1 \input texinfo @c -*-Texinfo-*-
2 @c Copyright (c) 1991, 92, 93, 94, 95, 96, 97, 1998
3 @c Free Software Foundation, Inc.
4 @c UPDATE!! On future updates--
5 @c (1) check for new machine-dep cmdline options in
6 @c md_parse_option definitions in config/tc-*.c
7 @c (2) for platform-specific directives, examine md_pseudo_op
9 @c (3) for object-format specific directives, examine obj_pseudo_op
11 @c (4) portable directives in potable[] in read.c
15 @c defaults, config file may override:
18 @include asconfig.texi
21 @c common OR combinations of conditions
41 @set abnormal-separator
45 @settitle Using @value{AS}
48 @settitle Using @value{AS} (@value{TARGET})
50 @setchapternewpage odd
55 @c WARE! Some of the machine-dependent sections contain tables of machine
56 @c instructions. Except in multi-column format, these tables look silly.
57 @c Unfortunately, Texinfo doesn't have a general-purpose multi-col format, so
58 @c the multi-col format is faked within @example sections.
60 @c Again unfortunately, the natural size that fits on a page, for these tables,
61 @c is different depending on whether or not smallbook is turned on.
62 @c This matters, because of order: text flow switches columns at each page
65 @c The format faked in this source works reasonably well for smallbook,
66 @c not well for the default large-page format. This manual expects that if you
67 @c turn on @smallbook, you will also uncomment the "@set SMALL" to enable the
68 @c tables in question. You can turn on one without the other at your
69 @c discretion, of course.
72 @c the insn tables look just as silly in info files regardless of smallbook,
73 @c might as well show 'em anyways.
79 * As: (as). The GNU assembler.
88 This file documents the GNU Assembler "@value{AS}".
90 Copyright (C) 1991, 92, 93, 94, 95, 96, 97, 1998 Free Software Foundation, Inc.
92 Permission is granted to make and distribute verbatim copies of
93 this manual provided the copyright notice and this permission notice
94 are preserved on all copies.
97 Permission is granted to process this file through Tex and print the
98 results, provided the printed document carries copying permission
99 notice identical to this one except for the removal of this paragraph
100 (this paragraph not being relevant to the printed manual).
103 Permission is granted to copy and distribute modified versions of this manual
104 under the conditions for verbatim copying, provided that the entire resulting
105 derived work is distributed under the terms of a permission notice identical to
108 Permission is granted to copy and distribute translations of this manual
109 into another language, under the above conditions for modified versions.
113 @title Using @value{AS}
114 @subtitle The @sc{gnu} Assembler
116 @subtitle for the @value{TARGET} family
119 @subtitle Version @value{VERSION}
122 The Free Software Foundation Inc. thanks The Nice Computer
123 Company of Australia for loaning Dean Elsner to write the
124 first (Vax) version of @code{as} for Project @sc{gnu}.
125 The proprietors, management and staff of TNCCA thank FSF for
126 distracting the boss while they got some work
129 @author Dean Elsner, Jay Fenlason & friends
133 \hfill {\it Using {\tt @value{AS}}}\par
134 \hfill Edited by Cygnus Support\par
136 %"boxit" macro for figures:
137 %Modified from Knuth's ``boxit'' macro from TeXbook (answer to exercise 21.3)
138 \gdef\boxit#1#2{\vbox{\hrule\hbox{\vrule\kern3pt
139 \vbox{\parindent=0pt\parskip=0pt\hsize=#1\kern3pt\strut\hfil
140 #2\hfil\strut\kern3pt}\kern3pt\vrule}\hrule}}%box with visible outline
141 \gdef\ibox#1#2{\hbox to #1{#2\hfil}\kern8pt}% invisible box
144 @vskip 0pt plus 1filll
145 Copyright @copyright{} 1991, 92, 93, 94, 95, 96, 97, 1998 Free Software Foundation, Inc.
147 Permission is granted to make and distribute verbatim copies of
148 this manual provided the copyright notice and this permission notice
149 are preserved on all copies.
151 Permission is granted to copy and distribute modified versions of this manual
152 under the conditions for verbatim copying, provided that the entire resulting
153 derived work is distributed under the terms of a permission notice identical to
156 Permission is granted to copy and distribute translations of this manual
157 into another language, under the above conditions for modified versions.
162 @top Using @value{AS}
164 This file is a user guide to the @sc{gnu} assembler @code{@value{AS}} version
167 This version of the file describes @code{@value{AS}} configured to generate
168 code for @value{TARGET} architectures.
171 * Overview:: Overview
172 * Invoking:: Command-Line Options
174 * Sections:: Sections and Relocation
176 * Expressions:: Expressions
177 * Pseudo Ops:: Assembler Directives
178 * Machine Dependencies:: Machine Dependent Features
179 * Reporting Bugs:: Reporting Bugs
180 * Acknowledgements:: Who Did What
188 This manual is a user guide to the @sc{gnu} assembler @code{@value{AS}}.
190 This version of the manual describes @code{@value{AS}} configured to generate
191 code for @value{TARGET} architectures.
195 @cindex invocation summary
196 @cindex option summary
197 @cindex summary of options
198 Here is a brief summary of how to invoke @code{@value{AS}}. For details,
199 @pxref{Invoking,,Comand-Line Options}.
201 @c We don't use deffn and friends for the following because they seem
202 @c to be limited to one line for the header.
204 @value{AS} [ -a[cdhlns][=file] ] [ -D ] [ --defsym @var{sym}=@var{val} ]
205 [ -f ] [ --gstabs ] [ --help ] [ -I @var{dir} ] [ -J ] [ -K ] [ -L ]
206 [ --keep-locals ] [ -o @var{objfile} ] [ -R ] [ --statistics ] [ -v ]
207 [ -version ] [ --version ] [ -W ] [ -w ] [ -x ] [ -Z ]
209 @c am29k has no machine-dependent assembler options
212 [ -mbig-endian | -mlittle-endian ]
215 [ -m[arm]1 | -m[arm]2 | -m[arm]250 | -m[arm]3 | -m[arm]6 | -m[arm]7[t][[d]m[i]] ]
216 [ -m[arm]v2 | -m[arm]v2a | -m[arm]v3 | -m[arm]v3m | -m[arm]v4 | -m[arm]v4t ]
218 [ -mfpa10 | -mfpa11 | -mfpe-old | -mno-fpu ]
220 [ -mapcs-32 | -mapcs-26 ]
229 @c Hitachi family chips have no machine-dependent assembler options
232 @c HPPA has no machine-dependent assembler options (yet).
235 @c The order here is important. See c-sparc.texi.
236 [ -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
237 -Av8plus | -Av8plusa | -Av9 | -Av9a ]
238 [ -xarch=v8plus | -xarch=v8plusa ] [ -bump ] [ -32 | -64 ]
241 @c Z8000 has no machine-dependent assembler options
244 @c see md_parse_option in tc-i960.c
245 [ -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC ]
249 [ -l ] [ -m68000 | -m68010 | -m68020 | ... ]
252 [ -jsri2bsr ] [ -sifilter ] [ -relax ]
255 [ -nocpp ] [ -EL ] [ -EB ] [ -G @var{num} ] [ -mcpu=@var{CPU} ]
256 [ -mips1 ] [ -mips2 ] [ -mips3 ] [ -m4650 ] [ -no-m4650 ]
257 [ --trap ] [ --break ]
258 [ --emulation=@var{name} ]
260 [ -- | @var{files} @dots{} ]
265 Turn on listings, in any of a variety of ways:
269 omit false conditionals
272 omit debugging directives
275 include high-level source
281 include macro expansions
284 omit forms processing
290 set the name of the listing file
293 You may combine these options; for example, use @samp{-aln} for assembly
294 listing without forms processing. The @samp{=file} option, if used, must be
295 the last one. By itself, @samp{-a} defaults to @samp{-ahls}.
298 Ignored. This option is accepted for script compatibility with calls to
301 @item --defsym @var{sym}=@var{value}
302 Define the symbol @var{sym} to be @var{value} before assembling the input file.
303 @var{value} must be an integer constant. As in C, a leading @samp{0x}
304 indicates a hexadecimal value, and a leading @samp{0} indicates an octal value.
307 ``fast''---skip whitespace and comment preprocessing (assume source is
311 Generate stabs debugging information for each assembler line. This
312 may help debugging assembler code, if the debugger can handle it.
315 Print a summary of the command line options and exit.
318 Add directory @var{dir} to the search list for @code{.include} directives.
321 Don't warn about signed overflow.
324 @ifclear DIFF-TBL-KLUGE
325 This option is accepted but has no effect on the @value{TARGET} family.
327 @ifset DIFF-TBL-KLUGE
328 Issue warnings when difference tables altered for long displacements.
333 Keep (in the symbol table) local symbols. On traditional a.out systems
334 these start with @samp{L}, but different systems have different local
337 @item -o @var{objfile}
338 Name the object-file output from @code{@value{AS}} @var{objfile}.
341 Fold the data section into the text section.
344 Print the maximum space (in bytes) and total time (in seconds) used by
347 @item --strip-local-absolute
348 Remove local absolute symbols from the outgoing symbol table.
352 Print the @code{as} version.
355 Print the @code{as} version and exit.
358 Suppress warning messages.
367 Generate an object file even after errors.
369 @item -- | @var{files} @dots{}
370 Standard input, or source files to assemble.
375 The following options are available when @value{AS} is configured for
380 @cindex ARC endianness
381 @cindex endianness, ARC
382 @cindex big endian output, ARC
384 Generate ``big endian'' format output.
386 @cindex little endian output, ARC
387 @item -mlittle-endian
388 Generate ``little endian'' format output.
394 The following options are available when @value{AS} is configured for the ARM
398 @item -m[arm]1 | -m[arm]2 | -m[arm]250 | -m[arm]3 | -m[arm]6 | -m[arm]7[t][[d]m] | -m[arm]v2 | -m[arm]v2a | -m[arm]v3 | -m[arm]v3m | -m[arm]v4 | -m[arm]v4t
399 Specify which variant of the ARM architecture is the target.
400 @item -mthumb | -mall
401 Enable or disable Thumb only instruction decoding.
402 @item -mfpa10 | -mfpa11 | -mfpe-old | -mno-fpu
403 Select which Floating Point architcture is the target.
404 @item -mapcs-32 | -mapcs-26
405 Select which procedure calling convention is in use.
407 Select either big-endian (-EB) or little-endian (-EL) output.
412 The following options are available when @value{AS} is configured for
415 @cindex D10V optimization
416 @cindex optimization, D10V
418 Optimize output by parallelizing instructions.
423 The following options are available when @value{AS} is configured for a D30V
426 @cindex D30V optimization
427 @cindex optimization, D30V
429 Optimize output by parallelizing instructions.
433 Warn when nops are generated.
435 @cindex D30V nops after 32-bit multiply
437 Warn when a nop after a 32-bit multiply instruction is generated.
442 The following options are available when @value{AS} is configured for the
443 Intel 80960 processor.
446 @item -ACA | -ACA_A | -ACB | -ACC | -AKA | -AKB | -AKC | -AMC
447 Specify which variant of the 960 architecture is the target.
450 Add code to collect statistics about branches taken.
453 Do not alter compare-and-branch instructions for long displacements;
461 The following options are available when @value{AS} is configured for the
462 Motorola 68000 series.
467 Shorten references to undefined symbols, to one word instead of two.
469 @item -m68000 | -m68008 | -m68010 | -m68020 | -m68030 | -m68040 | -m68060
470 @itemx | -m68302 | -m68331 | -m68332 | -m68333 | -m68340 | -mcpu32 | -m5200
471 Specify what processor in the 68000 family is the target. The default
472 is normally the 68020, but this can be changed at configuration time.
474 @item -m68881 | -m68882 | -mno-68881 | -mno-68882
475 The target machine does (or does not) have a floating-point coprocessor.
476 The default is to assume a coprocessor for 68020, 68030, and cpu32. Although
477 the basic 68000 is not compatible with the 68881, a combination of the
478 two can be specified, since it's possible to do emulation of the
479 coprocessor instructions with the main processor.
481 @item -m68851 | -mno-68851
482 The target machine does (or does not) have a memory-management
483 unit coprocessor. The default is to assume an MMU for 68020 and up.
489 The following options are available when @code{@value{AS}} is configured
490 for the SPARC architecture:
493 @item -Av6 | -Av7 | -Av8 | -Asparclet | -Asparclite
494 @itemx -Av8plus | -Av8plusa | -Av9 | -Av9a
495 Explicitly select a variant of the SPARC architecture.
497 @samp{-Av8plus} and @samp{-Av8plusa} select a 32 bit environment.
498 @samp{-Av9} and @samp{-Av9a} select a 64 bit environment.
500 @samp{-Av8plusa} and @samp{-Av9a} enable the SPARC V9 instruction set with
501 UltraSPARC extensions.
503 @item -xarch=v8plus | -xarch=v8plusa
504 For compatibility with the Solaris v9 assembler. These options are
505 equivalent to -Av8plus and -Av8plusa, respectively.
508 Warn when the assembler switches to another architecture.
513 The following options are available when @value{AS} is configured for
518 This option sets the largest size of an object that can be referenced
519 implicitly with the @code{gp} register. It is only accepted for targets that
520 use ECOFF format, such as a DECstation running Ultrix. The default value is 8.
522 @cindex MIPS endianness
523 @cindex endianness, MIPS
524 @cindex big endian output, MIPS
526 Generate ``big endian'' format output.
528 @cindex little endian output, MIPS
530 Generate ``little endian'' format output.
536 Generate code for a particular MIPS Instruction Set Architecture level.
537 @samp{-mips1} corresponds to the @sc{r2000} and @sc{r3000} processors,
538 @samp{-mips2} to the @sc{r6000} processor, and @samp{-mips3} to the @sc{r4000}
543 Generate code for the MIPS @sc{r4650} chip. This tells the assembler to accept
544 the @samp{mad} and @samp{madu} instruction, and to not schedule @samp{nop}
545 instructions around accesses to the @samp{HI} and @samp{LO} registers.
546 @samp{-no-m4650} turns off this option.
548 @item -mcpu=@var{CPU}
549 Generate code for a particular MIPS cpu. This has little effect on the
550 assembler, but it is passed by @code{@value{GCC}}.
553 @item --emulation=@var{name}
554 This option causes @code{@value{AS}} to emulate @code{@value{AS}} configured
555 for some other target, in all respects, including output format (choosing
556 between ELF and ECOFF only), handling of pseudo-opcodes which may generate
557 debugging information or store symbol table information, and default
558 endianness. The available configuration names are: @samp{mipsecoff},
559 @samp{mipself}, @samp{mipslecoff}, @samp{mipsbecoff}, @samp{mipslelf},
560 @samp{mipsbelf}. The first two do not alter the default endianness from that
561 of the primary target for which the assembler was configured; the others change
562 the default to little- or big-endian as indicated by the @samp{b} or @samp{l}
563 in the name. Using @samp{-EB} or @samp{-EL} will override the endianness
564 selection in any case.
566 This option is currently supported only when the primary target
567 @code{@value{AS}} is configured for is a MIPS ELF or ECOFF target.
568 Furthermore, the primary target or others specified with
569 @samp{--enable-targets=@dots{}} at configuration time must include support for
570 the other format, if both are to be available. For example, the Irix 5
571 configuration includes support for both.
573 Eventually, this option will support more configurations, with more
574 fine-grained control over the assembler's behavior, and will be supported for
578 @code{@value{AS}} ignores this option. It is accepted for compatibility with
586 Control how to deal with multiplication overflow and division by zero.
587 @samp{--trap} or @samp{--no-break} (which are synonyms) take a trap exception
588 (and only work for Instruction Set Architecture level 2 and higher);
589 @samp{--break} or @samp{--no-trap} (also synonyms, and the default) take a
595 The following options are available when @value{AS} is configured for
601 Enable or disable the JSRI to BSR transformation. By default this is enabled.
602 The command line option @samp{-nojsri2bsr} can be used to disable it.
606 Enable or disable the silicon filter behaviour. By default this is disabled.
607 The default can be overidden by the @samp{-sifilter} command line option.
610 Alter jump instructions for long displacements.
617 * Manual:: Structure of this Manual
618 * GNU Assembler:: The GNU Assembler
619 * Object Formats:: Object File Formats
620 * Command Line:: Command Line
621 * Input Files:: Input Files
622 * Object:: Output (Object) File
623 * Errors:: Error and Warning Messages
627 @section Structure of this Manual
629 @cindex manual, structure and purpose
630 This manual is intended to describe what you need to know to use
631 @sc{gnu} @code{@value{AS}}. We cover the syntax expected in source files, including
632 notation for symbols, constants, and expressions; the directives that
633 @code{@value{AS}} understands; and of course how to invoke @code{@value{AS}}.
636 We also cover special features in the @value{TARGET}
637 configuration of @code{@value{AS}}, including assembler directives.
640 This manual also describes some of the machine-dependent features of
641 various flavors of the assembler.
644 @cindex machine instructions (not covered)
645 On the other hand, this manual is @emph{not} intended as an introduction
646 to programming in assembly language---let alone programming in general!
647 In a similar vein, we make no attempt to introduce the machine
648 architecture; we do @emph{not} describe the instruction set, standard
649 mnemonics, registers or addressing modes that are standard to a
650 particular architecture.
652 You may want to consult the manufacturer's
653 machine architecture manual for this information.
657 For information on the H8/300 machine instruction set, see @cite{H8/300
658 Series Programming Manual} (Hitachi ADE--602--025). For the H8/300H,
659 see @cite{H8/300H Series Programming Manual} (Hitachi).
662 For information on the H8/500 machine instruction set, see @cite{H8/500
663 Series Programming Manual} (Hitachi M21T001).
666 For information on the Hitachi SH machine instruction set, see
667 @cite{SH-Microcomputer User's Manual} (Hitachi Micro Systems, Inc.).
670 For information on the Z8000 machine instruction set, see @cite{Z8000 CPU Technical Manual}
674 @c I think this is premature---doc@cygnus.com, 17jan1991
676 Throughout this manual, we assume that you are running @dfn{GNU},
677 the portable operating system from the @dfn{Free Software
678 Foundation, Inc.}. This restricts our attention to certain kinds of
679 computer (in particular, the kinds of computers that @sc{gnu} can run on);
680 once this assumption is granted examples and definitions need less
683 @code{@value{AS}} is part of a team of programs that turn a high-level
684 human-readable series of instructions into a low-level
685 computer-readable series of instructions. Different versions of
686 @code{@value{AS}} are used for different kinds of computer.
689 @c There used to be a section "Terminology" here, which defined
690 @c "contents", "byte", "word", and "long". Defining "word" to any
691 @c particular size is confusing when the .word directive may generate 16
692 @c bits on one machine and 32 bits on another; in general, for the user
693 @c version of this manual, none of these terms seem essential to define.
694 @c They were used very little even in the former draft of the manual;
695 @c this draft makes an effort to avoid them (except in names of
699 @section The GNU Assembler
701 @sc{gnu} @code{as} is really a family of assemblers.
703 This manual describes @code{@value{AS}}, a member of that family which is
704 configured for the @value{TARGET} architectures.
706 If you use (or have used) the @sc{gnu} assembler on one architecture, you
707 should find a fairly similar environment when you use it on another
708 architecture. Each version has much in common with the others,
709 including object file formats, most assembler directives (often called
710 @dfn{pseudo-ops}) and assembler syntax.@refill
712 @cindex purpose of @sc{gnu} assembler
713 @code{@value{AS}} is primarily intended to assemble the output of the
714 @sc{gnu} C compiler @code{@value{GCC}} for use by the linker
715 @code{@value{LD}}. Nevertheless, we've tried to make @code{@value{AS}}
716 assemble correctly everything that other assemblers for the same
717 machine would assemble.
719 Any exceptions are documented explicitly (@pxref{Machine Dependencies}).
722 @c This remark should appear in generic version of manual; assumption
723 @c here is that generic version sets M680x0.
724 This doesn't mean @code{@value{AS}} always uses the same syntax as another
725 assembler for the same architecture; for example, we know of several
726 incompatible versions of 680x0 assembly language syntax.
729 Unlike older assemblers, @code{@value{AS}} is designed to assemble a source
730 program in one pass of the source file. This has a subtle impact on the
731 @kbd{.org} directive (@pxref{Org,,@code{.org}}).
734 @section Object File Formats
736 @cindex object file format
737 The @sc{gnu} assembler can be configured to produce several alternative
738 object file formats. For the most part, this does not affect how you
739 write assembly language programs; but directives for debugging symbols
740 are typically different in different file formats. @xref{Symbol
741 Attributes,,Symbol Attributes}.
744 On the @value{TARGET}, @code{@value{AS}} is configured to produce
745 @value{OBJ-NAME} format object files.
747 @c The following should exhaust all configs that set MULTI-OBJ, ideally
749 On the @value{TARGET}, @code{@value{AS}} can be configured to produce either
750 @code{a.out} or COFF format object files.
753 On the @value{TARGET}, @code{@value{AS}} can be configured to produce either
754 @code{b.out} or COFF format object files.
757 On the @value{TARGET}, @code{@value{AS}} can be configured to produce either
758 SOM or ELF format object files.
763 @section Command Line
765 @cindex command line conventions
766 After the program name @code{@value{AS}}, the command line may contain
767 options and file names. Options may appear in any order, and may be
768 before, after, or between file names. The order of file names is
771 @cindex standard input, as input file
773 @file{--} (two hyphens) by itself names the standard input file
774 explicitly, as one of the files for @code{@value{AS}} to assemble.
776 @cindex options, command line
777 Except for @samp{--} any command line argument that begins with a
778 hyphen (@samp{-}) is an option. Each option changes the behavior of
779 @code{@value{AS}}. No option changes the way another option works. An
780 option is a @samp{-} followed by one or more letters; the case of
781 the letter is important. All options are optional.
783 Some options expect exactly one file name to follow them. The file
784 name may either immediately follow the option's letter (compatible
785 with older assemblers) or it may be the next command argument (@sc{gnu}
786 standard). These two command lines are equivalent:
789 @value{AS} -o my-object-file.o mumble.s
790 @value{AS} -omy-object-file.o mumble.s
797 @cindex source program
799 We use the phrase @dfn{source program}, abbreviated @dfn{source}, to
800 describe the program input to one run of @code{@value{AS}}. The program may
801 be in one or more files; how the source is partitioned into files
802 doesn't change the meaning of the source.
804 @c I added "con" prefix to "catenation" just to prove I can overcome my
805 @c APL training... doc@cygnus.com
806 The source program is a concatenation of the text in all the files, in the
809 Each time you run @code{@value{AS}} it assembles exactly one source
810 program. The source program is made up of one or more files.
811 (The standard input is also a file.)
813 You give @code{@value{AS}} a command line that has zero or more input file
814 names. The input files are read (from left file name to right). A
815 command line argument (in any position) that has no special meaning
816 is taken to be an input file name.
818 If you give @code{@value{AS}} no file names it attempts to read one input file
819 from the @code{@value{AS}} standard input, which is normally your terminal. You
820 may have to type @key{ctl-D} to tell @code{@value{AS}} there is no more program
823 Use @samp{--} if you need to explicitly name the standard input file
824 in your command line.
826 If the source is empty, @code{@value{AS}} produces a small, empty object
829 @subheading Filenames and Line-numbers
831 @cindex input file linenumbers
832 @cindex line numbers, in input files
833 There are two ways of locating a line in the input file (or files) and
834 either may be used in reporting error messages. One way refers to a line
835 number in a physical file; the other refers to a line number in a
836 ``logical'' file. @xref{Errors, ,Error and Warning Messages}.
838 @dfn{Physical files} are those files named in the command line given
839 to @code{@value{AS}}.
841 @dfn{Logical files} are simply names declared explicitly by assembler
842 directives; they bear no relation to physical files. Logical file names help
843 error messages reflect the original source file, when @code{@value{AS}} source
844 is itself synthesized from other files. @code{@value{AS}} understands the
845 @samp{#} directives emitted by the @code{@value{GCC}} preprocessor. See also
846 @ref{File,,@code{.file}}.
849 @section Output (Object) File
855 Every time you run @code{@value{AS}} it produces an output file, which is
856 your assembly language program translated into numbers. This file
857 is the object file. Its default name is
865 @code{b.out} when @code{@value{AS}} is configured for the Intel 80960.
867 You can give it another name by using the @code{-o} option. Conventionally,
868 object file names end with @file{.o}. The default name is used for historical
869 reasons: older assemblers were capable of assembling self-contained programs
870 directly into a runnable program. (For some formats, this isn't currently
871 possible, but it can be done for the @code{a.out} format.)
875 The object file is meant for input to the linker @code{@value{LD}}. It contains
876 assembled program code, information to help @code{@value{LD}} integrate
877 the assembled program into a runnable file, and (optionally) symbolic
878 information for the debugger.
880 @c link above to some info file(s) like the description of a.out.
881 @c don't forget to describe @sc{gnu} info as well as Unix lossage.
884 @section Error and Warning Messages
886 @cindex error messsages
887 @cindex warning messages
888 @cindex messages from assembler
889 @code{@value{AS}} may write warnings and error messages to the standard error
890 file (usually your terminal). This should not happen when a compiler
891 runs @code{@value{AS}} automatically. Warnings report an assumption made so
892 that @code{@value{AS}} could keep assembling a flawed program; errors report a
893 grave problem that stops the assembly.
895 @cindex format of warning messages
896 Warning messages have the format
899 file_name:@b{NNN}:Warning Message Text
903 @cindex line numbers, in warnings/errors
904 (where @b{NNN} is a line number). If a logical file name has been given
905 (@pxref{File,,@code{.file}}) it is used for the filename, otherwise the name of
906 the current input file is used. If a logical line number was given
908 (@pxref{Line,,@code{.line}})
912 (@pxref{Line,,@code{.line}})
915 (@pxref{Ln,,@code{.ln}})
918 then it is used to calculate the number printed,
919 otherwise the actual line in the current source file is printed. The
920 message text is intended to be self explanatory (in the grand Unix
923 @cindex format of error messages
924 Error messages have the format
926 file_name:@b{NNN}:FATAL:Error Message Text
928 The file name and line number are derived as for warning
929 messages. The actual message text may be rather less explanatory
930 because many of them aren't supposed to happen.
933 @chapter Command-Line Options
935 @cindex options, all versions of assembler
936 This chapter describes command-line options available in @emph{all}
937 versions of the @sc{gnu} assembler; @pxref{Machine Dependencies}, for options specific
939 to the @value{TARGET}.
942 to particular machine architectures.
945 If you are invoking @code{@value{AS}} via the @sc{gnu} C compiler (version 2),
946 you can use the @samp{-Wa} option to pass arguments through to the assembler.
947 The assembler arguments must be separated from each other (and the @samp{-Wa})
948 by commas. For example:
951 gcc -c -g -O -Wa,-alh,-L file.c
955 This passes two options to the assembler: @samp{-alh} (emit a listing to
956 standard output with with high-level and assembly source) and @samp{-L} (retain
957 local symbols in the symbol table).
959 Usually you do not need to use this @samp{-Wa} mechanism, since many compiler
960 command-line options are automatically passed to the assembler by the compiler.
961 (You can call the @sc{gnu} compiler driver with the @samp{-v} option to see
962 precisely what options it passes to each compilation pass, including the
966 * a:: -a[cdhlns] enable listings
967 * D:: -D for compatibility
968 * f:: -f to work faster
969 * I:: -I for .include search path
970 @ifclear DIFF-TBL-KLUGE
971 * K:: -K for compatibility
973 @ifset DIFF-TBL-KLUGE
974 * K:: -K for difference tables
977 * L:: -L to retain local labels
978 * M:: -M or --mri to assemble in MRI compatibility mode
979 * MD:: --MD for dependency tracking
980 * o:: -o to name the object file
981 * R:: -R to join data and text sections
982 * statistics:: --statistics to see statistics about assembly
983 * traditional-format:: --traditional-format for compatible output
984 * v:: -v to announce version
985 * W:: -W to suppress warnings
986 * Z:: -Z to make object file even after errors
990 @section Enable Listings: @code{-a[cdhlns]}
999 @cindex listings, enabling
1000 @cindex assembly listings, enabling
1002 These options enable listing output from the assembler. By itself,
1003 @samp{-a} requests high-level, assembly, and symbols listing.
1004 You can use other letters to select specific options for the list:
1005 @samp{-ah} requests a high-level language listing,
1006 @samp{-al} requests an output-program assembly listing, and
1007 @samp{-as} requests a symbol table listing.
1008 High-level listings require that a compiler debugging option like
1009 @samp{-g} be used, and that assembly listings (@samp{-al}) be requested
1012 Use the @samp{-ac} option to omit false conditionals from a listing. Any lines
1013 which are not assembled because of a false @code{.if} (or @code{.ifdef}, or any
1014 other conditional), or a true @code{.if} followed by an @code{.else}, will be
1015 omitted from the listing.
1017 Use the @samp{-ad} option to omit debugging directives from the
1020 Once you have specified one of these options, you can further control
1021 listing output and its appearance using the directives @code{.list},
1022 @code{.nolist}, @code{.psize}, @code{.eject}, @code{.title}, and
1024 The @samp{-an} option turns off all forms processing.
1025 If you do not request listing output with one of the @samp{-a} options, the
1026 listing-control directives have no effect.
1028 The letters after @samp{-a} may be combined into one option,
1029 @emph{e.g.}, @samp{-aln}.
1035 This option has no effect whatsoever, but it is accepted to make it more
1036 likely that scripts written for other assemblers also work with
1040 @section Work Faster: @code{-f}
1043 @cindex trusted compiler
1044 @cindex faster processing (@code{-f})
1045 @samp{-f} should only be used when assembling programs written by a
1046 (trusted) compiler. @samp{-f} stops the assembler from doing whitespace
1047 and comment preprocessing on
1048 the input file(s) before assembling them. @xref{Preprocessing,
1052 @emph{Warning:} if you use @samp{-f} when the files actually need to be
1053 preprocessed (if they contain comments, for example), @code{@value{AS}} does
1058 @section @code{.include} search path: @code{-I} @var{path}
1060 @kindex -I @var{path}
1061 @cindex paths for @code{.include}
1062 @cindex search path for @code{.include}
1063 @cindex @code{include} directive search path
1064 Use this option to add a @var{path} to the list of directories
1065 @code{@value{AS}} searches for files specified in @code{.include}
1066 directives (@pxref{Include,,@code{.include}}). You may use @code{-I} as
1067 many times as necessary to include a variety of paths. The current
1068 working directory is always searched first; after that, @code{@value{AS}}
1069 searches any @samp{-I} directories in the same order as they were
1070 specified (left to right) on the command line.
1073 @section Difference Tables: @code{-K}
1076 @ifclear DIFF-TBL-KLUGE
1077 On the @value{TARGET} family, this option is allowed, but has no effect. It is
1078 permitted for compatibility with the @sc{gnu} assembler on other platforms,
1079 where it can be used to warn when the assembler alters the machine code
1080 generated for @samp{.word} directives in difference tables. The @value{TARGET}
1081 family does not have the addressing limitations that sometimes lead to this
1082 alteration on other platforms.
1085 @ifset DIFF-TBL-KLUGE
1086 @cindex difference tables, warning
1087 @cindex warning for altered difference tables
1088 @code{@value{AS}} sometimes alters the code emitted for directives of the form
1089 @samp{.word @var{sym1}-@var{sym2}}; @pxref{Word,,@code{.word}}.
1090 You can use the @samp{-K} option if you want a warning issued when this
1095 @section Include Local Labels: @code{-L}
1098 @cindex local labels, retaining in output
1099 Labels beginning with @samp{L} (upper case only) are called @dfn{local
1100 labels}. @xref{Symbol Names}. Normally you do not see such labels when
1101 debugging, because they are intended for the use of programs (like
1102 compilers) that compose assembler programs, not for your notice.
1103 Normally both @code{@value{AS}} and @code{@value{LD}} discard such labels, so you do not
1104 normally debug with them.
1106 This option tells @code{@value{AS}} to retain those @samp{L@dots{}} symbols
1107 in the object file. Usually if you do this you also tell the linker
1108 @code{@value{LD}} to preserve symbols whose names begin with @samp{L}.
1110 By default, a local label is any label beginning with @samp{L}, but each
1111 target is allowed to redefine the local label prefix.
1113 On the HPPA local labels begin with @samp{L$}.
1116 @samp{;} for the ARM family;
1120 @section Assemble in MRI Compatibility Mode: @code{-M}
1123 @cindex MRI compatibility mode
1124 The @code{-M} or @code{--mri} option selects MRI compatibility mode. This
1125 changes the syntax and pseudo-op handling of @code{@value{AS}} to make it
1126 compatible with the @code{ASM68K} or the @code{ASM960} (depending upon the
1127 configured target) assembler from Microtec Research. The exact nature of the
1128 MRI syntax will not be documented here; see the MRI manuals for more
1129 information. Note in particular that the handling of macros and macro
1130 arguments is somewhat different. The purpose of this option is to permit
1131 assembling existing MRI assembler code using @code{@value{AS}}.
1133 The MRI compatibility is not complete. Certain operations of the MRI assembler
1134 depend upon its object file format, and can not be supported using other object
1135 file formats. Supporting these would require enhancing each object file format
1136 individually. These are:
1139 @item global symbols in common section
1141 The m68k MRI assembler supports common sections which are merged by the linker.
1142 Other object file formats do not support this. @code{@value{AS}} handles
1143 common sections by treating them as a single common symbol. It permits local
1144 symbols to be defined within a common section, but it can not support global
1145 symbols, since it has no way to describe them.
1147 @item complex relocations
1149 The MRI assemblers support relocations against a negated section address, and
1150 relocations which combine the start addresses of two or more sections. These
1151 are not support by other object file formats.
1153 @item @code{END} pseudo-op specifying start address
1155 The MRI @code{END} pseudo-op permits the specification of a start address.
1156 This is not supported by other object file formats. The start address may
1157 instead be specified using the @code{-e} option to the linker, or in a linker
1160 @item @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops
1162 The MRI @code{IDNT}, @code{.ident} and @code{NAME} pseudo-ops assign a module
1163 name to the output file. This is not supported by other object file formats.
1165 @item @code{ORG} pseudo-op
1167 The m68k MRI @code{ORG} pseudo-op begins an absolute section at a given
1168 address. This differs from the usual @code{@value{AS}} @code{.org} pseudo-op,
1169 which changes the location within the current section. Absolute sections are
1170 not supported by other object file formats. The address of a section may be
1171 assigned within a linker script.
1174 There are some other features of the MRI assembler which are not supported by
1175 @code{@value{AS}}, typically either because they are difficult or because they
1176 seem of little consequence. Some of these may be supported in future releases.
1180 @item EBCDIC strings
1182 EBCDIC strings are not supported.
1184 @item packed binary coded decimal
1186 Packed binary coded decimal is not supported. This means that the @code{DC.P}
1187 and @code{DCB.P} pseudo-ops are not supported.
1189 @item @code{FEQU} pseudo-op
1191 The m68k @code{FEQU} pseudo-op is not supported.
1193 @item @code{NOOBJ} pseudo-op
1195 The m68k @code{NOOBJ} pseudo-op is not supported.
1197 @item @code{OPT} branch control options
1199 The m68k @code{OPT} branch control options---@code{B}, @code{BRS}, @code{BRB},
1200 @code{BRL}, and @code{BRW}---are ignored. @code{@value{AS}} automatically
1201 relaxes all branches, whether forward or backward, to an appropriate size, so
1202 these options serve no purpose.
1204 @item @code{OPT} list control options
1206 The following m68k @code{OPT} list control options are ignored: @code{C},
1207 @code{CEX}, @code{CL}, @code{CRE}, @code{E}, @code{G}, @code{I}, @code{M},
1208 @code{MEX}, @code{MC}, @code{MD}, @code{X}.
1210 @item other @code{OPT} options
1212 The following m68k @code{OPT} options are ignored: @code{NEST}, @code{O},
1213 @code{OLD}, @code{OP}, @code{P}, @code{PCO}, @code{PCR}, @code{PCS}, @code{R}.
1215 @item @code{OPT} @code{D} option is default
1217 The m68k @code{OPT} @code{D} option is the default, unlike the MRI assembler.
1218 @code{OPT NOD} may be used to turn it off.
1220 @item @code{XREF} pseudo-op.
1222 The m68k @code{XREF} pseudo-op is ignored.
1224 @item @code{.debug} pseudo-op
1226 The i960 @code{.debug} pseudo-op is not supported.
1228 @item @code{.extended} pseudo-op
1230 The i960 @code{.extended} pseudo-op is not supported.
1232 @item @code{.list} pseudo-op.
1234 The various options of the i960 @code{.list} pseudo-op are not supported.
1236 @item @code{.optimize} pseudo-op
1238 The i960 @code{.optimize} pseudo-op is not supported.
1240 @item @code{.output} pseudo-op
1242 The i960 @code{.output} pseudo-op is not supported.
1244 @item @code{.setreal} pseudo-op
1246 The i960 @code{.setreal} pseudo-op is not supported.
1251 @section Dependency tracking: @code{--MD}
1254 @cindex dependency tracking
1257 @code{@value{AS}} can generate a dependency file for the file it creates. This
1258 file consists of a single rule suitable for @code{make} describing the
1259 dependencies of the main source file.
1261 The rule is written to the file named in its argument.
1263 This feature is used in the automatic updating of makefiles.
1266 @section Name the Object File: @code{-o}
1269 @cindex naming object file
1270 @cindex object file name
1271 There is always one object file output when you run @code{@value{AS}}. By
1272 default it has the name
1275 @file{a.out} (or @file{b.out}, for Intel 960 targets only).
1289 You use this option (which takes exactly one filename) to give the
1290 object file a different name.
1292 Whatever the object file is called, @code{@value{AS}} overwrites any
1293 existing file of the same name.
1296 @section Join Data and Text Sections: @code{-R}
1299 @cindex data and text sections, joining
1300 @cindex text and data sections, joining
1301 @cindex joining text and data sections
1302 @cindex merging text and data sections
1303 @code{-R} tells @code{@value{AS}} to write the object file as if all
1304 data-section data lives in the text section. This is only done at
1305 the very last moment: your binary data are the same, but data
1306 section parts are relocated differently. The data section part of
1307 your object file is zero bytes long because all its bytes are
1308 appended to the text section. (@xref{Sections,,Sections and Relocation}.)
1310 When you specify @code{-R} it would be possible to generate shorter
1311 address displacements (because we do not have to cross between text and
1312 data section). We refrain from doing this simply for compatibility with
1313 older versions of @code{@value{AS}}. In future, @code{-R} may work this way.
1316 When @code{@value{AS}} is configured for COFF output,
1317 this option is only useful if you use sections named @samp{.text} and
1322 @code{-R} is not supported for any of the HPPA targets. Using
1323 @code{-R} generates a warning from @code{@value{AS}}.
1327 @section Display Assembly Statistics: @code{--statistics}
1329 @kindex --statistics
1330 @cindex statistics, about assembly
1331 @cindex time, total for assembly
1332 @cindex space used, maximum for assembly
1333 Use @samp{--statistics} to display two statistics about the resources used by
1334 @code{@value{AS}}: the maximum amount of space allocated during the assembly
1335 (in bytes), and the total execution time taken for the assembly (in @sc{cpu}
1338 @node traditional-format
1339 @section Compatible output: @code{--traditional-format}
1341 @kindex --traditional-format
1342 For some targets, the output of @code{@value{AS}} is different in some ways
1343 from the output of some existing assembler. This switch requests
1344 @code{@value{AS}} to use the traditional format instead.
1346 For example, it disables the exception frame optimizations which
1347 @code{@value{AS}} normally does by default on @code{@value{GCC}} output.
1350 @section Announce Version: @code{-v}
1354 @cindex assembler version
1355 @cindex version of assembler
1356 You can find out what version of as is running by including the
1357 option @samp{-v} (which you can also spell as @samp{-version}) on the
1361 @section Suppress Warnings: @code{-W}
1364 @cindex suppressing warnings
1365 @cindex warnings, suppressing
1366 @code{@value{AS}} should never give a warning or error message when
1367 assembling compiler output. But programs written by people often
1368 cause @code{@value{AS}} to give a warning that a particular assumption was
1369 made. All such warnings are directed to the standard error file.
1370 If you use this option, no warnings are issued. This option only
1371 affects the warning messages: it does not change any particular of how
1372 @code{@value{AS}} assembles your file. Errors, which stop the assembly, are
1376 @section Generate Object File in Spite of Errors: @code{-Z}
1377 @cindex object file, after errors
1378 @cindex errors, continuing after
1379 After an error message, @code{@value{AS}} normally produces no output. If for
1380 some reason you are interested in object file output even after
1381 @code{@value{AS}} gives an error message on your program, use the @samp{-Z}
1382 option. If there are any errors, @code{@value{AS}} continues anyways, and
1383 writes an object file after a final warning message of the form @samp{@var{n}
1384 errors, @var{m} warnings, generating bad object file.}
1389 @cindex machine-independent syntax
1390 @cindex syntax, machine-independent
1391 This chapter describes the machine-independent syntax allowed in a
1392 source file. @code{@value{AS}} syntax is similar to what many other
1393 assemblers use; it is inspired by the BSD 4.2
1398 assembler, except that @code{@value{AS}} does not assemble Vax bit-fields.
1402 * Preprocessing:: Preprocessing
1403 * Whitespace:: Whitespace
1404 * Comments:: Comments
1405 * Symbol Intro:: Symbols
1406 * Statements:: Statements
1407 * Constants:: Constants
1411 @section Preprocessing
1413 @cindex preprocessing
1414 The @code{@value{AS}} internal preprocessor:
1416 @cindex whitespace, removed by preprocessor
1418 adjusts and removes extra whitespace. It leaves one space or tab before
1419 the keywords on a line, and turns any other whitespace on the line into
1422 @cindex comments, removed by preprocessor
1424 removes all comments, replacing them with a single space, or an
1425 appropriate number of newlines.
1427 @cindex constants, converted by preprocessor
1429 converts character constants into the appropriate numeric values.
1432 It does not do macro processing, include file handling, or
1433 anything else you may get from your C compiler's preprocessor. You can
1434 do include file processing with the @code{.include} directive
1435 (@pxref{Include,,@code{.include}}). You can use the @sc{gnu} C compiler driver
1436 to get other ``CPP'' style preprocessing, by giving the input file a
1437 @samp{.S} suffix. @xref{Overall Options,, Options Controlling the Kind of
1438 Output, gcc.info, Using GNU CC}.
1440 Excess whitespace, comments, and character constants
1441 cannot be used in the portions of the input text that are not
1444 @cindex turning preprocessing on and off
1445 @cindex preprocessing, turning on and off
1448 If the first line of an input file is @code{#NO_APP} or if you use the
1449 @samp{-f} option, whitespace and comments are not removed from the input file.
1450 Within an input file, you can ask for whitespace and comment removal in
1451 specific portions of the by putting a line that says @code{#APP} before the
1452 text that may contain whitespace or comments, and putting a line that says
1453 @code{#NO_APP} after this text. This feature is mainly intend to support
1454 @code{asm} statements in compilers whose output is otherwise free of comments
1461 @dfn{Whitespace} is one or more blanks or tabs, in any order.
1462 Whitespace is used to separate symbols, and to make programs neater for
1463 people to read. Unless within character constants
1464 (@pxref{Characters,,Character Constants}), any whitespace means the same
1465 as exactly one space.
1471 There are two ways of rendering comments to @code{@value{AS}}. In both
1472 cases the comment is equivalent to one space.
1474 Anything from @samp{/*} through the next @samp{*/} is a comment.
1475 This means you may not nest these comments.
1479 The only way to include a newline ('\n') in a comment
1480 is to use this sort of comment.
1483 /* This sort of comment does not nest. */
1486 @cindex line comment character
1487 Anything from the @dfn{line comment} character to the next newline
1488 is considered a comment and is ignored. The line comment character is
1490 @samp{;} for the AMD 29K family;
1493 @samp{;} on the ARC;
1496 @samp{;} for the H8/300 family;
1499 @samp{!} for the H8/500 family;
1502 @samp{;} for the HPPA;
1505 @samp{#} on the i960;
1508 @samp{!} for the Hitachi SH;
1511 @samp{!} on the SPARC;
1514 @samp{#} on the m32r;
1517 @samp{|} on the 680x0;
1520 @samp{#} on the Vax;
1523 @samp{!} for the Z8000;
1526 @samp{#} on the V850;
1528 see @ref{Machine Dependencies}. @refill
1529 @c FIXME What about i386, m88k, i860?
1532 On some machines there are two different line comment characters. One
1533 character only begins a comment if it is the first non-whitespace character on
1534 a line, while the other always begins a comment.
1538 The V850 assembler also supports a double dash as starting a comment that
1539 extends to the end of the line.
1545 @cindex lines starting with @code{#}
1546 @cindex logical line numbers
1547 To be compatible with past assemblers, lines that begin with @samp{#} have a
1548 special interpretation. Following the @samp{#} should be an absolute
1549 expression (@pxref{Expressions}): the logical line number of the @emph{next}
1550 line. Then a string (@pxref{Strings,, Strings}) is allowed: if present it is a
1551 new logical file name. The rest of the line, if any, should be whitespace.
1553 If the first non-whitespace characters on the line are not numeric,
1554 the line is ignored. (Just like a comment.)
1557 # This is an ordinary comment.
1558 # 42-6 "new_file_name" # New logical file name
1559 # This is logical line # 36.
1561 This feature is deprecated, and may disappear from future versions
1562 of @code{@value{AS}}.
1567 @cindex characters used in symbols
1568 @ifclear SPECIAL-SYMS
1569 A @dfn{symbol} is one or more characters chosen from the set of all
1570 letters (both upper and lower case), digits and the three characters
1576 A @dfn{symbol} is one or more characters chosen from the set of all
1577 letters (both upper and lower case), digits and the three characters
1578 @samp{._$}. (Save that, on the H8/300 only, you may not use @samp{$} in
1584 On most machines, you can also use @code{$} in symbol names; exceptions
1585 are noted in @ref{Machine Dependencies}.
1587 No symbol may begin with a digit. Case is significant.
1588 There is no length limit: all characters are significant. Symbols are
1589 delimited by characters not in that set, or by the beginning of a file
1590 (since the source program must end with a newline, the end of a file is
1591 not a possible symbol delimiter). @xref{Symbols}.
1592 @cindex length of symbols
1597 @cindex statements, structure of
1598 @cindex line separator character
1599 @cindex statement separator character
1601 @ifclear abnormal-separator
1602 A @dfn{statement} ends at a newline character (@samp{\n}) or at a
1603 semicolon (@samp{;}). The newline or semicolon is considered part of
1604 the preceding statement. Newlines and semicolons within character
1605 constants are an exception: they do not end statements.
1607 @ifset abnormal-separator
1609 A @dfn{statement} ends at a newline character (@samp{\n}) or an ``at''
1610 sign (@samp{@@}). The newline or at sign is considered part of the
1611 preceding statement. Newlines and at signs within character constants
1612 are an exception: they do not end statements.
1615 A @dfn{statement} ends at a newline character (@samp{\n}) or an exclamation
1616 point (@samp{!}). The newline or exclamation point is considered part of the
1617 preceding statement. Newlines and exclamation points within character
1618 constants are an exception: they do not end statements.
1621 A @dfn{statement} ends at a newline character (@samp{\n}); or (for the
1622 H8/300) a dollar sign (@samp{$}); or (for the
1625 (@samp{;}). The newline or separator character is considered part of
1626 the preceding statement. Newlines and separators within character
1627 constants are an exception: they do not end statements.
1632 A @dfn{statement} ends at a newline character (@samp{\n}) or line
1633 separator character. (The line separator is usually @samp{;}, unless
1634 this conflicts with the comment character; @pxref{Machine Dependencies}.) The
1635 newline or separator character is considered part of the preceding
1636 statement. Newlines and separators within character constants are an
1637 exception: they do not end statements.
1640 @cindex newline, required at file end
1641 @cindex EOF, newline must precede
1642 It is an error to end any statement with end-of-file: the last
1643 character of any input file should be a newline.@refill
1645 An empty statement is allowed, and may include whitespace. It is ignored.
1647 @cindex instructions and directives
1648 @cindex directives and instructions
1649 @c "key symbol" is not used elsewhere in the document; seems pedantic to
1650 @c @defn{} it in that case, as was done previously... doc@cygnus.com,
1652 A statement begins with zero or more labels, optionally followed by a
1653 key symbol which determines what kind of statement it is. The key
1654 symbol determines the syntax of the rest of the statement. If the
1655 symbol begins with a dot @samp{.} then the statement is an assembler
1656 directive: typically valid for any computer. If the symbol begins with
1657 a letter the statement is an assembly language @dfn{instruction}: it
1658 assembles into a machine language instruction.
1660 Different versions of @code{@value{AS}} for different computers
1661 recognize different instructions. In fact, the same symbol may
1662 represent a different instruction in a different computer's assembly
1666 @cindex @code{:} (label)
1667 @cindex label (@code{:})
1668 A label is a symbol immediately followed by a colon (@code{:}).
1669 Whitespace before a label or after a colon is permitted, but you may not
1670 have whitespace between a label's symbol and its colon. @xref{Labels}.
1673 For HPPA targets, labels need not be immediately followed by a colon, but
1674 the definition of a label must begin in column zero. This also implies that
1675 only one label may be defined on each line.
1679 label: .directive followed by something
1680 another_label: # This is an empty statement.
1681 instruction operand_1, operand_2, @dots{}
1688 A constant is a number, written so that its value is known by
1689 inspection, without knowing any context. Like this:
1692 .byte 74, 0112, 092, 0x4A, 0X4a, 'J, '\J # All the same value.
1693 .ascii "Ring the bell\7" # A string constant.
1694 .octa 0x123456789abcdef0123456789ABCDEF0 # A bignum.
1695 .float 0f-314159265358979323846264338327\
1696 95028841971.693993751E-40 # - pi, a flonum.
1701 * Characters:: Character Constants
1702 * Numbers:: Number Constants
1706 @subsection Character Constants
1708 @cindex character constants
1709 @cindex constants, character
1710 There are two kinds of character constants. A @dfn{character} stands
1711 for one character in one byte and its value may be used in
1712 numeric expressions. String constants (properly called string
1713 @emph{literals}) are potentially many bytes and their values may not be
1714 used in arithmetic expressions.
1718 * Chars:: Characters
1722 @subsubsection Strings
1724 @cindex string constants
1725 @cindex constants, string
1726 A @dfn{string} is written between double-quotes. It may contain
1727 double-quotes or null characters. The way to get special characters
1728 into a string is to @dfn{escape} these characters: precede them with
1729 a backslash @samp{\} character. For example @samp{\\} represents
1730 one backslash: the first @code{\} is an escape which tells
1731 @code{@value{AS}} to interpret the second character literally as a backslash
1732 (which prevents @code{@value{AS}} from recognizing the second @code{\} as an
1733 escape character). The complete list of escapes follows.
1735 @cindex escape codes, character
1736 @cindex character escape codes
1739 @c Mnemonic for ACKnowledge; for ASCII this is octal code 007.
1741 @cindex @code{\b} (backspace character)
1742 @cindex backspace (@code{\b})
1744 Mnemonic for backspace; for ASCII this is octal code 010.
1747 @c Mnemonic for EOText; for ASCII this is octal code 004.
1749 @cindex @code{\f} (formfeed character)
1750 @cindex formfeed (@code{\f})
1752 Mnemonic for FormFeed; for ASCII this is octal code 014.
1754 @cindex @code{\n} (newline character)
1755 @cindex newline (@code{\n})
1757 Mnemonic for newline; for ASCII this is octal code 012.
1760 @c Mnemonic for prefix; for ASCII this is octal code 033, usually known as @code{escape}.
1762 @cindex @code{\r} (carriage return character)
1763 @cindex carriage return (@code{\r})
1765 Mnemonic for carriage-Return; for ASCII this is octal code 015.
1768 @c Mnemonic for space; for ASCII this is octal code 040. Included for compliance with
1769 @c other assemblers.
1771 @cindex @code{\t} (tab)
1772 @cindex tab (@code{\t})
1774 Mnemonic for horizontal Tab; for ASCII this is octal code 011.
1777 @c Mnemonic for Vertical tab; for ASCII this is octal code 013.
1778 @c @item \x @var{digit} @var{digit} @var{digit}
1779 @c A hexadecimal character code. The numeric code is 3 hexadecimal digits.
1781 @cindex @code{\@var{ddd}} (octal character code)
1782 @cindex octal character code (@code{\@var{ddd}})
1783 @item \ @var{digit} @var{digit} @var{digit}
1784 An octal character code. The numeric code is 3 octal digits.
1785 For compatibility with other Unix systems, 8 and 9 are accepted as digits:
1786 for example, @code{\008} has the value 010, and @code{\009} the value 011.
1788 @cindex @code{\@var{xd...}} (hex character code)
1789 @cindex hex character code (@code{\@var{xd...}})
1790 @item \@code{x} @var{hex-digits...}
1791 A hex character code. All trailing hex digits are combined. Either upper or
1792 lower case @code{x} works.
1794 @cindex @code{\\} (@samp{\} character)
1795 @cindex backslash (@code{\\})
1797 Represents one @samp{\} character.
1800 @c Represents one @samp{'} (accent acute) character.
1801 @c This is needed in single character literals
1802 @c (@xref{Characters,,Character Constants}.) to represent
1805 @cindex @code{\"} (doublequote character)
1806 @cindex doublequote (@code{\"})
1808 Represents one @samp{"} character. Needed in strings to represent
1809 this character, because an unescaped @samp{"} would end the string.
1811 @item \ @var{anything-else}
1812 Any other character when escaped by @kbd{\} gives a warning, but
1813 assembles as if the @samp{\} was not present. The idea is that if
1814 you used an escape sequence you clearly didn't want the literal
1815 interpretation of the following character. However @code{@value{AS}} has no
1816 other interpretation, so @code{@value{AS}} knows it is giving you the wrong
1817 code and warns you of the fact.
1820 Which characters are escapable, and what those escapes represent,
1821 varies widely among assemblers. The current set is what we think
1822 the BSD 4.2 assembler recognizes, and is a subset of what most C
1823 compilers recognize. If you are in doubt, do not use an escape
1827 @subsubsection Characters
1829 @cindex single character constant
1830 @cindex character, single
1831 @cindex constant, single character
1832 A single character may be written as a single quote immediately
1833 followed by that character. The same escapes apply to characters as
1834 to strings. So if you want to write the character backslash, you
1835 must write @kbd{'\\} where the first @code{\} escapes the second
1836 @code{\}. As you can see, the quote is an acute accent, not a
1837 grave accent. A newline
1839 @ifclear abnormal-separator
1840 (or semicolon @samp{;})
1842 @ifset abnormal-separator
1844 (or at sign @samp{@@})
1847 (or dollar sign @samp{$}, for the H8/300; or semicolon @samp{;} for the
1853 immediately following an acute accent is taken as a literal character
1854 and does not count as the end of a statement. The value of a character
1855 constant in a numeric expression is the machine's byte-wide code for
1856 that character. @code{@value{AS}} assumes your character code is ASCII:
1857 @kbd{'A} means 65, @kbd{'B} means 66, and so on. @refill
1860 @subsection Number Constants
1862 @cindex constants, number
1863 @cindex number constants
1864 @code{@value{AS}} distinguishes three kinds of numbers according to how they
1865 are stored in the target machine. @emph{Integers} are numbers that
1866 would fit into an @code{int} in the C language. @emph{Bignums} are
1867 integers, but they are stored in more than 32 bits. @emph{Flonums}
1868 are floating point numbers, described below.
1871 * Integers:: Integers
1876 * Bit Fields:: Bit Fields
1882 @subsubsection Integers
1884 @cindex constants, integer
1886 @cindex binary integers
1887 @cindex integers, binary
1888 A binary integer is @samp{0b} or @samp{0B} followed by zero or more of
1889 the binary digits @samp{01}.
1891 @cindex octal integers
1892 @cindex integers, octal
1893 An octal integer is @samp{0} followed by zero or more of the octal
1894 digits (@samp{01234567}).
1896 @cindex decimal integers
1897 @cindex integers, decimal
1898 A decimal integer starts with a non-zero digit followed by zero or
1899 more digits (@samp{0123456789}).
1901 @cindex hexadecimal integers
1902 @cindex integers, hexadecimal
1903 A hexadecimal integer is @samp{0x} or @samp{0X} followed by one or
1904 more hexadecimal digits chosen from @samp{0123456789abcdefABCDEF}.
1906 Integers have the usual values. To denote a negative integer, use
1907 the prefix operator @samp{-} discussed under expressions
1908 (@pxref{Prefix Ops,,Prefix Operators}).
1911 @subsubsection Bignums
1914 @cindex constants, bignum
1915 A @dfn{bignum} has the same syntax and semantics as an integer
1916 except that the number (or its negative) takes more than 32 bits to
1917 represent in binary. The distinction is made because in some places
1918 integers are permitted while bignums are not.
1921 @subsubsection Flonums
1923 @cindex floating point numbers
1924 @cindex constants, floating point
1926 @cindex precision, floating point
1927 A @dfn{flonum} represents a floating point number. The translation is
1928 indirect: a decimal floating point number from the text is converted by
1929 @code{@value{AS}} to a generic binary floating point number of more than
1930 sufficient precision. This generic floating point number is converted
1931 to a particular computer's floating point format (or formats) by a
1932 portion of @code{@value{AS}} specialized to that computer.
1934 A flonum is written by writing (in order)
1939 (@samp{0} is optional on the HPPA.)
1943 A letter, to tell @code{@value{AS}} the rest of the number is a flonum.
1945 @kbd{e} is recommended. Case is not important.
1947 @c FIXME: verify if flonum syntax really this vague for most cases
1948 (Any otherwise illegal letter works here, but that might be changed. Vax BSD
1949 4.2 assembler seems to allow any of @samp{defghDEFGH}.)
1952 On the H8/300, H8/500,
1954 and AMD 29K architectures, the letter must be
1955 one of the letters @samp{DFPRSX} (in upper or lower case).
1957 On the ARC, the letter must be one of the letters @samp{DFRS}
1958 (in upper or lower case).
1960 On the Intel 960 architecture, the letter must be
1961 one of the letters @samp{DFT} (in upper or lower case).
1963 On the HPPA architecture, the letter must be @samp{E} (upper case only).
1967 One of the letters @samp{DFPRSX} (in upper or lower case).
1970 One of the letters @samp{DFRS} (in upper or lower case).
1973 One of the letters @samp{DFPRSX} (in upper or lower case).
1976 The letter @samp{E} (upper case only).
1979 One of the letters @samp{DFT} (in upper or lower case).
1984 An optional sign: either @samp{+} or @samp{-}.
1987 An optional @dfn{integer part}: zero or more decimal digits.
1990 An optional @dfn{fractional part}: @samp{.} followed by zero
1991 or more decimal digits.
1994 An optional exponent, consisting of:
1998 An @samp{E} or @samp{e}.
1999 @c I can't find a config where "EXP_CHARS" is other than 'eE', but in
2000 @c principle this can perfectly well be different on different targets.
2002 Optional sign: either @samp{+} or @samp{-}.
2004 One or more decimal digits.
2009 At least one of the integer part or the fractional part must be
2010 present. The floating point number has the usual base-10 value.
2012 @code{@value{AS}} does all processing using integers. Flonums are computed
2013 independently of any floating point hardware in the computer running
2018 @c Bit fields are written as a general facility but are also controlled
2019 @c by a conditional-compilation flag---which is as of now (21mar91)
2020 @c turned on only by the i960 config of GAS.
2022 @subsubsection Bit Fields
2025 @cindex constants, bit field
2026 You can also define numeric constants as @dfn{bit fields}.
2027 specify two numbers separated by a colon---
2029 @var{mask}:@var{value}
2032 @code{@value{AS}} applies a bitwise @sc{and} between @var{mask} and
2035 The resulting number is then packed
2037 @c this conditional paren in case bit fields turned on elsewhere than 960
2038 (in host-dependent byte order)
2040 into a field whose width depends on which assembler directive has the
2041 bit-field as its argument. Overflow (a result from the bitwise and
2042 requiring more binary digits to represent) is not an error; instead,
2043 more constants are generated, of the specified width, beginning with the
2044 least significant digits.@refill
2046 The directives @code{.byte}, @code{.hword}, @code{.int}, @code{.long},
2047 @code{.short}, and @code{.word} accept bit-field arguments.
2052 @chapter Sections and Relocation
2057 * Secs Background:: Background
2058 * Ld Sections:: Linker Sections
2059 * As Sections:: Assembler Internal Sections
2060 * Sub-Sections:: Sub-Sections
2064 @node Secs Background
2067 Roughly, a section is a range of addresses, with no gaps; all data
2068 ``in'' those addresses is treated the same for some particular purpose.
2069 For example there may be a ``read only'' section.
2071 @cindex linker, and assembler
2072 @cindex assembler, and linker
2073 The linker @code{@value{LD}} reads many object files (partial programs) and
2074 combines their contents to form a runnable program. When @code{@value{AS}}
2075 emits an object file, the partial program is assumed to start at address 0.
2076 @code{@value{LD}} assigns the final addresses for the partial program, so that
2077 different partial programs do not overlap. This is actually an
2078 oversimplification, but it suffices to explain how @code{@value{AS}} uses
2081 @code{@value{LD}} moves blocks of bytes of your program to their run-time
2082 addresses. These blocks slide to their run-time addresses as rigid
2083 units; their length does not change and neither does the order of bytes
2084 within them. Such a rigid unit is called a @emph{section}. Assigning
2085 run-time addresses to sections is called @dfn{relocation}. It includes
2086 the task of adjusting mentions of object-file addresses so they refer to
2087 the proper run-time addresses.
2089 For the H8/300 and H8/500,
2090 and for the Hitachi SH,
2091 @code{@value{AS}} pads sections if needed to
2092 ensure they end on a word (sixteen bit) boundary.
2095 @cindex standard assembler sections
2096 An object file written by @code{@value{AS}} has at least three sections, any
2097 of which may be empty. These are named @dfn{text}, @dfn{data} and
2102 When it generates COFF output,
2104 @code{@value{AS}} can also generate whatever other named sections you specify
2105 using the @samp{.section} directive (@pxref{Section,,@code{.section}}).
2106 If you do not use any directives that place output in the @samp{.text}
2107 or @samp{.data} sections, these sections still exist, but are empty.
2112 When @code{@value{AS}} generates SOM or ELF output for the HPPA,
2114 @code{@value{AS}} can also generate whatever other named sections you
2115 specify using the @samp{.space} and @samp{.subspace} directives. See
2116 @cite{HP9000 Series 800 Assembly Language Reference Manual}
2117 (HP 92432-90001) for details on the @samp{.space} and @samp{.subspace}
2118 assembler directives.
2121 Additionally, @code{@value{AS}} uses different names for the standard
2122 text, data, and bss sections when generating SOM output. Program text
2123 is placed into the @samp{$CODE$} section, data into @samp{$DATA$}, and
2124 BSS into @samp{$BSS$}.
2128 Within the object file, the text section starts at address @code{0}, the
2129 data section follows, and the bss section follows the data section.
2132 When generating either SOM or ELF output files on the HPPA, the text
2133 section starts at address @code{0}, the data section at address
2134 @code{0x4000000}, and the bss section follows the data section.
2137 To let @code{@value{LD}} know which data changes when the sections are
2138 relocated, and how to change that data, @code{@value{AS}} also writes to the
2139 object file details of the relocation needed. To perform relocation
2140 @code{@value{LD}} must know, each time an address in the object
2144 Where in the object file is the beginning of this reference to
2147 How long (in bytes) is this reference?
2149 Which section does the address refer to? What is the numeric value of
2151 (@var{address}) @minus{} (@var{start-address of section})?
2154 Is the reference to an address ``Program-Counter relative''?
2157 @cindex addresses, format of
2158 @cindex section-relative addressing
2159 In fact, every address @code{@value{AS}} ever uses is expressed as
2161 (@var{section}) + (@var{offset into section})
2164 Further, most expressions @code{@value{AS}} computes have this section-relative
2167 (For some object formats, such as SOM for the HPPA, some expressions are
2168 symbol-relative instead.)
2171 In this manual we use the notation @{@var{secname} @var{N}@} to mean ``offset
2172 @var{N} into section @var{secname}.''
2174 Apart from text, data and bss sections you need to know about the
2175 @dfn{absolute} section. When @code{@value{LD}} mixes partial programs,
2176 addresses in the absolute section remain unchanged. For example, address
2177 @code{@{absolute 0@}} is ``relocated'' to run-time address 0 by
2178 @code{@value{LD}}. Although the linker never arranges two partial programs'
2179 data sections with overlapping addresses after linking, @emph{by definition}
2180 their absolute sections must overlap. Address @code{@{absolute@ 239@}} in one
2181 part of a program is always the same address when the program is running as
2182 address @code{@{absolute@ 239@}} in any other part of the program.
2184 The idea of sections is extended to the @dfn{undefined} section. Any
2185 address whose section is unknown at assembly time is by definition
2186 rendered @{undefined @var{U}@}---where @var{U} is filled in later.
2187 Since numbers are always defined, the only way to generate an undefined
2188 address is to mention an undefined symbol. A reference to a named
2189 common block would be such a symbol: its value is unknown at assembly
2190 time so it has section @emph{undefined}.
2192 By analogy the word @emph{section} is used to describe groups of sections in
2193 the linked program. @code{@value{LD}} puts all partial programs' text
2194 sections in contiguous addresses in the linked program. It is
2195 customary to refer to the @emph{text section} of a program, meaning all
2196 the addresses of all partial programs' text sections. Likewise for
2197 data and bss sections.
2199 Some sections are manipulated by @code{@value{LD}}; others are invented for
2200 use of @code{@value{AS}} and have no meaning except during assembly.
2203 @section Linker Sections
2204 @code{@value{LD}} deals with just four kinds of sections, summarized below.
2209 @cindex named sections
2210 @cindex sections, named
2211 @item named sections
2214 @cindex text section
2215 @cindex data section
2219 These sections hold your program. @code{@value{AS}} and @code{@value{LD}} treat them as
2220 separate but equal sections. Anything you can say of one section is
2223 When the program is running, however, it is
2224 customary for the text section to be unalterable. The
2225 text section is often shared among processes: it contains
2226 instructions, constants and the like. The data section of a running
2227 program is usually alterable: for example, C variables would be stored
2228 in the data section.
2233 This section contains zeroed bytes when your program begins running. It
2234 is used to hold unitialized variables or common storage. The length of
2235 each partial program's bss section is important, but because it starts
2236 out containing zeroed bytes there is no need to store explicit zero
2237 bytes in the object file. The bss section was invented to eliminate
2238 those explicit zeros from object files.
2240 @cindex absolute section
2241 @item absolute section
2242 Address 0 of this section is always ``relocated'' to runtime address 0.
2243 This is useful if you want to refer to an address that @code{@value{LD}} must
2244 not change when relocating. In this sense we speak of absolute
2245 addresses being ``unrelocatable'': they do not change during relocation.
2247 @cindex undefined section
2248 @item undefined section
2249 This ``section'' is a catch-all for address references to objects not in
2250 the preceding sections.
2251 @c FIXME: ref to some other doc on obj-file formats could go here.
2254 @cindex relocation example
2255 An idealized example of three relocatable sections follows.
2257 The example uses the traditional section names @samp{.text} and @samp{.data}.
2259 Memory addresses are on the horizontal axis.
2263 @c END TEXI2ROFF-KILL
2266 partial program # 1: |ttttt|dddd|00|
2273 partial program # 2: |TTT|DDD|000|
2276 +--+---+-----+--+----+---+-----+~~
2277 linked program: | |TTT|ttttt| |dddd|DDD|00000|
2278 +--+---+-----+--+----+---+-----+~~
2280 addresses: 0 @dots{}
2287 \line{\it Partial program \#1: \hfil}
2288 \line{\ibox{2.5cm}{\tt text}\ibox{2cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2289 \line{\boxit{2.5cm}{\tt ttttt}\boxit{2cm}{\tt dddd}\boxit{1cm}{\tt 00}\hfil}
2291 \line{\it Partial program \#2: \hfil}
2292 \line{\ibox{1cm}{\tt text}\ibox{1.5cm}{\tt data}\ibox{1cm}{\tt bss}\hfil}
2293 \line{\boxit{1cm}{\tt TTT}\boxit{1.5cm}{\tt DDDD}\boxit{1cm}{\tt 000}\hfil}
2295 \line{\it linked program: \hfil}
2296 \line{\ibox{.5cm}{}\ibox{1cm}{\tt text}\ibox{2.5cm}{}\ibox{.75cm}{}\ibox{2cm}{\tt data}\ibox{1.5cm}{}\ibox{2cm}{\tt bss}\hfil}
2297 \line{\boxit{.5cm}{}\boxit{1cm}{\tt TTT}\boxit{2.5cm}{\tt
2298 ttttt}\boxit{.75cm}{}\boxit{2cm}{\tt dddd}\boxit{1.5cm}{\tt
2299 DDDD}\boxit{2cm}{\tt 00000}\ \dots\hfil}
2301 \line{\it addresses: \hfil}
2305 @c END TEXI2ROFF-KILL
2308 @section Assembler Internal Sections
2310 @cindex internal assembler sections
2311 @cindex sections in messages, internal
2312 These sections are meant only for the internal use of @code{@value{AS}}. They
2313 have no meaning at run-time. You do not really need to know about these
2314 sections for most purposes; but they can be mentioned in @code{@value{AS}}
2315 warning messages, so it might be helpful to have an idea of their
2316 meanings to @code{@value{AS}}. These sections are used to permit the
2317 value of every expression in your assembly language program to be a
2318 section-relative address.
2321 @cindex assembler internal logic error
2322 @item ASSEMBLER-INTERNAL-LOGIC-ERROR!
2323 An internal assembler logic error has been found. This means there is a
2324 bug in the assembler.
2326 @cindex expr (internal section)
2328 The assembler stores complex expression internally as combinations of
2329 symbols. When it needs to represent an expression as a symbol, it puts
2330 it in the expr section.
2332 @c FIXME item transfer[t] vector preload
2333 @c FIXME item transfer[t] vector postload
2334 @c FIXME item register
2338 @section Sub-Sections
2340 @cindex numbered subsections
2341 @cindex grouping data
2347 fall into two sections: text and data.
2349 You may have separate groups of
2351 data in named sections
2355 data in named sections
2361 that you want to end up near to each other in the object file, even though they
2362 are not contiguous in the assembler source. @code{@value{AS}} allows you to
2363 use @dfn{subsections} for this purpose. Within each section, there can be
2364 numbered subsections with values from 0 to 8192. Objects assembled into the
2365 same subsection go into the object file together with other objects in the same
2366 subsection. For example, a compiler might want to store constants in the text
2367 section, but might not want to have them interspersed with the program being
2368 assembled. In this case, the compiler could issue a @samp{.text 0} before each
2369 section of code being output, and a @samp{.text 1} before each group of
2370 constants being output.
2372 Subsections are optional. If you do not use subsections, everything
2373 goes in subsection number zero.
2376 Each subsection is zero-padded up to a multiple of four bytes.
2377 (Subsections may be padded a different amount on different flavors
2378 of @code{@value{AS}}.)
2382 On the H8/300 and H8/500 platforms, each subsection is zero-padded to a word
2383 boundary (two bytes).
2384 The same is true on the Hitachi SH.
2387 @c FIXME section padding (alignment)?
2388 @c Rich Pixley says padding here depends on target obj code format; that
2389 @c doesn't seem particularly useful to say without further elaboration,
2390 @c so for now I say nothing about it. If this is a generic BFD issue,
2391 @c these paragraphs might need to vanish from this manual, and be
2392 @c discussed in BFD chapter of binutils (or some such).
2395 On the AMD 29K family, no particular padding is added to section or
2396 subsection sizes; @value{AS} forces no alignment on this platform.
2400 Subsections appear in your object file in numeric order, lowest numbered
2401 to highest. (All this to be compatible with other people's assemblers.)
2402 The object file contains no representation of subsections; @code{@value{LD}} and
2403 other programs that manipulate object files see no trace of them.
2404 They just see all your text subsections as a text section, and all your
2405 data subsections as a data section.
2407 To specify which subsection you want subsequent statements assembled
2408 into, use a numeric argument to specify it, in a @samp{.text
2409 @var{expression}} or a @samp{.data @var{expression}} statement.
2412 When generating COFF output, you
2417 can also use an extra subsection
2418 argument with arbitrary named sections: @samp{.section @var{name},
2421 @var{Expression} should be an absolute expression.
2422 (@xref{Expressions}.) If you just say @samp{.text} then @samp{.text 0}
2423 is assumed. Likewise @samp{.data} means @samp{.data 0}. Assembly
2424 begins in @code{text 0}. For instance:
2426 .text 0 # The default subsection is text 0 anyway.
2427 .ascii "This lives in the first text subsection. *"
2429 .ascii "But this lives in the second text subsection."
2431 .ascii "This lives in the data section,"
2432 .ascii "in the first data subsection."
2434 .ascii "This lives in the first text section,"
2435 .ascii "immediately following the asterisk (*)."
2438 Each section has a @dfn{location counter} incremented by one for every byte
2439 assembled into that section. Because subsections are merely a convenience
2440 restricted to @code{@value{AS}} there is no concept of a subsection location
2441 counter. There is no way to directly manipulate a location counter---but the
2442 @code{.align} directive changes it, and any label definition captures its
2443 current value. The location counter of the section where statements are being
2444 assembled is said to be the @dfn{active} location counter.
2447 @section bss Section
2450 @cindex common variable storage
2451 The bss section is used for local common variable storage.
2452 You may allocate address space in the bss section, but you may
2453 not dictate data to load into it before your program executes. When
2454 your program starts running, all the contents of the bss
2455 section are zeroed bytes.
2457 The @code{.lcomm} pseudo-op defines a symbol in the bss section; see
2458 @ref{Lcomm,,@code{.lcomm}}.
2460 The @code{.comm} pseudo-op may be used to declare a common symbol, which is
2461 another form of uninitialized symbol; see @xref{Comm,,@code{.comm}}.
2464 When assembling for a target which supports multiple sections, such as ELF or
2465 COFF, you may switch into the @code{.bss} section and define symbols as usual;
2466 see @ref{Section,,@code{.section}}. You may only assemble zero values into the
2467 section. Typically the section will only contain symbol definitions and
2468 @code{.skip} directives (@pxref{Skip,,@code{.skip}}).
2475 Symbols are a central concept: the programmer uses symbols to name
2476 things, the linker uses symbols to link, and the debugger uses symbols
2480 @cindex debuggers, and symbol order
2481 @emph{Warning:} @code{@value{AS}} does not place symbols in the object file in
2482 the same order they were declared. This may break some debuggers.
2487 * Setting Symbols:: Giving Symbols Other Values
2488 * Symbol Names:: Symbol Names
2489 * Dot:: The Special Dot Symbol
2490 * Symbol Attributes:: Symbol Attributes
2497 A @dfn{label} is written as a symbol immediately followed by a colon
2498 @samp{:}. The symbol then represents the current value of the
2499 active location counter, and is, for example, a suitable instruction
2500 operand. You are warned if you use the same symbol to represent two
2501 different locations: the first definition overrides any other
2505 On the HPPA, the usual form for a label need not be immediately followed by a
2506 colon, but instead must start in column zero. Only one label may be defined on
2507 a single line. To work around this, the HPPA version of @code{@value{AS}} also
2508 provides a special directive @code{.label} for defining labels more flexibly.
2511 @node Setting Symbols
2512 @section Giving Symbols Other Values
2514 @cindex assigning values to symbols
2515 @cindex symbol values, assigning
2516 A symbol can be given an arbitrary value by writing a symbol, followed
2517 by an equals sign @samp{=}, followed by an expression
2518 (@pxref{Expressions}). This is equivalent to using the @code{.set}
2519 directive. @xref{Set,,@code{.set}}.
2522 @section Symbol Names
2524 @cindex symbol names
2525 @cindex names, symbol
2526 @ifclear SPECIAL-SYMS
2527 Symbol names begin with a letter or with one of @samp{._}. On most
2528 machines, you can also use @code{$} in symbol names; exceptions are
2529 noted in @ref{Machine Dependencies}. That character may be followed by any
2530 string of digits, letters, dollar signs (unless otherwise noted in
2531 @ref{Machine Dependencies}), and underscores.
2534 For the AMD 29K family, @samp{?} is also allowed in the
2535 body of a symbol name, though not at its beginning.
2540 Symbol names begin with a letter or with one of @samp{._}. On the
2542 H8/500, you can also use @code{$} in symbol names. That character may
2543 be followed by any string of digits, letters, dollar signs (save on the
2544 H8/300), and underscores.
2548 Case of letters is significant: @code{foo} is a different symbol name
2551 Each symbol has exactly one name. Each name in an assembly language program
2552 refers to exactly one symbol. You may use that symbol name any number of times
2555 @subheading Local Symbol Names
2557 @cindex local symbol names
2558 @cindex symbol names, local
2559 @cindex temporary symbol names
2560 @cindex symbol names, temporary
2561 Local symbols help compilers and programmers use names temporarily.
2562 There are ten local symbol names, which are re-used throughout the
2563 program. You may refer to them using the names @samp{0} @samp{1}
2564 @dots{} @samp{9}. To define a local symbol, write a label of the form
2565 @samp{@b{N}:} (where @b{N} represents any digit). To refer to the most
2566 recent previous definition of that symbol write @samp{@b{N}b}, using the
2567 same digit as when you defined the label. To refer to the next
2568 definition of a local label, write @samp{@b{N}f}---where @b{N} gives you
2569 a choice of 10 forward references. The @samp{b} stands for
2570 ``backwards'' and the @samp{f} stands for ``forwards''.
2572 Local symbols are not emitted by the current @sc{gnu} C compiler.
2574 There is no restriction on how you can use these labels, but
2575 remember that at any point in the assembly you can refer to at most
2576 10 prior local labels and to at most 10 forward local labels.
2578 Local symbol names are only a notation device. They are immediately
2579 transformed into more conventional symbol names before the assembler
2580 uses them. The symbol names stored in the symbol table, appearing in
2581 error messages and optionally emitted to the object file have these
2586 All local labels begin with @samp{L}. Normally both @code{@value{AS}} and
2587 @code{@value{LD}} forget symbols that start with @samp{L}. These labels are
2588 used for symbols you are never intended to see. If you use the
2589 @samp{-L} option then @code{@value{AS}} retains these symbols in the
2590 object file. If you also instruct @code{@value{LD}} to retain these symbols,
2591 you may use them in debugging.
2594 If the label is written @samp{0:} then the digit is @samp{0}.
2595 If the label is written @samp{1:} then the digit is @samp{1}.
2596 And so on up through @samp{9:}.
2599 This unusual character is included so you do not accidentally invent
2600 a symbol of the same name. The character has ASCII value
2603 @item @emph{ordinal number}
2604 This is a serial number to keep the labels distinct. The first
2605 @samp{0:} gets the number @samp{1}; The 15th @samp{0:} gets the
2606 number @samp{15}; @emph{etc.}. Likewise for the other labels @samp{1:}
2610 For instance, the first @code{1:} is named @code{L1@kbd{C-A}1}, the 44th
2611 @code{3:} is named @code{L3@kbd{C-A}44}.
2614 @section The Special Dot Symbol
2616 @cindex dot (symbol)
2617 @cindex @code{.} (symbol)
2618 @cindex current address
2619 @cindex location counter
2620 The special symbol @samp{.} refers to the current address that
2621 @code{@value{AS}} is assembling into. Thus, the expression @samp{melvin:
2622 .long .} defines @code{melvin} to contain its own address.
2623 Assigning a value to @code{.} is treated the same as a @code{.org}
2624 directive. Thus, the expression @samp{.=.+4} is the same as saying
2625 @ifclear no-space-dir
2634 @node Symbol Attributes
2635 @section Symbol Attributes
2637 @cindex symbol attributes
2638 @cindex attributes, symbol
2639 Every symbol has, as well as its name, the attributes ``Value'' and
2640 ``Type''. Depending on output format, symbols can also have auxiliary
2643 The detailed definitions are in @file{a.out.h}.
2646 If you use a symbol without defining it, @code{@value{AS}} assumes zero for
2647 all these attributes, and probably won't warn you. This makes the
2648 symbol an externally defined symbol, which is generally what you
2652 * Symbol Value:: Value
2653 * Symbol Type:: Type
2656 * a.out Symbols:: Symbol Attributes: @code{a.out}
2660 * a.out Symbols:: Symbol Attributes: @code{a.out}
2663 * a.out Symbols:: Symbol Attributes: @code{a.out}, @code{b.out}
2668 * COFF Symbols:: Symbol Attributes for COFF
2671 * SOM Symbols:: Symbol Attributes for SOM
2678 @cindex value of a symbol
2679 @cindex symbol value
2680 The value of a symbol is (usually) 32 bits. For a symbol which labels a
2681 location in the text, data, bss or absolute sections the value is the
2682 number of addresses from the start of that section to the label.
2683 Naturally for text, data and bss sections the value of a symbol changes
2684 as @code{@value{LD}} changes section base addresses during linking. Absolute
2685 symbols' values do not change during linking: that is why they are
2688 The value of an undefined symbol is treated in a special way. If it is
2689 0 then the symbol is not defined in this assembler source file, and
2690 @code{@value{LD}} tries to determine its value from other files linked into the
2691 same program. You make this kind of symbol simply by mentioning a symbol
2692 name without defining it. A non-zero value represents a @code{.comm}
2693 common declaration. The value is how much common storage to reserve, in
2694 bytes (addresses). The symbol refers to the first address of the
2700 @cindex type of a symbol
2702 The type attribute of a symbol contains relocation (section)
2703 information, any flag settings indicating that a symbol is external, and
2704 (optionally), other information for linkers and debuggers. The exact
2705 format depends on the object-code output format in use.
2710 @c The following avoids a "widow" subsection title. @group would be
2711 @c better if it were available outside examples.
2714 @subsection Symbol Attributes: @code{a.out}, @code{b.out}
2716 @cindex @code{b.out} symbol attributes
2717 @cindex symbol attributes, @code{b.out}
2718 These symbol attributes appear only when @code{@value{AS}} is configured for
2719 one of the Berkeley-descended object output formats---@code{a.out} or
2725 @subsection Symbol Attributes: @code{a.out}
2727 @cindex @code{a.out} symbol attributes
2728 @cindex symbol attributes, @code{a.out}
2734 @subsection Symbol Attributes: @code{a.out}
2736 @cindex @code{a.out} symbol attributes
2737 @cindex symbol attributes, @code{a.out}
2741 * Symbol Desc:: Descriptor
2742 * Symbol Other:: Other
2746 @subsubsection Descriptor
2748 @cindex descriptor, of @code{a.out} symbol
2749 This is an arbitrary 16-bit value. You may establish a symbol's
2750 descriptor value by using a @code{.desc} statement
2751 (@pxref{Desc,,@code{.desc}}). A descriptor value means nothing to
2755 @subsubsection Other
2757 @cindex other attribute, of @code{a.out} symbol
2758 This is an arbitrary 8-bit value. It means nothing to @code{@value{AS}}.
2763 @subsection Symbol Attributes for COFF
2765 @cindex COFF symbol attributes
2766 @cindex symbol attributes, COFF
2768 The COFF format supports a multitude of auxiliary symbol attributes;
2769 like the primary symbol attributes, they are set between @code{.def} and
2770 @code{.endef} directives.
2772 @subsubsection Primary Attributes
2774 @cindex primary attributes, COFF symbols
2775 The symbol name is set with @code{.def}; the value and type,
2776 respectively, with @code{.val} and @code{.type}.
2778 @subsubsection Auxiliary Attributes
2780 @cindex auxiliary attributes, COFF symbols
2781 The @code{@value{AS}} directives @code{.dim}, @code{.line}, @code{.scl},
2782 @code{.size}, and @code{.tag} can generate auxiliary symbol table
2783 information for COFF.
2788 @subsection Symbol Attributes for SOM
2790 @cindex SOM symbol attributes
2791 @cindex symbol attributes, SOM
2793 The SOM format for the HPPA supports a multitude of symbol attributes set with
2794 the @code{.EXPORT} and @code{.IMPORT} directives.
2796 The attributes are described in @cite{HP9000 Series 800 Assembly
2797 Language Reference Manual} (HP 92432-90001) under the @code{IMPORT} and
2798 @code{EXPORT} assembler directive documentation.
2802 @chapter Expressions
2806 @cindex numeric values
2807 An @dfn{expression} specifies an address or numeric value.
2808 Whitespace may precede and/or follow an expression.
2810 The result of an expression must be an absolute number, or else an offset into
2811 a particular section. If an expression is not absolute, and there is not
2812 enough information when @code{@value{AS}} sees the expression to know its
2813 section, a second pass over the source program might be necessary to interpret
2814 the expression---but the second pass is currently not implemented.
2815 @code{@value{AS}} aborts with an error message in this situation.
2818 * Empty Exprs:: Empty Expressions
2819 * Integer Exprs:: Integer Expressions
2823 @section Empty Expressions
2825 @cindex empty expressions
2826 @cindex expressions, empty
2827 An empty expression has no value: it is just whitespace or null.
2828 Wherever an absolute expression is required, you may omit the
2829 expression, and @code{@value{AS}} assumes a value of (absolute) 0. This
2830 is compatible with other assemblers.
2833 @section Integer Expressions
2835 @cindex integer expressions
2836 @cindex expressions, integer
2837 An @dfn{integer expression} is one or more @emph{arguments} delimited
2838 by @emph{operators}.
2841 * Arguments:: Arguments
2842 * Operators:: Operators
2843 * Prefix Ops:: Prefix Operators
2844 * Infix Ops:: Infix Operators
2848 @subsection Arguments
2850 @cindex expression arguments
2851 @cindex arguments in expressions
2852 @cindex operands in expressions
2853 @cindex arithmetic operands
2854 @dfn{Arguments} are symbols, numbers or subexpressions. In other
2855 contexts arguments are sometimes called ``arithmetic operands''. In
2856 this manual, to avoid confusing them with the ``instruction operands'' of
2857 the machine language, we use the term ``argument'' to refer to parts of
2858 expressions only, reserving the word ``operand'' to refer only to machine
2859 instruction operands.
2861 Symbols are evaluated to yield @{@var{section} @var{NNN}@} where
2862 @var{section} is one of text, data, bss, absolute,
2863 or undefined. @var{NNN} is a signed, 2's complement 32 bit
2866 Numbers are usually integers.
2868 A number can be a flonum or bignum. In this case, you are warned
2869 that only the low order 32 bits are used, and @code{@value{AS}} pretends
2870 these 32 bits are an integer. You may write integer-manipulating
2871 instructions that act on exotic constants, compatible with other
2874 @cindex subexpressions
2875 Subexpressions are a left parenthesis @samp{(} followed by an integer
2876 expression, followed by a right parenthesis @samp{)}; or a prefix
2877 operator followed by an argument.
2880 @subsection Operators
2882 @cindex operators, in expressions
2883 @cindex arithmetic functions
2884 @cindex functions, in expressions
2885 @dfn{Operators} are arithmetic functions, like @code{+} or @code{%}. Prefix
2886 operators are followed by an argument. Infix operators appear
2887 between their arguments. Operators may be preceded and/or followed by
2891 @subsection Prefix Operator
2893 @cindex prefix operators
2894 @code{@value{AS}} has the following @dfn{prefix operators}. They each take
2895 one argument, which must be absolute.
2897 @c the tex/end tex stuff surrounding this small table is meant to make
2898 @c it align, on the printed page, with the similar table in the next
2899 @c section (which is inside an enumerate).
2901 \global\advance\leftskip by \itemindent
2906 @dfn{Negation}. Two's complement negation.
2908 @dfn{Complementation}. Bitwise not.
2912 \global\advance\leftskip by -\itemindent
2916 @subsection Infix Operators
2918 @cindex infix operators
2919 @cindex operators, permitted arguments
2920 @dfn{Infix operators} take two arguments, one on either side. Operators
2921 have precedence, but operations with equal precedence are performed left
2922 to right. Apart from @code{+} or @code{-}, both arguments must be
2923 absolute, and the result is absolute.
2926 @cindex operator precedence
2927 @cindex precedence of operators
2934 @dfn{Multiplication}.
2937 @dfn{Division}. Truncation is the same as the C operator @samp{/}
2944 @dfn{Shift Left}. Same as the C operator @samp{<<}.
2948 @dfn{Shift Right}. Same as the C operator @samp{>>}.
2952 Intermediate precedence
2957 @dfn{Bitwise Inclusive Or}.
2963 @dfn{Bitwise Exclusive Or}.
2966 @dfn{Bitwise Or Not}.
2973 @cindex addition, permitted arguments
2974 @cindex plus, permitted arguments
2975 @cindex arguments for addition
2977 @dfn{Addition}. If either argument is absolute, the result has the section of
2978 the other argument. You may not add together arguments from different
2981 @cindex subtraction, permitted arguments
2982 @cindex minus, permitted arguments
2983 @cindex arguments for subtraction
2985 @dfn{Subtraction}. If the right argument is absolute, the
2986 result has the section of the left argument.
2987 If both arguments are in the same section, the result is absolute.
2988 You may not subtract arguments from different sections.
2989 @c FIXME is there still something useful to say about undefined - undefined ?
2993 In short, it's only meaningful to add or subtract the @emph{offsets} in an
2994 address; you can only have a defined section in one of the two arguments.
2997 @chapter Assembler Directives
2999 @cindex directives, machine independent
3000 @cindex pseudo-ops, machine independent
3001 @cindex machine independent directives
3002 All assembler directives have names that begin with a period (@samp{.}).
3003 The rest of the name is letters, usually in lower case.
3005 This chapter discusses directives that are available regardless of the
3006 target machine configuration for the @sc{gnu} assembler.
3008 Some machine configurations provide additional directives.
3009 @xref{Machine Dependencies}.
3012 @ifset machine-directives
3013 @xref{Machine Dependencies} for additional directives.
3018 * Abort:: @code{.abort}
3020 * ABORT:: @code{.ABORT}
3023 * Align:: @code{.align @var{abs-expr} , @var{abs-expr}}
3024 * Ascii:: @code{.ascii "@var{string}"}@dots{}
3025 * Asciz:: @code{.asciz "@var{string}"}@dots{}
3026 * Balign:: @code{.balign @var{abs-expr} , @var{abs-expr}}
3027 * Byte:: @code{.byte @var{expressions}}
3028 * Comm:: @code{.comm @var{symbol} , @var{length} }
3029 * Data:: @code{.data @var{subsection}}
3031 * Def:: @code{.def @var{name}}
3034 * Desc:: @code{.desc @var{symbol}, @var{abs-expression}}
3040 * Double:: @code{.double @var{flonums}}
3041 * Eject:: @code{.eject}
3042 * Else:: @code{.else}
3045 * Endef:: @code{.endef}
3048 * Endfunc:: @code{.endfunc}
3049 * Endif:: @code{.endif}
3050 * Equ:: @code{.equ @var{symbol}, @var{expression}}
3051 * Equiv:: @code{.equiv @var{symbol}, @var{expression}}
3053 * Exitm:: @code{.exitm}
3054 * Extern:: @code{.extern}
3055 * Fail:: @code{.fail}
3056 @ifclear no-file-dir
3057 * File:: @code{.file @var{string}}
3060 * Fill:: @code{.fill @var{repeat} , @var{size} , @var{value}}
3061 * Float:: @code{.float @var{flonums}}
3062 * Func:: @code{.func}
3063 * Global:: @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3064 * hword:: @code{.hword @var{expressions}}
3065 * Ident:: @code{.ident}
3066 * If:: @code{.if @var{absolute expression}}
3067 * Include:: @code{.include "@var{file}"}
3068 * Int:: @code{.int @var{expressions}}
3069 * Irp:: @code{.irp @var{symbol},@var{values}}@dots{}
3070 * Irpc:: @code{.irpc @var{symbol},@var{values}}@dots{}
3071 * Lcomm:: @code{.lcomm @var{symbol} , @var{length}}
3072 * Lflags:: @code{.lflags}
3073 @ifclear no-line-dir
3074 * Line:: @code{.line @var{line-number}}
3077 * Ln:: @code{.ln @var{line-number}}
3078 * Linkonce:: @code{.linkonce [@var{type}]}
3079 * List:: @code{.list}
3080 * Long:: @code{.long @var{expressions}}
3082 * Lsym:: @code{.lsym @var{symbol}, @var{expression}}
3085 * Macro:: @code{.macro @var{name} @var{args}}@dots{}
3086 * MRI:: @code{.mri @var{val}}
3088 * Nolist:: @code{.nolist}
3089 * Octa:: @code{.octa @var{bignums}}
3090 * Org:: @code{.org @var{new-lc} , @var{fill}}
3091 * P2align:: @code{.p2align @var{abs-expr} , @var{abs-expr}}
3092 * Print:: @code{.print @var{string}}
3093 * Psize:: @code{.psize @var{lines}, @var{columns}}
3094 * Purgem:: @code{.purgem @var{name}}
3095 * Quad:: @code{.quad @var{bignums}}
3096 * Rept:: @code{.rept @var{count}}
3097 * Sbttl:: @code{.sbttl "@var{subheading}"}
3099 * Scl:: @code{.scl @var{class}}
3100 * Section:: @code{.section @var{name}, @var{subsection}}
3103 * Set:: @code{.set @var{symbol}, @var{expression}}
3104 * Short:: @code{.short @var{expressions}}
3105 * Single:: @code{.single @var{flonums}}
3107 * Size:: @code{.size}
3110 * Skip:: @code{.skip @var{size} , @var{fill}}
3111 * Sleb128:: @code{.sleb128 @var{expressions}}
3112 * Space:: @code{.space @var{size} , @var{fill}}
3114 * Stab:: @code{.stabd, .stabn, .stabs}
3117 * String:: @code{.string "@var{str}"}
3118 * Struct:: @code{.struct @var{expression}}
3120 * Symver:: @code{.symver @var{name},@var{name2@@nodename}}
3123 * Tag:: @code{.tag @var{structname}}
3126 * Text:: @code{.text @var{subsection}}
3127 * Title:: @code{.title "@var{heading}"}
3129 * Type:: @code{.type @var{int}}
3130 * Val:: @code{.val @var{addr}}
3133 * Uleb128:: @code{.uleb128 @var{expressions}}
3134 * Word:: @code{.word @var{expressions}}
3135 * Deprecated:: Deprecated Directives
3139 @section @code{.abort}
3141 @cindex @code{abort} directive
3142 @cindex stopping the assembly
3143 This directive stops the assembly immediately. It is for
3144 compatibility with other assemblers. The original idea was that the
3145 assembly language source would be piped into the assembler. If the sender
3146 of the source quit, it could use this directive tells @code{@value{AS}} to
3147 quit also. One day @code{.abort} will not be supported.
3151 @section @code{.ABORT}
3153 @cindex @code{ABORT} directive
3154 When producing COFF output, @code{@value{AS}} accepts this directive as a
3155 synonym for @samp{.abort}.
3158 When producing @code{b.out} output, @code{@value{AS}} accepts this directive,
3164 @section @code{.align @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3166 @cindex padding the location counter
3167 @cindex @code{align} directive
3168 Pad the location counter (in the current subsection) to a particular storage
3169 boundary. The first expression (which must be absolute) is the alignment
3170 required, as described below.
3172 The second expression (also absolute) gives the fill value to be stored in the
3173 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3174 padding bytes are normally zero. However, on some systems, if the section is
3175 marked as containing code and the fill value is omitted, the space is filled
3176 with no-op instructions.
3178 The third expression is also absolute, and is also optional. If it is present,
3179 it is the maximum number of bytes that should be skipped by this alignment
3180 directive. If doing the alignment would require skipping more bytes than the
3181 specified maximum, then the alignment is not done at all. You can omit the
3182 fill value (the second argument) entirely by simply using two commas after the
3183 required alignment; this can be useful if you want the alignment to be filled
3184 with no-op instructions when appropriate.
3186 The way the required alignment is specified varies from system to system.
3187 For the a29k, hppa, m68k, m88k, w65, sparc, and Hitachi SH, and i386 using ELF
3189 the first expression is the
3190 alignment request in bytes. For example @samp{.align 8} advances
3191 the location counter until it is a multiple of 8. If the location counter
3192 is already a multiple of 8, no change is needed.
3194 For other systems, including the i386 using a.out format, it is the
3195 number of low-order zero bits the location counter must have after
3196 advancement. For example @samp{.align 3} advances the location
3197 counter until it a multiple of 8. If the location counter is already a
3198 multiple of 8, no change is needed.
3200 This inconsistency is due to the different behaviors of the various
3201 native assemblers for these systems which GAS must emulate.
3202 GAS also provides @code{.balign} and @code{.p2align} directives,
3203 described later, which have a consistent behavior across all
3204 architectures (but are specific to GAS).
3207 @section @code{.ascii "@var{string}"}@dots{}
3209 @cindex @code{ascii} directive
3210 @cindex string literals
3211 @code{.ascii} expects zero or more string literals (@pxref{Strings})
3212 separated by commas. It assembles each string (with no automatic
3213 trailing zero byte) into consecutive addresses.
3216 @section @code{.asciz "@var{string}"}@dots{}
3218 @cindex @code{asciz} directive
3219 @cindex zero-terminated strings
3220 @cindex null-terminated strings
3221 @code{.asciz} is just like @code{.ascii}, but each string is followed by
3222 a zero byte. The ``z'' in @samp{.asciz} stands for ``zero''.
3225 @section @code{.balign[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
3227 @cindex padding the location counter given number of bytes
3228 @cindex @code{balign} directive
3229 Pad the location counter (in the current subsection) to a particular
3230 storage boundary. The first expression (which must be absolute) is the
3231 alignment request in bytes. For example @samp{.balign 8} advances
3232 the location counter until it is a multiple of 8. If the location counter
3233 is already a multiple of 8, no change is needed.
3235 The second expression (also absolute) gives the fill value to be stored in the
3236 padding bytes. It (and the comma) may be omitted. If it is omitted, the
3237 padding bytes are normally zero. However, on some systems, if the section is
3238 marked as containing code and the fill value is omitted, the space is filled
3239 with no-op instructions.
3241 The third expression is also absolute, and is also optional. If it is present,
3242 it is the maximum number of bytes that should be skipped by this alignment
3243 directive. If doing the alignment would require skipping more bytes than the
3244 specified maximum, then the alignment is not done at all. You can omit the
3245 fill value (the second argument) entirely by simply using two commas after the
3246 required alignment; this can be useful if you want the alignment to be filled
3247 with no-op instructions when appropriate.
3249 @cindex @code{balignw} directive
3250 @cindex @code{balignl} directive
3251 The @code{.balignw} and @code{.balignl} directives are variants of the
3252 @code{.balign} directive. The @code{.balignw} directive treats the fill
3253 pattern as a two byte word value. The @code{.balignl} directives treats the
3254 fill pattern as a four byte longword value. For example, @code{.balignw
3255 4,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
3256 filled in with the value 0x368d (the exact placement of the bytes depends upon
3257 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
3261 @section @code{.byte @var{expressions}}
3263 @cindex @code{byte} directive
3264 @cindex integers, one byte
3265 @code{.byte} expects zero or more expressions, separated by commas.
3266 Each expression is assembled into the next byte.
3269 @section @code{.comm @var{symbol} , @var{length} }
3271 @cindex @code{comm} directive
3272 @cindex symbol, common
3273 @code{.comm} declares a common symbol named @var{symbol}. When linking, a
3274 common symbol in one object file may be merged with a defined or common symbol
3275 of the same name in another object file. If @code{@value{LD}} does not see a
3276 definition for the symbol--just one or more common symbols--then it will
3277 allocate @var{length} bytes of uninitialized memory. @var{length} must be an
3278 absolute expression. If @code{@value{LD}} sees multiple common symbols with
3279 the same name, and they do not all have the same size, it will allocate space
3280 using the largest size.
3283 When using ELF, the @code{.comm} directive takes an optional third argument.
3284 This is the desired alignment of the symbol, specified as a byte boundary (for
3285 example, an alignment of 16 means that the least significant 4 bits of the
3286 address should be zero). The alignment must be an absolute expression, and it
3287 must be a power of two. If @code{@value{LD}} allocates uninitialized memory
3288 for the common symbol, it will use the alignment when placing the symbol. If
3289 no alignment is specified, @code{@value{AS}} will set the alignment to the
3290 largest power of two less than or equal to the size of the symbol, up to a
3295 The syntax for @code{.comm} differs slightly on the HPPA. The syntax is
3296 @samp{@var{symbol} .comm, @var{length}}; @var{symbol} is optional.
3300 @section @code{.data @var{subsection}}
3302 @cindex @code{data} directive
3303 @code{.data} tells @code{@value{AS}} to assemble the following statements onto the
3304 end of the data subsection numbered @var{subsection} (which is an
3305 absolute expression). If @var{subsection} is omitted, it defaults
3310 @section @code{.def @var{name}}
3312 @cindex @code{def} directive
3313 @cindex COFF symbols, debugging
3314 @cindex debugging COFF symbols
3315 Begin defining debugging information for a symbol @var{name}; the
3316 definition extends until the @code{.endef} directive is encountered.
3319 This directive is only observed when @code{@value{AS}} is configured for COFF
3320 format output; when producing @code{b.out}, @samp{.def} is recognized,
3327 @section @code{.desc @var{symbol}, @var{abs-expression}}
3329 @cindex @code{desc} directive
3330 @cindex COFF symbol descriptor
3331 @cindex symbol descriptor, COFF
3332 This directive sets the descriptor of the symbol (@pxref{Symbol Attributes})
3333 to the low 16 bits of an absolute expression.
3336 The @samp{.desc} directive is not available when @code{@value{AS}} is
3337 configured for COFF output; it is only for @code{a.out} or @code{b.out}
3338 object format. For the sake of compatibility, @code{@value{AS}} accepts
3339 it, but produces no output, when configured for COFF.
3345 @section @code{.dim}
3347 @cindex @code{dim} directive
3348 @cindex COFF auxiliary symbol information
3349 @cindex auxiliary symbol information, COFF
3350 This directive is generated by compilers to include auxiliary debugging
3351 information in the symbol table. It is only permitted inside
3352 @code{.def}/@code{.endef} pairs.
3355 @samp{.dim} is only meaningful when generating COFF format output; when
3356 @code{@value{AS}} is generating @code{b.out}, it accepts this directive but
3362 @section @code{.double @var{flonums}}
3364 @cindex @code{double} directive
3365 @cindex floating point numbers (double)
3366 @code{.double} expects zero or more flonums, separated by commas. It
3367 assembles floating point numbers.
3369 The exact kind of floating point numbers emitted depends on how
3370 @code{@value{AS}} is configured. @xref{Machine Dependencies}.
3374 On the @value{TARGET} family @samp{.double} emits 64-bit floating-point numbers
3375 in @sc{ieee} format.
3380 @section @code{.eject}
3382 @cindex @code{eject} directive
3383 @cindex new page, in listings
3384 @cindex page, in listings
3385 @cindex listing control: new page
3386 Force a page break at this point, when generating assembly listings.
3389 @section @code{.else}
3391 @cindex @code{else} directive
3392 @code{.else} is part of the @code{@value{AS}} support for conditional
3393 assembly; @pxref{If,,@code{.if}}. It marks the beginning of a section
3394 of code to be assembled if the condition for the preceding @code{.if}
3398 @section @code{.end}
3400 @cindex @code{end} directive
3401 @code{.end} marks the end of the assembly file. @code{@value{AS}} does not
3402 process anything in the file past the @code{.end} directive.
3406 @section @code{.endef}
3408 @cindex @code{endef} directive
3409 This directive flags the end of a symbol definition begun with
3413 @samp{.endef} is only meaningful when generating COFF format output; if
3414 @code{@value{AS}} is configured to generate @code{b.out}, it accepts this
3415 directive but ignores it.
3420 @section @code{.endfunc}
3421 @cindex @code{endfunc} directive
3422 @code{.endfunc} marks the end of a function specified with @code{.func}.
3425 @section @code{.endif}
3427 @cindex @code{endif} directive
3428 @code{.endif} is part of the @code{@value{AS}} support for conditional assembly;
3429 it marks the end of a block of code that is only assembled
3430 conditionally. @xref{If,,@code{.if}}.
3433 @section @code{.equ @var{symbol}, @var{expression}}
3435 @cindex @code{equ} directive
3436 @cindex assigning values to symbols
3437 @cindex symbols, assigning values to
3438 This directive sets the value of @var{symbol} to @var{expression}.
3439 It is synonymous with @samp{.set}; @pxref{Set,,@code{.set}}.
3442 The syntax for @code{equ} on the HPPA is
3443 @samp{@var{symbol} .equ @var{expression}}.
3447 @section @code{.equiv @var{symbol}, @var{expression}}
3448 @cindex @code{equiv} directive
3449 The @code{.equiv} directive is like @code{.equ} and @code{.set}, except that
3450 the assembler will signal an error if @var{symbol} is already defined.
3452 Except for the contents of the error message, this is roughly equivalent to
3461 @section @code{.err}
3462 @cindex @code{err} directive
3463 If @code{@value{AS}} assembles a @code{.err} directive, it will print an error
3464 message and, unless the @code{-Z} option was used, it will not generate an
3465 object file. This can be used to signal error an conditionally compiled code.
3468 @section @code{.exitm}
3469 Exit early from the current macro definition. @xref{Macro}.
3472 @section @code{.extern}
3474 @cindex @code{extern} directive
3475 @code{.extern} is accepted in the source program---for compatibility
3476 with other assemblers---but it is ignored. @code{@value{AS}} treats
3477 all undefined symbols as external.
3480 @section @code{.fail @var{expression}}
3482 @cindex @code{fail} directive
3483 Generates an error or a warning. If the value of the @var{expression} is 500
3484 or more, @code{@value{AS}} will print a warning message. If the value is less
3485 than 500, @code{@value{AS}} will print an error message. The message will
3486 include the value of @var{expression}. This can occasionally be useful inside
3487 complex nested macros or conditional assembly.
3489 @ifclear no-file-dir
3491 @section @code{.file @var{string}}
3493 @cindex @code{file} directive
3494 @cindex logical file name
3495 @cindex file name, logical
3496 @code{.file} tells @code{@value{AS}} that we are about to start a new logical
3497 file. @var{string} is the new file name. In general, the filename is
3498 recognized whether or not it is surrounded by quotes @samp{"}; but if you wish
3499 to specify an empty file name, you must give the quotes--@code{""}. This
3500 statement may go away in future: it is only recognized to be compatible with
3501 old @code{@value{AS}} programs.
3503 In some configurations of @code{@value{AS}}, @code{.file} has already been
3504 removed to avoid conflicts with other assemblers. @xref{Machine Dependencies}.
3509 @section @code{.fill @var{repeat} , @var{size} , @var{value}}
3511 @cindex @code{fill} directive
3512 @cindex writing patterns in memory
3513 @cindex patterns, writing in memory
3514 @var{result}, @var{size} and @var{value} are absolute expressions.
3515 This emits @var{repeat} copies of @var{size} bytes. @var{Repeat}
3516 may be zero or more. @var{Size} may be zero or more, but if it is
3517 more than 8, then it is deemed to have the value 8, compatible with
3518 other people's assemblers. The contents of each @var{repeat} bytes
3519 is taken from an 8-byte number. The highest order 4 bytes are
3520 zero. The lowest order 4 bytes are @var{value} rendered in the
3521 byte-order of an integer on the computer @code{@value{AS}} is assembling for.
3522 Each @var{size} bytes in a repetition is taken from the lowest order
3523 @var{size} bytes of this number. Again, this bizarre behavior is
3524 compatible with other people's assemblers.
3526 @var{size} and @var{value} are optional.
3527 If the second comma and @var{value} are absent, @var{value} is
3528 assumed zero. If the first comma and following tokens are absent,
3529 @var{size} is assumed to be 1.
3532 @section @code{.float @var{flonums}}
3534 @cindex floating point numbers (single)
3535 @cindex @code{float} directive
3536 This directive assembles zero or more flonums, separated by commas. It
3537 has the same effect as @code{.single}.
3539 The exact kind of floating point numbers emitted depends on how
3540 @code{@value{AS}} is configured.
3541 @xref{Machine Dependencies}.
3545 On the @value{TARGET} family, @code{.float} emits 32-bit floating point numbers
3546 in @sc{ieee} format.
3551 @section @code{.func @var{name}[,@var{label}]}
3552 @cindex @code{func} directive
3553 @code{.func} emits debugging information to denote function @var{name}, and
3554 is ignored unless the file is assembled with debugging enabled.
3555 Only @samp{--gstabs} is currently supported.
3556 @var{label} is the entry point of the function and if omitted @var{name}
3557 prepended with the @samp{leading char} is used.
3558 @samp{leading char} is usually @code{_} or nothing, depending on the target.
3559 All functions are currently defined to have @code{void} return type.
3560 The function must be terminated with @code{.endfunc}.
3563 @section @code{.global @var{symbol}}, @code{.globl @var{symbol}}
3565 @cindex @code{global} directive
3566 @cindex symbol, making visible to linker
3567 @code{.global} makes the symbol visible to @code{@value{LD}}. If you define
3568 @var{symbol} in your partial program, its value is made available to
3569 other partial programs that are linked with it. Otherwise,
3570 @var{symbol} takes its attributes from a symbol of the same name
3571 from another file linked into the same program.
3573 Both spellings (@samp{.globl} and @samp{.global}) are accepted, for
3574 compatibility with other assemblers.
3577 On the HPPA, @code{.global} is not always enough to make it accessible to other
3578 partial programs. You may need the HPPA-only @code{.EXPORT} directive as well.
3579 @xref{HPPA Directives,, HPPA Assembler Directives}.
3583 @section @code{.hword @var{expressions}}
3585 @cindex @code{hword} directive
3586 @cindex integers, 16-bit
3587 @cindex numbers, 16-bit
3588 @cindex sixteen bit integers
3589 This expects zero or more @var{expressions}, and emits
3590 a 16 bit number for each.
3593 This directive is a synonym for @samp{.short}; depending on the target
3594 architecture, it may also be a synonym for @samp{.word}.
3598 This directive is a synonym for @samp{.short}.
3601 This directive is a synonym for both @samp{.short} and @samp{.word}.
3606 @section @code{.ident}
3608 @cindex @code{ident} directive
3609 This directive is used by some assemblers to place tags in object files.
3610 @code{@value{AS}} simply accepts the directive for source-file
3611 compatibility with such assemblers, but does not actually emit anything
3615 @section @code{.if @var{absolute expression}}
3617 @cindex conditional assembly
3618 @cindex @code{if} directive
3619 @code{.if} marks the beginning of a section of code which is only
3620 considered part of the source program being assembled if the argument
3621 (which must be an @var{absolute expression}) is non-zero. The end of
3622 the conditional section of code must be marked by @code{.endif}
3623 (@pxref{Endif,,@code{.endif}}); optionally, you may include code for the
3624 alternative condition, flagged by @code{.else} (@pxref{Else,,@code{.else}}).
3626 The following variants of @code{.if} are also supported:
3628 @cindex @code{ifdef} directive
3629 @item .ifdef @var{symbol}
3630 Assembles the following section of code if the specified @var{symbol}
3633 @cindex @code{ifc} directive
3634 @item .ifc @var{string1},@var{string2}
3635 Assembles the following section of code if the two strings are the same. The
3636 strings may be optionally quoted with single quotes. If they are not quoted,
3637 the first string stops at the first comma, and the second string stops at the
3638 end of the line. Strings which contain whitespace should be quoted. The
3639 string comparison is case sensitive.
3641 @cindex @code{ifeq} directive
3642 @item .ifeq @var{absolute expression}
3643 Assembles the following section of code if the argument is zero.
3645 @cindex @code{ifeqs} directive
3646 @item .ifeqs @var{string1},@var{string2}
3647 Another form of @code{.ifc}. The strings must be quoted using double quotes.
3649 @cindex @code{ifge} directive
3650 @item .ifge @var{absolute expression}
3651 Assembles the following section of code if the argument is greater than or
3654 @cindex @code{ifgt} directive
3655 @item .ifgt @var{absolute expression}
3656 Assembles the following section of code if the argument is greater than zero.
3658 @cindex @code{ifle} directive
3659 @item .ifle @var{absolute expression}
3660 Assembles the following section of code if the argument is less than or equal
3663 @cindex @code{iflt} directive
3664 @item .iflt @var{absolute expression}
3665 Assembles the following section of code if the argument is less than zero.
3667 @cindex @code{ifnc} directive
3668 @item .ifnc @var{string1},@var{string2}.
3669 Like @code{.ifc}, but the sense of the test is reversed: this assembles the
3670 following section of code if the two strings are not the same.
3672 @cindex @code{ifndef} directive
3673 @cindex @code{ifnotdef} directive
3674 @item .ifndef @var{symbol}
3675 @itemx .ifnotdef @var{symbol}
3676 Assembles the following section of code if the specified @var{symbol}
3677 has not been defined. Both spelling variants are equivalent.
3679 @cindex @code{ifne} directive
3680 @item .ifne @var{absolute expression}
3681 Assembles the following section of code if the argument is not equal to zero
3682 (in other words, this is equivalent to @code{.if}).
3684 @cindex @code{ifnes} directive
3685 @item .ifnes @var{string1},@var{string2}
3686 Like @code{.ifeqs}, but the sense of the test is reversed: this assembles the
3687 following section of code if the two strings are not the same.
3691 @section @code{.include "@var{file}"}
3693 @cindex @code{include} directive
3694 @cindex supporting files, including
3695 @cindex files, including
3696 This directive provides a way to include supporting files at specified
3697 points in your source program. The code from @var{file} is assembled as
3698 if it followed the point of the @code{.include}; when the end of the
3699 included file is reached, assembly of the original file continues. You
3700 can control the search paths used with the @samp{-I} command-line option
3701 (@pxref{Invoking,,Command-Line Options}). Quotation marks are required
3705 @section @code{.int @var{expressions}}
3707 @cindex @code{int} directive
3708 @cindex integers, 32-bit
3709 Expect zero or more @var{expressions}, of any section, separated by commas.
3710 For each expression, emit a number that, at run time, is the value of that
3711 expression. The byte order and bit size of the number depends on what kind
3712 of target the assembly is for.
3716 On the H8/500 and most forms of the H8/300, @code{.int} emits 16-bit
3717 integers. On the H8/300H and the Hitachi SH, however, @code{.int} emits
3723 @section @code{.irp @var{symbol},@var{values}}@dots{}
3725 @cindex @code{irp} directive
3726 Evaluate a sequence of statements assigning different values to @var{symbol}.
3727 The sequence of statements starts at the @code{.irp} directive, and is
3728 terminated by an @code{.endr} directive. For each @var{value}, @var{symbol} is
3729 set to @var{value}, and the sequence of statements is assembled. If no
3730 @var{value} is listed, the sequence of statements is assembled once, with
3731 @var{symbol} set to the null string. To refer to @var{symbol} within the
3732 sequence of statements, use @var{\symbol}.
3734 For example, assembling
3742 is equivalent to assembling
3751 @section @code{.irpc @var{symbol},@var{values}}@dots{}
3753 @cindex @code{irpc} directive
3754 Evaluate a sequence of statements assigning different values to @var{symbol}.
3755 The sequence of statements starts at the @code{.irpc} directive, and is
3756 terminated by an @code{.endr} directive. For each character in @var{value},
3757 @var{symbol} is set to the character, and the sequence of statements is
3758 assembled. If no @var{value} is listed, the sequence of statements is
3759 assembled once, with @var{symbol} set to the null string. To refer to
3760 @var{symbol} within the sequence of statements, use @var{\symbol}.
3762 For example, assembling
3770 is equivalent to assembling
3779 @section @code{.lcomm @var{symbol} , @var{length}}
3781 @cindex @code{lcomm} directive
3782 @cindex local common symbols
3783 @cindex symbols, local common
3784 Reserve @var{length} (an absolute expression) bytes for a local common
3785 denoted by @var{symbol}. The section and value of @var{symbol} are
3786 those of the new local common. The addresses are allocated in the bss
3787 section, so that at run-time the bytes start off zeroed. @var{Symbol}
3788 is not declared global (@pxref{Global,,@code{.global}}), so is normally
3789 not visible to @code{@value{LD}}.
3792 Some targets permit a third argument to be used with @code{.lcomm}. This
3793 argument specifies the desired alignment of the symbol in the bss section.
3797 The syntax for @code{.lcomm} differs slightly on the HPPA. The syntax is
3798 @samp{@var{symbol} .lcomm, @var{length}}; @var{symbol} is optional.
3802 @section @code{.lflags}
3804 @cindex @code{lflags} directive (ignored)
3805 @code{@value{AS}} accepts this directive, for compatibility with other
3806 assemblers, but ignores it.
3808 @ifclear no-line-dir
3810 @section @code{.line @var{line-number}}
3812 @cindex @code{line} directive
3816 @section @code{.ln @var{line-number}}
3818 @cindex @code{ln} directive
3820 @cindex logical line number
3822 Change the logical line number. @var{line-number} must be an absolute
3823 expression. The next line has that logical line number. Therefore any other
3824 statements on the current line (after a statement separator character) are
3825 reported as on logical line number @var{line-number} @minus{} 1. One day
3826 @code{@value{AS}} will no longer support this directive: it is recognized only
3827 for compatibility with existing assembler programs.
3831 @emph{Warning:} In the AMD29K configuration of @value{AS}, this command is
3832 not available; use the synonym @code{.ln} in that context.
3837 @ifclear no-line-dir
3838 Even though this is a directive associated with the @code{a.out} or
3839 @code{b.out} object-code formats, @code{@value{AS}} still recognizes it
3840 when producing COFF output, and treats @samp{.line} as though it
3841 were the COFF @samp{.ln} @emph{if} it is found outside a
3842 @code{.def}/@code{.endef} pair.
3844 Inside a @code{.def}, @samp{.line} is, instead, one of the directives
3845 used by compilers to generate auxiliary symbol information for
3850 @section @code{.linkonce [@var{type}]}
3852 @cindex @code{linkonce} directive
3853 @cindex common sections
3854 Mark the current section so that the linker only includes a single copy of it.
3855 This may be used to include the same section in several different object files,
3856 but ensure that the linker will only include it once in the final output file.
3857 The @code{.linkonce} pseudo-op must be used for each instance of the section.
3858 Duplicate sections are detected based on the section name, so it should be
3861 This directive is only supported by a few object file formats; as of this
3862 writing, the only object file format which supports it is the Portable
3863 Executable format used on Windows NT.
3865 The @var{type} argument is optional. If specified, it must be one of the
3866 following strings. For example:
3870 Not all types may be supported on all object file formats.
3874 Silently discard duplicate sections. This is the default.
3877 Warn if there are duplicate sections, but still keep only one copy.
3880 Warn if any of the duplicates have different sizes.
3883 Warn if any of the duplicates do not have exactly the same contents.
3887 @section @code{.ln @var{line-number}}
3889 @cindex @code{ln} directive
3890 @ifclear no-line-dir
3891 @samp{.ln} is a synonym for @samp{.line}.
3894 Tell @code{@value{AS}} to change the logical line number. @var{line-number}
3895 must be an absolute expression. The next line has that logical
3896 line number, so any other statements on the current line (after a
3897 statement separator character @code{;}) are reported as on logical
3898 line number @var{line-number} @minus{} 1.
3901 This directive is accepted, but ignored, when @code{@value{AS}} is
3902 configured for @code{b.out}; its effect is only associated with COFF
3908 @section @code{.mri @var{val}}
3910 @cindex @code{mri} directive
3911 @cindex MRI mode, temporarily
3912 If @var{val} is non-zero, this tells @code{@value{AS}} to enter MRI mode. If
3913 @var{val} is zero, this tells @code{@value{AS}} to exit MRI mode. This change
3914 affects code assembled until the next @code{.mri} directive, or until the end
3915 of the file. @xref{M, MRI mode, MRI mode}.
3918 @section @code{.list}
3920 @cindex @code{list} directive
3921 @cindex listing control, turning on
3922 Control (in conjunction with the @code{.nolist} directive) whether or
3923 not assembly listings are generated. These two directives maintain an
3924 internal counter (which is zero initially). @code{.list} increments the
3925 counter, and @code{.nolist} decrements it. Assembly listings are
3926 generated whenever the counter is greater than zero.
3928 By default, listings are disabled. When you enable them (with the
3929 @samp{-a} command line option; @pxref{Invoking,,Command-Line Options}),
3930 the initial value of the listing counter is one.
3933 @section @code{.long @var{expressions}}
3935 @cindex @code{long} directive
3936 @code{.long} is the same as @samp{.int}, @pxref{Int,,@code{.int}}.
3939 @c no one seems to know what this is for or whether this description is
3940 @c what it really ought to do
3942 @section @code{.lsym @var{symbol}, @var{expression}}
3944 @cindex @code{lsym} directive
3945 @cindex symbol, not referenced in assembly
3946 @code{.lsym} creates a new symbol named @var{symbol}, but does not put it in
3947 the hash table, ensuring it cannot be referenced by name during the
3948 rest of the assembly. This sets the attributes of the symbol to be
3949 the same as the expression value:
3951 @var{other} = @var{descriptor} = 0
3952 @var{type} = @r{(section of @var{expression})}
3953 @var{value} = @var{expression}
3956 The new symbol is not flagged as external.
3960 @section @code{.macro}
3963 The commands @code{.macro} and @code{.endm} allow you to define macros that
3964 generate assembly output. For example, this definition specifies a macro
3965 @code{sum} that puts a sequence of numbers into memory:
3968 .macro sum from=0, to=5
3977 With that definition, @samp{SUM 0,5} is equivalent to this assembly input:
3989 @item .macro @var{macname}
3990 @itemx .macro @var{macname} @var{macargs} @dots{}
3991 @cindex @code{macro} directive
3992 Begin the definition of a macro called @var{macname}. If your macro
3993 definition requires arguments, specify their names after the macro name,
3994 separated by commas or spaces. You can supply a default value for any
3995 macro argument by following the name with @samp{=@var{deflt}}. For
3996 example, these are all valid @code{.macro} statements:
4000 Begin the definition of a macro called @code{comm}, which takes no
4003 @item .macro plus1 p, p1
4004 @itemx .macro plus1 p p1
4005 Either statement begins the definition of a macro called @code{plus1},
4006 which takes two arguments; within the macro definition, write
4007 @samp{\p} or @samp{\p1} to evaluate the arguments.
4009 @item .macro reserve_str p1=0 p2
4010 Begin the definition of a macro called @code{reserve_str}, with two
4011 arguments. The first argument has a default value, but not the second.
4012 After the definition is complete, you can call the macro either as
4013 @samp{reserve_str @var{a},@var{b}} (with @samp{\p1} evaluating to
4014 @var{a} and @samp{\p2} evaluating to @var{b}), or as @samp{reserve_str
4015 ,@var{b}} (with @samp{\p1} evaluating as the default, in this case
4016 @samp{0}, and @samp{\p2} evaluating to @var{b}).
4019 When you call a macro, you can specify the argument values either by
4020 position, or by keyword. For example, @samp{sum 9,17} is equivalent to
4021 @samp{sum to=17, from=9}.
4024 @cindex @code{endm} directive
4025 Mark the end of a macro definition.
4028 @cindex @code{exitm} directive
4029 Exit early from the current macro definition.
4031 @cindex number of macros executed
4032 @cindex macros, count executed
4034 @code{@value{AS}} maintains a counter of how many macros it has
4035 executed in this pseudo-variable; you can copy that number to your
4036 output with @samp{\@@}, but @emph{only within a macro definition}.
4039 @item LOCAL @var{name} [ , @dots{} ]
4040 @emph{Warning: @code{LOCAL} is only available if you select ``alternate
4041 macro syntax'' with @samp{-a} or @samp{--alternate}.} @xref{Alternate,,
4042 Alternate macro syntax}.
4044 Generate a string replacement for each of the @var{name} arguments, and
4045 replace any instances of @var{name} in each macro expansion. The
4046 replacement string is unique in the assembly, and different for each
4047 separate macro expansion. @code{LOCAL} allows you to write macros that
4048 define symbols, without fear of conflict between separate macro expansions.
4053 @section @code{.nolist}
4055 @cindex @code{nolist} directive
4056 @cindex listing control, turning off
4057 Control (in conjunction with the @code{.list} directive) whether or
4058 not assembly listings are generated. These two directives maintain an
4059 internal counter (which is zero initially). @code{.list} increments the
4060 counter, and @code{.nolist} decrements it. Assembly listings are
4061 generated whenever the counter is greater than zero.
4064 @section @code{.octa @var{bignums}}
4066 @c FIXME: double size emitted for "octa" on i960, others? Or warn?
4067 @cindex @code{octa} directive
4068 @cindex integer, 16-byte
4069 @cindex sixteen byte integer
4070 This directive expects zero or more bignums, separated by commas. For each
4071 bignum, it emits a 16-byte integer.
4073 The term ``octa'' comes from contexts in which a ``word'' is two bytes;
4074 hence @emph{octa}-word for 16 bytes.
4077 @section @code{.org @var{new-lc} , @var{fill}}
4079 @cindex @code{org} directive
4080 @cindex location counter, advancing
4081 @cindex advancing location counter
4082 @cindex current address, advancing
4083 Advance the location counter of the current section to
4084 @var{new-lc}. @var{new-lc} is either an absolute expression or an
4085 expression with the same section as the current subsection. That is,
4086 you can't use @code{.org} to cross sections: if @var{new-lc} has the
4087 wrong section, the @code{.org} directive is ignored. To be compatible
4088 with former assemblers, if the section of @var{new-lc} is absolute,
4089 @code{@value{AS}} issues a warning, then pretends the section of @var{new-lc}
4090 is the same as the current subsection.
4092 @code{.org} may only increase the location counter, or leave it
4093 unchanged; you cannot use @code{.org} to move the location counter
4096 @c double negative used below "not undefined" because this is a specific
4097 @c reference to "undefined" (as SEG_UNKNOWN is called in this manual)
4098 @c section. doc@cygnus.com 18feb91
4099 Because @code{@value{AS}} tries to assemble programs in one pass, @var{new-lc}
4100 may not be undefined. If you really detest this restriction we eagerly await
4101 a chance to share your improved assembler.
4103 Beware that the origin is relative to the start of the section, not
4104 to the start of the subsection. This is compatible with other
4105 people's assemblers.
4107 When the location counter (of the current subsection) is advanced, the
4108 intervening bytes are filled with @var{fill} which should be an
4109 absolute expression. If the comma and @var{fill} are omitted,
4110 @var{fill} defaults to zero.
4113 @section @code{.p2align[wl] @var{abs-expr}, @var{abs-expr}, @var{abs-expr}}
4115 @cindex padding the location counter given a power of two
4116 @cindex @code{p2align} directive
4117 Pad the location counter (in the current subsection) to a particular
4118 storage boundary. The first expression (which must be absolute) is the
4119 number of low-order zero bits the location counter must have after
4120 advancement. For example @samp{.p2align 3} advances the location
4121 counter until it a multiple of 8. If the location counter is already a
4122 multiple of 8, no change is needed.
4124 The second expression (also absolute) gives the fill value to be stored in the
4125 padding bytes. It (and the comma) may be omitted. If it is omitted, the
4126 padding bytes are normally zero. However, on some systems, if the section is
4127 marked as containing code and the fill value is omitted, the space is filled
4128 with no-op instructions.
4130 The third expression is also absolute, and is also optional. If it is present,
4131 it is the maximum number of bytes that should be skipped by this alignment
4132 directive. If doing the alignment would require skipping more bytes than the
4133 specified maximum, then the alignment is not done at all. You can omit the
4134 fill value (the second argument) entirely by simply using two commas after the
4135 required alignment; this can be useful if you want the alignment to be filled
4136 with no-op instructions when appropriate.
4138 @cindex @code{p2alignw} directive
4139 @cindex @code{p2alignl} directive
4140 The @code{.p2alignw} and @code{.p2alignl} directives are variants of the
4141 @code{.p2align} directive. The @code{.p2alignw} directive treats the fill
4142 pattern as a two byte word value. The @code{.p2alignl} directives treats the
4143 fill pattern as a four byte longword value. For example, @code{.p2alignw
4144 2,0x368d} will align to a multiple of 4. If it skips two bytes, they will be
4145 filled in with the value 0x368d (the exact placement of the bytes depends upon
4146 the endianness of the processor). If it skips 1 or 3 bytes, the fill value is
4150 @section @code{.print @var{string}}
4152 @cindex @code{print} directive
4153 @code{@value{AS}} will print @var{string} on the standard output during
4154 assembly. You must put @var{string} in double quotes.
4157 @section @code{.psize @var{lines} , @var{columns}}
4159 @cindex @code{psize} directive
4160 @cindex listing control: paper size
4161 @cindex paper size, for listings
4162 Use this directive to declare the number of lines---and, optionally, the
4163 number of columns---to use for each page, when generating listings.
4165 If you do not use @code{.psize}, listings use a default line-count
4166 of 60. You may omit the comma and @var{columns} specification; the
4167 default width is 200 columns.
4169 @code{@value{AS}} generates formfeeds whenever the specified number of
4170 lines is exceeded (or whenever you explicitly request one, using
4173 If you specify @var{lines} as @code{0}, no formfeeds are generated save
4174 those explicitly specified with @code{.eject}.
4177 @section @code{.purgem @var{name}}
4179 @cindex @code{purgem} directive
4180 Undefine the macro @var{name}, so that later uses of the string will not be
4181 expanded. @xref{Macro}.
4184 @section @code{.quad @var{bignums}}
4186 @cindex @code{quad} directive
4187 @code{.quad} expects zero or more bignums, separated by commas. For
4188 each bignum, it emits
4190 an 8-byte integer. If the bignum won't fit in 8 bytes, it prints a
4191 warning message; and just takes the lowest order 8 bytes of the bignum.
4192 @cindex eight-byte integer
4193 @cindex integer, 8-byte
4195 The term ``quad'' comes from contexts in which a ``word'' is two bytes;
4196 hence @emph{quad}-word for 8 bytes.
4199 a 16-byte integer. If the bignum won't fit in 16 bytes, it prints a
4200 warning message; and just takes the lowest order 16 bytes of the bignum.
4201 @cindex sixteen-byte integer
4202 @cindex integer, 16-byte
4206 @section @code{.rept @var{count}}
4208 @cindex @code{rept} directive
4209 Repeat the sequence of lines between the @code{.rept} directive and the next
4210 @code{.endr} directive @var{count} times.
4212 For example, assembling
4220 is equivalent to assembling
4229 @section @code{.sbttl "@var{subheading}"}
4231 @cindex @code{sbttl} directive
4232 @cindex subtitles for listings
4233 @cindex listing control: subtitle
4234 Use @var{subheading} as the title (third line, immediately after the
4235 title line) when generating assembly listings.
4237 This directive affects subsequent pages, as well as the current page if
4238 it appears within ten lines of the top of a page.
4242 @section @code{.scl @var{class}}
4244 @cindex @code{scl} directive
4245 @cindex symbol storage class (COFF)
4246 @cindex COFF symbol storage class
4247 Set the storage-class value for a symbol. This directive may only be
4248 used inside a @code{.def}/@code{.endef} pair. Storage class may flag
4249 whether a symbol is static or external, or it may record further
4250 symbolic debugging information.
4253 The @samp{.scl} directive is primarily associated with COFF output; when
4254 configured to generate @code{b.out} output format, @code{@value{AS}}
4255 accepts this directive but ignores it.
4260 @section @code{.section @var{name}}
4262 @cindex @code{section} directive
4263 @cindex named section
4264 Use the @code{.section} directive to assemble the following code into a section
4267 This directive is only supported for targets that actually support arbitrarily
4268 named sections; on @code{a.out} targets, for example, it is not accepted, even
4269 with a standard @code{a.out} section name.
4272 For COFF targets, the @code{.section} directive is used in one of the following
4275 .section @var{name}[, "@var{flags}"]
4276 .section @var{name}[, @var{subsegment}]
4279 If the optional argument is quoted, it is taken as flags to use for the
4280 section. Each flag is a single character. The following flags are recognized:
4283 bss section (uninitialized data)
4285 section is not loaded
4296 If no flags are specified, the default flags depend upon the section name. If
4297 the section name is not recognized, the default will be for the section to be
4298 loaded and writable.
4300 If the optional argument to the @code{.section} directive is not quoted, it is
4301 taken as a subsegment number (@pxref{Sub-Sections}).
4305 For ELF targets, the @code{.section} directive is used like this:
4307 .section @var{name}[, "@var{flags}"[, @@@var{type}]]
4309 The optional @var{flags} argument is a quoted string which may contain any
4310 combintion of the following characters:
4313 section is allocatable
4317 section is executable
4320 The optional @var{type} argument may contain one of the following constants:
4323 section contains data
4325 section does not contain data (i.e., section only occupies space)
4328 If no flags are specified, the default flags depend upon the section name. If
4329 the section name is not recognized, the default will be for the section to have
4330 none of the above flags: it will not be allocated in memory, nor writable, nor
4331 executable. The section will contain data.
4333 For ELF targets, the assembler supports another type of @code{.section}
4334 directive for compatibility with the Solaris assembler:
4336 .section "@var{name}"[, @var{flags}...]
4338 Note that the section name is quoted. There may be a sequence of comma
4342 section is allocatable
4346 section is executable
4351 @section @code{.set @var{symbol}, @var{expression}}
4353 @cindex @code{set} directive
4354 @cindex symbol value, setting
4355 Set the value of @var{symbol} to @var{expression}. This
4356 changes @var{symbol}'s value and type to conform to
4357 @var{expression}. If @var{symbol} was flagged as external, it remains
4358 flagged (@pxref{Symbol Attributes}).
4360 You may @code{.set} a symbol many times in the same assembly.
4362 If you @code{.set} a global symbol, the value stored in the object
4363 file is the last value stored into it.
4366 The syntax for @code{set} on the HPPA is
4367 @samp{@var{symbol} .set @var{expression}}.
4371 @section @code{.short @var{expressions}}
4373 @cindex @code{short} directive
4375 @code{.short} is normally the same as @samp{.word}.
4376 @xref{Word,,@code{.word}}.
4378 In some configurations, however, @code{.short} and @code{.word} generate
4379 numbers of different lengths; @pxref{Machine Dependencies}.
4383 @code{.short} is the same as @samp{.word}. @xref{Word,,@code{.word}}.
4386 This expects zero or more @var{expressions}, and emits
4387 a 16 bit number for each.
4392 @section @code{.single @var{flonums}}
4394 @cindex @code{single} directive
4395 @cindex floating point numbers (single)
4396 This directive assembles zero or more flonums, separated by commas. It
4397 has the same effect as @code{.float}.
4399 The exact kind of floating point numbers emitted depends on how
4400 @code{@value{AS}} is configured. @xref{Machine Dependencies}.
4404 On the @value{TARGET} family, @code{.single} emits 32-bit floating point
4405 numbers in @sc{ieee} format.
4411 @section @code{.size}
4413 @cindex @code{size} directive
4414 This directive is generated by compilers to include auxiliary debugging
4415 information in the symbol table. It is only permitted inside
4416 @code{.def}/@code{.endef} pairs.
4419 @samp{.size} is only meaningful when generating COFF format output; when
4420 @code{@value{AS}} is generating @code{b.out}, it accepts this directive but
4426 @section @code{.sleb128 @var{expressions}}
4428 @cindex @code{sleb128} directive
4429 @var{sleb128} stands for ``signed little endian base 128.'' This is a
4430 compact, variable length representation of numbers used by the DWARF
4431 symbolic debugging format. @xref{Uleb128,@code{.uleb128}}.
4433 @ifclear no-space-dir
4435 @section @code{.skip @var{size} , @var{fill}}
4437 @cindex @code{skip} directive
4438 @cindex filling memory
4439 This directive emits @var{size} bytes, each of value @var{fill}. Both
4440 @var{size} and @var{fill} are absolute expressions. If the comma and
4441 @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same as
4445 @section @code{.space @var{size} , @var{fill}}
4447 @cindex @code{space} directive
4448 @cindex filling memory
4449 This directive emits @var{size} bytes, each of value @var{fill}. Both
4450 @var{size} and @var{fill} are absolute expressions. If the comma
4451 and @var{fill} are omitted, @var{fill} is assumed to be zero. This is the same
4456 @emph{Warning:} @code{.space} has a completely different meaning for HPPA
4457 targets; use @code{.block} as a substitute. See @cite{HP9000 Series 800
4458 Assembly Language Reference Manual} (HP 92432-90001) for the meaning of the
4459 @code{.space} directive. @xref{HPPA Directives,,HPPA Assembler Directives},
4468 @section @code{.space}
4469 @cindex @code{space} directive
4471 On the AMD 29K, this directive is ignored; it is accepted for
4472 compatibility with other AMD 29K assemblers.
4475 @emph{Warning:} In most versions of the @sc{gnu} assembler, the directive
4476 @code{.space} has the effect of @code{.block} @xref{Machine Dependencies}.
4482 @section @code{.stabd, .stabn, .stabs}
4484 @cindex symbolic debuggers, information for
4485 @cindex @code{stab@var{x}} directives
4486 There are three directives that begin @samp{.stab}.
4487 All emit symbols (@pxref{Symbols}), for use by symbolic debuggers.
4488 The symbols are not entered in the @code{@value{AS}} hash table: they
4489 cannot be referenced elsewhere in the source file.
4490 Up to five fields are required:
4494 This is the symbol's name. It may contain any character except
4495 @samp{\000}, so is more general than ordinary symbol names. Some
4496 debuggers used to code arbitrarily complex structures into symbol names
4500 An absolute expression. The symbol's type is set to the low 8 bits of
4501 this expression. Any bit pattern is permitted, but @code{@value{LD}}
4502 and debuggers choke on silly bit patterns.
4505 An absolute expression. The symbol's ``other'' attribute is set to the
4506 low 8 bits of this expression.
4509 An absolute expression. The symbol's descriptor is set to the low 16
4510 bits of this expression.
4513 An absolute expression which becomes the symbol's value.
4516 If a warning is detected while reading a @code{.stabd}, @code{.stabn},
4517 or @code{.stabs} statement, the symbol has probably already been created;
4518 you get a half-formed symbol in your object file. This is
4519 compatible with earlier assemblers!
4522 @cindex @code{stabd} directive
4523 @item .stabd @var{type} , @var{other} , @var{desc}
4525 The ``name'' of the symbol generated is not even an empty string.
4526 It is a null pointer, for compatibility. Older assemblers used a
4527 null pointer so they didn't waste space in object files with empty
4530 The symbol's value is set to the location counter,
4531 relocatably. When your program is linked, the value of this symbol
4532 is the address of the location counter when the @code{.stabd} was
4535 @cindex @code{stabn} directive
4536 @item .stabn @var{type} , @var{other} , @var{desc} , @var{value}
4537 The name of the symbol is set to the empty string @code{""}.
4539 @cindex @code{stabs} directive
4540 @item .stabs @var{string} , @var{type} , @var{other} , @var{desc} , @var{value}
4541 All five fields are specified.
4547 @section @code{.string} "@var{str}"
4549 @cindex string, copying to object file
4550 @cindex @code{string} directive
4552 Copy the characters in @var{str} to the object file. You may specify more than
4553 one string to copy, separated by commas. Unless otherwise specified for a
4554 particular machine, the assembler marks the end of each string with a 0 byte.
4555 You can use any of the escape sequences described in @ref{Strings,,Strings}.
4558 @section @code{.struct @var{expression}}
4560 @cindex @code{struct} directive
4561 Switch to the absolute section, and set the section offset to @var{expression},
4562 which must be an absolute expression. You might use this as follows:
4571 This would define the symbol @code{field1} to have the value 0, the symbol
4572 @code{field2} to have the value 4, and the symbol @code{field3} to have the
4573 value 8. Assembly would be left in the absolute section, and you would need to
4574 use a @code{.section} directive of some sort to change to some other section
4575 before further assembly.
4579 @section @code{.symver}
4580 @cindex @code{symver} directive
4581 @cindex symbol versioning
4582 @cindex versions of symbols
4583 Use the @code{.symver} directive to bind symbols to specific version nodes
4584 within a source file. This is only supported on ELF platforms, and is
4585 typically used when assembling files to be linked into a shared library.
4586 There are cases where it may make sense to use this in objects to be bound
4587 into an application itself so as to override a versioned symbol from a
4590 For ELF targets, the @code{.symver} directive is used like this:
4592 .symver @var{name}, @var{name2@@nodename}
4594 In this case, the symbol @var{name} must exist and be defined within the file
4595 being assembled. The @code{.versym} directive effectively creates a symbol
4596 alias with the name @var{name2@@nodename}, and in fact the main reason that we
4597 just don't try and create a regular alias is that the @var{@@} character isn't
4598 permitted in symbol names. The @var{name2} part of the name is the actual name
4599 of the symbol by which it will be externally referenced. The name @var{name}
4600 itself is merely a name of convenience that is used so that it is possible to
4601 have definitions for multiple versions of a function within a single source
4602 file, and so that the compiler can unambiguously know which version of a
4603 function is being mentioned. The @var{nodename} portion of the alias should be
4604 the name of a node specified in the version script supplied to the linker when
4605 building a shared library. If you are attempting to override a versioned
4606 symbol from a shared library, then @var{nodename} should correspond to the
4607 nodename of the symbol you are trying to override.
4612 @section @code{.tag @var{structname}}
4614 @cindex COFF structure debugging
4615 @cindex structure debugging, COFF
4616 @cindex @code{tag} directive
4617 This directive is generated by compilers to include auxiliary debugging
4618 information in the symbol table. It is only permitted inside
4619 @code{.def}/@code{.endef} pairs. Tags are used to link structure
4620 definitions in the symbol table with instances of those structures.
4623 @samp{.tag} is only used when generating COFF format output; when
4624 @code{@value{AS}} is generating @code{b.out}, it accepts this directive but
4630 @section @code{.text @var{subsection}}
4632 @cindex @code{text} directive
4633 Tells @code{@value{AS}} to assemble the following statements onto the end of
4634 the text subsection numbered @var{subsection}, which is an absolute
4635 expression. If @var{subsection} is omitted, subsection number zero
4639 @section @code{.title "@var{heading}"}
4641 @cindex @code{title} directive
4642 @cindex listing control: title line
4643 Use @var{heading} as the title (second line, immediately after the
4644 source file name and pagenumber) when generating assembly listings.
4646 This directive affects subsequent pages, as well as the current page if
4647 it appears within ten lines of the top of a page.
4651 @section @code{.type @var{int}}
4653 @cindex COFF symbol type
4654 @cindex symbol type, COFF
4655 @cindex @code{type} directive
4656 This directive, permitted only within @code{.def}/@code{.endef} pairs,
4657 records the integer @var{int} as the type attribute of a symbol table entry.
4660 @samp{.type} is associated only with COFF format output; when
4661 @code{@value{AS}} is configured for @code{b.out} output, it accepts this
4662 directive but ignores it.
4668 @section @code{.val @var{addr}}
4670 @cindex @code{val} directive
4671 @cindex COFF value attribute
4672 @cindex value attribute, COFF
4673 This directive, permitted only within @code{.def}/@code{.endef} pairs,
4674 records the address @var{addr} as the value attribute of a symbol table
4678 @samp{.val} is used only for COFF output; when @code{@value{AS}} is
4679 configured for @code{b.out}, it accepts this directive but ignores it.
4684 @section @code{.uleb128 @var{expressions}}
4686 @cindex @code{uleb128} directive
4687 @var{uleb128} stands for ``unsigned little endian base 128.'' This is a
4688 compact, variable length representation of numbers used by the DWARF
4689 symbolic debugging format. @xref{Sleb128,@code{.sleb128}}.
4692 @section @code{.word @var{expressions}}
4694 @cindex @code{word} directive
4695 This directive expects zero or more @var{expressions}, of any section,
4696 separated by commas.
4699 For each expression, @code{@value{AS}} emits a 32-bit number.
4702 For each expression, @code{@value{AS}} emits a 16-bit number.
4707 The size of the number emitted, and its byte order,
4708 depend on what target computer the assembly is for.
4711 @c on amd29k, i960, sparc the "special treatment to support compilers" doesn't
4712 @c happen---32-bit addressability, period; no long/short jumps.
4713 @ifset DIFF-TBL-KLUGE
4714 @cindex difference tables altered
4715 @cindex altered difference tables
4717 @emph{Warning: Special Treatment to support Compilers}
4721 Machines with a 32-bit address space, but that do less than 32-bit
4722 addressing, require the following special treatment. If the machine of
4723 interest to you does 32-bit addressing (or doesn't require it;
4724 @pxref{Machine Dependencies}), you can ignore this issue.
4727 In order to assemble compiler output into something that works,
4728 @code{@value{AS}} occasionlly does strange things to @samp{.word} directives.
4729 Directives of the form @samp{.word sym1-sym2} are often emitted by
4730 compilers as part of jump tables. Therefore, when @code{@value{AS}} assembles a
4731 directive of the form @samp{.word sym1-sym2}, and the difference between
4732 @code{sym1} and @code{sym2} does not fit in 16 bits, @code{@value{AS}}
4733 creates a @dfn{secondary jump table}, immediately before the next label.
4734 This secondary jump table is preceded by a short-jump to the
4735 first byte after the secondary table. This short-jump prevents the flow
4736 of control from accidentally falling into the new table. Inside the
4737 table is a long-jump to @code{sym2}. The original @samp{.word}
4738 contains @code{sym1} minus the address of the long-jump to
4741 If there were several occurrences of @samp{.word sym1-sym2} before the
4742 secondary jump table, all of them are adjusted. If there was a
4743 @samp{.word sym3-sym4}, that also did not fit in sixteen bits, a
4744 long-jump to @code{sym4} is included in the secondary jump table,
4745 and the @code{.word} directives are adjusted to contain @code{sym3}
4746 minus the address of the long-jump to @code{sym4}; and so on, for as many
4747 entries in the original jump table as necessary.
4750 @emph{This feature may be disabled by compiling @code{@value{AS}} with the
4751 @samp{-DWORKING_DOT_WORD} option.} This feature is likely to confuse
4752 assembly language programmers.
4755 @c end DIFF-TBL-KLUGE
4758 @section Deprecated Directives
4760 @cindex deprecated directives
4761 @cindex obsolescent directives
4762 One day these directives won't work.
4763 They are included for compatibility with older assemblers.
4770 @node Machine Dependencies
4771 @chapter Machine Dependent Features
4773 @cindex machine dependencies
4774 The machine instruction sets are (almost by definition) different on
4775 each machine where @code{@value{AS}} runs. Floating point representations
4776 vary as well, and @code{@value{AS}} often supports a few additional
4777 directives or command-line options for compatibility with other
4778 assemblers on a particular platform. Finally, some versions of
4779 @code{@value{AS}} support special pseudo-instructions for branch
4782 This chapter discusses most of these differences, though it does not
4783 include details on any machine's instruction set. For details on that
4784 subject, see the hardware manufacturer's manual.
4788 * AMD29K-Dependent:: AMD 29K Dependent Features
4791 * ARC-Dependent:: ARC Dependent Features
4794 * ARM-Dependent:: ARM Dependent Features
4797 * D10V-Dependent:: D10V Dependent Features
4800 * D30V-Dependent:: D30V Dependent Features
4803 * H8/300-Dependent:: Hitachi H8/300 Dependent Features
4806 * H8/500-Dependent:: Hitachi H8/500 Dependent Features
4809 * HPPA-Dependent:: HPPA Dependent Features
4812 * i386-Dependent:: Intel 80386 Dependent Features
4815 * i960-Dependent:: Intel 80960 Dependent Features
4818 * M68K-Dependent:: M680x0 Dependent Features
4821 * MIPS-Dependent:: MIPS Dependent Features
4824 * SH-Dependent:: Hitachi SH Dependent Features
4827 * Sparc-Dependent:: SPARC Dependent Features
4830 * V850-Dependent:: V850 Dependent Features
4833 * Z8000-Dependent:: Z8000 Dependent Features
4836 * Vax-Dependent:: VAX Dependent Features
4843 @c The following major nodes are *sections* in the GENERIC version, *chapters*
4844 @c in single-cpu versions. This is mainly achieved by @lowersections. There is a
4845 @c peculiarity: to preserve cross-references, there must be a node called
4846 @c "Machine Dependencies". Hence the conditional nodenames in each
4847 @c major node below. Node defaulting in makeinfo requires adjacency of
4848 @c node and sectioning commands; hence the repetition of @chapter BLAH
4849 @c in both conditional blocks.
4855 @chapter ARC Dependent Features
4858 @node Machine Dependencies
4859 @chapter ARC Dependent Features
4864 * ARC-Opts:: Options
4865 * ARC-Float:: Floating Point
4866 * ARC-Directives:: Sparc Machine Directives
4872 @cindex options for ARC
4874 @cindex architectures, ARC
4875 @cindex ARC architectures
4876 The ARC chip family includes several successive levels (or other
4877 variants) of chip, using the same core instruction set, but including
4878 a few additional instructions at each level.
4880 By default, @code{@value{AS}} assumes the core instruction set (ARC
4881 base). The @code{.cpu} pseudo-op is intended to be used to select
4885 @cindex @code{-mbig-endian} option (ARC)
4886 @cindex @code{-mlittle-endian} option (ARC)
4887 @cindex ARC big-endian output
4888 @cindex ARC little-endian output
4889 @cindex big-endian output, ARC
4890 @cindex little-endian output, ARC
4892 @itemx -mlittle-endian
4893 Any @sc{arc} configuration of @code{@value{AS}} can select big-endian or
4894 little-endian output at run time (unlike most other @sc{gnu} development
4895 tools, which must be configured for one or the other). Use
4896 @samp{-mbig-endian} to select big-endian output, and @samp{-mlittle-endian}
4901 @section Floating Point
4903 @cindex floating point, ARC (@sc{ieee})
4904 @cindex ARC floating point (@sc{ieee})
4905 The ARC cpu family currently does not have hardware floating point
4906 support. Software floating point support is provided by @code{GCC}
4907 and uses @sc{ieee} floating-point numbers.
4909 @node ARC-Directives
4910 @section ARC Machine Directives
4912 @cindex ARC machine directives
4913 @cindex machine directives, ARC
4914 The ARC version of @code{@value{AS}} supports the following additional
4919 @cindex @code{cpu} directive, SPARC
4920 This must be followed by the desired cpu.
4921 The ARC is intended to be customizable, @code{.cpu} is used to
4922 select the desired variant [though currently there are none].
4929 @include c-a29k.texi
4938 @node Machine Dependencies
4939 @chapter Machine Dependent Features
4941 The machine instruction sets are different on each Hitachi chip family,
4942 and there are also some syntax differences among the families. This
4943 chapter describes the specific @code{@value{AS}} features for each
4947 * H8/300-Dependent:: Hitachi H8/300 Dependent Features
4948 * H8/500-Dependent:: Hitachi H8/500 Dependent Features
4949 * SH-Dependent:: Hitachi SH Dependent Features
4956 @include c-d10v.texi
4960 @include c-d30v.texi
4964 @include c-h8300.texi
4968 @include c-h8500.texi
4972 @include c-hppa.texi
4976 @include c-i386.texi
4980 @include c-i960.texi
4985 @include c-m68k.texi
4989 @include c-mips.texi
4993 @include c-ns32k.texi
5001 @include c-sparc.texi
5013 @include c-v850.texi
5017 @c reverse effect of @down at top of generic Machine-Dep chapter
5021 @node Reporting Bugs
5022 @chapter Reporting Bugs
5023 @cindex bugs in assembler
5024 @cindex reporting bugs in assembler
5026 Your bug reports play an essential role in making @code{@value{AS}} reliable.
5028 Reporting a bug may help you by bringing a solution to your problem, or it may
5029 not. But in any case the principal function of a bug report is to help the
5030 entire community by making the next version of @code{@value{AS}} work better.
5031 Bug reports are your contribution to the maintenance of @code{@value{AS}}.
5033 In order for a bug report to serve its purpose, you must include the
5034 information that enables us to fix the bug.
5037 * Bug Criteria:: Have you found a bug?
5038 * Bug Reporting:: How to report bugs
5042 @section Have you found a bug?
5043 @cindex bug criteria
5045 If you are not sure whether you have found a bug, here are some guidelines:
5048 @cindex fatal signal
5049 @cindex assembler crash
5050 @cindex crash of assembler
5052 If the assembler gets a fatal signal, for any input whatever, that is a
5053 @code{@value{AS}} bug. Reliable assemblers never crash.
5055 @cindex error on valid input
5057 If @code{@value{AS}} produces an error message for valid input, that is a bug.
5059 @cindex invalid input
5061 If @code{@value{AS}} does not produce an error message for invalid input, that
5062 is a bug. However, you should note that your idea of ``invalid input'' might
5063 be our idea of ``an extension'' or ``support for traditional practice''.
5066 If you are an experienced user of assemblers, your suggestions for improvement
5067 of @code{@value{AS}} are welcome in any case.
5071 @section How to report bugs
5073 @cindex assembler bugs, reporting
5075 A number of companies and individuals offer support for @sc{gnu} products. If
5076 you obtained @code{@value{AS}} from a support organization, we recommend you
5077 contact that organization first.
5079 You can find contact information for many support companies and
5080 individuals in the file @file{etc/SERVICE} in the @sc{gnu} Emacs
5083 In any event, we also recommend that you send bug reports for @code{@value{AS}}
5084 to @samp{bug-gnu-utils@@gnu.org}.
5086 The fundamental principle of reporting bugs usefully is this:
5087 @strong{report all the facts}. If you are not sure whether to state a
5088 fact or leave it out, state it!
5090 Often people omit facts because they think they know what causes the problem
5091 and assume that some details do not matter. Thus, you might assume that the
5092 name of a symbol you use in an example does not matter. Well, probably it does
5093 not, but one cannot be sure. Perhaps the bug is a stray memory reference which
5094 happens to fetch from the location where that name is stored in memory;
5095 perhaps, if the name were different, the contents of that location would fool
5096 the assembler into doing the right thing despite the bug. Play it safe and
5097 give a specific, complete example. That is the easiest thing for you to do,
5098 and the most helpful.
5100 Keep in mind that the purpose of a bug report is to enable us to fix the bug if
5101 it is new to us. Therefore, always write your bug reports on the assumption
5102 that the bug has not been reported previously.
5104 Sometimes people give a few sketchy facts and ask, ``Does this ring a
5105 bell?'' Those bug reports are useless, and we urge everyone to
5106 @emph{refuse to respond to them} except to chide the sender to report
5109 To enable us to fix the bug, you should include all these things:
5113 The version of @code{@value{AS}}. @code{@value{AS}} announces it if you start
5114 it with the @samp{--version} argument.
5116 Without this, we will not know whether there is any point in looking for
5117 the bug in the current version of @code{@value{AS}}.
5120 Any patches you may have applied to the @code{@value{AS}} source.
5123 The type of machine you are using, and the operating system name and
5127 What compiler (and its version) was used to compile @code{@value{AS}}---e.g.
5131 The command arguments you gave the assembler to assemble your example and
5132 observe the bug. To guarantee you will not omit something important, list them
5133 all. A copy of the Makefile (or the output from make) is sufficient.
5135 If we were to try to guess the arguments, we would probably guess wrong
5136 and then we might not encounter the bug.
5139 A complete input file that will reproduce the bug. If the bug is observed when
5140 the assembler is invoked via a compiler, send the assembler source, not the
5141 high level language source. Most compilers will produce the assembler source
5142 when run with the @samp{-S} option. If you are using @code{@value{GCC}}, use
5143 the options @samp{-v --save-temps}; this will save the assembler source in a
5144 file with an extension of @file{.s}, and also show you exactly how
5145 @code{@value{AS}} is being run.
5148 A description of what behavior you observe that you believe is
5149 incorrect. For example, ``It gets a fatal signal.''
5151 Of course, if the bug is that @code{@value{AS}} gets a fatal signal, then we
5152 will certainly notice it. But if the bug is incorrect output, we might not
5153 notice unless it is glaringly wrong. You might as well not give us a chance to
5156 Even if the problem you experience is a fatal signal, you should still say so
5157 explicitly. Suppose something strange is going on, such as, your copy of
5158 @code{@value{AS}} is out of synch, or you have encountered a bug in the C
5159 library on your system. (This has happened!) Your copy might crash and ours
5160 would not. If you told us to expect a crash, then when ours fails to crash, we
5161 would know that the bug was not happening for us. If you had not told us to
5162 expect a crash, then we would not be able to draw any conclusion from our
5166 If you wish to suggest changes to the @code{@value{AS}} source, send us context
5167 diffs, as generated by @code{diff} with the @samp{-u}, @samp{-c}, or @samp{-p}
5168 option. Always send diffs from the old file to the new file. If you even
5169 discuss something in the @code{@value{AS}} source, refer to it by context, not
5172 The line numbers in our development sources will not match those in your
5173 sources. Your line numbers would convey no useful information to us.
5176 Here are some things that are not necessary:
5180 A description of the envelope of the bug.
5182 Often people who encounter a bug spend a lot of time investigating
5183 which changes to the input file will make the bug go away and which
5184 changes will not affect it.
5186 This is often time consuming and not very useful, because the way we
5187 will find the bug is by running a single example under the debugger
5188 with breakpoints, not by pure deduction from a series of examples.
5189 We recommend that you save your time for something else.
5191 Of course, if you can find a simpler example to report @emph{instead}
5192 of the original one, that is a convenience for us. Errors in the
5193 output will be easier to spot, running under the debugger will take
5194 less time, and so on.
5196 However, simplification is not vital; if you do not want to do this,
5197 report the bug anyway and send us the entire test case you used.
5200 A patch for the bug.
5202 A patch for the bug does help us if it is a good one. But do not omit
5203 the necessary information, such as the test case, on the assumption that
5204 a patch is all we need. We might see problems with your patch and decide
5205 to fix the problem another way, or we might not understand it at all.
5207 Sometimes with a program as complicated as @code{@value{AS}} it is very hard to
5208 construct an example that will make the program follow a certain path through
5209 the code. If you do not send us the example, we will not be able to construct
5210 one, so we will not be able to verify that the bug is fixed.
5212 And if we cannot understand what bug you are trying to fix, or why your
5213 patch should be an improvement, we will not install it. A test case will
5214 help us to understand.
5217 A guess about what the bug is or what it depends on.
5219 Such guesses are usually wrong. Even we cannot guess right about such
5220 things without first using the debugger to find the facts.
5223 @node Acknowledgements
5224 @chapter Acknowledgements
5226 If you have contributed to @code{@value{AS}} and your name isn't listed here,
5227 it is not meant as a slight. We just don't know about it. Send mail to the
5228 maintainer, and we'll correct the situation. Currently
5230 the maintainer is Ken Raeburn (email address @code{raeburn@@cygnus.com}).
5232 Dean Elsner wrote the original @sc{gnu} assembler for the VAX.@footnote{Any
5235 Jay Fenlason maintained GAS for a while, adding support for GDB-specific debug
5236 information and the 68k series machines, most of the preprocessing pass, and
5237 extensive changes in @file{messages.c}, @file{input-file.c}, @file{write.c}.
5239 K. Richard Pixley maintained GAS for a while, adding various enhancements and
5240 many bug fixes, including merging support for several processors, breaking GAS
5241 up to handle multiple object file format back ends (including heavy rewrite,
5242 testing, an integration of the coff and b.out back ends), adding configuration
5243 including heavy testing and verification of cross assemblers and file splits
5244 and renaming, converted GAS to strictly ANSI C including full prototypes, added
5245 support for m680[34]0 and cpu32, did considerable work on i960 including a COFF
5246 port (including considerable amounts of reverse engineering), a SPARC opcode
5247 file rewrite, DECstation, rs6000, and hp300hpux host ports, updated ``know''
5248 assertions and made them work, much other reorganization, cleanup, and lint.
5250 Ken Raeburn wrote the high-level BFD interface code to replace most of the code
5251 in format-specific I/O modules.
5253 The original VMS support was contributed by David L. Kashtan. Eric Youngdale
5254 has done much work with it since.
5256 The Intel 80386 machine description was written by Eliot Dresselhaus.
5258 Minh Tran-Le at IntelliCorp contributed some AIX 386 support.
5260 The Motorola 88k machine description was contributed by Devon Bowen of Buffalo
5261 University and Torbjorn Granlund of the Swedish Institute of Computer Science.
5263 Keith Knowles at the Open Software Foundation wrote the original MIPS back end
5264 (@file{tc-mips.c}, @file{tc-mips.h}), and contributed Rose format support
5265 (which hasn't been merged in yet). Ralph Campbell worked with the MIPS code to
5266 support a.out format.
5268 Support for the Zilog Z8k and Hitachi H8/300 and H8/500 processors (tc-z8k,
5269 tc-h8300, tc-h8500), and IEEE 695 object file format (obj-ieee), was written by
5270 Steve Chamberlain of Cygnus Support. Steve also modified the COFF back end to
5271 use BFD for some low-level operations, for use with the H8/300 and AMD 29k
5274 John Gilmore built the AMD 29000 support, added @code{.include} support, and
5275 simplified the configuration of which versions accept which directives. He
5276 updated the 68k machine description so that Motorola's opcodes always produced
5277 fixed-size instructions (e.g. @code{jsr}), while synthetic instructions
5278 remained shrinkable (@code{jbsr}). John fixed many bugs, including true tested
5279 cross-compilation support, and one bug in relaxation that took a week and
5280 required the proverbial one-bit fix.
5282 Ian Lance Taylor of Cygnus Support merged the Motorola and MIT syntax for the
5283 68k, completed support for some COFF targets (68k, i386 SVR3, and SCO Unix),
5284 added support for MIPS ECOFF and ELF targets, wrote the initial RS/6000 and
5285 PowerPC assembler, and made a few other minor patches.
5287 Steve Chamberlain made @code{@value{AS}} able to generate listings.
5289 Hewlett-Packard contributed support for the HP9000/300.
5291 Jeff Law wrote GAS and BFD support for the native HPPA object format (SOM)
5292 along with a fairly extensive HPPA testsuite (for both SOM and ELF object
5293 formats). This work was supported by both the Center for Software Science at
5294 the University of Utah and Cygnus Support.
5296 Support for ELF format files has been worked on by Mark Eichin of Cygnus
5297 Support (original, incomplete implementation for SPARC), Pete Hoogenboom and
5298 Jeff Law at the University of Utah (HPPA mainly), Michael Meissner of the Open
5299 Software Foundation (i386 mainly), and Ken Raeburn of Cygnus Support (sparc,
5300 and some initial 64-bit support).
5302 Richard Henderson rewrote the Alpha assembler. Klaus Kaempf wrote GAS and BFD
5303 support for openVMS/Alpha.
5305 Several engineers at Cygnus Support have also provided many small bug fixes and
5306 configuration enhancements.
5308 Many others have contributed large or small bugfixes and enhancements. If
5309 you have contributed significant work and are not mentioned on this list, and
5310 want to be, let us know. Some of the history has been lost; we are not
5311 intentionally leaving anyone out.