 This directive replaces the current subsection with @code{name}. The current
6368 section is not changed. The replaced subsection is put onto the section stack
6369 in place of the then current top of stack subsection.
6374 @section @code{.symver}
6375 @cindex @code{symver} directive
6376 @cindex symbol versioning
6377 @cindex versions of symbols
6378 Use the @code{.symver} directive to bind symbols to specific version nodes
6379 within a source file. This is only supported on ELF platforms, and is
6380 typically used when assembling files to be linked into a shared library.
6381 There are cases where it may make sense to use this in objects to be bound
6382 into an application itself so as to override a versioned symbol from a
6385 For ELF targets, the @code{.symver} directive can be used like this:
6387 .symver @var{name}, @var{name2@@nodename}
6389 If the symbol @var{name} is defined within the file
6390 being assembled, the @code{.symver} directive effectively creates a symbol
6391 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6392 just don't try and create a regular alias is that the @var{@@} character isn't
6393 permitted in symbol names. The @var{name2} part of the name is the actual name
6394 of the symbol by which it will be externally referenced. The name @var{name}
6395 itself is merely a name of convenience that is used so that it is possible to
6396 have definitions for multiple versions of a function within a single source
6397 file, and so that the compiler can unambiguously know which version of a
6398 function is being mentioned. The @var{nodename} portion of the alias should be
6399 the name of a node specified in the version script supplied to the linker when
6400 building a shared library. If you are attempting to override a versioned
6401 symbol from a shared library, then @var{nodename} should correspond to the
6402 nodename of the symbol you are trying to override.
6404 If the symbol @var{name} is not defined within the file being assembled, all
6405 references to @var{name} will be changed to @var{name2@@nodename}. If no
6406 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6409 Another usage of the @code{.symver} directive is:
6411 .symver @var{name}, @var{name2@@@@nodename}
6413 In this case, the symbol @var{name} must exist and be defined within
6414 the file being assembled. It is similar to @var{name2@@nodename}. The
6415 difference is @var{name2@@@@nodename} will also be used to resolve
6416 references to @var{name2} by the linker.
6418 The third usage of the @code{.symver} directive is:
6420 .symver @var{name}, @var{name2@@@@@@nodename}
6422 When @var{name} is not defined within the
6423 file being assembled, it is treated as @var{name2@@nodename}. When
6424 @var{name} is defined within the file being assembled, the symbol
6425 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6430 @section @code{.tag @var{structname}}
6432 @cindex COFF structure debugging
6433 @cindex structure debugging, COFF
6434 @cindex @code{tag} directive
6435 This directive is generated by compilers to include auxiliary debugging
6436 information in the symbol table. It is only permitted inside
6437 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6438 definitions in the symbol table with instances of those structures.
6441 @samp{.tag} is only used when generating COFF format output; when
6442 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6448 @section @code{.text @var{subsection}}
6450 @cindex @code{text} directive
6451 Tells @command{@value{AS}} to assemble the following statements onto the end of
6452 the text subsection numbered @var{subsection}, which is an absolute
6453 expression. If @var{subsection} is omitted, subsection number zero
6457 @section @code{.title "@var{heading}"}
6459 @cindex @code{title} directive
6460 @cindex listing control: title line
6461 Use @var{heading} as the title (second line, immediately after the
6462 source file name and pagenumber) when generating assembly listings.
6464 This directive affects subsequent pages, as well as the current page if
6465 it appears within ten lines of the top of a page.
6469 @section @code{.type}
6471 This directive is used to set the type of a symbol.
6475 @c only print the extra heading if both COFF and ELF are set
6476 @subheading COFF Version
6479 @cindex COFF symbol type
6480 @cindex symbol type, COFF
6481 @cindex @code{type} directive (COFF version)
6482 For COFF targets, this directive is permitted only within
6483 @code{.def}/@code{.endef} pairs. It is used like this:
6489 This records the integer @var{int} as the type attribute of a symbol table
6493 @samp{.type} is associated only with COFF format output; when
6494 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6495 directive but ignores it.
6501 @c only print the extra heading if both COFF and ELF are set
6502 @subheading ELF Version
6505 @cindex ELF symbol type
6506 @cindex symbol type, ELF
6507 @cindex @code{type} directive (ELF version)
6508 For ELF targets, the @code{.type} directive is used like this:
6511 .type @var{name} , @var{type description}
6514 This sets the type of symbol @var{name} to be either a
6515 function symbol or an object symbol. There are five different syntaxes
6516 supported for the @var{type description} field, in order to provide
6517 compatibility with various other assemblers.
6519 Because some of the characters used in these syntaxes (such as @samp{@@} and
6520 @samp{#}) are comment characters for some architectures, some of the syntaxes
6521 below do not work on all architectures. The first variant will be accepted by
6522 the GNU assembler on all architectures so that variant should be used for
6523 maximum portability, if you do not need to assemble your code with other
6526 The syntaxes supported are:
6529 .type <name> STT_<TYPE_IN_UPPER_CASE>
6530 .type <name>,#<type>
6531 .type <name>,@@<type>
6532 .type <name>,%<type>
6533 .type <name>,"<type>"
6536 The types supported are:
6541 Mark the symbol as being a function name.
6544 @itemx gnu_indirect_function
6545 Mark the symbol as an indirect function when evaluated during reloc
6546 processing. (This is only supported on assemblers targeting GNU systems).
6550 Mark the symbol as being a data object.
6554 Mark the symbol as being a thead-local data object.
6558 Mark the symbol as being a common data object.
6562 Does not mark the symbol in any way. It is supported just for completeness.
6564 @item gnu_unique_object
6565 Marks the symbol as being a globally unique data object. The dynamic linker
6566 will make sure that in the entire process there is just one symbol with this
6567 name and type in use. (This is only supported on assemblers targeting GNU
6572 Note: Some targets support extra types in addition to those listed above.
6578 @section @code{.uleb128 @var{expressions}}
6580 @cindex @code{uleb128} directive
6581 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6582 compact, variable length representation of numbers used by the DWARF
6583 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
6587 @section @code{.val @var{addr}}
6589 @cindex @code{val} directive
6590 @cindex COFF value attribute
6591 @cindex value attribute, COFF
6592 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6593 records the address @var{addr} as the value attribute of a symbol table
6597 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6598 configured for @code{b.out}, it accepts this directive but ignores it.
6604 @section @code{.version "@var{string}"}
6606 @cindex @code{version} directive
6607 This directive creates a @code{.note} section and places into it an ELF
6608 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6613 @section @code{.vtable_entry @var{table}, @var{offset}}
6615 @cindex @code{vtable_entry} directive
6616 This directive finds or creates a symbol @code{table} and creates a
6617 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6620 @section @code{.vtable_inherit @var{child}, @var{parent}}
6622 @cindex @code{vtable_inherit} directive
6623 This directive finds the symbol @code{child} and finds or creates the symbol
6624 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6625 parent whose addend is the value of the child symbol. As a special case the
6626 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
6630 @section @code{.warning "@var{string}"}
6631 @cindex warning directive
6632 Similar to the directive @code{.error}
6633 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6636 @section @code{.weak @var{names}}
6638 @cindex @code{weak} directive
6639 This directive sets the weak attribute on the comma separated list of symbol
6640 @code{names}. If the symbols do not already exist, they will be created.
6642 On COFF targets other than PE, weak symbols are a GNU extension. This
6643 directive sets the weak attribute on the comma separated list of symbol
6644 @code{names}. If the symbols do not already exist, they will be created.
6646 On the PE target, weak symbols are supported natively as weak aliases.
6647 When a weak symbol is created that is not an alias, GAS creates an
6648 alternate symbol to hold the default value.
6651 @section @code{.weakref @var{alias}, @var{target}}
6653 @cindex @code{weakref} directive
6654 This directive creates an alias to the target symbol that enables the symbol to
6655 be referenced with weak-symbol semantics, but without actually making it weak.
6656 If direct references or definitions of the symbol are present, then the symbol
6657 will not be weak, but if all references to it are through weak references, the
6658 symbol will be marked as weak in the symbol table.
6660 The effect is equivalent to moving all references to the alias to a separate
6661 assembly source file, renaming the alias to the symbol in it, declaring the
6662 symbol as weak there, and running a reloadable link to merge the object files
6663 resulting from the assembly of the new source file and the old source file that
6664 had the references to the alias removed.
6666 The alias itself never makes to the symbol table, and is entirely handled
6667 within the assembler.
6670 @section @code{.word @var{expressions}}
6672 @cindex @code{word} directive
6673 This directive expects zero or more @var{expressions}, of any section,
6674 separated by commas.
6677 For each expression, @command{@value{AS}} emits a 32-bit number.
6680 For each expression, @command{@value{AS}} emits a 16-bit number.
6685 The size of the number emitted, and its byte order,
6686 depend on what target computer the assembly is for.
6689 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6690 @c happen---32-bit addressability, period; no long/short jumps.
6691 @ifset DIFF-TBL-KLUGE
6692 @cindex difference tables altered
6693 @cindex altered difference tables
6695 @emph{Warning: Special Treatment to support Compilers}
6699 Machines with a 32-bit address space, but that do less than 32-bit
6700 addressing, require the following special treatment. If the machine of
6701 interest to you does 32-bit addressing (or doesn't require it;
6702 @pxref{Machine Dependencies}), you can ignore this issue.
6705 In order to assemble compiler output into something that works,
6706 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6707 Directives of the form @samp{.word sym1-sym2} are often emitted by
6708 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6709 directive of the form @samp{.word sym1-sym2}, and the difference between
6710 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6711 creates a @dfn{secondary jump table}, immediately before the next label.
6712 This secondary jump table is preceded by a short-jump to the
6713 first byte after the secondary table. This short-jump prevents the flow
6714 of control from accidentally falling into the new table. Inside the
6715 table is a long-jump to @code{sym2}. The original @samp{.word}
6716 contains @code{sym1} minus the address of the long-jump to
6719 If there were several occurrences of @samp{.word sym1-sym2} before the
6720 secondary jump table, all of them are adjusted. If there was a
6721 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6722 long-jump to @code{sym4} is included in the secondary jump table,
6723 and the @code{.word} directives are adjusted to contain @code{sym3}
6724 minus the address of the long-jump to @code{sym4}; and so on, for as many
6725 entries in the original jump table as necessary.
6728 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6729 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6730 assembly language programmers.
6733 @c end DIFF-TBL-KLUGE
6736 @section Deprecated Directives
6738 @cindex deprecated directives
6739 @cindex obsolescent directives
6740 One day these directives won't work.
6741 They are included for compatibility with older assemblers.
6748 @node Object Attributes
6749 @chapter Object Attributes
6750 @cindex object attributes
6752 @command{@value{AS}} assembles source files written for a specific architecture
6753 into object files for that architecture. But not all object files are alike.
6754 Many architectures support incompatible variations. For instance, floating
6755 point arguments might be passed in floating point registers if the object file
6756 requires hardware floating point support---or floating point arguments might be
6757 passed in integer registers if the object file supports processors with no
6758 hardware floating point unit. Or, if two objects are built for different
6759 generations of the same architecture, the combination may require the
6760 newer generation at run-time.
6762 This information is useful during and after linking. At link time,
6763 @command{@value{LD}} can warn about incompatible object files. After link
6764 time, tools like @command{gdb} can use it to process the linked file
6767 Compatibility information is recorded as a series of object attributes. Each
6768 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
6769 string, and indicates who sets the meaning of the tag. The tag is an integer,
6770 and indicates what property the attribute describes. The value may be a string
6771 or an integer, and indicates how the property affects this object. Missing
6772 attributes are the same as attributes with a zero value or empty string value.
6774 Object attributes were developed as part of the ABI for the ARM Architecture.
6775 The file format is documented in @cite{ELF for the ARM Architecture}.
6778 * GNU Object Attributes:: @sc{gnu} Object Attributes
6779 * Defining New Object Attributes:: Defining New Object Attributes
6782 @node GNU Object Attributes
6783 @section @sc{gnu} Object Attributes
6785 The @code{.gnu_attribute} directive records an object attribute
6786 with vendor @samp{gnu}.
6788 Except for @samp{Tag_compatibility}, which has both an integer and a string for
6789 its value, @sc{gnu} attributes have a string value if the tag number is odd and
6790 an integer value if the tag number is even. The second bit (@code{@var{tag} &
6791 2} is set for architecture-independent attributes and clear for
6792 architecture-dependent ones.
6794 @subsection Common @sc{gnu} attributes
6796 These attributes are valid on all architectures.
6799 @item Tag_compatibility (32)
6800 The compatibility attribute takes an integer flag value and a vendor name. If
6801 the flag value is 0, the file is compatible with other toolchains. If it is 1,
6802 then the file is only compatible with the named toolchain. If it is greater
6803 than 1, the file can only be processed by other toolchains under some private
6804 arrangement indicated by the flag value and the vendor name.
6807 @subsection MIPS Attributes
6810 @item Tag_GNU_MIPS_ABI_FP (4)
6811 The floating-point ABI used by this object file. The value will be:
6815 0 for files not affected by the floating-point ABI.
6817 1 for files using the hardware floating-point with a standard double-precision
6820 2 for files using the hardware floating-point ABI with a single-precision FPU.
6822 3 for files using the software floating-point ABI.
6824 4 for files using the hardware floating-point ABI with 64-bit wide
6825 double-precision floating-point registers and 32-bit wide general
6830 @subsection PowerPC Attributes
6833 @item Tag_GNU_Power_ABI_FP (4)
6834 The floating-point ABI used by this object file. The value will be:
6838 0 for files not affected by the floating-point ABI.
6840 1 for files using double-precision hardware floating-point ABI.
6842 2 for files using the software floating-point ABI.
6844 3 for files using single-precision hardware floating-point ABI.
6847 @item Tag_GNU_Power_ABI_Vector (8)
6848 The vector ABI used by this object file. The value will be:
6852 0 for files not affected by the vector ABI.
6854 1 for files using general purpose registers to pass vectors.
6856 2 for files using AltiVec registers to pass vectors.
6858 3 for files using SPE registers to pass vectors.
6862 @node Defining New Object Attributes
6863 @section Defining New Object Attributes
6865 If you want to define a new @sc{gnu} object attribute, here are the places you
6866 will need to modify. New attributes should be discussed on the @samp{binutils}
6871 This manual, which is the official register of attributes.
6873 The header for your architecture @file{include/elf}, to define the tag.
6875 The @file{bfd} support file for your architecture, to merge the attribute
6876 and issue any appropriate link warnings.
6878 Test cases in @file{ld/testsuite} for merging and link warnings.
6880 @file{binutils/readelf.c} to display your attribute.
6882 GCC, if you want the compiler to mark the attribute automatically.
6888 @node Machine Dependencies
6889 @chapter Machine Dependent Features
6891 @cindex machine dependencies
6892 The machine instruction sets are (almost by definition) different on
6893 each machine where @command{@value{AS}} runs. Floating point representations
6894 vary as well, and @command{@value{AS}} often supports a few additional
6895 directives or command-line options for compatibility with other
6896 assemblers on a particular platform. Finally, some versions of
6897 @command{@value{AS}} support special pseudo-instructions for branch
6900 This chapter discusses most of these differences, though it does not
6901 include details on any machine's instruction set. For details on that
6902 subject, see the hardware manufacturer's manual.
6906 * Alpha-Dependent:: Alpha Dependent Features
6909 * ARC-Dependent:: ARC Dependent Features
6912 * ARM-Dependent:: ARM Dependent Features
6915 * AVR-Dependent:: AVR Dependent Features
6918 * Blackfin-Dependent:: Blackfin Dependent Features
6921 * CR16-Dependent:: CR16 Dependent Features
6924 * CRIS-Dependent:: CRIS Dependent Features
6927 * D10V-Dependent:: D10V Dependent Features
6930 * D30V-Dependent:: D30V Dependent Features
6933 * Epiphany-Dependent:: EPIPHANY Dependent Features
6936 * H8/300-Dependent:: Renesas H8/300 Dependent Features
6939 * HPPA-Dependent:: HPPA Dependent Features
6942 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
6945 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
6948 * i860-Dependent:: Intel 80860 Dependent Features
6951 * i960-Dependent:: Intel 80960 Dependent Features
6954 * IA-64-Dependent:: Intel IA-64 Dependent Features
6957 * IP2K-Dependent:: IP2K Dependent Features
6960 * LM32-Dependent:: LM32 Dependent Features
6963 * M32C-Dependent:: M32C Dependent Features
6966 * M32R-Dependent:: M32R Dependent Features
6969 * M68K-Dependent:: M680x0 Dependent Features
6972 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
6975 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
6978 * MIPS-Dependent:: MIPS Dependent Features
6981 * MMIX-Dependent:: MMIX Dependent Features
6984 * MSP430-Dependent:: MSP430 Dependent Features
6987 * NS32K-Dependent:: NS32K Dependent Features
6990 * SH-Dependent:: Renesas / SuperH SH Dependent Features
6991 * SH64-Dependent:: SuperH SH64 Dependent Features
6994 * PDP-11-Dependent:: PDP-11 Dependent Features
6997 * PJ-Dependent:: picoJava Dependent Features
7000 * PPC-Dependent:: PowerPC Dependent Features
7003 * RL78-Dependent:: RL78 Dependent Features
7006 * RX-Dependent:: RX Dependent Features
7009 * S/390-Dependent:: IBM S/390 Dependent Features
7012 * SCORE-Dependent:: SCORE Dependent Features
7015 * Sparc-Dependent:: SPARC Dependent Features
7018 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7021 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7024 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7027 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7030 * V850-Dependent:: V850 Dependent Features
7033 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7036 * Xtensa-Dependent:: Xtensa Dependent Features
7039 * Z80-Dependent:: Z80 Dependent Features
7042 * Z8000-Dependent:: Z8000 Dependent Features
7045 * Vax-Dependent:: VAX Dependent Features
7052 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7053 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7054 @c peculiarity: to preserve cross-references, there must be a node called
7055 @c "Machine Dependencies". Hence the conditional nodenames in each
7056 @c major node below. Node defaulting in makeinfo requires adjacency of
7057 @c node and sectioning commands; hence the repetition of @chapter BLAH
7058 @c in both conditional blocks.
7061 @include c-alpha.texi
7077 @include c-bfin.texi
7081 @include c-cr16.texi
7085 @include c-cris.texi
7090 @node Machine Dependencies
7091 @chapter Machine Dependent Features
7093 The machine instruction sets are different on each Renesas chip family,
7094 and there are also some syntax differences among the families. This
7095 chapter describes the specific @command{@value{AS}} features for each
7099 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7100 * SH-Dependent:: Renesas SH Dependent Features
7107 @include c-d10v.texi
7111 @include c-d30v.texi
7115 @include c-epiphany.texi
7119 @include c-h8300.texi
7123 @include c-hppa.texi
7127 @include c-i370.texi
7131 @include c-i386.texi
7135 @include c-i860.texi
7139 @include c-i960.texi
7143 @include c-ia64.texi
7147 @include c-ip2k.texi
7151 @include c-lm32.texi
7155 @include c-m32c.texi
7159 @include c-m32r.texi
7163 @include c-m68k.texi
7167 @include c-m68hc11.texi
7171 @include c-microblaze.texi
7175 @include c-mips.texi
7179 @include c-mmix.texi
7183 @include c-msp430.texi
7187 @include c-ns32k.texi
7191 @include c-pdp11.texi
7203 @include c-rl78.texi
7211 @include c-s390.texi
7215 @include c-score.texi
7220 @include c-sh64.texi
7224 @include c-sparc.texi
7228 @include c-tic54x.texi
7232 @include c-tic6x.texi
7236 @include c-tilegx.texi
7240 @include c-tilepro.texi
7256 @include c-v850.texi
7260 @include c-xstormy16.texi
7264 @include c-xtensa.texi
7268 @c reverse effect of @down at top of generic Machine-Dep chapter
7272 @node Reporting Bugs
7273 @chapter Reporting Bugs
7274 @cindex bugs in assembler
7275 @cindex reporting bugs in assembler
7277 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7279 Reporting a bug may help you by bringing a solution to your problem, or it may
7280 not. But in any case the principal function of a bug report is to help the
7281 entire community by making the next version of @command{@value{AS}} work better.
7282 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7284 In order for a bug report to serve its purpose, you must include the
7285 information that enables us to fix the bug.
7288 * Bug Criteria:: Have you found a bug?
7289 * Bug Reporting:: How to report bugs
7293 @section Have You Found a Bug?
7294 @cindex bug criteria
7296 If you are not sure whether you have found a bug, here are some guidelines:
7299 @cindex fatal signal
7300 @cindex assembler crash
7301 @cindex crash of assembler
7303 If the assembler gets a fatal signal, for any input whatever, that is a
7304 @command{@value{AS}} bug. Reliable assemblers never crash.
7306 @cindex error on valid input
7308 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7310 @cindex invalid input
7312 If @command{@value{AS}} does not produce an error message for invalid input, that
7313 is a bug. However, you should note that your idea of ``invalid input'' might
7314 be our idea of ``an extension'' or ``support for traditional practice''.
7317 If you are an experienced user of assemblers, your suggestions for improvement
7318 of @command{@value{AS}} are welcome in any case.
7322 @section How to Report Bugs
7324 @cindex assembler bugs, reporting
7326 A number of companies and individuals offer support for @sc{gnu} products. If
7327 you obtained @command{@value{AS}} from a support organization, we recommend you
7328 contact that organization first.
7330 You can find contact information for many support companies and
7331 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7335 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7339 The fundamental principle of reporting bugs usefully is this:
7340 @strong{report all the facts}. If you are not sure whether to state a
7341 fact or leave it out, state it!
7343 Often people omit facts because they think they know what causes the problem
7344 and assume that some details do not matter. Thus, you might assume that the
7345 name of a symbol you use in an example does not matter. Well, probably it does
7346 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7347 happens to fetch from the location where that name is stored in memory;
7348 perhaps, if the name were different, the contents of that location would fool
7349 the assembler into doing the right thing despite the bug. Play it safe and
7350 give a specific, complete example. That is the easiest thing for you to do,
7351 and the most helpful.
7353 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7354 it is new to us. Therefore, always write your bug reports on the assumption
7355 that the bug has not been reported previously.
7357 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7358 bell?'' This cannot help us fix a bug, so it is basically useless. We
7359 respond by asking for enough details to enable us to investigate.
7360 You might as well expedite matters by sending them to begin with.
7362 To enable us to fix the bug, you should include all these things:
7366 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
7367 it with the @samp{--version} argument.
7369 Without this, we will not know whether there is any point in looking for
7370 the bug in the current version of @command{@value{AS}}.
7373 Any patches you may have applied to the @command{@value{AS}} source.
7376 The type of machine you are using, and the operating system name and
7380 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
7384 The command arguments you gave the assembler to assemble your example and
7385 observe the bug. To guarantee you will not omit something important, list them
7386 all. A copy of the Makefile (or the output from make) is sufficient.
7388 If we were to try to guess the arguments, we would probably guess wrong
7389 and then we might not encounter the bug.
7392 A complete input file that will reproduce the bug. If the bug is observed when
7393 the assembler is invoked via a compiler, send the assembler source, not the
7394 high level language source. Most compilers will produce the assembler source
7395 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
7396 the options @samp{-v --save-temps}; this will save the assembler source in a
7397 file with an extension of @file{.s}, and also show you exactly how
7398 @command{@value{AS}} is being run.
7401 A description of what behavior you observe that you believe is
7402 incorrect. For example, ``It gets a fatal signal.''
7404 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
7405 will certainly notice it. But if the bug is incorrect output, we might not
7406 notice unless it is glaringly wrong. You might as well not give us a chance to
7409 Even if the problem you experience is a fatal signal, you should still say so
7410 explicitly. Suppose something strange is going on, such as, your copy of
7411 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
7412 library on your system. (This has happened!) Your copy might crash and ours
7413 would not. If you told us to expect a crash, then when ours fails to crash, we
7414 would know that the bug was not happening for us. If you had not told us to
7415 expect a crash, then we would not be able to draw any conclusion from our
7419 If you wish to suggest changes to the @command{@value{AS}} source, send us context
7420 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
7421 option. Always send diffs from the old file to the new file. If you even
7422 discuss something in the @command{@value{AS}} source, refer to it by context, not
7425 The line numbers in our development sources will not match those in your
7426 sources. Your line numbers would convey no useful information to us.
7429 Here are some things that are not necessary:
7433 A description of the envelope of the bug.
7435 Often people who encounter a bug spend a lot of time investigating
7436 which changes to the input file will make the bug go away and which
7437 changes will not affect it.
7439 This is often time consuming and not very useful, because the way we
7440 will find the bug is by running a single example under the debugger
7441 with breakpoints, not by pure deduction from a series of examples.
7442 We recommend that you save your time for something else.
7444 Of course, if you can find a simpler example to report @emph{instead}
7445 of the original one, that is a convenience for us. Errors in the
7446 output will be easier to spot, running under the debugger will take
7447 less time, and so on.
7449 However, simplification is not vital; if you do not want to do this,
7450 report the bug anyway and send us the entire test case you used.
7453 A patch for the bug.
7455 A patch for the bug does help us if it is a good one. But do not omit
7456 the necessary information, such as the test case, on the assumption that
7457 a patch is all we need. We might see problems with your patch and decide
7458 to fix the problem another way, or we might not understand it at all.
7460 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
7461 construct an example that will make the program follow a certain path through
7462 the code. If you do not send us the example, we will not be able to construct
7463 one, so we will not be able to verify that the bug is fixed.
7465 And if we cannot understand what bug you are trying to fix, or why your
7466 patch should be an improvement, we will not install it. A test case will
7467 help us to understand.
7470 A guess about what the bug is or what it depends on.
7472 Such guesses are usually wrong. Even we cannot guess right about such
7473 things without first using the debugger to find the facts.
7476 @node Acknowledgements
7477 @chapter Acknowledgements
7479 If you have contributed to GAS and your name isn't listed here,
7480 it is not meant as a slight. We just don't know about it. Send mail to the
7481 maintainer, and we'll correct the situation. Currently
7483 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
7485 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
7488 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
7489 information and the 68k series machines, most of the preprocessing pass, and
7490 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
7492 K. Richard Pixley maintained GAS for a while, adding various enhancements and
7493 many bug fixes, including merging support for several processors, breaking GAS
7494 up to handle multiple object file format back ends (including heavy rewrite,
7495 testing, an integration of the coff and b.out back ends), adding configuration
7496 including heavy testing and verification of cross assemblers and file splits
7497 and renaming, converted GAS to strictly ANSI C including full prototypes, added
7498 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
7499 port (including considerable amounts of reverse engineering), a SPARC opcode
7500 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
7501 assertions and made them work, much other reorganization, cleanup, and lint.
7503 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
7504 in format-specific I/O modules.
7506 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
7507 has done much work with it since.
7509 The Intel 80386 machine description was written by Eliot Dresselhaus.
7511 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
7513 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
7514 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
7516 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
7517 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
7518 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
7519 support a.out format.
7521 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
7522 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
7523 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
7524 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
7527 John Gilmore built the AMD 29000 support, added @code{.include} support, and
7528 simplified the configuration of which versions accept which directives. He
7529 updated the 68k machine description so that Motorola's opcodes always produced
7530 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
7531 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
7532 cross-compilation support, and one bug in relaxation that took a week and
7533 required the proverbial one-bit fix.
7535 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
7536 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
7537 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
7538 PowerPC assembler, and made a few other minor patches.
7540 Steve Chamberlain made GAS able to generate listings.
7542 Hewlett-Packard contributed support for the HP9000/300.
7544 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
7545 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
7546 formats). This work was supported by both the Center for Software Science at
7547 the University of Utah and Cygnus Support.
7549 Support for ELF format files has been worked on by Mark Eichin of Cygnus
7550 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
7551 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
7552 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
7553 and some initial 64-bit support).
7555 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
7557 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
7558 support for openVMS/Alpha.
7560 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
7563 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
7564 Inc.@: added support for Xtensa processors.
7566 Several engineers at Cygnus Support have also provided many small bug fixes and
7567 configuration enhancements.
7569 Jon Beniston added support for the Lattice Mico32 architecture.
7571 Many others have contributed large or small bugfixes and enhancements. If
7572 you have contributed significant work and are not mentioned on this list, and
7573 want to be, let us know. Some of the history has been lost; we are not
7574 intentionally leaving anyone out.
7576 @node GNU Free Documentation License
7577 @appendix GNU Free Documentation License
7581 @unnumbered AS Index