1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright (C) 1991-2014 Free Software Foundation, Inc.
3 @c UPDATE!! On future updates--
4 @c (1) check for new machine-dep cmdline options in
5 @c md_parse_option definitions in config/tc-*.c
6 @c (2) for platform-specific directives, examine md_pseudo_op
8 @c (3) for object-format specific directives, examine obj_pseudo_op
10 @c (4) portable directives in potable[] in read.c
14 @macro gcctabopt{body}
17 @c defaults, config file may override:
22 @include asconfig.texi
27 @c common OR combinations of conditions
53 @set abnormal-separator
57 @settitle Using @value{AS}
60 @settitle Using @value{AS} (@value{TARGET})
62 @setchapternewpage odd
67 @c WARE! Some of the machine-dependent sections contain tables of machine
68 @c instructions. Except in multi-column format, these tables look silly.
69 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
70 @c the multi-col format is faked within @example sections.
72 @c Again unfortunately, the natural size that fits on a page, for these tables,
73 @c is different depending on whether or not smallbook is turned on.
74 @c This matters, because of order: text flow switches columns at each page
77 @c The format faked in this source works reasonably well for smallbook,
78 @c not well for the default large-page format. This manual expects that if you
79 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
80 @c tables in question. You can turn on one without the other at your
81 @c discretion, of course.
84 @c the insn tables look just as silly in info files regardless of smallbook,
85 @c might as well show 'em anyways.
89 @dircategory Software development
91 * As: (as). The GNU assembler.
92 * Gas: (as). The GNU assembler.
100 This file documents the GNU Assembler "@value{AS}".
102 @c man begin COPYRIGHT
103 Copyright @copyright{} 1991-2014 Free Software Foundation, Inc.
105 Permission is granted to copy, distribute and/or modify this document
106 under the terms of the GNU Free Documentation License, Version 1.3
107 or any later version published by the Free Software Foundation;
108 with no Invariant Sections, with no Front-Cover Texts, and with no
109 Back-Cover Texts. A copy of the license is included in the
110 section entitled ``GNU Free Documentation License''.
116 @title Using @value{AS}
117 @subtitle The @sc{gnu} Assembler
119 @subtitle for the @value{TARGET} family
121 @ifset VERSION_PACKAGE
123 @subtitle @value{VERSION_PACKAGE}
126 @subtitle Version @value{VERSION}
129 The Free Software Foundation Inc.@: thanks The Nice Computer
130 Company of Australia for loaning Dean Elsner to write the
131 first (Vax) version of @command{as} for Project @sc{gnu}.
132 The proprietors, management and staff of TNCCA thank FSF for
133 distracting the boss while they got some work
136 @author Dean Elsner, Jay Fenlason & friends
140 \hfill {\it Using {\tt @value{AS}}}\par
141 \hfill Edited by Cygnus Support\par
143 %"boxit" macro for figures:
144 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
145 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
146 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
147 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
148 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
151 @vskip 0pt plus 1filll
152 Copyright @copyright{} 1991-2014 Free Software Foundation, Inc.
154 Permission is granted to copy, distribute and/or modify this document
155 under the terms of the GNU Free Documentation License, Version 1.3
156 or any later version published by the Free Software Foundation;
157 with no Invariant Sections, with no Front-Cover Texts, and with no
158 Back-Cover Texts. A copy of the license is included in the
159 section entitled ``GNU Free Documentation License''.
166 @top Using @value{AS}
168 This file is a user guide to the @sc{gnu} assembler @command{@value{AS}}
169 @ifset VERSION_PACKAGE
170 @value{VERSION_PACKAGE}
172 version @value{VERSION}.
174 This version of the file describes @command{@value{AS}} configured to generate
175 code for @value{TARGET} architectures.
178 This document is distributed under the terms of the GNU Free
179 Documentation License. A copy of the license is included in the
180 section entitled ``GNU Free Documentation License''.
183 * Overview:: Overview
184 * Invoking:: Command-Line Options
186 * Sections:: Sections and Relocation
188 * Expressions:: Expressions
189 * Pseudo Ops:: Assembler Directives
191 * Object Attributes:: Object Attributes
193 * Machine Dependencies:: Machine Dependent Features
194 * Reporting Bugs:: Reporting Bugs
195 * Acknowledgements:: Who Did What
196 * GNU Free Documentation License:: GNU Free Documentation License
197 * AS Index:: AS Index
204 This manual is a user guide to the @sc{gnu} assembler @command{@value{AS}}.
206 This version of the manual describes @command{@value{AS}} configured to generate
207 code for @value{TARGET} architectures.
211 @cindex invocation summary
212 @cindex option summary
213 @cindex summary of options
214 Here is a brief summary of how to invoke @command{@value{AS}}. For details,
215 see @ref{Invoking,,Command-Line Options}.
217 @c man title AS the portable GNU assembler.
221 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
225 @c We don't use deffn and friends for the following because they seem
226 @c to be limited to one line for the header.
228 @c man begin SYNOPSIS
229 @value{AS} [@b{-a}[@b{cdghlns}][=@var{file}]] [@b{--alternate}] [@b{-D}]
230 [@b{--compress-debug-sections}] [@b{--nocompress-debug-sections}]
231 [@b{--debug-prefix-map} @var{old}=@var{new}]
232 [@b{--defsym} @var{sym}=@var{val}] [@b{-f}] [@b{-g}] [@b{--gstabs}]
233 [@b{--gstabs+}] [@b{--gdwarf-2}] [@b{--gdwarf-sections}]
234 [@b{--help}] [@b{-I} @var{dir}] [@b{-J}]
235 [@b{-K}] [@b{-L}] [@b{--listing-lhs-width}=@var{NUM}]
236 [@b{--listing-lhs-width2}=@var{NUM}] [@b{--listing-rhs-width}=@var{NUM}]
237 [@b{--listing-cont-lines}=@var{NUM}] [@b{--keep-locals}] [@b{-o}
238 @var{objfile}] [@b{-R}] [@b{--reduce-memory-overheads}] [@b{--statistics}]
239 [@b{-v}] [@b{-version}] [@b{--version}] [@b{-W}] [@b{--warn}]
240 [@b{--fatal-warnings}] [@b{-w}] [@b{-x}] [@b{-Z}] [@b{@@@var{FILE}}]
241 [@b{--size-check=[error|warning]}]
242 [@b{--target-help}] [@var{target-options}]
243 [@b{--}|@var{files} @dots{}]
245 @c Target dependent options are listed below. Keep the list sorted.
246 @c Add an empty line for separation.
249 @emph{Target AArch64 options:}
251 [@b{-mabi}=@var{ABI}]
255 @emph{Target Alpha options:}
257 [@b{-mdebug} | @b{-no-mdebug}]
258 [@b{-replace} | @b{-noreplace}]
259 [@b{-relax}] [@b{-g}] [@b{-G@var{size}}]
260 [@b{-F}] [@b{-32addr}]
264 @emph{Target ARC options:}
270 @emph{Target ARM options:}
271 @c Don't document the deprecated options
272 [@b{-mcpu}=@var{processor}[+@var{extension}@dots{}]]
273 [@b{-march}=@var{architecture}[+@var{extension}@dots{}]]
274 [@b{-mfpu}=@var{floating-point-format}]
275 [@b{-mfloat-abi}=@var{abi}]
276 [@b{-meabi}=@var{ver}]
279 [@b{-mapcs-32}|@b{-mapcs-26}|@b{-mapcs-float}|
280 @b{-mapcs-reentrant}]
281 [@b{-mthumb-interwork}] [@b{-k}]
285 @emph{Target Blackfin options:}
286 [@b{-mcpu}=@var{processor}[-@var{sirevision}]]
293 @emph{Target CRIS options:}
294 [@b{--underscore} | @b{--no-underscore}]
296 [@b{--emulation=criself} | @b{--emulation=crisaout}]
297 [@b{--march=v0_v10} | @b{--march=v10} | @b{--march=v32} | @b{--march=common_v10_v32}]
298 @c Deprecated -- deliberately not documented.
303 @emph{Target D10V options:}
308 @emph{Target D30V options:}
309 [@b{-O}|@b{-n}|@b{-N}]
313 @emph{Target EPIPHANY options:}
314 [@b{-mepiphany}|@b{-mepiphany16}]
318 @emph{Target H8/300 options:}
322 @c HPPA has no machine-dependent assembler options (yet).
326 @emph{Target i386 options:}
327 [@b{--32}|@b{--x32}|@b{--64}] [@b{-n}]
328 [@b{-march}=@var{CPU}[+@var{EXTENSION}@dots{}]] [@b{-mtune}=@var{CPU}]
332 @emph{Target i960 options:}
333 @c see md_parse_option in tc-i960.c
334 [@b{-ACA}|@b{-ACA_A}|@b{-ACB}|@b{-ACC}|@b{-AKA}|@b{-AKB}|
336 [@b{-b}] [@b{-no-relax}]
340 @emph{Target IA-64 options:}
341 [@b{-mconstant-gp}|@b{-mauto-pic}]
342 [@b{-milp32}|@b{-milp64}|@b{-mlp64}|@b{-mp64}]
344 [@b{-mtune=itanium1}|@b{-mtune=itanium2}]
345 [@b{-munwind-check=warning}|@b{-munwind-check=error}]
346 [@b{-mhint.b=ok}|@b{-mhint.b=warning}|@b{-mhint.b=error}]
347 [@b{-x}|@b{-xexplicit}] [@b{-xauto}] [@b{-xdebug}]
351 @emph{Target IP2K options:}
352 [@b{-mip2022}|@b{-mip2022ext}]
356 @emph{Target M32C options:}
357 [@b{-m32c}|@b{-m16c}] [-relax] [-h-tick-hex]
361 @emph{Target M32R options:}
362 [@b{--m32rx}|@b{--[no-]warn-explicit-parallel-conflicts}|
367 @emph{Target M680X0 options:}
368 [@b{-l}] [@b{-m68000}|@b{-m68010}|@b{-m68020}|@dots{}]
372 @emph{Target M68HC11 options:}
373 [@b{-m68hc11}|@b{-m68hc12}|@b{-m68hcs12}|@b{-mm9s12x}|@b{-mm9s12xg}]
374 [@b{-mshort}|@b{-mlong}]
375 [@b{-mshort-double}|@b{-mlong-double}]
376 [@b{--force-long-branches}] [@b{--short-branches}]
377 [@b{--strict-direct-mode}] [@b{--print-insn-syntax}]
378 [@b{--print-opcodes}] [@b{--generate-example}]
382 @emph{Target MCORE options:}
383 [@b{-jsri2bsr}] [@b{-sifilter}] [@b{-relax}]
384 [@b{-mcpu=[210|340]}]
388 @emph{Target Meta options:}
389 [@b{-mcpu=@var{cpu}}] [@b{-mfpu=@var{cpu}}] [@b{-mdsp=@var{cpu}}]
392 @emph{Target MICROBLAZE options:}
393 @c MicroBlaze has no machine-dependent assembler options.
397 @emph{Target MIPS options:}
398 [@b{-nocpp}] [@b{-EL}] [@b{-EB}] [@b{-O}[@var{optimization level}]]
399 [@b{-g}[@var{debug level}]] [@b{-G} @var{num}] [@b{-KPIC}] [@b{-call_shared}]
400 [@b{-non_shared}] [@b{-xgot} [@b{-mvxworks-pic}]
401 [@b{-mabi}=@var{ABI}] [@b{-32}] [@b{-n32}] [@b{-64}] [@b{-mfp32}] [@b{-mgp32}]
402 [@b{-march}=@var{CPU}] [@b{-mtune}=@var{CPU}] [@b{-mips1}] [@b{-mips2}]
403 [@b{-mips3}] [@b{-mips4}] [@b{-mips5}] [@b{-mips32}] [@b{-mips32r2}]
404 [@b{-mips32r3}] [@b{-mips32r5}] [@b{-mips64}] [@b{-mips64r2}]
405 [@b{-mips64r3}] [@b{-mips64r5}]
406 [@b{-construct-floats}] [@b{-no-construct-floats}]
407 [@b{-mnan=@var{encoding}}]
408 [@b{-trap}] [@b{-no-break}] [@b{-break}] [@b{-no-trap}]
409 [@b{-mips16}] [@b{-no-mips16}]
410 [@b{-mmicromips}] [@b{-mno-micromips}]
411 [@b{-msmartmips}] [@b{-mno-smartmips}]
412 [@b{-mips3d}] [@b{-no-mips3d}]
413 [@b{-mdmx}] [@b{-no-mdmx}]
414 [@b{-mdsp}] [@b{-mno-dsp}]
415 [@b{-mdspr2}] [@b{-mno-dspr2}]
416 [@b{-mmsa}] [@b{-mno-msa}]
417 [@b{-mxpa}] [@b{-mno-xpa}]
418 [@b{-mmt}] [@b{-mno-mt}]
419 [@b{-mmcu}] [@b{-mno-mcu}]
420 [@b{-minsn32}] [@b{-mno-insn32}]
421 [@b{-mfix7000}] [@b{-mno-fix7000}]
422 [@b{-mfix-rm7000}] [@b{-mno-fix-rm7000}]
423 [@b{-mfix-vr4120}] [@b{-mno-fix-vr4120}]
424 [@b{-mfix-vr4130}] [@b{-mno-fix-vr4130}]
425 [@b{-mdebug}] [@b{-no-mdebug}]
426 [@b{-mpdr}] [@b{-mno-pdr}]
430 @emph{Target MMIX options:}
431 [@b{--fixed-special-register-names}] [@b{--globalize-symbols}]
432 [@b{--gnu-syntax}] [@b{--relax}] [@b{--no-predefined-symbols}]
433 [@b{--no-expand}] [@b{--no-merge-gregs}] [@b{-x}]
434 [@b{--linker-allocated-gregs}]
438 @emph{Target Nios II options:}
439 [@b{-relax-all}] [@b{-relax-section}] [@b{-no-relax}]
444 @emph{Target NDS32 options:}
445 [@b{-EL}] [@b{-EB}] [@b{-O}] [@b{-Os}] [@b{-mcpu=@var{cpu}}]
446 [@b{-misa=@var{isa}}] [@b{-mabi=@var{abi}}] [@b{-mall-ext}]
447 [@b{-m[no-]16-bit}] [@b{-m[no-]perf-ext}] [@b{-m[no-]perf2-ext}]
448 [@b{-m[no-]string-ext}] [@b{-m[no-]dsp-ext}] [@b{-m[no-]mac}] [@b{-m[no-]div}]
449 [@b{-m[no-]audio-isa-ext}] [@b{-m[no-]fpu-sp-ext}] [@b{-m[no-]fpu-dp-ext}]
450 [@b{-m[no-]fpu-fma}] [@b{-mfpu-freg=@var{FREG}}] [@b{-mreduced-regs}]
451 [@b{-mfull-regs}] [@b{-m[no-]dx-regs}] [@b{-mpic}] [@b{-mno-relax}]
456 @emph{Target PDP11 options:}
457 [@b{-mpic}|@b{-mno-pic}] [@b{-mall}] [@b{-mno-extensions}]
458 [@b{-m}@var{extension}|@b{-mno-}@var{extension}]
459 [@b{-m}@var{cpu}] [@b{-m}@var{machine}]
463 @emph{Target picoJava options:}
468 @emph{Target PowerPC options:}
470 [@b{-mpwrx}|@b{-mpwr2}|@b{-mpwr}|@b{-m601}|@b{-mppc}|@b{-mppc32}|@b{-m603}|@b{-m604}|@b{-m403}|@b{-m405}|
471 @b{-m440}|@b{-m464}|@b{-m476}|@b{-m7400}|@b{-m7410}|@b{-m7450}|@b{-m7455}|@b{-m750cl}|@b{-mppc64}|
472 @b{-m620}|@b{-me500}|@b{-e500x2}|@b{-me500mc}|@b{-me500mc64}|@b{-me5500}|@b{-me6500}|@b{-mppc64bridge}|
473 @b{-mbooke}|@b{-mpower4}|@b{-mpwr4}|@b{-mpower5}|@b{-mpwr5}|@b{-mpwr5x}|@b{-mpower6}|@b{-mpwr6}|
474 @b{-mpower7}|@b{-mpwr7}|@b{-mpower8}|@b{-mpwr8}|@b{-ma2}|@b{-mcell}|@b{-mspe}|@b{-mtitan}|@b{-me300}|@b{-mcom}]
475 [@b{-many}] [@b{-maltivec}|@b{-mvsx}|@b{-mhtm}|@b{-mvle}]
476 [@b{-mregnames}|@b{-mno-regnames}]
477 [@b{-mrelocatable}|@b{-mrelocatable-lib}|@b{-K PIC}] [@b{-memb}]
478 [@b{-mlittle}|@b{-mlittle-endian}|@b{-le}|@b{-mbig}|@b{-mbig-endian}|@b{-be}]
479 [@b{-msolaris}|@b{-mno-solaris}]
480 [@b{-nops=@var{count}}]
484 @emph{Target RX options:}
485 [@b{-mlittle-endian}|@b{-mbig-endian}]
486 [@b{-m32bit-doubles}|@b{-m64bit-doubles}]
487 [@b{-muse-conventional-section-names}]
488 [@b{-msmall-data-limit}]
491 [@b{-mint-register=@var{number}}]
492 [@b{-mgcc-abi}|@b{-mrx-abi}]
496 @emph{Target s390 options:}
497 [@b{-m31}|@b{-m64}] [@b{-mesa}|@b{-mzarch}] [@b{-march}=@var{CPU}]
498 [@b{-mregnames}|@b{-mno-regnames}]
499 [@b{-mwarn-areg-zero}]
503 @emph{Target SCORE options:}
504 [@b{-EB}][@b{-EL}][@b{-FIXDD}][@b{-NWARN}]
505 [@b{-SCORE5}][@b{-SCORE5U}][@b{-SCORE7}][@b{-SCORE3}]
506 [@b{-march=score7}][@b{-march=score3}]
507 [@b{-USE_R1}][@b{-KPIC}][@b{-O0}][@b{-G} @var{num}][@b{-V}]
511 @emph{Target SPARC options:}
512 @c The order here is important. See c-sparc.texi.
513 [@b{-Av6}|@b{-Av7}|@b{-Av8}|@b{-Asparclet}|@b{-Asparclite}
514 @b{-Av8plus}|@b{-Av8plusa}|@b{-Av9}|@b{-Av9a}]
515 [@b{-xarch=v8plus}|@b{-xarch=v8plusa}] [@b{-bump}]
520 @emph{Target TIC54X options:}
521 [@b{-mcpu=54[123589]}|@b{-mcpu=54[56]lp}] [@b{-mfar-mode}|@b{-mf}]
522 [@b{-merrors-to-file} @var{<filename>}|@b{-me} @var{<filename>}]
527 @emph{Target TIC6X options:}
528 [@b{-march=@var{arch}}] [@b{-mbig-endian}|@b{-mlittle-endian}]
529 [@b{-mdsbt}|@b{-mno-dsbt}] [@b{-mpid=no}|@b{-mpid=near}|@b{-mpid=far}]
530 [@b{-mpic}|@b{-mno-pic}]
534 @emph{Target TILE-Gx options:}
535 [@b{-m32}|@b{-m64}][@b{-EB}][@b{-EL}]
538 @c TILEPro has no machine-dependent assembler options
543 @emph{Target Xtensa options:}
544 [@b{--[no-]text-section-literals}] [@b{--[no-]absolute-literals}]
545 [@b{--[no-]target-align}] [@b{--[no-]longcalls}]
546 [@b{--[no-]transform}]
547 [@b{--rename-section} @var{oldname}=@var{newname}]
548 [@b{--[no-]trampolines}]
553 @emph{Target Z80 options:}
554 [@b{-z80}] [@b{-r800}]
555 [@b{ -ignore-undocumented-instructions}] [@b{-Wnud}]
556 [@b{ -ignore-unportable-instructions}] [@b{-Wnup}]
557 [@b{ -warn-undocumented-instructions}] [@b{-Wud}]
558 [@b{ -warn-unportable-instructions}] [@b{-Wup}]
559 [@b{ -forbid-undocumented-instructions}] [@b{-Fud}]
560 [@b{ -forbid-unportable-instructions}] [@b{-Fup}]
564 @c Z8000 has no machine-dependent assembler options
573 @include at-file.texi
576 Turn on listings, in any of a variety of ways:
580 omit false conditionals
583 omit debugging directives
586 include general information, like @value{AS} version and options passed
589 include high-level source
595 include macro expansions
598 omit forms processing
604 set the name of the listing file
607 You may combine these options; for example, use @samp{-aln} for assembly
608 listing without forms processing. The @samp{=file} option, if used, must be
609 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
612 Begin in alternate macro mode.
614 @xref{Altmacro,,@code{.altmacro}}.
617 @item --compress-debug-sections
618 Compress DWARF debug sections using zlib. The debug sections are renamed
619 to begin with @samp{.zdebug}, and the resulting object file may not be
620 compatible with older linkers and object file utilities.
622 @item --nocompress-debug-sections
623 Do not compress DWARF debug sections. This is the default.
626 Ignored. This option is accepted for script compatibility with calls to
629 @item --debug-prefix-map @var{old}=@var{new}
630 When assembling files in directory @file{@var{old}}, record debugging
631 information describing them as in @file{@var{new}} instead.
633 @item --defsym @var{sym}=@var{value}
634 Define the symbol @var{sym} to be @var{value} before assembling the input file.
635 @var{value} must be an integer constant. As in C, a leading @samp{0x}
636 indicates a hexadecimal value, and a leading @samp{0} indicates an octal
637 value. The value of the symbol can be overridden inside a source file via the
638 use of a @code{.set} pseudo-op.
641 ``fast''---skip whitespace and comment preprocessing (assume source is
646 Generate debugging information for each assembler source line using whichever
647 debug format is preferred by the target. This currently means either STABS,
651 Generate stabs debugging information for each assembler line. This
652 may help debugging assembler code, if the debugger can handle it.
655 Generate stabs debugging information for each assembler line, with GNU
656 extensions that probably only gdb can handle, and that could make other
657 debuggers crash or refuse to read your program. This
658 may help debugging assembler code. Currently the only GNU extension is
659 the location of the current working directory at assembling time.
662 Generate DWARF2 debugging information for each assembler line. This
663 may help debugging assembler code, if the debugger can handle it. Note---this
664 option is only supported by some targets, not all of them.
666 @item --gdwarf-sections
667 Instead of creating a .debug_line section, create a series of
668 .debug_line.@var{foo} sections where @var{foo} is the name of the
669 corresponding code section. For example a code section called @var{.text.func}
670 will have its dwarf line number information placed into a section called
671 @var{.debug_line.text.func}. If the code section is just called @var{.text}
672 then debug line section will still be called just @var{.debug_line} without any
675 @item --size-check=error
676 @itemx --size-check=warning
677 Issue an error or warning for invalid ELF .size directive.
680 Print a summary of the command line options and exit.
683 Print a summary of all target specific options and exit.
686 Add directory @var{dir} to the search list for @code{.include} directives.
689 Don't warn about signed overflow.
692 @ifclear DIFF-TBL-KLUGE
693 This option is accepted but has no effect on the @value{TARGET} family.
695 @ifset DIFF-TBL-KLUGE
696 Issue warnings when difference tables altered for long displacements.
701 Keep (in the symbol table) local symbols. These symbols start with
702 system-specific local label prefixes, typically @samp{.L} for ELF systems
703 or @samp{L} for traditional a.out systems.
708 @item --listing-lhs-width=@var{number}
709 Set the maximum width, in words, of the output data column for an assembler
710 listing to @var{number}.
712 @item --listing-lhs-width2=@var{number}
713 Set the maximum width, in words, of the output data column for continuation
714 lines in an assembler listing to @var{number}.
716 @item --listing-rhs-width=@var{number}
717 Set the maximum width of an input source line, as displayed in a listing, to
720 @item --listing-cont-lines=@var{number}
721 Set the maximum number of lines printed in a listing for a single line of input
724 @item -o @var{objfile}
725 Name the object-file output from @command{@value{AS}} @var{objfile}.
728 Fold the data section into the text section.
730 @kindex --hash-size=@var{number}
731 Set the default size of GAS's hash tables to a prime number close to
732 @var{number}. Increasing this value can reduce the length of time it takes the
733 assembler to perform its tasks, at the expense of increasing the assembler's
734 memory requirements. Similarly reducing this value can reduce the memory
735 requirements at the expense of speed.
737 @item --reduce-memory-overheads
738 This option reduces GAS's memory requirements, at the expense of making the
739 assembly processes slower. Currently this switch is a synonym for
740 @samp{--hash-size=4051}, but in the future it may have other effects as well.
743 Print the maximum space (in bytes) and total time (in seconds) used by
746 @item --strip-local-absolute
747 Remove local absolute symbols from the outgoing symbol table.
751 Print the @command{as} version.
754 Print the @command{as} version and exit.
758 Suppress warning messages.
760 @item --fatal-warnings
761 Treat warnings as errors.
764 Don't suppress warning messages or treat them as errors.
773 Generate an object file even after errors.
775 @item -- | @var{files} @dots{}
776 Standard input, or source files to assemble.
784 @xref{AArch64 Options}, for the options available when @value{AS} is configured
785 for the 64-bit mode of the ARM Architecture (AArch64).
790 The following options are available when @value{AS} is configured for the
791 64-bit mode of the ARM Architecture (AArch64).
794 @include c-aarch64.texi
795 @c ended inside the included file
803 @xref{Alpha Options}, for the options available when @value{AS} is configured
804 for an Alpha processor.
809 The following options are available when @value{AS} is configured for an Alpha
813 @include c-alpha.texi
814 @c ended inside the included file
821 The following options are available when @value{AS} is configured for
826 This option selects the core processor variant.
828 Select either big-endian (-EB) or little-endian (-EL) output.
833 The following options are available when @value{AS} is configured for the ARM
837 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
838 Specify which ARM processor variant is the target.
839 @item -march=@var{architecture}[+@var{extension}@dots{}]
840 Specify which ARM architecture variant is used by the target.
841 @item -mfpu=@var{floating-point-format}
842 Select which Floating Point architecture is the target.
843 @item -mfloat-abi=@var{abi}
844 Select which floating point ABI is in use.
846 Enable Thumb only instruction decoding.
847 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant
848 Select which procedure calling convention is in use.
850 Select either big-endian (-EB) or little-endian (-EL) output.
851 @item -mthumb-interwork
852 Specify that the code has been generated with interworking between Thumb and
855 Turns on CodeComposer Studio assembly syntax compatibility mode.
857 Specify that PIC code has been generated.
865 @xref{Blackfin Options}, for the options available when @value{AS} is
866 configured for the Blackfin processor family.
871 The following options are available when @value{AS} is configured for
872 the Blackfin processor family.
876 @c ended inside the included file
883 See the info pages for documentation of the CRIS-specific options.
887 The following options are available when @value{AS} is configured for
890 @cindex D10V optimization
891 @cindex optimization, D10V
893 Optimize output by parallelizing instructions.
898 The following options are available when @value{AS} is configured for a D30V
901 @cindex D30V optimization
902 @cindex optimization, D30V
904 Optimize output by parallelizing instructions.
908 Warn when nops are generated.
910 @cindex D30V nops after 32-bit multiply
912 Warn when a nop after a 32-bit multiply instruction is generated.
918 The following options are available when @value{AS} is configured for the
919 Adapteva EPIPHANY series.
922 @xref{Epiphany Options}, for the options available when @value{AS} is
923 configured for an Epiphany processor.
928 The following options are available when @value{AS} is configured for
929 an Epiphany processor.
932 @include c-epiphany.texi
933 @c ended inside the included file
941 @xref{H8/300 Options}, for the options available when @value{AS} is configured
942 for an H8/300 processor.
947 The following options are available when @value{AS} is configured for an H8/300
951 @include c-h8300.texi
952 @c ended inside the included file
960 @xref{i386-Options}, for the options available when @value{AS} is
961 configured for an i386 processor.
966 The following options are available when @value{AS} is configured for
971 @c ended inside the included file
978 The following options are available when @value{AS} is configured for the
979 Intel 80960 processor.
982 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
983 Specify which variant of the 960 architecture is the target.
986 Add code to collect statistics about branches taken.
989 Do not alter compare-and-branch instructions for long displacements;
996 The following options are available when @value{AS} is configured for the
1002 Specifies that the extended IP2022 instructions are allowed.
1005 Restores the default behaviour, which restricts the permitted instructions to
1006 just the basic IP2022 ones.
1012 The following options are available when @value{AS} is configured for the
1013 Renesas M32C and M16C processors.
1018 Assemble M32C instructions.
1021 Assemble M16C instructions (the default).
1024 Enable support for link-time relaxations.
1027 Support H'00 style hex constants in addition to 0x00 style.
1033 The following options are available when @value{AS} is configured for the
1034 Renesas M32R (formerly Mitsubishi M32R) series.
1039 Specify which processor in the M32R family is the target. The default
1040 is normally the M32R, but this option changes it to the M32RX.
1042 @item --warn-explicit-parallel-conflicts or --Wp
1043 Produce warning messages when questionable parallel constructs are
1046 @item --no-warn-explicit-parallel-conflicts or --Wnp
1047 Do not produce warning messages when questionable parallel constructs are
1054 The following options are available when @value{AS} is configured for the
1055 Motorola 68000 series.
1060 Shorten references to undefined symbols, to one word instead of two.
1062 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
1063 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
1064 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
1065 Specify what processor in the 68000 family is the target. The default
1066 is normally the 68020, but this can be changed at configuration time.
1068 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
1069 The target machine does (or does not) have a floating-point coprocessor.
1070 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
1071 the basic 68000 is not compatible with the 68881, a combination of the
1072 two can be specified, since it's possible to do emulation of the
1073 coprocessor instructions with the main processor.
1075 @item -m68851 | -mno-68851
1076 The target machine does (or does not) have a memory-management
1077 unit coprocessor. The default is to assume an MMU for 68020 and up.
1085 @xref{Nios II Options}, for the options available when @value{AS} is configured
1086 for an Altera Nios II processor.
1090 @c man begin OPTIONS
1091 The following options are available when @value{AS} is configured for an
1092 Altera Nios II processor.
1094 @c man begin INCLUDE
1095 @include c-nios2.texi
1096 @c ended inside the included file
1102 For details about the PDP-11 machine dependent features options,
1103 see @ref{PDP-11-Options}.
1106 @item -mpic | -mno-pic
1107 Generate position-independent (or position-dependent) code. The
1108 default is @option{-mpic}.
1111 @itemx -mall-extensions
1112 Enable all instruction set extensions. This is the default.
1114 @item -mno-extensions
1115 Disable all instruction set extensions.
1117 @item -m@var{extension} | -mno-@var{extension}
1118 Enable (or disable) a particular instruction set extension.
1121 Enable the instruction set extensions supported by a particular CPU, and
1122 disable all other extensions.
1124 @item -m@var{machine}
1125 Enable the instruction set extensions supported by a particular machine
1126 model, and disable all other extensions.
1132 The following options are available when @value{AS} is configured for
1133 a picoJava processor.
1137 @cindex PJ endianness
1138 @cindex endianness, PJ
1139 @cindex big endian output, PJ
1141 Generate ``big endian'' format output.
1143 @cindex little endian output, PJ
1145 Generate ``little endian'' format output.
1151 The following options are available when @value{AS} is configured for the
1152 Motorola 68HC11 or 68HC12 series.
1156 @item -m68hc11 | -m68hc12 | -m68hcs12 | -mm9s12x | -mm9s12xg
1157 Specify what processor is the target. The default is
1158 defined by the configuration option when building the assembler.
1160 @item --xgate-ramoffset
1161 Instruct the linker to offset RAM addresses from S12X address space into
1162 XGATE address space.
1165 Specify to use the 16-bit integer ABI.
1168 Specify to use the 32-bit integer ABI.
1170 @item -mshort-double
1171 Specify to use the 32-bit double ABI.
1174 Specify to use the 64-bit double ABI.
1176 @item --force-long-branches
1177 Relative branches are turned into absolute ones. This concerns
1178 conditional branches, unconditional branches and branches to a
1181 @item -S | --short-branches
1182 Do not turn relative branches into absolute ones
1183 when the offset is out of range.
1185 @item --strict-direct-mode
1186 Do not turn the direct addressing mode into extended addressing mode
1187 when the instruction does not support direct addressing mode.
1189 @item --print-insn-syntax
1190 Print the syntax of instruction in case of error.
1192 @item --print-opcodes
1193 Print the list of instructions with syntax and then exit.
1195 @item --generate-example
1196 Print an example of instruction for each possible instruction and then exit.
1197 This option is only useful for testing @command{@value{AS}}.
1203 The following options are available when @command{@value{AS}} is configured
1204 for the SPARC architecture:
1207 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
1208 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
1209 Explicitly select a variant of the SPARC architecture.
1211 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
1212 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
1214 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
1215 UltraSPARC extensions.
1217 @item -xarch=v8plus | -xarch=v8plusa
1218 For compatibility with the Solaris v9 assembler. These options are
1219 equivalent to -Av8plus and -Av8plusa, respectively.
1222 Warn when the assembler switches to another architecture.
1227 The following options are available when @value{AS} is configured for the 'c54x
1232 Enable extended addressing mode. All addresses and relocations will assume
1233 extended addressing (usually 23 bits).
1234 @item -mcpu=@var{CPU_VERSION}
1235 Sets the CPU version being compiled for.
1236 @item -merrors-to-file @var{FILENAME}
1237 Redirect error output to a file, for broken systems which don't support such
1238 behaviour in the shell.
1243 The following options are available when @value{AS} is configured for
1248 This option sets the largest size of an object that can be referenced
1249 implicitly with the @code{gp} register. It is only accepted for targets that
1250 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
1252 @cindex MIPS endianness
1253 @cindex endianness, MIPS
1254 @cindex big endian output, MIPS
1256 Generate ``big endian'' format output.
1258 @cindex little endian output, MIPS
1260 Generate ``little endian'' format output.
1276 Generate code for a particular MIPS Instruction Set Architecture level.
1277 @samp{-mips1} is an alias for @samp{-march=r3000}, @samp{-mips2} is an
1278 alias for @samp{-march=r6000}, @samp{-mips3} is an alias for
1279 @samp{-march=r4000} and @samp{-mips4} is an alias for @samp{-march=r8000}.
1280 @samp{-mips5}, @samp{-mips32}, @samp{-mips32r2}, @samp{-mips32r3},
1281 @samp{-mips32r5}, @samp{-mips64}, @samp{-mips64r2}, @samp{-mips64r3}, and
1282 @samp{-mips64r5} correspond to generic MIPS V, MIPS32, MIPS32 Release 2,
1283 MIPS32 Release 3, MIPS32 Release 5, MIPS64, MIPS64 Release 2,
1284 MIPS64 Release 3, and MIPS64 Release 5 ISA processors, respectively.
1286 @item -march=@var{cpu}
1287 Generate code for a particular MIPS CPU.
1289 @item -mtune=@var{cpu}
1290 Schedule and tune for a particular MIPS CPU.
1294 Cause nops to be inserted if the read of the destination register
1295 of an mfhi or mflo instruction occurs in the following two instructions.
1298 @itemx -mno-fix-rm7000
1299 Cause nops to be inserted if a dmult or dmultu instruction is
1300 followed by a load instruction.
1304 Cause stabs-style debugging output to go into an ECOFF-style .mdebug
1305 section instead of the standard ELF .stabs sections.
1309 Control generation of @code{.pdr} sections.
1313 The register sizes are normally inferred from the ISA and ABI, but these
1314 flags force a certain group of registers to be treated as 32 bits wide at
1315 all times. @samp{-mgp32} controls the size of general-purpose registers
1316 and @samp{-mfp32} controls the size of floating-point registers.
1320 Generate code for the MIPS 16 processor. This is equivalent to putting
1321 @code{.set mips16} at the start of the assembly file. @samp{-no-mips16}
1322 turns off this option.
1325 @itemx -mno-micromips
1326 Generate code for the microMIPS processor. This is equivalent to putting
1327 @code{.set micromips} at the start of the assembly file. @samp{-mno-micromips}
1328 turns off this option. This is equivalent to putting @code{.set nomicromips}
1329 at the start of the assembly file.
1332 @itemx -mno-smartmips
1333 Enables the SmartMIPS extension to the MIPS32 instruction set. This is
1334 equivalent to putting @code{.set smartmips} at the start of the assembly file.
1335 @samp{-mno-smartmips} turns off this option.
1339 Generate code for the MIPS-3D Application Specific Extension.
1340 This tells the assembler to accept MIPS-3D instructions.
1341 @samp{-no-mips3d} turns off this option.
1345 Generate code for the MDMX Application Specific Extension.
1346 This tells the assembler to accept MDMX instructions.
1347 @samp{-no-mdmx} turns off this option.
1351 Generate code for the DSP Release 1 Application Specific Extension.
1352 This tells the assembler to accept DSP Release 1 instructions.
1353 @samp{-mno-dsp} turns off this option.
1357 Generate code for the DSP Release 2 Application Specific Extension.
1358 This option implies -mdsp.
1359 This tells the assembler to accept DSP Release 2 instructions.
1360 @samp{-mno-dspr2} turns off this option.
1364 Generate code for the MIPS SIMD Architecture Extension.
1365 This tells the assembler to accept MSA instructions.
1366 @samp{-mno-msa} turns off this option.
1370 Generate code for the MIPS eXtended Physical Address (XPA) Extension.
1371 This tells the assembler to accept XPA instructions.
1372 @samp{-mno-xpa} turns off this option.
1376 Generate code for the MT Application Specific Extension.
1377 This tells the assembler to accept MT instructions.
1378 @samp{-mno-mt} turns off this option.
1382 Generate code for the MCU Application Specific Extension.
1383 This tells the assembler to accept MCU instructions.
1384 @samp{-mno-mcu} turns off this option.
1388 Only use 32-bit instruction encodings when generating code for the
1389 microMIPS processor. This option inhibits the use of any 16-bit
1390 instructions. This is equivalent to putting @code{.set insn32} at
1391 the start of the assembly file. @samp{-mno-insn32} turns off this
1392 option. This is equivalent to putting @code{.set noinsn32} at the
1393 start of the assembly file. By default @samp{-mno-insn32} is
1394 selected, allowing all instructions to be used.
1396 @item --construct-floats
1397 @itemx --no-construct-floats
1398 The @samp{--no-construct-floats} option disables the construction of
1399 double width floating point constants by loading the two halves of the
1400 value into the two single width floating point registers that make up
1401 the double width register. By default @samp{--construct-floats} is
1402 selected, allowing construction of these floating point constants.
1404 @item --relax-branch
1405 @itemx --no-relax-branch
1406 The @samp{--relax-branch} option enables the relaxation of out-of-range
1407 branches. By default @samp{--no-relax-branch} is selected, causing any
1408 out-of-range branches to produce an error.
1410 @item -mnan=@var{encoding}
1411 Select between the IEEE 754-2008 (@option{-mnan=2008}) or the legacy
1412 (@option{-mnan=legacy}) NaN encoding format. The latter is the default.
1415 @item --emulation=@var{name}
1416 This option was formerly used to switch between ELF and ECOFF output
1417 on targets like IRIX 5 that supported both. MIPS ECOFF support was
1418 removed in GAS 2.24, so the option now serves little purpose.
1419 It is retained for backwards compatibility.
1421 The available configuration names are: @samp{mipself}, @samp{mipslelf} and
1422 @samp{mipsbelf}. Choosing @samp{mipself} now has no effect, since the output
1423 is always ELF. @samp{mipslelf} and @samp{mipsbelf} select little- and
1424 big-endian output respectively, but @samp{-EL} and @samp{-EB} are now the
1425 preferred options instead.
1428 @command{@value{AS}} ignores this option. It is accepted for compatibility with
1435 Control how to deal with multiplication overflow and division by zero.
1436 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
1437 (and only work for Instruction Set Architecture level 2 and higher);
1438 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
1442 When this option is used, @command{@value{AS}} will issue a warning every
1443 time it generates a nop instruction from a macro.
1448 The following options are available when @value{AS} is configured for
1454 Enable or disable the JSRI to BSR transformation. By default this is enabled.
1455 The command line option @samp{-nojsri2bsr} can be used to disable it.
1459 Enable or disable the silicon filter behaviour. By default this is disabled.
1460 The default can be overridden by the @samp{-sifilter} command line option.
1463 Alter jump instructions for long displacements.
1465 @item -mcpu=[210|340]
1466 Select the cpu type on the target hardware. This controls which instructions
1470 Assemble for a big endian target.
1473 Assemble for a little endian target.
1482 @xref{Meta Options}, for the options available when @value{AS} is configured
1483 for a Meta processor.
1487 @c man begin OPTIONS
1488 The following options are available when @value{AS} is configured for a
1491 @c man begin INCLUDE
1492 @include c-metag.texi
1493 @c ended inside the included file
1498 @c man begin OPTIONS
1500 See the info pages for documentation of the MMIX-specific options.
1506 @xref{NDS32 Options}, for the options available when @value{AS} is configured
1507 for a NDS32 processor.
1509 @c ended inside the included file
1513 @c man begin OPTIONS
1514 The following options are available when @value{AS} is configured for a
1517 @c man begin INCLUDE
1518 @include c-nds32.texi
1519 @c ended inside the included file
1526 @xref{PowerPC-Opts}, for the options available when @value{AS} is configured
1527 for a PowerPC processor.
1531 @c man begin OPTIONS
1532 The following options are available when @value{AS} is configured for a
1535 @c man begin INCLUDE
1537 @c ended inside the included file
1542 @c man begin OPTIONS
1544 See the info pages for documentation of the RX-specific options.
1548 The following options are available when @value{AS} is configured for the s390
1554 Select the word size, either 31/32 bits or 64 bits.
1557 Select the architecture mode, either the Enterprise System
1558 Architecture (esa) or the z/Architecture mode (zarch).
1559 @item -march=@var{processor}
1560 Specify which s390 processor variant is the target, @samp{g6}, @samp{g6},
1561 @samp{z900}, @samp{z990}, @samp{z9-109}, @samp{z9-ec}, @samp{z10},
1562 @samp{z196}, or @samp{zEC12}.
1564 @itemx -mno-regnames
1565 Allow or disallow symbolic names for registers.
1566 @item -mwarn-areg-zero
1567 Warn whenever the operand for a base or index register has been specified
1568 but evaluates to zero.
1576 @xref{TIC6X Options}, for the options available when @value{AS} is configured
1577 for a TMS320C6000 processor.
1581 @c man begin OPTIONS
1582 The following options are available when @value{AS} is configured for a
1583 TMS320C6000 processor.
1585 @c man begin INCLUDE
1586 @include c-tic6x.texi
1587 @c ended inside the included file
1595 @xref{TILE-Gx Options}, for the options available when @value{AS} is configured
1596 for a TILE-Gx processor.
1600 @c man begin OPTIONS
1601 The following options are available when @value{AS} is configured for a TILE-Gx
1604 @c man begin INCLUDE
1605 @include c-tilegx.texi
1606 @c ended inside the included file
1614 @xref{Xtensa Options}, for the options available when @value{AS} is configured
1615 for an Xtensa processor.
1619 @c man begin OPTIONS
1620 The following options are available when @value{AS} is configured for an
1623 @c man begin INCLUDE
1624 @include c-xtensa.texi
1625 @c ended inside the included file
1630 @c man begin OPTIONS
1633 The following options are available when @value{AS} is configured for
1634 a Z80 family processor.
1637 Assemble for Z80 processor.
1639 Assemble for R800 processor.
1640 @item -ignore-undocumented-instructions
1642 Assemble undocumented Z80 instructions that also work on R800 without warning.
1643 @item -ignore-unportable-instructions
1645 Assemble all undocumented Z80 instructions without warning.
1646 @item -warn-undocumented-instructions
1648 Issue a warning for undocumented Z80 instructions that also work on R800.
1649 @item -warn-unportable-instructions
1651 Issue a warning for undocumented Z80 instructions that do not work on R800.
1652 @item -forbid-undocumented-instructions
1654 Treat all undocumented instructions as errors.
1655 @item -forbid-unportable-instructions
1657 Treat undocumented Z80 instructions that do not work on R800 as errors.
1664 * Manual:: Structure of this Manual
1665 * GNU Assembler:: The GNU Assembler
1666 * Object Formats:: Object File Formats
1667 * Command Line:: Command Line
1668 * Input Files:: Input Files
1669 * Object:: Output (Object) File
1670 * Errors:: Error and Warning Messages
1674 @section Structure of this Manual
1676 @cindex manual, structure and purpose
1677 This manual is intended to describe what you need to know to use
1678 @sc{gnu} @command{@value{AS}}. We cover the syntax expected in source files, including
1679 notation for symbols, constants, and expressions; the directives that
1680 @command{@value{AS}} understands; and of course how to invoke @command{@value{AS}}.
1683 We also cover special features in the @value{TARGET}
1684 configuration of @command{@value{AS}}, including assembler directives.
1687 This manual also describes some of the machine-dependent features of
1688 various flavors of the assembler.
1691 @cindex machine instructions (not covered)
1692 On the other hand, this manual is @emph{not} intended as an introduction
1693 to programming in assembly language---let alone programming in general!
1694 In a similar vein, we make no attempt to introduce the machine
1695 architecture; we do @emph{not} describe the instruction set, standard
1696 mnemonics, registers or addressing modes that are standard to a
1697 particular architecture.
1699 You may want to consult the manufacturer's
1700 machine architecture manual for this information.
1704 For information on the H8/300 machine instruction set, see @cite{H8/300
1705 Series Programming Manual}. For the H8/300H, see @cite{H8/300H Series
1706 Programming Manual} (Renesas).
1709 For information on the Renesas (formerly Hitachi) / SuperH SH machine instruction set,
1710 see @cite{SH-Microcomputer User's Manual} (Renesas) or
1711 @cite{SH-4 32-bit CPU Core Architecture} (SuperH) and
1712 @cite{SuperH (SH) 64-Bit RISC Series} (SuperH).
1715 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
1719 @c I think this is premature---doc@cygnus.com, 17jan1991
1721 Throughout this manual, we assume that you are running @dfn{GNU},
1722 the portable operating system from the @dfn{Free Software
1723 Foundation, Inc.}. This restricts our attention to certain kinds of
1724 computer (in particular, the kinds of computers that @sc{gnu} can run on);
1725 once this assumption is granted examples and definitions need less
1728 @command{@value{AS}} is part of a team of programs that turn a high-level
1729 human-readable series of instructions into a low-level
1730 computer-readable series of instructions. Different versions of
1731 @command{@value{AS}} are used for different kinds of computer.
1734 @c There used to be a section "Terminology" here, which defined
1735 @c "contents", "byte", "word", and "long". Defining "word" to any
1736 @c particular size is confusing when the .word directive may generate 16
1737 @c bits on one machine and 32 bits on another; in general, for the user
1738 @c version of this manual, none of these terms seem essential to define.
1739 @c They were used very little even in the former draft of the manual;
1740 @c this draft makes an effort to avoid them (except in names of
1744 @section The GNU Assembler
1746 @c man begin DESCRIPTION
1748 @sc{gnu} @command{as} is really a family of assemblers.
1750 This manual describes @command{@value{AS}}, a member of that family which is
1751 configured for the @value{TARGET} architectures.
1753 If you use (or have used) the @sc{gnu} assembler on one architecture, you
1754 should find a fairly similar environment when you use it on another
1755 architecture. Each version has much in common with the others,
1756 including object file formats, most assembler directives (often called
1757 @dfn{pseudo-ops}) and assembler syntax.@refill
1759 @cindex purpose of @sc{gnu} assembler
1760 @command{@value{AS}} is primarily intended to assemble the output of the
1761 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
1762 @code{@value{LD}}. Nevertheless, we've tried to make @command{@value{AS}}
1763 assemble correctly everything that other assemblers for the same
1764 machine would assemble.
1766 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
1769 @c This remark should appear in generic version of manual; assumption
1770 @c here is that generic version sets M680x0.
1771 This doesn't mean @command{@value{AS}} always uses the same syntax as another
1772 assembler for the same architecture; for example, we know of several
1773 incompatible versions of 680x0 assembly language syntax.
1778 Unlike older assemblers, @command{@value{AS}} is designed to assemble a source
1779 program in one pass of the source file. This has a subtle impact on the
1780 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
1782 @node Object Formats
1783 @section Object File Formats
1785 @cindex object file format
1786 The @sc{gnu} assembler can be configured to produce several alternative
1787 object file formats. For the most part, this does not affect how you
1788 write assembly language programs; but directives for debugging symbols
1789 are typically different in different file formats. @xref{Symbol
1790 Attributes,,Symbol Attributes}.
1793 For the @value{TARGET} target, @command{@value{AS}} is configured to produce
1794 @value{OBJ-NAME} format object files.
1796 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1798 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1799 @code{b.out} or COFF format object files.
1802 On the @value{TARGET}, @command{@value{AS}} can be configured to produce either
1803 SOM or ELF format object files.
1808 @section Command Line
1810 @cindex command line conventions
1812 After the program name @command{@value{AS}}, the command line may contain
1813 options and file names. Options may appear in any order, and may be
1814 before, after, or between file names. The order of file names is
1817 @cindex standard input, as input file
1819 @file{--} (two hyphens) by itself names the standard input file
1820 explicitly, as one of the files for @command{@value{AS}} to assemble.
1822 @cindex options, command line
1823 Except for @samp{--} any command line argument that begins with a
1824 hyphen (@samp{-}) is an option. Each option changes the behavior of
1825 @command{@value{AS}}. No option changes the way another option works. An
1826 option is a @samp{-} followed by one or more letters; the case of
1827 the letter is important. All options are optional.
1829 Some options expect exactly one file name to follow them. The file
1830 name may either immediately follow the option's letter (compatible
1831 with older assemblers) or it may be the next command argument (@sc{gnu}
1832 standard). These two command lines are equivalent:
1835 @value{AS} -o my-object-file.o mumble.s
1836 @value{AS} -omy-object-file.o mumble.s
1840 @section Input Files
1843 @cindex source program
1844 @cindex files, input
1845 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1846 describe the program input to one run of @command{@value{AS}}. The program may
1847 be in one or more files; how the source is partitioned into files
1848 doesn't change the meaning of the source.
1850 @c I added "con" prefix to "catenation" just to prove I can overcome my
1851 @c APL training... doc@cygnus.com
1852 The source program is a concatenation of the text in all the files, in the
1855 @c man begin DESCRIPTION
1856 Each time you run @command{@value{AS}} it assembles exactly one source
1857 program. The source program is made up of one or more files.
1858 (The standard input is also a file.)
1860 You give @command{@value{AS}} a command line that has zero or more input file
1861 names. The input files are read (from left file name to right). A
1862 command line argument (in any position) that has no special meaning
1863 is taken to be an input file name.
1865 If you give @command{@value{AS}} no file names it attempts to read one input file
1866 from the @command{@value{AS}} standard input, which is normally your terminal. You
1867 may have to type @key{ctl-D} to tell @command{@value{AS}} there is no more program
1870 Use @samp{--} if you need to explicitly name the standard input file
1871 in your command line.
1873 If the source is empty, @command{@value{AS}} produces a small, empty object
1878 @subheading Filenames and Line-numbers
1880 @cindex input file linenumbers
1881 @cindex line numbers, in input files
1882 There are two ways of locating a line in the input file (or files) and
1883 either may be used in reporting error messages. One way refers to a line
1884 number in a physical file; the other refers to a line number in a
1885 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1887 @dfn{Physical files} are those files named in the command line given
1888 to @command{@value{AS}}.
1890 @dfn{Logical files} are simply names declared explicitly by assembler
1891 directives; they bear no relation to physical files. Logical file names help
1892 error messages reflect the original source file, when @command{@value{AS}} source
1893 is itself synthesized from other files. @command{@value{AS}} understands the
1894 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1895 @ref{File,,@code{.file}}.
1898 @section Output (Object) File
1904 Every time you run @command{@value{AS}} it produces an output file, which is
1905 your assembly language program translated into numbers. This file
1906 is the object file. Its default name is
1914 @code{b.out} when @command{@value{AS}} is configured for the Intel 80960.
1916 You can give it another name by using the @option{-o} option. Conventionally,
1917 object file names end with @file{.o}. The default name is used for historical
1918 reasons: older assemblers were capable of assembling self-contained programs
1919 directly into a runnable program. (For some formats, this isn't currently
1920 possible, but it can be done for the @code{a.out} format.)
1924 The object file is meant for input to the linker @code{@value{LD}}. It contains
1925 assembled program code, information to help @code{@value{LD}} integrate
1926 the assembled program into a runnable file, and (optionally) symbolic
1927 information for the debugger.
1929 @c link above to some info file(s) like the description of a.out.
1930 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1933 @section Error and Warning Messages
1935 @c man begin DESCRIPTION
1937 @cindex error messages
1938 @cindex warning messages
1939 @cindex messages from assembler
1940 @command{@value{AS}} may write warnings and error messages to the standard error
1941 file (usually your terminal). This should not happen when a compiler
1942 runs @command{@value{AS}} automatically. Warnings report an assumption made so
1943 that @command{@value{AS}} could keep assembling a flawed program; errors report a
1944 grave problem that stops the assembly.
1948 @cindex format of warning messages
1949 Warning messages have the format
1952 file_name:@b{NNN}:Warning Message Text
1956 @cindex line numbers, in warnings/errors
1957 (where @b{NNN} is a line number). If a logical file name has been given
1958 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1959 the current input file is used. If a logical line number was given
1961 (@pxref{Line,,@code{.line}})
1963 then it is used to calculate the number printed,
1964 otherwise the actual line in the current source file is printed. The
1965 message text is intended to be self explanatory (in the grand Unix
1968 @cindex format of error messages
1969 Error messages have the format
1971 file_name:@b{NNN}:FATAL:Error Message Text
1973 The file name and line number are derived as for warning
1974 messages. The actual message text may be rather less explanatory
1975 because many of them aren't supposed to happen.
1978 @chapter Command-Line Options
1980 @cindex options, all versions of assembler
1981 This chapter describes command-line options available in @emph{all}
1982 versions of the @sc{gnu} assembler; see @ref{Machine Dependencies},
1983 for options specific
1985 to the @value{TARGET} target.
1988 to particular machine architectures.
1991 @c man begin DESCRIPTION
1993 If you are invoking @command{@value{AS}} via the @sc{gnu} C compiler,
1994 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1995 The assembler arguments must be separated from each other (and the @samp{-Wa})
1996 by commas. For example:
1999 gcc -c -g -O -Wa,-alh,-L file.c
2003 This passes two options to the assembler: @samp{-alh} (emit a listing to
2004 standard output with high-level and assembly source) and @samp{-L} (retain
2005 local symbols in the symbol table).
2007 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
2008 command-line options are automatically passed to the assembler by the compiler.
2009 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
2010 precisely what options it passes to each compilation pass, including the
2016 * a:: -a[cdghlns] enable listings
2017 * alternate:: --alternate enable alternate macro syntax
2018 * D:: -D for compatibility
2019 * f:: -f to work faster
2020 * I:: -I for .include search path
2021 @ifclear DIFF-TBL-KLUGE
2022 * K:: -K for compatibility
2024 @ifset DIFF-TBL-KLUGE
2025 * K:: -K for difference tables
2028 * L:: -L to retain local symbols
2029 * listing:: --listing-XXX to configure listing output
2030 * M:: -M or --mri to assemble in MRI compatibility mode
2031 * MD:: --MD for dependency tracking
2032 * o:: -o to name the object file
2033 * R:: -R to join data and text sections
2034 * statistics:: --statistics to see statistics about assembly
2035 * traditional-format:: --traditional-format for compatible output
2036 * v:: -v to announce version
2037 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
2038 * Z:: -Z to make object file even after errors
2042 @section Enable Listings: @option{-a[cdghlns]}
2052 @cindex listings, enabling
2053 @cindex assembly listings, enabling
2055 These options enable listing output from the assembler. By itself,
2056 @samp{-a} requests high-level, assembly, and symbols listing.
2057 You can use other letters to select specific options for the list:
2058 @samp{-ah} requests a high-level language listing,
2059 @samp{-al} requests an output-program assembly listing, and
2060 @samp{-as} requests a symbol table listing.
2061 High-level listings require that a compiler debugging option like
2062 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
2065 Use the @samp{-ag} option to print a first section with general assembly
2066 information, like @value{AS} version, switches passed, or time stamp.
2068 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
2069 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
2070 other conditional), or a true @code{.if} followed by an @code{.else}, will be
2071 omitted from the listing.
2073 Use the @samp{-ad} option to omit debugging directives from the
2076 Once you have specified one of these options, you can further control
2077 listing output and its appearance using the directives @code{.list},
2078 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
2080 The @samp{-an} option turns off all forms processing.
2081 If you do not request listing output with one of the @samp{-a} options, the
2082 listing-control directives have no effect.
2084 The letters after @samp{-a} may be combined into one option,
2085 @emph{e.g.}, @samp{-aln}.
2087 Note if the assembler source is coming from the standard input (e.g.,
2089 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
2090 is being used) then the listing will not contain any comments or preprocessor
2091 directives. This is because the listing code buffers input source lines from
2092 stdin only after they have been preprocessed by the assembler. This reduces
2093 memory usage and makes the code more efficient.
2096 @section @option{--alternate}
2099 Begin in alternate macro mode, see @ref{Altmacro,,@code{.altmacro}}.
2102 @section @option{-D}
2105 This option has no effect whatsoever, but it is accepted to make it more
2106 likely that scripts written for other assemblers also work with
2107 @command{@value{AS}}.
2110 @section Work Faster: @option{-f}
2113 @cindex trusted compiler
2114 @cindex faster processing (@option{-f})
2115 @samp{-f} should only be used when assembling programs written by a
2116 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
2117 and comment preprocessing on
2118 the input file(s) before assembling them. @xref{Preprocessing,
2122 @emph{Warning:} if you use @samp{-f} when the files actually need to be
2123 preprocessed (if they contain comments, for example), @command{@value{AS}} does
2128 @section @code{.include} Search Path: @option{-I} @var{path}
2130 @kindex -I @var{path}
2131 @cindex paths for @code{.include}
2132 @cindex search path for @code{.include}
2133 @cindex @code{include} directive search path
2134 Use this option to add a @var{path} to the list of directories
2135 @command{@value{AS}} searches for files specified in @code{.include}
2136 directives (@pxref{Include,,@code{.include}}). You may use @option{-I} as
2137 many times as necessary to include a variety of paths. The current
2138 working directory is always searched first; after that, @command{@value{AS}}
2139 searches any @samp{-I} directories in the same order as they were
2140 specified (left to right) on the command line.
2143 @section Difference Tables: @option{-K}
2146 @ifclear DIFF-TBL-KLUGE
2147 On the @value{TARGET} family, this option is allowed, but has no effect. It is
2148 permitted for compatibility with the @sc{gnu} assembler on other platforms,
2149 where it can be used to warn when the assembler alters the machine code
2150 generated for @samp{.word} directives in difference tables. The @value{TARGET}
2151 family does not have the addressing limitations that sometimes lead to this
2152 alteration on other platforms.
2155 @ifset DIFF-TBL-KLUGE
2156 @cindex difference tables, warning
2157 @cindex warning for altered difference tables
2158 @command{@value{AS}} sometimes alters the code emitted for directives of the
2159 form @samp{.word @var{sym1}-@var{sym2}}. @xref{Word,,@code{.word}}.
2160 You can use the @samp{-K} option if you want a warning issued when this
2165 @section Include Local Symbols: @option{-L}
2168 @cindex local symbols, retaining in output
2169 Symbols beginning with system-specific local label prefixes, typically
2170 @samp{.L} for ELF systems or @samp{L} for traditional a.out systems, are
2171 called @dfn{local symbols}. @xref{Symbol Names}. Normally you do not see
2172 such symbols when debugging, because they are intended for the use of
2173 programs (like compilers) that compose assembler programs, not for your
2174 notice. Normally both @command{@value{AS}} and @code{@value{LD}} discard
2175 such symbols, so you do not normally debug with them.
2177 This option tells @command{@value{AS}} to retain those local symbols
2178 in the object file. Usually if you do this you also tell the linker
2179 @code{@value{LD}} to preserve those symbols.
2182 @section Configuring listing output: @option{--listing}
2184 The listing feature of the assembler can be enabled via the command line switch
2185 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
2186 hex dump of the corresponding locations in the output object file, and displays
2187 them as a listing file. The format of this listing can be controlled by
2188 directives inside the assembler source (i.e., @code{.list} (@pxref{List}),
2189 @code{.title} (@pxref{Title}), @code{.sbttl} (@pxref{Sbttl}),
2190 @code{.psize} (@pxref{Psize}), and
2191 @code{.eject} (@pxref{Eject}) and also by the following switches:
2194 @item --listing-lhs-width=@samp{number}
2195 @kindex --listing-lhs-width
2196 @cindex Width of first line disassembly output
2197 Sets the maximum width, in words, of the first line of the hex byte dump. This
2198 dump appears on the left hand side of the listing output.
2200 @item --listing-lhs-width2=@samp{number}
2201 @kindex --listing-lhs-width2
2202 @cindex Width of continuation lines of disassembly output
2203 Sets the maximum width, in words, of any further lines of the hex byte dump for
2204 a given input source line. If this value is not specified, it defaults to being
2205 the same as the value specified for @samp{--listing-lhs-width}. If neither
2206 switch is used the default is to one.
2208 @item --listing-rhs-width=@samp{number}
2209 @kindex --listing-rhs-width
2210 @cindex Width of source line output
2211 Sets the maximum width, in characters, of the source line that is displayed
2212 alongside the hex dump. The default value for this parameter is 100. The
2213 source line is displayed on the right hand side of the listing output.
2215 @item --listing-cont-lines=@samp{number}
2216 @kindex --listing-cont-lines
2217 @cindex Maximum number of continuation lines
2218 Sets the maximum number of continuation lines of hex dump that will be
2219 displayed for a given single line of source input. The default value is 4.
2223 @section Assemble in MRI Compatibility Mode: @option{-M}
2226 @cindex MRI compatibility mode
2227 The @option{-M} or @option{--mri} option selects MRI compatibility mode. This
2228 changes the syntax and pseudo-op handling of @command{@value{AS}} to make it
2229 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
2230 configured target) assembler from Microtec Research. The exact nature of the
2231 MRI syntax will not be documented here; see the MRI manuals for more
2232 information. Note in particular that the handling of macros and macro
2233 arguments is somewhat different. The purpose of this option is to permit
2234 assembling existing MRI assembler code using @command{@value{AS}}.
2236 The MRI compatibility is not complete. Certain operations of the MRI assembler
2237 depend upon its object file format, and can not be supported using other object
2238 file formats. Supporting these would require enhancing each object file format
2239 individually. These are:
2242 @item global symbols in common section
2244 The m68k MRI assembler supports common sections which are merged by the linker.
2245 Other object file formats do not support this. @command{@value{AS}} handles
2246 common sections by treating them as a single common symbol. It permits local
2247 symbols to be defined within a common section, but it can not support global
2248 symbols, since it has no way to describe them.
2250 @item complex relocations
2252 The MRI assemblers support relocations against a negated section address, and
2253 relocations which combine the start addresses of two or more sections. These
2254 are not support by other object file formats.
2256 @item @code{END} pseudo-op specifying start address
2258 The MRI @code{END} pseudo-op permits the specification of a start address.
2259 This is not supported by other object file formats. The start address may
2260 instead be specified using the @option{-e} option to the linker, or in a linker
2263 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
2265 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
2266 name to the output file. This is not supported by other object file formats.
2268 @item @code{ORG} pseudo-op
2270 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
2271 address. This differs from the usual @command{@value{AS}} @code{.org} pseudo-op,
2272 which changes the location within the current section. Absolute sections are
2273 not supported by other object file formats. The address of a section may be
2274 assigned within a linker script.
2277 There are some other features of the MRI assembler which are not supported by
2278 @command{@value{AS}}, typically either because they are difficult or because they
2279 seem of little consequence. Some of these may be supported in future releases.
2283 @item EBCDIC strings
2285 EBCDIC strings are not supported.
2287 @item packed binary coded decimal
2289 Packed binary coded decimal is not supported. This means that the @code{DC.P}
2290 and @code{DCB.P} pseudo-ops are not supported.
2292 @item @code{FEQU} pseudo-op
2294 The m68k @code{FEQU} pseudo-op is not supported.
2296 @item @code{NOOBJ} pseudo-op
2298 The m68k @code{NOOBJ} pseudo-op is not supported.
2300 @item @code{OPT} branch control options
2302 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
2303 @code{BRL}, and @code{BRW}---are ignored. @command{@value{AS}} automatically
2304 relaxes all branches, whether forward or backward, to an appropriate size, so
2305 these options serve no purpose.
2307 @item @code{OPT} list control options
2309 The following m68k @code{OPT} list control options are ignored: @code{C},
2310 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
2311 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
2313 @item other @code{OPT} options
2315 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
2316 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
2318 @item @code{OPT} @code{D} option is default
2320 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
2321 @code{OPT NOD} may be used to turn it off.
2323 @item @code{XREF} pseudo-op.
2325 The m68k @code{XREF} pseudo-op is ignored.
2327 @item @code{.debug} pseudo-op
2329 The i960 @code{.debug} pseudo-op is not supported.
2331 @item @code{.extended} pseudo-op
2333 The i960 @code{.extended} pseudo-op is not supported.
2335 @item @code{.list} pseudo-op.
2337 The various options of the i960 @code{.list} pseudo-op are not supported.
2339 @item @code{.optimize} pseudo-op
2341 The i960 @code{.optimize} pseudo-op is not supported.
2343 @item @code{.output} pseudo-op
2345 The i960 @code{.output} pseudo-op is not supported.
2347 @item @code{.setreal} pseudo-op
2349 The i960 @code{.setreal} pseudo-op is not supported.
2354 @section Dependency Tracking: @option{--MD}
2357 @cindex dependency tracking
2360 @command{@value{AS}} can generate a dependency file for the file it creates. This
2361 file consists of a single rule suitable for @code{make} describing the
2362 dependencies of the main source file.
2364 The rule is written to the file named in its argument.
2366 This feature is used in the automatic updating of makefiles.
2369 @section Name the Object File: @option{-o}
2372 @cindex naming object file
2373 @cindex object file name
2374 There is always one object file output when you run @command{@value{AS}}. By
2375 default it has the name
2378 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
2392 You use this option (which takes exactly one filename) to give the
2393 object file a different name.
2395 Whatever the object file is called, @command{@value{AS}} overwrites any
2396 existing file of the same name.
2399 @section Join Data and Text Sections: @option{-R}
2402 @cindex data and text sections, joining
2403 @cindex text and data sections, joining
2404 @cindex joining text and data sections
2405 @cindex merging text and data sections
2406 @option{-R} tells @command{@value{AS}} to write the object file as if all
2407 data-section data lives in the text section. This is only done at
2408 the very last moment: your binary data are the same, but data
2409 section parts are relocated differently. The data section part of
2410 your object file is zero bytes long because all its bytes are
2411 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
2413 When you specify @option{-R} it would be possible to generate shorter
2414 address displacements (because we do not have to cross between text and
2415 data section). We refrain from doing this simply for compatibility with
2416 older versions of @command{@value{AS}}. In future, @option{-R} may work this way.
2419 When @command{@value{AS}} is configured for COFF or ELF output,
2420 this option is only useful if you use sections named @samp{.text} and
2425 @option{-R} is not supported for any of the HPPA targets. Using
2426 @option{-R} generates a warning from @command{@value{AS}}.
2430 @section Display Assembly Statistics: @option{--statistics}
2432 @kindex --statistics
2433 @cindex statistics, about assembly
2434 @cindex time, total for assembly
2435 @cindex space used, maximum for assembly
2436 Use @samp{--statistics} to display two statistics about the resources used by
2437 @command{@value{AS}}: the maximum amount of space allocated during the assembly
2438 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
2441 @node traditional-format
2442 @section Compatible Output: @option{--traditional-format}
2444 @kindex --traditional-format
2445 For some targets, the output of @command{@value{AS}} is different in some ways
2446 from the output of some existing assembler. This switch requests
2447 @command{@value{AS}} to use the traditional format instead.
2449 For example, it disables the exception frame optimizations which
2450 @command{@value{AS}} normally does by default on @code{@value{GCC}} output.
2453 @section Announce Version: @option{-v}
2457 @cindex assembler version
2458 @cindex version of assembler
2459 You can find out what version of as is running by including the
2460 option @samp{-v} (which you can also spell as @samp{-version}) on the
2464 @section Control Warnings: @option{-W}, @option{--warn}, @option{--no-warn}, @option{--fatal-warnings}
2466 @command{@value{AS}} should never give a warning or error message when
2467 assembling compiler output. But programs written by people often
2468 cause @command{@value{AS}} to give a warning that a particular assumption was
2469 made. All such warnings are directed to the standard error file.
2473 @cindex suppressing warnings
2474 @cindex warnings, suppressing
2475 If you use the @option{-W} and @option{--no-warn} options, no warnings are issued.
2476 This only affects the warning messages: it does not change any particular of
2477 how @command{@value{AS}} assembles your file. Errors, which stop the assembly,
2480 @kindex --fatal-warnings
2481 @cindex errors, caused by warnings
2482 @cindex warnings, causing error
2483 If you use the @option{--fatal-warnings} option, @command{@value{AS}} considers
2484 files that generate warnings to be in error.
2487 @cindex warnings, switching on
2488 You can switch these options off again by specifying @option{--warn}, which
2489 causes warnings to be output as usual.
2492 @section Generate Object File in Spite of Errors: @option{-Z}
2493 @cindex object file, after errors
2494 @cindex errors, continuing after
2495 After an error message, @command{@value{AS}} normally produces no output. If for
2496 some reason you are interested in object file output even after
2497 @command{@value{AS}} gives an error message on your program, use the @samp{-Z}
2498 option. If there are any errors, @command{@value{AS}} continues anyways, and
2499 writes an object file after a final warning message of the form @samp{@var{n}
2500 errors, @var{m} warnings, generating bad object file.}
2505 @cindex machine-independent syntax
2506 @cindex syntax, machine-independent
2507 This chapter describes the machine-independent syntax allowed in a
2508 source file. @command{@value{AS}} syntax is similar to what many other
2509 assemblers use; it is inspired by the BSD 4.2
2514 assembler, except that @command{@value{AS}} does not assemble Vax bit-fields.
2518 * Preprocessing:: Preprocessing
2519 * Whitespace:: Whitespace
2520 * Comments:: Comments
2521 * Symbol Intro:: Symbols
2522 * Statements:: Statements
2523 * Constants:: Constants
2527 @section Preprocessing
2529 @cindex preprocessing
2530 The @command{@value{AS}} internal preprocessor:
2532 @cindex whitespace, removed by preprocessor
2534 adjusts and removes extra whitespace. It leaves one space or tab before
2535 the keywords on a line, and turns any other whitespace on the line into
2538 @cindex comments, removed by preprocessor
2540 removes all comments, replacing them with a single space, or an
2541 appropriate number of newlines.
2543 @cindex constants, converted by preprocessor
2545 converts character constants into the appropriate numeric values.
2548 It does not do macro processing, include file handling, or
2549 anything else you may get from your C compiler's preprocessor. You can
2550 do include file processing with the @code{.include} directive
2551 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
2552 to get other ``CPP'' style preprocessing by giving the input file a
2553 @samp{.S} suffix. @xref{Overall Options, ,Options Controlling the Kind of
2554 Output, gcc.info, Using GNU CC}.
2556 Excess whitespace, comments, and character constants
2557 cannot be used in the portions of the input text that are not
2560 @cindex turning preprocessing on and off
2561 @cindex preprocessing, turning on and off
2564 If the first line of an input file is @code{#NO_APP} or if you use the
2565 @samp{-f} option, whitespace and comments are not removed from the input file.
2566 Within an input file, you can ask for whitespace and comment removal in
2567 specific portions of the by putting a line that says @code{#APP} before the
2568 text that may contain whitespace or comments, and putting a line that says
2569 @code{#NO_APP} after this text. This feature is mainly intend to support
2570 @code{asm} statements in compilers whose output is otherwise free of comments
2577 @dfn{Whitespace} is one or more blanks or tabs, in any order.
2578 Whitespace is used to separate symbols, and to make programs neater for
2579 people to read. Unless within character constants
2580 (@pxref{Characters,,Character Constants}), any whitespace means the same
2581 as exactly one space.
2587 There are two ways of rendering comments to @command{@value{AS}}. In both
2588 cases the comment is equivalent to one space.
2590 Anything from @samp{/*} through the next @samp{*/} is a comment.
2591 This means you may not nest these comments.
2595 The only way to include a newline ('\n') in a comment
2596 is to use this sort of comment.
2599 /* This sort of comment does not nest. */
2602 @cindex line comment character
2603 Anything from a @dfn{line comment} character up to the next newline is
2604 considered a comment and is ignored. The line comment character is target
2605 specific, and some targets multiple comment characters. Some targets also have
2606 line comment characters that only work if they are the first character on a
2607 line. Some targets use a sequence of two characters to introduce a line
2608 comment. Some targets can also change their line comment characters depending
2609 upon command line options that have been used. For more details see the
2610 @emph{Syntax} section in the documentation for individual targets.
2612 If the line comment character is the hash sign (@samp{#}) then it still has the
2613 special ability to enable and disable preprocessing (@pxref{Preprocessing}) and
2614 to specify logical line numbers:
2617 @cindex lines starting with @code{#}
2618 @cindex logical line numbers
2619 To be compatible with past assemblers, lines that begin with @samp{#} have a
2620 special interpretation. Following the @samp{#} should be an absolute
2621 expression (@pxref{Expressions}): the logical line number of the @emph{next}
2622 line. Then a string (@pxref{Strings, ,Strings}) is allowed: if present it is a
2623 new logical file name. The rest of the line, if any, should be whitespace.
2625 If the first non-whitespace characters on the line are not numeric,
2626 the line is ignored. (Just like a comment.)
2629 # This is an ordinary comment.
2630 # 42-6 "new_file_name" # New logical file name
2631 # This is logical line # 36.
2633 This feature is deprecated, and may disappear from future versions
2634 of @command{@value{AS}}.
2639 @cindex characters used in symbols
2640 @ifclear SPECIAL-SYMS
2641 A @dfn{symbol} is one or more characters chosen from the set of all
2642 letters (both upper and lower case), digits and the three characters
2648 A @dfn{symbol} is one or more characters chosen from the set of all
2649 letters (both upper and lower case), digits and the three characters
2650 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
2656 On most machines, you can also use @code{$} in symbol names; exceptions
2657 are noted in @ref{Machine Dependencies}.
2659 No symbol may begin with a digit. Case is significant.
2660 There is no length limit: all characters are significant. Multibyte characters
2661 are supported. Symbols are delimited by characters not in that set, or by the
2662 beginning of a file (since the source program must end with a newline, the end
2663 of a file is not a possible symbol delimiter). @xref{Symbols}.
2664 @cindex length of symbols
2669 @cindex statements, structure of
2670 @cindex line separator character
2671 @cindex statement separator character
2673 A @dfn{statement} ends at a newline character (@samp{\n}) or a
2674 @dfn{line separator character}. The line separator character is target
2675 specific and described in the @emph{Syntax} section of each
2676 target's documentation. Not all targets support a line separator character.
2677 The newline or line separator character is considered to be part of the
2678 preceding statement. Newlines and separators within character constants are an
2679 exception: they do not end statements.
2681 @cindex newline, required at file end
2682 @cindex EOF, newline must precede
2683 It is an error to end any statement with end-of-file: the last
2684 character of any input file should be a newline.@refill
2686 An empty statement is allowed, and may include whitespace. It is ignored.
2688 @cindex instructions and directives
2689 @cindex directives and instructions
2690 @c "key symbol" is not used elsewhere in the document; seems pedantic to
2691 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
2693 A statement begins with zero or more labels, optionally followed by a
2694 key symbol which determines what kind of statement it is. The key
2695 symbol determines the syntax of the rest of the statement. If the
2696 symbol begins with a dot @samp{.} then the statement is an assembler
2697 directive: typically valid for any computer. If the symbol begins with
2698 a letter the statement is an assembly language @dfn{instruction}: it
2699 assembles into a machine language instruction.
2701 Different versions of @command{@value{AS}} for different computers
2702 recognize different instructions. In fact, the same symbol may
2703 represent a different instruction in a different computer's assembly
2707 @cindex @code{:} (label)
2708 @cindex label (@code{:})
2709 A label is a symbol immediately followed by a colon (@code{:}).
2710 Whitespace before a label or after a colon is permitted, but you may not
2711 have whitespace between a label's symbol and its colon. @xref{Labels}.
2714 For HPPA targets, labels need not be immediately followed by a colon, but
2715 the definition of a label must begin in column zero. This also implies that
2716 only one label may be defined on each line.
2720 label: .directive followed by something
2721 another_label: # This is an empty statement.
2722 instruction operand_1, operand_2, @dots{}
2729 A constant is a number, written so that its value is known by
2730 inspection, without knowing any context. Like this:
2733 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2734 .ascii "Ring the bell\7" # A string constant.
2735 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2736 .float 0f-314159265358979323846264338327\
2737 95028841971.693993751E-40 # - pi, a flonum.
2742 * Characters:: Character Constants
2743 * Numbers:: Number Constants
2747 @subsection Character Constants
2749 @cindex character constants
2750 @cindex constants, character
2751 There are two kinds of character constants. A @dfn{character} stands
2752 for one character in one byte and its value may be used in
2753 numeric expressions. String constants (properly called string
2754 @emph{literals}) are potentially many bytes and their values may not be
2755 used in arithmetic expressions.
2759 * Chars:: Characters
2763 @subsubsection Strings
2765 @cindex string constants
2766 @cindex constants, string
2767 A @dfn{string} is written between double-quotes. It may contain
2768 double-quotes or null characters. The way to get special characters
2769 into a string is to @dfn{escape} these characters: precede them with
2770 a backslash @samp{\} character. For example @samp{\\} represents
2771 one backslash: the first @code{\} is an escape which tells
2772 @command{@value{AS}} to interpret the second character literally as a backslash
2773 (which prevents @command{@value{AS}} from recognizing the second @code{\} as an
2774 escape character). The complete list of escapes follows.
2776 @cindex escape codes, character
2777 @cindex character escape codes
2780 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2782 @cindex @code{\b} (backspace character)
2783 @cindex backspace (@code{\b})
2785 Mnemonic for backspace; for ASCII this is octal code 010.
2788 @c Mnemonic for EOText; for ASCII this is octal code 004.
2790 @cindex @code{\f} (formfeed character)
2791 @cindex formfeed (@code{\f})
2793 Mnemonic for FormFeed; for ASCII this is octal code 014.
2795 @cindex @code{\n} (newline character)
2796 @cindex newline (@code{\n})
2798 Mnemonic for newline; for ASCII this is octal code 012.
2801 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2803 @cindex @code{\r} (carriage return character)
2804 @cindex carriage return (@code{\r})
2806 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2809 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2810 @c other assemblers.
2812 @cindex @code{\t} (tab)
2813 @cindex tab (@code{\t})
2815 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2818 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2819 @c @item \x @var{digit} @var{digit} @var{digit}
2820 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2822 @cindex @code{\@var{ddd}} (octal character code)
2823 @cindex octal character code (@code{\@var{ddd}})
2824 @item \ @var{digit} @var{digit} @var{digit}
2825 An octal character code. The numeric code is 3 octal digits.
2826 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2827 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2829 @cindex @code{\@var{xd...}} (hex character code)
2830 @cindex hex character code (@code{\@var{xd...}})
2831 @item \@code{x} @var{hex-digits...}
2832 A hex character code. All trailing hex digits are combined. Either upper or
2833 lower case @code{x} works.
2835 @cindex @code{\\} (@samp{\} character)
2836 @cindex backslash (@code{\\})
2838 Represents one @samp{\} character.
2841 @c Represents one @samp{'} (accent acute) character.
2842 @c This is needed in single character literals
2843 @c (@xref{Characters,,Character Constants}.) to represent
2846 @cindex @code{\"} (doublequote character)
2847 @cindex doublequote (@code{\"})
2849 Represents one @samp{"} character. Needed in strings to represent
2850 this character, because an unescaped @samp{"} would end the string.
2852 @item \ @var{anything-else}
2853 Any other character when escaped by @kbd{\} gives a warning, but
2854 assembles as if the @samp{\} was not present. The idea is that if
2855 you used an escape sequence you clearly didn't want the literal
2856 interpretation of the following character. However @command{@value{AS}} has no
2857 other interpretation, so @command{@value{AS}} knows it is giving you the wrong
2858 code and warns you of the fact.
2861 Which characters are escapable, and what those escapes represent,
2862 varies widely among assemblers. The current set is what we think
2863 the BSD 4.2 assembler recognizes, and is a subset of what most C
2864 compilers recognize. If you are in doubt, do not use an escape
2868 @subsubsection Characters
2870 @cindex single character constant
2871 @cindex character, single
2872 @cindex constant, single character
2873 A single character may be written as a single quote immediately
2874 followed by that character. The same escapes apply to characters as
2875 to strings. So if you want to write the character backslash, you
2876 must write @kbd{'\\} where the first @code{\} escapes the second
2877 @code{\}. As you can see, the quote is an acute accent, not a
2878 grave accent. A newline
2880 @ifclear abnormal-separator
2881 (or semicolon @samp{;})
2883 @ifset abnormal-separator
2885 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2890 immediately following an acute accent is taken as a literal character
2891 and does not count as the end of a statement. The value of a character
2892 constant in a numeric expression is the machine's byte-wide code for
2893 that character. @command{@value{AS}} assumes your character code is ASCII:
2894 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2897 @subsection Number Constants
2899 @cindex constants, number
2900 @cindex number constants
2901 @command{@value{AS}} distinguishes three kinds of numbers according to how they
2902 are stored in the target machine. @emph{Integers} are numbers that
2903 would fit into an @code{int} in the C language. @emph{Bignums} are
2904 integers, but they are stored in more than 32 bits. @emph{Flonums}
2905 are floating point numbers, described below.
2908 * Integers:: Integers
2913 * Bit Fields:: Bit Fields
2919 @subsubsection Integers
2921 @cindex constants, integer
2923 @cindex binary integers
2924 @cindex integers, binary
2925 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2926 the binary digits @samp{01}.
2928 @cindex octal integers
2929 @cindex integers, octal
2930 An octal integer is @samp{0} followed by zero or more of the octal
2931 digits (@samp{01234567}).
2933 @cindex decimal integers
2934 @cindex integers, decimal
2935 A decimal integer starts with a non-zero digit followed by zero or
2936 more digits (@samp{0123456789}).
2938 @cindex hexadecimal integers
2939 @cindex integers, hexadecimal
2940 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2941 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2943 Integers have the usual values. To denote a negative integer, use
2944 the prefix operator @samp{-} discussed under expressions
2945 (@pxref{Prefix Ops,,Prefix Operators}).
2948 @subsubsection Bignums
2951 @cindex constants, bignum
2952 A @dfn{bignum} has the same syntax and semantics as an integer
2953 except that the number (or its negative) takes more than 32 bits to
2954 represent in binary. The distinction is made because in some places
2955 integers are permitted while bignums are not.
2958 @subsubsection Flonums
2960 @cindex floating point numbers
2961 @cindex constants, floating point
2963 @cindex precision, floating point
2964 A @dfn{flonum} represents a floating point number. The translation is
2965 indirect: a decimal floating point number from the text is converted by
2966 @command{@value{AS}} to a generic binary floating point number of more than
2967 sufficient precision. This generic floating point number is converted
2968 to a particular computer's floating point format (or formats) by a
2969 portion of @command{@value{AS}} specialized to that computer.
2971 A flonum is written by writing (in order)
2976 (@samp{0} is optional on the HPPA.)
2980 A letter, to tell @command{@value{AS}} the rest of the number is a flonum.
2982 @kbd{e} is recommended. Case is not important.
2984 @c FIXME: verify if flonum syntax really this vague for most cases
2985 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2986 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2989 On the H8/300, Renesas / SuperH SH,
2990 and AMD 29K architectures, the letter must be
2991 one of the letters @samp{DFPRSX} (in upper or lower case).
2993 On the ARC, the letter must be one of the letters @samp{DFRS}
2994 (in upper or lower case).
2996 On the Intel 960 architecture, the letter must be
2997 one of the letters @samp{DFT} (in upper or lower case).
2999 On the HPPA architecture, the letter must be @samp{E} (upper case only).
3003 One of the letters @samp{DFRS} (in upper or lower case).
3006 One of the letters @samp{DFPRSX} (in upper or lower case).
3009 The letter @samp{E} (upper case only).
3012 One of the letters @samp{DFT} (in upper or lower case).
3017 An optional sign: either @samp{+} or @samp{-}.
3020 An optional @dfn{integer part}: zero or more decimal digits.
3023 An optional @dfn{fractional part}: @samp{.} followed by zero
3024 or more decimal digits.
3027 An optional exponent, consisting of:
3031 An @samp{E} or @samp{e}.
3032 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
3033 @c principle this can perfectly well be different on different targets.
3035 Optional sign: either @samp{+} or @samp{-}.
3037 One or more decimal digits.
3042 At least one of the integer part or the fractional part must be
3043 present. The floating point number has the usual base-10 value.
3045 @command{@value{AS}} does all processing using integers. Flonums are computed
3046 independently of any floating point hardware in the computer running
3047 @command{@value{AS}}.
3051 @c Bit fields are written as a general facility but are also controlled
3052 @c by a conditional-compilation flag---which is as of now (21mar91)
3053 @c turned on only by the i960 config of GAS.
3055 @subsubsection Bit Fields
3058 @cindex constants, bit field
3059 You can also define numeric constants as @dfn{bit fields}.
3060 Specify two numbers separated by a colon---
3062 @var{mask}:@var{value}
3065 @command{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
3068 The resulting number is then packed
3070 @c this conditional paren in case bit fields turned on elsewhere than 960
3071 (in host-dependent byte order)
3073 into a field whose width depends on which assembler directive has the
3074 bit-field as its argument. Overflow (a result from the bitwise and
3075 requiring more binary digits to represent) is not an error; instead,
3076 more constants are generated, of the specified width, beginning with the
3077 least significant digits.@refill
3079 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
3080 @code{.short}, and @code{.word} accept bit-field arguments.
3085 @chapter Sections and Relocation
3090 * Secs Background:: Background
3091 * Ld Sections:: Linker Sections
3092 * As Sections:: Assembler Internal Sections
3093 * Sub-Sections:: Sub-Sections
3097 @node Secs Background
3100 Roughly, a section is a range of addresses, with no gaps; all data
3101 ``in'' those addresses is treated the same for some particular purpose.
3102 For example there may be a ``read only'' section.
3104 @cindex linker, and assembler
3105 @cindex assembler, and linker
3106 The linker @code{@value{LD}} reads many object files (partial programs) and
3107 combines their contents to form a runnable program. When @command{@value{AS}}
3108 emits an object file, the partial program is assumed to start at address 0.
3109 @code{@value{LD}} assigns the final addresses for the partial program, so that
3110 different partial programs do not overlap. This is actually an
3111 oversimplification, but it suffices to explain how @command{@value{AS}} uses
3114 @code{@value{LD}} moves blocks of bytes of your program to their run-time
3115 addresses. These blocks slide to their run-time addresses as rigid
3116 units; their length does not change and neither does the order of bytes
3117 within them. Such a rigid unit is called a @emph{section}. Assigning
3118 run-time addresses to sections is called @dfn{relocation}. It includes
3119 the task of adjusting mentions of object-file addresses so they refer to
3120 the proper run-time addresses.
3122 For the H8/300, and for the Renesas / SuperH SH,
3123 @command{@value{AS}} pads sections if needed to
3124 ensure they end on a word (sixteen bit) boundary.
3127 @cindex standard assembler sections
3128 An object file written by @command{@value{AS}} has at least three sections, any
3129 of which may be empty. These are named @dfn{text}, @dfn{data} and
3134 When it generates COFF or ELF output,
3136 @command{@value{AS}} can also generate whatever other named sections you specify
3137 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
3138 If you do not use any directives that place output in the @samp{.text}
3139 or @samp{.data} sections, these sections still exist, but are empty.
3144 When @command{@value{AS}} generates SOM or ELF output for the HPPA,
3146 @command{@value{AS}} can also generate whatever other named sections you
3147 specify using the @samp{.space} and @samp{.subspace} directives. See
3148 @cite{HP9000 Series 800 Assembly Language Reference Manual}
3149 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
3150 assembler directives.
3153 Additionally, @command{@value{AS}} uses different names for the standard
3154 text, data, and bss sections when generating SOM output. Program text
3155 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
3156 BSS into @samp{$BSS$}.
3160 Within the object file, the text section starts at address @code{0}, the
3161 data section follows, and the bss section follows the data section.
3164 When generating either SOM or ELF output files on the HPPA, the text
3165 section starts at address @code{0}, the data section at address
3166 @code{0x4000000}, and the bss section follows the data section.
3169 To let @code{@value{LD}} know which data changes when the sections are
3170 relocated, and how to change that data, @command{@value{AS}} also writes to the
3171 object file details of the relocation needed. To perform relocation
3172 @code{@value{LD}} must know, each time an address in the object
3176 Where in the object file is the beginning of this reference to
3179 How long (in bytes) is this reference?
3181 Which section does the address refer to? What is the numeric value of
3183 (@var{address}) @minus{} (@var{start-address of section})?
3186 Is the reference to an address ``Program-Counter relative''?
3189 @cindex addresses, format of
3190 @cindex section-relative addressing
3191 In fact, every address @command{@value{AS}} ever uses is expressed as
3193 (@var{section}) + (@var{offset into section})
3196 Further, most expressions @command{@value{AS}} computes have this section-relative
3199 (For some object formats, such as SOM for the HPPA, some expressions are
3200 symbol-relative instead.)
3203 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
3204 @var{N} into section @var{secname}.''
3206 Apart from text, data and bss sections you need to know about the
3207 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
3208 addresses in the absolute section remain unchanged. For example, address
3209 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
3210 @code{@value{LD}}. Although the linker never arranges two partial programs'
3211 data sections with overlapping addresses after linking, @emph{by definition}
3212 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
3213 part of a program is always the same address when the program is running as
3214 address @code{@{absolute@ 239@}} in any other part of the program.
3216 The idea of sections is extended to the @dfn{undefined} section. Any
3217 address whose section is unknown at assembly time is by definition
3218 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
3219 Since numbers are always defined, the only way to generate an undefined
3220 address is to mention an undefined symbol. A reference to a named
3221 common block would be such a symbol: its value is unknown at assembly
3222 time so it has section @emph{undefined}.
3224 By analogy the word @emph{section} is used to describe groups of sections in
3225 the linked program. @code{@value{LD}} puts all partial programs' text
3226 sections in contiguous addresses in the linked program. It is
3227 customary to refer to the @emph{text section} of a program, meaning all
3228 the addresses of all partial programs' text sections. Likewise for
3229 data and bss sections.
3231 Some sections are manipulated by @code{@value{LD}}; others are invented for
3232 use of @command{@value{AS}} and have no meaning except during assembly.
3235 @section Linker Sections
3236 @code{@value{LD}} deals with just four kinds of sections, summarized below.
3241 @cindex named sections
3242 @cindex sections, named
3243 @item named sections
3246 @cindex text section
3247 @cindex data section
3251 These sections hold your program. @command{@value{AS}} and @code{@value{LD}} treat them as
3252 separate but equal sections. Anything you can say of one section is
3255 When the program is running, however, it is
3256 customary for the text section to be unalterable. The
3257 text section is often shared among processes: it contains
3258 instructions, constants and the like. The data section of a running
3259 program is usually alterable: for example, C variables would be stored
3260 in the data section.
3265 This section contains zeroed bytes when your program begins running. It
3266 is used to hold uninitialized variables or common storage. The length of
3267 each partial program's bss section is important, but because it starts
3268 out containing zeroed bytes there is no need to store explicit zero
3269 bytes in the object file. The bss section was invented to eliminate
3270 those explicit zeros from object files.
3272 @cindex absolute section
3273 @item absolute section
3274 Address 0 of this section is always ``relocated'' to runtime address 0.
3275 This is useful if you want to refer to an address that @code{@value{LD}} must
3276 not change when relocating. In this sense we speak of absolute
3277 addresses being ``unrelocatable'': they do not change during relocation.
3279 @cindex undefined section
3280 @item undefined section
3281 This ``section'' is a catch-all for address references to objects not in
3282 the preceding sections.
3283 @c FIXME: ref to some other doc on obj-file formats could go here.
3286 @cindex relocation example
3287 An idealized example of three relocatable sections follows.
3289 The example uses the traditional section names @samp{.text} and @samp{.data}.
3291 Memory addresses are on the horizontal axis.
3295 @c END TEXI2ROFF-KILL
3298 partial program # 1: |ttttt|dddd|00|
3305 partial program # 2: |TTT|DDD|000|
3308 +--+---+-----+--+----+---+-----+~~
3309 linked program: | |TTT|ttttt| |dddd|DDD|00000|
3310 +--+---+-----+--+----+---+-----+~~
3312 addresses: 0 @dots{}
3319 \line{\it Partial program \#1: \hfil}
3320 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3321 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
3323 \line{\it Partial program \#2: \hfil}
3324 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
3325 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
3327 \line{\it linked program: \hfil}
3328 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
3329 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
3330 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
3331 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
3333 \line{\it addresses: \hfil}
3337 @c END TEXI2ROFF-KILL
3340 @section Assembler Internal Sections
3342 @cindex internal assembler sections
3343 @cindex sections in messages, internal
3344 These sections are meant only for the internal use of @command{@value{AS}}. They
3345 have no meaning at run-time. You do not really need to know about these
3346 sections for most purposes; but they can be mentioned in @command{@value{AS}}
3347 warning messages, so it might be helpful to have an idea of their
3348 meanings to @command{@value{AS}}. These sections are used to permit the
3349 value of every expression in your assembly language program to be a
3350 section-relative address.
3353 @cindex assembler internal logic error
3354 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
3355 An internal assembler logic error has been found. This means there is a
3356 bug in the assembler.
3358 @cindex expr (internal section)
3360 The assembler stores complex expression internally as combinations of
3361 symbols. When it needs to represent an expression as a symbol, it puts
3362 it in the expr section.
3364 @c FIXME item transfer[t] vector preload
3365 @c FIXME item transfer[t] vector postload
3366 @c FIXME item register
3370 @section Sub-Sections
3372 @cindex numbered subsections
3373 @cindex grouping data
3379 fall into two sections: text and data.
3381 You may have separate groups of
3383 data in named sections
3387 data in named sections
3393 that you want to end up near to each other in the object file, even though they
3394 are not contiguous in the assembler source. @command{@value{AS}} allows you to
3395 use @dfn{subsections} for this purpose. Within each section, there can be
3396 numbered subsections with values from 0 to 8192. Objects assembled into the
3397 same subsection go into the object file together with other objects in the same
3398 subsection. For example, a compiler might want to store constants in the text
3399 section, but might not want to have them interspersed with the program being
3400 assembled. In this case, the compiler could issue a @samp{.text 0} before each
3401 section of code being output, and a @samp{.text 1} before each group of
3402 constants being output.
3404 Subsections are optional. If you do not use subsections, everything
3405 goes in subsection number zero.
3408 Each subsection is zero-padded up to a multiple of four bytes.
3409 (Subsections may be padded a different amount on different flavors
3410 of @command{@value{AS}}.)
3414 On the H8/300 platform, each subsection is zero-padded to a word
3415 boundary (two bytes).
3416 The same is true on the Renesas SH.
3419 @c FIXME section padding (alignment)?
3420 @c Rich Pixley says padding here depends on target obj code format; that
3421 @c doesn't seem particularly useful to say without further elaboration,
3422 @c so for now I say nothing about it. If this is a generic BFD issue,
3423 @c these paragraphs might need to vanish from this manual, and be
3424 @c discussed in BFD chapter of binutils (or some such).
3428 Subsections appear in your object file in numeric order, lowest numbered
3429 to highest. (All this to be compatible with other people's assemblers.)
3430 The object file contains no representation of subsections; @code{@value{LD}} and
3431 other programs that manipulate object files see no trace of them.
3432 They just see all your text subsections as a text section, and all your
3433 data subsections as a data section.
3435 To specify which subsection you want subsequent statements assembled
3436 into, use a numeric argument to specify it, in a @samp{.text
3437 @var{expression}} or a @samp{.data @var{expression}} statement.
3440 When generating COFF output, you
3445 can also use an extra subsection
3446 argument with arbitrary named sections: @samp{.section @var{name},
3451 When generating ELF output, you
3456 can also use the @code{.subsection} directive (@pxref{SubSection})
3457 to specify a subsection: @samp{.subsection @var{expression}}.
3459 @var{Expression} should be an absolute expression
3460 (@pxref{Expressions}). If you just say @samp{.text} then @samp{.text 0}
3461 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
3462 begins in @code{text 0}. For instance:
3464 .text 0 # The default subsection is text 0 anyway.
3465 .ascii "This lives in the first text subsection. *"
3467 .ascii "But this lives in the second text subsection."
3469 .ascii "This lives in the data section,"
3470 .ascii "in the first data subsection."
3472 .ascii "This lives in the first text section,"
3473 .ascii "immediately following the asterisk (*)."
3476 Each section has a @dfn{location counter} incremented by one for every byte
3477 assembled into that section. Because subsections are merely a convenience
3478 restricted to @command{@value{AS}} there is no concept of a subsection location
3479 counter. There is no way to directly manipulate a location counter---but the
3480 @code{.align} directive changes it, and any label definition captures its
3481 current value. The location counter of the section where statements are being
3482 assembled is said to be the @dfn{active} location counter.
3485 @section bss Section
3488 @cindex common variable storage
3489 The bss section is used for local common variable storage.
3490 You may allocate address space in the bss section, but you may
3491 not dictate data to load into it before your program executes. When
3492 your program starts running, all the contents of the bss
3493 section are zeroed bytes.
3495 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
3496 @ref{Lcomm,,@code{.lcomm}}.
3498 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
3499 another form of uninitialized symbol; see @ref{Comm,,@code{.comm}}.
3502 When assembling for a target which supports multiple sections, such as ELF or
3503 COFF, you may switch into the @code{.bss} section and define symbols as usual;
3504 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
3505 section. Typically the section will only contain symbol definitions and
3506 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
3513 Symbols are a central concept: the programmer uses symbols to name
3514 things, the linker uses symbols to link, and the debugger uses symbols
3518 @cindex debuggers, and symbol order
3519 @emph{Warning:} @command{@value{AS}} does not place symbols in the object file in
3520 the same order they were declared. This may break some debuggers.
3525 * Setting Symbols:: Giving Symbols Other Values
3526 * Symbol Names:: Symbol Names
3527 * Dot:: The Special Dot Symbol
3528 * Symbol Attributes:: Symbol Attributes
3535 A @dfn{label} is written as a symbol immediately followed by a colon
3536 @samp{:}. The symbol then represents the current value of the
3537 active location counter, and is, for example, a suitable instruction
3538 operand. You are warned if you use the same symbol to represent two
3539 different locations: the first definition overrides any other
3543 On the HPPA, the usual form for a label need not be immediately followed by a
3544 colon, but instead must start in column zero. Only one label may be defined on
3545 a single line. To work around this, the HPPA version of @command{@value{AS}} also
3546 provides a special directive @code{.label} for defining labels more flexibly.
3549 @node Setting Symbols
3550 @section Giving Symbols Other Values
3552 @cindex assigning values to symbols
3553 @cindex symbol values, assigning
3554 A symbol can be given an arbitrary value by writing a symbol, followed
3555 by an equals sign @samp{=}, followed by an expression
3556 (@pxref{Expressions}). This is equivalent to using the @code{.set}
3557 directive. @xref{Set,,@code{.set}}. In the same way, using a double
3558 equals sign @samp{=}@samp{=} here represents an equivalent of the
3559 @code{.eqv} directive. @xref{Eqv,,@code{.eqv}}.
3562 Blackfin does not support symbol assignment with @samp{=}.
3566 @section Symbol Names
3568 @cindex symbol names
3569 @cindex names, symbol
3570 @ifclear SPECIAL-SYMS
3571 Symbol names begin with a letter or with one of @samp{._}. On most
3572 machines, you can also use @code{$} in symbol names; exceptions are
3573 noted in @ref{Machine Dependencies}. That character may be followed by any
3574 string of digits, letters, dollar signs (unless otherwise noted for a
3575 particular target machine), and underscores.
3579 Symbol names begin with a letter or with one of @samp{._}. On the
3580 Renesas SH you can also use @code{$} in symbol names. That
3581 character may be followed by any string of digits, letters, dollar signs (save
3582 on the H8/300), and underscores.
3586 Case of letters is significant: @code{foo} is a different symbol name
3589 Multibyte characters are supported. To generate a symbol name containing
3590 multibyte characters enclose it within double quotes and use escape codes. cf
3591 @xref{Strings}. Generating a multibyte symbol name from a label is not
3592 currently supported.
3594 Each symbol has exactly one name. Each name in an assembly language program
3595 refers to exactly one symbol. You may use that symbol name any number of times
3598 @subheading Local Symbol Names
3600 @cindex local symbol names
3601 @cindex symbol names, local
3602 A local symbol is any symbol beginning with certain local label prefixes.
3603 By default, the local label prefix is @samp{.L} for ELF systems or
3604 @samp{L} for traditional a.out systems, but each target may have its own
3605 set of local label prefixes.
3607 On the HPPA local symbols begin with @samp{L$}.
3610 Local symbols are defined and used within the assembler, but they are
3611 normally not saved in object files. Thus, they are not visible when debugging.
3612 You may use the @samp{-L} option (@pxref{L, ,Include Local Symbols:
3613 @option{-L}}) to retain the local symbols in the object files.
3615 @subheading Local Labels
3617 @cindex local labels
3618 @cindex temporary symbol names
3619 @cindex symbol names, temporary
3620 Local labels help compilers and programmers use names temporarily.
3621 They create symbols which are guaranteed to be unique over the entire scope of
3622 the input source code and which can be referred to by a simple notation.
3623 To define a local label, write a label of the form @samp{@b{N}:} (where @b{N}
3624 represents any positive integer). To refer to the most recent previous
3625 definition of that label write @samp{@b{N}b}, using the same number as when
3626 you defined the label. To refer to the next definition of a local label, write
3627 @samp{@b{N}f}---the @samp{b} stands for ``backwards'' and the @samp{f} stands
3630 There is no restriction on how you can use these labels, and you can reuse them
3631 too. So that it is possible to repeatedly define the same local label (using
3632 the same number @samp{@b{N}}), although you can only refer to the most recently
3633 defined local label of that number (for a backwards reference) or the next
3634 definition of a specific local label for a forward reference. It is also worth
3635 noting that the first 10 local labels (@samp{@b{0:}}@dots{}@samp{@b{9:}}) are
3636 implemented in a slightly more efficient manner than the others.
3647 Which is the equivalent of:
3650 label_1: branch label_3
3651 label_2: branch label_1
3652 label_3: branch label_4
3653 label_4: branch label_3
3656 Local label names are only a notational device. They are immediately
3657 transformed into more conventional symbol names before the assembler uses them.
3658 The symbol names are stored in the symbol table, appear in error messages, and
3659 are optionally emitted to the object file. The names are constructed using
3663 @item @emph{local label prefix}
3664 All local symbols begin with the system-specific local label prefix.
3665 Normally both @command{@value{AS}} and @code{@value{LD}} forget symbols
3666 that start with the local label prefix. These labels are
3667 used for symbols you are never intended to see. If you use the
3668 @samp{-L} option then @command{@value{AS}} retains these symbols in the
3669 object file. If you also instruct @code{@value{LD}} to retain these symbols,
3670 you may use them in debugging.
3673 This is the number that was used in the local label definition. So if the
3674 label is written @samp{55:} then the number is @samp{55}.
3677 This unusual character is included so you do not accidentally invent a symbol
3678 of the same name. The character has ASCII value of @samp{\002} (control-B).
3680 @item @emph{ordinal number}
3681 This is a serial number to keep the labels distinct. The first definition of
3682 @samp{0:} gets the number @samp{1}. The 15th definition of @samp{0:} gets the
3683 number @samp{15}, and so on. Likewise the first definition of @samp{1:} gets
3684 the number @samp{1} and its 15th definition gets @samp{15} as well.
3687 So for example, the first @code{1:} may be named @code{.L1@kbd{C-B}1}, and
3688 the 44th @code{3:} may be named @code{.L3@kbd{C-B}44}.
3690 @subheading Dollar Local Labels
3691 @cindex dollar local symbols
3693 @code{@value{AS}} also supports an even more local form of local labels called
3694 dollar labels. These labels go out of scope (i.e., they become undefined) as
3695 soon as a non-local label is defined. Thus they remain valid for only a small
3696 region of the input source code. Normal local labels, by contrast, remain in
3697 scope for the entire file, or until they are redefined by another occurrence of
3698 the same local label.
3700 Dollar labels are defined in exactly the same way as ordinary local labels,
3701 except that they have a dollar sign suffix to their numeric value, e.g.,
3704 They can also be distinguished from ordinary local labels by their transformed
3705 names which use ASCII character @samp{\001} (control-A) as the magic character
3706 to distinguish them from ordinary labels. For example, the fifth definition of
3707 @samp{6$} may be named @samp{.L6@kbd{C-A}5}.
3710 @section The Special Dot Symbol
3712 @cindex dot (symbol)
3713 @cindex @code{.} (symbol)
3714 @cindex current address
3715 @cindex location counter
3716 The special symbol @samp{.} refers to the current address that
3717 @command{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
3718 .long .} defines @code{melvin} to contain its own address.
3719 Assigning a value to @code{.} is treated the same as a @code{.org}
3721 @ifclear no-space-dir
3722 Thus, the expression @samp{.=.+4} is the same as saying
3726 @node Symbol Attributes
3727 @section Symbol Attributes
3729 @cindex symbol attributes
3730 @cindex attributes, symbol
3731 Every symbol has, as well as its name, the attributes ``Value'' and
3732 ``Type''. Depending on output format, symbols can also have auxiliary
3735 The detailed definitions are in @file{a.out.h}.
3738 If you use a symbol without defining it, @command{@value{AS}} assumes zero for
3739 all these attributes, and probably won't warn you. This makes the
3740 symbol an externally defined symbol, which is generally what you
3744 * Symbol Value:: Value
3745 * Symbol Type:: Type
3748 * a.out Symbols:: Symbol Attributes: @code{a.out}
3752 * a.out Symbols:: Symbol Attributes: @code{a.out}
3755 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3760 * COFF Symbols:: Symbol Attributes for COFF
3763 * SOM Symbols:: Symbol Attributes for SOM
3770 @cindex value of a symbol
3771 @cindex symbol value
3772 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3773 location in the text, data, bss or absolute sections the value is the
3774 number of addresses from the start of that section to the label.
3775 Naturally for text, data and bss sections the value of a symbol changes
3776 as @code{@value{LD}} changes section base addresses during linking. Absolute
3777 symbols' values do not change during linking: that is why they are
3780 The value of an undefined symbol is treated in a special way. If it is
3781 0 then the symbol is not defined in this assembler source file, and
3782 @code{@value{LD}} tries to determine its value from other files linked into the
3783 same program. You make this kind of symbol simply by mentioning a symbol
3784 name without defining it. A non-zero value represents a @code{.comm}
3785 common declaration. The value is how much common storage to reserve, in
3786 bytes (addresses). The symbol refers to the first address of the
3792 @cindex type of a symbol
3794 The type attribute of a symbol contains relocation (section)
3795 information, any flag settings indicating that a symbol is external, and
3796 (optionally), other information for linkers and debuggers. The exact
3797 format depends on the object-code output format in use.
3802 @c The following avoids a "widow" subsection title. @group would be
3803 @c better if it were available outside examples.
3806 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3808 @cindex @code{b.out} symbol attributes
3809 @cindex symbol attributes, @code{b.out}
3810 These symbol attributes appear only when @command{@value{AS}} is configured for
3811 one of the Berkeley-descended object output formats---@code{a.out} or
3817 @subsection Symbol Attributes: @code{a.out}
3819 @cindex @code{a.out} symbol attributes
3820 @cindex symbol attributes, @code{a.out}
3826 @subsection Symbol Attributes: @code{a.out}
3828 @cindex @code{a.out} symbol attributes
3829 @cindex symbol attributes, @code{a.out}
3833 * Symbol Desc:: Descriptor
3834 * Symbol Other:: Other
3838 @subsubsection Descriptor
3840 @cindex descriptor, of @code{a.out} symbol
3841 This is an arbitrary 16-bit value. You may establish a symbol's
3842 descriptor value by using a @code{.desc} statement
3843 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3844 @command{@value{AS}}.
3847 @subsubsection Other
3849 @cindex other attribute, of @code{a.out} symbol
3850 This is an arbitrary 8-bit value. It means nothing to @command{@value{AS}}.
3855 @subsection Symbol Attributes for COFF
3857 @cindex COFF symbol attributes
3858 @cindex symbol attributes, COFF
3860 The COFF format supports a multitude of auxiliary symbol attributes;
3861 like the primary symbol attributes, they are set between @code{.def} and
3862 @code{.endef} directives.
3864 @subsubsection Primary Attributes
3866 @cindex primary attributes, COFF symbols
3867 The symbol name is set with @code{.def}; the value and type,
3868 respectively, with @code{.val} and @code{.type}.
3870 @subsubsection Auxiliary Attributes
3872 @cindex auxiliary attributes, COFF symbols
3873 The @command{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3874 @code{.size}, @code{.tag}, and @code{.weak} can generate auxiliary symbol
3875 table information for COFF.
3880 @subsection Symbol Attributes for SOM
3882 @cindex SOM symbol attributes
3883 @cindex symbol attributes, SOM
3885 The SOM format for the HPPA supports a multitude of symbol attributes set with
3886 the @code{.EXPORT} and @code{.IMPORT} directives.
3888 The attributes are described in @cite{HP9000 Series 800 Assembly
3889 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3890 @code{EXPORT} assembler directive documentation.
3894 @chapter Expressions
3898 @cindex numeric values
3899 An @dfn{expression} specifies an address or numeric value.
3900 Whitespace may precede and/or follow an expression.
3902 The result of an expression must be an absolute number, or else an offset into
3903 a particular section. If an expression is not absolute, and there is not
3904 enough information when @command{@value{AS}} sees the expression to know its
3905 section, a second pass over the source program might be necessary to interpret
3906 the expression---but the second pass is currently not implemented.
3907 @command{@value{AS}} aborts with an error message in this situation.
3910 * Empty Exprs:: Empty Expressions
3911 * Integer Exprs:: Integer Expressions
3915 @section Empty Expressions
3917 @cindex empty expressions
3918 @cindex expressions, empty
3919 An empty expression has no value: it is just whitespace or null.
3920 Wherever an absolute expression is required, you may omit the
3921 expression, and @command{@value{AS}} assumes a value of (absolute) 0. This
3922 is compatible with other assemblers.
3925 @section Integer Expressions
3927 @cindex integer expressions
3928 @cindex expressions, integer
3929 An @dfn{integer expression} is one or more @emph{arguments} delimited
3930 by @emph{operators}.
3933 * Arguments:: Arguments
3934 * Operators:: Operators
3935 * Prefix Ops:: Prefix Operators
3936 * Infix Ops:: Infix Operators
3940 @subsection Arguments
3942 @cindex expression arguments
3943 @cindex arguments in expressions
3944 @cindex operands in expressions
3945 @cindex arithmetic operands
3946 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3947 contexts arguments are sometimes called ``arithmetic operands''. In
3948 this manual, to avoid confusing them with the ``instruction operands'' of
3949 the machine language, we use the term ``argument'' to refer to parts of
3950 expressions only, reserving the word ``operand'' to refer only to machine
3951 instruction operands.
3953 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3954 @var{section} is one of text, data, bss, absolute,
3955 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3958 Numbers are usually integers.
3960 A number can be a flonum or bignum. In this case, you are warned
3961 that only the low order 32 bits are used, and @command{@value{AS}} pretends
3962 these 32 bits are an integer. You may write integer-manipulating
3963 instructions that act on exotic constants, compatible with other
3966 @cindex subexpressions
3967 Subexpressions are a left parenthesis @samp{(} followed by an integer
3968 expression, followed by a right parenthesis @samp{)}; or a prefix
3969 operator followed by an argument.
3972 @subsection Operators
3974 @cindex operators, in expressions
3975 @cindex arithmetic functions
3976 @cindex functions, in expressions
3977 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3978 operators are followed by an argument. Infix operators appear
3979 between their arguments. Operators may be preceded and/or followed by
3983 @subsection Prefix Operator
3985 @cindex prefix operators
3986 @command{@value{AS}} has the following @dfn{prefix operators}. They each take
3987 one argument, which must be absolute.
3989 @c the tex/end tex stuff surrounding this small table is meant to make
3990 @c it align, on the printed page, with the similar table in the next
3991 @c section (which is inside an enumerate).
3993 \global\advance\leftskip by \itemindent
3998 @dfn{Negation}. Two's complement negation.
4000 @dfn{Complementation}. Bitwise not.
4004 \global\advance\leftskip by -\itemindent
4008 @subsection Infix Operators
4010 @cindex infix operators
4011 @cindex operators, permitted arguments
4012 @dfn{Infix operators} take two arguments, one on either side. Operators
4013 have precedence, but operations with equal precedence are performed left
4014 to right. Apart from @code{+} or @option{-}, both arguments must be
4015 absolute, and the result is absolute.
4018 @cindex operator precedence
4019 @cindex precedence of operators
4026 @dfn{Multiplication}.
4029 @dfn{Division}. Truncation is the same as the C operator @samp{/}
4035 @dfn{Shift Left}. Same as the C operator @samp{<<}.
4038 @dfn{Shift Right}. Same as the C operator @samp{>>}.
4042 Intermediate precedence
4047 @dfn{Bitwise Inclusive Or}.
4053 @dfn{Bitwise Exclusive Or}.
4056 @dfn{Bitwise Or Not}.
4063 @cindex addition, permitted arguments
4064 @cindex plus, permitted arguments
4065 @cindex arguments for addition
4067 @dfn{Addition}. If either argument is absolute, the result has the section of
4068 the other argument. You may not add together arguments from different
4071 @cindex subtraction, permitted arguments
4072 @cindex minus, permitted arguments
4073 @cindex arguments for subtraction
4075 @dfn{Subtraction}. If the right argument is absolute, the
4076 result has the section of the left argument.
4077 If both arguments are in the same section, the result is absolute.
4078 You may not subtract arguments from different sections.
4079 @c FIXME is there still something useful to say about undefined - undefined ?
4081 @cindex comparison expressions
4082 @cindex expressions, comparison
4087 @dfn{Is Not Equal To}
4091 @dfn{Is Greater Than}
4093 @dfn{Is Greater Than Or Equal To}
4095 @dfn{Is Less Than Or Equal To}
4097 The comparison operators can be used as infix operators. A true results has a
4098 value of -1 whereas a false result has a value of 0. Note, these operators
4099 perform signed comparisons.
4102 @item Lowest Precedence
4111 These two logical operations can be used to combine the results of sub
4112 expressions. Note, unlike the comparison operators a true result returns a
4113 value of 1 but a false results does still return 0. Also note that the logical
4114 or operator has a slightly lower precedence than logical and.
4119 In short, it's only meaningful to add or subtract the @emph{offsets} in an
4120 address; you can only have a defined section in one of the two arguments.
4123 @chapter Assembler Directives
4125 @cindex directives, machine independent
4126 @cindex pseudo-ops, machine independent
4127 @cindex machine independent directives
4128 All assembler directives have names that begin with a period (@samp{.}).
4129 The rest of the name is letters, usually in lower case.
4131 This chapter discusses directives that are available regardless of the
4132 target machine configuration for the @sc{gnu} assembler.
4134 Some machine configurations provide additional directives.
4135 @xref{Machine Dependencies}.
4138 @ifset machine-directives
4139 @xref{Machine Dependencies}, for additional directives.
4144 * Abort:: @code{.abort}
4146 * ABORT (COFF):: @code{.ABORT}
4149 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
4150 * Altmacro:: @code{.altmacro}
4151 * Ascii:: @code{.ascii "@var{string}"}@dots{}
4152 * Asciz:: @code{.asciz "@var{string}"}@dots{}
4153 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
4154 * Bundle directives:: @code{.bundle_align_mode @var{abs-expr}}, @code{.bundle_lock}, @code{.bundle_unlock}
4155 * Byte:: @code{.byte @var{expressions}}
4156 * CFI directives:: @code{.cfi_startproc [simple]}, @code{.cfi_endproc}, etc.
4157 * Comm:: @code{.comm @var{symbol} , @var{length} }
4158 * Data:: @code{.data @var{subsection}}
4160 * Def:: @code{.def @var{name}}
4163 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
4169 * Double:: @code{.double @var{flonums}}
4170 * Eject:: @code{.eject}
4171 * Else:: @code{.else}
4172 * Elseif:: @code{.elseif}
4175 * Endef:: @code{.endef}
4178 * Endfunc:: @code{.endfunc}
4179 * Endif:: @code{.endif}
4180 * Equ:: @code{.equ @var{symbol}, @var{expression}}
4181 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
4182 * Eqv:: @code{.eqv @var{symbol}, @var{expression}}
4184 * Error:: @code{.error @var{string}}
4185 * Exitm:: @code{.exitm}
4186 * Extern:: @code{.extern}
4187 * Fail:: @code{.fail}
4188 * File:: @code{.file}
4189 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
4190 * Float:: @code{.float @var{flonums}}
4191 * Func:: @code{.func}
4192 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
4194 * Gnu_attribute:: @code{.gnu_attribute @var{tag},@var{value}}
4195 * Hidden:: @code{.hidden @var{names}}
4198 * hword:: @code{.hword @var{expressions}}
4199 * Ident:: @code{.ident}
4200 * If:: @code{.if @var{absolute expression}}
4201 * Incbin:: @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
4202 * Include:: @code{.include "@var{file}"}
4203 * Int:: @code{.int @var{expressions}}
4205 * Internal:: @code{.internal @var{names}}
4208 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
4209 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
4210 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
4211 * Lflags:: @code{.lflags}
4212 @ifclear no-line-dir
4213 * Line:: @code{.line @var{line-number}}
4216 * Linkonce:: @code{.linkonce [@var{type}]}
4217 * List:: @code{.list}
4218 * Ln:: @code{.ln @var{line-number}}
4219 * Loc:: @code{.loc @var{fileno} @var{lineno}}
4220 * Loc_mark_labels:: @code{.loc_mark_labels @var{enable}}
4222 * Local:: @code{.local @var{names}}
4225 * Long:: @code{.long @var{expressions}}
4227 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
4230 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
4231 * MRI:: @code{.mri @var{val}}
4232 * Noaltmacro:: @code{.noaltmacro}
4233 * Nolist:: @code{.nolist}
4234 * Octa:: @code{.octa @var{bignums}}
4235 * Offset:: @code{.offset @var{loc}}
4236 * Org:: @code{.org @var{new-lc}, @var{fill}}
4237 * P2align:: @code{.p2align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4239 * PopSection:: @code{.popsection}
4240 * Previous:: @code{.previous}
4243 * Print:: @code{.print @var{string}}
4245 * Protected:: @code{.protected @var{names}}
4248 * Psize:: @code{.psize @var{lines}, @var{columns}}
4249 * Purgem:: @code{.purgem @var{name}}
4251 * PushSection:: @code{.pushsection @var{name}}
4254 * Quad:: @code{.quad @var{bignums}}
4255 * Reloc:: @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
4256 * Rept:: @code{.rept @var{count}}
4257 * Sbttl:: @code{.sbttl "@var{subheading}"}
4259 * Scl:: @code{.scl @var{class}}
4262 * Section:: @code{.section @var{name}[, @var{flags}]}
4265 * Set:: @code{.set @var{symbol}, @var{expression}}
4266 * Short:: @code{.short @var{expressions}}
4267 * Single:: @code{.single @var{flonums}}
4269 * Size:: @code{.size [@var{name} , @var{expression}]}
4271 @ifclear no-space-dir
4272 * Skip:: @code{.skip @var{size} , @var{fill}}
4275 * Sleb128:: @code{.sleb128 @var{expressions}}
4276 @ifclear no-space-dir
4277 * Space:: @code{.space @var{size} , @var{fill}}
4280 * Stab:: @code{.stabd, .stabn, .stabs}
4283 * String:: @code{.string "@var{str}"}, @code{.string8 "@var{str}"}, @code{.string16 "@var{str}"}, @code{.string32 "@var{str}"}, @code{.string64 "@var{str}"}
4284 * Struct:: @code{.struct @var{expression}}
4286 * SubSection:: @code{.subsection}
4287 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
4291 * Tag:: @code{.tag @var{structname}}
4294 * Text:: @code{.text @var{subsection}}
4295 * Title:: @code{.title "@var{heading}"}
4297 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
4300 * Uleb128:: @code{.uleb128 @var{expressions}}
4302 * Val:: @code{.val @var{addr}}
4306 * Version:: @code{.version "@var{string}"}
4307 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
4308 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
4311 * Warning:: @code{.warning @var{string}}
4312 * Weak:: @code{.weak @var{names}}
4313 * Weakref:: @code{.weakref @var{alias}, @var{symbol}}
4314 * Word:: @code{.word @var{expressions}}
4315 * Deprecated:: Deprecated Directives
4319 @section @code{.abort}
4321 @cindex @code{abort} directive
4322 @cindex stopping the assembly
4323 This directive stops the assembly immediately. It is for
4324 compatibility with other assemblers. The original idea was that the
4325 assembly language source would be piped into the assembler. If the sender
4326 of the source quit, it could use this directive tells @command{@value{AS}} to
4327 quit also. One day @code{.abort} will not be supported.
4331 @section @code{.ABORT} (COFF)
4333 @cindex @code{ABORT} directive
4334 When producing COFF output, @command{@value{AS}} accepts this directive as a
4335 synonym for @samp{.abort}.
4338 When producing @code{b.out} output, @command{@value{AS}} accepts this directive,
4344 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4346 @cindex padding the location counter
4347 @cindex @code{align} directive
4348 Pad the location counter (in the current subsection) to a particular storage
4349 boundary. The first expression (which must be absolute) is the alignment
4350 required, as described below.
4352 The second expression (also absolute) gives the fill value to be stored in the
4353 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4354 padding bytes are normally zero. However, on some systems, if the section is
4355 marked as containing code and the fill value is omitted, the space is filled
4356 with no-op instructions.
4358 The third expression is also absolute, and is also optional. If it is present,
4359 it is the maximum number of bytes that should be skipped by this alignment
4360 directive. If doing the alignment would require skipping more bytes than the
4361 specified maximum, then the alignment is not done at all. You can omit the
4362 fill value (the second argument) entirely by simply using two commas after the
4363 required alignment; this can be useful if you want the alignment to be filled
4364 with no-op instructions when appropriate.
4366 The way the required alignment is specified varies from system to system.
4367 For the arc, hppa, i386 using ELF, i860, iq2000, m68k, or1k,
4368 s390, sparc, tic4x, tic80 and xtensa, the first expression is the
4369 alignment request in bytes. For example @samp{.align 8} advances
4370 the location counter until it is a multiple of 8. If the location counter
4371 is already a multiple of 8, no change is needed. For the tic54x, the
4372 first expression is the alignment request in words.
4374 For other systems, including ppc, i386 using a.out format, arm and
4375 strongarm, it is the
4376 number of low-order zero bits the location counter must have after
4377 advancement. For example @samp{.align 3} advances the location
4378 counter until it a multiple of 8. If the location counter is already a
4379 multiple of 8, no change is needed.
4381 This inconsistency is due to the different behaviors of the various
4382 native assemblers for these systems which GAS must emulate.
4383 GAS also provides @code{.balign} and @code{.p2align} directives,
4384 described later, which have a consistent behavior across all
4385 architectures (but are specific to GAS).
4388 @section @code{.altmacro}
4389 Enable alternate macro mode, enabling:
4392 @item LOCAL @var{name} [ , @dots{} ]
4393 One additional directive, @code{LOCAL}, is available. It is used to
4394 generate a string replacement for each of the @var{name} arguments, and
4395 replace any instances of @var{name} in each macro expansion. The
4396 replacement string is unique in the assembly, and different for each
4397 separate macro expansion. @code{LOCAL} allows you to write macros that
4398 define symbols, without fear of conflict between separate macro expansions.
4400 @item String delimiters
4401 You can write strings delimited in these other ways besides
4402 @code{"@var{string}"}:
4405 @item '@var{string}'
4406 You can delimit strings with single-quote characters.
4408 @item <@var{string}>
4409 You can delimit strings with matching angle brackets.
4412 @item single-character string escape
4413 To include any single character literally in a string (even if the
4414 character would otherwise have some special meaning), you can prefix the
4415 character with @samp{!} (an exclamation mark). For example, you can
4416 write @samp{<4.3 !> 5.4!!>} to get the literal text @samp{4.3 > 5.4!}.
4418 @item Expression results as strings
4419 You can write @samp{%@var{expr}} to evaluate the expression @var{expr}
4420 and use the result as a string.
4424 @section @code{.ascii "@var{string}"}@dots{}
4426 @cindex @code{ascii} directive
4427 @cindex string literals
4428 @code{.ascii} expects zero or more string literals (@pxref{Strings})
4429 separated by commas. It assembles each string (with no automatic
4430 trailing zero byte) into consecutive addresses.
4433 @section @code{.asciz "@var{string}"}@dots{}
4435 @cindex @code{asciz} directive
4436 @cindex zero-terminated strings
4437 @cindex null-terminated strings
4438 @code{.asciz} is just like @code{.ascii}, but each string is followed by
4439 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
4442 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4444 @cindex padding the location counter given number of bytes
4445 @cindex @code{balign} directive
4446 Pad the location counter (in the current subsection) to a particular
4447 storage boundary. The first expression (which must be absolute) is the
4448 alignment request in bytes. For example @samp{.balign 8} advances
4449 the location counter until it is a multiple of 8. If the location counter
4450 is already a multiple of 8, no change is needed.
4452 The second expression (also absolute) gives the fill value to be stored in the
4453 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4454 padding bytes are normally zero. However, on some systems, if the section is
4455 marked as containing code and the fill value is omitted, the space is filled
4456 with no-op instructions.
4458 The third expression is also absolute, and is also optional. If it is present,
4459 it is the maximum number of bytes that should be skipped by this alignment
4460 directive. If doing the alignment would require skipping more bytes than the
4461 specified maximum, then the alignment is not done at all. You can omit the
4462 fill value (the second argument) entirely by simply using two commas after the
4463 required alignment; this can be useful if you want the alignment to be filled
4464 with no-op instructions when appropriate.
4466 @cindex @code{balignw} directive
4467 @cindex @code{balignl} directive
4468 The @code{.balignw} and @code{.balignl} directives are variants of the
4469 @code{.balign} directive. The @code{.balignw} directive treats the fill
4470 pattern as a two byte word value. The @code{.balignl} directives treats the
4471 fill pattern as a four byte longword value. For example, @code{.balignw
4472 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4473 filled in with the value 0x368d (the exact placement of the bytes depends upon
4474 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4477 @node Bundle directives
4478 @section @code{.bundle_align_mode @var{abs-expr}}
4479 @cindex @code{bundle_align_mode} directive
4481 @cindex instruction bundle
4482 @cindex aligned instruction bundle
4483 @code{.bundle_align_mode} enables or disables @dfn{aligned instruction
4484 bundle} mode. In this mode, sequences of adjacent instructions are grouped
4485 into fixed-sized @dfn{bundles}. If the argument is zero, this mode is
4486 disabled (which is the default state). If the argument it not zero, it
4487 gives the size of an instruction bundle as a power of two (as for the
4488 @code{.p2align} directive, @pxref{P2align}).
4490 For some targets, it's an ABI requirement that no instruction may span a
4491 certain aligned boundary. A @dfn{bundle} is simply a sequence of
4492 instructions that starts on an aligned boundary. For example, if
4493 @var{abs-expr} is @code{5} then the bundle size is 32, so each aligned
4494 chunk of 32 bytes is a bundle. When aligned instruction bundle mode is in
4495 effect, no single instruction may span a boundary between bundles. If an
4496 instruction would start too close to the end of a bundle for the length of
4497 that particular instruction to fit within the bundle, then the space at the
4498 end of that bundle is filled with no-op instructions so the instruction
4499 starts in the next bundle. As a corollary, it's an error if any single
4500 instruction's encoding is longer than the bundle size.
4502 @section @code{.bundle_lock} and @code{.bundle_unlock}
4503 @cindex @code{bundle_lock} directive
4504 @cindex @code{bundle_unlock} directive
4505 The @code{.bundle_lock} and directive @code{.bundle_unlock} directives
4506 allow explicit control over instruction bundle padding. These directives
4507 are only valid when @code{.bundle_align_mode} has been used to enable
4508 aligned instruction bundle mode. It's an error if they appear when
4509 @code{.bundle_align_mode} has not been used at all, or when the last
4510 directive was @w{@code{.bundle_align_mode 0}}.
4512 @cindex bundle-locked
4513 For some targets, it's an ABI requirement that certain instructions may
4514 appear only as part of specified permissible sequences of multiple
4515 instructions, all within the same bundle. A pair of @code{.bundle_lock}
4516 and @code{.bundle_unlock} directives define a @dfn{bundle-locked}
4517 instruction sequence. For purposes of aligned instruction bundle mode, a
4518 sequence starting with @code{.bundle_lock} and ending with
4519 @code{.bundle_unlock} is treated as a single instruction. That is, the
4520 entire sequence must fit into a single bundle and may not span a bundle
4521 boundary. If necessary, no-op instructions will be inserted before the
4522 first instruction of the sequence so that the whole sequence starts on an
4523 aligned bundle boundary. It's an error if the sequence is longer than the
4526 For convenience when using @code{.bundle_lock} and @code{.bundle_unlock}
4527 inside assembler macros (@pxref{Macro}), bundle-locked sequences may be
4528 nested. That is, a second @code{.bundle_lock} directive before the next
4529 @code{.bundle_unlock} directive has no effect except that it must be
4530 matched by another closing @code{.bundle_unlock} so that there is the
4531 same number of @code{.bundle_lock} and @code{.bundle_unlock} directives.
4534 @section @code{.byte @var{expressions}}
4536 @cindex @code{byte} directive
4537 @cindex integers, one byte
4538 @code{.byte} expects zero or more expressions, separated by commas.
4539 Each expression is assembled into the next byte.
4541 @node CFI directives
4542 @section @code{.cfi_sections @var{section_list}}
4543 @cindex @code{cfi_sections} directive
4544 @code{.cfi_sections} may be used to specify whether CFI directives
4545 should emit @code{.eh_frame} section and/or @code{.debug_frame} section.
4546 If @var{section_list} is @code{.eh_frame}, @code{.eh_frame} is emitted,
4547 if @var{section_list} is @code{.debug_frame}, @code{.debug_frame} is emitted.
4548 To emit both use @code{.eh_frame, .debug_frame}. The default if this
4549 directive is not used is @code{.cfi_sections .eh_frame}.
4551 @section @code{.cfi_startproc [simple]}
4552 @cindex @code{cfi_startproc} directive
4553 @code{.cfi_startproc} is used at the beginning of each function that
4554 should have an entry in @code{.eh_frame}. It initializes some internal
4555 data structures. Don't forget to close the function by
4556 @code{.cfi_endproc}.
4558 Unless @code{.cfi_startproc} is used along with parameter @code{simple}
4559 it also emits some architecture dependent initial CFI instructions.
4561 @section @code{.cfi_endproc}
4562 @cindex @code{cfi_endproc} directive
4563 @code{.cfi_endproc} is used at the end of a function where it closes its
4564 unwind entry previously opened by
4565 @code{.cfi_startproc}, and emits it to @code{.eh_frame}.
4567 @section @code{.cfi_personality @var{encoding} [, @var{exp}]}
4568 @code{.cfi_personality} defines personality routine and its encoding.
4569 @var{encoding} must be a constant determining how the personality
4570 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4571 argument is not present, otherwise second argument should be
4572 a constant or a symbol name. When using indirect encodings,
4573 the symbol provided should be the location where personality
4574 can be loaded from, not the personality routine itself.
4575 The default after @code{.cfi_startproc} is @code{.cfi_personality 0xff},
4576 no personality routine.
4578 @section @code{.cfi_lsda @var{encoding} [, @var{exp}]}
4579 @code{.cfi_lsda} defines LSDA and its encoding.
4580 @var{encoding} must be a constant determining how the LSDA
4581 should be encoded. If it is 255 (@code{DW_EH_PE_omit}), second
4582 argument is not present, otherwise second argument should be a constant
4583 or a symbol name. The default after @code{.cfi_startproc} is @code{.cfi_lsda 0xff},
4586 @section @code{.cfi_def_cfa @var{register}, @var{offset}}
4587 @code{.cfi_def_cfa} defines a rule for computing CFA as: @i{take
4588 address from @var{register} and add @var{offset} to it}.
4590 @section @code{.cfi_def_cfa_register @var{register}}
4591 @code{.cfi_def_cfa_register} modifies a rule for computing CFA. From
4592 now on @var{register} will be used instead of the old one. Offset
4595 @section @code{.cfi_def_cfa_offset @var{offset}}
4596 @code{.cfi_def_cfa_offset} modifies a rule for computing CFA. Register
4597 remains the same, but @var{offset} is new. Note that it is the
4598 absolute offset that will be added to a defined register to compute
4601 @section @code{.cfi_adjust_cfa_offset @var{offset}}
4602 Same as @code{.cfi_def_cfa_offset} but @var{offset} is a relative
4603 value that is added/substracted from the previous offset.
4605 @section @code{.cfi_offset @var{register}, @var{offset}}
4606 Previous value of @var{register} is saved at offset @var{offset} from
4609 @section @code{.cfi_rel_offset @var{register}, @var{offset}}
4610 Previous value of @var{register} is saved at offset @var{offset} from
4611 the current CFA register. This is transformed to @code{.cfi_offset}
4612 using the known displacement of the CFA register from the CFA.
4613 This is often easier to use, because the number will match the
4614 code it's annotating.
4616 @section @code{.cfi_register @var{register1}, @var{register2}}
4617 Previous value of @var{register1} is saved in register @var{register2}.
4619 @section @code{.cfi_restore @var{register}}
4620 @code{.cfi_restore} says that the rule for @var{register} is now the
4621 same as it was at the beginning of the function, after all initial
4622 instruction added by @code{.cfi_startproc} were executed.
4624 @section @code{.cfi_undefined @var{register}}
4625 From now on the previous value of @var{register} can't be restored anymore.
4627 @section @code{.cfi_same_value @var{register}}
4628 Current value of @var{register} is the same like in the previous frame,
4629 i.e. no restoration needed.
4631 @section @code{.cfi_remember_state},
4632 First save all current rules for all registers by @code{.cfi_remember_state},
4633 then totally screw them up by subsequent @code{.cfi_*} directives and when
4634 everything is hopelessly bad, use @code{.cfi_restore_state} to restore
4635 the previous saved state.
4637 @section @code{.cfi_return_column @var{register}}
4638 Change return column @var{register}, i.e. the return address is either
4639 directly in @var{register} or can be accessed by rules for @var{register}.
4641 @section @code{.cfi_signal_frame}
4642 Mark current function as signal trampoline.
4644 @section @code{.cfi_window_save}
4645 SPARC register window has been saved.
4647 @section @code{.cfi_escape} @var{expression}[, @dots{}]
4648 Allows the user to add arbitrary bytes to the unwind info. One
4649 might use this to add OS-specific CFI opcodes, or generic CFI
4650 opcodes that GAS does not yet support.
4652 @section @code{.cfi_val_encoded_addr @var{register}, @var{encoding}, @var{label}}
4653 The current value of @var{register} is @var{label}. The value of @var{label}
4654 will be encoded in the output file according to @var{encoding}; see the
4655 description of @code{.cfi_personality} for details on this encoding.
4657 The usefulness of equating a register to a fixed label is probably
4658 limited to the return address register. Here, it can be useful to
4659 mark a code segment that has only one return address which is reached
4660 by a direct branch and no copy of the return address exists in memory
4661 or another register.
4664 @section @code{.comm @var{symbol} , @var{length} }
4666 @cindex @code{comm} directive
4667 @cindex symbol, common
4668 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
4669 common symbol in one object file may be merged with a defined or common symbol
4670 of the same name in another object file. If @code{@value{LD}} does not see a
4671 definition for the symbol--just one or more common symbols--then it will
4672 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
4673 absolute expression. If @code{@value{LD}} sees multiple common symbols with
4674 the same name, and they do not all have the same size, it will allocate space
4675 using the largest size.
4678 When using ELF or (as a GNU extension) PE, the @code{.comm} directive takes
4679 an optional third argument. This is the desired alignment of the symbol,
4680 specified for ELF as a byte boundary (for example, an alignment of 16 means
4681 that the least significant 4 bits of the address should be zero), and for PE
4682 as a power of two (for example, an alignment of 5 means aligned to a 32-byte
4683 boundary). The alignment must be an absolute expression, and it must be a
4684 power of two. If @code{@value{LD}} allocates uninitialized memory for the
4685 common symbol, it will use the alignment when placing the symbol. If no
4686 alignment is specified, @command{@value{AS}} will set the alignment to the
4687 largest power of two less than or equal to the size of the symbol, up to a
4688 maximum of 16 on ELF, or the default section alignment of 4 on PE@footnote{This
4689 is not the same as the executable image file alignment controlled by @code{@value{LD}}'s
4690 @samp{--section-alignment} option; image file sections in PE are aligned to
4691 multiples of 4096, which is far too large an alignment for ordinary variables.
4692 It is rather the default alignment for (non-debug) sections within object
4693 (@samp{*.o}) files, which are less strictly aligned.}.
4697 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
4698 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
4702 @section @code{.data @var{subsection}}
4704 @cindex @code{data} directive
4705 @code{.data} tells @command{@value{AS}} to assemble the following statements onto the
4706 end of the data subsection numbered @var{subsection} (which is an
4707 absolute expression). If @var{subsection} is omitted, it defaults
4712 @section @code{.def @var{name}}
4714 @cindex @code{def} directive
4715 @cindex COFF symbols, debugging
4716 @cindex debugging COFF symbols
4717 Begin defining debugging information for a symbol @var{name}; the
4718 definition extends until the @code{.endef} directive is encountered.
4721 This directive is only observed when @command{@value{AS}} is configured for COFF
4722 format output; when producing @code{b.out}, @samp{.def} is recognized,
4729 @section @code{.desc @var{symbol}, @var{abs-expression}}
4731 @cindex @code{desc} directive
4732 @cindex COFF symbol descriptor
4733 @cindex symbol descriptor, COFF
4734 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
4735 to the low 16 bits of an absolute expression.
4738 The @samp{.desc} directive is not available when @command{@value{AS}} is
4739 configured for COFF output; it is only for @code{a.out} or @code{b.out}
4740 object format. For the sake of compatibility, @command{@value{AS}} accepts
4741 it, but produces no output, when configured for COFF.
4747 @section @code{.dim}
4749 @cindex @code{dim} directive
4750 @cindex COFF auxiliary symbol information
4751 @cindex auxiliary symbol information, COFF
4752 This directive is generated by compilers to include auxiliary debugging
4753 information in the symbol table. It is only permitted inside
4754 @code{.def}/@code{.endef} pairs.
4757 @samp{.dim} is only meaningful when generating COFF format output; when
4758 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
4764 @section @code{.double @var{flonums}}
4766 @cindex @code{double} directive
4767 @cindex floating point numbers (double)
4768 @code{.double} expects zero or more flonums, separated by commas. It
4769 assembles floating point numbers.
4771 The exact kind of floating point numbers emitted depends on how
4772 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
4776 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
4777 in @sc{ieee} format.
4782 @section @code{.eject}
4784 @cindex @code{eject} directive
4785 @cindex new page, in listings
4786 @cindex page, in listings
4787 @cindex listing control: new page
4788 Force a page break at this point, when generating assembly listings.
4791 @section @code{.else}
4793 @cindex @code{else} directive
4794 @code{.else} is part of the @command{@value{AS}} support for conditional
4795 assembly; see @ref{If,,@code{.if}}. It marks the beginning of a section
4796 of code to be assembled if the condition for the preceding @code{.if}
4800 @section @code{.elseif}
4802 @cindex @code{elseif} directive
4803 @code{.elseif} is part of the @command{@value{AS}} support for conditional
4804 assembly; see @ref{If,,@code{.if}}. It is shorthand for beginning a new
4805 @code{.if} block that would otherwise fill the entire @code{.else} section.
4808 @section @code{.end}
4810 @cindex @code{end} directive
4811 @code{.end} marks the end of the assembly file. @command{@value{AS}} does not
4812 process anything in the file past the @code{.end} directive.
4816 @section @code{.endef}
4818 @cindex @code{endef} directive
4819 This directive flags the end of a symbol definition begun with
4823 @samp{.endef} is only meaningful when generating COFF format output; if
4824 @command{@value{AS}} is configured to generate @code{b.out}, it accepts this
4825 directive but ignores it.
4830 @section @code{.endfunc}
4831 @cindex @code{endfunc} directive
4832 @code{.endfunc} marks the end of a function specified with @code{.func}.
4835 @section @code{.endif}
4837 @cindex @code{endif} directive
4838 @code{.endif} is part of the @command{@value{AS}} support for conditional assembly;
4839 it marks the end of a block of code that is only assembled
4840 conditionally. @xref{If,,@code{.if}}.
4843 @section @code{.equ @var{symbol}, @var{expression}}
4845 @cindex @code{equ} directive
4846 @cindex assigning values to symbols
4847 @cindex symbols, assigning values to
4848 This directive sets the value of @var{symbol} to @var{expression}.
4849 It is synonymous with @samp{.set}; see @ref{Set,,@code{.set}}.
4852 The syntax for @code{equ} on the HPPA is
4853 @samp{@var{symbol} .equ @var{expression}}.
4857 The syntax for @code{equ} on the Z80 is
4858 @samp{@var{symbol} equ @var{expression}}.
4859 On the Z80 it is an eror if @var{symbol} is already defined,
4860 but the symbol is not protected from later redefinition.
4861 Compare @ref{Equiv}.
4865 @section @code{.equiv @var{symbol}, @var{expression}}
4866 @cindex @code{equiv} directive
4867 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
4868 the assembler will signal an error if @var{symbol} is already defined. Note a
4869 symbol which has been referenced but not actually defined is considered to be
4872 Except for the contents of the error message, this is roughly equivalent to
4879 plus it protects the symbol from later redefinition.
4882 @section @code{.eqv @var{symbol}, @var{expression}}
4883 @cindex @code{eqv} directive
4884 The @code{.eqv} directive is like @code{.equiv}, but no attempt is made to
4885 evaluate the expression or any part of it immediately. Instead each time
4886 the resulting symbol is used in an expression, a snapshot of its current
4890 @section @code{.err}
4891 @cindex @code{err} directive
4892 If @command{@value{AS}} assembles a @code{.err} directive, it will print an error
4893 message and, unless the @option{-Z} option was used, it will not generate an
4894 object file. This can be used to signal an error in conditionally compiled code.
4897 @section @code{.error "@var{string}"}
4898 @cindex error directive
4900 Similarly to @code{.err}, this directive emits an error, but you can specify a
4901 string that will be emitted as the error message. If you don't specify the
4902 message, it defaults to @code{".error directive invoked in source file"}.
4903 @xref{Errors, ,Error and Warning Messages}.
4906 .error "This code has not been assembled and tested."
4910 @section @code{.exitm}
4911 Exit early from the current macro definition. @xref{Macro}.
4914 @section @code{.extern}
4916 @cindex @code{extern} directive
4917 @code{.extern} is accepted in the source program---for compatibility
4918 with other assemblers---but it is ignored. @command{@value{AS}} treats
4919 all undefined symbols as external.
4922 @section @code{.fail @var{expression}}
4924 @cindex @code{fail} directive
4925 Generates an error or a warning. If the value of the @var{expression} is 500
4926 or more, @command{@value{AS}} will print a warning message. If the value is less
4927 than 500, @command{@value{AS}} will print an error message. The message will
4928 include the value of @var{expression}. This can occasionally be useful inside
4929 complex nested macros or conditional assembly.
4932 @section @code{.file}
4933 @cindex @code{file} directive
4935 @ifclear no-file-dir
4936 There are two different versions of the @code{.file} directive. Targets
4937 that support DWARF2 line number information use the DWARF2 version of
4938 @code{.file}. Other targets use the default version.
4940 @subheading Default Version
4942 @cindex logical file name
4943 @cindex file name, logical
4944 This version of the @code{.file} directive tells @command{@value{AS}} that we
4945 are about to start a new logical file. The syntax is:
4951 @var{string} is the new file name. In general, the filename is
4952 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
4953 to specify an empty file name, you must give the quotes--@code{""}. This
4954 statement may go away in future: it is only recognized to be compatible with
4955 old @command{@value{AS}} programs.
4957 @subheading DWARF2 Version
4960 When emitting DWARF2 line number information, @code{.file} assigns filenames
4961 to the @code{.debug_line} file name table. The syntax is:
4964 .file @var{fileno} @var{filename}
4967 The @var{fileno} operand should be a unique positive integer to use as the
4968 index of the entry in the table. The @var{filename} operand is a C string
4971 The detail of filename indices is exposed to the user because the filename
4972 table is shared with the @code{.debug_info} section of the DWARF2 debugging
4973 information, and thus the user must know the exact indices that table
4977 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
4979 @cindex @code{fill} directive
4980 @cindex writing patterns in memory
4981 @cindex patterns, writing in memory
4982 @var{repeat}, @var{size} and @var{value} are absolute expressions.
4983 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
4984 may be zero or more. @var{Size} may be zero or more, but if it is
4985 more than 8, then it is deemed to have the value 8, compatible with
4986 other people's assemblers. The contents of each @var{repeat} bytes
4987 is taken from an 8-byte number. The highest order 4 bytes are
4988 zero. The lowest order 4 bytes are @var{value} rendered in the
4989 byte-order of an integer on the computer @command{@value{AS}} is assembling for.
4990 Each @var{size} bytes in a repetition is taken from the lowest order
4991 @var{size} bytes of this number. Again, this bizarre behavior is
4992 compatible with other people's assemblers.
4994 @var{size} and @var{value} are optional.
4995 If the second comma and @var{value} are absent, @var{value} is
4996 assumed zero. If the first comma and following tokens are absent,
4997 @var{size} is assumed to be 1.
5000 @section @code{.float @var{flonums}}
5002 @cindex floating point numbers (single)
5003 @cindex @code{float} directive
5004 This directive assembles zero or more flonums, separated by commas. It
5005 has the same effect as @code{.single}.
5007 The exact kind of floating point numbers emitted depends on how
5008 @command{@value{AS}} is configured.
5009 @xref{Machine Dependencies}.
5013 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
5014 in @sc{ieee} format.
5019 @section @code{.func @var{name}[,@var{label}]}
5020 @cindex @code{func} directive
5021 @code{.func} emits debugging information to denote function @var{name}, and
5022 is ignored unless the file is assembled with debugging enabled.
5023 Only @samp{--gstabs[+]} is currently supported.
5024 @var{label} is the entry point of the function and if omitted @var{name}
5025 prepended with the @samp{leading char} is used.
5026 @samp{leading char} is usually @code{_} or nothing, depending on the target.
5027 All functions are currently defined to have @code{void} return type.
5028 The function must be terminated with @code{.endfunc}.
5031 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
5033 @cindex @code{global} directive
5034 @cindex symbol, making visible to linker
5035 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
5036 @var{symbol} in your partial program, its value is made available to
5037 other partial programs that are linked with it. Otherwise,
5038 @var{symbol} takes its attributes from a symbol of the same name
5039 from another file linked into the same program.
5041 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
5042 compatibility with other assemblers.
5045 On the HPPA, @code{.global} is not always enough to make it accessible to other
5046 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
5047 @xref{HPPA Directives, ,HPPA Assembler Directives}.
5052 @section @code{.gnu_attribute @var{tag},@var{value}}
5053 Record a @sc{gnu} object attribute for this file. @xref{Object Attributes}.
5056 @section @code{.hidden @var{names}}
5058 @cindex @code{hidden} directive
5060 This is one of the ELF visibility directives. The other two are
5061 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
5062 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5064 This directive overrides the named symbols default visibility (which is set by
5065 their binding: local, global or weak). The directive sets the visibility to
5066 @code{hidden} which means that the symbols are not visible to other components.
5067 Such symbols are always considered to be @code{protected} as well.
5071 @section @code{.hword @var{expressions}}
5073 @cindex @code{hword} directive
5074 @cindex integers, 16-bit
5075 @cindex numbers, 16-bit
5076 @cindex sixteen bit integers
5077 This expects zero or more @var{expressions}, and emits
5078 a 16 bit number for each.
5081 This directive is a synonym for @samp{.short}; depending on the target
5082 architecture, it may also be a synonym for @samp{.word}.
5086 This directive is a synonym for @samp{.short}.
5089 This directive is a synonym for both @samp{.short} and @samp{.word}.
5094 @section @code{.ident}
5096 @cindex @code{ident} directive
5098 This directive is used by some assemblers to place tags in object files. The
5099 behavior of this directive varies depending on the target. When using the
5100 a.out object file format, @command{@value{AS}} simply accepts the directive for
5101 source-file compatibility with existing assemblers, but does not emit anything
5102 for it. When using COFF, comments are emitted to the @code{.comment} or
5103 @code{.rdata} section, depending on the target. When using ELF, comments are
5104 emitted to the @code{.comment} section.
5107 @section @code{.if @var{absolute expression}}
5109 @cindex conditional assembly
5110 @cindex @code{if} directive
5111 @code{.if} marks the beginning of a section of code which is only
5112 considered part of the source program being assembled if the argument
5113 (which must be an @var{absolute expression}) is non-zero. The end of
5114 the conditional section of code must be marked by @code{.endif}
5115 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
5116 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
5117 If you have several conditions to check, @code{.elseif} may be used to avoid
5118 nesting blocks if/else within each subsequent @code{.else} block.
5120 The following variants of @code{.if} are also supported:
5122 @cindex @code{ifdef} directive
5123 @item .ifdef @var{symbol}
5124 Assembles the following section of code if the specified @var{symbol}
5125 has been defined. Note a symbol which has been referenced but not yet defined
5126 is considered to be undefined.
5128 @cindex @code{ifb} directive
5129 @item .ifb @var{text}
5130 Assembles the following section of code if the operand is blank (empty).
5132 @cindex @code{ifc} directive
5133 @item .ifc @var{string1},@var{string2}
5134 Assembles the following section of code if the two strings are the same. The
5135 strings may be optionally quoted with single quotes. If they are not quoted,
5136 the first string stops at the first comma, and the second string stops at the
5137 end of the line. Strings which contain whitespace should be quoted. The
5138 string comparison is case sensitive.
5140 @cindex @code{ifeq} directive
5141 @item .ifeq @var{absolute expression}
5142 Assembles the following section of code if the argument is zero.
5144 @cindex @code{ifeqs} directive
5145 @item .ifeqs @var{string1},@var{string2}
5146 Another form of @code{.ifc}. The strings must be quoted using double quotes.
5148 @cindex @code{ifge} directive
5149 @item .ifge @var{absolute expression}
5150 Assembles the following section of code if the argument is greater than or
5153 @cindex @code{ifgt} directive
5154 @item .ifgt @var{absolute expression}
5155 Assembles the following section of code if the argument is greater than zero.
5157 @cindex @code{ifle} directive
5158 @item .ifle @var{absolute expression}
5159 Assembles the following section of code if the argument is less than or equal
5162 @cindex @code{iflt} directive
5163 @item .iflt @var{absolute expression}
5164 Assembles the following section of code if the argument is less than zero.
5166 @cindex @code{ifnb} directive
5167 @item .ifnb @var{text}
5168 Like @code{.ifb}, but the sense of the test is reversed: this assembles the
5169 following section of code if the operand is non-blank (non-empty).
5171 @cindex @code{ifnc} directive
5172 @item .ifnc @var{string1},@var{string2}.
5173 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
5174 following section of code if the two strings are not the same.
5176 @cindex @code{ifndef} directive
5177 @cindex @code{ifnotdef} directive
5178 @item .ifndef @var{symbol}
5179 @itemx .ifnotdef @var{symbol}
5180 Assembles the following section of code if the specified @var{symbol}
5181 has not been defined. Both spelling variants are equivalent. Note a symbol
5182 which has been referenced but not yet defined is considered to be undefined.
5184 @cindex @code{ifne} directive
5185 @item .ifne @var{absolute expression}
5186 Assembles the following section of code if the argument is not equal to zero
5187 (in other words, this is equivalent to @code{.if}).
5189 @cindex @code{ifnes} directive
5190 @item .ifnes @var{string1},@var{string2}
5191 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
5192 following section of code if the two strings are not the same.
5196 @section @code{.incbin "@var{file}"[,@var{skip}[,@var{count}]]}
5198 @cindex @code{incbin} directive
5199 @cindex binary files, including
5200 The @code{incbin} directive includes @var{file} verbatim at the current
5201 location. You can control the search paths used with the @samp{-I} command-line
5202 option (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5205 The @var{skip} argument skips a number of bytes from the start of the
5206 @var{file}. The @var{count} argument indicates the maximum number of bytes to
5207 read. Note that the data is not aligned in any way, so it is the user's
5208 responsibility to make sure that proper alignment is provided both before and
5209 after the @code{incbin} directive.
5212 @section @code{.include "@var{file}"}
5214 @cindex @code{include} directive
5215 @cindex supporting files, including
5216 @cindex files, including
5217 This directive provides a way to include supporting files at specified
5218 points in your source program. The code from @var{file} is assembled as
5219 if it followed the point of the @code{.include}; when the end of the
5220 included file is reached, assembly of the original file continues. You
5221 can control the search paths used with the @samp{-I} command-line option
5222 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
5226 @section @code{.int @var{expressions}}
5228 @cindex @code{int} directive
5229 @cindex integers, 32-bit
5230 Expect zero or more @var{expressions}, of any section, separated by commas.
5231 For each expression, emit a number that, at run time, is the value of that
5232 expression. The byte order and bit size of the number depends on what kind
5233 of target the assembly is for.
5237 On most forms of the H8/300, @code{.int} emits 16-bit
5238 integers. On the H8/300H and the Renesas SH, however, @code{.int} emits
5245 @section @code{.internal @var{names}}
5247 @cindex @code{internal} directive
5249 This is one of the ELF visibility directives. The other two are
5250 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
5251 @code{.protected} (@pxref{Protected,,@code{.protected}}).
5253 This directive overrides the named symbols default visibility (which is set by
5254 their binding: local, global or weak). The directive sets the visibility to
5255 @code{internal} which means that the symbols are considered to be @code{hidden}
5256 (i.e., not visible to other components), and that some extra, processor specific
5257 processing must also be performed upon the symbols as well.
5261 @section @code{.irp @var{symbol},@var{values}}@dots{}
5263 @cindex @code{irp} directive
5264 Evaluate a sequence of statements assigning different values to @var{symbol}.
5265 The sequence of statements starts at the @code{.irp} directive, and is
5266 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
5267 set to @var{value}, and the sequence of statements is assembled. If no
5268 @var{value} is listed, the sequence of statements is assembled once, with
5269 @var{symbol} set to the null string. To refer to @var{symbol} within the
5270 sequence of statements, use @var{\symbol}.
5272 For example, assembling
5280 is equivalent to assembling
5288 For some caveats with the spelling of @var{symbol}, see also @ref{Macro}.
5291 @section @code{.irpc @var{symbol},@var{values}}@dots{}
5293 @cindex @code{irpc} directive
5294 Evaluate a sequence of statements assigning different values to @var{symbol}.
5295 The sequence of statements starts at the @code{.irpc} directive, and is
5296 terminated by an @code{.endr} directive. For each character in @var{value},
5297 @var{symbol} is set to the character, and the sequence of statements is
5298 assembled. If no @var{value} is listed, the sequence of statements is
5299 assembled once, with @var{symbol} set to the null string. To refer to
5300 @var{symbol} within the sequence of statements, use @var{\symbol}.
5302 For example, assembling
5310 is equivalent to assembling
5318 For some caveats with the spelling of @var{symbol}, see also the discussion
5322 @section @code{.lcomm @var{symbol} , @var{length}}
5324 @cindex @code{lcomm} directive
5325 @cindex local common symbols
5326 @cindex symbols, local common
5327 Reserve @var{length} (an absolute expression) bytes for a local common
5328 denoted by @var{symbol}. The section and value of @var{symbol} are
5329 those of the new local common. The addresses are allocated in the bss
5330 section, so that at run-time the bytes start off zeroed. @var{Symbol}
5331 is not declared global (@pxref{Global,,@code{.global}}), so is normally
5332 not visible to @code{@value{LD}}.
5335 Some targets permit a third argument to be used with @code{.lcomm}. This
5336 argument specifies the desired alignment of the symbol in the bss section.
5340 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
5341 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
5345 @section @code{.lflags}
5347 @cindex @code{lflags} directive (ignored)
5348 @command{@value{AS}} accepts this directive, for compatibility with other
5349 assemblers, but ignores it.
5351 @ifclear no-line-dir
5353 @section @code{.line @var{line-number}}
5355 @cindex @code{line} directive
5356 @cindex logical line number
5358 Change the logical line number. @var{line-number} must be an absolute
5359 expression. The next line has that logical line number. Therefore any other
5360 statements on the current line (after a statement separator character) are
5361 reported as on logical line number @var{line-number} @minus{} 1. One day
5362 @command{@value{AS}} will no longer support this directive: it is recognized only
5363 for compatibility with existing assembler programs.
5366 Even though this is a directive associated with the @code{a.out} or
5367 @code{b.out} object-code formats, @command{@value{AS}} still recognizes it
5368 when producing COFF output, and treats @samp{.line} as though it
5369 were the COFF @samp{.ln} @emph{if} it is found outside a
5370 @code{.def}/@code{.endef} pair.
5372 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
5373 used by compilers to generate auxiliary symbol information for
5378 @section @code{.linkonce [@var{type}]}
5380 @cindex @code{linkonce} directive
5381 @cindex common sections
5382 Mark the current section so that the linker only includes a single copy of it.
5383 This may be used to include the same section in several different object files,
5384 but ensure that the linker will only include it once in the final output file.
5385 The @code{.linkonce} pseudo-op must be used for each instance of the section.
5386 Duplicate sections are detected based on the section name, so it should be
5389 This directive is only supported by a few object file formats; as of this
5390 writing, the only object file format which supports it is the Portable
5391 Executable format used on Windows NT.
5393 The @var{type} argument is optional. If specified, it must be one of the
5394 following strings. For example:
5398 Not all types may be supported on all object file formats.
5402 Silently discard duplicate sections. This is the default.
5405 Warn if there are duplicate sections, but still keep only one copy.
5408 Warn if any of the duplicates have different sizes.
5411 Warn if any of the duplicates do not have exactly the same contents.
5415 @section @code{.list}
5417 @cindex @code{list} directive
5418 @cindex listing control, turning on
5419 Control (in conjunction with the @code{.nolist} directive) whether or
5420 not assembly listings are generated. These two directives maintain an
5421 internal counter (which is zero initially). @code{.list} increments the
5422 counter, and @code{.nolist} decrements it. Assembly listings are
5423 generated whenever the counter is greater than zero.
5425 By default, listings are disabled. When you enable them (with the
5426 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
5427 the initial value of the listing counter is one.
5430 @section @code{.ln @var{line-number}}
5432 @cindex @code{ln} directive
5433 @ifclear no-line-dir
5434 @samp{.ln} is a synonym for @samp{.line}.
5437 Tell @command{@value{AS}} to change the logical line number. @var{line-number}
5438 must be an absolute expression. The next line has that logical
5439 line number, so any other statements on the current line (after a
5440 statement separator character @code{;}) are reported as on logical
5441 line number @var{line-number} @minus{} 1.
5444 This directive is accepted, but ignored, when @command{@value{AS}} is
5445 configured for @code{b.out}; its effect is only associated with COFF
5451 @section @code{.loc @var{fileno} @var{lineno} [@var{column}] [@var{options}]}
5452 @cindex @code{loc} directive
5453 When emitting DWARF2 line number information,
5454 the @code{.loc} directive will add a row to the @code{.debug_line} line
5455 number matrix corresponding to the immediately following assembly
5456 instruction. The @var{fileno}, @var{lineno}, and optional @var{column}
5457 arguments will be applied to the @code{.debug_line} state machine before
5460 The @var{options} are a sequence of the following tokens in any order:
5464 This option will set the @code{basic_block} register in the
5465 @code{.debug_line} state machine to @code{true}.
5468 This option will set the @code{prologue_end} register in the
5469 @code{.debug_line} state machine to @code{true}.
5471 @item epilogue_begin
5472 This option will set the @code{epilogue_begin} register in the
5473 @code{.debug_line} state machine to @code{true}.
5475 @item is_stmt @var{value}
5476 This option will set the @code{is_stmt} register in the
5477 @code{.debug_line} state machine to @code{value}, which must be
5480 @item isa @var{value}
5481 This directive will set the @code{isa} register in the @code{.debug_line}
5482 state machine to @var{value}, which must be an unsigned integer.
5484 @item discriminator @var{value}
5485 This directive will set the @code{discriminator} register in the @code{.debug_line}
5486 state machine to @var{value}, which must be an unsigned integer.
5490 @node Loc_mark_labels
5491 @section @code{.loc_mark_labels @var{enable}}
5492 @cindex @code{loc_mark_labels} directive
5493 When emitting DWARF2 line number information,
5494 the @code{.loc_mark_labels} directive makes the assembler emit an entry
5495 to the @code{.debug_line} line number matrix with the @code{basic_block}
5496 register in the state machine set whenever a code label is seen.
5497 The @var{enable} argument should be either 1 or 0, to enable or disable
5498 this function respectively.
5502 @section @code{.local @var{names}}
5504 @cindex @code{local} directive
5505 This directive, which is available for ELF targets, marks each symbol in
5506 the comma-separated list of @code{names} as a local symbol so that it
5507 will not be externally visible. If the symbols do not already exist,
5508 they will be created.
5510 For targets where the @code{.lcomm} directive (@pxref{Lcomm}) does not
5511 accept an alignment argument, which is the case for most ELF targets,
5512 the @code{.local} directive can be used in combination with @code{.comm}
5513 (@pxref{Comm}) to define aligned local common data.
5517 @section @code{.long @var{expressions}}
5519 @cindex @code{long} directive
5520 @code{.long} is the same as @samp{.int}. @xref{Int,,@code{.int}}.
5523 @c no one seems to know what this is for or whether this description is
5524 @c what it really ought to do
5526 @section @code{.lsym @var{symbol}, @var{expression}}
5528 @cindex @code{lsym} directive
5529 @cindex symbol, not referenced in assembly
5530 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
5531 the hash table, ensuring it cannot be referenced by name during the
5532 rest of the assembly. This sets the attributes of the symbol to be
5533 the same as the expression value:
5535 @var{other} = @var{descriptor} = 0
5536 @var{type} = @r{(section of @var{expression})}
5537 @var{value} = @var{expression}
5540 The new symbol is not flagged as external.
5544 @section @code{.macro}
5547 The commands @code{.macro} and @code{.endm} allow you to define macros that
5548 generate assembly output. For example, this definition specifies a macro
5549 @code{sum} that puts a sequence of numbers into memory:
5552 .macro sum from=0, to=5
5561 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
5573 @item .macro @var{macname}
5574 @itemx .macro @var{macname} @var{macargs} @dots{}
5575 @cindex @code{macro} directive
5576 Begin the definition of a macro called @var{macname}. If your macro
5577 definition requires arguments, specify their names after the macro name,
5578 separated by commas or spaces. You can qualify the macro argument to
5579 indicate whether all invocations must specify a non-blank value (through
5580 @samp{:@code{req}}), or whether it takes all of the remaining arguments
5581 (through @samp{:@code{vararg}}). You can supply a default value for any
5582 macro argument by following the name with @samp{=@var{deflt}}. You
5583 cannot define two macros with the same @var{macname} unless it has been
5584 subject to the @code{.purgem} directive (@pxref{Purgem}) between the two
5585 definitions. For example, these are all valid @code{.macro} statements:
5589 Begin the definition of a macro called @code{comm}, which takes no
5592 @item .macro plus1 p, p1
5593 @itemx .macro plus1 p p1
5594 Either statement begins the definition of a macro called @code{plus1},
5595 which takes two arguments; within the macro definition, write
5596 @samp{\p} or @samp{\p1} to evaluate the arguments.
5598 @item .macro reserve_str p1=0 p2
5599 Begin the definition of a macro called @code{reserve_str}, with two
5600 arguments. The first argument has a default value, but not the second.
5601 After the definition is complete, you can call the macro either as
5602 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
5603 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
5604 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
5605 @samp{0}, and @samp{\p2} evaluating to @var{b}).
5607 @item .macro m p1:req, p2=0, p3:vararg
5608 Begin the definition of a macro called @code{m}, with at least three
5609 arguments. The first argument must always have a value specified, but
5610 not the second, which instead has a default value. The third formal
5611 will get assigned all remaining arguments specified at invocation time.
5613 When you call a macro, you can specify the argument values either by
5614 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
5615 @samp{sum to=17, from=9}.
5619 Note that since each of the @var{macargs} can be an identifier exactly
5620 as any other one permitted by the target architecture, there may be
5621 occasional problems if the target hand-crafts special meanings to certain
5622 characters when they occur in a special position. For example, if the colon
5623 (@code{:}) is generally permitted to be part of a symbol name, but the
5624 architecture specific code special-cases it when occurring as the final
5625 character of a symbol (to denote a label), then the macro parameter
5626 replacement code will have no way of knowing that and consider the whole
5627 construct (including the colon) an identifier, and check only this
5628 identifier for being the subject to parameter substitution. So for example
5629 this macro definition:
5637 might not work as expected. Invoking @samp{label foo} might not create a label
5638 called @samp{foo} but instead just insert the text @samp{\l:} into the
5639 assembler source, probably generating an error about an unrecognised
5642 Similarly problems might occur with the period character (@samp{.})
5643 which is often allowed inside opcode names (and hence identifier names). So
5644 for example constructing a macro to build an opcode from a base name and a
5645 length specifier like this:
5648 .macro opcode base length
5653 and invoking it as @samp{opcode store l} will not create a @samp{store.l}
5654 instruction but instead generate some kind of error as the assembler tries to
5655 interpret the text @samp{\base.\length}.
5657 There are several possible ways around this problem:
5660 @item Insert white space
5661 If it is possible to use white space characters then this is the simplest
5670 @item Use @samp{\()}
5671 The string @samp{\()} can be used to separate the end of a macro argument from
5672 the following text. eg:
5675 .macro opcode base length
5680 @item Use the alternate macro syntax mode
5681 In the alternative macro syntax mode the ampersand character (@samp{&}) can be
5682 used as a separator. eg:
5692 Note: this problem of correctly identifying string parameters to pseudo ops
5693 also applies to the identifiers used in @code{.irp} (@pxref{Irp})
5694 and @code{.irpc} (@pxref{Irpc}) as well.
5697 @cindex @code{endm} directive
5698 Mark the end of a macro definition.
5701 @cindex @code{exitm} directive
5702 Exit early from the current macro definition.
5704 @cindex number of macros executed
5705 @cindex macros, count executed
5707 @command{@value{AS}} maintains a counter of how many macros it has
5708 executed in this pseudo-variable; you can copy that number to your
5709 output with @samp{\@@}, but @emph{only within a macro definition}.
5711 @item LOCAL @var{name} [ , @dots{} ]
5712 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
5713 macro syntax'' with @samp{--alternate} or @code{.altmacro}.}
5714 @xref{Altmacro,,@code{.altmacro}}.
5718 @section @code{.mri @var{val}}
5720 @cindex @code{mri} directive
5721 @cindex MRI mode, temporarily
5722 If @var{val} is non-zero, this tells @command{@value{AS}} to enter MRI mode. If
5723 @var{val} is zero, this tells @command{@value{AS}} to exit MRI mode. This change
5724 affects code assembled until the next @code{.mri} directive, or until the end
5725 of the file. @xref{M, MRI mode, MRI mode}.
5728 @section @code{.noaltmacro}
5729 Disable alternate macro mode. @xref{Altmacro}.
5732 @section @code{.nolist}
5734 @cindex @code{nolist} directive
5735 @cindex listing control, turning off
5736 Control (in conjunction with the @code{.list} directive) whether or
5737 not assembly listings are generated. These two directives maintain an
5738 internal counter (which is zero initially). @code{.list} increments the
5739 counter, and @code{.nolist} decrements it. Assembly listings are
5740 generated whenever the counter is greater than zero.
5743 @section @code{.octa @var{bignums}}
5745 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
5746 @cindex @code{octa} directive
5747 @cindex integer, 16-byte
5748 @cindex sixteen byte integer
5749 This directive expects zero or more bignums, separated by commas. For each
5750 bignum, it emits a 16-byte integer.
5752 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
5753 hence @emph{octa}-word for 16 bytes.
5756 @section @code{.offset @var{loc}}
5758 @cindex @code{offset} directive
5759 Set the location counter to @var{loc} in the absolute section. @var{loc} must
5760 be an absolute expression. This directive may be useful for defining
5761 symbols with absolute values. Do not confuse it with the @code{.org}
5765 @section @code{.org @var{new-lc} , @var{fill}}
5767 @cindex @code{org} directive
5768 @cindex location counter, advancing
5769 @cindex advancing location counter
5770 @cindex current address, advancing
5771 Advance the location counter of the current section to
5772 @var{new-lc}. @var{new-lc} is either an absolute expression or an
5773 expression with the same section as the current subsection. That is,
5774 you can't use @code{.org} to cross sections: if @var{new-lc} has the
5775 wrong section, the @code{.org} directive is ignored. To be compatible
5776 with former assemblers, if the section of @var{new-lc} is absolute,
5777 @command{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
5778 is the same as the current subsection.
5780 @code{.org} may only increase the location counter, or leave it
5781 unchanged; you cannot use @code{.org} to move the location counter
5784 @c double negative used below "not undefined" because this is a specific
5785 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
5786 @c section. doc@cygnus.com 18feb91
5787 Because @command{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
5788 may not be undefined. If you really detest this restriction we eagerly await
5789 a chance to share your improved assembler.
5791 Beware that the origin is relative to the start of the section, not
5792 to the start of the subsection. This is compatible with other
5793 people's assemblers.
5795 When the location counter (of the current subsection) is advanced, the
5796 intervening bytes are filled with @var{fill} which should be an
5797 absolute expression. If the comma and @var{fill} are omitted,
5798 @var{fill} defaults to zero.
5801 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
5803 @cindex padding the location counter given a power of two
5804 @cindex @code{p2align} directive
5805 Pad the location counter (in the current subsection) to a particular
5806 storage boundary. The first expression (which must be absolute) is the
5807 number of low-order zero bits the location counter must have after
5808 advancement. For example @samp{.p2align 3} advances the location
5809 counter until it a multiple of 8. If the location counter is already a
5810 multiple of 8, no change is needed.
5812 The second expression (also absolute) gives the fill value to be stored in the
5813 padding bytes. It (and the comma) may be omitted. If it is omitted, the
5814 padding bytes are normally zero. However, on some systems, if the section is
5815 marked as containing code and the fill value is omitted, the space is filled
5816 with no-op instructions.
5818 The third expression is also absolute, and is also optional. If it is present,
5819 it is the maximum number of bytes that should be skipped by this alignment
5820 directive. If doing the alignment would require skipping more bytes than the
5821 specified maximum, then the alignment is not done at all. You can omit the
5822 fill value (the second argument) entirely by simply using two commas after the
5823 required alignment; this can be useful if you want the alignment to be filled
5824 with no-op instructions when appropriate.
5826 @cindex @code{p2alignw} directive
5827 @cindex @code{p2alignl} directive
5828 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
5829 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
5830 pattern as a two byte word value. The @code{.p2alignl} directives treats the
5831 fill pattern as a four byte longword value. For example, @code{.p2alignw
5832 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
5833 filled in with the value 0x368d (the exact placement of the bytes depends upon
5834 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
5839 @section @code{.popsection}
5841 @cindex @code{popsection} directive
5842 @cindex Section Stack
5843 This is one of the ELF section stack manipulation directives. The others are
5844 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5845 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
5848 This directive replaces the current section (and subsection) with the top
5849 section (and subsection) on the section stack. This section is popped off the
5855 @section @code{.previous}
5857 @cindex @code{previous} directive
5858 @cindex Section Stack
5859 This is one of the ELF section stack manipulation directives. The others are
5860 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5861 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
5862 (@pxref{PopSection}).
5864 This directive swaps the current section (and subsection) with most recently
5865 referenced section/subsection pair prior to this one. Multiple
5866 @code{.previous} directives in a row will flip between two sections (and their
5867 subsections). For example:
5879 Will place 0x1234 and 0x9abc into subsection 1 and 0x5678 into subsection 2 of
5885 # Now in section A subsection 1
5889 # Now in section B subsection 0
5892 # Now in section B subsection 1
5895 # Now in section B subsection 0
5899 Will place 0x1234 into section A, 0x5678 and 0xdef0 into subsection 0 of
5900 section B and 0x9abc into subsection 1 of section B.
5902 In terms of the section stack, this directive swaps the current section with
5903 the top section on the section stack.
5907 @section @code{.print @var{string}}
5909 @cindex @code{print} directive
5910 @command{@value{AS}} will print @var{string} on the standard output during
5911 assembly. You must put @var{string} in double quotes.
5915 @section @code{.protected @var{names}}
5917 @cindex @code{protected} directive
5919 This is one of the ELF visibility directives. The other two are
5920 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
5922 This directive overrides the named symbols default visibility (which is set by
5923 their binding: local, global or weak). The directive sets the visibility to
5924 @code{protected} which means that any references to the symbols from within the
5925 components that defines them must be resolved to the definition in that
5926 component, even if a definition in another component would normally preempt
5931 @section @code{.psize @var{lines} , @var{columns}}
5933 @cindex @code{psize} directive
5934 @cindex listing control: paper size
5935 @cindex paper size, for listings
5936 Use this directive to declare the number of lines---and, optionally, the
5937 number of columns---to use for each page, when generating listings.
5939 If you do not use @code{.psize}, listings use a default line-count
5940 of 60. You may omit the comma and @var{columns} specification; the
5941 default width is 200 columns.
5943 @command{@value{AS}} generates formfeeds whenever the specified number of
5944 lines is exceeded (or whenever you explicitly request one, using
5947 If you specify @var{lines} as @code{0}, no formfeeds are generated save
5948 those explicitly specified with @code{.eject}.
5951 @section @code{.purgem @var{name}}
5953 @cindex @code{purgem} directive
5954 Undefine the macro @var{name}, so that later uses of the string will not be
5955 expanded. @xref{Macro}.
5959 @section @code{.pushsection @var{name} [, @var{subsection}] [, "@var{flags}"[, @@@var{type}[,@var{arguments}]]]}
5961 @cindex @code{pushsection} directive
5962 @cindex Section Stack
5963 This is one of the ELF section stack manipulation directives. The others are
5964 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
5965 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5968 This directive pushes the current section (and subsection) onto the
5969 top of the section stack, and then replaces the current section and
5970 subsection with @code{name} and @code{subsection}. The optional
5971 @code{flags}, @code{type} and @code{arguments} are treated the same
5972 as in the @code{.section} (@pxref{Section}) directive.
5976 @section @code{.quad @var{bignums}}
5978 @cindex @code{quad} directive
5979 @code{.quad} expects zero or more bignums, separated by commas. For
5980 each bignum, it emits
5982 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
5983 warning message; and just takes the lowest order 8 bytes of the bignum.
5984 @cindex eight-byte integer
5985 @cindex integer, 8-byte
5987 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
5988 hence @emph{quad}-word for 8 bytes.
5991 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
5992 warning message; and just takes the lowest order 16 bytes of the bignum.
5993 @cindex sixteen-byte integer
5994 @cindex integer, 16-byte
5998 @section @code{.reloc @var{offset}, @var{reloc_name}[, @var{expression}]}
6000 @cindex @code{reloc} directive
6001 Generate a relocation at @var{offset} of type @var{reloc_name} with value
6002 @var{expression}. If @var{offset} is a number, the relocation is generated in
6003 the current section. If @var{offset} is an expression that resolves to a
6004 symbol plus offset, the relocation is generated in the given symbol's section.
6005 @var{expression}, if present, must resolve to a symbol plus addend or to an
6006 absolute value, but note that not all targets support an addend. e.g. ELF REL
6007 targets such as i386 store an addend in the section contents rather than in the
6008 relocation. This low level interface does not support addends stored in the
6012 @section @code{.rept @var{count}}
6014 @cindex @code{rept} directive
6015 Repeat the sequence of lines between the @code{.rept} directive and the next
6016 @code{.endr} directive @var{count} times.
6018 For example, assembling
6026 is equivalent to assembling
6035 @section @code{.sbttl "@var{subheading}"}
6037 @cindex @code{sbttl} directive
6038 @cindex subtitles for listings
6039 @cindex listing control: subtitle
6040 Use @var{subheading} as the title (third line, immediately after the
6041 title line) when generating assembly listings.
6043 This directive affects subsequent pages, as well as the current page if
6044 it appears within ten lines of the top of a page.
6048 @section @code{.scl @var{class}}
6050 @cindex @code{scl} directive
6051 @cindex symbol storage class (COFF)
6052 @cindex COFF symbol storage class
6053 Set the storage-class value for a symbol. This directive may only be
6054 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
6055 whether a symbol is static or external, or it may record further
6056 symbolic debugging information.
6059 The @samp{.scl} directive is primarily associated with COFF output; when
6060 configured to generate @code{b.out} output format, @command{@value{AS}}
6061 accepts this directive but ignores it.
6067 @section @code{.section @var{name}}
6069 @cindex named section
6070 Use the @code{.section} directive to assemble the following code into a section
6073 This directive is only supported for targets that actually support arbitrarily
6074 named sections; on @code{a.out} targets, for example, it is not accepted, even
6075 with a standard @code{a.out} section name.
6079 @c only print the extra heading if both COFF and ELF are set
6080 @subheading COFF Version
6083 @cindex @code{section} directive (COFF version)
6084 For COFF targets, the @code{.section} directive is used in one of the following
6088 .section @var{name}[, "@var{flags}"]
6089 .section @var{name}[, @var{subsection}]
6092 If the optional argument is quoted, it is taken as flags to use for the
6093 section. Each flag is a single character. The following flags are recognized:
6096 bss section (uninitialized data)
6098 section is not loaded
6104 exclude section from linking
6110 shared section (meaningful for PE targets)
6112 ignored. (For compatibility with the ELF version)
6114 section is not readable (meaningful for PE targets)
6116 single-digit power-of-two section alignment (GNU extension)
6119 If no flags are specified, the default flags depend upon the section name. If
6120 the section name is not recognized, the default will be for the section to be
6121 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
6122 from the section, rather than adding them, so if they are used on their own it
6123 will be as if no flags had been specified at all.
6125 If the optional argument to the @code{.section} directive is not quoted, it is
6126 taken as a subsection number (@pxref{Sub-Sections}).
6131 @c only print the extra heading if both COFF and ELF are set
6132 @subheading ELF Version
6135 @cindex Section Stack
6136 This is one of the ELF section stack manipulation directives. The others are
6137 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
6138 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
6139 @code{.previous} (@pxref{Previous}).
6141 @cindex @code{section} directive (ELF version)
6142 For ELF targets, the @code{.section} directive is used like this:
6145 .section @var{name} [, "@var{flags}"[, @@@var{type}[,@var{flag_specific_arguments}]]]
6148 The optional @var{flags} argument is a quoted string which may contain any
6149 combination of the following characters:
6152 section is allocatable
6154 section is excluded from executable and shared library.
6158 section is executable
6160 section is mergeable
6162 section contains zero terminated strings
6164 section is a member of a section group
6166 section is used for thread-local-storage
6168 section is a member of the previously-current section's group, if any
6171 The optional @var{type} argument may contain one of the following constants:
6174 section contains data
6176 section does not contain data (i.e., section only occupies space)
6178 section contains data which is used by things other than the program
6180 section contains an array of pointers to init functions
6182 section contains an array of pointers to finish functions
6183 @item @@preinit_array
6184 section contains an array of pointers to pre-init functions
6187 Many targets only support the first three section types.
6189 Note on targets where the @code{@@} character is the start of a comment (eg
6190 ARM) then another character is used instead. For example the ARM port uses the
6193 If @var{flags} contains the @code{M} symbol then the @var{type} argument must
6194 be specified as well as an extra argument---@var{entsize}---like this:
6197 .section @var{name} , "@var{flags}"M, @@@var{type}, @var{entsize}
6200 Sections with the @code{M} flag but not @code{S} flag must contain fixed size
6201 constants, each @var{entsize} octets long. Sections with both @code{M} and
6202 @code{S} must contain zero terminated strings where each character is
6203 @var{entsize} bytes long. The linker may remove duplicates within sections with
6204 the same name, same entity size and same flags. @var{entsize} must be an
6205 absolute expression. For sections with both @code{M} and @code{S}, a string
6206 which is a suffix of a larger string is considered a duplicate. Thus
6207 @code{"def"} will be merged with @code{"abcdef"}; A reference to the first
6208 @code{"def"} will be changed to a reference to @code{"abcdef"+3}.
6210 If @var{flags} contains the @code{G} symbol then the @var{type} argument must
6211 be present along with an additional field like this:
6214 .section @var{name} , "@var{flags}"G, @@@var{type}, @var{GroupName}[, @var{linkage}]
6217 The @var{GroupName} field specifies the name of the section group to which this
6218 particular section belongs. The optional linkage field can contain:
6221 indicates that only one copy of this section should be retained
6226 Note: if both the @var{M} and @var{G} flags are present then the fields for
6227 the Merge flag should come first, like this:
6230 .section @var{name} , "@var{flags}"MG, @@@var{type}, @var{entsize}, @var{GroupName}[, @var{linkage}]
6233 If @var{flags} contains the @code{?} symbol then it may not also contain the
6234 @code{G} symbol and the @var{GroupName} or @var{linkage} fields should not be
6235 present. Instead, @code{?} says to consider the section that's current before
6236 this directive. If that section used @code{G}, then the new section will use
6237 @code{G} with those same @var{GroupName} and @var{linkage} fields implicitly.
6238 If not, then the @code{?} symbol has no effect.
6240 If no flags are specified, the default flags depend upon the section name. If
6241 the section name is not recognized, the default will be for the section to have
6242 none of the above flags: it will not be allocated in memory, nor writable, nor
6243 executable. The section will contain data.
6245 For ELF targets, the assembler supports another type of @code{.section}
6246 directive for compatibility with the Solaris assembler:
6249 .section "@var{name}"[, @var{flags}...]
6252 Note that the section name is quoted. There may be a sequence of comma
6256 section is allocatable
6260 section is executable
6262 section is excluded from executable and shared library.
6264 section is used for thread local storage
6267 This directive replaces the current section and subsection. See the
6268 contents of the gas testsuite directory @code{gas/testsuite/gas/elf} for
6269 some examples of how this directive and the other section stack directives
6275 @section @code{.set @var{symbol}, @var{expression}}
6277 @cindex @code{set} directive
6278 @cindex symbol value, setting
6279 Set the value of @var{symbol} to @var{expression}. This
6280 changes @var{symbol}'s value and type to conform to
6281 @var{expression}. If @var{symbol} was flagged as external, it remains
6282 flagged (@pxref{Symbol Attributes}).
6284 You may @code{.set} a symbol many times in the same assembly.
6286 If you @code{.set} a global symbol, the value stored in the object
6287 file is the last value stored into it.
6290 On Z80 @code{set} is a real instruction, use
6291 @samp{@var{symbol} defl @var{expression}} instead.
6295 @section @code{.short @var{expressions}}
6297 @cindex @code{short} directive
6299 @code{.short} is normally the same as @samp{.word}.
6300 @xref{Word,,@code{.word}}.
6302 In some configurations, however, @code{.short} and @code{.word} generate
6303 numbers of different lengths. @xref{Machine Dependencies}.
6307 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
6310 This expects zero or more @var{expressions}, and emits
6311 a 16 bit number for each.
6316 @section @code{.single @var{flonums}}
6318 @cindex @code{single} directive
6319 @cindex floating point numbers (single)
6320 This directive assembles zero or more flonums, separated by commas. It
6321 has the same effect as @code{.float}.
6323 The exact kind of floating point numbers emitted depends on how
6324 @command{@value{AS}} is configured. @xref{Machine Dependencies}.
6328 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
6329 numbers in @sc{ieee} format.
6335 @section @code{.size}
6337 This directive is used to set the size associated with a symbol.
6341 @c only print the extra heading if both COFF and ELF are set
6342 @subheading COFF Version
6345 @cindex @code{size} directive (COFF version)
6346 For COFF targets, the @code{.size} directive is only permitted inside
6347 @code{.def}/@code{.endef} pairs. It is used like this:
6350 .size @var{expression}
6354 @samp{.size} is only meaningful when generating COFF format output; when
6355 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6362 @c only print the extra heading if both COFF and ELF are set
6363 @subheading ELF Version
6366 @cindex @code{size} directive (ELF version)
6367 For ELF targets, the @code{.size} directive is used like this:
6370 .size @var{name} , @var{expression}
6373 This directive sets the size associated with a symbol @var{name}.
6374 The size in bytes is computed from @var{expression} which can make use of label
6375 arithmetic. This directive is typically used to set the size of function
6380 @ifclear no-space-dir
6382 @section @code{.skip @var{size} , @var{fill}}
6384 @cindex @code{skip} directive
6385 @cindex filling memory
6386 This directive emits @var{size} bytes, each of value @var{fill}. Both
6387 @var{size} and @var{fill} are absolute expressions. If the comma and
6388 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
6393 @section @code{.sleb128 @var{expressions}}
6395 @cindex @code{sleb128} directive
6396 @var{sleb128} stands for ``signed little endian base 128.'' This is a
6397 compact, variable length representation of numbers used by the DWARF
6398 symbolic debugging format. @xref{Uleb128, ,@code{.uleb128}}.
6400 @ifclear no-space-dir
6402 @section @code{.space @var{size} , @var{fill}}
6404 @cindex @code{space} directive
6405 @cindex filling memory
6406 This directive emits @var{size} bytes, each of value @var{fill}. Both
6407 @var{size} and @var{fill} are absolute expressions. If the comma
6408 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
6413 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
6414 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
6415 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
6416 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
6424 @section @code{.stabd, .stabn, .stabs}
6426 @cindex symbolic debuggers, information for
6427 @cindex @code{stab@var{x}} directives
6428 There are three directives that begin @samp{.stab}.
6429 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
6430 The symbols are not entered in the @command{@value{AS}} hash table: they
6431 cannot be referenced elsewhere in the source file.
6432 Up to five fields are required:
6436 This is the symbol's name. It may contain any character except
6437 @samp{\000}, so is more general than ordinary symbol names. Some
6438 debuggers used to code arbitrarily complex structures into symbol names
6442 An absolute expression. The symbol's type is set to the low 8 bits of
6443 this expression. Any bit pattern is permitted, but @code{@value{LD}}
6444 and debuggers choke on silly bit patterns.
6447 An absolute expression. The symbol's ``other'' attribute is set to the
6448 low 8 bits of this expression.
6451 An absolute expression. The symbol's descriptor is set to the low 16
6452 bits of this expression.
6455 An absolute expression which becomes the symbol's value.
6458 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
6459 or @code{.stabs} statement, the symbol has probably already been created;
6460 you get a half-formed symbol in your object file. This is
6461 compatible with earlier assemblers!
6464 @cindex @code{stabd} directive
6465 @item .stabd @var{type} , @var{other} , @var{desc}
6467 The ``name'' of the symbol generated is not even an empty string.
6468 It is a null pointer, for compatibility. Older assemblers used a
6469 null pointer so they didn't waste space in object files with empty
6472 The symbol's value is set to the location counter,
6473 relocatably. When your program is linked, the value of this symbol
6474 is the address of the location counter when the @code{.stabd} was
6477 @cindex @code{stabn} directive
6478 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
6479 The name of the symbol is set to the empty string @code{""}.
6481 @cindex @code{stabs} directive
6482 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
6483 All five fields are specified.
6489 @section @code{.string} "@var{str}", @code{.string8} "@var{str}", @code{.string16}
6490 "@var{str}", @code{.string32} "@var{str}", @code{.string64} "@var{str}"
6492 @cindex string, copying to object file
6493 @cindex string8, copying to object file
6494 @cindex string16, copying to object file
6495 @cindex string32, copying to object file
6496 @cindex string64, copying to object file
6497 @cindex @code{string} directive
6498 @cindex @code{string8} directive
6499 @cindex @code{string16} directive
6500 @cindex @code{string32} directive
6501 @cindex @code{string64} directive
6503 Copy the characters in @var{str} to the object file. You may specify more than
6504 one string to copy, separated by commas. Unless otherwise specified for a
6505 particular machine, the assembler marks the end of each string with a 0 byte.
6506 You can use any of the escape sequences described in @ref{Strings,,Strings}.
6508 The variants @code{string16}, @code{string32} and @code{string64} differ from
6509 the @code{string} pseudo opcode in that each 8-bit character from @var{str} is
6510 copied and expanded to 16, 32 or 64 bits respectively. The expanded characters
6511 are stored in target endianness byte order.
6517 .string "B\0\0\0Y\0\0\0E\0\0\0" /* On little endian targets. */
6518 .string "\0\0\0B\0\0\0Y\0\0\0E" /* On big endian targets. */
6523 @section @code{.struct @var{expression}}
6525 @cindex @code{struct} directive
6526 Switch to the absolute section, and set the section offset to @var{expression},
6527 which must be an absolute expression. You might use this as follows:
6536 This would define the symbol @code{field1} to have the value 0, the symbol
6537 @code{field2} to have the value 4, and the symbol @code{field3} to have the
6538 value 8. Assembly would be left in the absolute section, and you would need to
6539 use a @code{.section} directive of some sort to change to some other section
6540 before further assembly.
6544 @section @code{.subsection @var{name}}
6546 @cindex @code{subsection} directive
6547 @cindex Section Stack
6548 This is one of the ELF section stack manipulation directives. The others are
6549 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
6550 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
6553 This directive replaces the current subsection with @code{name}. The current
6554 section is not changed. The replaced subsection is put onto the section stack
6555 in place of the then current top of stack subsection.
6560 @section @code{.symver}
6561 @cindex @code{symver} directive
6562 @cindex symbol versioning
6563 @cindex versions of symbols
6564 Use the @code{.symver} directive to bind symbols to specific version nodes
6565 within a source file. This is only supported on ELF platforms, and is
6566 typically used when assembling files to be linked into a shared library.
6567 There are cases where it may make sense to use this in objects to be bound
6568 into an application itself so as to override a versioned symbol from a
6571 For ELF targets, the @code{.symver} directive can be used like this:
6573 .symver @var{name}, @var{name2@@nodename}
6575 If the symbol @var{name} is defined within the file
6576 being assembled, the @code{.symver} directive effectively creates a symbol
6577 alias with the name @var{name2@@nodename}, and in fact the main reason that we
6578 just don't try and create a regular alias is that the @var{@@} character isn't
6579 permitted in symbol names. The @var{name2} part of the name is the actual name
6580 of the symbol by which it will be externally referenced. The name @var{name}
6581 itself is merely a name of convenience that is used so that it is possible to
6582 have definitions for multiple versions of a function within a single source
6583 file, and so that the compiler can unambiguously know which version of a
6584 function is being mentioned. The @var{nodename} portion of the alias should be
6585 the name of a node specified in the version script supplied to the linker when
6586 building a shared library. If you are attempting to override a versioned
6587 symbol from a shared library, then @var{nodename} should correspond to the
6588 nodename of the symbol you are trying to override.
6590 If the symbol @var{name} is not defined within the file being assembled, all
6591 references to @var{name} will be changed to @var{name2@@nodename}. If no
6592 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
6595 Another usage of the @code{.symver} directive is:
6597 .symver @var{name}, @var{name2@@@@nodename}
6599 In this case, the symbol @var{name} must exist and be defined within
6600 the file being assembled. It is similar to @var{name2@@nodename}. The
6601 difference is @var{name2@@@@nodename} will also be used to resolve
6602 references to @var{name2} by the linker.
6604 The third usage of the @code{.symver} directive is:
6606 .symver @var{name}, @var{name2@@@@@@nodename}
6608 When @var{name} is not defined within the
6609 file being assembled, it is treated as @var{name2@@nodename}. When
6610 @var{name} is defined within the file being assembled, the symbol
6611 name, @var{name}, will be changed to @var{name2@@@@nodename}.
6616 @section @code{.tag @var{structname}}
6618 @cindex COFF structure debugging
6619 @cindex structure debugging, COFF
6620 @cindex @code{tag} directive
6621 This directive is generated by compilers to include auxiliary debugging
6622 information in the symbol table. It is only permitted inside
6623 @code{.def}/@code{.endef} pairs. Tags are used to link structure
6624 definitions in the symbol table with instances of those structures.
6627 @samp{.tag} is only used when generating COFF format output; when
6628 @command{@value{AS}} is generating @code{b.out}, it accepts this directive but
6634 @section @code{.text @var{subsection}}
6636 @cindex @code{text} directive
6637 Tells @command{@value{AS}} to assemble the following statements onto the end of
6638 the text subsection numbered @var{subsection}, which is an absolute
6639 expression. If @var{subsection} is omitted, subsection number zero
6643 @section @code{.title "@var{heading}"}
6645 @cindex @code{title} directive
6646 @cindex listing control: title line
6647 Use @var{heading} as the title (second line, immediately after the
6648 source file name and pagenumber) when generating assembly listings.
6650 This directive affects subsequent pages, as well as the current page if
6651 it appears within ten lines of the top of a page.
6655 @section @code{.type}
6657 This directive is used to set the type of a symbol.
6661 @c only print the extra heading if both COFF and ELF are set
6662 @subheading COFF Version
6665 @cindex COFF symbol type
6666 @cindex symbol type, COFF
6667 @cindex @code{type} directive (COFF version)
6668 For COFF targets, this directive is permitted only within
6669 @code{.def}/@code{.endef} pairs. It is used like this:
6675 This records the integer @var{int} as the type attribute of a symbol table
6679 @samp{.type} is associated only with COFF format output; when
6680 @command{@value{AS}} is configured for @code{b.out} output, it accepts this
6681 directive but ignores it.
6687 @c only print the extra heading if both COFF and ELF are set
6688 @subheading ELF Version
6691 @cindex ELF symbol type
6692 @cindex symbol type, ELF
6693 @cindex @code{type} directive (ELF version)
6694 For ELF targets, the @code{.type} directive is used like this:
6697 .type @var{name} , @var{type description}
6700 This sets the type of symbol @var{name} to be either a
6701 function symbol or an object symbol. There are five different syntaxes
6702 supported for the @var{type description} field, in order to provide
6703 compatibility with various other assemblers.
6705 Because some of the characters used in these syntaxes (such as @samp{@@} and
6706 @samp{#}) are comment characters for some architectures, some of the syntaxes
6707 below do not work on all architectures. The first variant will be accepted by
6708 the GNU assembler on all architectures so that variant should be used for
6709 maximum portability, if you do not need to assemble your code with other
6712 The syntaxes supported are:
6715 .type <name> STT_<TYPE_IN_UPPER_CASE>
6716 .type <name>,#<type>
6717 .type <name>,@@<type>
6718 .type <name>,%<type>
6719 .type <name>,"<type>"
6722 The types supported are:
6727 Mark the symbol as being a function name.
6730 @itemx gnu_indirect_function
6731 Mark the symbol as an indirect function when evaluated during reloc
6732 processing. (This is only supported on assemblers targeting GNU systems).
6736 Mark the symbol as being a data object.
6740 Mark the symbol as being a thead-local data object.
6744 Mark the symbol as being a common data object.
6748 Does not mark the symbol in any way. It is supported just for completeness.
6750 @item gnu_unique_object
6751 Marks the symbol as being a globally unique data object. The dynamic linker
6752 will make sure that in the entire process there is just one symbol with this
6753 name and type in use. (This is only supported on assemblers targeting GNU
6758 Note: Some targets support extra types in addition to those listed above.
6764 @section @code{.uleb128 @var{expressions}}
6766 @cindex @code{uleb128} directive
6767 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
6768 compact, variable length representation of numbers used by the DWARF
6769 symbolic debugging format. @xref{Sleb128, ,@code{.sleb128}}.
6773 @section @code{.val @var{addr}}
6775 @cindex @code{val} directive
6776 @cindex COFF value attribute
6777 @cindex value attribute, COFF
6778 This directive, permitted only within @code{.def}/@code{.endef} pairs,
6779 records the address @var{addr} as the value attribute of a symbol table
6783 @samp{.val} is used only for COFF output; when @command{@value{AS}} is
6784 configured for @code{b.out}, it accepts this directive but ignores it.
6790 @section @code{.version "@var{string}"}
6792 @cindex @code{version} directive
6793 This directive creates a @code{.note} section and places into it an ELF
6794 formatted note of type NT_VERSION. The note's name is set to @code{string}.
6799 @section @code{.vtable_entry @var{table}, @var{offset}}
6801 @cindex @code{vtable_entry} directive
6802 This directive finds or creates a symbol @code{table} and creates a
6803 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
6806 @section @code{.vtable_inherit @var{child}, @var{parent}}
6808 @cindex @code{vtable_inherit} directive
6809 This directive finds the symbol @code{child} and finds or creates the symbol
6810 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
6811 parent whose addend is the value of the child symbol. As a special case the
6812 parent name of @code{0} is treated as referring to the @code{*ABS*} section.
6816 @section @code{.warning "@var{string}"}
6817 @cindex warning directive
6818 Similar to the directive @code{.error}
6819 (@pxref{Error,,@code{.error "@var{string}"}}), but just emits a warning.
6822 @section @code{.weak @var{names}}
6824 @cindex @code{weak} directive
6825 This directive sets the weak attribute on the comma separated list of symbol
6826 @code{names}. If the symbols do not already exist, they will be created.
6828 On COFF targets other than PE, weak symbols are a GNU extension. This
6829 directive sets the weak attribute on the comma separated list of symbol
6830 @code{names}. If the symbols do not already exist, they will be created.
6832 On the PE target, weak symbols are supported natively as weak aliases.
6833 When a weak symbol is created that is not an alias, GAS creates an
6834 alternate symbol to hold the default value.
6837 @section @code{.weakref @var{alias}, @var{target}}
6839 @cindex @code{weakref} directive
6840 This directive creates an alias to the target symbol that enables the symbol to
6841 be referenced with weak-symbol semantics, but without actually making it weak.
6842 If direct references or definitions of the symbol are present, then the symbol
6843 will not be weak, but if all references to it are through weak references, the
6844 symbol will be marked as weak in the symbol table.
6846 The effect is equivalent to moving all references to the alias to a separate
6847 assembly source file, renaming the alias to the symbol in it, declaring the
6848 symbol as weak there, and running a reloadable link to merge the object files
6849 resulting from the assembly of the new source file and the old source file that
6850 had the references to the alias removed.
6852 The alias itself never makes to the symbol table, and is entirely handled
6853 within the assembler.
6856 @section @code{.word @var{expressions}}
6858 @cindex @code{word} directive
6859 This directive expects zero or more @var{expressions}, of any section,
6860 separated by commas.
6863 For each expression, @command{@value{AS}} emits a 32-bit number.
6866 For each expression, @command{@value{AS}} emits a 16-bit number.
6871 The size of the number emitted, and its byte order,
6872 depend on what target computer the assembly is for.
6875 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
6876 @c happen---32-bit addressability, period; no long/short jumps.
6877 @ifset DIFF-TBL-KLUGE
6878 @cindex difference tables altered
6879 @cindex altered difference tables
6881 @emph{Warning: Special Treatment to support Compilers}
6885 Machines with a 32-bit address space, but that do less than 32-bit
6886 addressing, require the following special treatment. If the machine of
6887 interest to you does 32-bit addressing (or doesn't require it;
6888 @pxref{Machine Dependencies}), you can ignore this issue.
6891 In order to assemble compiler output into something that works,
6892 @command{@value{AS}} occasionally does strange things to @samp{.word} directives.
6893 Directives of the form @samp{.word sym1-sym2} are often emitted by
6894 compilers as part of jump tables. Therefore, when @command{@value{AS}} assembles a
6895 directive of the form @samp{.word sym1-sym2}, and the difference between
6896 @code{sym1} and @code{sym2} does not fit in 16 bits, @command{@value{AS}}
6897 creates a @dfn{secondary jump table}, immediately before the next label.
6898 This secondary jump table is preceded by a short-jump to the
6899 first byte after the secondary table. This short-jump prevents the flow
6900 of control from accidentally falling into the new table. Inside the
6901 table is a long-jump to @code{sym2}. The original @samp{.word}
6902 contains @code{sym1} minus the address of the long-jump to
6905 If there were several occurrences of @samp{.word sym1-sym2} before the
6906 secondary jump table, all of them are adjusted. If there was a
6907 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
6908 long-jump to @code{sym4} is included in the secondary jump table,
6909 and the @code{.word} directives are adjusted to contain @code{sym3}
6910 minus the address of the long-jump to @code{sym4}; and so on, for as many
6911 entries in the original jump table as necessary.
6914 @emph{This feature may be disabled by compiling @command{@value{AS}} with the
6915 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
6916 assembly language programmers.
6919 @c end DIFF-TBL-KLUGE
6922 @section Deprecated Directives
6924 @cindex deprecated directives
6925 @cindex obsolescent directives
6926 One day these directives won't work.
6927 They are included for compatibility with older assemblers.
6934 @node Object Attributes
6935 @chapter Object Attributes
6936 @cindex object attributes
6938 @command{@value{AS}} assembles source files written for a specific architecture
6939 into object files for that architecture. But not all object files are alike.
6940 Many architectures support incompatible variations. For instance, floating
6941 point arguments might be passed in floating point registers if the object file
6942 requires hardware floating point support---or floating point arguments might be
6943 passed in integer registers if the object file supports processors with no
6944 hardware floating point unit. Or, if two objects are built for different
6945 generations of the same architecture, the combination may require the
6946 newer generation at run-time.
6948 This information is useful during and after linking. At link time,
6949 @command{@value{LD}} can warn about incompatible object files. After link
6950 time, tools like @command{gdb} can use it to process the linked file
6953 Compatibility information is recorded as a series of object attributes. Each
6954 attribute has a @dfn{vendor}, @dfn{tag}, and @dfn{value}. The vendor is a
6955 string, and indicates who sets the meaning of the tag. The tag is an integer,
6956 and indicates what property the attribute describes. The value may be a string
6957 or an integer, and indicates how the property affects this object. Missing
6958 attributes are the same as attributes with a zero value or empty string value.
6960 Object attributes were developed as part of the ABI for the ARM Architecture.
6961 The file format is documented in @cite{ELF for the ARM Architecture}.
6964 * GNU Object Attributes:: @sc{gnu} Object Attributes
6965 * Defining New Object Attributes:: Defining New Object Attributes
6968 @node GNU Object Attributes
6969 @section @sc{gnu} Object Attributes
6971 The @code{.gnu_attribute} directive records an object attribute
6972 with vendor @samp{gnu}.
6974 Except for @samp{Tag_compatibility}, which has both an integer and a string for
6975 its value, @sc{gnu} attributes have a string value if the tag number is odd and
6976 an integer value if the tag number is even. The second bit (@code{@var{tag} &
6977 2} is set for architecture-independent attributes and clear for
6978 architecture-dependent ones.
6980 @subsection Common @sc{gnu} attributes
6982 These attributes are valid on all architectures.
6985 @item Tag_compatibility (32)
6986 The compatibility attribute takes an integer flag value and a vendor name. If
6987 the flag value is 0, the file is compatible with other toolchains. If it is 1,
6988 then the file is only compatible with the named toolchain. If it is greater
6989 than 1, the file can only be processed by other toolchains under some private
6990 arrangement indicated by the flag value and the vendor name.
6993 @subsection MIPS Attributes
6996 @item Tag_GNU_MIPS_ABI_FP (4)
6997 The floating-point ABI used by this object file. The value will be:
7001 0 for files not affected by the floating-point ABI.
7003 1 for files using the hardware floating-point with a standard double-precision
7006 2 for files using the hardware floating-point ABI with a single-precision FPU.
7008 3 for files using the software floating-point ABI.
7010 4 for files using the hardware floating-point ABI with 64-bit wide
7011 double-precision floating-point registers and 32-bit wide general
7016 @subsection PowerPC Attributes
7019 @item Tag_GNU_Power_ABI_FP (4)
7020 The floating-point ABI used by this object file. The value will be:
7024 0 for files not affected by the floating-point ABI.
7026 1 for files using double-precision hardware floating-point ABI.
7028 2 for files using the software floating-point ABI.
7030 3 for files using single-precision hardware floating-point ABI.
7033 @item Tag_GNU_Power_ABI_Vector (8)
7034 The vector ABI used by this object file. The value will be:
7038 0 for files not affected by the vector ABI.
7040 1 for files using general purpose registers to pass vectors.
7042 2 for files using AltiVec registers to pass vectors.
7044 3 for files using SPE registers to pass vectors.
7048 @node Defining New Object Attributes
7049 @section Defining New Object Attributes
7051 If you want to define a new @sc{gnu} object attribute, here are the places you
7052 will need to modify. New attributes should be discussed on the @samp{binutils}
7057 This manual, which is the official register of attributes.
7059 The header for your architecture @file{include/elf}, to define the tag.
7061 The @file{bfd} support file for your architecture, to merge the attribute
7062 and issue any appropriate link warnings.
7064 Test cases in @file{ld/testsuite} for merging and link warnings.
7066 @file{binutils/readelf.c} to display your attribute.
7068 GCC, if you want the compiler to mark the attribute automatically.
7074 @node Machine Dependencies
7075 @chapter Machine Dependent Features
7077 @cindex machine dependencies
7078 The machine instruction sets are (almost by definition) different on
7079 each machine where @command{@value{AS}} runs. Floating point representations
7080 vary as well, and @command{@value{AS}} often supports a few additional
7081 directives or command-line options for compatibility with other
7082 assemblers on a particular platform. Finally, some versions of
7083 @command{@value{AS}} support special pseudo-instructions for branch
7086 This chapter discusses most of these differences, though it does not
7087 include details on any machine's instruction set. For details on that
7088 subject, see the hardware manufacturer's manual.
7092 * AArch64-Dependent:: AArch64 Dependent Features
7095 * Alpha-Dependent:: Alpha Dependent Features
7098 * ARC-Dependent:: ARC Dependent Features
7101 * ARM-Dependent:: ARM Dependent Features
7104 * AVR-Dependent:: AVR Dependent Features
7107 * Blackfin-Dependent:: Blackfin Dependent Features
7110 * CR16-Dependent:: CR16 Dependent Features
7113 * CRIS-Dependent:: CRIS Dependent Features
7116 * D10V-Dependent:: D10V Dependent Features
7119 * D30V-Dependent:: D30V Dependent Features
7122 * Epiphany-Dependent:: EPIPHANY Dependent Features
7125 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7128 * HPPA-Dependent:: HPPA Dependent Features
7131 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
7134 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
7137 * i860-Dependent:: Intel 80860 Dependent Features
7140 * i960-Dependent:: Intel 80960 Dependent Features
7143 * IA-64-Dependent:: Intel IA-64 Dependent Features
7146 * IP2K-Dependent:: IP2K Dependent Features
7149 * LM32-Dependent:: LM32 Dependent Features
7152 * M32C-Dependent:: M32C Dependent Features
7155 * M32R-Dependent:: M32R Dependent Features
7158 * M68K-Dependent:: M680x0 Dependent Features
7161 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
7164 * Meta-Dependent :: Meta Dependent Features
7167 * MicroBlaze-Dependent:: MICROBLAZE Dependent Features
7170 * MIPS-Dependent:: MIPS Dependent Features
7173 * MMIX-Dependent:: MMIX Dependent Features
7176 * MSP430-Dependent:: MSP430 Dependent Features
7179 * NDS32-Dependent:: Andes NDS32 Dependent Features
7182 * NiosII-Dependent:: Altera Nios II Dependent Features
7185 * NS32K-Dependent:: NS32K Dependent Features
7188 * SH-Dependent:: Renesas / SuperH SH Dependent Features
7189 * SH64-Dependent:: SuperH SH64 Dependent Features
7192 * PDP-11-Dependent:: PDP-11 Dependent Features
7195 * PJ-Dependent:: picoJava Dependent Features
7198 * PPC-Dependent:: PowerPC Dependent Features
7201 * RL78-Dependent:: RL78 Dependent Features
7204 * RX-Dependent:: RX Dependent Features
7207 * S/390-Dependent:: IBM S/390 Dependent Features
7210 * SCORE-Dependent:: SCORE Dependent Features
7213 * Sparc-Dependent:: SPARC Dependent Features
7216 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
7219 * TIC6X-Dependent :: TI TMS320C6x Dependent Features
7222 * TILE-Gx-Dependent :: Tilera TILE-Gx Dependent Features
7225 * TILEPro-Dependent :: Tilera TILEPro Dependent Features
7228 * V850-Dependent:: V850 Dependent Features
7231 * XGATE-Dependent:: XGATE Features
7234 * XSTORMY16-Dependent:: XStormy16 Dependent Features
7237 * Xtensa-Dependent:: Xtensa Dependent Features
7240 * Z80-Dependent:: Z80 Dependent Features
7243 * Z8000-Dependent:: Z8000 Dependent Features
7246 * Vax-Dependent:: VAX Dependent Features
7253 @c The following major nodes are *sections* in the GENERIC version, *chapters*
7254 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
7255 @c peculiarity: to preserve cross-references, there must be a node called
7256 @c "Machine Dependencies". Hence the conditional nodenames in each
7257 @c major node below. Node defaulting in makeinfo requires adjacency of
7258 @c node and sectioning commands; hence the repetition of @chapter BLAH
7259 @c in both conditional blocks.
7262 @include c-aarch64.texi
7266 @include c-alpha.texi
7282 @include c-bfin.texi
7286 @include c-cr16.texi
7290 @include c-cris.texi
7295 @node Machine Dependencies
7296 @chapter Machine Dependent Features
7298 The machine instruction sets are different on each Renesas chip family,
7299 and there are also some syntax differences among the families. This
7300 chapter describes the specific @command{@value{AS}} features for each
7304 * H8/300-Dependent:: Renesas H8/300 Dependent Features
7305 * SH-Dependent:: Renesas SH Dependent Features
7312 @include c-d10v.texi
7316 @include c-d30v.texi
7320 @include c-epiphany.texi
7324 @include c-h8300.texi
7328 @include c-hppa.texi
7332 @include c-i370.texi
7336 @include c-i386.texi
7340 @include c-i860.texi
7344 @include c-i960.texi
7348 @include c-ia64.texi
7352 @include c-ip2k.texi
7356 @include c-lm32.texi
7360 @include c-m32c.texi
7364 @include c-m32r.texi
7368 @include c-m68k.texi
7372 @include c-m68hc11.texi
7376 @include c-metag.texi
7380 @include c-microblaze.texi
7384 @include c-mips.texi
7388 @include c-mmix.texi
7392 @include c-msp430.texi
7396 @include c-nds32.texi
7400 @include c-nios2.texi
7404 @include c-ns32k.texi
7408 @include c-pdp11.texi
7420 @include c-rl78.texi
7428 @include c-s390.texi
7432 @include c-score.texi
7437 @include c-sh64.texi
7441 @include c-sparc.texi
7445 @include c-tic54x.texi
7449 @include c-tic6x.texi
7453 @include c-tilegx.texi
7457 @include c-tilepro.texi
7473 @include c-v850.texi
7477 @include c-xgate.texi
7481 @include c-xstormy16.texi
7485 @include c-xtensa.texi
7489 @c reverse effect of @down at top of generic Machine-Dep chapter
7493 @node Reporting Bugs
7494 @chapter Reporting Bugs
7495 @cindex bugs in assembler
7496 @cindex reporting bugs in assembler
7498 Your bug reports play an essential role in making @command{@value{AS}} reliable.
7500 Reporting a bug may help you by bringing a solution to your problem, or it may
7501 not. But in any case the principal function of a bug report is to help the
7502 entire community by making the next version of @command{@value{AS}} work better.
7503 Bug reports are your contribution to the maintenance of @command{@value{AS}}.
7505 In order for a bug report to serve its purpose, you must include the
7506 information that enables us to fix the bug.
7509 * Bug Criteria:: Have you found a bug?
7510 * Bug Reporting:: How to report bugs
7514 @section Have You Found a Bug?
7515 @cindex bug criteria
7517 If you are not sure whether you have found a bug, here are some guidelines:
7520 @cindex fatal signal
7521 @cindex assembler crash
7522 @cindex crash of assembler
7524 If the assembler gets a fatal signal, for any input whatever, that is a
7525 @command{@value{AS}} bug. Reliable assemblers never crash.
7527 @cindex error on valid input
7529 If @command{@value{AS}} produces an error message for valid input, that is a bug.
7531 @cindex invalid input
7533 If @command{@value{AS}} does not produce an error message for invalid input, that
7534 is a bug. However, you should note that your idea of ``invalid input'' might
7535 be our idea of ``an extension'' or ``support for traditional practice''.
7538 If you are an experienced user of assemblers, your suggestions for improvement
7539 of @command{@value{AS}} are welcome in any case.
7543 @section How to Report Bugs
7545 @cindex assembler bugs, reporting
7547 A number of companies and individuals offer support for @sc{gnu} products. If
7548 you obtained @command{@value{AS}} from a support organization, we recommend you
7549 contact that organization first.
7551 You can find contact information for many support companies and
7552 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
7556 In any event, we also recommend that you send bug reports for @command{@value{AS}}
7560 The fundamental principle of reporting bugs usefully is this:
7561 @strong{report all the facts}. If you are not sure whether to state a
7562 fact or leave it out, state it!
7564 Often people omit facts because they think they know what causes the problem
7565 and assume that some details do not matter. Thus, you might assume that the
7566 name of a symbol you use in an example does not matter. Well, probably it does
7567 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
7568 happens to fetch from the location where that name is stored in memory;
7569 perhaps, if the name were different, the contents of that location would fool
7570 the assembler into doing the right thing despite the bug. Play it safe and
7571 give a specific, complete example. That is the easiest thing for you to do,
7572 and the most helpful.
7574 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
7575 it is new to us. Therefore, always write your bug reports on the assumption
7576 that the bug has not been reported previously.
7578 Sometimes people give a few sketchy facts and ask, ``Does this ring a
7579 bell?'' This cannot help us fix a bug, so it is basically useless. We
7580 respond by asking for enough details to enable us to investigate.
7581 You might as well expedite matters by sending them to begin with.
7583 To enable us to fix the bug, you should include all these things:
7587 The version of @command{@value{AS}}. @command{@value{AS}} announces it if you start
7588 it with the @samp{--version} argument.
7590 Without this, we will not know whether there is any point in looking for
7591 the bug in the current version of @command{@value{AS}}.
7594 Any patches you may have applied to the @command{@value{AS}} source.
7597 The type of machine you are using, and the operating system name and
7601 What compiler (and its version) was used to compile @command{@value{AS}}---e.g.
7605 The command arguments you gave the assembler to assemble your example and
7606 observe the bug. To guarantee you will not omit something important, list them
7607 all. A copy of the Makefile (or the output from make) is sufficient.
7609 If we were to try to guess the arguments, we would probably guess wrong
7610 and then we might not encounter the bug.
7613 A complete input file that will reproduce the bug. If the bug is observed when
7614 the assembler is invoked via a compiler, send the assembler source, not the
7615 high level language source. Most compilers will produce the assembler source
7616 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
7617 the options @samp{-v --save-temps}; this will save the assembler source in a
7618 file with an extension of @file{.s}, and also show you exactly how
7619 @command{@value{AS}} is being run.
7622 A description of what behavior you observe that you believe is
7623 incorrect. For example, ``It gets a fatal signal.''
7625 Of course, if the bug is that @command{@value{AS}} gets a fatal signal, then we
7626 will certainly notice it. But if the bug is incorrect output, we might not
7627 notice unless it is glaringly wrong. You might as well not give us a chance to
7630 Even if the problem you experience is a fatal signal, you should still say so
7631 explicitly. Suppose something strange is going on, such as, your copy of
7632 @command{@value{AS}} is out of sync, or you have encountered a bug in the C
7633 library on your system. (This has happened!) Your copy might crash and ours
7634 would not. If you told us to expect a crash, then when ours fails to crash, we
7635 would know that the bug was not happening for us. If you had not told us to
7636 expect a crash, then we would not be able to draw any conclusion from our
7640 If you wish to suggest changes to the @command{@value{AS}} source, send us context
7641 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
7642 option. Always send diffs from the old file to the new file. If you even
7643 discuss something in the @command{@value{AS}} source, refer to it by context, not
7646 The line numbers in our development sources will not match those in your
7647 sources. Your line numbers would convey no useful information to us.
7650 Here are some things that are not necessary:
7654 A description of the envelope of the bug.
7656 Often people who encounter a bug spend a lot of time investigating
7657 which changes to the input file will make the bug go away and which
7658 changes will not affect it.
7660 This is often time consuming and not very useful, because the way we
7661 will find the bug is by running a single example under the debugger
7662 with breakpoints, not by pure deduction from a series of examples.
7663 We recommend that you save your time for something else.
7665 Of course, if you can find a simpler example to report @emph{instead}
7666 of the original one, that is a convenience for us. Errors in the
7667 output will be easier to spot, running under the debugger will take
7668 less time, and so on.
7670 However, simplification is not vital; if you do not want to do this,
7671 report the bug anyway and send us the entire test case you used.
7674 A patch for the bug.
7676 A patch for the bug does help us if it is a good one. But do not omit
7677 the necessary information, such as the test case, on the assumption that
7678 a patch is all we need. We might see problems with your patch and decide
7679 to fix the problem another way, or we might not understand it at all.
7681 Sometimes with a program as complicated as @command{@value{AS}} it is very hard to
7682 construct an example that will make the program follow a certain path through
7683 the code. If you do not send us the example, we will not be able to construct
7684 one, so we will not be able to verify that the bug is fixed.
7686 And if we cannot understand what bug you are trying to fix, or why your
7687 patch should be an improvement, we will not install it. A test case will
7688 help us to understand.
7691 A guess about what the bug is or what it depends on.
7693 Such guesses are usually wrong. Even we cannot guess right about such
7694 things without first using the debugger to find the facts.
7697 @node Acknowledgements
7698 @chapter Acknowledgements
7700 If you have contributed to GAS and your name isn't listed here,
7701 it is not meant as a slight. We just don't know about it. Send mail to the
7702 maintainer, and we'll correct the situation. Currently
7704 the maintainer is Nick Clifton (email address @code{nickc@@redhat.com}).
7706 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
7709 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
7710 information and the 68k series machines, most of the preprocessing pass, and
7711 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
7713 K. Richard Pixley maintained GAS for a while, adding various enhancements and
7714 many bug fixes, including merging support for several processors, breaking GAS
7715 up to handle multiple object file format back ends (including heavy rewrite,
7716 testing, an integration of the coff and b.out back ends), adding configuration
7717 including heavy testing and verification of cross assemblers and file splits
7718 and renaming, converted GAS to strictly ANSI C including full prototypes, added
7719 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
7720 port (including considerable amounts of reverse engineering), a SPARC opcode
7721 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
7722 assertions and made them work, much other reorganization, cleanup, and lint.
7724 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
7725 in format-specific I/O modules.
7727 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
7728 has done much work with it since.
7730 The Intel 80386 machine description was written by Eliot Dresselhaus.
7732 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
7734 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
7735 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
7737 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
7738 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
7739 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
7740 support a.out format.
7742 Support for the Zilog Z8k and Renesas H8/300 processors (tc-z8k,
7743 tc-h8300), and IEEE 695 object file format (obj-ieee), was written by
7744 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
7745 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
7748 John Gilmore built the AMD 29000 support, added @code{.include} support, and
7749 simplified the configuration of which versions accept which directives. He
7750 updated the 68k machine description so that Motorola's opcodes always produced
7751 fixed-size instructions (e.g., @code{jsr}), while synthetic instructions
7752 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
7753 cross-compilation support, and one bug in relaxation that took a week and
7754 required the proverbial one-bit fix.
7756 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
7757 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
7758 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
7759 PowerPC assembler, and made a few other minor patches.
7761 Steve Chamberlain made GAS able to generate listings.
7763 Hewlett-Packard contributed support for the HP9000/300.
7765 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
7766 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
7767 formats). This work was supported by both the Center for Software Science at
7768 the University of Utah and Cygnus Support.
7770 Support for ELF format files has been worked on by Mark Eichin of Cygnus
7771 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
7772 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
7773 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
7774 and some initial 64-bit support).
7776 Linas Vepstas added GAS support for the ESA/390 ``IBM 370'' architecture.
7778 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
7779 support for openVMS/Alpha.
7781 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
7784 David Heine, Sterling Augustine, Bob Wilson and John Ruttenberg from Tensilica,
7785 Inc.@: added support for Xtensa processors.
7787 Several engineers at Cygnus Support have also provided many small bug fixes and
7788 configuration enhancements.
7790 Jon Beniston added support for the Lattice Mico32 architecture.
7792 Many others have contributed large or small bugfixes and enhancements. If
7793 you have contributed significant work and are not mentioned on this list, and
7794 want to be, let us know. Some of the history has been lost; we are not
7795 intentionally leaving anyone out.
7797 @node GNU Free Documentation License
7798 @appendix GNU Free Documentation License
7802 @unnumbered AS Index
projects / deliverable / binutils-gdb.git / blob