1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
4 @c Free Software Foundation, Inc.
5 @c UPDATE!! On future updates--
6 @c (1) check for new machine-dep cmdline options in
7 @c md_parse_option definitions in config/tc-*.c
8 @c (2) for platform-specific directives, examine md_pseudo_op
10 @c (3) for object-format specific directives, examine obj_pseudo_op
12 @c (4) portable directives in potable[] in read.c
16 @c defaults, config file may override:
19 @include asconfig.texi
24 @c Configure for the generation of man pages
54 @c common OR combinations of conditions
74 @set abnormal-separator
78 @settitle Using @value{AS}
81 @settitle Using @value{AS} (@value{TARGET})
83 @setchapternewpage odd
88 @c WARE! Some of the machine-dependent sections contain tables of machine
89 @c instructions. Except in multi-column format, these tables look silly.
90 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
91 @c the multi-col format is faked within @example sections.
93 @c Again unfortunately, the natural size that fits on a page, for these tables,
94 @c is different depending on whether or not smallbook is turned on.
95 @c This matters, because of order: text flow switches columns at each page
98 @c The format faked in this source works reasonably well for smallbook,
99 @c not well for the default large-page format. This manual expects that if you
100 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
101 @c tables in question. You can turn on one without the other at your
102 @c discretion, of course.
105 @c the insn tables look just as silly in info files regardless of smallbook,
106 @c might as well show 'em anyways.
112 * As: (as). The GNU assembler.
121 This file documents the GNU Assembler "@value{AS}".
123 @c man begin COPYRIGHT
124 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001 Free Software Foundation, Inc.
126 Permission is granted to copy, distribute and/or modify this document
127 under the terms of the GNU Free Documentation License, Version 1.1
128 or any later version published by the Free Software Foundation;
129 with no Invariant Sections, with no Front-Cover Texts, and with no
130 Back-Cover Texts. A copy of the license is included in the
131 section entitled "GNU Free Documentation License".
136 Permission is granted to process this file through Tex and print the
137 results, provided the printed document carries copying permission
138 notice identical to this one except for the removal of this paragraph
139 (this paragraph not being relevant to the printed manual).
145 @title Using @value{AS}
146 @subtitle The @sc{gnu} Assembler
148 @subtitle for the @value{TARGET} family
151 @subtitle Version @value{VERSION}
154 The Free Software Foundation Inc. thanks The Nice Computer
155 Company of Australia for loaning Dean Elsner to write the
156 first (Vax) version of @code{as} for Project @sc{gnu}.
157 The proprietors, management and staff of TNCCA thank FSF for
158 distracting the boss while they got some work
161 @author Dean Elsner, Jay Fenlason & friends
165 \hfill {\it Using {\tt @value{AS}}}\par
166 \hfill Edited by Cygnus Support\par
168 %"boxit" macro for figures:
169 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
170 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
171 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
172 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
173 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
176 @vskip 0pt plus 1filll
177 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 98, 99, 2000, 2001 Free Software Foundation, Inc.
179 Permission is granted to copy, distribute and/or modify this document
180 under the terms of the GNU Free Documentation License, Version 1.1
181 or any later version published by the Free Software Foundation;
182 with no Invariant Sections, with no Front-Cover Texts, and with no
183 Back-Cover Texts. A copy of the license is included in the
184 section entitled "GNU Free Documentation License".
190 @top Using @value{AS}
192 This file is a user guide to the @sc{gnu} assembler @code{@value{AS}} version
195 This version of the file describes @code{@value{AS}} configured to generate
196 code for @value{TARGET} architectures.
199 This document is distributed under the terms of the GNU Free
200 Documentation License. A copy of the license is included in the
201 section entitled "GNU Free Documentation License".
204 * Overview:: Overview
205 * Invoking:: Command-Line Options
207 * Sections:: Sections and Relocation
209 * Expressions:: Expressions
210 * Pseudo Ops:: Assembler Directives
211 * Machine Dependencies:: Machine Dependent Features
212 * Reporting Bugs:: Reporting Bugs
213 * Acknowledgements:: Who Did What
214 * GNU Free Documentation License:: GNU Free Documentation License
222 This manual is a user guide to the @sc{gnu} assembler @code{@value{AS}}.
224 This version of the manual describes @code{@value{AS}} configured to generate
225 code for @value{TARGET} architectures.
229 @cindex invocation summary
230 @cindex option summary
231 @cindex summary of options
232 Here is a brief summary of how to invoke @code{@value{AS}}. For details,
233 @pxref{Invoking,,Comand-Line Options}.
235 @c man title AS the portable GNU assembler.
238 gcc(1), ld(1), and the Info entries for @file{binutils} and @file{ld}.
241 @c We don't use deffn and friends for the following because they seem
242 @c to be limited to one line for the header.
244 @c man begin SYNOPSIS
245 @value{AS} [ -a[cdhlns][=file] ] [ -D ] [ --defsym @var{sym}=@var{val} ]
246 [ -f ] [ --gstabs ] [ --gdwarf2 ] [ --help ] [ -I @var{dir} ]
248 [ --listing--lhs-width=NUM ][ --listing-lhs-width2=NUM ]
249 [ --listing-rhs-width=NUM ][ --listing-cont-lines=NUM ]
250 [ --keep-locals ] [ -o @var{objfile} ] [ -R ] [ --statistics ] [ -v ]
251 [ -version ] [ --version ] [ -W ] [ --warn ] [ --fatal-warnings ]
252 [ -w ] [ -x ] [ -Z ] [ --target-help ]
254 @c am29k has no machine-dependent assembler options
261 [ -m[arm]1 | -m[arm]2 | -m[arm]250 | -m[arm]3 |
262 -m[arm]6 | -m[arm]60 | -m[arm]600 | -m[arm]610 |
263 -m[arm]620 | -m[arm]7[t][[d]m[i]][fe] | -m[arm]70 |
264 -m[arm]700 | -m[arm]710[c] | -m[arm]7100 |
265 -m[arm]7500 | -m[arm]8 | -m[arm]810 | -m[arm]9 |
266 -m[arm]920 | -m[arm]920t | -m[arm]9tdmi |
267 -mstrongarm | -mstrongarm110 | -mstrongarm1100 ]
268 [ -m[arm]v2 | -m[arm]v2a | -m[arm]v3 | -m[arm]v3m |
269 -m[arm]v4 | -m[arm]v4t | -m[arm]v5 | -[arm]v5t |
272 [ -mfpa10 | -mfpa11 | -mfpe-old | -mno-fpu ]
274 [ -mapcs-32 | -mapcs-26 | -mapcs-float |
276 [ -mthumb-interwork ] [ -moabi ] [ -k ]
285 @c Hitachi family chips have no machine-dependent assembler options
288 @c HPPA has no machine-dependent assembler options (yet).
294 @c The order here is important. See c-sparc.texi.
295 [ -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
296 -Av8plus | -Av8plusa | -Av9 | -Av9a ]
297 [ -xarch=v8plus | -xarch=v8plusa ] [ -bump ]
301 [ -mcpu=54[123589] | -mcpu=54[56]lp ] [ -mfar-mode | -mf ]
302 [ -merrors-to-file <filename> | -me <filename> ]
305 @c Z8000 has no machine-dependent assembler options
308 @c see md_parse_option in tc-i960.c
309 [ -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB |
314 [ --m32rx | --[no-]warn-explicit-parallel-conflicts |
318 [ -l ] [ -m68000 | -m68010 | -m68020 | ... ]
321 [ -jsri2bsr ] [ -sifilter ] [ -relax ]
325 [ -m68hc11 | -m68hc12 ]
326 [ --force-long-branchs ] [ --short-branchs ]
327 [ --strict-direct-mode ] [ --print-insn-syntax ]
328 [ --print-opcodes ] [ --generate-example ]
331 [ -nocpp ] [ -EL ] [ -EB ] [ -G @var{num} ] [ -mcpu=@var{CPU} ]
332 [ -mips1 ] [ -mips2 ] [ -mips3 ] [ -mips4 ] [ -mips5 ]
333 [ -mips32 ] [ -mips64 ]
334 [ -m4650 ] [ -no-m4650 ]
335 [ --trap ] [ --break ] [ -n ]
336 [ --emulation=@var{name} ]
338 [ -- | @var{files} @dots{} ]
346 Turn on listings, in any of a variety of ways:
350 omit false conditionals
353 omit debugging directives
356 include high-level source
362 include macro expansions
365 omit forms processing
371 set the name of the listing file
374 You may combine these options; for example, use @samp{-aln} for assembly
375 listing without forms processing. The @samp{=file} option, if used, must be
376 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
379 Ignored. This option is accepted for script compatibility with calls to
382 @item --defsym @var{sym}=@var{value}
383 Define the symbol @var{sym} to be @var{value} before assembling the input file.
384 @var{value} must be an integer constant. As in C, a leading @samp{0x}
385 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
388 ``fast''---skip whitespace and comment preprocessing (assume source is
392 Generate stabs debugging information for each assembler line. This
393 may help debugging assembler code, if the debugger can handle it.
396 Generate DWARF2 debugging information for each assembler line. This
397 may help debugging assembler code, if the debugger can handle it. Note - this
398 option is only supported by some targets, not all of them.
401 Print a summary of the command line options and exit.
404 Print a summary of all target specific options and exit.
407 Add directory @var{dir} to the search list for @code{.include} directives.
410 Don't warn about signed overflow.
413 @ifclear DIFF-TBL-KLUGE
414 This option is accepted but has no effect on the @value{TARGET} family.
416 @ifset DIFF-TBL-KLUGE
417 Issue warnings when difference tables altered for long displacements.
422 Keep (in the symbol table) local symbols. On traditional a.out systems
423 these start with @samp{L}, but different systems have different local
426 @item --listing-lhs-width=@var{number}
427 Set the maximum width, in words, of the output data column for an assembler
428 listing to @var{number}.
430 @item --listing-lhs-width2=@var{number}
431 Set the maximum width, in words, of the output data column for continuation
432 lines in an assembler listing to @var{number}.
434 @item --listing-rhs-width=@var{number}
435 Set the maximum width of an input source line, as displayed in a listing, to
438 @item --listing-cont-lines=@var{number}
439 Set the maximum number of lines printed in a listing for a single line of input
442 @item -o @var{objfile}
443 Name the object-file output from @code{@value{AS}} @var{objfile}.
446 Fold the data section into the text section.
449 Print the maximum space (in bytes) and total time (in seconds) used by
452 @item --strip-local-absolute
453 Remove local absolute symbols from the outgoing symbol table.
457 Print the @code{as} version.
460 Print the @code{as} version and exit.
464 Suppress warning messages.
466 @item --fatal-warnings
467 Treat warnings as errors.
470 Don't suppress warning messages or treat them as errors.
479 Generate an object file even after errors.
481 @item -- | @var{files} @dots{}
482 Standard input, or source files to assemble.
487 The following options are available when @value{AS} is configured for
492 This option selects the core processor variant.
494 Select either big-endian (-EB) or little-endian (-EL) output.
499 The following options are available when @value{AS} is configured for the ARM
503 @item -m[arm][1|2|3|6|7|8|9][...]
504 Specify which ARM processor variant is the target.
505 @item -m[arm]v[2|2a|3|3m|4|4t|5|5t]
506 Specify which ARM architecture variant is used by the target.
507 @item -mthumb | -mall
508 Enable or disable Thumb only instruction decoding.
509 @item -mfpa10 | -mfpa11 | -mfpe-old | -mno-fpu
510 Select which Floating Point architecture is the target.
511 @item -mapcs-32 | -mapcs-26 | -mapcs-float | -mapcs-reentrant | -moabi
512 Select which procedure calling convention is in use.
514 Select either big-endian (-EB) or little-endian (-EL) output.
515 @item -mthumb-interwork
516 Specify that the code has been generated with interworking between Thumb and
519 Specify that PIC code has been generated.
524 The following options are available when @value{AS} is configured for
527 @cindex D10V optimization
528 @cindex optimization, D10V
530 Optimize output by parallelizing instructions.
535 The following options are available when @value{AS} is configured for a D30V
538 @cindex D30V optimization
539 @cindex optimization, D30V
541 Optimize output by parallelizing instructions.
545 Warn when nops are generated.
547 @cindex D30V nops after 32-bit multiply
549 Warn when a nop after a 32-bit multiply instruction is generated.
554 The following options are available when @value{AS} is configured for the
555 Intel 80960 processor.
558 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
559 Specify which variant of the 960 architecture is the target.
562 Add code to collect statistics about branches taken.
565 Do not alter compare-and-branch instructions for long displacements;
572 The following options are available when @value{AS} is configured for the
573 Mitsubishi M32R series.
578 Specify which processor in the M32R family is the target. The default
579 is normally the M32R, but this option changes it to the M32RX.
581 @item --warn-explicit-parallel-conflicts or --Wp
582 Produce warning messages when questionable parallel constructs are
585 @item --no-warn-explicit-parallel-conflicts or --Wnp
586 Do not produce warning messages when questionable parallel constructs are
593 The following options are available when @value{AS} is configured for the
594 Motorola 68000 series.
599 Shorten references to undefined symbols, to one word instead of two.
601 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030
602 @itemx | -m68040 | -m68060 | -m68302 | -m68331 | -m68332
603 @itemx | -m68333 | -m68340 | -mcpu32 | -m5200
604 Specify what processor in the 68000 family is the target. The default
605 is normally the 68020, but this can be changed at configuration time.
607 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
608 The target machine does (or does not) have a floating-point coprocessor.
609 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
610 the basic 68000 is not compatible with the 68881, a combination of the
611 two can be specified, since it's possible to do emulation of the
612 coprocessor instructions with the main processor.
614 @item -m68851 | -mno-68851
615 The target machine does (or does not) have a memory-management
616 unit coprocessor. The default is to assume an MMU for 68020 and up.
623 For details about the PDP-11 machine dependent features options,
624 see @ref{PDP-11-Options}.
627 @item -mpic | -mno-pic
628 Generate position-independent (or position-dependent) code. The
629 default is @code{-mpic}.
632 @itemx -mall-extensions
633 Enable all instruction set extensions. This is the default.
635 @item -mno-extensions
636 Disable all instruction set extensions.
638 @item -m@var{extension} | -mno-@var{extension}
639 Enable (or disable) a particular instruction set extension.
642 Enable the instruction set extensions supported by a particular CPU, and
643 disable all other extensions.
645 @item -m@var{machine}
646 Enable the instruction set extensions supported by a particular machine
647 model, and disable all other extensions.
653 The following options are available when @value{AS} is configured for
654 a picoJava processor.
658 @cindex PJ endianness
659 @cindex endianness, PJ
660 @cindex big endian output, PJ
662 Generate ``big endian'' format output.
664 @cindex little endian output, PJ
666 Generate ``little endian'' format output.
672 The following options are available when @value{AS} is configured for the
673 Motorola 68HC11 or 68HC12 series.
677 @item -m68hc11 | -m68hc12
678 Specify what processor is the target. The default is
679 defined by the configuration option when building the assembler.
681 @item --force-long-branchs
682 Relative branches are turned into absolute ones. This concerns
683 conditional branches, unconditional branches and branches to a
686 @item -S | --short-branchs
687 Do not turn relative branchs into absolute ones
688 when the offset is out of range.
690 @item --strict-direct-mode
691 Do not turn the direct addressing mode into extended addressing mode
692 when the instruction does not support direct addressing mode.
694 @item --print-insn-syntax
695 Print the syntax of instruction in case of error.
697 @item --print-opcodes
698 print the list of instructions with syntax and then exit.
700 @item --generate-example
701 print an example of instruction for each possible instruction and then exit.
702 This option is only useful for testing @code{@value{AS}}.
708 The following options are available when @code{@value{AS}} is configured
709 for the SPARC architecture:
712 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
713 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
714 Explicitly select a variant of the SPARC architecture.
716 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
717 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
719 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
720 UltraSPARC extensions.
722 @item -xarch=v8plus | -xarch=v8plusa
723 For compatibility with the Solaris v9 assembler. These options are
724 equivalent to -Av8plus and -Av8plusa, respectively.
727 Warn when the assembler switches to another architecture.
732 The following options are available when @value{AS} is configured for the 'c54x
737 Enable extended addressing mode. All addresses and relocations will assume
738 extended addressing (usually 23 bits).
739 @item -mcpu=@var{CPU_VERSION}
740 Sets the CPU version being compiled for.
741 @item -merrors-to-file @var{FILENAME}
742 Redirect error output to a file, for broken systems which don't support such
743 behaviour in the shell.
748 The following options are available when @value{AS} is configured for
753 This option sets the largest size of an object that can be referenced
754 implicitly with the @code{gp} register. It is only accepted for targets that
755 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
757 @cindex MIPS endianness
758 @cindex endianness, MIPS
759 @cindex big endian output, MIPS
761 Generate ``big endian'' format output.
763 @cindex little endian output, MIPS
765 Generate ``little endian'' format output.
773 Generate code for a particular MIPS Instruction Set Architecture level.
774 @samp{-mips1} corresponds to the @sc{r2000} and @sc{r3000} processors,
775 @samp{-mips2} to the @sc{r6000} processor, and @samp{-mips3} to the @sc{r4000}
777 @samp{-mips5}, @samp{-mips32}, and @samp{-mips64} correspond
778 to generic @sc{MIPS V}, @sc{MIPS32}, and @sc{MIPS64} ISA
779 processors, respectively.
783 Generate code for the MIPS @sc{r4650} chip. This tells the assembler to accept
784 the @samp{mad} and @samp{madu} instruction, and to not schedule @samp{nop}
785 instructions around accesses to the @samp{HI} and @samp{LO} registers.
786 @samp{-no-m4650} turns off this option.
788 @item -mcpu=@var{CPU}
789 Generate code for a particular MIPS cpu. It is exactly equivalent to
790 @samp{-m@var{cpu}}, except that there are more value of @var{cpu}
794 @item --emulation=@var{name}
795 This option causes @code{@value{AS}} to emulate @code{@value{AS}} configured
796 for some other target, in all respects, including output format (choosing
797 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
798 debugging information or store symbol table information, and default
799 endianness. The available configuration names are: @samp{mipsecoff},
800 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
801 @samp{mipsbelf}. The first two do not alter the default endianness from that
802 of the primary target for which the assembler was configured; the others change
803 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
804 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
805 selection in any case.
807 This option is currently supported only when the primary target
808 @code{@value{AS}} is configured for is a MIPS ELF or ECOFF target.
809 Furthermore, the primary target or others specified with
810 @samp{--enable-targets=@dots{}} at configuration time must include support for
811 the other format, if both are to be available. For example, the Irix 5
812 configuration includes support for both.
814 Eventually, this option will support more configurations, with more
815 fine-grained control over the assembler's behavior, and will be supported for
819 @code{@value{AS}} ignores this option. It is accepted for compatibility with
827 Control how to deal with multiplication overflow and division by zero.
828 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
829 (and only work for Instruction Set Architecture level 2 and higher);
830 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
834 When this option is used, @code{@value{AS}} will issue a warning every
835 time it generates a nop instruction from a macro.
840 The following options are available when @value{AS} is configured for
846 Enable or disable the JSRI to BSR transformation. By default this is enabled.
847 The command line option @samp{-nojsri2bsr} can be used to disable it.
851 Enable or disable the silicon filter behaviour. By default this is disabled.
852 The default can be overridden by the @samp{-sifilter} command line option.
855 Alter jump instructions for long displacements.
857 @item -mcpu=[210|340]
858 Select the cpu type on the target hardware. This controls which instructions
862 Assemble for a big endian target.
865 Assemble for a little endian target.
873 * Manual:: Structure of this Manual
874 * GNU Assembler:: The GNU Assembler
875 * Object Formats:: Object File Formats
876 * Command Line:: Command Line
877 * Input Files:: Input Files
878 * Object:: Output (Object) File
879 * Errors:: Error and Warning Messages
883 @section Structure of this Manual
885 @cindex manual, structure and purpose
886 This manual is intended to describe what you need to know to use
887 @sc{gnu} @code{@value{AS}}. We cover the syntax expected in source files, including
888 notation for symbols, constants, and expressions; the directives that
889 @code{@value{AS}} understands; and of course how to invoke @code{@value{AS}}.
892 We also cover special features in the @value{TARGET}
893 configuration of @code{@value{AS}}, including assembler directives.
896 This manual also describes some of the machine-dependent features of
897 various flavors of the assembler.
900 @cindex machine instructions (not covered)
901 On the other hand, this manual is @emph{not} intended as an introduction
902 to programming in assembly language---let alone programming in general!
903 In a similar vein, we make no attempt to introduce the machine
904 architecture; we do @emph{not} describe the instruction set, standard
905 mnemonics, registers or addressing modes that are standard to a
906 particular architecture.
908 You may want to consult the manufacturer's
909 machine architecture manual for this information.
913 For information on the H8/300 machine instruction set, see @cite{H8/300
914 Series Programming Manual} (Hitachi ADE--602--025). For the H8/300H,
915 see @cite{H8/300H Series Programming Manual} (Hitachi).
918 For information on the H8/500 machine instruction set, see @cite{H8/500
919 Series Programming Manual} (Hitachi M21T001).
922 For information on the Hitachi SH machine instruction set, see
923 @cite{SH-Microcomputer User's Manual} (Hitachi Micro Systems, Inc.).
926 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
930 @c I think this is premature---doc@cygnus.com, 17jan1991
932 Throughout this manual, we assume that you are running @dfn{GNU},
933 the portable operating system from the @dfn{Free Software
934 Foundation, Inc.}. This restricts our attention to certain kinds of
935 computer (in particular, the kinds of computers that @sc{gnu} can run on);
936 once this assumption is granted examples and definitions need less
939 @code{@value{AS}} is part of a team of programs that turn a high-level
940 human-readable series of instructions into a low-level
941 computer-readable series of instructions. Different versions of
942 @code{@value{AS}} are used for different kinds of computer.
945 @c There used to be a section "Terminology" here, which defined
946 @c "contents", "byte", "word", and "long". Defining "word" to any
947 @c particular size is confusing when the .word directive may generate 16
948 @c bits on one machine and 32 bits on another; in general, for the user
949 @c version of this manual, none of these terms seem essential to define.
950 @c They were used very little even in the former draft of the manual;
951 @c this draft makes an effort to avoid them (except in names of
955 @section The GNU Assembler
957 @c man begin DESCRIPTION
959 @sc{gnu} @code{as} is really a family of assemblers.
961 This manual describes @code{@value{AS}}, a member of that family which is
962 configured for the @value{TARGET} architectures.
964 If you use (or have used) the @sc{gnu} assembler on one architecture, you
965 should find a fairly similar environment when you use it on another
966 architecture. Each version has much in common with the others,
967 including object file formats, most assembler directives (often called
968 @dfn{pseudo-ops}) and assembler syntax.@refill
970 @cindex purpose of @sc{gnu} assembler
971 @code{@value{AS}} is primarily intended to assemble the output of the
972 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
973 @code{@value{LD}}. Nevertheless, we've tried to make @code{@value{AS}}
974 assemble correctly everything that other assemblers for the same
975 machine would assemble.
977 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
980 @c This remark should appear in generic version of manual; assumption
981 @c here is that generic version sets M680x0.
982 This doesn't mean @code{@value{AS}} always uses the same syntax as another
983 assembler for the same architecture; for example, we know of several
984 incompatible versions of 680x0 assembly language syntax.
989 Unlike older assemblers, @code{@value{AS}} is designed to assemble a source
990 program in one pass of the source file. This has a subtle impact on the
991 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
994 @section Object File Formats
996 @cindex object file format
997 The @sc{gnu} assembler can be configured to produce several alternative
998 object file formats. For the most part, this does not affect how you
999 write assembly language programs; but directives for debugging symbols
1000 are typically different in different file formats. @xref{Symbol
1001 Attributes,,Symbol Attributes}.
1004 On the @value{TARGET}, @code{@value{AS}} is configured to produce
1005 @value{OBJ-NAME} format object files.
1007 @c The following should exhaust all configs that set MULTI-OBJ, ideally
1009 On the @value{TARGET}, @code{@value{AS}} can be configured to produce either
1010 @code{a.out} or COFF format object files.
1013 On the @value{TARGET}, @code{@value{AS}} can be configured to produce either
1014 @code{b.out} or COFF format object files.
1017 On the @value{TARGET}, @code{@value{AS}} can be configured to produce either
1018 SOM or ELF format object files.
1023 @section Command Line
1025 @cindex command line conventions
1027 After the program name @code{@value{AS}}, the command line may contain
1028 options and file names. Options may appear in any order, and may be
1029 before, after, or between file names. The order of file names is
1032 @cindex standard input, as input file
1034 @file{--} (two hyphens) by itself names the standard input file
1035 explicitly, as one of the files for @code{@value{AS}} to assemble.
1037 @cindex options, command line
1038 Except for @samp{--} any command line argument that begins with a
1039 hyphen (@samp{-}) is an option. Each option changes the behavior of
1040 @code{@value{AS}}. No option changes the way another option works. An
1041 option is a @samp{-} followed by one or more letters; the case of
1042 the letter is important. All options are optional.
1044 Some options expect exactly one file name to follow them. The file
1045 name may either immediately follow the option's letter (compatible
1046 with older assemblers) or it may be the next command argument (@sc{gnu}
1047 standard). These two command lines are equivalent:
1050 @value{AS} -o my-object-file.o mumble.s
1051 @value{AS} -omy-object-file.o mumble.s
1055 @section Input Files
1058 @cindex source program
1059 @cindex files, input
1060 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
1061 describe the program input to one run of @code{@value{AS}}. The program may
1062 be in one or more files; how the source is partitioned into files
1063 doesn't change the meaning of the source.
1065 @c I added "con" prefix to "catenation" just to prove I can overcome my
1066 @c APL training... doc@cygnus.com
1067 The source program is a concatenation of the text in all the files, in the
1070 @c man begin DESCRIPTION
1071 Each time you run @code{@value{AS}} it assembles exactly one source
1072 program. The source program is made up of one or more files.
1073 (The standard input is also a file.)
1075 You give @code{@value{AS}} a command line that has zero or more input file
1076 names. The input files are read (from left file name to right). A
1077 command line argument (in any position) that has no special meaning
1078 is taken to be an input file name.
1080 If you give @code{@value{AS}} no file names it attempts to read one input file
1081 from the @code{@value{AS}} standard input, which is normally your terminal. You
1082 may have to type @key{ctl-D} to tell @code{@value{AS}} there is no more program
1085 Use @samp{--} if you need to explicitly name the standard input file
1086 in your command line.
1088 If the source is empty, @code{@value{AS}} produces a small, empty object
1093 @subheading Filenames and Line-numbers
1095 @cindex input file linenumbers
1096 @cindex line numbers, in input files
1097 There are two ways of locating a line in the input file (or files) and
1098 either may be used in reporting error messages. One way refers to a line
1099 number in a physical file; the other refers to a line number in a
1100 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
1102 @dfn{Physical files} are those files named in the command line given
1103 to @code{@value{AS}}.
1105 @dfn{Logical files} are simply names declared explicitly by assembler
1106 directives; they bear no relation to physical files. Logical file names help
1107 error messages reflect the original source file, when @code{@value{AS}} source
1108 is itself synthesized from other files. @code{@value{AS}} understands the
1109 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
1110 @ref{File,,@code{.file}}.
1113 @section Output (Object) File
1119 Every time you run @code{@value{AS}} it produces an output file, which is
1120 your assembly language program translated into numbers. This file
1121 is the object file. Its default name is
1129 @code{b.out} when @code{@value{AS}} is configured for the Intel 80960.
1131 You can give it another name by using the @code{-o} option. Conventionally,
1132 object file names end with @file{.o}. The default name is used for historical
1133 reasons: older assemblers were capable of assembling self-contained programs
1134 directly into a runnable program. (For some formats, this isn't currently
1135 possible, but it can be done for the @code{a.out} format.)
1139 The object file is meant for input to the linker @code{@value{LD}}. It contains
1140 assembled program code, information to help @code{@value{LD}} integrate
1141 the assembled program into a runnable file, and (optionally) symbolic
1142 information for the debugger.
1144 @c link above to some info file(s) like the description of a.out.
1145 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
1148 @section Error and Warning Messages
1150 @c man begin DESCRIPTION
1152 @cindex error messages
1153 @cindex warning messages
1154 @cindex messages from assembler
1155 @code{@value{AS}} may write warnings and error messages to the standard error
1156 file (usually your terminal). This should not happen when a compiler
1157 runs @code{@value{AS}} automatically. Warnings report an assumption made so
1158 that @code{@value{AS}} could keep assembling a flawed program; errors report a
1159 grave problem that stops the assembly.
1163 @cindex format of warning messages
1164 Warning messages have the format
1167 file_name:@b{NNN}:Warning Message Text
1171 @cindex line numbers, in warnings/errors
1172 (where @b{NNN} is a line number). If a logical file name has been given
1173 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
1174 the current input file is used. If a logical line number was given
1176 (@pxref{Line,,@code{.line}})
1180 (@pxref{Line,,@code{.line}})
1183 (@pxref{Ln,,@code{.ln}})
1186 then it is used to calculate the number printed,
1187 otherwise the actual line in the current source file is printed. The
1188 message text is intended to be self explanatory (in the grand Unix
1191 @cindex format of error messages
1192 Error messages have the format
1194 file_name:@b{NNN}:FATAL:Error Message Text
1196 The file name and line number are derived as for warning
1197 messages. The actual message text may be rather less explanatory
1198 because many of them aren't supposed to happen.
1201 @chapter Command-Line Options
1203 @cindex options, all versions of assembler
1204 This chapter describes command-line options available in @emph{all}
1205 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
1207 to the @value{TARGET}.
1210 to particular machine architectures.
1213 @c man begin DESCRIPTION
1215 If you are invoking @code{@value{AS}} via the @sc{gnu} C compiler (version 2),
1216 you can use the @samp{-Wa} option to pass arguments through to the assembler.
1217 The assembler arguments must be separated from each other (and the @samp{-Wa})
1218 by commas. For example:
1221 gcc -c -g -O -Wa,-alh,-L file.c
1225 This passes two options to the assembler: @samp{-alh} (emit a listing to
1226 standard output with with high-level and assembly source) and @samp{-L} (retain
1227 local symbols in the symbol table).
1229 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
1230 command-line options are automatically passed to the assembler by the compiler.
1231 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
1232 precisely what options it passes to each compilation pass, including the
1238 * a:: -a[cdhlns] enable listings
1239 * D:: -D for compatibility
1240 * f:: -f to work faster
1241 * I:: -I for .include search path
1242 @ifclear DIFF-TBL-KLUGE
1243 * K:: -K for compatibility
1245 @ifset DIFF-TBL-KLUGE
1246 * K:: -K for difference tables
1249 * L:: -L to retain local labels
1250 * listing:: --listing-XXX to configure listing output
1251 * M:: -M or --mri to assemble in MRI compatibility mode
1252 * MD:: --MD for dependency tracking
1253 * o:: -o to name the object file
1254 * R:: -R to join data and text sections
1255 * statistics:: --statistics to see statistics about assembly
1256 * traditional-format:: --traditional-format for compatible output
1257 * v:: -v to announce version
1258 * W:: -W, --no-warn, --warn, --fatal-warnings to control warnings
1259 * Z:: -Z to make object file even after errors
1263 @section Enable Listings: @code{-a[cdhlns]}
1272 @cindex listings, enabling
1273 @cindex assembly listings, enabling
1275 These options enable listing output from the assembler. By itself,
1276 @samp{-a} requests high-level, assembly, and symbols listing.
1277 You can use other letters to select specific options for the list:
1278 @samp{-ah} requests a high-level language listing,
1279 @samp{-al} requests an output-program assembly listing, and
1280 @samp{-as} requests a symbol table listing.
1281 High-level listings require that a compiler debugging option like
1282 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1285 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1286 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1287 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1288 omitted from the listing.
1290 Use the @samp{-ad} option to omit debugging directives from the
1293 Once you have specified one of these options, you can further control
1294 listing output and its appearance using the directives @code{.list},
1295 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1297 The @samp{-an} option turns off all forms processing.
1298 If you do not request listing output with one of the @samp{-a} options, the
1299 listing-control directives have no effect.
1301 The letters after @samp{-a} may be combined into one option,
1302 @emph{e.g.}, @samp{-aln}.
1304 Note if the assembler source is coming from the standard input (eg because it
1305 is being created by @code{@value{GCC}} and the @samp{-pipe} command line switch
1306 is being used) then the listing will not contain any comments or preprocessor
1307 directives. This is because the listing code buffers input source lines from
1308 stdin only after they have been preprocessed by the assembler. This reduces
1309 memory usage and makes the code more efficient.
1315 This option has no effect whatsoever, but it is accepted to make it more
1316 likely that scripts written for other assemblers also work with
1320 @section Work Faster: @code{-f}
1323 @cindex trusted compiler
1324 @cindex faster processing (@code{-f})
1325 @samp{-f} should only be used when assembling programs written by a
1326 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1327 and comment preprocessing on
1328 the input file(s) before assembling them. @xref{Preprocessing,
1332 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1333 preprocessed (if they contain comments, for example), @code{@value{AS}} does
1338 @section @code{.include} search path: @code{-I} @var{path}
1340 @kindex -I @var{path}
1341 @cindex paths for @code{.include}
1342 @cindex search path for @code{.include}
1343 @cindex @code{include} directive search path
1344 Use this option to add a @var{path} to the list of directories
1345 @code{@value{AS}} searches for files specified in @code{.include}
1346 directives (@pxref{Include,,@code{.include}}). You may use @code{-I} as
1347 many times as necessary to include a variety of paths. The current
1348 working directory is always searched first; after that, @code{@value{AS}}
1349 searches any @samp{-I} directories in the same order as they were
1350 specified (left to right) on the command line.
1353 @section Difference Tables: @code{-K}
1356 @ifclear DIFF-TBL-KLUGE
1357 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1358 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1359 where it can be used to warn when the assembler alters the machine code
1360 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1361 family does not have the addressing limitations that sometimes lead to this
1362 alteration on other platforms.
1365 @ifset DIFF-TBL-KLUGE
1366 @cindex difference tables, warning
1367 @cindex warning for altered difference tables
1368 @code{@value{AS}} sometimes alters the code emitted for directives of the form
1369 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1370 You can use the @samp{-K} option if you want a warning issued when this
1375 @section Include Local Labels: @code{-L}
1378 @cindex local labels, retaining in output
1379 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1380 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1381 debugging, because they are intended for the use of programs (like
1382 compilers) that compose assembler programs, not for your notice.
1383 Normally both @code{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1384 normally debug with them.
1386 This option tells @code{@value{AS}} to retain those @samp{L@dots{}} symbols
1387 in the object file. Usually if you do this you also tell the linker
1388 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1390 By default, a local label is any label beginning with @samp{L}, but each
1391 target is allowed to redefine the local label prefix.
1393 On the HPPA local labels begin with @samp{L$}.
1397 @section Configuringh listing output: @code{--listing}
1399 The listing feature of the assembler can be enabled via the command line switch
1400 @samp{-a} (@pxref{a}). This feature combines the input source file(s) with a
1401 hex dump of the corresponding locations in the output object file, and displays
1402 them as a listing file. The format of this listing can be controlled by pseudo
1403 ops inside the assembler source (@pxref{List} @pxref{Title} @pxref{Sbttl}
1404 @pxref{Psize} @pxref{Eject}) and also by the following switches:
1407 @item --listing-lhs-width=@samp{number}
1408 @kindex --listing-lhs-width
1409 @cindex Width of first line disassembly output
1410 Sets the maximum width, in words, of the first line of the hex byte dump. This
1411 dump appears on the left hand side of the listing output.
1413 @item --listing-lhs-width2=@samp{number}
1414 @kindex --listing-lhs-width2
1415 @cindex Width of continuation lines of disassembly output
1416 Sets the maximum width, in words, of any further lines of the hex byte dump for
1417 a given inut source line. If this value is not specified, it defaults to being
1418 the same as the value specified for @samp{--listing-lhs-width}. If neither
1419 switch is used the default is to one.
1421 @item --listing-rhs-width=@samp{number}
1422 @kindex --listing-rhs-width
1423 @cindex Width of source line output
1424 Sets the maximum width, in characters, of the source line that is displayed
1425 alongside the hex dump. The default value for this parameter is 100. The
1426 source line is displayed on the right hand side of the listing output.
1428 @item --listing-cont-lines=@samp{number}
1429 @kindex --listing-cont-lines
1430 @cindex Maximum number of continuation lines
1431 Sets the maximum number of continuation lines of hex dump that will be
1432 displayed for a given single line of source input. The default value is 4.
1436 @section Assemble in MRI Compatibility Mode: @code{-M}
1439 @cindex MRI compatibility mode
1440 The @code{-M} or @code{--mri} option selects MRI compatibility mode. This
1441 changes the syntax and pseudo-op handling of @code{@value{AS}} to make it
1442 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1443 configured target) assembler from Microtec Research. The exact nature of the
1444 MRI syntax will not be documented here; see the MRI manuals for more
1445 information. Note in particular that the handling of macros and macro
1446 arguments is somewhat different. The purpose of this option is to permit
1447 assembling existing MRI assembler code using @code{@value{AS}}.
1449 The MRI compatibility is not complete. Certain operations of the MRI assembler
1450 depend upon its object file format, and can not be supported using other object
1451 file formats. Supporting these would require enhancing each object file format
1452 individually. These are:
1455 @item global symbols in common section
1457 The m68k MRI assembler supports common sections which are merged by the linker.
1458 Other object file formats do not support this. @code{@value{AS}} handles
1459 common sections by treating them as a single common symbol. It permits local
1460 symbols to be defined within a common section, but it can not support global
1461 symbols, since it has no way to describe them.
1463 @item complex relocations
1465 The MRI assemblers support relocations against a negated section address, and
1466 relocations which combine the start addresses of two or more sections. These
1467 are not support by other object file formats.
1469 @item @code{END} pseudo-op specifying start address
1471 The MRI @code{END} pseudo-op permits the specification of a start address.
1472 This is not supported by other object file formats. The start address may
1473 instead be specified using the @code{-e} option to the linker, or in a linker
1476 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1478 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1479 name to the output file. This is not supported by other object file formats.
1481 @item @code{ORG} pseudo-op
1483 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1484 address. This differs from the usual @code{@value{AS}} @code{.org} pseudo-op,
1485 which changes the location within the current section. Absolute sections are
1486 not supported by other object file formats. The address of a section may be
1487 assigned within a linker script.
1490 There are some other features of the MRI assembler which are not supported by
1491 @code{@value{AS}}, typically either because they are difficult or because they
1492 seem of little consequence. Some of these may be supported in future releases.
1496 @item EBCDIC strings
1498 EBCDIC strings are not supported.
1500 @item packed binary coded decimal
1502 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1503 and @code{DCB.P} pseudo-ops are not supported.
1505 @item @code{FEQU} pseudo-op
1507 The m68k @code{FEQU} pseudo-op is not supported.
1509 @item @code{NOOBJ} pseudo-op
1511 The m68k @code{NOOBJ} pseudo-op is not supported.
1513 @item @code{OPT} branch control options
1515 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1516 @code{BRL}, and @code{BRW}---are ignored. @code{@value{AS}} automatically
1517 relaxes all branches, whether forward or backward, to an appropriate size, so
1518 these options serve no purpose.
1520 @item @code{OPT} list control options
1522 The following m68k @code{OPT} list control options are ignored: @code{C},
1523 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1524 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1526 @item other @code{OPT} options
1528 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1529 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1531 @item @code{OPT} @code{D} option is default
1533 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1534 @code{OPT NOD} may be used to turn it off.
1536 @item @code{XREF} pseudo-op.
1538 The m68k @code{XREF} pseudo-op is ignored.
1540 @item @code{.debug} pseudo-op
1542 The i960 @code{.debug} pseudo-op is not supported.
1544 @item @code{.extended} pseudo-op
1546 The i960 @code{.extended} pseudo-op is not supported.
1548 @item @code{.list} pseudo-op.
1550 The various options of the i960 @code{.list} pseudo-op are not supported.
1552 @item @code{.optimize} pseudo-op
1554 The i960 @code{.optimize} pseudo-op is not supported.
1556 @item @code{.output} pseudo-op
1558 The i960 @code{.output} pseudo-op is not supported.
1560 @item @code{.setreal} pseudo-op
1562 The i960 @code{.setreal} pseudo-op is not supported.
1567 @section Dependency tracking: @code{--MD}
1570 @cindex dependency tracking
1573 @code{@value{AS}} can generate a dependency file for the file it creates. This
1574 file consists of a single rule suitable for @code{make} describing the
1575 dependencies of the main source file.
1577 The rule is written to the file named in its argument.
1579 This feature is used in the automatic updating of makefiles.
1582 @section Name the Object File: @code{-o}
1585 @cindex naming object file
1586 @cindex object file name
1587 There is always one object file output when you run @code{@value{AS}}. By
1588 default it has the name
1591 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1605 You use this option (which takes exactly one filename) to give the
1606 object file a different name.
1608 Whatever the object file is called, @code{@value{AS}} overwrites any
1609 existing file of the same name.
1612 @section Join Data and Text Sections: @code{-R}
1615 @cindex data and text sections, joining
1616 @cindex text and data sections, joining
1617 @cindex joining text and data sections
1618 @cindex merging text and data sections
1619 @code{-R} tells @code{@value{AS}} to write the object file as if all
1620 data-section data lives in the text section. This is only done at
1621 the very last moment: your binary data are the same, but data
1622 section parts are relocated differently. The data section part of
1623 your object file is zero bytes long because all its bytes are
1624 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1626 When you specify @code{-R} it would be possible to generate shorter
1627 address displacements (because we do not have to cross between text and
1628 data section). We refrain from doing this simply for compatibility with
1629 older versions of @code{@value{AS}}. In future, @code{-R} may work this way.
1632 When @code{@value{AS}} is configured for COFF output,
1633 this option is only useful if you use sections named @samp{.text} and
1638 @code{-R} is not supported for any of the HPPA targets. Using
1639 @code{-R} generates a warning from @code{@value{AS}}.
1643 @section Display Assembly Statistics: @code{--statistics}
1645 @kindex --statistics
1646 @cindex statistics, about assembly
1647 @cindex time, total for assembly
1648 @cindex space used, maximum for assembly
1649 Use @samp{--statistics} to display two statistics about the resources used by
1650 @code{@value{AS}}: the maximum amount of space allocated during the assembly
1651 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1654 @node traditional-format
1655 @section Compatible output: @code{--traditional-format}
1657 @kindex --traditional-format
1658 For some targets, the output of @code{@value{AS}} is different in some ways
1659 from the output of some existing assembler. This switch requests
1660 @code{@value{AS}} to use the traditional format instead.
1662 For example, it disables the exception frame optimizations which
1663 @code{@value{AS}} normally does by default on @code{@value{GCC}} output.
1666 @section Announce Version: @code{-v}
1670 @cindex assembler version
1671 @cindex version of assembler
1672 You can find out what version of as is running by including the
1673 option @samp{-v} (which you can also spell as @samp{-version}) on the
1677 @section Control Warnings: @code{-W}, @code{--warn}, @code{--no-warn}, @code{--fatal-warnings}
1679 @code{@value{AS}} should never give a warning or error message when
1680 assembling compiler output. But programs written by people often
1681 cause @code{@value{AS}} to give a warning that a particular assumption was
1682 made. All such warnings are directed to the standard error file.
1685 @kindex @samp{--no-warn}
1686 @cindex suppressing warnings
1687 @cindex warnings, suppressing
1688 If you use the @code{-W} and @code{--no-warn} options, no warnings are issued.
1689 This only affects the warning messages: it does not change any particular of
1690 how @code{@value{AS}} assembles your file. Errors, which stop the assembly,
1693 @kindex @samp{--fatal-warnings}
1694 @cindex errors, caused by warnings
1695 @cindex warnings, causing error
1696 If you use the @code{--fatal-warnings} option, @code{@value{AS}} considers
1697 files that generate warnings to be in error.
1699 @kindex @samp{--warn}
1700 @cindex warnings, switching on
1701 You can switch these options off again by specifying @code{--warn}, which
1702 causes warnings to be output as usual.
1705 @section Generate Object File in Spite of Errors: @code{-Z}
1706 @cindex object file, after errors
1707 @cindex errors, continuing after
1708 After an error message, @code{@value{AS}} normally produces no output. If for
1709 some reason you are interested in object file output even after
1710 @code{@value{AS}} gives an error message on your program, use the @samp{-Z}
1711 option. If there are any errors, @code{@value{AS}} continues anyways, and
1712 writes an object file after a final warning message of the form @samp{@var{n}
1713 errors, @var{m} warnings, generating bad object file.}
1718 @cindex machine-independent syntax
1719 @cindex syntax, machine-independent
1720 This chapter describes the machine-independent syntax allowed in a
1721 source file. @code{@value{AS}} syntax is similar to what many other
1722 assemblers use; it is inspired by the BSD 4.2
1727 assembler, except that @code{@value{AS}} does not assemble Vax bit-fields.
1731 * Preprocessing:: Preprocessing
1732 * Whitespace:: Whitespace
1733 * Comments:: Comments
1734 * Symbol Intro:: Symbols
1735 * Statements:: Statements
1736 * Constants:: Constants
1740 @section Preprocessing
1742 @cindex preprocessing
1743 The @code{@value{AS}} internal preprocessor:
1745 @cindex whitespace, removed by preprocessor
1747 adjusts and removes extra whitespace. It leaves one space or tab before
1748 the keywords on a line, and turns any other whitespace on the line into
1751 @cindex comments, removed by preprocessor
1753 removes all comments, replacing them with a single space, or an
1754 appropriate number of newlines.
1756 @cindex constants, converted by preprocessor
1758 converts character constants into the appropriate numeric values.
1761 It does not do macro processing, include file handling, or
1762 anything else you may get from your C compiler's preprocessor. You can
1763 do include file processing with the @code{.include} directive
1764 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
1765 to get other ``CPP'' style preprocessing, by giving the input file a
1766 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
1767 Output, gcc.info, Using GNU CC}.
1769 Excess whitespace, comments, and character constants
1770 cannot be used in the portions of the input text that are not
1773 @cindex turning preprocessing on and off
1774 @cindex preprocessing, turning on and off
1777 If the first line of an input file is @code{#NO_APP} or if you use the
1778 @samp{-f} option, whitespace and comments are not removed from the input file.
1779 Within an input file, you can ask for whitespace and comment removal in
1780 specific portions of the by putting a line that says @code{#APP} before the
1781 text that may contain whitespace or comments, and putting a line that says
1782 @code{#NO_APP} after this text. This feature is mainly intend to support
1783 @code{asm} statements in compilers whose output is otherwise free of comments
1790 @dfn{Whitespace} is one or more blanks or tabs, in any order.
1791 Whitespace is used to separate symbols, and to make programs neater for
1792 people to read. Unless within character constants
1793 (@pxref{Characters,,Character Constants}), any whitespace means the same
1794 as exactly one space.
1800 There are two ways of rendering comments to @code{@value{AS}}. In both
1801 cases the comment is equivalent to one space.
1803 Anything from @samp{/*} through the next @samp{*/} is a comment.
1804 This means you may not nest these comments.
1808 The only way to include a newline ('\n') in a comment
1809 is to use this sort of comment.
1812 /* This sort of comment does not nest. */
1815 @cindex line comment character
1816 Anything from the @dfn{line comment} character to the next newline
1817 is considered a comment and is ignored. The line comment character is
1819 @samp{;} for the AMD 29K family;
1822 @samp{;} on the ARC;
1825 @samp{@@} on the ARM;
1828 @samp{;} for the H8/300 family;
1831 @samp{!} for the H8/500 family;
1834 @samp{;} for the HPPA;
1837 @samp{#} on the i386 and x86-64;
1840 @samp{#} on the i960;
1843 @samp{;} for the PDP-11;
1846 @samp{;} for picoJava;
1849 @samp{!} for the Hitachi SH;
1852 @samp{!} on the SPARC;
1855 @samp{#} on the m32r;
1858 @samp{|} on the 680x0;
1861 @samp{#} on the 68HC11 and 68HC12;
1864 @samp{#} on the Vax;
1867 @samp{!} for the Z8000;
1870 @samp{#} on the V850;
1872 see @ref{Machine Dependencies}. @refill
1873 @c FIXME What about m88k, i860?
1876 On some machines there are two different line comment characters. One
1877 character only begins a comment if it is the first non-whitespace character on
1878 a line, while the other always begins a comment.
1882 The V850 assembler also supports a double dash as starting a comment that
1883 extends to the end of the line.
1889 @cindex lines starting with @code{#}
1890 @cindex logical line numbers
1891 To be compatible with past assemblers, lines that begin with @samp{#} have a
1892 special interpretation. Following the @samp{#} should be an absolute
1893 expression (@pxref{Expressions}): the logical line number of the @emph{next}
1894 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
1895 new logical file name. The rest of the line, if any, should be whitespace.
1897 If the first non-whitespace characters on the line are not numeric,
1898 the line is ignored. (Just like a comment.)
1901 # This is an ordinary comment.
1902 # 42-6 "new_file_name" # New logical file name
1903 # This is logical line # 36.
1905 This feature is deprecated, and may disappear from future versions
1906 of @code{@value{AS}}.
1911 @cindex characters used in symbols
1912 @ifclear SPECIAL-SYMS
1913 A @dfn{symbol} is one or more characters chosen from the set of all
1914 letters (both upper and lower case), digits and the three characters
1920 A @dfn{symbol} is one or more characters chosen from the set of all
1921 letters (both upper and lower case), digits and the three characters
1922 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
1928 On most machines, you can also use @code{$} in symbol names; exceptions
1929 are noted in @ref{Machine Dependencies}.
1931 No symbol may begin with a digit. Case is significant.
1932 There is no length limit: all characters are significant. Symbols are
1933 delimited by characters not in that set, or by the beginning of a file
1934 (since the source program must end with a newline, the end of a file is
1935 not a possible symbol delimiter). @xref{Symbols}.
1936 @cindex length of symbols
1941 @cindex statements, structure of
1942 @cindex line separator character
1943 @cindex statement separator character
1945 @ifclear abnormal-separator
1946 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
1947 semicolon (@samp{;}). The newline or semicolon is considered part of
1948 the preceding statement. Newlines and semicolons within character
1949 constants are an exception: they do not end statements.
1951 @ifset abnormal-separator
1953 A @dfn{statement} ends at a newline character (@samp{\n}) or an ``at''
1954 sign (@samp{@@}). The newline or at sign is considered part of the
1955 preceding statement. Newlines and at signs within character constants
1956 are an exception: they do not end statements.
1959 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
1960 point (@samp{!}). The newline or exclamation point is considered part of the
1961 preceding statement. Newlines and exclamation points within character
1962 constants are an exception: they do not end statements.
1965 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
1966 H8/300) a dollar sign (@samp{$}); or (for the
1969 (@samp{;}). The newline or separator character is considered part of
1970 the preceding statement. Newlines and separators within character
1971 constants are an exception: they do not end statements.
1976 A @dfn{statement} ends at a newline character (@samp{\n}) or line
1977 separator character. (The line separator is usually @samp{;}, unless
1978 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
1979 newline or separator character is considered part of the preceding
1980 statement. Newlines and separators within character constants are an
1981 exception: they do not end statements.
1984 @cindex newline, required at file end
1985 @cindex EOF, newline must precede
1986 It is an error to end any statement with end-of-file: the last
1987 character of any input file should be a newline.@refill
1989 An empty statement is allowed, and may include whitespace. It is ignored.
1991 @cindex instructions and directives
1992 @cindex directives and instructions
1993 @c "key symbol" is not used elsewhere in the document; seems pedantic to
1994 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
1996 A statement begins with zero or more labels, optionally followed by a
1997 key symbol which determines what kind of statement it is. The key
1998 symbol determines the syntax of the rest of the statement. If the
1999 symbol begins with a dot @samp{.} then the statement is an assembler
2000 directive: typically valid for any computer. If the symbol begins with
2001 a letter the statement is an assembly language @dfn{instruction}: it
2002 assembles into a machine language instruction.
2004 Different versions of @code{@value{AS}} for different computers
2005 recognize different instructions. In fact, the same symbol may
2006 represent a different instruction in a different computer's assembly
2010 @cindex @code{:} (label)
2011 @cindex label (@code{:})
2012 A label is a symbol immediately followed by a colon (@code{:}).
2013 Whitespace before a label or after a colon is permitted, but you may not
2014 have whitespace between a label's symbol and its colon. @xref{Labels}.
2017 For HPPA targets, labels need not be immediately followed by a colon, but
2018 the definition of a label must begin in column zero. This also implies that
2019 only one label may be defined on each line.
2023 label: .directive followed by something
2024 another_label: # This is an empty statement.
2025 instruction operand_1, operand_2, @dots{}
2032 A constant is a number, written so that its value is known by
2033 inspection, without knowing any context. Like this:
2036 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
2037 .ascii "Ring the bell\7" # A string constant.
2038 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
2039 .float 0f-314159265358979323846264338327\
2040 95028841971.693993751E-40 # - pi, a flonum.
2045 * Characters:: Character Constants
2046 * Numbers:: Number Constants
2050 @subsection Character Constants
2052 @cindex character constants
2053 @cindex constants, character
2054 There are two kinds of character constants. A @dfn{character} stands
2055 for one character in one byte and its value may be used in
2056 numeric expressions. String constants (properly called string
2057 @emph{literals}) are potentially many bytes and their values may not be
2058 used in arithmetic expressions.
2062 * Chars:: Characters
2066 @subsubsection Strings
2068 @cindex string constants
2069 @cindex constants, string
2070 A @dfn{string} is written between double-quotes. It may contain
2071 double-quotes or null characters. The way to get special characters
2072 into a string is to @dfn{escape} these characters: precede them with
2073 a backslash @samp{\} character. For example @samp{\\} represents
2074 one backslash: the first @code{\} is an escape which tells
2075 @code{@value{AS}} to interpret the second character literally as a backslash
2076 (which prevents @code{@value{AS}} from recognizing the second @code{\} as an
2077 escape character). The complete list of escapes follows.
2079 @cindex escape codes, character
2080 @cindex character escape codes
2083 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
2085 @cindex @code{\b} (backspace character)
2086 @cindex backspace (@code{\b})
2088 Mnemonic for backspace; for ASCII this is octal code 010.
2091 @c Mnemonic for EOText; for ASCII this is octal code 004.
2093 @cindex @code{\f} (formfeed character)
2094 @cindex formfeed (@code{\f})
2096 Mnemonic for FormFeed; for ASCII this is octal code 014.
2098 @cindex @code{\n} (newline character)
2099 @cindex newline (@code{\n})
2101 Mnemonic for newline; for ASCII this is octal code 012.
2104 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
2106 @cindex @code{\r} (carriage return character)
2107 @cindex carriage return (@code{\r})
2109 Mnemonic for carriage-Return; for ASCII this is octal code 015.
2112 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
2113 @c other assemblers.
2115 @cindex @code{\t} (tab)
2116 @cindex tab (@code{\t})
2118 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
2121 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
2122 @c @item \x @var{digit} @var{digit} @var{digit}
2123 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
2125 @cindex @code{\@var{ddd}} (octal character code)
2126 @cindex octal character code (@code{\@var{ddd}})
2127 @item \ @var{digit} @var{digit} @var{digit}
2128 An octal character code. The numeric code is 3 octal digits.
2129 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
2130 for example, @code{\008} has the value 010, and @code{\009} the value 011.
2132 @cindex @code{\@var{xd...}} (hex character code)
2133 @cindex hex character code (@code{\@var{xd...}})
2134 @item \@code{x} @var{hex-digits...}
2135 A hex character code. All trailing hex digits are combined. Either upper or
2136 lower case @code{x} works.
2138 @cindex @code{\\} (@samp{\} character)
2139 @cindex backslash (@code{\\})
2141 Represents one @samp{\} character.
2144 @c Represents one @samp{'} (accent acute) character.
2145 @c This is needed in single character literals
2146 @c (@xref{Characters,,Character Constants}.) to represent
2149 @cindex @code{\"} (doublequote character)
2150 @cindex doublequote (@code{\"})
2152 Represents one @samp{"} character. Needed in strings to represent
2153 this character, because an unescaped @samp{"} would end the string.
2155 @item \ @var{anything-else}
2156 Any other character when escaped by @kbd{\} gives a warning, but
2157 assembles as if the @samp{\} was not present. The idea is that if
2158 you used an escape sequence you clearly didn't want the literal
2159 interpretation of the following character. However @code{@value{AS}} has no
2160 other interpretation, so @code{@value{AS}} knows it is giving you the wrong
2161 code and warns you of the fact.
2164 Which characters are escapable, and what those escapes represent,
2165 varies widely among assemblers. The current set is what we think
2166 the BSD 4.2 assembler recognizes, and is a subset of what most C
2167 compilers recognize. If you are in doubt, do not use an escape
2171 @subsubsection Characters
2173 @cindex single character constant
2174 @cindex character, single
2175 @cindex constant, single character
2176 A single character may be written as a single quote immediately
2177 followed by that character. The same escapes apply to characters as
2178 to strings. So if you want to write the character backslash, you
2179 must write @kbd{'\\} where the first @code{\} escapes the second
2180 @code{\}. As you can see, the quote is an acute accent, not a
2181 grave accent. A newline
2183 @ifclear abnormal-separator
2184 (or semicolon @samp{;})
2186 @ifset abnormal-separator
2188 (or at sign @samp{@@})
2191 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
2197 immediately following an acute accent is taken as a literal character
2198 and does not count as the end of a statement. The value of a character
2199 constant in a numeric expression is the machine's byte-wide code for
2200 that character. @code{@value{AS}} assumes your character code is ASCII:
2201 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
2204 @subsection Number Constants
2206 @cindex constants, number
2207 @cindex number constants
2208 @code{@value{AS}} distinguishes three kinds of numbers according to how they
2209 are stored in the target machine. @emph{Integers} are numbers that
2210 would fit into an @code{int} in the C language. @emph{Bignums} are
2211 integers, but they are stored in more than 32 bits. @emph{Flonums}
2212 are floating point numbers, described below.
2215 * Integers:: Integers
2220 * Bit Fields:: Bit Fields
2226 @subsubsection Integers
2228 @cindex constants, integer
2230 @cindex binary integers
2231 @cindex integers, binary
2232 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
2233 the binary digits @samp{01}.
2235 @cindex octal integers
2236 @cindex integers, octal
2237 An octal integer is @samp{0} followed by zero or more of the octal
2238 digits (@samp{01234567}).
2240 @cindex decimal integers
2241 @cindex integers, decimal
2242 A decimal integer starts with a non-zero digit followed by zero or
2243 more digits (@samp{0123456789}).
2245 @cindex hexadecimal integers
2246 @cindex integers, hexadecimal
2247 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
2248 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
2250 Integers have the usual values. To denote a negative integer, use
2251 the prefix operator @samp{-} discussed under expressions
2252 (@pxref{Prefix Ops,,Prefix Operators}).
2255 @subsubsection Bignums
2258 @cindex constants, bignum
2259 A @dfn{bignum} has the same syntax and semantics as an integer
2260 except that the number (or its negative) takes more than 32 bits to
2261 represent in binary. The distinction is made because in some places
2262 integers are permitted while bignums are not.
2265 @subsubsection Flonums
2267 @cindex floating point numbers
2268 @cindex constants, floating point
2270 @cindex precision, floating point
2271 A @dfn{flonum} represents a floating point number. The translation is
2272 indirect: a decimal floating point number from the text is converted by
2273 @code{@value{AS}} to a generic binary floating point number of more than
2274 sufficient precision. This generic floating point number is converted
2275 to a particular computer's floating point format (or formats) by a
2276 portion of @code{@value{AS}} specialized to that computer.
2278 A flonum is written by writing (in order)
2283 (@samp{0} is optional on the HPPA.)
2287 A letter, to tell @code{@value{AS}} the rest of the number is a flonum.
2289 @kbd{e} is recommended. Case is not important.
2291 @c FIXME: verify if flonum syntax really this vague for most cases
2292 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
2293 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
2296 On the H8/300, H8/500,
2298 and AMD 29K architectures, the letter must be
2299 one of the letters @samp{DFPRSX} (in upper or lower case).
2301 On the ARC, the letter must be one of the letters @samp{DFRS}
2302 (in upper or lower case).
2304 On the Intel 960 architecture, the letter must be
2305 one of the letters @samp{DFT} (in upper or lower case).
2307 On the HPPA architecture, the letter must be @samp{E} (upper case only).
2311 One of the letters @samp{DFPRSX} (in upper or lower case).
2314 One of the letters @samp{DFRS} (in upper or lower case).
2317 One of the letters @samp{DFPRSX} (in upper or lower case).
2320 The letter @samp{E} (upper case only).
2323 One of the letters @samp{DFT} (in upper or lower case).
2328 An optional sign: either @samp{+} or @samp{-}.
2331 An optional @dfn{integer part}: zero or more decimal digits.
2334 An optional @dfn{fractional part}: @samp{.} followed by zero
2335 or more decimal digits.
2338 An optional exponent, consisting of:
2342 An @samp{E} or @samp{e}.
2343 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2344 @c principle this can perfectly well be different on different targets.
2346 Optional sign: either @samp{+} or @samp{-}.
2348 One or more decimal digits.
2353 At least one of the integer part or the fractional part must be
2354 present. The floating point number has the usual base-10 value.
2356 @code{@value{AS}} does all processing using integers. Flonums are computed
2357 independently of any floating point hardware in the computer running
2362 @c Bit fields are written as a general facility but are also controlled
2363 @c by a conditional-compilation flag---which is as of now (21mar91)
2364 @c turned on only by the i960 config of GAS.
2366 @subsubsection Bit Fields
2369 @cindex constants, bit field
2370 You can also define numeric constants as @dfn{bit fields}.
2371 specify two numbers separated by a colon---
2373 @var{mask}:@var{value}
2376 @code{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2379 The resulting number is then packed
2381 @c this conditional paren in case bit fields turned on elsewhere than 960
2382 (in host-dependent byte order)
2384 into a field whose width depends on which assembler directive has the
2385 bit-field as its argument. Overflow (a result from the bitwise and
2386 requiring more binary digits to represent) is not an error; instead,
2387 more constants are generated, of the specified width, beginning with the
2388 least significant digits.@refill
2390 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2391 @code{.short}, and @code{.word} accept bit-field arguments.
2396 @chapter Sections and Relocation
2401 * Secs Background:: Background
2402 * Ld Sections:: Linker Sections
2403 * As Sections:: Assembler Internal Sections
2404 * Sub-Sections:: Sub-Sections
2408 @node Secs Background
2411 Roughly, a section is a range of addresses, with no gaps; all data
2412 ``in'' those addresses is treated the same for some particular purpose.
2413 For example there may be a ``read only'' section.
2415 @cindex linker, and assembler
2416 @cindex assembler, and linker
2417 The linker @code{@value{LD}} reads many object files (partial programs) and
2418 combines their contents to form a runnable program. When @code{@value{AS}}
2419 emits an object file, the partial program is assumed to start at address 0.
2420 @code{@value{LD}} assigns the final addresses for the partial program, so that
2421 different partial programs do not overlap. This is actually an
2422 oversimplification, but it suffices to explain how @code{@value{AS}} uses
2425 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2426 addresses. These blocks slide to their run-time addresses as rigid
2427 units; their length does not change and neither does the order of bytes
2428 within them. Such a rigid unit is called a @emph{section}. Assigning
2429 run-time addresses to sections is called @dfn{relocation}. It includes
2430 the task of adjusting mentions of object-file addresses so they refer to
2431 the proper run-time addresses.
2433 For the H8/300 and H8/500,
2434 and for the Hitachi SH,
2435 @code{@value{AS}} pads sections if needed to
2436 ensure they end on a word (sixteen bit) boundary.
2439 @cindex standard assembler sections
2440 An object file written by @code{@value{AS}} has at least three sections, any
2441 of which may be empty. These are named @dfn{text}, @dfn{data} and
2446 When it generates COFF output,
2448 @code{@value{AS}} can also generate whatever other named sections you specify
2449 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2450 If you do not use any directives that place output in the @samp{.text}
2451 or @samp{.data} sections, these sections still exist, but are empty.
2456 When @code{@value{AS}} generates SOM or ELF output for the HPPA,
2458 @code{@value{AS}} can also generate whatever other named sections you
2459 specify using the @samp{.space} and @samp{.subspace} directives. See
2460 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2461 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2462 assembler directives.
2465 Additionally, @code{@value{AS}} uses different names for the standard
2466 text, data, and bss sections when generating SOM output. Program text
2467 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2468 BSS into @samp{$BSS$}.
2472 Within the object file, the text section starts at address @code{0}, the
2473 data section follows, and the bss section follows the data section.
2476 When generating either SOM or ELF output files on the HPPA, the text
2477 section starts at address @code{0}, the data section at address
2478 @code{0x4000000}, and the bss section follows the data section.
2481 To let @code{@value{LD}} know which data changes when the sections are
2482 relocated, and how to change that data, @code{@value{AS}} also writes to the
2483 object file details of the relocation needed. To perform relocation
2484 @code{@value{LD}} must know, each time an address in the object
2488 Where in the object file is the beginning of this reference to
2491 How long (in bytes) is this reference?
2493 Which section does the address refer to? What is the numeric value of
2495 (@var{address}) @minus{} (@var{start-address of section})?
2498 Is the reference to an address ``Program-Counter relative''?
2501 @cindex addresses, format of
2502 @cindex section-relative addressing
2503 In fact, every address @code{@value{AS}} ever uses is expressed as
2505 (@var{section}) + (@var{offset into section})
2508 Further, most expressions @code{@value{AS}} computes have this section-relative
2511 (For some object formats, such as SOM for the HPPA, some expressions are
2512 symbol-relative instead.)
2515 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2516 @var{N} into section @var{secname}.''
2518 Apart from text, data and bss sections you need to know about the
2519 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2520 addresses in the absolute section remain unchanged. For example, address
2521 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2522 @code{@value{LD}}. Although the linker never arranges two partial programs'
2523 data sections with overlapping addresses after linking, @emph{by definition}
2524 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2525 part of a program is always the same address when the program is running as
2526 address @code{@{absolute@ 239@}} in any other part of the program.
2528 The idea of sections is extended to the @dfn{undefined} section. Any
2529 address whose section is unknown at assembly time is by definition
2530 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2531 Since numbers are always defined, the only way to generate an undefined
2532 address is to mention an undefined symbol. A reference to a named
2533 common block would be such a symbol: its value is unknown at assembly
2534 time so it has section @emph{undefined}.
2536 By analogy the word @emph{section} is used to describe groups of sections in
2537 the linked program. @code{@value{LD}} puts all partial programs' text
2538 sections in contiguous addresses in the linked program. It is
2539 customary to refer to the @emph{text section} of a program, meaning all
2540 the addresses of all partial programs' text sections. Likewise for
2541 data and bss sections.
2543 Some sections are manipulated by @code{@value{LD}}; others are invented for
2544 use of @code{@value{AS}} and have no meaning except during assembly.
2547 @section Linker Sections
2548 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2553 @cindex named sections
2554 @cindex sections, named
2555 @item named sections
2558 @cindex text section
2559 @cindex data section
2563 These sections hold your program. @code{@value{AS}} and @code{@value{LD}} treat them as
2564 separate but equal sections. Anything you can say of one section is
2567 When the program is running, however, it is
2568 customary for the text section to be unalterable. The
2569 text section is often shared among processes: it contains
2570 instructions, constants and the like. The data section of a running
2571 program is usually alterable: for example, C variables would be stored
2572 in the data section.
2577 This section contains zeroed bytes when your program begins running. It
2578 is used to hold uninitialized variables or common storage. The length of
2579 each partial program's bss section is important, but because it starts
2580 out containing zeroed bytes there is no need to store explicit zero
2581 bytes in the object file. The bss section was invented to eliminate
2582 those explicit zeros from object files.
2584 @cindex absolute section
2585 @item absolute section
2586 Address 0 of this section is always ``relocated'' to runtime address 0.
2587 This is useful if you want to refer to an address that @code{@value{LD}} must
2588 not change when relocating. In this sense we speak of absolute
2589 addresses being ``unrelocatable'': they do not change during relocation.
2591 @cindex undefined section
2592 @item undefined section
2593 This ``section'' is a catch-all for address references to objects not in
2594 the preceding sections.
2595 @c FIXME: ref to some other doc on obj-file formats could go here.
2598 @cindex relocation example
2599 An idealized example of three relocatable sections follows.
2601 The example uses the traditional section names @samp{.text} and @samp{.data}.
2603 Memory addresses are on the horizontal axis.
2607 @c END TEXI2ROFF-KILL
2610 partial program # 1: |ttttt|dddd|00|
2617 partial program # 2: |TTT|DDD|000|
2620 +--+---+-----+--+----+---+-----+~~
2621 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2622 +--+---+-----+--+----+---+-----+~~
2624 addresses: 0 @dots{}
2631 \line{\it Partial program \#1: \hfil}
2632 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2633 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2635 \line{\it Partial program \#2: \hfil}
2636 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2637 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2639 \line{\it linked program: \hfil}
2640 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2641 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2642 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2643 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2645 \line{\it addresses: \hfil}
2649 @c END TEXI2ROFF-KILL
2652 @section Assembler Internal Sections
2654 @cindex internal assembler sections
2655 @cindex sections in messages, internal
2656 These sections are meant only for the internal use of @code{@value{AS}}. They
2657 have no meaning at run-time. You do not really need to know about these
2658 sections for most purposes; but they can be mentioned in @code{@value{AS}}
2659 warning messages, so it might be helpful to have an idea of their
2660 meanings to @code{@value{AS}}. These sections are used to permit the
2661 value of every expression in your assembly language program to be a
2662 section-relative address.
2665 @cindex assembler internal logic error
2666 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2667 An internal assembler logic error has been found. This means there is a
2668 bug in the assembler.
2670 @cindex expr (internal section)
2672 The assembler stores complex expression internally as combinations of
2673 symbols. When it needs to represent an expression as a symbol, it puts
2674 it in the expr section.
2676 @c FIXME item transfer[t] vector preload
2677 @c FIXME item transfer[t] vector postload
2678 @c FIXME item register
2682 @section Sub-Sections
2684 @cindex numbered subsections
2685 @cindex grouping data
2691 fall into two sections: text and data.
2693 You may have separate groups of
2695 data in named sections
2699 data in named sections
2705 that you want to end up near to each other in the object file, even though they
2706 are not contiguous in the assembler source. @code{@value{AS}} allows you to
2707 use @dfn{subsections} for this purpose. Within each section, there can be
2708 numbered subsections with values from 0 to 8192. Objects assembled into the
2709 same subsection go into the object file together with other objects in the same
2710 subsection. For example, a compiler might want to store constants in the text
2711 section, but might not want to have them interspersed with the program being
2712 assembled. In this case, the compiler could issue a @samp{.text 0} before each
2713 section of code being output, and a @samp{.text 1} before each group of
2714 constants being output.
2716 Subsections are optional. If you do not use subsections, everything
2717 goes in subsection number zero.
2720 Each subsection is zero-padded up to a multiple of four bytes.
2721 (Subsections may be padded a different amount on different flavors
2722 of @code{@value{AS}}.)
2726 On the H8/300 and H8/500 platforms, each subsection is zero-padded to a word
2727 boundary (two bytes).
2728 The same is true on the Hitachi SH.
2731 @c FIXME section padding (alignment)?
2732 @c Rich Pixley says padding here depends on target obj code format; that
2733 @c doesn't seem particularly useful to say without further elaboration,
2734 @c so for now I say nothing about it. If this is a generic BFD issue,
2735 @c these paragraphs might need to vanish from this manual, and be
2736 @c discussed in BFD chapter of binutils (or some such).
2739 On the AMD 29K family, no particular padding is added to section or
2740 subsection sizes; @value{AS} forces no alignment on this platform.
2744 Subsections appear in your object file in numeric order, lowest numbered
2745 to highest. (All this to be compatible with other people's assemblers.)
2746 The object file contains no representation of subsections; @code{@value{LD}} and
2747 other programs that manipulate object files see no trace of them.
2748 They just see all your text subsections as a text section, and all your
2749 data subsections as a data section.
2751 To specify which subsection you want subsequent statements assembled
2752 into, use a numeric argument to specify it, in a @samp{.text
2753 @var{expression}} or a @samp{.data @var{expression}} statement.
2756 When generating COFF output, you
2761 can also use an extra subsection
2762 argument with arbitrary named sections: @samp{.section @var{name},
2765 @var{Expression} should be an absolute expression.
2766 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
2767 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
2768 begins in @code{text 0}. For instance:
2770 .text 0 # The default subsection is text 0 anyway.
2771 .ascii "This lives in the first text subsection. *"
2773 .ascii "But this lives in the second text subsection."
2775 .ascii "This lives in the data section,"
2776 .ascii "in the first data subsection."
2778 .ascii "This lives in the first text section,"
2779 .ascii "immediately following the asterisk (*)."
2782 Each section has a @dfn{location counter} incremented by one for every byte
2783 assembled into that section. Because subsections are merely a convenience
2784 restricted to @code{@value{AS}} there is no concept of a subsection location
2785 counter. There is no way to directly manipulate a location counter---but the
2786 @code{.align} directive changes it, and any label definition captures its
2787 current value. The location counter of the section where statements are being
2788 assembled is said to be the @dfn{active} location counter.
2791 @section bss Section
2794 @cindex common variable storage
2795 The bss section is used for local common variable storage.
2796 You may allocate address space in the bss section, but you may
2797 not dictate data to load into it before your program executes. When
2798 your program starts running, all the contents of the bss
2799 section are zeroed bytes.
2801 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
2802 @ref{Lcomm,,@code{.lcomm}}.
2804 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
2805 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
2808 When assembling for a target which supports multiple sections, such as ELF or
2809 COFF, you may switch into the @code{.bss} section and define symbols as usual;
2810 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
2811 section. Typically the section will only contain symbol definitions and
2812 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
2819 Symbols are a central concept: the programmer uses symbols to name
2820 things, the linker uses symbols to link, and the debugger uses symbols
2824 @cindex debuggers, and symbol order
2825 @emph{Warning:} @code{@value{AS}} does not place symbols in the object file in
2826 the same order they were declared. This may break some debuggers.
2831 * Setting Symbols:: Giving Symbols Other Values
2832 * Symbol Names:: Symbol Names
2833 * Dot:: The Special Dot Symbol
2834 * Symbol Attributes:: Symbol Attributes
2841 A @dfn{label} is written as a symbol immediately followed by a colon
2842 @samp{:}. The symbol then represents the current value of the
2843 active location counter, and is, for example, a suitable instruction
2844 operand. You are warned if you use the same symbol to represent two
2845 different locations: the first definition overrides any other
2849 On the HPPA, the usual form for a label need not be immediately followed by a
2850 colon, but instead must start in column zero. Only one label may be defined on
2851 a single line. To work around this, the HPPA version of @code{@value{AS}} also
2852 provides a special directive @code{.label} for defining labels more flexibly.
2855 @node Setting Symbols
2856 @section Giving Symbols Other Values
2858 @cindex assigning values to symbols
2859 @cindex symbol values, assigning
2860 A symbol can be given an arbitrary value by writing a symbol, followed
2861 by an equals sign @samp{=}, followed by an expression
2862 (@pxref{Expressions}). This is equivalent to using the @code{.set}
2863 directive. @xref{Set,,@code{.set}}.
2866 @section Symbol Names
2868 @cindex symbol names
2869 @cindex names, symbol
2870 @ifclear SPECIAL-SYMS
2871 Symbol names begin with a letter or with one of @samp{._}. On most
2872 machines, you can also use @code{$} in symbol names; exceptions are
2873 noted in @ref{Machine Dependencies}. That character may be followed by any
2874 string of digits, letters, dollar signs (unless otherwise noted in
2875 @ref{Machine Dependencies}), and underscores.
2878 For the AMD 29K family, @samp{?} is also allowed in the
2879 body of a symbol name, though not at its beginning.
2884 Symbol names begin with a letter or with one of @samp{._}. On the
2886 H8/500, you can also use @code{$} in symbol names. That character may
2887 be followed by any string of digits, letters, dollar signs (save on the
2888 H8/300), and underscores.
2892 Case of letters is significant: @code{foo} is a different symbol name
2895 Each symbol has exactly one name. Each name in an assembly language program
2896 refers to exactly one symbol. You may use that symbol name any number of times
2899 @subheading Local Symbol Names
2901 @cindex local symbol names
2902 @cindex symbol names, local
2903 @cindex temporary symbol names
2904 @cindex symbol names, temporary
2905 Local symbols help compilers and programmers use names temporarily.
2906 There are ten local symbol names, which are re-used throughout the
2907 program. You may refer to them using the names @samp{0} @samp{1}
2908 @dots{} @samp{9}. To define a local symbol, write a label of the form
2909 @samp{@b{N}:} (where @b{N} represents any digit). To refer to the most
2910 recent previous definition of that symbol write @samp{@b{N}b}, using the
2911 same digit as when you defined the label. To refer to the next
2912 definition of a local label, write @samp{@b{N}f}---where @b{N} gives you
2913 a choice of 10 forward references. The @samp{b} stands for
2914 ``backwards'' and the @samp{f} stands for ``forwards''.
2916 Local symbols are not emitted by the current @sc{gnu} C compiler.
2918 There is no restriction on how you can use these labels, but
2919 remember that at any point in the assembly you can refer to at most
2920 10 prior local labels and to at most 10 forward local labels.
2922 Local symbol names are only a notation device. They are immediately
2923 transformed into more conventional symbol names before the assembler
2924 uses them. The symbol names stored in the symbol table, appearing in
2925 error messages and optionally emitted to the object file have these
2930 All local labels begin with @samp{L}. Normally both @code{@value{AS}} and
2931 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
2932 used for symbols you are never intended to see. If you use the
2933 @samp{-L} option then @code{@value{AS}} retains these symbols in the
2934 object file. If you also instruct @code{@value{LD}} to retain these symbols,
2935 you may use them in debugging.
2938 If the label is written @samp{0:} then the digit is @samp{0}.
2939 If the label is written @samp{1:} then the digit is @samp{1}.
2940 And so on up through @samp{9:}.
2943 This unusual character is included so you do not accidentally invent
2944 a symbol of the same name. The character has ASCII value
2947 @item @emph{ordinal number}
2948 This is a serial number to keep the labels distinct. The first
2949 @samp{0:} gets the number @samp{1}; The 15th @samp{0:} gets the
2950 number @samp{15}; @emph{etc.}. Likewise for the other labels @samp{1:}
2954 For instance, the first @code{1:} is named @code{L1@kbd{C-A}1}, the 44th
2955 @code{3:} is named @code{L3@kbd{C-A}44}.
2958 @section The Special Dot Symbol
2960 @cindex dot (symbol)
2961 @cindex @code{.} (symbol)
2962 @cindex current address
2963 @cindex location counter
2964 The special symbol @samp{.} refers to the current address that
2965 @code{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
2966 .long .} defines @code{melvin} to contain its own address.
2967 Assigning a value to @code{.} is treated the same as a @code{.org}
2968 directive. Thus, the expression @samp{.=.+4} is the same as saying
2969 @ifclear no-space-dir
2978 @node Symbol Attributes
2979 @section Symbol Attributes
2981 @cindex symbol attributes
2982 @cindex attributes, symbol
2983 Every symbol has, as well as its name, the attributes ``Value'' and
2984 ``Type''. Depending on output format, symbols can also have auxiliary
2987 The detailed definitions are in @file{a.out.h}.
2990 If you use a symbol without defining it, @code{@value{AS}} assumes zero for
2991 all these attributes, and probably won't warn you. This makes the
2992 symbol an externally defined symbol, which is generally what you
2996 * Symbol Value:: Value
2997 * Symbol Type:: Type
3000 * a.out Symbols:: Symbol Attributes: @code{a.out}
3004 * a.out Symbols:: Symbol Attributes: @code{a.out}
3007 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
3012 * COFF Symbols:: Symbol Attributes for COFF
3015 * SOM Symbols:: Symbol Attributes for SOM
3022 @cindex value of a symbol
3023 @cindex symbol value
3024 The value of a symbol is (usually) 32 bits. For a symbol which labels a
3025 location in the text, data, bss or absolute sections the value is the
3026 number of addresses from the start of that section to the label.
3027 Naturally for text, data and bss sections the value of a symbol changes
3028 as @code{@value{LD}} changes section base addresses during linking. Absolute
3029 symbols' values do not change during linking: that is why they are
3032 The value of an undefined symbol is treated in a special way. If it is
3033 0 then the symbol is not defined in this assembler source file, and
3034 @code{@value{LD}} tries to determine its value from other files linked into the
3035 same program. You make this kind of symbol simply by mentioning a symbol
3036 name without defining it. A non-zero value represents a @code{.comm}
3037 common declaration. The value is how much common storage to reserve, in
3038 bytes (addresses). The symbol refers to the first address of the
3044 @cindex type of a symbol
3046 The type attribute of a symbol contains relocation (section)
3047 information, any flag settings indicating that a symbol is external, and
3048 (optionally), other information for linkers and debuggers. The exact
3049 format depends on the object-code output format in use.
3054 @c The following avoids a "widow" subsection title. @group would be
3055 @c better if it were available outside examples.
3058 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
3060 @cindex @code{b.out} symbol attributes
3061 @cindex symbol attributes, @code{b.out}
3062 These symbol attributes appear only when @code{@value{AS}} is configured for
3063 one of the Berkeley-descended object output formats---@code{a.out} or
3069 @subsection Symbol Attributes: @code{a.out}
3071 @cindex @code{a.out} symbol attributes
3072 @cindex symbol attributes, @code{a.out}
3078 @subsection Symbol Attributes: @code{a.out}
3080 @cindex @code{a.out} symbol attributes
3081 @cindex symbol attributes, @code{a.out}
3085 * Symbol Desc:: Descriptor
3086 * Symbol Other:: Other
3090 @subsubsection Descriptor
3092 @cindex descriptor, of @code{a.out} symbol
3093 This is an arbitrary 16-bit value. You may establish a symbol's
3094 descriptor value by using a @code{.desc} statement
3095 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
3099 @subsubsection Other
3101 @cindex other attribute, of @code{a.out} symbol
3102 This is an arbitrary 8-bit value. It means nothing to @code{@value{AS}}.
3107 @subsection Symbol Attributes for COFF
3109 @cindex COFF symbol attributes
3110 @cindex symbol attributes, COFF
3112 The COFF format supports a multitude of auxiliary symbol attributes;
3113 like the primary symbol attributes, they are set between @code{.def} and
3114 @code{.endef} directives.
3116 @subsubsection Primary Attributes
3118 @cindex primary attributes, COFF symbols
3119 The symbol name is set with @code{.def}; the value and type,
3120 respectively, with @code{.val} and @code{.type}.
3122 @subsubsection Auxiliary Attributes
3124 @cindex auxiliary attributes, COFF symbols
3125 The @code{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
3126 @code{.size}, and @code{.tag} can generate auxiliary symbol table
3127 information for COFF.
3132 @subsection Symbol Attributes for SOM
3134 @cindex SOM symbol attributes
3135 @cindex symbol attributes, SOM
3137 The SOM format for the HPPA supports a multitude of symbol attributes set with
3138 the @code{.EXPORT} and @code{.IMPORT} directives.
3140 The attributes are described in @cite{HP9000 Series 800 Assembly
3141 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
3142 @code{EXPORT} assembler directive documentation.
3146 @chapter Expressions
3150 @cindex numeric values
3151 An @dfn{expression} specifies an address or numeric value.
3152 Whitespace may precede and/or follow an expression.
3154 The result of an expression must be an absolute number, or else an offset into
3155 a particular section. If an expression is not absolute, and there is not
3156 enough information when @code{@value{AS}} sees the expression to know its
3157 section, a second pass over the source program might be necessary to interpret
3158 the expression---but the second pass is currently not implemented.
3159 @code{@value{AS}} aborts with an error message in this situation.
3162 * Empty Exprs:: Empty Expressions
3163 * Integer Exprs:: Integer Expressions
3167 @section Empty Expressions
3169 @cindex empty expressions
3170 @cindex expressions, empty
3171 An empty expression has no value: it is just whitespace or null.
3172 Wherever an absolute expression is required, you may omit the
3173 expression, and @code{@value{AS}} assumes a value of (absolute) 0. This
3174 is compatible with other assemblers.
3177 @section Integer Expressions
3179 @cindex integer expressions
3180 @cindex expressions, integer
3181 An @dfn{integer expression} is one or more @emph{arguments} delimited
3182 by @emph{operators}.
3185 * Arguments:: Arguments
3186 * Operators:: Operators
3187 * Prefix Ops:: Prefix Operators
3188 * Infix Ops:: Infix Operators
3192 @subsection Arguments
3194 @cindex expression arguments
3195 @cindex arguments in expressions
3196 @cindex operands in expressions
3197 @cindex arithmetic operands
3198 @dfn{Arguments} are symbols, numbers or subexpressions. In other
3199 contexts arguments are sometimes called ``arithmetic operands''. In
3200 this manual, to avoid confusing them with the ``instruction operands'' of
3201 the machine language, we use the term ``argument'' to refer to parts of
3202 expressions only, reserving the word ``operand'' to refer only to machine
3203 instruction operands.
3205 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
3206 @var{section} is one of text, data, bss, absolute,
3207 or undefined. @var{NNN} is a signed, 2's complement 32 bit
3210 Numbers are usually integers.
3212 A number can be a flonum or bignum. In this case, you are warned
3213 that only the low order 32 bits are used, and @code{@value{AS}} pretends
3214 these 32 bits are an integer. You may write integer-manipulating
3215 instructions that act on exotic constants, compatible with other
3218 @cindex subexpressions
3219 Subexpressions are a left parenthesis @samp{(} followed by an integer
3220 expression, followed by a right parenthesis @samp{)}; or a prefix
3221 operator followed by an argument.
3224 @subsection Operators
3226 @cindex operators, in expressions
3227 @cindex arithmetic functions
3228 @cindex functions, in expressions
3229 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
3230 operators are followed by an argument. Infix operators appear
3231 between their arguments. Operators may be preceded and/or followed by
3235 @subsection Prefix Operator
3237 @cindex prefix operators
3238 @code{@value{AS}} has the following @dfn{prefix operators}. They each take
3239 one argument, which must be absolute.
3241 @c the tex/end tex stuff surrounding this small table is meant to make
3242 @c it align, on the printed page, with the similar table in the next
3243 @c section (which is inside an enumerate).
3245 \global\advance\leftskip by \itemindent
3250 @dfn{Negation}. Two's complement negation.
3252 @dfn{Complementation}. Bitwise not.
3256 \global\advance\leftskip by -\itemindent
3260 @subsection Infix Operators
3262 @cindex infix operators
3263 @cindex operators, permitted arguments
3264 @dfn{Infix operators} take two arguments, one on either side. Operators
3265 have precedence, but operations with equal precedence are performed left
3266 to right. Apart from @code{+} or @code{-}, both arguments must be
3267 absolute, and the result is absolute.
3270 @cindex operator precedence
3271 @cindex precedence of operators
3278 @dfn{Multiplication}.
3281 @dfn{Division}. Truncation is the same as the C operator @samp{/}
3288 @dfn{Shift Left}. Same as the C operator @samp{<<}.
3292 @dfn{Shift Right}. Same as the C operator @samp{>>}.
3296 Intermediate precedence
3301 @dfn{Bitwise Inclusive Or}.
3307 @dfn{Bitwise Exclusive Or}.
3310 @dfn{Bitwise Or Not}.
3317 @cindex addition, permitted arguments
3318 @cindex plus, permitted arguments
3319 @cindex arguments for addition
3321 @dfn{Addition}. If either argument is absolute, the result has the section of
3322 the other argument. You may not add together arguments from different
3325 @cindex subtraction, permitted arguments
3326 @cindex minus, permitted arguments
3327 @cindex arguments for subtraction
3329 @dfn{Subtraction}. If the right argument is absolute, the
3330 result has the section of the left argument.
3331 If both arguments are in the same section, the result is absolute.
3332 You may not subtract arguments from different sections.
3333 @c FIXME is there still something useful to say about undefined - undefined ?
3337 In short, it's only meaningful to add or subtract the @emph{offsets} in an
3338 address; you can only have a defined section in one of the two arguments.
3341 @chapter Assembler Directives
3343 @cindex directives, machine independent
3344 @cindex pseudo-ops, machine independent
3345 @cindex machine independent directives
3346 All assembler directives have names that begin with a period (@samp{.}).
3347 The rest of the name is letters, usually in lower case.
3349 This chapter discusses directives that are available regardless of the
3350 target machine configuration for the @sc{gnu} assembler.
3352 Some machine configurations provide additional directives.
3353 @xref{Machine Dependencies}.
3356 @ifset machine-directives
3357 @xref{Machine Dependencies} for additional directives.
3362 * Abort:: @code{.abort}
3364 * ABORT:: @code{.ABORT}
3367 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3368 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3369 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3370 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3371 * Byte:: @code{.byte @var{expressions}}
3372 * Comm:: @code{.comm @var{symbol} , @var{length} }
3373 * Data:: @code{.data @var{subsection}}
3375 * Def:: @code{.def @var{name}}
3378 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3384 * Double:: @code{.double @var{flonums}}
3385 * Eject:: @code{.eject}
3386 * Else:: @code{.else}
3387 * Elseif:: @code{.elseif}
3390 * Endef:: @code{.endef}
3393 * Endfunc:: @code{.endfunc}
3394 * Endif:: @code{.endif}
3395 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3396 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3398 * Exitm:: @code{.exitm}
3399 * Extern:: @code{.extern}
3400 * Fail:: @code{.fail}
3401 @ifclear no-file-dir
3402 * File:: @code{.file @var{string}}
3405 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3406 * Float:: @code{.float @var{flonums}}
3407 * Func:: @code{.func}
3408 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3410 * Hidden:: @code{.hidden @var{names}}
3413 * hword:: @code{.hword @var{expressions}}
3414 * Ident:: @code{.ident}
3415 * If:: @code{.if @var{absolute expression}}
3416 * Include:: @code{.include "@var{file}"}
3417 * Int:: @code{.int @var{expressions}}
3419 * Internal:: @code{.internal @var{names}}
3422 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3423 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3424 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3425 * Lflags:: @code{.lflags}
3426 @ifclear no-line-dir
3427 * Line:: @code{.line @var{line-number}}
3430 * Ln:: @code{.ln @var{line-number}}
3431 * Linkonce:: @code{.linkonce [@var{type}]}
3432 * List:: @code{.list}
3433 * Long:: @code{.long @var{expressions}}
3435 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3438 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3439 * MRI:: @code{.mri @var{val}}
3440 * Nolist:: @code{.nolist}
3441 * Octa:: @code{.octa @var{bignums}}
3442 * Org:: @code{.org @var{new-lc} , @var{fill}}
3443 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3445 * PopSection:: @code{.popsection}
3446 * Previous:: @code{.previous}
3449 * Print:: @code{.print @var{string}}
3451 * Protected:: @code{.protected @var{names}}
3454 * Psize:: @code{.psize @var{lines}, @var{columns}}
3455 * Purgem:: @code{.purgem @var{name}}
3457 * PushSection:: @code{.pushsection @var{name}}
3460 * Quad:: @code{.quad @var{bignums}}
3461 * Rept:: @code{.rept @var{count}}
3462 * Sbttl:: @code{.sbttl "@var{subheading}"}
3464 * Scl:: @code{.scl @var{class}}
3465 * Section:: @code{.section @var{name}, @var{subsection}}
3468 * Set:: @code{.set @var{symbol}, @var{expression}}
3469 * Short:: @code{.short @var{expressions}}
3470 * Single:: @code{.single @var{flonums}}
3471 * Size:: @code{.size [@var{name} , @var{expression}]}
3472 * Skip:: @code{.skip @var{size} , @var{fill}}
3473 * Sleb128:: @code{.sleb128 @var{expressions}}
3474 * Space:: @code{.space @var{size} , @var{fill}}
3476 * Stab:: @code{.stabd, .stabn, .stabs}
3479 * String:: @code{.string "@var{str}"}
3480 * Struct:: @code{.struct @var{expression}}
3482 * SubSection:: @code{.subsection}
3483 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3487 * Tag:: @code{.tag @var{structname}}
3490 * Text:: @code{.text @var{subsection}}
3491 * Title:: @code{.title "@var{heading}"}
3492 * Type:: @code{.type <@var{int} | @var{name} , @var{type description}>}
3493 * Uleb128:: @code{.uleb128 @var{expressions}}
3495 * Val:: @code{.val @var{addr}}
3499 * Version:: @code{.version "@var{string}"}
3500 * VTableEntry:: @code{.vtable_entry @var{table}, @var{offset}}
3501 * VTableInherit:: @code{.vtable_inherit @var{child}, @var{parent}}
3502 * Weak:: @code{.weak @var{names}}
3505 * Word:: @code{.word @var{expressions}}
3506 * Deprecated:: Deprecated Directives
3510 @section @code{.abort}
3512 @cindex @code{abort} directive
3513 @cindex stopping the assembly
3514 This directive stops the assembly immediately. It is for
3515 compatibility with other assemblers. The original idea was that the
3516 assembly language source would be piped into the assembler. If the sender
3517 of the source quit, it could use this directive tells @code{@value{AS}} to
3518 quit also. One day @code{.abort} will not be supported.
3522 @section @code{.ABORT}
3524 @cindex @code{ABORT} directive
3525 When producing COFF output, @code{@value{AS}} accepts this directive as a
3526 synonym for @samp{.abort}.
3529 When producing @code{b.out} output, @code{@value{AS}} accepts this directive,
3535 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3537 @cindex padding the location counter
3538 @cindex @code{align} directive
3539 Pad the location counter (in the current subsection) to a particular storage
3540 boundary. The first expression (which must be absolute) is the alignment
3541 required, as described below.
3543 The second expression (also absolute) gives the fill value to be stored in the
3544 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3545 padding bytes are normally zero. However, on some systems, if the section is
3546 marked as containing code and the fill value is omitted, the space is filled
3547 with no-op instructions.
3549 The third expression is also absolute, and is also optional. If it is present,
3550 it is the maximum number of bytes that should be skipped by this alignment
3551 directive. If doing the alignment would require skipping more bytes than the
3552 specified maximum, then the alignment is not done at all. You can omit the
3553 fill value (the second argument) entirely by simply using two commas after the
3554 required alignment; this can be useful if you want the alignment to be filled
3555 with no-op instructions when appropriate.
3557 The way the required alignment is specified varies from system to system.
3558 For the a29k, hppa, m68k, m88k, w65, sparc, and Hitachi SH, and i386 using ELF
3560 the first expression is the
3561 alignment request in bytes. For example @samp{.align 8} advances
3562 the location counter until it is a multiple of 8. If the location counter
3563 is already a multiple of 8, no change is needed.
3565 For other systems, including the i386 using a.out format, and the arm and
3566 strongarm, it is the
3567 number of low-order zero bits the location counter must have after
3568 advancement. For example @samp{.align 3} advances the location
3569 counter until it a multiple of 8. If the location counter is already a
3570 multiple of 8, no change is needed.
3572 This inconsistency is due to the different behaviors of the various
3573 native assemblers for these systems which GAS must emulate.
3574 GAS also provides @code{.balign} and @code{.p2align} directives,
3575 described later, which have a consistent behavior across all
3576 architectures (but are specific to GAS).
3579 @section @code{.ascii "@var{string}"}@dots{}
3581 @cindex @code{ascii} directive
3582 @cindex string literals
3583 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3584 separated by commas. It assembles each string (with no automatic
3585 trailing zero byte) into consecutive addresses.
3588 @section @code{.asciz "@var{string}"}@dots{}
3590 @cindex @code{asciz} directive
3591 @cindex zero-terminated strings
3592 @cindex null-terminated strings
3593 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3594 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3597 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3599 @cindex padding the location counter given number of bytes
3600 @cindex @code{balign} directive
3601 Pad the location counter (in the current subsection) to a particular
3602 storage boundary. The first expression (which must be absolute) is the
3603 alignment request in bytes. For example @samp{.balign 8} advances
3604 the location counter until it is a multiple of 8. If the location counter
3605 is already a multiple of 8, no change is needed.
3607 The second expression (also absolute) gives the fill value to be stored in the
3608 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3609 padding bytes are normally zero. However, on some systems, if the section is
3610 marked as containing code and the fill value is omitted, the space is filled
3611 with no-op instructions.
3613 The third expression is also absolute, and is also optional. If it is present,
3614 it is the maximum number of bytes that should be skipped by this alignment
3615 directive. If doing the alignment would require skipping more bytes than the
3616 specified maximum, then the alignment is not done at all. You can omit the
3617 fill value (the second argument) entirely by simply using two commas after the
3618 required alignment; this can be useful if you want the alignment to be filled
3619 with no-op instructions when appropriate.
3621 @cindex @code{balignw} directive
3622 @cindex @code{balignl} directive
3623 The @code{.balignw} and @code{.balignl} directives are variants of the
3624 @code{.balign} directive. The @code{.balignw} directive treats the fill
3625 pattern as a two byte word value. The @code{.balignl} directives treats the
3626 fill pattern as a four byte longword value. For example, @code{.balignw
3627 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
3628 filled in with the value 0x368d (the exact placement of the bytes depends upon
3629 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
3633 @section @code{.byte @var{expressions}}
3635 @cindex @code{byte} directive
3636 @cindex integers, one byte
3637 @code{.byte} expects zero or more expressions, separated by commas.
3638 Each expression is assembled into the next byte.
3641 @section @code{.comm @var{symbol} , @var{length} }
3643 @cindex @code{comm} directive
3644 @cindex symbol, common
3645 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
3646 common symbol in one object file may be merged with a defined or common symbol
3647 of the same name in another object file. If @code{@value{LD}} does not see a
3648 definition for the symbol--just one or more common symbols--then it will
3649 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
3650 absolute expression. If @code{@value{LD}} sees multiple common symbols with
3651 the same name, and they do not all have the same size, it will allocate space
3652 using the largest size.
3655 When using ELF, the @code{.comm} directive takes an optional third argument.
3656 This is the desired alignment of the symbol, specified as a byte boundary (for
3657 example, an alignment of 16 means that the least significant 4 bits of the
3658 address should be zero). The alignment must be an absolute expression, and it
3659 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
3660 for the common symbol, it will use the alignment when placing the symbol. If
3661 no alignment is specified, @code{@value{AS}} will set the alignment to the
3662 largest power of two less than or equal to the size of the symbol, up to a
3667 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
3668 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
3672 @section @code{.data @var{subsection}}
3674 @cindex @code{data} directive
3675 @code{.data} tells @code{@value{AS}} to assemble the following statements onto the
3676 end of the data subsection numbered @var{subsection} (which is an
3677 absolute expression). If @var{subsection} is omitted, it defaults
3682 @section @code{.def @var{name}}
3684 @cindex @code{def} directive
3685 @cindex COFF symbols, debugging
3686 @cindex debugging COFF symbols
3687 Begin defining debugging information for a symbol @var{name}; the
3688 definition extends until the @code{.endef} directive is encountered.
3691 This directive is only observed when @code{@value{AS}} is configured for COFF
3692 format output; when producing @code{b.out}, @samp{.def} is recognized,
3699 @section @code{.desc @var{symbol}, @var{abs-expression}}
3701 @cindex @code{desc} directive
3702 @cindex COFF symbol descriptor
3703 @cindex symbol descriptor, COFF
3704 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
3705 to the low 16 bits of an absolute expression.
3708 The @samp{.desc} directive is not available when @code{@value{AS}} is
3709 configured for COFF output; it is only for @code{a.out} or @code{b.out}
3710 object format. For the sake of compatibility, @code{@value{AS}} accepts
3711 it, but produces no output, when configured for COFF.
3717 @section @code{.dim}
3719 @cindex @code{dim} directive
3720 @cindex COFF auxiliary symbol information
3721 @cindex auxiliary symbol information, COFF
3722 This directive is generated by compilers to include auxiliary debugging
3723 information in the symbol table. It is only permitted inside
3724 @code{.def}/@code{.endef} pairs.
3727 @samp{.dim} is only meaningful when generating COFF format output; when
3728 @code{@value{AS}} is generating @code{b.out}, it accepts this directive but
3734 @section @code{.double @var{flonums}}
3736 @cindex @code{double} directive
3737 @cindex floating point numbers (double)
3738 @code{.double} expects zero or more flonums, separated by commas. It
3739 assembles floating point numbers.
3741 The exact kind of floating point numbers emitted depends on how
3742 @code{@value{AS}} is configured. @xref{Machine Dependencies}.
3746 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
3747 in @sc{ieee} format.
3752 @section @code{.eject}
3754 @cindex @code{eject} directive
3755 @cindex new page, in listings
3756 @cindex page, in listings
3757 @cindex listing control: new page
3758 Force a page break at this point, when generating assembly listings.
3761 @section @code{.else}
3763 @cindex @code{else} directive
3764 @code{.else} is part of the @code{@value{AS}} support for conditional
3765 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
3766 of code to be assembled if the condition for the preceding @code{.if}
3770 @section @code{.elseif}
3772 @cindex @code{elseif} directive
3773 @code{.elseif} is part of the @code{@value{AS}} support for conditional
3774 assembly; @pxref{If,,@code{.if}}. It is shorthand for beginning a new
3775 @code{.if} block that would otherwise fill the entire @code{.else} section.
3778 @section @code{.end}
3780 @cindex @code{end} directive
3781 @code{.end} marks the end of the assembly file. @code{@value{AS}} does not
3782 process anything in the file past the @code{.end} directive.
3786 @section @code{.endef}
3788 @cindex @code{endef} directive
3789 This directive flags the end of a symbol definition begun with
3793 @samp{.endef} is only meaningful when generating COFF format output; if
3794 @code{@value{AS}} is configured to generate @code{b.out}, it accepts this
3795 directive but ignores it.
3800 @section @code{.endfunc}
3801 @cindex @code{endfunc} directive
3802 @code{.endfunc} marks the end of a function specified with @code{.func}.
3805 @section @code{.endif}
3807 @cindex @code{endif} directive
3808 @code{.endif} is part of the @code{@value{AS}} support for conditional assembly;
3809 it marks the end of a block of code that is only assembled
3810 conditionally. @xref{If,,@code{.if}}.
3813 @section @code{.equ @var{symbol}, @var{expression}}
3815 @cindex @code{equ} directive
3816 @cindex assigning values to symbols
3817 @cindex symbols, assigning values to
3818 This directive sets the value of @var{symbol} to @var{expression}.
3819 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
3822 The syntax for @code{equ} on the HPPA is
3823 @samp{@var{symbol} .equ @var{expression}}.
3827 @section @code{.equiv @var{symbol}, @var{expression}}
3828 @cindex @code{equiv} directive
3829 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
3830 the assembler will signal an error if @var{symbol} is already defined.
3832 Except for the contents of the error message, this is roughly equivalent to
3841 @section @code{.err}
3842 @cindex @code{err} directive
3843 If @code{@value{AS}} assembles a @code{.err} directive, it will print an error
3844 message and, unless the @code{-Z} option was used, it will not generate an
3845 object file. This can be used to signal error an conditionally compiled code.
3848 @section @code{.exitm}
3849 Exit early from the current macro definition. @xref{Macro}.
3852 @section @code{.extern}
3854 @cindex @code{extern} directive
3855 @code{.extern} is accepted in the source program---for compatibility
3856 with other assemblers---but it is ignored. @code{@value{AS}} treats
3857 all undefined symbols as external.
3860 @section @code{.fail @var{expression}}
3862 @cindex @code{fail} directive
3863 Generates an error or a warning. If the value of the @var{expression} is 500
3864 or more, @code{@value{AS}} will print a warning message. If the value is less
3865 than 500, @code{@value{AS}} will print an error message. The message will
3866 include the value of @var{expression}. This can occasionally be useful inside
3867 complex nested macros or conditional assembly.
3869 @ifclear no-file-dir
3871 @section @code{.file @var{string}}
3873 @cindex @code{file} directive
3874 @cindex logical file name
3875 @cindex file name, logical
3876 @code{.file} tells @code{@value{AS}} that we are about to start a new logical
3877 file. @var{string} is the new file name. In general, the filename is
3878 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
3879 to specify an empty file name, you must give the quotes--@code{""}. This
3880 statement may go away in future: it is only recognized to be compatible with
3881 old @code{@value{AS}} programs.
3883 In some configurations of @code{@value{AS}}, @code{.file} has already been
3884 removed to avoid conflicts with other assemblers. @xref{Machine Dependencies}.
3889 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
3891 @cindex @code{fill} directive
3892 @cindex writing patterns in memory
3893 @cindex patterns, writing in memory
3894 @var{repeat}, @var{size} and @var{value} are absolute expressions.
3895 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
3896 may be zero or more. @var{Size} may be zero or more, but if it is
3897 more than 8, then it is deemed to have the value 8, compatible with
3898 other people's assemblers. The contents of each @var{repeat} bytes
3899 is taken from an 8-byte number. The highest order 4 bytes are
3900 zero. The lowest order 4 bytes are @var{value} rendered in the
3901 byte-order of an integer on the computer @code{@value{AS}} is assembling for.
3902 Each @var{size} bytes in a repetition is taken from the lowest order
3903 @var{size} bytes of this number. Again, this bizarre behavior is
3904 compatible with other people's assemblers.
3906 @var{size} and @var{value} are optional.
3907 If the second comma and @var{value} are absent, @var{value} is
3908 assumed zero. If the first comma and following tokens are absent,
3909 @var{size} is assumed to be 1.
3912 @section @code{.float @var{flonums}}
3914 @cindex floating point numbers (single)
3915 @cindex @code{float} directive
3916 This directive assembles zero or more flonums, separated by commas. It
3917 has the same effect as @code{.single}.
3919 The exact kind of floating point numbers emitted depends on how
3920 @code{@value{AS}} is configured.
3921 @xref{Machine Dependencies}.
3925 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
3926 in @sc{ieee} format.
3931 @section @code{.func @var{name}[,@var{label}]}
3932 @cindex @code{func} directive
3933 @code{.func} emits debugging information to denote function @var{name}, and
3934 is ignored unless the file is assembled with debugging enabled.
3935 Only @samp{--gstabs} is currently supported.
3936 @var{label} is the entry point of the function and if omitted @var{name}
3937 prepended with the @samp{leading char} is used.
3938 @samp{leading char} is usually @code{_} or nothing, depending on the target.
3939 All functions are currently defined to have @code{void} return type.
3940 The function must be terminated with @code{.endfunc}.
3943 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3945 @cindex @code{global} directive
3946 @cindex symbol, making visible to linker
3947 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
3948 @var{symbol} in your partial program, its value is made available to
3949 other partial programs that are linked with it. Otherwise,
3950 @var{symbol} takes its attributes from a symbol of the same name
3951 from another file linked into the same program.
3953 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
3954 compatibility with other assemblers.
3957 On the HPPA, @code{.global} is not always enough to make it accessible to other
3958 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
3959 @xref{HPPA Directives,, HPPA Assembler Directives}.
3964 @section @code{.hidden @var{names}}
3966 @cindex @code{.hidden} directive
3968 This one of the ELF visibility directives. The other two are
3969 @code{.internal} (@pxref{Internal,,@code{.internal}}) and
3970 @code{.protected} (@pxref{Protected,,@code{.protected}}).
3972 This directive overrides the named symbols default visibility (which is set by
3973 their binding: local, global or weak). The directive sets the visibility to
3974 @code{hidden} which means that the symbols are not visible to other components.
3975 Such symbols are always considered to be @code{protected} as well.
3979 @section @code{.hword @var{expressions}}
3981 @cindex @code{hword} directive
3982 @cindex integers, 16-bit
3983 @cindex numbers, 16-bit
3984 @cindex sixteen bit integers
3985 This expects zero or more @var{expressions}, and emits
3986 a 16 bit number for each.
3989 This directive is a synonym for @samp{.short}; depending on the target
3990 architecture, it may also be a synonym for @samp{.word}.
3994 This directive is a synonym for @samp{.short}.
3997 This directive is a synonym for both @samp{.short} and @samp{.word}.
4002 @section @code{.ident}
4004 @cindex @code{ident} directive
4005 This directive is used by some assemblers to place tags in object files.
4006 @code{@value{AS}} simply accepts the directive for source-file
4007 compatibility with such assemblers, but does not actually emit anything
4011 @section @code{.if @var{absolute expression}}
4013 @cindex conditional assembly
4014 @cindex @code{if} directive
4015 @code{.if} marks the beginning of a section of code which is only
4016 considered part of the source program being assembled if the argument
4017 (which must be an @var{absolute expression}) is non-zero. The end of
4018 the conditional section of code must be marked by @code{.endif}
4019 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
4020 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
4021 If you have several conditions to check, @code{.elseif} may be used to avoid
4022 nesting blocks if/else within each subsequent @code{.else} block.
4024 The following variants of @code{.if} are also supported:
4026 @cindex @code{ifdef} directive
4027 @item .ifdef @var{symbol}
4028 Assembles the following section of code if the specified @var{symbol}
4031 @cindex @code{ifc} directive
4032 @item .ifc @var{string1},@var{string2}
4033 Assembles the following section of code if the two strings are the same. The
4034 strings may be optionally quoted with single quotes. If they are not quoted,
4035 the first string stops at the first comma, and the second string stops at the
4036 end of the line. Strings which contain whitespace should be quoted. The
4037 string comparison is case sensitive.
4039 @cindex @code{ifeq} directive
4040 @item .ifeq @var{absolute expression}
4041 Assembles the following section of code if the argument is zero.
4043 @cindex @code{ifeqs} directive
4044 @item .ifeqs @var{string1},@var{string2}
4045 Another form of @code{.ifc}. The strings must be quoted using double quotes.
4047 @cindex @code{ifge} directive
4048 @item .ifge @var{absolute expression}
4049 Assembles the following section of code if the argument is greater than or
4052 @cindex @code{ifgt} directive
4053 @item .ifgt @var{absolute expression}
4054 Assembles the following section of code if the argument is greater than zero.
4056 @cindex @code{ifle} directive
4057 @item .ifle @var{absolute expression}
4058 Assembles the following section of code if the argument is less than or equal
4061 @cindex @code{iflt} directive
4062 @item .iflt @var{absolute expression}
4063 Assembles the following section of code if the argument is less than zero.
4065 @cindex @code{ifnc} directive
4066 @item .ifnc @var{string1},@var{string2}.
4067 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
4068 following section of code if the two strings are not the same.
4070 @cindex @code{ifndef} directive
4071 @cindex @code{ifnotdef} directive
4072 @item .ifndef @var{symbol}
4073 @itemx .ifnotdef @var{symbol}
4074 Assembles the following section of code if the specified @var{symbol}
4075 has not been defined. Both spelling variants are equivalent.
4077 @cindex @code{ifne} directive
4078 @item .ifne @var{absolute expression}
4079 Assembles the following section of code if the argument is not equal to zero
4080 (in other words, this is equivalent to @code{.if}).
4082 @cindex @code{ifnes} directive
4083 @item .ifnes @var{string1},@var{string2}
4084 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
4085 following section of code if the two strings are not the same.
4089 @section @code{.include "@var{file}"}
4091 @cindex @code{include} directive
4092 @cindex supporting files, including
4093 @cindex files, including
4094 This directive provides a way to include supporting files at specified
4095 points in your source program. The code from @var{file} is assembled as
4096 if it followed the point of the @code{.include}; when the end of the
4097 included file is reached, assembly of the original file continues. You
4098 can control the search paths used with the @samp{-I} command-line option
4099 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
4103 @section @code{.int @var{expressions}}
4105 @cindex @code{int} directive
4106 @cindex integers, 32-bit
4107 Expect zero or more @var{expressions}, of any section, separated by commas.
4108 For each expression, emit a number that, at run time, is the value of that
4109 expression. The byte order and bit size of the number depends on what kind
4110 of target the assembly is for.
4114 On the H8/500 and most forms of the H8/300, @code{.int} emits 16-bit
4115 integers. On the H8/300H and the Hitachi SH, however, @code{.int} emits
4122 @section @code{.internal @var{names}}
4124 @cindex @code{.internal} directive
4126 This one of the ELF visibility directives. The other two are
4127 @code{.hidden} (@pxref{Hidden,,@code{.hidden}}) and
4128 @code{.protected} (@pxref{Protected,,@code{.protected}}).
4130 This directive overrides the named symbols default visibility (which is set by
4131 their binding: local, global or weak). The directive sets the visibility to
4132 @code{internal} which means that the symbols are considered to be @code{hidden}
4133 (ie not visible to other components), and that some extra, processor specific
4134 processing must also be performed upon the symbols as well.
4138 @section @code{.irp @var{symbol},@var{values}}@dots{}
4140 @cindex @code{irp} directive
4141 Evaluate a sequence of statements assigning different values to @var{symbol}.
4142 The sequence of statements starts at the @code{.irp} directive, and is
4143 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
4144 set to @var{value}, and the sequence of statements is assembled. If no
4145 @var{value} is listed, the sequence of statements is assembled once, with
4146 @var{symbol} set to the null string. To refer to @var{symbol} within the
4147 sequence of statements, use @var{\symbol}.
4149 For example, assembling
4157 is equivalent to assembling
4166 @section @code{.irpc @var{symbol},@var{values}}@dots{}
4168 @cindex @code{irpc} directive
4169 Evaluate a sequence of statements assigning different values to @var{symbol}.
4170 The sequence of statements starts at the @code{.irpc} directive, and is
4171 terminated by an @code{.endr} directive. For each character in @var{value},
4172 @var{symbol} is set to the character, and the sequence of statements is
4173 assembled. If no @var{value} is listed, the sequence of statements is
4174 assembled once, with @var{symbol} set to the null string. To refer to
4175 @var{symbol} within the sequence of statements, use @var{\symbol}.
4177 For example, assembling
4185 is equivalent to assembling
4194 @section @code{.lcomm @var{symbol} , @var{length}}
4196 @cindex @code{lcomm} directive
4197 @cindex local common symbols
4198 @cindex symbols, local common
4199 Reserve @var{length} (an absolute expression) bytes for a local common
4200 denoted by @var{symbol}. The section and value of @var{symbol} are
4201 those of the new local common. The addresses are allocated in the bss
4202 section, so that at run-time the bytes start off zeroed. @var{Symbol}
4203 is not declared global (@pxref{Global,,@code{.global}}), so is normally
4204 not visible to @code{@value{LD}}.
4207 Some targets permit a third argument to be used with @code{.lcomm}. This
4208 argument specifies the desired alignment of the symbol in the bss section.
4212 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
4213 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
4217 @section @code{.lflags}
4219 @cindex @code{lflags} directive (ignored)
4220 @code{@value{AS}} accepts this directive, for compatibility with other
4221 assemblers, but ignores it.
4223 @ifclear no-line-dir
4225 @section @code{.line @var{line-number}}
4227 @cindex @code{line} directive
4231 @section @code{.ln @var{line-number}}
4233 @cindex @code{ln} directive
4235 @cindex logical line number
4237 Change the logical line number. @var{line-number} must be an absolute
4238 expression. The next line has that logical line number. Therefore any other
4239 statements on the current line (after a statement separator character) are
4240 reported as on logical line number @var{line-number} @minus{} 1. One day
4241 @code{@value{AS}} will no longer support this directive: it is recognized only
4242 for compatibility with existing assembler programs.
4246 @emph{Warning:} In the AMD29K configuration of @value{AS}, this command is
4247 not available; use the synonym @code{.ln} in that context.
4252 @ifclear no-line-dir
4253 Even though this is a directive associated with the @code{a.out} or
4254 @code{b.out} object-code formats, @code{@value{AS}} still recognizes it
4255 when producing COFF output, and treats @samp{.line} as though it
4256 were the COFF @samp{.ln} @emph{if} it is found outside a
4257 @code{.def}/@code{.endef} pair.
4259 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
4260 used by compilers to generate auxiliary symbol information for
4265 @section @code{.linkonce [@var{type}]}
4267 @cindex @code{linkonce} directive
4268 @cindex common sections
4269 Mark the current section so that the linker only includes a single copy of it.
4270 This may be used to include the same section in several different object files,
4271 but ensure that the linker will only include it once in the final output file.
4272 The @code{.linkonce} pseudo-op must be used for each instance of the section.
4273 Duplicate sections are detected based on the section name, so it should be
4276 This directive is only supported by a few object file formats; as of this
4277 writing, the only object file format which supports it is the Portable
4278 Executable format used on Windows NT.
4280 The @var{type} argument is optional. If specified, it must be one of the
4281 following strings. For example:
4285 Not all types may be supported on all object file formats.
4289 Silently discard duplicate sections. This is the default.
4292 Warn if there are duplicate sections, but still keep only one copy.
4295 Warn if any of the duplicates have different sizes.
4298 Warn if any of the duplicates do not have exactly the same contents.
4302 @section @code{.ln @var{line-number}}
4304 @cindex @code{ln} directive
4305 @ifclear no-line-dir
4306 @samp{.ln} is a synonym for @samp{.line}.
4309 Tell @code{@value{AS}} to change the logical line number. @var{line-number}
4310 must be an absolute expression. The next line has that logical
4311 line number, so any other statements on the current line (after a
4312 statement separator character @code{;}) are reported as on logical
4313 line number @var{line-number} @minus{} 1.
4316 This directive is accepted, but ignored, when @code{@value{AS}} is
4317 configured for @code{b.out}; its effect is only associated with COFF
4323 @section @code{.mri @var{val}}
4325 @cindex @code{mri} directive
4326 @cindex MRI mode, temporarily
4327 If @var{val} is non-zero, this tells @code{@value{AS}} to enter MRI mode. If
4328 @var{val} is zero, this tells @code{@value{AS}} to exit MRI mode. This change
4329 affects code assembled until the next @code{.mri} directive, or until the end
4330 of the file. @xref{M, MRI mode, MRI mode}.
4333 @section @code{.list}
4335 @cindex @code{list} directive
4336 @cindex listing control, turning on
4337 Control (in conjunction with the @code{.nolist} directive) whether or
4338 not assembly listings are generated. These two directives maintain an
4339 internal counter (which is zero initially). @code{.list} increments the
4340 counter, and @code{.nolist} decrements it. Assembly listings are
4341 generated whenever the counter is greater than zero.
4343 By default, listings are disabled. When you enable them (with the
4344 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
4345 the initial value of the listing counter is one.
4348 @section @code{.long @var{expressions}}
4350 @cindex @code{long} directive
4351 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
4354 @c no one seems to know what this is for or whether this description is
4355 @c what it really ought to do
4357 @section @code{.lsym @var{symbol}, @var{expression}}
4359 @cindex @code{lsym} directive
4360 @cindex symbol, not referenced in assembly
4361 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
4362 the hash table, ensuring it cannot be referenced by name during the
4363 rest of the assembly. This sets the attributes of the symbol to be
4364 the same as the expression value:
4366 @var{other} = @var{descriptor} = 0
4367 @var{type} = @r{(section of @var{expression})}
4368 @var{value} = @var{expression}
4371 The new symbol is not flagged as external.
4375 @section @code{.macro}
4378 The commands @code{.macro} and @code{.endm} allow you to define macros that
4379 generate assembly output. For example, this definition specifies a macro
4380 @code{sum} that puts a sequence of numbers into memory:
4383 .macro sum from=0, to=5
4392 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
4404 @item .macro @var{macname}
4405 @itemx .macro @var{macname} @var{macargs} @dots{}
4406 @cindex @code{macro} directive
4407 Begin the definition of a macro called @var{macname}. If your macro
4408 definition requires arguments, specify their names after the macro name,
4409 separated by commas or spaces. You can supply a default value for any
4410 macro argument by following the name with @samp{=@var{deflt}}. For
4411 example, these are all valid @code{.macro} statements:
4415 Begin the definition of a macro called @code{comm}, which takes no
4418 @item .macro plus1 p, p1
4419 @itemx .macro plus1 p p1
4420 Either statement begins the definition of a macro called @code{plus1},
4421 which takes two arguments; within the macro definition, write
4422 @samp{\p} or @samp{\p1} to evaluate the arguments.
4424 @item .macro reserve_str p1=0 p2
4425 Begin the definition of a macro called @code{reserve_str}, with two
4426 arguments. The first argument has a default value, but not the second.
4427 After the definition is complete, you can call the macro either as
4428 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4429 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4430 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4431 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4434 When you call a macro, you can specify the argument values either by
4435 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
4436 @samp{sum to=17, from=9}.
4439 @cindex @code{endm} directive
4440 Mark the end of a macro definition.
4443 @cindex @code{exitm} directive
4444 Exit early from the current macro definition.
4446 @cindex number of macros executed
4447 @cindex macros, count executed
4449 @code{@value{AS}} maintains a counter of how many macros it has
4450 executed in this pseudo-variable; you can copy that number to your
4451 output with @samp{\@@}, but @emph{only within a macro definition}.
4454 @item LOCAL @var{name} [ , @dots{} ]
4455 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
4456 macro syntax'' with @samp{-a} or @samp{--alternate}.} @xref{Alternate,,
4457 Alternate macro syntax}.
4459 Generate a string replacement for each of the @var{name} arguments, and
4460 replace any instances of @var{name} in each macro expansion. The
4461 replacement string is unique in the assembly, and different for each
4462 separate macro expansion. @code{LOCAL} allows you to write macros that
4463 define symbols, without fear of conflict between separate macro expansions.
4468 @section @code{.nolist}
4470 @cindex @code{nolist} directive
4471 @cindex listing control, turning off
4472 Control (in conjunction with the @code{.list} directive) whether or
4473 not assembly listings are generated. These two directives maintain an
4474 internal counter (which is zero initially). @code{.list} increments the
4475 counter, and @code{.nolist} decrements it. Assembly listings are
4476 generated whenever the counter is greater than zero.
4479 @section @code{.octa @var{bignums}}
4481 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
4482 @cindex @code{octa} directive
4483 @cindex integer, 16-byte
4484 @cindex sixteen byte integer
4485 This directive expects zero or more bignums, separated by commas. For each
4486 bignum, it emits a 16-byte integer.
4488 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
4489 hence @emph{octa}-word for 16 bytes.
4492 @section @code{.org @var{new-lc} , @var{fill}}
4494 @cindex @code{org} directive
4495 @cindex location counter, advancing
4496 @cindex advancing location counter
4497 @cindex current address, advancing
4498 Advance the location counter of the current section to
4499 @var{new-lc}. @var{new-lc} is either an absolute expression or an
4500 expression with the same section as the current subsection. That is,
4501 you can't use @code{.org} to cross sections: if @var{new-lc} has the
4502 wrong section, the @code{.org} directive is ignored. To be compatible
4503 with former assemblers, if the section of @var{new-lc} is absolute,
4504 @code{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
4505 is the same as the current subsection.
4507 @code{.org} may only increase the location counter, or leave it
4508 unchanged; you cannot use @code{.org} to move the location counter
4511 @c double negative used below "not undefined" because this is a specific
4512 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
4513 @c section. doc@cygnus.com 18feb91
4514 Because @code{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
4515 may not be undefined. If you really detest this restriction we eagerly await
4516 a chance to share your improved assembler.
4518 Beware that the origin is relative to the start of the section, not
4519 to the start of the subsection. This is compatible with other
4520 people's assemblers.
4522 When the location counter (of the current subsection) is advanced, the
4523 intervening bytes are filled with @var{fill} which should be an
4524 absolute expression. If the comma and @var{fill} are omitted,
4525 @var{fill} defaults to zero.
4528 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4530 @cindex padding the location counter given a power of two
4531 @cindex @code{p2align} directive
4532 Pad the location counter (in the current subsection) to a particular
4533 storage boundary. The first expression (which must be absolute) is the
4534 number of low-order zero bits the location counter must have after
4535 advancement. For example @samp{.p2align 3} advances the location
4536 counter until it a multiple of 8. If the location counter is already a
4537 multiple of 8, no change is needed.
4539 The second expression (also absolute) gives the fill value to be stored in the
4540 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4541 padding bytes are normally zero. However, on some systems, if the section is
4542 marked as containing code and the fill value is omitted, the space is filled
4543 with no-op instructions.
4545 The third expression is also absolute, and is also optional. If it is present,
4546 it is the maximum number of bytes that should be skipped by this alignment
4547 directive. If doing the alignment would require skipping more bytes than the
4548 specified maximum, then the alignment is not done at all. You can omit the
4549 fill value (the second argument) entirely by simply using two commas after the
4550 required alignment; this can be useful if you want the alignment to be filled
4551 with no-op instructions when appropriate.
4553 @cindex @code{p2alignw} directive
4554 @cindex @code{p2alignl} directive
4555 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
4556 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
4557 pattern as a two byte word value. The @code{.p2alignl} directives treats the
4558 fill pattern as a four byte longword value. For example, @code{.p2alignw
4559 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4560 filled in with the value 0x368d (the exact placement of the bytes depends upon
4561 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4566 @section @code{.previous}
4568 @cindex @code{.previous} directive
4569 @cindex Section Stack
4570 This is one of the ELF section stack manipulation directives. The others are
4571 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4572 @code{.pushsection} (@pxref{PushSection}), and @code{.popsection}
4573 (@pxref{PopSection}).
4575 This directive swaps the current section (and subsection) with most recently
4576 referenced section (and subsection) prior to this one. Multiple
4577 @code{.previous} directives in a row will flip between two sections (and their
4580 In terms of the section stack, this directive swaps the current section with
4581 the top section on the section stack.
4586 @section @code{.popsection}
4588 @cindex @code{.popsection} directive
4589 @cindex Section Stack
4590 This is one of the ELF section stack manipulation directives. The others are
4591 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4592 @code{.pushsection} (@pxref{PushSection}), and @code{.previous}
4595 This directive replaces the current section (and subsection) with the top
4596 section (and subsection) on the section stack. This section is popped off the
4601 @section @code{.print @var{string}}
4603 @cindex @code{print} directive
4604 @code{@value{AS}} will print @var{string} on the standard output during
4605 assembly. You must put @var{string} in double quotes.
4609 @section @code{.protected @var{names}}
4611 @cindex @code{.protected} directive
4613 This one of the ELF visibility directives. The other two are
4614 @code{.hidden} (@pxref{Hidden}) and @code{.internal} (@pxref{Internal}).
4616 This directive overrides the named symbols default visibility (which is set by
4617 their binding: local, global or weak). The directive sets the visibility to
4618 @code{protected} which means that any references to the symbols from within the
4619 components that defines them must be resolved to the definition in that
4620 component, even if a definition in another component would normally preempt
4625 @section @code{.psize @var{lines} , @var{columns}}
4627 @cindex @code{psize} directive
4628 @cindex listing control: paper size
4629 @cindex paper size, for listings
4630 Use this directive to declare the number of lines---and, optionally, the
4631 number of columns---to use for each page, when generating listings.
4633 If you do not use @code{.psize}, listings use a default line-count
4634 of 60. You may omit the comma and @var{columns} specification; the
4635 default width is 200 columns.
4637 @code{@value{AS}} generates formfeeds whenever the specified number of
4638 lines is exceeded (or whenever you explicitly request one, using
4641 If you specify @var{lines} as @code{0}, no formfeeds are generated save
4642 those explicitly specified with @code{.eject}.
4645 @section @code{.purgem @var{name}}
4647 @cindex @code{purgem} directive
4648 Undefine the macro @var{name}, so that later uses of the string will not be
4649 expanded. @xref{Macro}.
4653 @section @code{.pushsection @var{name} , @var{subsection}}
4655 @cindex @code{.pushsection} directive
4656 @cindex Section Stack
4657 This is one of the ELF section stack manipulation directives. The others are
4658 @code{.section} (@pxref{Section}), @code{.subsection} (@pxref{SubSection}),
4659 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
4662 This directive is a synonym for @code{.section}. It pushes the current section
4663 (and subsection) onto the top of the section stack, and then replaces the
4664 current section and subsection with @code{name} and @code{subsection}.
4668 @section @code{.quad @var{bignums}}
4670 @cindex @code{quad} directive
4671 @code{.quad} expects zero or more bignums, separated by commas. For
4672 each bignum, it emits
4674 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
4675 warning message; and just takes the lowest order 8 bytes of the bignum.
4676 @cindex eight-byte integer
4677 @cindex integer, 8-byte
4679 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
4680 hence @emph{quad}-word for 8 bytes.
4683 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
4684 warning message; and just takes the lowest order 16 bytes of the bignum.
4685 @cindex sixteen-byte integer
4686 @cindex integer, 16-byte
4690 @section @code{.rept @var{count}}
4692 @cindex @code{rept} directive
4693 Repeat the sequence of lines between the @code{.rept} directive and the next
4694 @code{.endr} directive @var{count} times.
4696 For example, assembling
4704 is equivalent to assembling
4713 @section @code{.sbttl "@var{subheading}"}
4715 @cindex @code{sbttl} directive
4716 @cindex subtitles for listings
4717 @cindex listing control: subtitle
4718 Use @var{subheading} as the title (third line, immediately after the
4719 title line) when generating assembly listings.
4721 This directive affects subsequent pages, as well as the current page if
4722 it appears within ten lines of the top of a page.
4726 @section @code{.scl @var{class}}
4728 @cindex @code{scl} directive
4729 @cindex symbol storage class (COFF)
4730 @cindex COFF symbol storage class
4731 Set the storage-class value for a symbol. This directive may only be
4732 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
4733 whether a symbol is static or external, or it may record further
4734 symbolic debugging information.
4737 The @samp{.scl} directive is primarily associated with COFF output; when
4738 configured to generate @code{b.out} output format, @code{@value{AS}}
4739 accepts this directive but ignores it.
4744 @section @code{.section @var{name}} (COFF version)
4746 @cindex @code{section} directive
4747 @cindex named section
4748 Use the @code{.section} directive to assemble the following code into a section
4751 This directive is only supported for targets that actually support arbitrarily
4752 named sections; on @code{a.out} targets, for example, it is not accepted, even
4753 with a standard @code{a.out} section name.
4755 For COFF targets, the @code{.section} directive is used in one of the following
4759 .section @var{name}[, "@var{flags}"]
4760 .section @var{name}[, @var{subsegment}]
4763 If the optional argument is quoted, it is taken as flags to use for the
4764 section. Each flag is a single character. The following flags are recognized:
4767 bss section (uninitialized data)
4769 section is not loaded
4779 shared section (meaningful for PE targets)
4782 If no flags are specified, the default flags depend upon the section name. If
4783 the section name is not recognized, the default will be for the section to be
4784 loaded and writable. Note the @code{n} and @code{w} flags remove attributes
4785 from the section, rather than adding them, so if they are used on their own it
4786 will be as if no flags had been specified at all.
4788 If the optional argument to the @code{.section} directive is not quoted, it is
4789 taken as a subsegment number (@pxref{Sub-Sections}).
4792 @section @code{.section @var{name}} (ELF version)
4794 @cindex @code{section} directive
4795 @cindex named section
4797 @cindex Section Stack
4798 This is one of the ELF section stack manipulation directives. The others are
4799 @code{.subsection} (@pxref{SubSection}), @code{.pushsection}
4800 (@pxref{PushSection}), @code{.popsection} (@pxref{PopSection}), and
4801 @code{.previous} (@pxref{Previous}).
4804 For ELF targets, the @code{.section} directive is used like this:
4807 .section @var{name} [, "@var{flags}"[, @@@var{type}]]
4810 The optional @var{flags} argument is a quoted string which may contain any
4811 combination of the following characters:
4814 section is allocatable
4818 section is executable
4821 The optional @var{type} argument may contain one of the following constants:
4824 section contains data
4826 section does not contain data (i.e., section only occupies space)
4829 If no flags are specified, the default flags depend upon the section name. If
4830 the section name is not recognized, the default will be for the section to have
4831 none of the above flags: it will not be allocated in memory, nor writable, nor
4832 executable. The section will contain data.
4834 For ELF targets, the assembler supports another type of @code{.section}
4835 directive for compatibility with the Solaris assembler:
4838 .section "@var{name}"[, @var{flags}...]
4841 Note that the section name is quoted. There may be a sequence of comma
4845 section is allocatable
4849 section is executable
4852 This directive replaces the current section and subsection. The replaced
4853 section and subsection are pushed onto the section stack. See the contents of
4854 the gas testsuite directory @code{gas/testsuite/gas/elf} for some examples of
4855 how this directive and the other section stack directives work.
4858 @section @code{.set @var{symbol}, @var{expression}}
4860 @cindex @code{set} directive
4861 @cindex symbol value, setting
4862 Set the value of @var{symbol} to @var{expression}. This
4863 changes @var{symbol}'s value and type to conform to
4864 @var{expression}. If @var{symbol} was flagged as external, it remains
4865 flagged (@pxref{Symbol Attributes}).
4867 You may @code{.set} a symbol many times in the same assembly.
4869 If you @code{.set} a global symbol, the value stored in the object
4870 file is the last value stored into it.
4873 The syntax for @code{set} on the HPPA is
4874 @samp{@var{symbol} .set @var{expression}}.
4878 @section @code{.short @var{expressions}}
4880 @cindex @code{short} directive
4882 @code{.short} is normally the same as @samp{.word}.
4883 @xref{Word,,@code{.word}}.
4885 In some configurations, however, @code{.short} and @code{.word} generate
4886 numbers of different lengths; @pxref{Machine Dependencies}.
4890 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
4893 This expects zero or more @var{expressions}, and emits
4894 a 16 bit number for each.
4899 @section @code{.single @var{flonums}}
4901 @cindex @code{single} directive
4902 @cindex floating point numbers (single)
4903 This directive assembles zero or more flonums, separated by commas. It
4904 has the same effect as @code{.float}.
4906 The exact kind of floating point numbers emitted depends on how
4907 @code{@value{AS}} is configured. @xref{Machine Dependencies}.
4911 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
4912 numbers in @sc{ieee} format.
4917 @section @code{.size} (COFF version)
4919 @cindex @code{size} directive
4920 This directive is generated by compilers to include auxiliary debugging
4921 information in the symbol table. It is only permitted inside
4922 @code{.def}/@code{.endef} pairs.
4925 @samp{.size} is only meaningful when generating COFF format output; when
4926 @code{@value{AS}} is generating @code{b.out}, it accepts this directive but
4930 @section @code{.size @var{name} , @var{expression}} (ELF version)
4931 @cindex @code{size} directive
4933 This directive is used to set the size associated with a symbol @var{name}.
4934 The size in bytes is computed from @var{expression} which can make use of label
4935 arithmetic. This directive is typically used to set the size of function
4939 @section @code{.sleb128 @var{expressions}}
4941 @cindex @code{sleb128} directive
4942 @var{sleb128} stands for ``signed little endian base 128.'' This is a
4943 compact, variable length representation of numbers used by the DWARF
4944 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
4946 @ifclear no-space-dir
4948 @section @code{.skip @var{size} , @var{fill}}
4950 @cindex @code{skip} directive
4951 @cindex filling memory
4952 This directive emits @var{size} bytes, each of value @var{fill}. Both
4953 @var{size} and @var{fill} are absolute expressions. If the comma and
4954 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
4958 @section @code{.space @var{size} , @var{fill}}
4960 @cindex @code{space} directive
4961 @cindex filling memory
4962 This directive emits @var{size} bytes, each of value @var{fill}. Both
4963 @var{size} and @var{fill} are absolute expressions. If the comma
4964 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
4969 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
4970 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
4971 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
4972 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
4981 @section @code{.space}
4982 @cindex @code{space} directive
4984 On the AMD 29K, this directive is ignored; it is accepted for
4985 compatibility with other AMD 29K assemblers.
4988 @emph{Warning:} In most versions of the @sc{gnu} assembler, the directive
4989 @code{.space} has the effect of @code{.block} @xref{Machine Dependencies}.
4995 @section @code{.stabd, .stabn, .stabs}
4997 @cindex symbolic debuggers, information for
4998 @cindex @code{stab@var{x}} directives
4999 There are three directives that begin @samp{.stab}.
5000 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
5001 The symbols are not entered in the @code{@value{AS}} hash table: they
5002 cannot be referenced elsewhere in the source file.
5003 Up to five fields are required:
5007 This is the symbol's name. It may contain any character except
5008 @samp{\000}, so is more general than ordinary symbol names. Some
5009 debuggers used to code arbitrarily complex structures into symbol names
5013 An absolute expression. The symbol's type is set to the low 8 bits of
5014 this expression. Any bit pattern is permitted, but @code{@value{LD}}
5015 and debuggers choke on silly bit patterns.
5018 An absolute expression. The symbol's ``other'' attribute is set to the
5019 low 8 bits of this expression.
5022 An absolute expression. The symbol's descriptor is set to the low 16
5023 bits of this expression.
5026 An absolute expression which becomes the symbol's value.
5029 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
5030 or @code{.stabs} statement, the symbol has probably already been created;
5031 you get a half-formed symbol in your object file. This is
5032 compatible with earlier assemblers!
5035 @cindex @code{stabd} directive
5036 @item .stabd @var{type} , @var{other} , @var{desc}
5038 The ``name'' of the symbol generated is not even an empty string.
5039 It is a null pointer, for compatibility. Older assemblers used a
5040 null pointer so they didn't waste space in object files with empty
5043 The symbol's value is set to the location counter,
5044 relocatably. When your program is linked, the value of this symbol
5045 is the address of the location counter when the @code{.stabd} was
5048 @cindex @code{stabn} directive
5049 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
5050 The name of the symbol is set to the empty string @code{""}.
5052 @cindex @code{stabs} directive
5053 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
5054 All five fields are specified.
5060 @section @code{.string} "@var{str}"
5062 @cindex string, copying to object file
5063 @cindex @code{string} directive
5065 Copy the characters in @var{str} to the object file. You may specify more than
5066 one string to copy, separated by commas. Unless otherwise specified for a
5067 particular machine, the assembler marks the end of each string with a 0 byte.
5068 You can use any of the escape sequences described in @ref{Strings,,Strings}.
5071 @section @code{.struct @var{expression}}
5073 @cindex @code{struct} directive
5074 Switch to the absolute section, and set the section offset to @var{expression},
5075 which must be an absolute expression. You might use this as follows:
5084 This would define the symbol @code{field1} to have the value 0, the symbol
5085 @code{field2} to have the value 4, and the symbol @code{field3} to have the
5086 value 8. Assembly would be left in the absolute section, and you would need to
5087 use a @code{.section} directive of some sort to change to some other section
5088 before further assembly.
5092 @section @code{.subsection @var{name}}
5094 @cindex @code{.subsection} directive
5095 @cindex Section Stack
5096 This is one of the ELF section stack manipulation directives. The others are
5097 @code{.section} (@pxref{Section}), @code{.pushsection} (@pxref{PushSection}),
5098 @code{.popsection} (@pxref{PopSection}), and @code{.previous}
5101 This directive replaces the current subsection with @code{name}. The current
5102 section is not changed. The replaced subsection is put onto the section stack
5103 in place of the then current top of stack subsection.
5108 @section @code{.symver}
5109 @cindex @code{symver} directive
5110 @cindex symbol versioning
5111 @cindex versions of symbols
5112 Use the @code{.symver} directive to bind symbols to specific version nodes
5113 within a source file. This is only supported on ELF platforms, and is
5114 typically used when assembling files to be linked into a shared library.
5115 There are cases where it may make sense to use this in objects to be bound
5116 into an application itself so as to override a versioned symbol from a
5119 For ELF targets, the @code{.symver} directive can be used like this:
5121 .symver @var{name}, @var{name2@@nodename}
5123 If the symbol @var{name} is defined within the file
5124 being assembled, the @code{.symver} directive effectively creates a symbol
5125 alias with the name @var{name2@@nodename}, and in fact the main reason that we
5126 just don't try and create a regular alias is that the @var{@@} character isn't
5127 permitted in symbol names. The @var{name2} part of the name is the actual name
5128 of the symbol by which it will be externally referenced. The name @var{name}
5129 itself is merely a name of convenience that is used so that it is possible to
5130 have definitions for multiple versions of a function within a single source
5131 file, and so that the compiler can unambiguously know which version of a
5132 function is being mentioned. The @var{nodename} portion of the alias should be
5133 the name of a node specified in the version script supplied to the linker when
5134 building a shared library. If you are attempting to override a versioned
5135 symbol from a shared library, then @var{nodename} should correspond to the
5136 nodename of the symbol you are trying to override.
5138 If the symbol @var{name} is not defined within the file being assembled, all
5139 references to @var{name} will be changed to @var{name2@@nodename}. If no
5140 reference to @var{name} is made, @var{name2@@nodename} will be removed from the
5143 Another usage of the @code{.symver} directive is:
5145 .symver @var{name}, @var{name2@@@@nodename}
5147 In this case, the symbol @var{name} must exist and be defined within
5148 the file being assembled. It is similar to @var{name2@@nodename}. The
5149 difference is @var{name2@@@@nodename} will also be used to resolve
5150 references to @var{name2} by the linker.
5152 The third usage of the @code{.symver} directive is:
5154 .symver @var{name}, @var{name2@@@@@@nodename}
5156 When @var{name} is not defined within the
5157 file being assembled, it is treated as @var{name2@@nodename}. When
5158 @var{name} is defined within the file being assembled, the symbol
5159 name, @var{name}, will be changed to @var{name2@@@@nodename}.
5164 @section @code{.tag @var{structname}}
5166 @cindex COFF structure debugging
5167 @cindex structure debugging, COFF
5168 @cindex @code{tag} directive
5169 This directive is generated by compilers to include auxiliary debugging
5170 information in the symbol table. It is only permitted inside
5171 @code{.def}/@code{.endef} pairs. Tags are used to link structure
5172 definitions in the symbol table with instances of those structures.
5175 @samp{.tag} is only used when generating COFF format output; when
5176 @code{@value{AS}} is generating @code{b.out}, it accepts this directive but
5182 @section @code{.text @var{subsection}}
5184 @cindex @code{text} directive
5185 Tells @code{@value{AS}} to assemble the following statements onto the end of
5186 the text subsection numbered @var{subsection}, which is an absolute
5187 expression. If @var{subsection} is omitted, subsection number zero
5191 @section @code{.title "@var{heading}"}
5193 @cindex @code{title} directive
5194 @cindex listing control: title line
5195 Use @var{heading} as the title (second line, immediately after the
5196 source file name and pagenumber) when generating assembly listings.
5198 This directive affects subsequent pages, as well as the current page if
5199 it appears within ten lines of the top of a page.
5202 @section @code{.type @var{int}} (COFF version)
5204 @cindex COFF symbol type
5205 @cindex symbol type, COFF
5206 @cindex @code{type} directive
5207 This directive, permitted only within @code{.def}/@code{.endef} pairs,
5208 records the integer @var{int} as the type attribute of a symbol table entry.
5211 @samp{.type} is associated only with COFF format output; when
5212 @code{@value{AS}} is configured for @code{b.out} output, it accepts this
5213 directive but ignores it.
5216 @section @code{.type @var{name} , @var{type description}} (ELF version)
5218 @cindex ELF symbol type
5219 @cindex symbol type, ELF
5220 @cindex @code{type} directive
5221 This directive is used to set the type of symbol @var{name} to be either a
5222 function symbol or an object symbol. There are five different syntaxes
5223 supported for the @var{type description} field, in order to provide
5224 compatibility with various other assemblers. The syntaxes supported are:
5227 .type <name>,#function
5228 .type <name>,#object
5230 .type <name>,@@function
5231 .type <name>,@@object
5233 .type <name>,%function
5234 .type <name>,%object
5236 .type <name>,"function"
5237 .type <name>,"object"
5239 .type <name> STT_FUNCTION
5240 .type <name> STT_OBJECT
5244 @section @code{.uleb128 @var{expressions}}
5246 @cindex @code{uleb128} directive
5247 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
5248 compact, variable length representation of numbers used by the DWARF
5249 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
5253 @section @code{.val @var{addr}}
5255 @cindex @code{val} directive
5256 @cindex COFF value attribute
5257 @cindex value attribute, COFF
5258 This directive, permitted only within @code{.def}/@code{.endef} pairs,
5259 records the address @var{addr} as the value attribute of a symbol table
5263 @samp{.val} is used only for COFF output; when @code{@value{AS}} is
5264 configured for @code{b.out}, it accepts this directive but ignores it.
5270 @section @code{.version "@var{string}"}
5272 @cindex @code{.version}
5273 This directive creates a @code{.note} section and places into it an ELF
5274 formatted note of type NT_VERSION. The note's name is set to @code{string}.
5279 @section @code{.vtable_entry @var{table}, @var{offset}}
5281 @cindex @code{.vtable_entry}
5282 This directive finds or creates a symbol @code{table} and creates a
5283 @code{VTABLE_ENTRY} relocation for it with an addend of @code{offset}.
5286 @section @code{.vtable_inherit @var{child}, @var{parent}}
5288 @cindex @code{.vtable_inherit}
5289 This directive finds the symbol @code{child} and finds or creates the symbol
5290 @code{parent} and then creates a @code{VTABLE_INHERIT} relocation for the
5291 parent whose addend is the value of the child symbol. As a special case the
5292 parent name of @code{0} is treated as refering the @code{*ABS*} section.
5297 @section @code{.weak @var{names}}
5299 @cindex @code{.weak}
5300 This directive sets the weak attribute on the comma separated list of symbol
5301 @code{names}. If the symbols do not already exist, they will be created.
5305 @section @code{.word @var{expressions}}
5307 @cindex @code{word} directive
5308 This directive expects zero or more @var{expressions}, of any section,
5309 separated by commas.
5312 For each expression, @code{@value{AS}} emits a 32-bit number.
5315 For each expression, @code{@value{AS}} emits a 16-bit number.
5320 The size of the number emitted, and its byte order,
5321 depend on what target computer the assembly is for.
5324 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
5325 @c happen---32-bit addressability, period; no long/short jumps.
5326 @ifset DIFF-TBL-KLUGE
5327 @cindex difference tables altered
5328 @cindex altered difference tables
5330 @emph{Warning: Special Treatment to support Compilers}
5334 Machines with a 32-bit address space, but that do less than 32-bit
5335 addressing, require the following special treatment. If the machine of
5336 interest to you does 32-bit addressing (or doesn't require it;
5337 @pxref{Machine Dependencies}), you can ignore this issue.
5340 In order to assemble compiler output into something that works,
5341 @code{@value{AS}} occasionally does strange things to @samp{.word} directives.
5342 Directives of the form @samp{.word sym1-sym2} are often emitted by
5343 compilers as part of jump tables. Therefore, when @code{@value{AS}} assembles a
5344 directive of the form @samp{.word sym1-sym2}, and the difference between
5345 @code{sym1} and @code{sym2} does not fit in 16 bits, @code{@value{AS}}
5346 creates a @dfn{secondary jump table}, immediately before the next label.
5347 This secondary jump table is preceded by a short-jump to the
5348 first byte after the secondary table. This short-jump prevents the flow
5349 of control from accidentally falling into the new table. Inside the
5350 table is a long-jump to @code{sym2}. The original @samp{.word}
5351 contains @code{sym1} minus the address of the long-jump to
5354 If there were several occurrences of @samp{.word sym1-sym2} before the
5355 secondary jump table, all of them are adjusted. If there was a
5356 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
5357 long-jump to @code{sym4} is included in the secondary jump table,
5358 and the @code{.word} directives are adjusted to contain @code{sym3}
5359 minus the address of the long-jump to @code{sym4}; and so on, for as many
5360 entries in the original jump table as necessary.
5363 @emph{This feature may be disabled by compiling @code{@value{AS}} with the
5364 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
5365 assembly language programmers.
5368 @c end DIFF-TBL-KLUGE
5371 @section Deprecated Directives
5373 @cindex deprecated directives
5374 @cindex obsolescent directives
5375 One day these directives won't work.
5376 They are included for compatibility with older assemblers.
5383 @node Machine Dependencies
5384 @chapter Machine Dependent Features
5386 @cindex machine dependencies
5387 The machine instruction sets are (almost by definition) different on
5388 each machine where @code{@value{AS}} runs. Floating point representations
5389 vary as well, and @code{@value{AS}} often supports a few additional
5390 directives or command-line options for compatibility with other
5391 assemblers on a particular platform. Finally, some versions of
5392 @code{@value{AS}} support special pseudo-instructions for branch
5395 This chapter discusses most of these differences, though it does not
5396 include details on any machine's instruction set. For details on that
5397 subject, see the hardware manufacturer's manual.
5401 * AMD29K-Dependent:: AMD 29K Dependent Features
5404 * ARC-Dependent:: ARC Dependent Features
5407 * ARM-Dependent:: ARM Dependent Features
5410 * D10V-Dependent:: D10V Dependent Features
5413 * D30V-Dependent:: D30V Dependent Features
5416 * H8/300-Dependent:: Hitachi H8/300 Dependent Features
5419 * H8/500-Dependent:: Hitachi H8/500 Dependent Features
5422 * HPPA-Dependent:: HPPA Dependent Features
5425 * ESA/390-Dependent:: IBM ESA/390 Dependent Features
5428 * i386-Dependent:: Intel 80386 and AMD x86-64 Dependent Features
5431 * i860-Dependent:: Intel 80860 Dependent Features
5434 * i960-Dependent:: Intel 80960 Dependent Features
5437 * M32R-Dependent:: M32R Dependent Features
5440 * M68K-Dependent:: M680x0 Dependent Features
5443 * M68HC11-Dependent:: M68HC11 and 68HC12 Dependent Features
5446 * MIPS-Dependent:: MIPS Dependent Features
5449 * SH-Dependent:: Hitachi SH Dependent Features
5452 * PDP-11-Dependent:: PDP-11 Dependent Features
5455 * PJ-Dependent:: picoJava Dependent Features
5458 * Sparc-Dependent:: SPARC Dependent Features
5461 * TIC54X-Dependent:: TI TMS320C54x Dependent Features
5464 * V850-Dependent:: V850 Dependent Features
5467 * Z8000-Dependent:: Z8000 Dependent Features
5470 * Vax-Dependent:: VAX Dependent Features
5477 @c The following major nodes are *sections* in the GENERIC version, *chapters*
5478 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
5479 @c peculiarity: to preserve cross-references, there must be a node called
5480 @c "Machine Dependencies". Hence the conditional nodenames in each
5481 @c major node below. Node defaulting in makeinfo requires adjacency of
5482 @c node and sectioning commands; hence the repetition of @chapter BLAH
5483 @c in both conditional blocks.
5490 @include c-a29k.texi
5499 @node Machine Dependencies
5500 @chapter Machine Dependent Features
5502 The machine instruction sets are different on each Hitachi chip family,
5503 and there are also some syntax differences among the families. This
5504 chapter describes the specific @code{@value{AS}} features for each
5508 * H8/300-Dependent:: Hitachi H8/300 Dependent Features
5509 * H8/500-Dependent:: Hitachi H8/500 Dependent Features
5510 * SH-Dependent:: Hitachi SH Dependent Features
5517 @include c-d10v.texi
5521 @include c-d30v.texi
5525 @include c-h8300.texi
5529 @include c-h8500.texi
5533 @include c-hppa.texi
5537 @include c-i370.texi
5541 @include c-i386.texi
5545 @include c-i860.texi
5549 @include c-i960.texi
5553 @include c-m32r.texi
5557 @include c-m68k.texi
5561 @include c-m68hc11.texi
5565 @include c-mips.texi
5569 @include c-ns32k.texi
5573 @include c-pdp11.texi
5585 @include c-sparc.texi
5589 @include c-tic54x.texi
5601 @include c-v850.texi
5605 @c reverse effect of @down at top of generic Machine-Dep chapter
5609 @node Reporting Bugs
5610 @chapter Reporting Bugs
5611 @cindex bugs in assembler
5612 @cindex reporting bugs in assembler
5614 Your bug reports play an essential role in making @code{@value{AS}} reliable.
5616 Reporting a bug may help you by bringing a solution to your problem, or it may
5617 not. But in any case the principal function of a bug report is to help the
5618 entire community by making the next version of @code{@value{AS}} work better.
5619 Bug reports are your contribution to the maintenance of @code{@value{AS}}.
5621 In order for a bug report to serve its purpose, you must include the
5622 information that enables us to fix the bug.
5625 * Bug Criteria:: Have you found a bug?
5626 * Bug Reporting:: How to report bugs
5630 @section Have you found a bug?
5631 @cindex bug criteria
5633 If you are not sure whether you have found a bug, here are some guidelines:
5636 @cindex fatal signal
5637 @cindex assembler crash
5638 @cindex crash of assembler
5640 If the assembler gets a fatal signal, for any input whatever, that is a
5641 @code{@value{AS}} bug. Reliable assemblers never crash.
5643 @cindex error on valid input
5645 If @code{@value{AS}} produces an error message for valid input, that is a bug.
5647 @cindex invalid input
5649 If @code{@value{AS}} does not produce an error message for invalid input, that
5650 is a bug. However, you should note that your idea of ``invalid input'' might
5651 be our idea of ``an extension'' or ``support for traditional practice''.
5654 If you are an experienced user of assemblers, your suggestions for improvement
5655 of @code{@value{AS}} are welcome in any case.
5659 @section How to report bugs
5661 @cindex assembler bugs, reporting
5663 A number of companies and individuals offer support for @sc{gnu} products. If
5664 you obtained @code{@value{AS}} from a support organization, we recommend you
5665 contact that organization first.
5667 You can find contact information for many support companies and
5668 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
5671 In any event, we also recommend that you send bug reports for @code{@value{AS}}
5672 to @samp{bug-binutils@@gnu.org}.
5674 The fundamental principle of reporting bugs usefully is this:
5675 @strong{report all the facts}. If you are not sure whether to state a
5676 fact or leave it out, state it!
5678 Often people omit facts because they think they know what causes the problem
5679 and assume that some details do not matter. Thus, you might assume that the
5680 name of a symbol you use in an example does not matter. Well, probably it does
5681 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
5682 happens to fetch from the location where that name is stored in memory;
5683 perhaps, if the name were different, the contents of that location would fool
5684 the assembler into doing the right thing despite the bug. Play it safe and
5685 give a specific, complete example. That is the easiest thing for you to do,
5686 and the most helpful.
5688 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
5689 it is new to us. Therefore, always write your bug reports on the assumption
5690 that the bug has not been reported previously.
5692 Sometimes people give a few sketchy facts and ask, ``Does this ring a
5693 bell?'' Those bug reports are useless, and we urge everyone to
5694 @emph{refuse to respond to them} except to chide the sender to report
5697 To enable us to fix the bug, you should include all these things:
5701 The version of @code{@value{AS}}. @code{@value{AS}} announces it if you start
5702 it with the @samp{--version} argument.
5704 Without this, we will not know whether there is any point in looking for
5705 the bug in the current version of @code{@value{AS}}.
5708 Any patches you may have applied to the @code{@value{AS}} source.
5711 The type of machine you are using, and the operating system name and
5715 What compiler (and its version) was used to compile @code{@value{AS}}---e.g.
5719 The command arguments you gave the assembler to assemble your example and
5720 observe the bug. To guarantee you will not omit something important, list them
5721 all. A copy of the Makefile (or the output from make) is sufficient.
5723 If we were to try to guess the arguments, we would probably guess wrong
5724 and then we might not encounter the bug.
5727 A complete input file that will reproduce the bug. If the bug is observed when
5728 the assembler is invoked via a compiler, send the assembler source, not the
5729 high level language source. Most compilers will produce the assembler source
5730 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
5731 the options @samp{-v --save-temps}; this will save the assembler source in a
5732 file with an extension of @file{.s}, and also show you exactly how
5733 @code{@value{AS}} is being run.
5736 A description of what behavior you observe that you believe is
5737 incorrect. For example, ``It gets a fatal signal.''
5739 Of course, if the bug is that @code{@value{AS}} gets a fatal signal, then we
5740 will certainly notice it. But if the bug is incorrect output, we might not
5741 notice unless it is glaringly wrong. You might as well not give us a chance to
5744 Even if the problem you experience is a fatal signal, you should still say so
5745 explicitly. Suppose something strange is going on, such as, your copy of
5746 @code{@value{AS}} is out of synch, or you have encountered a bug in the C
5747 library on your system. (This has happened!) Your copy might crash and ours
5748 would not. If you told us to expect a crash, then when ours fails to crash, we
5749 would know that the bug was not happening for us. If you had not told us to
5750 expect a crash, then we would not be able to draw any conclusion from our
5754 If you wish to suggest changes to the @code{@value{AS}} source, send us context
5755 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
5756 option. Always send diffs from the old file to the new file. If you even
5757 discuss something in the @code{@value{AS}} source, refer to it by context, not
5760 The line numbers in our development sources will not match those in your
5761 sources. Your line numbers would convey no useful information to us.
5764 Here are some things that are not necessary:
5768 A description of the envelope of the bug.
5770 Often people who encounter a bug spend a lot of time investigating
5771 which changes to the input file will make the bug go away and which
5772 changes will not affect it.
5774 This is often time consuming and not very useful, because the way we
5775 will find the bug is by running a single example under the debugger
5776 with breakpoints, not by pure deduction from a series of examples.
5777 We recommend that you save your time for something else.
5779 Of course, if you can find a simpler example to report @emph{instead}
5780 of the original one, that is a convenience for us. Errors in the
5781 output will be easier to spot, running under the debugger will take
5782 less time, and so on.
5784 However, simplification is not vital; if you do not want to do this,
5785 report the bug anyway and send us the entire test case you used.
5788 A patch for the bug.
5790 A patch for the bug does help us if it is a good one. But do not omit
5791 the necessary information, such as the test case, on the assumption that
5792 a patch is all we need. We might see problems with your patch and decide
5793 to fix the problem another way, or we might not understand it at all.
5795 Sometimes with a program as complicated as @code{@value{AS}} it is very hard to
5796 construct an example that will make the program follow a certain path through
5797 the code. If you do not send us the example, we will not be able to construct
5798 one, so we will not be able to verify that the bug is fixed.
5800 And if we cannot understand what bug you are trying to fix, or why your
5801 patch should be an improvement, we will not install it. A test case will
5802 help us to understand.
5805 A guess about what the bug is or what it depends on.
5807 Such guesses are usually wrong. Even we cannot guess right about such
5808 things without first using the debugger to find the facts.
5811 @node Acknowledgements
5812 @chapter Acknowledgements
5814 If you have contributed to @code{@value{AS}} and your name isn't listed here,
5815 it is not meant as a slight. We just don't know about it. Send mail to the
5816 maintainer, and we'll correct the situation. Currently
5818 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
5820 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
5823 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
5824 information and the 68k series machines, most of the preprocessing pass, and
5825 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
5827 K. Richard Pixley maintained GAS for a while, adding various enhancements and
5828 many bug fixes, including merging support for several processors, breaking GAS
5829 up to handle multiple object file format back ends (including heavy rewrite,
5830 testing, an integration of the coff and b.out back ends), adding configuration
5831 including heavy testing and verification of cross assemblers and file splits
5832 and renaming, converted GAS to strictly ANSI C including full prototypes, added
5833 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
5834 port (including considerable amounts of reverse engineering), a SPARC opcode
5835 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
5836 assertions and made them work, much other reorganization, cleanup, and lint.
5838 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
5839 in format-specific I/O modules.
5841 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
5842 has done much work with it since.
5844 The Intel 80386 machine description was written by Eliot Dresselhaus.
5846 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
5848 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
5849 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
5851 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
5852 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
5853 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
5854 support a.out format.
5856 Support for the Zilog Z8k and Hitachi H8/300 and H8/500 processors (tc-z8k,
5857 tc-h8300, tc-h8500), and IEEE 695 object file format (obj-ieee), was written by
5858 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
5859 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
5862 John Gilmore built the AMD 29000 support, added @code{.include} support, and
5863 simplified the configuration of which versions accept which directives. He
5864 updated the 68k machine description so that Motorola's opcodes always produced
5865 fixed-size instructions (e.g. @code{jsr}), while synthetic instructions
5866 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
5867 cross-compilation support, and one bug in relaxation that took a week and
5868 required the proverbial one-bit fix.
5870 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
5871 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
5872 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
5873 PowerPC assembler, and made a few other minor patches.
5875 Steve Chamberlain made @code{@value{AS}} able to generate listings.
5877 Hewlett-Packard contributed support for the HP9000/300.
5879 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
5880 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
5881 formats). This work was supported by both the Center for Software Science at
5882 the University of Utah and Cygnus Support.
5884 Support for ELF format files has been worked on by Mark Eichin of Cygnus
5885 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
5886 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
5887 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
5888 and some initial 64-bit support).
5890 Linas Vepstas added GAS support for the ESA/390 "IBM 370" architecture.
5892 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
5893 support for openVMS/Alpha.
5895 Timothy Wall, Michael Hayes, and Greg Smart contributed to the various tic*
5898 Several engineers at Cygnus Support have also provided many small bug fixes and
5899 configuration enhancements.
5901 Many others have contributed large or small bugfixes and enhancements. If
5902 you have contributed significant work and are not mentioned on this list, and
5903 want to be, let us know. Some of the history has been lost; we are not
5904 intentionally leaving anyone out.
5906 @node GNU Free Documentation License
5907 @chapter GNU Free Documentation License
5909 GNU Free Documentation License
5911 Version 1.1, March 2000
5913 Copyright (C) 2000 Free Software Foundation, Inc.
5914 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
5916 Everyone is permitted to copy and distribute verbatim copies
5917 of this license document, but changing it is not allowed.
5922 The purpose of this License is to make a manual, textbook, or other
5923 written document "free" in the sense of freedom: to assure everyone
5924 the effective freedom to copy and redistribute it, with or without
5925 modifying it, either commercially or noncommercially. Secondarily,
5926 this License preserves for the author and publisher a way to get
5927 credit for their work, while not being considered responsible for
5928 modifications made by others.
5930 This License is a kind of "copyleft", which means that derivative
5931 works of the document must themselves be free in the same sense. It
5932 complements the GNU General Public License, which is a copyleft
5933 license designed for free software.
5935 We have designed this License in order to use it for manuals for free
5936 software, because free software needs free documentation: a free
5937 program should come with manuals providing the same freedoms that the
5938 software does. But this License is not limited to software manuals;
5939 it can be used for any textual work, regardless of subject matter or
5940 whether it is published as a printed book. We recommend this License
5941 principally for works whose purpose is instruction or reference.
5944 1. APPLICABILITY AND DEFINITIONS
5946 This License applies to any manual or other work that contains a
5947 notice placed by the copyright holder saying it can be distributed
5948 under the terms of this License. The "Document", below, refers to any
5949 such manual or work. Any member of the public is a licensee, and is
5952 A "Modified Version" of the Document means any work containing the
5953 Document or a portion of it, either copied verbatim, or with
5954 modifications and/or translated into another language.
5956 A "Secondary Section" is a named appendix or a front-matter section of
5957 the Document that deals exclusively with the relationship of the
5958 publishers or authors of the Document to the Document's overall subject
5959 (or to related matters) and contains nothing that could fall directly
5960 within that overall subject. (For example, if the Document is in part a
5961 textbook of mathematics, a Secondary Section may not explain any
5962 mathematics.) The relationship could be a matter of historical
5963 connection with the subject or with related matters, or of legal,
5964 commercial, philosophical, ethical or political position regarding
5967 The "Invariant Sections" are certain Secondary Sections whose titles
5968 are designated, as being those of Invariant Sections, in the notice
5969 that says that the Document is released under this License.
5971 The "Cover Texts" are certain short passages of text that are listed,
5972 as Front-Cover Texts or Back-Cover Texts, in the notice that says that
5973 the Document is released under this License.
5975 A "Transparent" copy of the Document means a machine-readable copy,
5976 represented in a format whose specification is available to the
5977 general public, whose contents can be viewed and edited directly and
5978 straightforwardly with generic text editors or (for images composed of
5979 pixels) generic paint programs or (for drawings) some widely available
5980 drawing editor, and that is suitable for input to text formatters or
5981 for automatic translation to a variety of formats suitable for input
5982 to text formatters. A copy made in an otherwise Transparent file
5983 format whose markup has been designed to thwart or discourage
5984 subsequent modification by readers is not Transparent. A copy that is
5985 not "Transparent" is called "Opaque".
5987 Examples of suitable formats for Transparent copies include plain
5988 ASCII without markup, Texinfo input format, LaTeX input format, SGML
5989 or XML using a publicly available DTD, and standard-conforming simple
5990 HTML designed for human modification. Opaque formats include
5991 PostScript, PDF, proprietary formats that can be read and edited only
5992 by proprietary word processors, SGML or XML for which the DTD and/or
5993 processing tools are not generally available, and the
5994 machine-generated HTML produced by some word processors for output
5997 The "Title Page" means, for a printed book, the title page itself,
5998 plus such following pages as are needed to hold, legibly, the material
5999 this License requires to appear in the title page. For works in
6000 formats which do not have any title page as such, "Title Page" means
6001 the text near the most prominent appearance of the work's title,
6002 preceding the beginning of the body of the text.
6007 You may copy and distribute the Document in any medium, either
6008 commercially or noncommercially, provided that this License, the
6009 copyright notices, and the license notice saying this License applies
6010 to the Document are reproduced in all copies, and that you add no other
6011 conditions whatsoever to those of this License. You may not use
6012 technical measures to obstruct or control the reading or further
6013 copying of the copies you make or distribute. However, you may accept
6014 compensation in exchange for copies. If you distribute a large enough
6015 number of copies you must also follow the conditions in section 3.
6017 You may also lend copies, under the same conditions stated above, and
6018 you may publicly display copies.
6021 3. COPYING IN QUANTITY
6023 If you publish printed copies of the Document numbering more than 100,
6024 and the Document's license notice requires Cover Texts, you must enclose
6025 the copies in covers that carry, clearly and legibly, all these Cover
6026 Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on
6027 the back cover. Both covers must also clearly and legibly identify
6028 you as the publisher of these copies. The front cover must present
6029 the full title with all words of the title equally prominent and
6030 visible. You may add other material on the covers in addition.
6031 Copying with changes limited to the covers, as long as they preserve
6032 the title of the Document and satisfy these conditions, can be treated
6033 as verbatim copying in other respects.
6035 If the required texts for either cover are too voluminous to fit
6036 legibly, you should put the first ones listed (as many as fit
6037 reasonably) on the actual cover, and continue the rest onto adjacent
6040 If you publish or distribute Opaque copies of the Document numbering
6041 more than 100, you must either include a machine-readable Transparent
6042 copy along with each Opaque copy, or state in or with each Opaque copy
6043 a publicly-accessible computer-network location containing a complete
6044 Transparent copy of the Document, free of added material, which the
6045 general network-using public has access to download anonymously at no
6046 charge using public-standard network protocols. If you use the latter
6047 option, you must take reasonably prudent steps, when you begin
6048 distribution of Opaque copies in quantity, to ensure that this
6049 Transparent copy will remain thus accessible at the stated location
6050 until at least one year after the last time you distribute an Opaque
6051 copy (directly or through your agents or retailers) of that edition to
6054 It is requested, but not required, that you contact the authors of the
6055 Document well before redistributing any large number of copies, to give
6056 them a chance to provide you with an updated version of the Document.
6061 You may copy and distribute a Modified Version of the Document under
6062 the conditions of sections 2 and 3 above, provided that you release
6063 the Modified Version under precisely this License, with the Modified
6064 Version filling the role of the Document, thus licensing distribution
6065 and modification of the Modified Version to whoever possesses a copy
6066 of it. In addition, you must do these things in the Modified Version:
6068 A. Use in the Title Page (and on the covers, if any) a title distinct
6069 from that of the Document, and from those of previous versions
6070 (which should, if there were any, be listed in the History section
6071 of the Document). You may use the same title as a previous version
6072 if the original publisher of that version gives permission.
6073 B. List on the Title Page, as authors, one or more persons or entities
6074 responsible for authorship of the modifications in the Modified
6075 Version, together with at least five of the principal authors of the
6076 Document (all of its principal authors, if it has less than five).
6077 C. State on the Title page the name of the publisher of the
6078 Modified Version, as the publisher.
6079 D. Preserve all the copyright notices of the Document.
6080 E. Add an appropriate copyright notice for your modifications
6081 adjacent to the other copyright notices.
6082 F. Include, immediately after the copyright notices, a license notice
6083 giving the public permission to use the Modified Version under the
6084 terms of this License, in the form shown in the Addendum below.
6085 G. Preserve in that license notice the full lists of Invariant Sections
6086 and required Cover Texts given in the Document's license notice.
6087 H. Include an unaltered copy of this License.
6088 I. Preserve the section entitled "History", and its title, and add to
6089 it an item stating at least the title, year, new authors, and
6090 publisher of the Modified Version as given on the Title Page. If
6091 there is no section entitled "History" in the Document, create one
6092 stating the title, year, authors, and publisher of the Document as
6093 given on its Title Page, then add an item describing the Modified
6094 Version as stated in the previous sentence.
6095 J. Preserve the network location, if any, given in the Document for
6096 public access to a Transparent copy of the Document, and likewise
6097 the network locations given in the Document for previous versions
6098 it was based on. These may be placed in the "History" section.
6099 You may omit a network location for a work that was published at
6100 least four years before the Document itself, or if the original
6101 publisher of the version it refers to gives permission.
6102 K. In any section entitled "Acknowledgements" or "Dedications",
6103 preserve the section's title, and preserve in the section all the
6104 substance and tone of each of the contributor acknowledgements
6105 and/or dedications given therein.
6106 L. Preserve all the Invariant Sections of the Document,
6107 unaltered in their text and in their titles. Section numbers
6108 or the equivalent are not considered part of the section titles.
6109 M. Delete any section entitled "Endorsements". Such a section
6110 may not be included in the Modified Version.
6111 N. Do not retitle any existing section as "Endorsements"
6112 or to conflict in title with any Invariant Section.
6114 If the Modified Version includes new front-matter sections or
6115 appendices that qualify as Secondary Sections and contain no material
6116 copied from the Document, you may at your option designate some or all
6117 of these sections as invariant. To do this, add their titles to the
6118 list of Invariant Sections in the Modified Version's license notice.
6119 These titles must be distinct from any other section titles.
6121 You may add a section entitled "Endorsements", provided it contains
6122 nothing but endorsements of your Modified Version by various
6123 parties--for example, statements of peer review or that the text has
6124 been approved by an organization as the authoritative definition of a
6127 You may add a passage of up to five words as a Front-Cover Text, and a
6128 passage of up to 25 words as a Back-Cover Text, to the end of the list
6129 of Cover Texts in the Modified Version. Only one passage of
6130 Front-Cover Text and one of Back-Cover Text may be added by (or
6131 through arrangements made by) any one entity. If the Document already
6132 includes a cover text for the same cover, previously added by you or
6133 by arrangement made by the same entity you are acting on behalf of,
6134 you may not add another; but you may replace the old one, on explicit
6135 permission from the previous publisher that added the old one.
6137 The author(s) and publisher(s) of the Document do not by this License
6138 give permission to use their names for publicity for or to assert or
6139 imply endorsement of any Modified Version.
6142 5. COMBINING DOCUMENTS
6144 You may combine the Document with other documents released under this
6145 License, under the terms defined in section 4 above for modified
6146 versions, provided that you include in the combination all of the
6147 Invariant Sections of all of the original documents, unmodified, and
6148 list them all as Invariant Sections of your combined work in its
6151 The combined work need only contain one copy of this License, and
6152 multiple identical Invariant Sections may be replaced with a single
6153 copy. If there are multiple Invariant Sections with the same name but
6154 different contents, make the title of each such section unique by
6155 adding at the end of it, in parentheses, the name of the original
6156 author or publisher of that section if known, or else a unique number.
6157 Make the same adjustment to the section titles in the list of
6158 Invariant Sections in the license notice of the combined work.
6160 In the combination, you must combine any sections entitled "History"
6161 in the various original documents, forming one section entitled
6162 "History"; likewise combine any sections entitled "Acknowledgements",
6163 and any sections entitled "Dedications". You must delete all sections
6164 entitled "Endorsements."
6167 6. COLLECTIONS OF DOCUMENTS
6169 You may make a collection consisting of the Document and other documents
6170 released under this License, and replace the individual copies of this
6171 License in the various documents with a single copy that is included in
6172 the collection, provided that you follow the rules of this License for
6173 verbatim copying of each of the documents in all other respects.
6175 You may extract a single document from such a collection, and distribute
6176 it individually under this License, provided you insert a copy of this
6177 License into the extracted document, and follow this License in all
6178 other respects regarding verbatim copying of that document.
6181 7. AGGREGATION WITH INDEPENDENT WORKS
6183 A compilation of the Document or its derivatives with other separate
6184 and independent documents or works, in or on a volume of a storage or
6185 distribution medium, does not as a whole count as a Modified Version
6186 of the Document, provided no compilation copyright is claimed for the
6187 compilation. Such a compilation is called an "aggregate", and this
6188 License does not apply to the other self-contained works thus compiled
6189 with the Document, on account of their being thus compiled, if they
6190 are not themselves derivative works of the Document.
6192 If the Cover Text requirement of section 3 is applicable to these
6193 copies of the Document, then if the Document is less than one quarter
6194 of the entire aggregate, the Document's Cover Texts may be placed on
6195 covers that surround only the Document within the aggregate.
6196 Otherwise they must appear on covers around the whole aggregate.
6201 Translation is considered a kind of modification, so you may
6202 distribute translations of the Document under the terms of section 4.
6203 Replacing Invariant Sections with translations requires special
6204 permission from their copyright holders, but you may include
6205 translations of some or all Invariant Sections in addition to the
6206 original versions of these Invariant Sections. You may include a
6207 translation of this License provided that you also include the
6208 original English version of this License. In case of a disagreement
6209 between the translation and the original English version of this
6210 License, the original English version will prevail.
6215 You may not copy, modify, sublicense, or distribute the Document except
6216 as expressly provided for under this License. Any other attempt to
6217 copy, modify, sublicense or distribute the Document is void, and will
6218 automatically terminate your rights under this License. However,
6219 parties who have received copies, or rights, from you under this
6220 License will not have their licenses terminated so long as such
6221 parties remain in full compliance.
6224 10. FUTURE REVISIONS OF THIS LICENSE
6226 The Free Software Foundation may publish new, revised versions
6227 of the GNU Free Documentation License from time to time. Such new
6228 versions will be similar in spirit to the present version, but may
6229 differ in detail to address new problems or concerns. See
6230 http://www.gnu.org/copyleft/.
6232 Each version of the License is given a distinguishing version number.
6233 If the Document specifies that a particular numbered version of this
6234 License "or any later version" applies to it, you have the option of
6235 following the terms and conditions either of that specified version or
6236 of any later version that has been published (not as a draft) by the
6237 Free Software Foundation. If the Document does not specify a version
6238 number of this License, you may choose any version ever published (not
6239 as a draft) by the Free Software Foundation.
6242 ADDENDUM: How to use this License for your documents
6244 To use this License in a document you have written, include a copy of
6245 the License in the document and put the following copyright and
6246 license notices just after the title page:
6249 Copyright (c) YEAR YOUR NAME.
6250 Permission is granted to copy, distribute and/or modify this document
6251 under the terms of the GNU Free Documentation License, Version 1.1
6252 or any later version published by the Free Software Foundation;
6253 with the Invariant Sections being LIST THEIR TITLES, with the
6254 Front-Cover Texts being LIST, and with the Back-Cover Texts being LIST.
6255 A copy of the license is included in the section entitled "GNU
6256 Free Documentation License".
6259 If you have no Invariant Sections, write "with no Invariant Sections"
6260 instead of saying which ones are invariant. If you have no
6261 Front-Cover Texts, write "no Front-Cover Texts" instead of
6262 "Front-Cover Texts being LIST"; likewise for Back-Cover Texts.
6264 If your document contains nontrivial examples of program code, we
6265 recommend releasing these examples in parallel under your choice of
6266 free software license, such as the GNU General Public License,
6267 to permit their use in free software.