1 \# --------------------------------------------------------------------------
3 \# Copyright 1996-2013 The NASM Authors - All Rights Reserved
4 \# See the file AUTHORS included with the NASM distribution for
5 \# the specific copyright holders.
7 \# Redistribution and use in source and binary forms, with or without
8 \# modification, are permitted provided that the following
11 \# * Redistributions of source code must retain the above copyright
12 \# notice, this list of conditions and the following disclaimer.
13 \# * Redistributions in binary form must reproduce the above
14 \# copyright notice, this list of conditions and the following
15 \# disclaimer in the documentation and/or other materials provided
16 \# with the distribution.
18 \# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
19 \# CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
20 \# INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
21 \# MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
22 \# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
23 \# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 \# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
25 \# NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
26 \# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27 \# HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
28 \# CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
29 \# OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
30 \# EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
32 \# --------------------------------------------------------------------------
34 \# Source code to NASM documentation
36 \M{category}{Programming}
37 \M{title}{NASM - The Netwide Assembler}
39 \M{author}{The NASM Development Team}
40 \M{copyright_tail}{-- All Rights Reserved}
41 \M{license}{This document is redistributable under the license given in the file "LICENSE" distributed in the NASM archive.}
42 \M{summary}{This file documents NASM, the Netwide Assembler: an assembler targetting the Intel x86 series of processors, with portable source.}
45 \M{infotitle}{The Netwide Assembler for x86}
46 \M{epslogo}{nasmlogo.eps}
53 \IR{-MD} \c{-MD} option
54 \IR{-MF} \c{-MF} option
55 \IR{-MG} \c{-MG} option
56 \IR{-MP} \c{-MP} option
57 \IR{-MQ} \c{-MQ} option
58 \IR{-MT} \c{-MT} option
79 \IR{!=} \c{!=} operator
80 \IR{$, here} \c{$}, Here token
81 \IR{$, prefix} \c{$}, prefix
84 \IR{%%} \c{%%} operator
85 \IR{%+1} \c{%+1} and \c{%-1} syntax
87 \IR{%0} \c{%0} parameter count
89 \IR{&&} \c{&&} operator
91 \IR{..@} \c{..@} symbol prefix
93 \IR{//} \c{//} operator
95 \IR{<<} \c{<<} operator
96 \IR{<=} \c{<=} operator
97 \IR{<>} \c{<>} operator
99 \IR{==} \c{==} operator
100 \IR{>} \c{>} operator
101 \IR{>=} \c{>=} operator
102 \IR{>>} \c{>>} operator
103 \IR{?} \c{?} MASM syntax
104 \IR{^} \c{^} operator
105 \IR{^^} \c{^^} operator
106 \IR{|} \c{|} operator
107 \IR{||} \c{||} operator
108 \IR{~} \c{~} operator
109 \IR{%$} \c{%$} and \c{%$$} prefixes
111 \IR{+ opaddition} \c{+} operator, binary
112 \IR{+ opunary} \c{+} operator, unary
113 \IR{+ modifier} \c{+} modifier
114 \IR{- opsubtraction} \c{-} operator, binary
115 \IR{- opunary} \c{-} operator, unary
116 \IR{! opunary} \c{!} operator, unary
117 \IR{alignment, in bin sections} alignment, in \c{bin} sections
118 \IR{alignment, in elf sections} alignment, in \c{elf} sections
119 \IR{alignment, in win32 sections} alignment, in \c{win32} sections
120 \IR{alignment, of elf common variables} alignment, of \c{elf} common
122 \IR{alignment, in obj sections} alignment, in \c{obj} sections
123 \IR{a.out, bsd version} \c{a.out}, BSD version
124 \IR{a.out, linux version} \c{a.out}, Linux version
125 \IR{autoconf} Autoconf
127 \IR{bitwise and} bitwise AND
128 \IR{bitwise or} bitwise OR
129 \IR{bitwise xor} bitwise XOR
130 \IR{block ifs} block IFs
131 \IR{borland pascal} Borland, Pascal
132 \IR{borland's win32 compilers} Borland, Win32 compilers
133 \IR{braces, after % sign} braces, after \c{%} sign
135 \IR{c calling convention} C calling convention
136 \IR{c symbol names} C symbol names
137 \IA{critical expressions}{critical expression}
138 \IA{command line}{command-line}
139 \IA{case sensitivity}{case sensitive}
140 \IA{case-sensitive}{case sensitive}
141 \IA{case-insensitive}{case sensitive}
142 \IA{character constants}{character constant}
143 \IR{common object file format} Common Object File Format
144 \IR{common variables, alignment in elf} common variables, alignment
146 \IR{common, elf extensions to} \c{COMMON}, \c{elf} extensions to
147 \IR{common, obj extensions to} \c{COMMON}, \c{obj} extensions to
148 \IR{declaring structure} declaring structures
149 \IR{default-wrt mechanism} default-\c{WRT} mechanism
152 \IR{dll symbols, exporting} DLL symbols, exporting
153 \IR{dll symbols, importing} DLL symbols, importing
155 \IR{dos archive} DOS archive
156 \IR{dos source archive} DOS source archive
157 \IA{effective address}{effective addresses}
158 \IA{effective-address}{effective addresses}
160 \IR{elf, 16-bit code and} ELF, 16-bit code and
161 \IR{elf shared libraries} ELF, shared libraries
164 \IR{elfx32} \c{elfx32}
165 \IR{executable and linkable format} Executable and Linkable Format
166 \IR{extern, obj extensions to} \c{EXTERN}, \c{obj} extensions to
167 \IR{extern, rdf extensions to} \c{EXTERN}, \c{rdf} extensions to
168 \IR{floating-point, constants} floating-point, constants
169 \IR{floating-point, packed bcd constants} floating-point, packed BCD constants
171 \IR{freelink} FreeLink
172 \IR{functions, c calling convention} functions, C calling convention
173 \IR{functions, pascal calling convention} functions, Pascal calling
175 \IR{global, aoutb extensions to} \c{GLOBAL}, \c{aoutb} extensions to
176 \IR{global, elf extensions to} \c{GLOBAL}, \c{elf} extensions to
177 \IR{global, rdf extensions to} \c{GLOBAL}, \c{rdf} extensions to
179 \IR{got relocations} \c{GOT} relocations
180 \IR{gotoff relocation} \c{GOTOFF} relocations
181 \IR{gotpc relocation} \c{GOTPC} relocations
182 \IR{intel number formats} Intel number formats
183 \IR{linux, elf} Linux, ELF
184 \IR{linux, a.out} Linux, \c{a.out}
185 \IR{linux, as86} Linux, \c{as86}
186 \IR{logical and} logical AND
187 \IR{logical or} logical OR
188 \IR{logical xor} logical XOR
189 \IR{mach object file format} Mach, object file format
191 \IR{macho32} \c{macho32}
192 \IR{macho64} \c{macho64}
195 \IA{memory reference}{memory references}
197 \IA{misc directory}{misc subdirectory}
198 \IR{misc subdirectory} \c{misc} subdirectory
199 \IR{microsoft omf} Microsoft OMF
200 \IR{mmx registers} MMX registers
201 \IA{modr/m}{modr/m byte}
202 \IR{modr/m byte} ModR/M byte
204 \IR{ms-dos device drivers} MS-DOS device drivers
205 \IR{multipush} \c{multipush} macro
207 \IR{nasm version} NASM version
211 \IR{operating system} operating system
213 \IR{pascal calling convention}Pascal calling convention
214 \IR{passes} passes, assembly
219 \IR{plt} \c{PLT} relocations
220 \IA{pre-defining macros}{pre-define}
221 \IA{preprocessor expressions}{preprocessor, expressions}
222 \IA{preprocessor loops}{preprocessor, loops}
223 \IA{preprocessor variables}{preprocessor, variables}
224 \IA{rdoff subdirectory}{rdoff}
225 \IR{rdoff} \c{rdoff} subdirectory
226 \IR{relocatable dynamic object file format} Relocatable Dynamic
228 \IR{relocations, pic-specific} relocations, PIC-specific
229 \IA{repeating}{repeating code}
230 \IR{section alignment, in elf} section alignment, in \c{elf}
231 \IR{section alignment, in bin} section alignment, in \c{bin}
232 \IR{section alignment, in obj} section alignment, in \c{obj}
233 \IR{section alignment, in win32} section alignment, in \c{win32}
234 \IR{section, elf extensions to} \c{SECTION}, \c{elf} extensions to
235 \IR{section, win32 extensions to} \c{SECTION}, \c{win32} extensions to
236 \IR{segment alignment, in bin} segment alignment, in \c{bin}
237 \IR{segment alignment, in obj} segment alignment, in \c{obj}
238 \IR{segment, obj extensions to} \c{SEGMENT}, \c{elf} extensions to
239 \IR{segment names, borland pascal} segment names, Borland Pascal
240 \IR{shift command} \c{shift} command
242 \IR{sib byte} SIB byte
243 \IR{align, smart} \c{ALIGN}, smart
244 \IA{sectalign}{sectalign}
245 \IR{solaris x86} Solaris x86
246 \IA{standard section names}{standardized section names}
247 \IR{symbols, exporting from dlls} symbols, exporting from DLLs
248 \IR{symbols, importing from dlls} symbols, importing from DLLs
249 \IR{test subdirectory} \c{test} subdirectory
251 \IR{underscore, in c symbols} underscore, in C symbols
257 \IA{sco unix}{unix, sco}
258 \IR{unix, sco} Unix, SCO
259 \IA{unix source archive}{unix, source archive}
260 \IR{unix, source archive} Unix, source archive
261 \IA{unix system v}{unix, system v}
262 \IR{unix, system v} Unix, System V
263 \IR{unixware} UnixWare
265 \IR{version number of nasm} version number of NASM
266 \IR{visual c++} Visual C++
267 \IR{www page} WWW page
271 \IR{windows 95} Windows 95
272 \IR{windows nt} Windows NT
273 \# \IC{program entry point}{entry point, program}
274 \# \IC{program entry point}{start point, program}
275 \# \IC{MS-DOS device drivers}{device drivers, MS-DOS}
276 \# \IC{16-bit mode, versus 32-bit mode}{32-bit mode, versus 16-bit mode}
277 \# \IC{c symbol names}{symbol names, in C}
280 \C{intro} Introduction
282 \H{whatsnasm} What Is NASM?
284 The Netwide Assembler, NASM, is an 80x86 and x86-64 assembler designed
285 for portability and modularity. It supports a range of object file
286 formats, including Linux and \c{*BSD} \c{a.out}, \c{ELF}, \c{COFF},
287 \c{Mach-O}, Microsoft 16-bit \c{OBJ}, \c{Win32} and \c{Win64}. It will
288 also output plain binary files. Its syntax is designed to be simple
289 and easy to understand, similar to Intel's but less complex. It
290 supports all currently known x86 architectural extensions, and has
291 strong support for macros.
294 \S{yaasm} Why Yet Another Assembler?
296 The Netwide Assembler grew out of an idea on \i\c{comp.lang.asm.x86}
297 (or possibly \i\c{alt.lang.asm} - I forget which), which was
298 essentially that there didn't seem to be a good \e{free} x86-series
299 assembler around, and that maybe someone ought to write one.
301 \b \i\c{a86} is good, but not free, and in particular you don't get any
302 32-bit capability until you pay. It's DOS only, too.
304 \b \i\c{gas} is free, and ports over to DOS and Unix, but it's not
305 very good, since it's designed to be a back end to \i\c{gcc}, which
306 always feeds it correct code. So its error checking is minimal. Also,
307 its syntax is horrible, from the point of view of anyone trying to
308 actually \e{write} anything in it. Plus you can't write 16-bit code in
311 \b \i\c{as86} is specific to Minix and Linux, and (my version at least)
312 doesn't seem to have much (or any) documentation.
314 \b \i\c{MASM} isn't very good, and it's (was) expensive, and it runs only under
317 \b \i\c{TASM} is better, but still strives for MASM compatibility,
318 which means millions of directives and tons of red tape. And its syntax
319 is essentially MASM's, with the contradictions and quirks that
320 entails (although it sorts out some of those by means of Ideal mode.)
321 It's expensive too. And it's DOS-only.
323 So here, for your coding pleasure, is NASM. At present it's
324 still in prototype stage - we don't promise that it can outperform
325 any of these assemblers. But please, \e{please} send us bug reports,
326 fixes, helpful information, and anything else you can get your hands
327 on (and thanks to the many people who've done this already! You all
328 know who you are), and we'll improve it out of all recognition.
332 \S{legal} \i{License} Conditions
334 Please see the file \c{LICENSE}, supplied as part of any NASM
335 distribution archive, for the license conditions under which you may
336 use NASM. NASM is now under the so-called 2-clause BSD license, also
337 known as the simplified BSD license.
339 Copyright 1996-2011 the NASM Authors - All rights reserved.
341 Redistribution and use in source and binary forms, with or without
342 modification, are permitted provided that the following conditions are
345 \b Redistributions of source code must retain the above copyright
346 notice, this list of conditions and the following disclaimer.
348 \b Redistributions in binary form must reproduce the above copyright
349 notice, this list of conditions and the following disclaimer in the
350 documentation and/or other materials provided with the distribution.
352 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
353 CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
354 INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
355 MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
356 DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
357 CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
358 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
359 NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
360 LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
361 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
362 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
363 OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
364 EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
367 \H{contact} Contact Information
369 The current version of NASM (since about 0.98.08) is maintained by a
370 team of developers, accessible through the \c{nasm-devel} mailing list
371 (see below for the link).
372 If you want to report a bug, please read \k{bugs} first.
374 NASM has a \i{website} at
375 \W{http://www.nasm.us/}\c{http://www.nasm.us/}. If it's not there,
378 \i{New releases}, \i{release candidates}, and \I{snapshots, daily
379 development}\i{daily development snapshots} of NASM are available from
380 the official web site.
382 Announcements are posted to
383 \W{news:comp.lang.asm.x86}\i\c{comp.lang.asm.x86},
385 \W{http://www.freshmeat.net/}\c{http://www.freshmeat.net/}.
387 If you want information about the current development status, please
388 subscribe to the \i\c{nasm-devel} email list; see link from the
392 \H{install} Installation
394 \S{instdos} \i{Installing} NASM under MS-\i{DOS} or Windows
396 Once you've obtained the appropriate archive for NASM,
397 \i\c{nasm-XXX-dos.zip} or \i\c{nasm-XXX-win32.zip} (where \c{XXX}
398 denotes the version number of NASM contained in the archive), unpack
399 it into its own directory (for example \c{c:\\nasm}).
401 The archive will contain a set of executable files: the NASM
402 executable file \i\c{nasm.exe}, the NDISASM executable file
403 \i\c{ndisasm.exe}, and possibly additional utilities to handle the
406 The only file NASM needs to run is its own executable, so copy
407 \c{nasm.exe} to a directory on your PATH, or alternatively edit
408 \i\c{autoexec.bat} to add the \c{nasm} directory to your
409 \i\c{PATH} (to do that under Windows XP, go to Start > Control Panel >
410 System > Advanced > Environment Variables; these instructions may work
411 under other versions of Windows as well.)
413 That's it - NASM is installed. You don't need the nasm directory
414 to be present to run NASM (unless you've added it to your \c{PATH}),
415 so you can delete it if you need to save space; however, you may
416 want to keep the documentation or test programs.
418 If you've downloaded the \i{DOS source archive}, \i\c{nasm-XXX.zip},
419 the \c{nasm} directory will also contain the full NASM \i{source
420 code}, and a selection of \i{Makefiles} you can (hopefully) use to
421 rebuild your copy of NASM from scratch. See the file \c{INSTALL} in
424 Note that a number of files are generated from other files by Perl
425 scripts. Although the NASM source distribution includes these
426 generated files, you will need to rebuild them (and hence, will need a
427 Perl interpreter) if you change insns.dat, standard.mac or the
428 documentation. It is possible future source distributions may not
429 include these files at all. Ports of \i{Perl} for a variety of
430 platforms, including DOS and Windows, are available from
431 \W{http://www.cpan.org/ports/}\i{www.cpan.org}.
434 \S{instdos} Installing NASM under \i{Unix}
436 Once you've obtained the \i{Unix source archive} for NASM,
437 \i\c{nasm-XXX.tar.gz} (where \c{XXX} denotes the version number of
438 NASM contained in the archive), unpack it into a directory such
439 as \c{/usr/local/src}. The archive, when unpacked, will create its
440 own subdirectory \c{nasm-XXX}.
442 NASM is an \I{Autoconf}\I\c{configure}auto-configuring package: once
443 you've unpacked it, \c{cd} to the directory it's been unpacked into
444 and type \c{./configure}. This shell script will find the best C
445 compiler to use for building NASM and set up \i{Makefiles}
448 Once NASM has auto-configured, you can type \i\c{make} to build the
449 \c{nasm} and \c{ndisasm} binaries, and then \c{make install} to
450 install them in \c{/usr/local/bin} and install the \i{man pages}
451 \i\c{nasm.1} and \i\c{ndisasm.1} in \c{/usr/local/man/man1}.
452 Alternatively, you can give options such as \c{--prefix} to the
453 configure script (see the file \i\c{INSTALL} for more details), or
454 install the programs yourself.
456 NASM also comes with a set of utilities for handling the \c{RDOFF}
457 custom object-file format, which are in the \i\c{rdoff} subdirectory
458 of the NASM archive. You can build these with \c{make rdf} and
459 install them with \c{make rdf_install}, if you want them.
462 \C{running} Running NASM
464 \H{syntax} NASM \i{Command-Line} Syntax
466 To assemble a file, you issue a command of the form
468 \c nasm -f <format> <filename> [-o <output>]
472 \c nasm -f elf myfile.asm
474 will assemble \c{myfile.asm} into an \c{ELF} object file \c{myfile.o}. And
476 \c nasm -f bin myfile.asm -o myfile.com
478 will assemble \c{myfile.asm} into a raw binary file \c{myfile.com}.
480 To produce a listing file, with the hex codes output from NASM
481 displayed on the left of the original sources, use the \c{-l} option
482 to give a listing file name, for example:
484 \c nasm -f coff myfile.asm -l myfile.lst
486 To get further usage instructions from NASM, try typing
490 As \c{-hf}, this will also list the available output file formats, and what they
493 If you use Linux but aren't sure whether your system is \c{a.out}
498 (in the directory in which you put the NASM binary when you
499 installed it). If it says something like
501 \c nasm: ELF 32-bit LSB executable i386 (386 and up) Version 1
503 then your system is \c{ELF}, and you should use the option \c{-f elf}
504 when you want NASM to produce Linux object files. If it says
506 \c nasm: Linux/i386 demand-paged executable (QMAGIC)
508 or something similar, your system is \c{a.out}, and you should use
509 \c{-f aout} instead (Linux \c{a.out} systems have long been obsolete,
510 and are rare these days.)
512 Like Unix compilers and assemblers, NASM is silent unless it
513 goes wrong: you won't see any output at all, unless it gives error
517 \S{opt-o} The \i\c{-o} Option: Specifying the Output File Name
519 NASM will normally choose the name of your output file for you;
520 precisely how it does this is dependent on the object file format.
521 For Microsoft object file formats (\c{obj}, \c{win32} and \c{win64}),
522 it will remove the \c{.asm} \i{extension} (or whatever extension you
523 like to use - NASM doesn't care) from your source file name and
524 substitute \c{.obj}. For Unix object file formats (\c{aout}, \c{as86},
525 \c{coff}, \c{elf32}, \c{elf64}, \c{elfx32}, \c{ieee}, \c{macho32} and
526 \c{macho64}) it will substitute \c{.o}. For \c{dbg}, \c{rdf}, \c{ith}
527 and \c{srec}, it will use \c{.dbg}, \c{.rdf}, \c{.ith} and \c{.srec},
528 respectively, and for the \c{bin} format it will simply remove the
529 extension, so that \c{myfile.asm} produces the output file \c{myfile}.
531 If the output file already exists, NASM will overwrite it, unless it
532 has the same name as the input file, in which case it will give a
533 warning and use \i\c{nasm.out} as the output file name instead.
535 For situations in which this behaviour is unacceptable, NASM
536 provides the \c{-o} command-line option, which allows you to specify
537 your desired output file name. You invoke \c{-o} by following it
538 with the name you wish for the output file, either with or without
539 an intervening space. For example:
541 \c nasm -f bin program.asm -o program.com
542 \c nasm -f bin driver.asm -odriver.sys
544 Note that this is a small o, and is different from a capital O , which
545 is used to specify the number of optimisation passes required. See \k{opt-O}.
548 \S{opt-f} The \i\c{-f} Option: Specifying the \i{Output File Format}
550 If you do not supply the \c{-f} option to NASM, it will choose an
551 output file format for you itself. In the distribution versions of
552 NASM, the default is always \i\c{bin}; if you've compiled your own
553 copy of NASM, you can redefine \i\c{OF_DEFAULT} at compile time and
554 choose what you want the default to be.
556 Like \c{-o}, the intervening space between \c{-f} and the output
557 file format is optional; so \c{-f elf} and \c{-felf} are both valid.
559 A complete list of the available output file formats can be given by
560 issuing the command \i\c{nasm -hf}.
563 \S{opt-l} The \i\c{-l} Option: Generating a \i{Listing File}
565 If you supply the \c{-l} option to NASM, followed (with the usual
566 optional space) by a file name, NASM will generate a
567 \i{source-listing file} for you, in which addresses and generated
568 code are listed on the left, and the actual source code, with
569 expansions of multi-line macros (except those which specifically
570 request no expansion in source listings: see \k{nolist}) on the
573 \c nasm -f elf myfile.asm -l myfile.lst
575 If a list file is selected, you may turn off listing for a
576 section of your source with \c{[list -]}, and turn it back on
577 with \c{[list +]}, (the default, obviously). There is no "user
578 form" (without the brackets). This can be used to list only
579 sections of interest, avoiding excessively long listings.
582 \S{opt-M} The \i\c{-M} Option: Generate \i{Makefile Dependencies}
584 This option can be used to generate makefile dependencies on stdout.
585 This can be redirected to a file for further processing. For example:
587 \c nasm -M myfile.asm > myfile.dep
590 \S{opt-MG} The \i\c{-MG} Option: Generate \i{Makefile Dependencies}
592 This option can be used to generate makefile dependencies on stdout.
593 This differs from the \c{-M} option in that if a nonexisting file is
594 encountered, it is assumed to be a generated file and is added to the
595 dependency list without a prefix.
598 \S{opt-MF} The \i\c\{-MF} Option: Set Makefile Dependency File
600 This option can be used with the \c{-M} or \c{-MG} options to send the
601 output to a file, rather than to stdout. For example:
603 \c nasm -M -MF myfile.dep myfile.asm
606 \S{opt-MD} The \i\c{-MD} Option: Assemble and Generate Dependencies
608 The \c{-MD} option acts as the combination of the \c{-M} and \c{-MF}
609 options (i.e. a filename has to be specified.) However, unlike the
610 \c{-M} or \c{-MG} options, \c{-MD} does \e{not} inhibit the normal
611 operation of the assembler. Use this to automatically generate
612 updated dependencies with every assembly session. For example:
614 \c nasm -f elf -o myfile.o -MD myfile.dep myfile.asm
617 \S{opt-MT} The \i\c{-MT} Option: Dependency Target Name
619 The \c{-MT} option can be used to override the default name of the
620 dependency target. This is normally the same as the output filename,
621 specified by the \c{-o} option.
624 \S{opt-MQ} The \i\c{-MQ} Option: Dependency Target Name (Quoted)
626 The \c{-MQ} option acts as the \c{-MT} option, except it tries to
627 quote characters that have special meaning in Makefile syntax. This
628 is not foolproof, as not all characters with special meaning are
629 quotable in Make. The default output (if no \c{-MT} or \c{-MQ} option
630 is specified) is automatically quoted.
633 \S{opt-MP} The \i\c{-MP} Option: Emit phony targets
635 When used with any of the dependency generation options, the \c{-MP}
636 option causes NASM to emit a phony target without dependencies for
637 each header file. This prevents Make from complaining if a header
638 file has been removed.
641 \S{opt-F} The \i\c{-F} Option: Selecting a \i{Debug Information Format}
643 This option is used to select the format of the debug information
644 emitted into the output file, to be used by a debugger (or \e{will}
645 be). Prior to version 2.03.01, the use of this switch did \e{not} enable
646 output of the selected debug info format. Use \c{-g}, see \k{opt-g},
647 to enable output. Versions 2.03.01 and later automatically enable \c{-g}
648 if \c{-F} is specified.
650 A complete list of the available debug file formats for an output
651 format can be seen by issuing the command \c{nasm -f <format> -y}. Not
652 all output formats currently support debugging output. See \k{opt-y}.
654 This should not be confused with the \c{-f dbg} output format option which
655 is not built into NASM by default. For information on how
656 to enable it when building from the sources, see \k{dbgfmt}.
659 \S{opt-g} The \i\c{-g} Option: Enabling \i{Debug Information}.
661 This option can be used to generate debugging information in the specified
662 format. See \k{opt-F}. Using \c{-g} without \c{-F} results in emitting
663 debug info in the default format, if any, for the selected output format.
664 If no debug information is currently implemented in the selected output
665 format, \c{-g} is \e{silently ignored}.
668 \S{opt-X} The \i\c{-X} Option: Selecting an \i{Error Reporting Format}
670 This option can be used to select an error reporting format for any
671 error messages that might be produced by NASM.
673 Currently, two error reporting formats may be selected. They are
674 the \c{-Xvc} option and the \c{-Xgnu} option. The GNU format is
675 the default and looks like this:
677 \c filename.asm:65: error: specific error message
679 where \c{filename.asm} is the name of the source file in which the
680 error was detected, \c{65} is the source file line number on which
681 the error was detected, \c{error} is the severity of the error (this
682 could be \c{warning}), and \c{specific error message} is a more
683 detailed text message which should help pinpoint the exact problem.
685 The other format, specified by \c{-Xvc} is the style used by Microsoft
686 Visual C++ and some other programs. It looks like this:
688 \c filename.asm(65) : error: specific error message
690 where the only difference is that the line number is in parentheses
691 instead of being delimited by colons.
693 See also the \c{Visual C++} output format, \k{win32fmt}.
695 \S{opt-Z} The \i\c{-Z} Option: Send Errors to a File
697 Under \I{DOS}\c{MS-DOS} it can be difficult (though there are ways) to
698 redirect the standard-error output of a program to a file. Since
699 NASM usually produces its warning and \i{error messages} on
700 \i\c{stderr}, this can make it hard to capture the errors if (for
701 example) you want to load them into an editor.
703 NASM therefore provides the \c{-Z} option, taking a filename argument
704 which causes errors to be sent to the specified files rather than
705 standard error. Therefore you can \I{redirecting errors}redirect
706 the errors into a file by typing
708 \c nasm -Z myfile.err -f obj myfile.asm
710 In earlier versions of NASM, this option was called \c{-E}, but it was
711 changed since \c{-E} is an option conventionally used for
712 preprocessing only, with disastrous results. See \k{opt-E}.
714 \S{opt-s} The \i\c{-s} Option: Send Errors to \i\c{stdout}
716 The \c{-s} option redirects \i{error messages} to \c{stdout} rather
717 than \c{stderr}, so it can be redirected under \I{DOS}\c{MS-DOS}. To
718 assemble the file \c{myfile.asm} and pipe its output to the \c{more}
719 program, you can type:
721 \c nasm -s -f obj myfile.asm | more
723 See also the \c{-Z} option, \k{opt-Z}.
726 \S{opt-i} The \i\c{-i}\I\c{-I} Option: Include File Search Directories
728 When NASM sees the \i\c{%include} or \i\c{%pathsearch} directive in a
729 source file (see \k{include}, \k{pathsearch} or \k{incbin}), it will
730 search for the given file not only in the current directory, but also
731 in any directories specified on the command line by the use of the
732 \c{-i} option. Therefore you can include files from a \i{macro
733 library}, for example, by typing
735 \c nasm -ic:\macrolib\ -f obj myfile.asm
737 (As usual, a space between \c{-i} and the path name is allowed, and
740 NASM, in the interests of complete source-code portability, does not
741 understand the file naming conventions of the OS it is running on;
742 the string you provide as an argument to the \c{-i} option will be
743 prepended exactly as written to the name of the include file.
744 Therefore the trailing backslash in the above example is necessary.
745 Under Unix, a trailing forward slash is similarly necessary.
747 (You can use this to your advantage, if you're really \i{perverse},
748 by noting that the option \c{-ifoo} will cause \c{%include "bar.i"}
749 to search for the file \c{foobar.i}...)
751 If you want to define a \e{standard} \i{include search path},
752 similar to \c{/usr/include} on Unix systems, you should place one or
753 more \c{-i} directives in the \c{NASMENV} environment variable (see
756 For Makefile compatibility with many C compilers, this option can also
757 be specified as \c{-I}.
760 \S{opt-p} The \i\c{-p}\I\c{-P} Option: \I{pre-including files}Pre-Include a File
762 \I\c{%include}NASM allows you to specify files to be
763 \e{pre-included} into your source file, by the use of the \c{-p}
766 \c nasm myfile.asm -p myinc.inc
768 is equivalent to running \c{nasm myfile.asm} and placing the
769 directive \c{%include "myinc.inc"} at the start of the file.
771 For consistency with the \c{-I}, \c{-D} and \c{-U} options, this
772 option can also be specified as \c{-P}.
775 \S{opt-d} The \i\c{-d}\I\c{-D} Option: \I{pre-defining macros}Pre-Define a Macro
777 \I\c{%define}Just as the \c{-p} option gives an alternative to placing
778 \c{%include} directives at the start of a source file, the \c{-d}
779 option gives an alternative to placing a \c{%define} directive. You
782 \c nasm myfile.asm -dFOO=100
784 as an alternative to placing the directive
788 at the start of the file. You can miss off the macro value, as well:
789 the option \c{-dFOO} is equivalent to coding \c{%define FOO}. This
790 form of the directive may be useful for selecting \i{assembly-time
791 options} which are then tested using \c{%ifdef}, for example
794 For Makefile compatibility with many C compilers, this option can also
795 be specified as \c{-D}.
798 \S{opt-u} The \i\c{-u}\I\c{-U} Option: \I{Undefining macros}Undefine a Macro
800 \I\c{%undef}The \c{-u} option undefines a macro that would otherwise
801 have been pre-defined, either automatically or by a \c{-p} or \c{-d}
802 option specified earlier on the command lines.
804 For example, the following command line:
806 \c nasm myfile.asm -dFOO=100 -uFOO
808 would result in \c{FOO} \e{not} being a predefined macro in the
809 program. This is useful to override options specified at a different
812 For Makefile compatibility with many C compilers, this option can also
813 be specified as \c{-U}.
816 \S{opt-E} The \i\c{-E}\I{-e} Option: Preprocess Only
818 NASM allows the \i{preprocessor} to be run on its own, up to a
819 point. Using the \c{-E} option (which requires no arguments) will
820 cause NASM to preprocess its input file, expand all the macro
821 references, remove all the comments and preprocessor directives, and
822 print the resulting file on standard output (or save it to a file,
823 if the \c{-o} option is also used).
825 This option cannot be applied to programs which require the
826 preprocessor to evaluate \I{preprocessor expressions}\i{expressions}
827 which depend on the values of symbols: so code such as
829 \c %assign tablesize ($-tablestart)
831 will cause an error in \i{preprocess-only mode}.
833 For compatiblity with older version of NASM, this option can also be
834 written \c{-e}. \c{-E} in older versions of NASM was the equivalent
835 of the current \c{-Z} option, \k{opt-Z}.
837 \S{opt-a} The \i\c{-a} Option: Don't Preprocess At All
839 If NASM is being used as the back end to a compiler, it might be
840 desirable to \I{suppressing preprocessing}suppress preprocessing
841 completely and assume the compiler has already done it, to save time
842 and increase compilation speeds. The \c{-a} option, requiring no
843 argument, instructs NASM to replace its powerful \i{preprocessor}
844 with a \i{stub preprocessor} which does nothing.
847 \S{opt-O} The \i\c{-O} Option: Specifying \i{Multipass Optimization}
849 Using the \c{-O} option, you can tell NASM to carry out different
850 levels of optimization. The syntax is:
852 \b \c{-O0}: No optimization. All operands take their long forms,
853 if a short form is not specified, except conditional jumps.
854 This is intended to match NASM 0.98 behavior.
856 \b \c{-O1}: Minimal optimization. As above, but immediate operands
857 which will fit in a signed byte are optimized,
858 unless the long form is specified. Conditional jumps default
859 to the long form unless otherwise specified.
861 \b \c{-Ox} (where \c{x} is the actual letter \c{x}): Multipass optimization.
862 Minimize branch offsets and signed immediate bytes,
863 overriding size specification unless the \c{strict} keyword
864 has been used (see \k{strict}). For compatibility with earlier
865 releases, the letter \c{x} may also be any number greater than
866 one. This number has no effect on the actual number of passes.
868 The \c{-Ox} mode is recommended for most uses, and is the default
871 Note that this is a capital \c{O}, and is different from a small \c{o}, which
872 is used to specify the output file name. See \k{opt-o}.
875 \S{opt-t} The \i\c{-t} Option: Enable TASM Compatibility Mode
877 NASM includes a limited form of compatibility with Borland's \i\c{TASM}.
878 When NASM's \c{-t} option is used, the following changes are made:
880 \b local labels may be prefixed with \c{@@} instead of \c{.}
882 \b size override is supported within brackets. In TASM compatible mode,
883 a size override inside square brackets changes the size of the operand,
884 and not the address type of the operand as it does in NASM syntax. E.g.
885 \c{mov eax,[DWORD val]} is valid syntax in TASM compatibility mode.
886 Note that you lose the ability to override the default address type for
889 \b unprefixed forms of some directives supported (\c{arg}, \c{elif},
890 \c{else}, \c{endif}, \c{if}, \c{ifdef}, \c{ifdifi}, \c{ifndef},
891 \c{include}, \c{local})
893 \S{opt-w} The \i\c{-w} and \i\c{-W} Options: Enable or Disable Assembly \i{Warnings}
895 NASM can observe many conditions during the course of assembly which
896 are worth mentioning to the user, but not a sufficiently severe
897 error to justify NASM refusing to generate an output file. These
898 conditions are reported like errors, but come up with the word
899 `warning' before the message. Warnings do not prevent NASM from
900 generating an output file and returning a success status to the
903 Some conditions are even less severe than that: they are only
904 sometimes worth mentioning to the user. Therefore NASM supports the
905 \c{-w} command-line option, which enables or disables certain
906 classes of assembly warning. Such warning classes are described by a
907 name, for example \c{orphan-labels}; you can enable warnings of
908 this class by the command-line option \c{-w+orphan-labels} and
909 disable it by \c{-w-orphan-labels}.
911 The \i{suppressible warning} classes are:
913 \b \i\c{macro-params} covers warnings about \i{multi-line macros}
914 being invoked with the wrong number of parameters. This warning
915 class is enabled by default; see \k{mlmacover} for an example of why
916 you might want to disable it.
918 \b \i\c{macro-selfref} warns if a macro references itself. This
919 warning class is disabled by default.
921 \b\i\c{macro-defaults} warns when a macro has more default
922 parameters than optional parameters. This warning class
923 is enabled by default; see \k{mlmacdef} for why you might want to disable it.
925 \b \i\c{orphan-labels} covers warnings about source lines which
926 contain no instruction but define a label without a trailing colon.
927 NASM warns about this somewhat obscure condition by default;
928 see \k{syntax} for more information.
930 \b \i\c{number-overflow} covers warnings about numeric constants which
931 don't fit in 64 bits. This warning class is enabled by default.
933 \b \i\c{gnu-elf-extensions} warns if 8-bit or 16-bit relocations
934 are used in \c{-f elf} format. The GNU extensions allow this.
935 This warning class is disabled by default.
937 \b \i\c{float-overflow} warns about floating point overflow.
940 \b \i\c{float-denorm} warns about floating point denormals.
943 \b \i\c{float-underflow} warns about floating point underflow.
946 \b \i\c{float-toolong} warns about too many digits in floating-point numbers.
949 \b \i\c{user} controls \c{%warning} directives (see \k{pperror}).
952 \b \i\c{lock} warns about \c{LOCK} prefixes on unlockable instructions.
955 \b \i\c{hle} warns about invalid use of the HLE \c{XACQUIRE} or \c{XRELEASE}
959 \b \i\c{error} causes warnings to be treated as errors. Disabled by
962 \b \i\c{all} is an alias for \e{all} suppressible warning classes (not
963 including \c{error}). Thus, \c{-w+all} enables all available warnings.
965 In addition, you can set warning classes across sections.
966 Warning classes may be enabled with \i\c{[warning +warning-name]},
967 disabled with \i\c{[warning -warning-name]} or reset to their
968 original value with \i\c{[warning *warning-name]}. No "user form"
969 (without the brackets) exists.
971 Since version 2.00, NASM has also supported the gcc-like syntax
972 \c{-Wwarning} and \c{-Wno-warning} instead of \c{-w+warning} and
973 \c{-w-warning}, respectively.
976 \S{opt-v} The \i\c{-v} Option: Display \i{Version} Info
978 Typing \c{NASM -v} will display the version of NASM which you are using,
979 and the date on which it was compiled.
981 You will need the version number if you report a bug.
983 \S{opt-y} The \i\c{-y} Option: Display Available Debug Info Formats
985 Typing \c{nasm -f <option> -y} will display a list of the available
986 debug info formats for the given output format. The default format
987 is indicated by an asterisk. For example:
991 \c valid debug formats for 'elf32' output format are
992 \c ('*' denotes default):
993 \c * stabs ELF32 (i386) stabs debug format for Linux
994 \c dwarf elf32 (i386) dwarf debug format for Linux
997 \S{opt-pfix} The \i\c{--prefix} and \i\c{--postfix} Options.
999 The \c{--prefix} and \c{--postfix} options prepend or append
1000 (respectively) the given argument to all \c{global} or
1001 \c{extern} variables. E.g. \c{--prefix _} will prepend the
1002 underscore to all global and external variables, as C sometimes
1003 (but not always) likes it.
1006 \S{nasmenv} The \i\c{NASMENV} \i{Environment} Variable
1008 If you define an environment variable called \c{NASMENV}, the program
1009 will interpret it as a list of extra command-line options, which are
1010 processed before the real command line. You can use this to define
1011 standard search directories for include files, by putting \c{-i}
1012 options in the \c{NASMENV} variable.
1014 The value of the variable is split up at white space, so that the
1015 value \c{-s -ic:\\nasmlib\\} will be treated as two separate options.
1016 However, that means that the value \c{-dNAME="my name"} won't do
1017 what you might want, because it will be split at the space and the
1018 NASM command-line processing will get confused by the two
1019 nonsensical words \c{-dNAME="my} and \c{name"}.
1021 To get round this, NASM provides a feature whereby, if you begin the
1022 \c{NASMENV} environment variable with some character that isn't a minus
1023 sign, then NASM will treat this character as the \i{separator
1024 character} for options. So setting the \c{NASMENV} variable to the
1025 value \c{!-s!-ic:\\nasmlib\\} is equivalent to setting it to \c{-s
1026 -ic:\\nasmlib\\}, but \c{!-dNAME="my name"} will work.
1028 This environment variable was previously called \c{NASM}. This was
1029 changed with version 0.98.31.
1032 \H{qstart} \i{Quick Start} for \i{MASM} Users
1034 If you're used to writing programs with MASM, or with \i{TASM} in
1035 MASM-compatible (non-Ideal) mode, or with \i\c{a86}, this section
1036 attempts to outline the major differences between MASM's syntax and
1037 NASM's. If you're not already used to MASM, it's probably worth
1038 skipping this section.
1041 \S{qscs} NASM Is \I{case sensitivity}Case-Sensitive
1043 One simple difference is that NASM is case-sensitive. It makes a
1044 difference whether you call your label \c{foo}, \c{Foo} or \c{FOO}.
1045 If you're assembling to \c{DOS} or \c{OS/2} \c{.OBJ} files, you can
1046 invoke the \i\c{UPPERCASE} directive (documented in \k{objfmt}) to
1047 ensure that all symbols exported to other code modules are forced
1048 to be upper case; but even then, \e{within} a single module, NASM
1049 will distinguish between labels differing only in case.
1052 \S{qsbrackets} NASM Requires \i{Square Brackets} For \i{Memory References}
1054 NASM was designed with simplicity of syntax in mind. One of the
1055 \i{design goals} of NASM is that it should be possible, as far as is
1056 practical, for the user to look at a single line of NASM code
1057 and tell what opcode is generated by it. You can't do this in MASM:
1058 if you declare, for example,
1063 then the two lines of code
1068 generate completely different opcodes, despite having
1069 identical-looking syntaxes.
1071 NASM avoids this undesirable situation by having a much simpler
1072 syntax for memory references. The rule is simply that any access to
1073 the \e{contents} of a memory location requires square brackets
1074 around the address, and any access to the \e{address} of a variable
1075 doesn't. So an instruction of the form \c{mov ax,foo} will
1076 \e{always} refer to a compile-time constant, whether it's an \c{EQU}
1077 or the address of a variable; and to access the \e{contents} of the
1078 variable \c{bar}, you must code \c{mov ax,[bar]}.
1080 This also means that NASM has no need for MASM's \i\c{OFFSET}
1081 keyword, since the MASM code \c{mov ax,offset bar} means exactly the
1082 same thing as NASM's \c{mov ax,bar}. If you're trying to get
1083 large amounts of MASM code to assemble sensibly under NASM, you
1084 can always code \c{%idefine offset} to make the preprocessor treat
1085 the \c{OFFSET} keyword as a no-op.
1087 This issue is even more confusing in \i\c{a86}, where declaring a
1088 label with a trailing colon defines it to be a `label' as opposed to
1089 a `variable' and causes \c{a86} to adopt NASM-style semantics; so in
1090 \c{a86}, \c{mov ax,var} has different behaviour depending on whether
1091 \c{var} was declared as \c{var: dw 0} (a label) or \c{var dw 0} (a
1092 word-size variable). NASM is very simple by comparison:
1093 \e{everything} is a label.
1095 NASM, in the interests of simplicity, also does not support the
1096 \i{hybrid syntaxes} supported by MASM and its clones, such as
1097 \c{mov ax,table[bx]}, where a memory reference is denoted by one
1098 portion outside square brackets and another portion inside. The
1099 correct syntax for the above is \c{mov ax,[table+bx]}. Likewise,
1100 \c{mov ax,es:[di]} is wrong and \c{mov ax,[es:di]} is right.
1103 \S{qstypes} NASM Doesn't Store \i{Variable Types}
1105 NASM, by design, chooses not to remember the types of variables you
1106 declare. Whereas MASM will remember, on seeing \c{var dw 0}, that
1107 you declared \c{var} as a word-size variable, and will then be able
1108 to fill in the \i{ambiguity} in the size of the instruction \c{mov
1109 var,2}, NASM will deliberately remember nothing about the symbol
1110 \c{var} except where it begins, and so you must explicitly code
1111 \c{mov word [var],2}.
1113 For this reason, NASM doesn't support the \c{LODS}, \c{MOVS},
1114 \c{STOS}, \c{SCAS}, \c{CMPS}, \c{INS}, or \c{OUTS} instructions,
1115 but only supports the forms such as \c{LODSB}, \c{MOVSW}, and
1116 \c{SCASD}, which explicitly specify the size of the components of
1117 the strings being manipulated.
1120 \S{qsassume} NASM Doesn't \i\c{ASSUME}
1122 As part of NASM's drive for simplicity, it also does not support the
1123 \c{ASSUME} directive. NASM will not keep track of what values you
1124 choose to put in your segment registers, and will never
1125 \e{automatically} generate a \i{segment override} prefix.
1128 \S{qsmodel} NASM Doesn't Support \i{Memory Models}
1130 NASM also does not have any directives to support different 16-bit
1131 memory models. The programmer has to keep track of which functions
1132 are supposed to be called with a \i{far call} and which with a
1133 \i{near call}, and is responsible for putting the correct form of
1134 \c{RET} instruction (\c{RETN} or \c{RETF}; NASM accepts \c{RET}
1135 itself as an alternate form for \c{RETN}); in addition, the
1136 programmer is responsible for coding CALL FAR instructions where
1137 necessary when calling \e{external} functions, and must also keep
1138 track of which external variable definitions are far and which are
1142 \S{qsfpu} \i{Floating-Point} Differences
1144 NASM uses different names to refer to floating-point registers from
1145 MASM: where MASM would call them \c{ST(0)}, \c{ST(1)} and so on, and
1146 \i\c{a86} would call them simply \c{0}, \c{1} and so on, NASM
1147 chooses to call them \c{st0}, \c{st1} etc.
1149 As of version 0.96, NASM now treats the instructions with
1150 \i{`nowait'} forms in the same way as MASM-compatible assemblers.
1151 The idiosyncratic treatment employed by 0.95 and earlier was based
1152 on a misunderstanding by the authors.
1155 \S{qsother} Other Differences
1157 For historical reasons, NASM uses the keyword \i\c{TWORD} where MASM
1158 and compatible assemblers use \i\c{TBYTE}.
1160 NASM does not declare \i{uninitialized storage} in the same way as
1161 MASM: where a MASM programmer might use \c{stack db 64 dup (?)},
1162 NASM requires \c{stack resb 64}, intended to be read as `reserve 64
1163 bytes'. For a limited amount of compatibility, since NASM treats
1164 \c{?} as a valid character in symbol names, you can code \c{? equ 0}
1165 and then writing \c{dw ?} will at least do something vaguely useful.
1166 \I\c{RESB}\i\c{DUP} is still not a supported syntax, however.
1168 In addition to all of this, macros and directives work completely
1169 differently to MASM. See \k{preproc} and \k{directive} for further
1173 \C{lang} The NASM Language
1175 \H{syntax} Layout of a NASM Source Line
1177 Like most assemblers, each NASM source line contains (unless it
1178 is a macro, a preprocessor directive or an assembler directive: see
1179 \k{preproc} and \k{directive}) some combination of the four fields
1181 \c label: instruction operands ; comment
1183 As usual, most of these fields are optional; the presence or absence
1184 of any combination of a label, an instruction and a comment is allowed.
1185 Of course, the operand field is either required or forbidden by the
1186 presence and nature of the instruction field.
1188 NASM uses backslash (\\) as the line continuation character; if a line
1189 ends with backslash, the next line is considered to be a part of the
1190 backslash-ended line.
1192 NASM places no restrictions on white space within a line: labels may
1193 have white space before them, or instructions may have no space
1194 before them, or anything. The \i{colon} after a label is also
1195 optional. (Note that this means that if you intend to code \c{lodsb}
1196 alone on a line, and type \c{lodab} by accident, then that's still a
1197 valid source line which does nothing but define a label. Running
1198 NASM with the command-line option
1199 \I{orphan-labels}\c{-w+orphan-labels} will cause it to warn you if
1200 you define a label alone on a line without a \i{trailing colon}.)
1202 \i{Valid characters} in labels are letters, numbers, \c{_}, \c{$},
1203 \c{#}, \c{@}, \c{~}, \c{.}, and \c{?}. The only characters which may
1204 be used as the \e{first} character of an identifier are letters,
1205 \c{.} (with special meaning: see \k{locallab}), \c{_} and \c{?}.
1206 An identifier may also be prefixed with a \I{$, prefix}\c{$} to
1207 indicate that it is intended to be read as an identifier and not a
1208 reserved word; thus, if some other module you are linking with
1209 defines a symbol called \c{eax}, you can refer to \c{$eax} in NASM
1210 code to distinguish the symbol from the register. Maximum length of
1211 an identifier is 4095 characters.
1213 The instruction field may contain any machine instruction: Pentium
1214 and P6 instructions, FPU instructions, MMX instructions and even
1215 undocumented instructions are all supported. The instruction may be
1216 prefixed by \c{LOCK}, \c{REP}, \c{REPE}/\c{REPZ}, \c{REPNE}/\c{REPNZ},
1217 \c{XACQUIRE}/\c{XRELEASE} or \c{BND}, in the usual way. Explicit
1218 \I{address-size prefixes}address-size and \i{operand-size prefixes} \i\c{A16},
1219 \i\c{A32}, \i\c{A64}, \i\c{O16} and \i\c{O32}, \i\c{O64} are provided - one example of their use
1220 is given in \k{mixsize}. You can also use the name of a \I{segment
1221 override}segment register as an instruction prefix: coding
1222 \c{es mov [bx],ax} is equivalent to coding \c{mov [es:bx],ax}. We
1223 recommend the latter syntax, since it is consistent with other
1224 syntactic features of the language, but for instructions such as
1225 \c{LODSB}, which has no operands and yet can require a segment
1226 override, there is no clean syntactic way to proceed apart from
1229 An instruction is not required to use a prefix: prefixes such as
1230 \c{CS}, \c{A32}, \c{LOCK} or \c{REPE} can appear on a line by
1231 themselves, and NASM will just generate the prefix bytes.
1233 In addition to actual machine instructions, NASM also supports a
1234 number of pseudo-instructions, described in \k{pseudop}.
1236 Instruction \i{operands} may take a number of forms: they can be
1237 registers, described simply by the register name (e.g. \c{ax},
1238 \c{bp}, \c{ebx}, \c{cr0}: NASM does not use the \c{gas}-style
1239 syntax in which register names must be prefixed by a \c{%} sign), or
1240 they can be \i{effective addresses} (see \k{effaddr}), constants
1241 (\k{const}) or expressions (\k{expr}).
1243 For x87 \i{floating-point} instructions, NASM accepts a wide range of
1244 syntaxes: you can use two-operand forms like MASM supports, or you
1245 can use NASM's native single-operand forms in most cases.
1247 \# all forms of each supported instruction are given in
1249 For example, you can code:
1251 \c fadd st1 ; this sets st0 := st0 + st1
1252 \c fadd st0,st1 ; so does this
1254 \c fadd st1,st0 ; this sets st1 := st1 + st0
1255 \c fadd to st1 ; so does this
1257 Almost any x87 floating-point instruction that references memory must
1258 use one of the prefixes \i\c{DWORD}, \i\c{QWORD} or \i\c{TWORD} to
1259 indicate what size of \i{memory operand} it refers to.
1262 \H{pseudop} \i{Pseudo-Instructions}
1264 Pseudo-instructions are things which, though not real x86 machine
1265 instructions, are used in the instruction field anyway because that's
1266 the most convenient place to put them. The current pseudo-instructions
1267 are \i\c{DB}, \i\c{DW}, \i\c{DD}, \i\c{DQ}, \i\c{DT}, \i\c{DO},
1268 \i\c{DY} and \i\c\{DZ}; their \i{uninitialized} counterparts
1269 \i\c{RESB}, \i\c{RESW}, \i\c{RESD}, \i\c{RESQ}, \i\c{REST},
1270 \i\c{RESO}, \i\c{RESY} and \i\c\{RESZ}; the \i\c{INCBIN} command, the
1271 \i\c{EQU} command, and the \i\c{TIMES} prefix.
1274 \S{db} \c{DB} and Friends: Declaring Initialized Data
1276 \i\c{DB}, \i\c{DW}, \i\c{DD}, \i\c{DQ}, \i\c{DT}, \i\c{DO}, \i\c{DY}
1277 and \i\c{DZ} are used, much as in MASM, to declare initialized data in
1278 the output file. They can be invoked in a wide range of ways:
1279 \I{floating-point}\I{character constant}\I{string constant}
1281 \c db 0x55 ; just the byte 0x55
1282 \c db 0x55,0x56,0x57 ; three bytes in succession
1283 \c db 'a',0x55 ; character constants are OK
1284 \c db 'hello',13,10,'$' ; so are string constants
1285 \c dw 0x1234 ; 0x34 0x12
1286 \c dw 'a' ; 0x61 0x00 (it's just a number)
1287 \c dw 'ab' ; 0x61 0x62 (character constant)
1288 \c dw 'abc' ; 0x61 0x62 0x63 0x00 (string)
1289 \c dd 0x12345678 ; 0x78 0x56 0x34 0x12
1290 \c dd 1.234567e20 ; floating-point constant
1291 \c dq 0x123456789abcdef0 ; eight byte constant
1292 \c dq 1.234567e20 ; double-precision float
1293 \c dt 1.234567e20 ; extended-precision float
1295 \c{DT}, \c{DO}, \c{DY} and \c{DZ} do not accept \i{numeric constants}
1299 \S{resb} \c{RESB} and Friends: Declaring \i{Uninitialized} Data
1301 \i\c{RESB}, \i\c{RESW}, \i\c{RESD}, \i\c{RESQ}, \i\c{REST},
1302 \i\c{RESO}, \i\c{RESY} and \i\c\{RESZ} are designed to be used in the
1303 BSS section of a module: they declare \e{uninitialized} storage
1304 space. Each takes a single operand, which is the number of bytes,
1305 words, doublewords or whatever to reserve. As stated in \k{qsother},
1306 NASM does not support the MASM/TASM syntax of reserving uninitialized
1307 space by writing \I\c{?}\c{DW ?} or similar things: this is what it
1308 does instead. The operand to a \c{RESB}-type pseudo-instruction is a
1309 \i\e{critical expression}: see \k{crit}.
1313 \c buffer: resb 64 ; reserve 64 bytes
1314 \c wordvar: resw 1 ; reserve a word
1315 \c realarray resq 10 ; array of ten reals
1316 \c ymmval: resy 1 ; one YMM register
1317 \c zmmvals: resz 32 ; 32 ZMM registers
1319 \S{incbin} \i\c{INCBIN}: Including External \i{Binary Files}
1321 \c{INCBIN} is borrowed from the old Amiga assembler \i{DevPac}: it
1322 includes a binary file verbatim into the output file. This can be
1323 handy for (for example) including \i{graphics} and \i{sound} data
1324 directly into a game executable file. It can be called in one of
1327 \c incbin "file.dat" ; include the whole file
1328 \c incbin "file.dat",1024 ; skip the first 1024 bytes
1329 \c incbin "file.dat",1024,512 ; skip the first 1024, and
1330 \c ; actually include at most 512
1332 \c{INCBIN} is both a directive and a standard macro; the standard
1333 macro version searches for the file in the include file search path
1334 and adds the file to the dependency lists. This macro can be
1335 overridden if desired.
1338 \S{equ} \i\c{EQU}: Defining Constants
1340 \c{EQU} defines a symbol to a given constant value: when \c{EQU} is
1341 used, the source line must contain a label. The action of \c{EQU} is
1342 to define the given label name to the value of its (only) operand.
1343 This definition is absolute, and cannot change later. So, for
1346 \c message db 'hello, world'
1347 \c msglen equ $-message
1349 defines \c{msglen} to be the constant 12. \c{msglen} may not then be
1350 redefined later. This is not a \i{preprocessor} definition either:
1351 the value of \c{msglen} is evaluated \e{once}, using the value of
1352 \c{$} (see \k{expr} for an explanation of \c{$}) at the point of
1353 definition, rather than being evaluated wherever it is referenced
1354 and using the value of \c{$} at the point of reference.
1357 \S{times} \i\c{TIMES}: \i{Repeating} Instructions or Data
1359 The \c{TIMES} prefix causes the instruction to be assembled multiple
1360 times. This is partly present as NASM's equivalent of the \i\c{DUP}
1361 syntax supported by \i{MASM}-compatible assemblers, in that you can
1364 \c zerobuf: times 64 db 0
1366 or similar things; but \c{TIMES} is more versatile than that. The
1367 argument to \c{TIMES} is not just a numeric constant, but a numeric
1368 \e{expression}, so you can do things like
1370 \c buffer: db 'hello, world'
1371 \c times 64-$+buffer db ' '
1373 which will store exactly enough spaces to make the total length of
1374 \c{buffer} up to 64. Finally, \c{TIMES} can be applied to ordinary
1375 instructions, so you can code trivial \i{unrolled loops} in it:
1379 Note that there is no effective difference between \c{times 100 resb
1380 1} and \c{resb 100}, except that the latter will be assembled about
1381 100 times faster due to the internal structure of the assembler.
1383 The operand to \c{TIMES} is a critical expression (\k{crit}).
1385 Note also that \c{TIMES} can't be applied to \i{macros}: the reason
1386 for this is that \c{TIMES} is processed after the macro phase, which
1387 allows the argument to \c{TIMES} to contain expressions such as
1388 \c{64-$+buffer} as above. To repeat more than one line of code, or a
1389 complex macro, use the preprocessor \i\c{%rep} directive.
1392 \H{effaddr} Effective Addresses
1394 An \i{effective address} is any operand to an instruction which
1395 \I{memory reference}references memory. Effective addresses, in NASM,
1396 have a very simple syntax: they consist of an expression evaluating
1397 to the desired address, enclosed in \i{square brackets}. For
1402 \c mov ax,[wordvar+1]
1403 \c mov ax,[es:wordvar+bx]
1405 Anything not conforming to this simple system is not a valid memory
1406 reference in NASM, for example \c{es:wordvar[bx]}.
1408 More complicated effective addresses, such as those involving more
1409 than one register, work in exactly the same way:
1411 \c mov eax,[ebx*2+ecx+offset]
1414 NASM is capable of doing \i{algebra} on these effective addresses,
1415 so that things which don't necessarily \e{look} legal are perfectly
1418 \c mov eax,[ebx*5] ; assembles as [ebx*4+ebx]
1419 \c mov eax,[label1*2-label2] ; ie [label1+(label1-label2)]
1421 Some forms of effective address have more than one assembled form;
1422 in most such cases NASM will generate the smallest form it can. For
1423 example, there are distinct assembled forms for the 32-bit effective
1424 addresses \c{[eax*2+0]} and \c{[eax+eax]}, and NASM will generally
1425 generate the latter on the grounds that the former requires four
1426 bytes to store a zero offset.
1428 NASM has a hinting mechanism which will cause \c{[eax+ebx]} and
1429 \c{[ebx+eax]} to generate different opcodes; this is occasionally
1430 useful because \c{[esi+ebp]} and \c{[ebp+esi]} have different
1431 default segment registers.
1433 However, you can force NASM to generate an effective address in a
1434 particular form by the use of the keywords \c{BYTE}, \c{WORD},
1435 \c{DWORD} and \c{NOSPLIT}. If you need \c{[eax+3]} to be assembled
1436 using a double-word offset field instead of the one byte NASM will
1437 normally generate, you can code \c{[dword eax+3]}. Similarly, you
1438 can force NASM to use a byte offset for a small value which it
1439 hasn't seen on the first pass (see \k{crit} for an example of such a
1440 code fragment) by using \c{[byte eax+offset]}. As special cases,
1441 \c{[byte eax]} will code \c{[eax+0]} with a byte offset of zero, and
1442 \c{[dword eax]} will code it with a double-word offset of zero. The
1443 normal form, \c{[eax]}, will be coded with no offset field.
1445 The form described in the previous paragraph is also useful if you
1446 are trying to access data in a 32-bit segment from within 16 bit code.
1447 For more information on this see the section on mixed-size addressing
1448 (\k{mixaddr}). In particular, if you need to access data with a known
1449 offset that is larger than will fit in a 16-bit value, if you don't
1450 specify that it is a dword offset, nasm will cause the high word of
1451 the offset to be lost.
1453 Similarly, NASM will split \c{[eax*2]} into \c{[eax+eax]} because
1454 that allows the offset field to be absent and space to be saved; in
1455 fact, it will also split \c{[eax*2+offset]} into
1456 \c{[eax+eax+offset]}. You can combat this behaviour by the use of
1457 the \c{NOSPLIT} keyword: \c{[nosplit eax*2]} will force
1458 \c{[eax*2+0]} to be generated literally.
1460 In 64-bit mode, NASM will by default generate absolute addresses. The
1461 \i\c{REL} keyword makes it produce \c{RIP}-relative addresses. Since
1462 this is frequently the normally desired behaviour, see the \c{DEFAULT}
1463 directive (\k{default}). The keyword \i\c{ABS} overrides \i\c{REL}.
1465 A new form of split effective addres syntax is also supported. This is
1466 mainly intended for mib operands as used by MPX instructions, but can
1467 be used for any memory reference. The basic concept of this form is
1468 splitting base and index.
1470 \c mov eax,[ebx+8,ecx*4] ; ebx=base, ecx=index, 4=scale, 8=disp
1472 For mib operands, there are several ways of writing effective address depending
1473 on the tools. NASM supports all currently possible ways of mib syntax:
1476 \c ; next 5 lines are parsed same
1477 \c ; base=rax, index=rbx, scale=1, displacement=3
1478 \c bndstx [rax+0x3,rbx], bnd0 ; NASM - split EA
1479 \c bndstx [rbx*1+rax+0x3], bnd0 ; GAS - '*1' indecates an index reg
1480 \c bndstx [rax+rbx+3], bnd0 ; GAS - without hints
1481 \c bndstx [rax+0x3], bnd0, rbx ; ICC-1
1482 \c bndstx [rax+0x3], rbx, bnd0 ; ICC-2
1484 When broadcasting decorator is used, the opsize keyword should match
1485 the size of each element.
1487 \c VDIVPS zmm4, zmm5, dword [rbx]{1to16} ; single-precision float
1488 \c VDIVPS zmm4, zmm5, zword [rbx] ; packed 512 bit memory
1491 \H{const} \i{Constants}
1493 NASM understands four different types of constant: numeric,
1494 character, string and floating-point.
1497 \S{numconst} \i{Numeric Constants}
1499 A numeric constant is simply a number. NASM allows you to specify
1500 numbers in a variety of number bases, in a variety of ways: you can
1501 suffix \c{H} or \c{X}, \c{D} or \c{T}, \c{Q} or \c{O}, and \c{B} or
1502 \c{Y} for \i{hexadecimal}, \i{decimal}, \i{octal} and \i{binary}
1503 respectively, or you can prefix \c{0x}, for hexadecimal in the style
1504 of C, or you can prefix \c{$} for hexadecimal in the style of Borland
1505 Pascal or Motorola Assemblers. Note, though, that the \I{$,
1506 prefix}\c{$} prefix does double duty as a prefix on identifiers (see
1507 \k{syntax}), so a hex number prefixed with a \c{$} sign must have a
1508 digit after the \c{$} rather than a letter. In addition, current
1509 versions of NASM accept the prefix \c{0h} for hexadecimal, \c{0d} or
1510 \c{0t} for decimal, \c{0o} or \c{0q} for octal, and \c{0b} or \c{0y}
1511 for binary. Please note that unlike C, a \c{0} prefix by itself does
1512 \e{not} imply an octal constant!
1514 Numeric constants can have underscores (\c{_}) interspersed to break
1517 Some examples (all producing exactly the same code):
1519 \c mov ax,200 ; decimal
1520 \c mov ax,0200 ; still decimal
1521 \c mov ax,0200d ; explicitly decimal
1522 \c mov ax,0d200 ; also decimal
1523 \c mov ax,0c8h ; hex
1524 \c mov ax,$0c8 ; hex again: the 0 is required
1525 \c mov ax,0xc8 ; hex yet again
1526 \c mov ax,0hc8 ; still hex
1527 \c mov ax,310q ; octal
1528 \c mov ax,310o ; octal again
1529 \c mov ax,0o310 ; octal yet again
1530 \c mov ax,0q310 ; octal yet again
1531 \c mov ax,11001000b ; binary
1532 \c mov ax,1100_1000b ; same binary constant
1533 \c mov ax,1100_1000y ; same binary constant once more
1534 \c mov ax,0b1100_1000 ; same binary constant yet again
1535 \c mov ax,0y1100_1000 ; same binary constant yet again
1537 \S{strings} \I{Strings}\i{Character Strings}
1539 A character string consists of up to eight characters enclosed in
1540 either single quotes (\c{'...'}), double quotes (\c{"..."}) or
1541 backquotes (\c{`...`}). Single or double quotes are equivalent to
1542 NASM (except of course that surrounding the constant with single
1543 quotes allows double quotes to appear within it and vice versa); the
1544 contents of those are represented verbatim. Strings enclosed in
1545 backquotes support C-style \c{\\}-escapes for special characters.
1548 The following \i{escape sequences} are recognized by backquoted strings:
1550 \c \' single quote (')
1551 \c \" double quote (")
1553 \c \\\ backslash (\)
1554 \c \? question mark (?)
1562 \c \e ESC (ASCII 27)
1563 \c \377 Up to 3 octal digits - literal byte
1564 \c \xFF Up to 2 hexadecimal digits - literal byte
1565 \c \u1234 4 hexadecimal digits - Unicode character
1566 \c \U12345678 8 hexadecimal digits - Unicode character
1568 All other escape sequences are reserved. Note that \c{\\0}, meaning a
1569 \c{NUL} character (ASCII 0), is a special case of the octal escape
1572 \i{Unicode} characters specified with \c{\\u} or \c{\\U} are converted to
1573 \i{UTF-8}. For example, the following lines are all equivalent:
1575 \c db `\u263a` ; UTF-8 smiley face
1576 \c db `\xe2\x98\xba` ; UTF-8 smiley face
1577 \c db 0E2h, 098h, 0BAh ; UTF-8 smiley face
1580 \S{chrconst} \i{Character Constants}
1582 A character constant consists of a string up to eight bytes long, used
1583 in an expression context. It is treated as if it was an integer.
1585 A character constant with more than one byte will be arranged
1586 with \i{little-endian} order in mind: if you code
1590 then the constant generated is not \c{0x61626364}, but
1591 \c{0x64636261}, so that if you were then to store the value into
1592 memory, it would read \c{abcd} rather than \c{dcba}. This is also
1593 the sense of character constants understood by the Pentium's
1594 \i\c{CPUID} instruction.
1597 \S{strconst} \i{String Constants}
1599 String constants are character strings used in the context of some
1600 pseudo-instructions, namely the
1601 \I\c{DW}\I\c{DD}\I\c{DQ}\I\c{DT}\I\c{DO}\I\c{DY}\i\c{DB} family and
1602 \i\c{INCBIN} (where it represents a filename.) They are also used in
1603 certain preprocessor directives.
1605 A string constant looks like a character constant, only longer. It
1606 is treated as a concatenation of maximum-size character constants
1607 for the conditions. So the following are equivalent:
1609 \c db 'hello' ; string constant
1610 \c db 'h','e','l','l','o' ; equivalent character constants
1612 And the following are also equivalent:
1614 \c dd 'ninechars' ; doubleword string constant
1615 \c dd 'nine','char','s' ; becomes three doublewords
1616 \c db 'ninechars',0,0,0 ; and really looks like this
1618 Note that when used in a string-supporting context, quoted strings are
1619 treated as a string constants even if they are short enough to be a
1620 character constant, because otherwise \c{db 'ab'} would have the same
1621 effect as \c{db 'a'}, which would be silly. Similarly, three-character
1622 or four-character constants are treated as strings when they are
1623 operands to \c{DW}, and so forth.
1625 \S{unicode} \I{UTF-16}\I{UTF-32}\i{Unicode} Strings
1627 The special operators \i\c{__utf16__}, \i\c{__utf16le__},
1628 \i\c{__utf16be__}, \i\c{__utf32__}, \i\c{__utf32le__} and
1629 \i\c{__utf32be__} allows definition of Unicode strings. They take a
1630 string in UTF-8 format and converts it to UTF-16 or UTF-32,
1631 respectively. Unless the \c{be} forms are specified, the output is
1636 \c %define u(x) __utf16__(x)
1637 \c %define w(x) __utf32__(x)
1639 \c dw u('C:\WINDOWS'), 0 ; Pathname in UTF-16
1640 \c dd w(`A + B = \u206a`), 0 ; String in UTF-32
1642 The UTF operators can be applied either to strings passed to the
1643 \c{DB} family instructions, or to character constants in an expression
1646 \S{fltconst} \I{floating-point, constants}Floating-Point Constants
1648 \i{Floating-point} constants are acceptable only as arguments to
1649 \i\c{DB}, \i\c{DW}, \i\c{DD}, \i\c{DQ}, \i\c{DT}, and \i\c{DO}, or as
1650 arguments to the special operators \i\c{__float8__},
1651 \i\c{__float16__}, \i\c{__float32__}, \i\c{__float64__},
1652 \i\c{__float80m__}, \i\c{__float80e__}, \i\c{__float128l__}, and
1653 \i\c{__float128h__}.
1655 Floating-point constants are expressed in the traditional form:
1656 digits, then a period, then optionally more digits, then optionally an
1657 \c{E} followed by an exponent. The period is mandatory, so that NASM
1658 can distinguish between \c{dd 1}, which declares an integer constant,
1659 and \c{dd 1.0} which declares a floating-point constant.
1661 NASM also support C99-style hexadecimal floating-point: \c{0x},
1662 hexadecimal digits, period, optionally more hexadeximal digits, then
1663 optionally a \c{P} followed by a \e{binary} (not hexadecimal) exponent
1664 in decimal notation. As an extension, NASM additionally supports the
1665 \c{0h} and \c{$} prefixes for hexadecimal, as well binary and octal
1666 floating-point, using the \c{0b} or \c{0y} and \c{0o} or \c{0q}
1667 prefixes, respectively.
1669 Underscores to break up groups of digits are permitted in
1670 floating-point constants as well.
1674 \c db -0.2 ; "Quarter precision"
1675 \c dw -0.5 ; IEEE 754r/SSE5 half precision
1676 \c dd 1.2 ; an easy one
1677 \c dd 1.222_222_222 ; underscores are permitted
1678 \c dd 0x1p+2 ; 1.0x2^2 = 4.0
1679 \c dq 0x1p+32 ; 1.0x2^32 = 4 294 967 296.0
1680 \c dq 1.e10 ; 10 000 000 000.0
1681 \c dq 1.e+10 ; synonymous with 1.e10
1682 \c dq 1.e-10 ; 0.000 000 000 1
1683 \c dt 3.141592653589793238462 ; pi
1684 \c do 1.e+4000 ; IEEE 754r quad precision
1686 The 8-bit "quarter-precision" floating-point format is
1687 sign:exponent:mantissa = 1:4:3 with an exponent bias of 7. This
1688 appears to be the most frequently used 8-bit floating-point format,
1689 although it is not covered by any formal standard. This is sometimes
1690 called a "\i{minifloat}."
1692 The special operators are used to produce floating-point numbers in
1693 other contexts. They produce the binary representation of a specific
1694 floating-point number as an integer, and can use anywhere integer
1695 constants are used in an expression. \c{__float80m__} and
1696 \c{__float80e__} produce the 64-bit mantissa and 16-bit exponent of an
1697 80-bit floating-point number, and \c{__float128l__} and
1698 \c{__float128h__} produce the lower and upper 64-bit halves of a 128-bit
1699 floating-point number, respectively.
1703 \c mov rax,__float64__(3.141592653589793238462)
1705 ... would assign the binary representation of pi as a 64-bit floating
1706 point number into \c{RAX}. This is exactly equivalent to:
1708 \c mov rax,0x400921fb54442d18
1710 NASM cannot do compile-time arithmetic on floating-point constants.
1711 This is because NASM is designed to be portable - although it always
1712 generates code to run on x86 processors, the assembler itself can
1713 run on any system with an ANSI C compiler. Therefore, the assembler
1714 cannot guarantee the presence of a floating-point unit capable of
1715 handling the \i{Intel number formats}, and so for NASM to be able to
1716 do floating arithmetic it would have to include its own complete set
1717 of floating-point routines, which would significantly increase the
1718 size of the assembler for very little benefit.
1720 The special tokens \i\c{__Infinity__}, \i\c{__QNaN__} (or
1721 \i\c{__NaN__}) and \i\c{__SNaN__} can be used to generate
1722 \I{infinity}infinities, quiet \i{NaN}s, and signalling NaNs,
1723 respectively. These are normally used as macros:
1725 \c %define Inf __Infinity__
1726 \c %define NaN __QNaN__
1728 \c dq +1.5, -Inf, NaN ; Double-precision constants
1730 The \c{%use fp} standard macro package contains a set of convenience
1731 macros. See \k{pkg_fp}.
1733 \S{bcdconst} \I{floating-point, packed BCD constants}Packed BCD Constants
1735 x87-style packed BCD constants can be used in the same contexts as
1736 80-bit floating-point numbers. They are suffixed with \c{p} or
1737 prefixed with \c{0p}, and can include up to 18 decimal digits.
1739 As with other numeric constants, underscores can be used to separate
1744 \c dt 12_345_678_901_245_678p
1745 \c dt -12_345_678_901_245_678p
1750 \H{expr} \i{Expressions}
1752 Expressions in NASM are similar in syntax to those in C. Expressions
1753 are evaluated as 64-bit integers which are then adjusted to the
1756 NASM supports two special tokens in expressions, allowing
1757 calculations to involve the current assembly position: the
1758 \I{$, here}\c{$} and \i\c{$$} tokens. \c{$} evaluates to the assembly
1759 position at the beginning of the line containing the expression; so
1760 you can code an \i{infinite loop} using \c{JMP $}. \c{$$} evaluates
1761 to the beginning of the current section; so you can tell how far
1762 into the section you are by using \c{($-$$)}.
1764 The arithmetic \i{operators} provided by NASM are listed here, in
1765 increasing order of \i{precedence}.
1768 \S{expor} \i\c{|}: \i{Bitwise OR} Operator
1770 The \c{|} operator gives a bitwise OR, exactly as performed by the
1771 \c{OR} machine instruction. Bitwise OR is the lowest-priority
1772 arithmetic operator supported by NASM.
1775 \S{expxor} \i\c{^}: \i{Bitwise XOR} Operator
1777 \c{^} provides the bitwise XOR operation.
1780 \S{expand} \i\c{&}: \i{Bitwise AND} Operator
1782 \c{&} provides the bitwise AND operation.
1785 \S{expshift} \i\c{<<} and \i\c{>>}: \i{Bit Shift} Operators
1787 \c{<<} gives a bit-shift to the left, just as it does in C. So \c{5<<3}
1788 evaluates to 5 times 8, or 40. \c{>>} gives a bit-shift to the
1789 right; in NASM, such a shift is \e{always} unsigned, so that
1790 the bits shifted in from the left-hand end are filled with zero
1791 rather than a sign-extension of the previous highest bit.
1794 \S{expplmi} \I{+ opaddition}\c{+} and \I{- opsubtraction}\c{-}:
1795 \i{Addition} and \i{Subtraction} Operators
1797 The \c{+} and \c{-} operators do perfectly ordinary addition and
1801 \S{expmul} \i\c{*}, \i\c{/}, \i\c{//}, \i\c{%} and \i\c{%%}:
1802 \i{Multiplication} and \i{Division}
1804 \c{*} is the multiplication operator. \c{/} and \c{//} are both
1805 division operators: \c{/} is \i{unsigned division} and \c{//} is
1806 \i{signed division}. Similarly, \c{%} and \c{%%} provide \I{unsigned
1807 modulo}\I{modulo operators}unsigned and
1808 \i{signed modulo} operators respectively.
1810 NASM, like ANSI C, provides no guarantees about the sensible
1811 operation of the signed modulo operator.
1813 Since the \c{%} character is used extensively by the macro
1814 \i{preprocessor}, you should ensure that both the signed and unsigned
1815 modulo operators are followed by white space wherever they appear.
1818 \S{expmul} \i{Unary Operators}
1820 The highest-priority operators in NASM's expression grammar are those
1821 which only apply to one argument. These are \I{+ opunary}\c{+}, \I{-
1822 opunary}\c{-}, \i\c{~}, \I{! opunary}\c{!}, \i\c{SEG}, and the
1823 \i{integer functions} operators.
1825 \c{-} negates its operand, \c{+} does nothing (it's provided for
1826 symmetry with \c{-}), \c{~} computes the \i{one's complement} of its
1827 operand, \c{!} is the \i{logical negation} operator.
1829 \c{SEG} provides the \i{segment address}
1830 of its operand (explained in more detail in \k{segwrt}).
1832 A set of additional operators with leading and trailing double
1833 underscores are used to implement the integer functions of the
1834 \c{ifunc} macro package, see \k{pkg_ifunc}.
1837 \H{segwrt} \i\c{SEG} and \i\c{WRT}
1839 When writing large 16-bit programs, which must be split into
1840 multiple \i{segments}, it is often necessary to be able to refer to
1841 the \I{segment address}segment part of the address of a symbol. NASM
1842 supports the \c{SEG} operator to perform this function.
1844 The \c{SEG} operator returns the \i\e{preferred} segment base of a
1845 symbol, defined as the segment base relative to which the offset of
1846 the symbol makes sense. So the code
1848 \c mov ax,seg symbol
1852 will load \c{ES:BX} with a valid pointer to the symbol \c{symbol}.
1854 Things can be more complex than this: since 16-bit segments and
1855 \i{groups} may \I{overlapping segments}overlap, you might occasionally
1856 want to refer to some symbol using a different segment base from the
1857 preferred one. NASM lets you do this, by the use of the \c{WRT}
1858 (With Reference To) keyword. So you can do things like
1860 \c mov ax,weird_seg ; weird_seg is a segment base
1862 \c mov bx,symbol wrt weird_seg
1864 to load \c{ES:BX} with a different, but functionally equivalent,
1865 pointer to the symbol \c{symbol}.
1867 NASM supports far (inter-segment) calls and jumps by means of the
1868 syntax \c{call segment:offset}, where \c{segment} and \c{offset}
1869 both represent immediate values. So to call a far procedure, you
1870 could code either of
1872 \c call (seg procedure):procedure
1873 \c call weird_seg:(procedure wrt weird_seg)
1875 (The parentheses are included for clarity, to show the intended
1876 parsing of the above instructions. They are not necessary in
1879 NASM supports the syntax \I\c{CALL FAR}\c{call far procedure} as a
1880 synonym for the first of the above usages. \c{JMP} works identically
1881 to \c{CALL} in these examples.
1883 To declare a \i{far pointer} to a data item in a data segment, you
1886 \c dw symbol, seg symbol
1888 NASM supports no convenient synonym for this, though you can always
1889 invent one using the macro processor.
1892 \H{strict} \i\c{STRICT}: Inhibiting Optimization
1894 When assembling with the optimizer set to level 2 or higher (see
1895 \k{opt-O}), NASM will use size specifiers (\c{BYTE}, \c{WORD},
1896 \c{DWORD}, \c{QWORD}, \c{TWORD}, \c{OWORD}, \c{YWORD} or \c{ZWORD}),
1897 but will give them the smallest possible size. The keyword \c{STRICT}
1898 can be used to inhibit optimization and force a particular operand to
1899 be emitted in the specified size. For example, with the optimizer on,
1900 and in \c{BITS 16} mode,
1904 is encoded in three bytes \c{66 6A 21}, whereas
1906 \c push strict dword 33
1908 is encoded in six bytes, with a full dword immediate operand \c{66 68
1911 With the optimizer off, the same code (six bytes) is generated whether
1912 the \c{STRICT} keyword was used or not.
1915 \H{crit} \i{Critical Expressions}
1917 Although NASM has an optional multi-pass optimizer, there are some
1918 expressions which must be resolvable on the first pass. These are
1919 called \e{Critical Expressions}.
1921 The first pass is used to determine the size of all the assembled
1922 code and data, so that the second pass, when generating all the
1923 code, knows all the symbol addresses the code refers to. So one
1924 thing NASM can't handle is code whose size depends on the value of a
1925 symbol declared after the code in question. For example,
1927 \c times (label-$) db 0
1928 \c label: db 'Where am I?'
1930 The argument to \i\c{TIMES} in this case could equally legally
1931 evaluate to anything at all; NASM will reject this example because
1932 it cannot tell the size of the \c{TIMES} line when it first sees it.
1933 It will just as firmly reject the slightly \I{paradox}paradoxical
1936 \c times (label-$+1) db 0
1937 \c label: db 'NOW where am I?'
1939 in which \e{any} value for the \c{TIMES} argument is by definition
1942 NASM rejects these examples by means of a concept called a
1943 \e{critical expression}, which is defined to be an expression whose
1944 value is required to be computable in the first pass, and which must
1945 therefore depend only on symbols defined before it. The argument to
1946 the \c{TIMES} prefix is a critical expression.
1948 \H{locallab} \i{Local Labels}
1950 NASM gives special treatment to symbols beginning with a \i{period}.
1951 A label beginning with a single period is treated as a \e{local}
1952 label, which means that it is associated with the previous non-local
1953 label. So, for example:
1955 \c label1 ; some code
1963 \c label2 ; some code
1971 In the above code fragment, each \c{JNE} instruction jumps to the
1972 line immediately before it, because the two definitions of \c{.loop}
1973 are kept separate by virtue of each being associated with the
1974 previous non-local label.
1976 This form of local label handling is borrowed from the old Amiga
1977 assembler \i{DevPac}; however, NASM goes one step further, in
1978 allowing access to local labels from other parts of the code. This
1979 is achieved by means of \e{defining} a local label in terms of the
1980 previous non-local label: the first definition of \c{.loop} above is
1981 really defining a symbol called \c{label1.loop}, and the second
1982 defines a symbol called \c{label2.loop}. So, if you really needed
1985 \c label3 ; some more code
1990 Sometimes it is useful - in a macro, for instance - to be able to
1991 define a label which can be referenced from anywhere but which
1992 doesn't interfere with the normal local-label mechanism. Such a
1993 label can't be non-local because it would interfere with subsequent
1994 definitions of, and references to, local labels; and it can't be
1995 local because the macro that defined it wouldn't know the label's
1996 full name. NASM therefore introduces a third type of label, which is
1997 probably only useful in macro definitions: if a label begins with
1998 the \I{label prefix}special prefix \i\c{..@}, then it does nothing
1999 to the local label mechanism. So you could code
2001 \c label1: ; a non-local label
2002 \c .local: ; this is really label1.local
2003 \c ..@foo: ; this is a special symbol
2004 \c label2: ; another non-local label
2005 \c .local: ; this is really label2.local
2007 \c jmp ..@foo ; this will jump three lines up
2009 NASM has the capacity to define other special symbols beginning with
2010 a double period: for example, \c{..start} is used to specify the
2011 entry point in the \c{obj} output format (see \k{dotdotstart}),
2012 \c{..imagebase} is used to find out the offset from a base address
2013 of the current image in the \c{win64} output format (see \k{win64pic}).
2014 So just keep in mind that symbols beginning with a double period are
2018 \C{preproc} The NASM \i{Preprocessor}
2020 NASM contains a powerful \i{macro processor}, which supports
2021 conditional assembly, multi-level file inclusion, two forms of macro
2022 (single-line and multi-line), and a `context stack' mechanism for
2023 extra macro power. Preprocessor directives all begin with a \c{%}
2026 The preprocessor collapses all lines which end with a backslash (\\)
2027 character into a single line. Thus:
2029 \c %define THIS_VERY_LONG_MACRO_NAME_IS_DEFINED_TO \\
2032 will work like a single-line macro without the backslash-newline
2035 \H{slmacro} \i{Single-Line Macros}
2037 \S{define} The Normal Way: \I\c{%idefine}\i\c{%define}
2039 Single-line macros are defined using the \c{%define} preprocessor
2040 directive. The definitions work in a similar way to C; so you can do
2043 \c %define ctrl 0x1F &
2044 \c %define param(a,b) ((a)+(a)*(b))
2046 \c mov byte [param(2,ebx)], ctrl 'D'
2048 which will expand to
2050 \c mov byte [(2)+(2)*(ebx)], 0x1F & 'D'
2052 When the expansion of a single-line macro contains tokens which
2053 invoke another macro, the expansion is performed at invocation time,
2054 not at definition time. Thus the code
2056 \c %define a(x) 1+b(x)
2061 will evaluate in the expected way to \c{mov ax,1+2*8}, even though
2062 the macro \c{b} wasn't defined at the time of definition of \c{a}.
2064 Macros defined with \c{%define} are \i{case sensitive}: after
2065 \c{%define foo bar}, only \c{foo} will expand to \c{bar}: \c{Foo} or
2066 \c{FOO} will not. By using \c{%idefine} instead of \c{%define} (the
2067 `i' stands for `insensitive') you can define all the case variants
2068 of a macro at once, so that \c{%idefine foo bar} would cause
2069 \c{foo}, \c{Foo}, \c{FOO}, \c{fOO} and so on all to expand to
2072 There is a mechanism which detects when a macro call has occurred as
2073 a result of a previous expansion of the same macro, to guard against
2074 \i{circular references} and infinite loops. If this happens, the
2075 preprocessor will only expand the first occurrence of the macro.
2078 \c %define a(x) 1+a(x)
2082 the macro \c{a(3)} will expand once, becoming \c{1+a(3)}, and will
2083 then expand no further. This behaviour can be useful: see \k{32c}
2084 for an example of its use.
2086 You can \I{overloading, single-line macros}overload single-line
2087 macros: if you write
2089 \c %define foo(x) 1+x
2090 \c %define foo(x,y) 1+x*y
2092 the preprocessor will be able to handle both types of macro call,
2093 by counting the parameters you pass; so \c{foo(3)} will become
2094 \c{1+3} whereas \c{foo(ebx,2)} will become \c{1+ebx*2}. However, if
2099 then no other definition of \c{foo} will be accepted: a macro with
2100 no parameters prohibits the definition of the same name as a macro
2101 \e{with} parameters, and vice versa.
2103 This doesn't prevent single-line macros being \e{redefined}: you can
2104 perfectly well define a macro with
2108 and then re-define it later in the same source file with
2112 Then everywhere the macro \c{foo} is invoked, it will be expanded
2113 according to the most recent definition. This is particularly useful
2114 when defining single-line macros with \c{%assign} (see \k{assign}).
2116 You can \i{pre-define} single-line macros using the `-d' option on
2117 the NASM command line: see \k{opt-d}.
2120 \S{xdefine} Resolving \c{%define}: \I\c{%ixdefine}\i\c{%xdefine}
2122 To have a reference to an embedded single-line macro resolved at the
2123 time that the embedding macro is \e{defined}, as opposed to when the
2124 embedding macro is \e{expanded}, you need a different mechanism to the
2125 one offered by \c{%define}. The solution is to use \c{%xdefine}, or
2126 it's \I{case sensitive}case-insensitive counterpart \c{%ixdefine}.
2128 Suppose you have the following code:
2131 \c %define isFalse isTrue
2140 In this case, \c{val1} is equal to 0, and \c{val2} is equal to 1.
2141 This is because, when a single-line macro is defined using
2142 \c{%define}, it is expanded only when it is called. As \c{isFalse}
2143 expands to \c{isTrue}, the expansion will be the current value of
2144 \c{isTrue}. The first time it is called that is 0, and the second
2147 If you wanted \c{isFalse} to expand to the value assigned to the
2148 embedded macro \c{isTrue} at the time that \c{isFalse} was defined,
2149 you need to change the above code to use \c{%xdefine}.
2151 \c %xdefine isTrue 1
2152 \c %xdefine isFalse isTrue
2153 \c %xdefine isTrue 0
2157 \c %xdefine isTrue 1
2161 Now, each time that \c{isFalse} is called, it expands to 1,
2162 as that is what the embedded macro \c{isTrue} expanded to at
2163 the time that \c{isFalse} was defined.
2166 \S{indmacro} \i{Macro Indirection}: \I\c{%[}\c{%[...]}
2168 The \c{%[...]} construct can be used to expand macros in contexts
2169 where macro expansion would otherwise not occur, including in the
2170 names other macros. For example, if you have a set of macros named
2171 \c{Foo16}, \c{Foo32} and \c{Foo64}, you could write:
2173 \c mov ax,Foo%[__BITS__] ; The Foo value
2175 to use the builtin macro \c{__BITS__} (see \k{bitsm}) to automatically
2176 select between them. Similarly, the two statements:
2178 \c %xdefine Bar Quux ; Expands due to %xdefine
2179 \c %define Bar %[Quux] ; Expands due to %[...]
2181 have, in fact, exactly the same effect.
2183 \c{%[...]} concatenates to adjacent tokens in the same way that
2184 multi-line macro parameters do, see \k{concat} for details.
2187 \S{concat%+} Concatenating Single Line Macro Tokens: \i\c{%+}
2189 Individual tokens in single line macros can be concatenated, to produce
2190 longer tokens for later processing. This can be useful if there are
2191 several similar macros that perform similar functions.
2193 Please note that a space is required after \c{%+}, in order to
2194 disambiguate it from the syntax \c{%+1} used in multiline macros.
2196 As an example, consider the following:
2198 \c %define BDASTART 400h ; Start of BIOS data area
2200 \c struc tBIOSDA ; its structure
2206 Now, if we need to access the elements of tBIOSDA in different places,
2209 \c mov ax,BDASTART + tBIOSDA.COM1addr
2210 \c mov bx,BDASTART + tBIOSDA.COM2addr
2212 This will become pretty ugly (and tedious) if used in many places, and
2213 can be reduced in size significantly by using the following macro:
2215 \c ; Macro to access BIOS variables by their names (from tBDA):
2217 \c %define BDA(x) BDASTART + tBIOSDA. %+ x
2219 Now the above code can be written as:
2221 \c mov ax,BDA(COM1addr)
2222 \c mov bx,BDA(COM2addr)
2224 Using this feature, we can simplify references to a lot of macros (and,
2225 in turn, reduce typing errors).
2228 \S{selfref%?} The Macro Name Itself: \i\c{%?} and \i\c{%??}
2230 The special symbols \c{%?} and \c{%??} can be used to reference the
2231 macro name itself inside a macro expansion, this is supported for both
2232 single-and multi-line macros. \c{%?} refers to the macro name as
2233 \e{invoked}, whereas \c{%??} refers to the macro name as
2234 \e{declared}. The two are always the same for case-sensitive
2235 macros, but for case-insensitive macros, they can differ.
2239 \c %idefine Foo mov %?,%??
2251 \c %idefine keyword $%?
2253 can be used to make a keyword "disappear", for example in case a new
2254 instruction has been used as a label in older code. For example:
2256 \c %idefine pause $%? ; Hide the PAUSE instruction
2259 \S{undef} Undefining Single-Line Macros: \i\c{%undef}
2261 Single-line macros can be removed with the \c{%undef} directive. For
2262 example, the following sequence:
2269 will expand to the instruction \c{mov eax, foo}, since after
2270 \c{%undef} the macro \c{foo} is no longer defined.
2272 Macros that would otherwise be pre-defined can be undefined on the
2273 command-line using the `-u' option on the NASM command line: see
2277 \S{assign} \i{Preprocessor Variables}: \i\c{%assign}
2279 An alternative way to define single-line macros is by means of the
2280 \c{%assign} command (and its \I{case sensitive}case-insensitive
2281 counterpart \i\c{%iassign}, which differs from \c{%assign} in
2282 exactly the same way that \c{%idefine} differs from \c{%define}).
2284 \c{%assign} is used to define single-line macros which take no
2285 parameters and have a numeric value. This value can be specified in
2286 the form of an expression, and it will be evaluated once, when the
2287 \c{%assign} directive is processed.
2289 Like \c{%define}, macros defined using \c{%assign} can be re-defined
2290 later, so you can do things like
2294 to increment the numeric value of a macro.
2296 \c{%assign} is useful for controlling the termination of \c{%rep}
2297 preprocessor loops: see \k{rep} for an example of this. Another
2298 use for \c{%assign} is given in \k{16c} and \k{32c}.
2300 The expression passed to \c{%assign} is a \i{critical expression}
2301 (see \k{crit}), and must also evaluate to a pure number (rather than
2302 a relocatable reference such as a code or data address, or anything
2303 involving a register).
2306 \S{defstr} Defining Strings: \I\c{%idefstr}\i\c{%defstr}
2308 \c{%defstr}, and its case-insensitive counterpart \c{%idefstr}, define
2309 or redefine a single-line macro without parameters but converts the
2310 entire right-hand side, after macro expansion, to a quoted string
2315 \c %defstr test TEST
2319 \c %define test 'TEST'
2321 This can be used, for example, with the \c{%!} construct (see
2324 \c %defstr PATH %!PATH ; The operating system PATH variable
2327 \S{deftok} Defining Tokens: \I\c{%ideftok}\i\c{%deftok}
2329 \c{%deftok}, and its case-insensitive counterpart \c{%ideftok}, define
2330 or redefine a single-line macro without parameters but converts the
2331 second parameter, after string conversion, to a sequence of tokens.
2335 \c %deftok test 'TEST'
2339 \c %define test TEST
2342 \H{strlen} \i{String Manipulation in Macros}
2344 It's often useful to be able to handle strings in macros. NASM
2345 supports a few simple string handling macro operators from which
2346 more complex operations can be constructed.
2348 All the string operators define or redefine a value (either a string
2349 or a numeric value) to a single-line macro. When producing a string
2350 value, it may change the style of quoting of the input string or
2351 strings, and possibly use \c{\\}-escapes inside \c{`}-quoted strings.
2353 \S{strcat} \i{Concatenating Strings}: \i\c{%strcat}
2355 The \c{%strcat} operator concatenates quoted strings and assign them to
2356 a single-line macro.
2360 \c %strcat alpha "Alpha: ", '12" screen'
2362 ... would assign the value \c{'Alpha: 12" screen'} to \c{alpha}.
2365 \c %strcat beta '"foo"\', "'bar'"
2367 ... would assign the value \c{`"foo"\\\\'bar'`} to \c{beta}.
2369 The use of commas to separate strings is permitted but optional.
2372 \S{strlen} \i{String Length}: \i\c{%strlen}
2374 The \c{%strlen} operator assigns the length of a string to a macro.
2377 \c %strlen charcnt 'my string'
2379 In this example, \c{charcnt} would receive the value 9, just as
2380 if an \c{%assign} had been used. In this example, \c{'my string'}
2381 was a literal string but it could also have been a single-line
2382 macro that expands to a string, as in the following example:
2384 \c %define sometext 'my string'
2385 \c %strlen charcnt sometext
2387 As in the first case, this would result in \c{charcnt} being
2388 assigned the value of 9.
2391 \S{substr} \i{Extracting Substrings}: \i\c{%substr}
2393 Individual letters or substrings in strings can be extracted using the
2394 \c{%substr} operator. An example of its use is probably more useful
2395 than the description:
2397 \c %substr mychar 'xyzw' 1 ; equivalent to %define mychar 'x'
2398 \c %substr mychar 'xyzw' 2 ; equivalent to %define mychar 'y'
2399 \c %substr mychar 'xyzw' 3 ; equivalent to %define mychar 'z'
2400 \c %substr mychar 'xyzw' 2,2 ; equivalent to %define mychar 'yz'
2401 \c %substr mychar 'xyzw' 2,-1 ; equivalent to %define mychar 'yzw'
2402 \c %substr mychar 'xyzw' 2,-2 ; equivalent to %define mychar 'yz'
2404 As with \c{%strlen} (see \k{strlen}), the first parameter is the
2405 single-line macro to be created and the second is the string. The
2406 third parameter specifies the first character to be selected, and the
2407 optional fourth parameter preceeded by comma) is the length. Note
2408 that the first index is 1, not 0 and the last index is equal to the
2409 value that \c{%strlen} would assign given the same string. Index
2410 values out of range result in an empty string. A negative length
2411 means "until N-1 characters before the end of string", i.e. \c{-1}
2412 means until end of string, \c{-2} until one character before, etc.
2415 \H{mlmacro} \i{Multi-Line Macros}: \I\c{%imacro}\i\c{%macro}
2417 Multi-line macros are much more like the type of macro seen in MASM
2418 and TASM: a multi-line macro definition in NASM looks something like
2421 \c %macro prologue 1
2429 This defines a C-like function prologue as a macro: so you would
2430 invoke the macro with a call such as
2432 \c myfunc: prologue 12
2434 which would expand to the three lines of code
2440 The number \c{1} after the macro name in the \c{%macro} line defines
2441 the number of parameters the macro \c{prologue} expects to receive.
2442 The use of \c{%1} inside the macro definition refers to the first
2443 parameter to the macro call. With a macro taking more than one
2444 parameter, subsequent parameters would be referred to as \c{%2},
2447 Multi-line macros, like single-line macros, are \i{case-sensitive},
2448 unless you define them using the alternative directive \c{%imacro}.
2450 If you need to pass a comma as \e{part} of a parameter to a
2451 multi-line macro, you can do that by enclosing the entire parameter
2452 in \I{braces, around macro parameters}braces. So you could code
2461 \c silly 'a', letter_a ; letter_a: db 'a'
2462 \c silly 'ab', string_ab ; string_ab: db 'ab'
2463 \c silly {13,10}, crlf ; crlf: db 13,10
2466 \S{mlmacover} Overloading Multi-Line Macros\I{overloading, multi-line macros}
2468 As with single-line macros, multi-line macros can be overloaded by
2469 defining the same macro name several times with different numbers of
2470 parameters. This time, no exception is made for macros with no
2471 parameters at all. So you could define
2473 \c %macro prologue 0
2480 to define an alternative form of the function prologue which
2481 allocates no local stack space.
2483 Sometimes, however, you might want to `overload' a machine
2484 instruction; for example, you might want to define
2493 so that you could code
2495 \c push ebx ; this line is not a macro call
2496 \c push eax,ecx ; but this one is
2498 Ordinarily, NASM will give a warning for the first of the above two
2499 lines, since \c{push} is now defined to be a macro, and is being
2500 invoked with a number of parameters for which no definition has been
2501 given. The correct code will still be generated, but the assembler
2502 will give a warning. This warning can be disabled by the use of the
2503 \c{-w-macro-params} command-line option (see \k{opt-w}).
2506 \S{maclocal} \i{Macro-Local Labels}
2508 NASM allows you to define labels within a multi-line macro
2509 definition in such a way as to make them local to the macro call: so
2510 calling the same macro multiple times will use a different label
2511 each time. You do this by prefixing \i\c{%%} to the label name. So
2512 you can invent an instruction which executes a \c{RET} if the \c{Z}
2513 flag is set by doing this:
2523 You can call this macro as many times as you want, and every time
2524 you call it NASM will make up a different `real' name to substitute
2525 for the label \c{%%skip}. The names NASM invents are of the form
2526 \c{..@2345.skip}, where the number 2345 changes with every macro
2527 call. The \i\c{..@} prefix prevents macro-local labels from
2528 interfering with the local label mechanism, as described in
2529 \k{locallab}. You should avoid defining your own labels in this form
2530 (the \c{..@} prefix, then a number, then another period) in case
2531 they interfere with macro-local labels.
2534 \S{mlmacgre} \i{Greedy Macro Parameters}
2536 Occasionally it is useful to define a macro which lumps its entire
2537 command line into one parameter definition, possibly after
2538 extracting one or two smaller parameters from the front. An example
2539 might be a macro to write a text string to a file in MS-DOS, where
2540 you might want to be able to write
2542 \c writefile [filehandle],"hello, world",13,10
2544 NASM allows you to define the last parameter of a macro to be
2545 \e{greedy}, meaning that if you invoke the macro with more
2546 parameters than it expects, all the spare parameters get lumped into
2547 the last defined one along with the separating commas. So if you
2550 \c %macro writefile 2+
2556 \c mov cx,%%endstr-%%str
2563 then the example call to \c{writefile} above will work as expected:
2564 the text before the first comma, \c{[filehandle]}, is used as the
2565 first macro parameter and expanded when \c{%1} is referred to, and
2566 all the subsequent text is lumped into \c{%2} and placed after the
2569 The greedy nature of the macro is indicated to NASM by the use of
2570 the \I{+ modifier}\c{+} sign after the parameter count on the
2573 If you define a greedy macro, you are effectively telling NASM how
2574 it should expand the macro given \e{any} number of parameters from
2575 the actual number specified up to infinity; in this case, for
2576 example, NASM now knows what to do when it sees a call to
2577 \c{writefile} with 2, 3, 4 or more parameters. NASM will take this
2578 into account when overloading macros, and will not allow you to
2579 define another form of \c{writefile} taking 4 parameters (for
2582 Of course, the above macro could have been implemented as a
2583 non-greedy macro, in which case the call to it would have had to
2586 \c writefile [filehandle], {"hello, world",13,10}
2588 NASM provides both mechanisms for putting \i{commas in macro
2589 parameters}, and you choose which one you prefer for each macro
2592 See \k{sectmac} for a better way to write the above macro.
2594 \S{mlmacrange} \i{Macro Parameters Range}
2596 NASM allows you to expand parameters via special construction \c{%\{x:y\}}
2597 where \c{x} is the first parameter index and \c{y} is the last. Any index can
2598 be either negative or positive but must never be zero.
2608 expands to \c{3,4,5} range.
2610 Even more, the parameters can be reversed so that
2618 expands to \c{5,4,3} range.
2620 But even this is not the last. The parameters can be addressed via negative
2621 indices so NASM will count them reversed. The ones who know Python may see
2630 expands to \c{6,5,4} range.
2632 Note that NASM uses \i{comma} to separate parameters being expanded.
2634 By the way, here is a trick - you might use the index \c{%{-1:-1}}
2635 which gives you the \i{last} argument passed to a macro.
2637 \S{mlmacdef} \i{Default Macro Parameters}
2639 NASM also allows you to define a multi-line macro with a \e{range}
2640 of allowable parameter counts. If you do this, you can specify
2641 defaults for \i{omitted parameters}. So, for example:
2643 \c %macro die 0-1 "Painful program death has occurred."
2651 This macro (which makes use of the \c{writefile} macro defined in
2652 \k{mlmacgre}) can be called with an explicit error message, which it
2653 will display on the error output stream before exiting, or it can be
2654 called with no parameters, in which case it will use the default
2655 error message supplied in the macro definition.
2657 In general, you supply a minimum and maximum number of parameters
2658 for a macro of this type; the minimum number of parameters are then
2659 required in the macro call, and then you provide defaults for the
2660 optional ones. So if a macro definition began with the line
2662 \c %macro foobar 1-3 eax,[ebx+2]
2664 then it could be called with between one and three parameters, and
2665 \c{%1} would always be taken from the macro call. \c{%2}, if not
2666 specified by the macro call, would default to \c{eax}, and \c{%3} if
2667 not specified would default to \c{[ebx+2]}.
2669 You can provide extra information to a macro by providing
2670 too many default parameters:
2672 \c %macro quux 1 something
2674 This will trigger a warning by default; see \k{opt-w} for
2676 When \c{quux} is invoked, it receives not one but two parameters.
2677 \c{something} can be referred to as \c{%2}. The difference
2678 between passing \c{something} this way and writing \c{something}
2679 in the macro body is that with this way \c{something} is evaluated
2680 when the macro is defined, not when it is expanded.
2682 You may omit parameter defaults from the macro definition, in which
2683 case the parameter default is taken to be blank. This can be useful
2684 for macros which can take a variable number of parameters, since the
2685 \i\c{%0} token (see \k{percent0}) allows you to determine how many
2686 parameters were really passed to the macro call.
2688 This defaulting mechanism can be combined with the greedy-parameter
2689 mechanism; so the \c{die} macro above could be made more powerful,
2690 and more useful, by changing the first line of the definition to
2692 \c %macro die 0-1+ "Painful program death has occurred.",13,10
2694 The maximum parameter count can be infinite, denoted by \c{*}. In
2695 this case, of course, it is impossible to provide a \e{full} set of
2696 default parameters. Examples of this usage are shown in \k{rotate}.
2699 \S{percent0} \i\c{%0}: \I{counting macro parameters}Macro Parameter Counter
2701 The parameter reference \c{%0} will return a numeric constant giving the
2702 number of parameters received, that is, if \c{%0} is n then \c{%}n is the
2703 last parameter. \c{%0} is mostly useful for macros that can take a variable
2704 number of parameters. It can be used as an argument to \c{%rep}
2705 (see \k{rep}) in order to iterate through all the parameters of a macro.
2706 Examples are given in \k{rotate}.
2709 \S{percent00} \i\c{%00}: \I{label preceeding macro}Label Preceeding Macro
2711 \c{%00} will return the label preceeding the macro invocation, if any. The
2712 label must be on the same line as the macro invocation, may be a local label
2713 (see \k{locallab}), and need not end in a colon.
2716 \S{rotate} \i\c{%rotate}: \i{Rotating Macro Parameters}
2718 Unix shell programmers will be familiar with the \I{shift
2719 command}\c{shift} shell command, which allows the arguments passed
2720 to a shell script (referenced as \c{$1}, \c{$2} and so on) to be
2721 moved left by one place, so that the argument previously referenced
2722 as \c{$2} becomes available as \c{$1}, and the argument previously
2723 referenced as \c{$1} is no longer available at all.
2725 NASM provides a similar mechanism, in the form of \c{%rotate}. As
2726 its name suggests, it differs from the Unix \c{shift} in that no
2727 parameters are lost: parameters rotated off the left end of the
2728 argument list reappear on the right, and vice versa.
2730 \c{%rotate} is invoked with a single numeric argument (which may be
2731 an expression). The macro parameters are rotated to the left by that
2732 many places. If the argument to \c{%rotate} is negative, the macro
2733 parameters are rotated to the right.
2735 \I{iterating over macro parameters}So a pair of macros to save and
2736 restore a set of registers might work as follows:
2738 \c %macro multipush 1-*
2747 This macro invokes the \c{PUSH} instruction on each of its arguments
2748 in turn, from left to right. It begins by pushing its first
2749 argument, \c{%1}, then invokes \c{%rotate} to move all the arguments
2750 one place to the left, so that the original second argument is now
2751 available as \c{%1}. Repeating this procedure as many times as there
2752 were arguments (achieved by supplying \c{%0} as the argument to
2753 \c{%rep}) causes each argument in turn to be pushed.
2755 Note also the use of \c{*} as the maximum parameter count,
2756 indicating that there is no upper limit on the number of parameters
2757 you may supply to the \i\c{multipush} macro.
2759 It would be convenient, when using this macro, to have a \c{POP}
2760 equivalent, which \e{didn't} require the arguments to be given in
2761 reverse order. Ideally, you would write the \c{multipush} macro
2762 call, then cut-and-paste the line to where the pop needed to be
2763 done, and change the name of the called macro to \c{multipop}, and
2764 the macro would take care of popping the registers in the opposite
2765 order from the one in which they were pushed.
2767 This can be done by the following definition:
2769 \c %macro multipop 1-*
2778 This macro begins by rotating its arguments one place to the
2779 \e{right}, so that the original \e{last} argument appears as \c{%1}.
2780 This is then popped, and the arguments are rotated right again, so
2781 the second-to-last argument becomes \c{%1}. Thus the arguments are
2782 iterated through in reverse order.
2785 \S{concat} \i{Concatenating Macro Parameters}
2787 NASM can concatenate macro parameters and macro indirection constructs
2788 on to other text surrounding them. This allows you to declare a family
2789 of symbols, for example, in a macro definition. If, for example, you
2790 wanted to generate a table of key codes along with offsets into the
2791 table, you could code something like
2793 \c %macro keytab_entry 2
2795 \c keypos%1 equ $-keytab
2801 \c keytab_entry F1,128+1
2802 \c keytab_entry F2,128+2
2803 \c keytab_entry Return,13
2805 which would expand to
2808 \c keyposF1 equ $-keytab
2810 \c keyposF2 equ $-keytab
2812 \c keyposReturn equ $-keytab
2815 You can just as easily concatenate text on to the other end of a
2816 macro parameter, by writing \c{%1foo}.
2818 If you need to append a \e{digit} to a macro parameter, for example
2819 defining labels \c{foo1} and \c{foo2} when passed the parameter
2820 \c{foo}, you can't code \c{%11} because that would be taken as the
2821 eleventh macro parameter. Instead, you must code
2822 \I{braces, after % sign}\c{%\{1\}1}, which will separate the first
2823 \c{1} (giving the number of the macro parameter) from the second
2824 (literal text to be concatenated to the parameter).
2826 This concatenation can also be applied to other preprocessor in-line
2827 objects, such as macro-local labels (\k{maclocal}) and context-local
2828 labels (\k{ctxlocal}). In all cases, ambiguities in syntax can be
2829 resolved by enclosing everything after the \c{%} sign and before the
2830 literal text in braces: so \c{%\{%foo\}bar} concatenates the text
2831 \c{bar} to the end of the real name of the macro-local label
2832 \c{%%foo}. (This is unnecessary, since the form NASM uses for the
2833 real names of macro-local labels means that the two usages
2834 \c{%\{%foo\}bar} and \c{%%foobar} would both expand to the same
2835 thing anyway; nevertheless, the capability is there.)
2837 The single-line macro indirection construct, \c{%[...]}
2838 (\k{indmacro}), behaves the same way as macro parameters for the
2839 purpose of concatenation.
2841 See also the \c{%+} operator, \k{concat%+}.
2844 \S{mlmaccc} \i{Condition Codes as Macro Parameters}
2846 NASM can give special treatment to a macro parameter which contains
2847 a condition code. For a start, you can refer to the macro parameter
2848 \c{%1} by means of the alternative syntax \i\c{%+1}, which informs
2849 NASM that this macro parameter is supposed to contain a condition
2850 code, and will cause the preprocessor to report an error message if
2851 the macro is called with a parameter which is \e{not} a valid
2854 Far more usefully, though, you can refer to the macro parameter by
2855 means of \i\c{%-1}, which NASM will expand as the \e{inverse}
2856 condition code. So the \c{retz} macro defined in \k{maclocal} can be
2857 replaced by a general \i{conditional-return macro} like this:
2867 This macro can now be invoked using calls like \c{retc ne}, which
2868 will cause the conditional-jump instruction in the macro expansion
2869 to come out as \c{JE}, or \c{retc po} which will make the jump a
2872 The \c{%+1} macro-parameter reference is quite happy to interpret
2873 the arguments \c{CXZ} and \c{ECXZ} as valid condition codes;
2874 however, \c{%-1} will report an error if passed either of these,
2875 because no inverse condition code exists.
2878 \S{nolist} \i{Disabling Listing Expansion}\I\c{.nolist}
2880 When NASM is generating a listing file from your program, it will
2881 generally expand multi-line macros by means of writing the macro
2882 call and then listing each line of the expansion. This allows you to
2883 see which instructions in the macro expansion are generating what
2884 code; however, for some macros this clutters the listing up
2887 NASM therefore provides the \c{.nolist} qualifier, which you can
2888 include in a macro definition to inhibit the expansion of the macro
2889 in the listing file. The \c{.nolist} qualifier comes directly after
2890 the number of parameters, like this:
2892 \c %macro foo 1.nolist
2896 \c %macro bar 1-5+.nolist a,b,c,d,e,f,g,h
2898 \S{unmacro} Undefining Multi-Line Macros: \i\c{%unmacro}
2900 Multi-line macros can be removed with the \c{%unmacro} directive.
2901 Unlike the \c{%undef} directive, however, \c{%unmacro} takes an
2902 argument specification, and will only remove \i{exact matches} with
2903 that argument specification.
2912 removes the previously defined macro \c{foo}, but
2919 does \e{not} remove the macro \c{bar}, since the argument
2920 specification does not match exactly.
2923 \H{condasm} \i{Conditional Assembly}\I\c{%if}
2925 Similarly to the C preprocessor, NASM allows sections of a source
2926 file to be assembled only if certain conditions are met. The general
2927 syntax of this feature looks like this:
2930 \c ; some code which only appears if <condition> is met
2931 \c %elif<condition2>
2932 \c ; only appears if <condition> is not met but <condition2> is
2934 \c ; this appears if neither <condition> nor <condition2> was met
2937 The inverse forms \i\c{%ifn} and \i\c{%elifn} are also supported.
2939 The \i\c{%else} clause is optional, as is the \i\c{%elif} clause.
2940 You can have more than one \c{%elif} clause as well.
2942 There are a number of variants of the \c{%if} directive. Each has its
2943 corresponding \c{%elif}, \c{%ifn}, and \c{%elifn} directives; for
2944 example, the equivalents to the \c{%ifdef} directive are \c{%elifdef},
2945 \c{%ifndef}, and \c{%elifndef}.
2947 \S{ifdef} \i\c{%ifdef}: Testing Single-Line Macro Existence\I{testing,
2948 single-line macro existence}
2950 Beginning a conditional-assembly block with the line \c{%ifdef
2951 MACRO} will assemble the subsequent code if, and only if, a
2952 single-line macro called \c{MACRO} is defined. If not, then the
2953 \c{%elif} and \c{%else} blocks (if any) will be processed instead.
2955 For example, when debugging a program, you might want to write code
2958 \c ; perform some function
2960 \c writefile 2,"Function performed successfully",13,10
2962 \c ; go and do something else
2964 Then you could use the command-line option \c{-dDEBUG} to create a
2965 version of the program which produced debugging messages, and remove
2966 the option to generate the final release version of the program.
2968 You can test for a macro \e{not} being defined by using
2969 \i\c{%ifndef} instead of \c{%ifdef}. You can also test for macro
2970 definitions in \c{%elif} blocks by using \i\c{%elifdef} and
2974 \S{ifmacro} \i\c{%ifmacro}: Testing Multi-Line Macro
2975 Existence\I{testing, multi-line macro existence}
2977 The \c{%ifmacro} directive operates in the same way as the \c{%ifdef}
2978 directive, except that it checks for the existence of a multi-line macro.
2980 For example, you may be working with a large project and not have control
2981 over the macros in a library. You may want to create a macro with one
2982 name if it doesn't already exist, and another name if one with that name
2985 The \c{%ifmacro} is considered true if defining a macro with the given name
2986 and number of arguments would cause a definitions conflict. For example:
2988 \c %ifmacro MyMacro 1-3
2990 \c %error "MyMacro 1-3" causes a conflict with an existing macro.
2994 \c %macro MyMacro 1-3
2996 \c ; insert code to define the macro
3002 This will create the macro "MyMacro 1-3" if no macro already exists which
3003 would conflict with it, and emits a warning if there would be a definition
3006 You can test for the macro not existing by using the \i\c{%ifnmacro} instead
3007 of \c{%ifmacro}. Additional tests can be performed in \c{%elif} blocks by using
3008 \i\c{%elifmacro} and \i\c{%elifnmacro}.
3011 \S{ifctx} \i\c{%ifctx}: Testing the Context Stack\I{testing, context
3014 The conditional-assembly construct \c{%ifctx} will cause the
3015 subsequent code to be assembled if and only if the top context on
3016 the preprocessor's context stack has the same name as one of the arguments.
3017 As with \c{%ifdef}, the inverse and \c{%elif} forms \i\c{%ifnctx},
3018 \i\c{%elifctx} and \i\c{%elifnctx} are also supported.
3020 For more details of the context stack, see \k{ctxstack}. For a
3021 sample use of \c{%ifctx}, see \k{blockif}.
3024 \S{if} \i\c{%if}: Testing Arbitrary Numeric Expressions\I{testing,
3025 arbitrary numeric expressions}
3027 The conditional-assembly construct \c{%if expr} will cause the
3028 subsequent code to be assembled if and only if the value of the
3029 numeric expression \c{expr} is non-zero. An example of the use of
3030 this feature is in deciding when to break out of a \c{%rep}
3031 preprocessor loop: see \k{rep} for a detailed example.
3033 The expression given to \c{%if}, and its counterpart \i\c{%elif}, is
3034 a critical expression (see \k{crit}).
3036 \c{%if} extends the normal NASM expression syntax, by providing a
3037 set of \i{relational operators} which are not normally available in
3038 expressions. The operators \i\c{=}, \i\c{<}, \i\c{>}, \i\c{<=},
3039 \i\c{>=} and \i\c{<>} test equality, less-than, greater-than,
3040 less-or-equal, greater-or-equal and not-equal respectively. The
3041 C-like forms \i\c{==} and \i\c{!=} are supported as alternative
3042 forms of \c{=} and \c{<>}. In addition, low-priority logical
3043 operators \i\c{&&}, \i\c{^^} and \i\c{||} are provided, supplying
3044 \i{logical AND}, \i{logical XOR} and \i{logical OR}. These work like
3045 the C logical operators (although C has no logical XOR), in that
3046 they always return either 0 or 1, and treat any non-zero input as 1
3047 (so that \c{^^}, for example, returns 1 if exactly one of its inputs
3048 is zero, and 0 otherwise). The relational operators also return 1
3049 for true and 0 for false.
3051 Like other \c{%if} constructs, \c{%if} has a counterpart
3052 \i\c{%elif}, and negative forms \i\c{%ifn} and \i\c{%elifn}.
3054 \S{ifidn} \i\c{%ifidn} and \i\c{%ifidni}: Testing Exact Text
3055 Identity\I{testing, exact text identity}
3057 The construct \c{%ifidn text1,text2} will cause the subsequent code
3058 to be assembled if and only if \c{text1} and \c{text2}, after
3059 expanding single-line macros, are identical pieces of text.
3060 Differences in white space are not counted.
3062 \c{%ifidni} is similar to \c{%ifidn}, but is \i{case-insensitive}.
3064 For example, the following macro pushes a register or number on the
3065 stack, and allows you to treat \c{IP} as a real register:
3067 \c %macro pushparam 1
3078 Like other \c{%if} constructs, \c{%ifidn} has a counterpart
3079 \i\c{%elifidn}, and negative forms \i\c{%ifnidn} and \i\c{%elifnidn}.
3080 Similarly, \c{%ifidni} has counterparts \i\c{%elifidni},
3081 \i\c{%ifnidni} and \i\c{%elifnidni}.
3083 \S{iftyp} \i\c{%ifid}, \i\c{%ifnum}, \i\c{%ifstr}: Testing Token
3084 Types\I{testing, token types}
3086 Some macros will want to perform different tasks depending on
3087 whether they are passed a number, a string, or an identifier. For
3088 example, a string output macro might want to be able to cope with
3089 being passed either a string constant or a pointer to an existing
3092 The conditional assembly construct \c{%ifid}, taking one parameter
3093 (which may be blank), assembles the subsequent code if and only if
3094 the first token in the parameter exists and is an identifier.
3095 \c{%ifnum} works similarly, but tests for the token being a numeric
3096 constant; \c{%ifstr} tests for it being a string.
3098 For example, the \c{writefile} macro defined in \k{mlmacgre} can be
3099 extended to take advantage of \c{%ifstr} in the following fashion:
3101 \c %macro writefile 2-3+
3110 \c %%endstr: mov dx,%%str
3111 \c mov cx,%%endstr-%%str
3122 Then the \c{writefile} macro can cope with being called in either of
3123 the following two ways:
3125 \c writefile [file], strpointer, length
3126 \c writefile [file], "hello", 13, 10
3128 In the first, \c{strpointer} is used as the address of an
3129 already-declared string, and \c{length} is used as its length; in
3130 the second, a string is given to the macro, which therefore declares
3131 it itself and works out the address and length for itself.
3133 Note the use of \c{%if} inside the \c{%ifstr}: this is to detect
3134 whether the macro was passed two arguments (so the string would be a
3135 single string constant, and \c{db %2} would be adequate) or more (in
3136 which case, all but the first two would be lumped together into
3137 \c{%3}, and \c{db %2,%3} would be required).
3139 The usual \I\c{%elifid}\I\c{%elifnum}\I\c{%elifstr}\c{%elif}...,
3140 \I\c{%ifnid}\I\c{%ifnnum}\I\c{%ifnstr}\c{%ifn}..., and
3141 \I\c{%elifnid}\I\c{%elifnnum}\I\c{%elifnstr}\c{%elifn}... versions
3142 exist for each of \c{%ifid}, \c{%ifnum} and \c{%ifstr}.
3144 \S{iftoken} \i\c{%iftoken}: Test for a Single Token
3146 Some macros will want to do different things depending on if it is
3147 passed a single token (e.g. paste it to something else using \c{%+})
3148 versus a multi-token sequence.
3150 The conditional assembly construct \c{%iftoken} assembles the
3151 subsequent code if and only if the expanded parameters consist of
3152 exactly one token, possibly surrounded by whitespace.
3158 will assemble the subsequent code, but
3162 will not, since \c{-1} contains two tokens: the unary minus operator
3163 \c{-}, and the number \c{1}.
3165 The usual \i\c{%eliftoken}, \i\c\{%ifntoken}, and \i\c{%elifntoken}
3166 variants are also provided.
3168 \S{ifempty} \i\c{%ifempty}: Test for Empty Expansion
3170 The conditional assembly construct \c{%ifempty} assembles the
3171 subsequent code if and only if the expanded parameters do not contain
3172 any tokens at all, whitespace excepted.
3174 The usual \i\c{%elifempty}, \i\c\{%ifnempty}, and \i\c{%elifnempty}
3175 variants are also provided.
3177 \S{ifenv} \i\c{%ifenv}: Test If Environment Variable Exists
3179 The conditional assembly construct \c{%ifenv} assembles the
3180 subsequent code if and only if the environment variable referenced by
3181 the \c{%!<env>} directive exists.
3183 The usual \i\c{%elifenv}, \i\c\{%ifnenv}, and \i\c{%elifnenv}
3184 variants are also provided.
3186 Just as for \c{%!<env>} the argument should be written as a string if
3187 it contains characters that would not be legal in an identifier. See
3190 \H{rep} \i{Preprocessor Loops}\I{repeating code}: \i\c{%rep}
3192 NASM's \c{TIMES} prefix, though useful, cannot be used to invoke a
3193 multi-line macro multiple times, because it is processed by NASM
3194 after macros have already been expanded. Therefore NASM provides
3195 another form of loop, this time at the preprocessor level: \c{%rep}.
3197 The directives \c{%rep} and \i\c{%endrep} (\c{%rep} takes a numeric
3198 argument, which can be an expression; \c{%endrep} takes no
3199 arguments) can be used to enclose a chunk of code, which is then
3200 replicated as many times as specified by the preprocessor:
3204 \c inc word [table+2*i]
3208 This will generate a sequence of 64 \c{INC} instructions,
3209 incrementing every word of memory from \c{[table]} to
3212 For more complex termination conditions, or to break out of a repeat
3213 loop part way along, you can use the \i\c{%exitrep} directive to
3214 terminate the loop, like this:
3229 \c fib_number equ ($-fibonacci)/2
3231 This produces a list of all the Fibonacci numbers that will fit in
3232 16 bits. Note that a maximum repeat count must still be given to
3233 \c{%rep}. This is to prevent the possibility of NASM getting into an
3234 infinite loop in the preprocessor, which (on multitasking or
3235 multi-user systems) would typically cause all the system memory to
3236 be gradually used up and other applications to start crashing.
3238 Note a maximum repeat count is limited by 62 bit number, though it
3239 is hardly possible that you ever need anything bigger.
3242 \H{files} Source Files and Dependencies
3244 These commands allow you to split your sources into multiple files.
3246 \S{include} \i\c{%include}: \i{Including Other Files}
3248 Using, once again, a very similar syntax to the C preprocessor,
3249 NASM's preprocessor lets you include other source files into your
3250 code. This is done by the use of the \i\c{%include} directive:
3252 \c %include "macros.mac"
3254 will include the contents of the file \c{macros.mac} into the source
3255 file containing the \c{%include} directive.
3257 Include files are \I{searching for include files}searched for in the
3258 current directory (the directory you're in when you run NASM, as
3259 opposed to the location of the NASM executable or the location of
3260 the source file), plus any directories specified on the NASM command
3261 line using the \c{-i} option.
3263 The standard C idiom for preventing a file being included more than
3264 once is just as applicable in NASM: if the file \c{macros.mac} has
3267 \c %ifndef MACROS_MAC
3268 \c %define MACROS_MAC
3269 \c ; now define some macros
3272 then including the file more than once will not cause errors,
3273 because the second time the file is included nothing will happen
3274 because the macro \c{MACROS_MAC} will already be defined.
3276 You can force a file to be included even if there is no \c{%include}
3277 directive that explicitly includes it, by using the \i\c{-p} option
3278 on the NASM command line (see \k{opt-p}).
3281 \S{pathsearch} \i\c{%pathsearch}: Search the Include Path
3283 The \c{%pathsearch} directive takes a single-line macro name and a
3284 filename, and declare or redefines the specified single-line macro to
3285 be the include-path-resolved version of the filename, if the file
3286 exists (otherwise, it is passed unchanged.)
3290 \c %pathsearch MyFoo "foo.bin"
3292 ... with \c{-Ibins/} in the include path may end up defining the macro
3293 \c{MyFoo} to be \c{"bins/foo.bin"}.
3296 \S{depend} \i\c{%depend}: Add Dependent Files
3298 The \c{%depend} directive takes a filename and adds it to the list of
3299 files to be emitted as dependency generation when the \c{-M} options
3300 and its relatives (see \k{opt-M}) are used. It produces no output.
3302 This is generally used in conjunction with \c{%pathsearch}. For
3303 example, a simplified version of the standard macro wrapper for the
3304 \c{INCBIN} directive looks like:
3306 \c %imacro incbin 1-2+ 0
3307 \c %pathsearch dep %1
3312 This first resolves the location of the file into the macro \c{dep},
3313 then adds it to the dependency lists, and finally issues the
3314 assembler-level \c{INCBIN} directive.
3317 \S{use} \i\c{%use}: Include Standard Macro Package
3319 The \c{%use} directive is similar to \c{%include}, but rather than
3320 including the contents of a file, it includes a named standard macro
3321 package. The standard macro packages are part of NASM, and are
3322 described in \k{macropkg}.
3324 Unlike the \c{%include} directive, package names for the \c{%use}
3325 directive do not require quotes, but quotes are permitted. In NASM
3326 2.04 and 2.05 the unquoted form would be macro-expanded; this is no
3327 longer true. Thus, the following lines are equivalent:
3332 Standard macro packages are protected from multiple inclusion. When a
3333 standard macro package is used, a testable single-line macro of the
3334 form \c{__USE_}\e{package}\c{__} is also defined, see \k{use_def}.
3336 \H{ctxstack} The \i{Context Stack}
3338 Having labels that are local to a macro definition is sometimes not
3339 quite powerful enough: sometimes you want to be able to share labels
3340 between several macro calls. An example might be a \c{REPEAT} ...
3341 \c{UNTIL} loop, in which the expansion of the \c{REPEAT} macro
3342 would need to be able to refer to a label which the \c{UNTIL} macro
3343 had defined. However, for such a macro you would also want to be
3344 able to nest these loops.
3346 NASM provides this level of power by means of a \e{context stack}.
3347 The preprocessor maintains a stack of \e{contexts}, each of which is
3348 characterized by a name. You add a new context to the stack using
3349 the \i\c{%push} directive, and remove one using \i\c{%pop}. You can
3350 define labels that are local to a particular context on the stack.
3353 \S{pushpop} \i\c{%push} and \i\c{%pop}: \I{creating
3354 contexts}\I{removing contexts}Creating and Removing Contexts
3356 The \c{%push} directive is used to create a new context and place it
3357 on the top of the context stack. \c{%push} takes an optional argument,
3358 which is the name of the context. For example:
3362 This pushes a new context called \c{foobar} on the stack. You can have
3363 several contexts on the stack with the same name: they can still be
3364 distinguished. If no name is given, the context is unnamed (this is
3365 normally used when both the \c{%push} and the \c{%pop} are inside a
3366 single macro definition.)
3368 The directive \c{%pop}, taking one optional argument, removes the top
3369 context from the context stack and destroys it, along with any
3370 labels associated with it. If an argument is given, it must match the
3371 name of the current context, otherwise it will issue an error.
3374 \S{ctxlocal} \i{Context-Local Labels}
3376 Just as the usage \c{%%foo} defines a label which is local to the
3377 particular macro call in which it is used, the usage \I{%$}\c{%$foo}
3378 is used to define a label which is local to the context on the top
3379 of the context stack. So the \c{REPEAT} and \c{UNTIL} example given
3380 above could be implemented by means of:
3396 and invoked by means of, for example,
3404 which would scan every fourth byte of a string in search of the byte
3407 If you need to define, or access, labels local to the context
3408 \e{below} the top one on the stack, you can use \I{%$$}\c{%$$foo}, or
3409 \c{%$$$foo} for the context below that, and so on.
3412 \S{ctxdefine} \i{Context-Local Single-Line Macros}
3414 NASM also allows you to define single-line macros which are local to
3415 a particular context, in just the same way:
3417 \c %define %$localmac 3
3419 will define the single-line macro \c{%$localmac} to be local to the
3420 top context on the stack. Of course, after a subsequent \c{%push},
3421 it can then still be accessed by the name \c{%$$localmac}.
3424 \S{ctxfallthrough} \i{Context Fall-Through Lookup}
3426 Context fall-through lookup (automatic searching of outer contexts)
3427 is a feature that was added in NASM version 0.98.03. Unfortunately,
3428 this feature is unintuitive and can result in buggy code that would
3429 have otherwise been prevented by NASM's error reporting. As a result,
3430 this feature has been \e{deprecated}. NASM version 2.09 will issue a
3431 warning when usage of this \e{deprecated} feature is detected. Starting
3432 with NASM version 2.10, usage of this \e{deprecated} feature will simply
3433 result in an \e{expression syntax error}.
3435 An example usage of this \e{deprecated} feature follows:
3439 \c %assign %$external 1
3441 \c %assign %$internal 1
3442 \c mov eax, %$external
3443 \c mov eax, %$internal
3448 As demonstrated, \c{%$external} is being defined in the \c{ctx1}
3449 context and referenced within the \c{ctx2} context. With context
3450 fall-through lookup, referencing an undefined context-local macro
3451 like this implicitly searches through all outer contexts until a match
3452 is made or isn't found in any context. As a result, \c{%$external}
3453 referenced within the \c{ctx2} context would implicitly use \c{%$external}
3454 as defined in \c{ctx1}. Most people would expect NASM to issue an error in
3455 this situation because \c{%$external} was never defined within \c{ctx2} and also
3456 isn't qualified with the proper context depth, \c{%$$external}.
3458 Here is a revision of the above example with proper context depth:
3462 \c %assign %$external 1
3464 \c %assign %$internal 1
3465 \c mov eax, %$$external
3466 \c mov eax, %$internal
3471 As demonstrated, \c{%$external} is still being defined in the \c{ctx1}
3472 context and referenced within the \c{ctx2} context. However, the
3473 reference to \c{%$external} within \c{ctx2} has been fully qualified with
3474 the proper context depth, \c{%$$external}, and thus is no longer ambiguous,
3475 unintuitive or erroneous.
3478 \S{ctxrepl} \i\c{%repl}: \I{renaming contexts}Renaming a Context
3480 If you need to change the name of the top context on the stack (in
3481 order, for example, to have it respond differently to \c{%ifctx}),
3482 you can execute a \c{%pop} followed by a \c{%push}; but this will
3483 have the side effect of destroying all context-local labels and
3484 macros associated with the context that was just popped.
3486 NASM provides the directive \c{%repl}, which \e{replaces} a context
3487 with a different name, without touching the associated macros and
3488 labels. So you could replace the destructive code
3493 with the non-destructive version \c{%repl newname}.
3496 \S{blockif} Example Use of the \i{Context Stack}: \i{Block IFs}
3498 This example makes use of almost all the context-stack features,
3499 including the conditional-assembly construct \i\c{%ifctx}, to
3500 implement a block IF statement as a set of macros.
3516 \c %error "expected `if' before `else'"
3530 \c %error "expected `if' or `else' before `endif'"
3535 This code is more robust than the \c{REPEAT} and \c{UNTIL} macros
3536 given in \k{ctxlocal}, because it uses conditional assembly to check
3537 that the macros are issued in the right order (for example, not
3538 calling \c{endif} before \c{if}) and issues a \c{%error} if they're
3541 In addition, the \c{endif} macro has to be able to cope with the two
3542 distinct cases of either directly following an \c{if}, or following
3543 an \c{else}. It achieves this, again, by using conditional assembly
3544 to do different things depending on whether the context on top of
3545 the stack is \c{if} or \c{else}.
3547 The \c{else} macro has to preserve the context on the stack, in
3548 order to have the \c{%$ifnot} referred to by the \c{if} macro be the
3549 same as the one defined by the \c{endif} macro, but has to change
3550 the context's name so that \c{endif} will know there was an
3551 intervening \c{else}. It does this by the use of \c{%repl}.
3553 A sample usage of these macros might look like:
3575 The block-\c{IF} macros handle nesting quite happily, by means of
3576 pushing another context, describing the inner \c{if}, on top of the
3577 one describing the outer \c{if}; thus \c{else} and \c{endif} always
3578 refer to the last unmatched \c{if} or \c{else}.
3581 \H{stackrel} \i{Stack Relative Preprocessor Directives}
3583 The following preprocessor directives provide a way to use
3584 labels to refer to local variables allocated on the stack.
3586 \b\c{%arg} (see \k{arg})
3588 \b\c{%stacksize} (see \k{stacksize})
3590 \b\c{%local} (see \k{local})
3593 \S{arg} \i\c{%arg} Directive
3595 The \c{%arg} directive is used to simplify the handling of
3596 parameters passed on the stack. Stack based parameter passing
3597 is used by many high level languages, including C, C++ and Pascal.
3599 While NASM has macros which attempt to duplicate this
3600 functionality (see \k{16cmacro}), the syntax is not particularly
3601 convenient to use and is not TASM compatible. Here is an example
3602 which shows the use of \c{%arg} without any external macros:
3606 \c %push mycontext ; save the current context
3607 \c %stacksize large ; tell NASM to use bp
3608 \c %arg i:word, j_ptr:word
3615 \c %pop ; restore original context
3617 This is similar to the procedure defined in \k{16cmacro} and adds
3618 the value in i to the value pointed to by j_ptr and returns the
3619 sum in the ax register. See \k{pushpop} for an explanation of
3620 \c{push} and \c{pop} and the use of context stacks.
3623 \S{stacksize} \i\c{%stacksize} Directive
3625 The \c{%stacksize} directive is used in conjunction with the
3626 \c{%arg} (see \k{arg}) and the \c{%local} (see \k{local}) directives.
3627 It tells NASM the default size to use for subsequent \c{%arg} and
3628 \c{%local} directives. The \c{%stacksize} directive takes one
3629 required argument which is one of \c{flat}, \c{flat64}, \c{large} or \c{small}.
3633 This form causes NASM to use stack-based parameter addressing
3634 relative to \c{ebp} and it assumes that a near form of call was used
3635 to get to this label (i.e. that \c{eip} is on the stack).
3637 \c %stacksize flat64
3639 This form causes NASM to use stack-based parameter addressing
3640 relative to \c{rbp} and it assumes that a near form of call was used
3641 to get to this label (i.e. that \c{rip} is on the stack).
3645 This form uses \c{bp} to do stack-based parameter addressing and
3646 assumes that a far form of call was used to get to this address
3647 (i.e. that \c{ip} and \c{cs} are on the stack).
3651 This form also uses \c{bp} to address stack parameters, but it is
3652 different from \c{large} because it also assumes that the old value
3653 of bp is pushed onto the stack (i.e. it expects an \c{ENTER}
3654 instruction). In other words, it expects that \c{bp}, \c{ip} and
3655 \c{cs} are on the top of the stack, underneath any local space which
3656 may have been allocated by \c{ENTER}. This form is probably most
3657 useful when used in combination with the \c{%local} directive
3661 \S{local} \i\c{%local} Directive
3663 The \c{%local} directive is used to simplify the use of local
3664 temporary stack variables allocated in a stack frame. Automatic
3665 local variables in C are an example of this kind of variable. The
3666 \c{%local} directive is most useful when used with the \c{%stacksize}
3667 (see \k{stacksize} and is also compatible with the \c{%arg} directive
3668 (see \k{arg}). It allows simplified reference to variables on the
3669 stack which have been allocated typically by using the \c{ENTER}
3671 \# (see \k{insENTER} for a description of that instruction).
3672 An example of its use is the following:
3676 \c %push mycontext ; save the current context
3677 \c %stacksize small ; tell NASM to use bp
3678 \c %assign %$localsize 0 ; see text for explanation
3679 \c %local old_ax:word, old_dx:word
3681 \c enter %$localsize,0 ; see text for explanation
3682 \c mov [old_ax],ax ; swap ax & bx
3683 \c mov [old_dx],dx ; and swap dx & cx
3688 \c leave ; restore old bp
3691 \c %pop ; restore original context
3693 The \c{%$localsize} variable is used internally by the
3694 \c{%local} directive and \e{must} be defined within the
3695 current context before the \c{%local} directive may be used.
3696 Failure to do so will result in one expression syntax error for
3697 each \c{%local} variable declared. It then may be used in
3698 the construction of an appropriately sized ENTER instruction
3699 as shown in the example.
3702 \H{pperror} Reporting \i{User-Defined Errors}: \i\c{%error}, \i\c{%warning}, \i\c{%fatal}
3704 The preprocessor directive \c{%error} will cause NASM to report an
3705 error if it occurs in assembled code. So if other users are going to
3706 try to assemble your source files, you can ensure that they define the
3707 right macros by means of code like this:
3712 \c ; do some different setup
3714 \c %error "Neither F1 nor F2 was defined."
3717 Then any user who fails to understand the way your code is supposed
3718 to be assembled will be quickly warned of their mistake, rather than
3719 having to wait until the program crashes on being run and then not
3720 knowing what went wrong.
3722 Similarly, \c{%warning} issues a warning, but allows assembly to continue:
3727 \c ; do some different setup
3729 \c %warning "Neither F1 nor F2 was defined, assuming F1."
3733 \c{%error} and \c{%warning} are issued only on the final assembly
3734 pass. This makes them safe to use in conjunction with tests that
3735 depend on symbol values.
3737 \c{%fatal} terminates assembly immediately, regardless of pass. This
3738 is useful when there is no point in continuing the assembly further,
3739 and doing so is likely just going to cause a spew of confusing error
3742 It is optional for the message string after \c{%error}, \c{%warning}
3743 or \c{%fatal} to be quoted. If it is \e{not}, then single-line macros
3744 are expanded in it, which can be used to display more information to
3745 the user. For example:
3748 \c %assign foo_over foo-64
3749 \c %error foo is foo_over bytes too large
3753 \H{otherpreproc} \i{Other Preprocessor Directives}
3755 NASM also has preprocessor directives which allow access to
3756 information from external sources. Currently they include:
3758 \b\c{%line} enables NASM to correctly handle the output of another
3759 preprocessor (see \k{line}).
3761 \b\c{%!} enables NASM to read in the value of an environment variable,
3762 which can then be used in your program (see \k{getenv}).
3764 \S{line} \i\c{%line} Directive
3766 The \c{%line} directive is used to notify NASM that the input line
3767 corresponds to a specific line number in another file. Typically
3768 this other file would be an original source file, with the current
3769 NASM input being the output of a pre-processor. The \c{%line}
3770 directive allows NASM to output messages which indicate the line
3771 number of the original source file, instead of the file that is being
3774 This preprocessor directive is not generally of use to programmers,
3775 by may be of interest to preprocessor authors. The usage of the
3776 \c{%line} preprocessor directive is as follows:
3778 \c %line nnn[+mmm] [filename]
3780 In this directive, \c{nnn} identifies the line of the original source
3781 file which this line corresponds to. \c{mmm} is an optional parameter
3782 which specifies a line increment value; each line of the input file
3783 read in is considered to correspond to \c{mmm} lines of the original
3784 source file. Finally, \c{filename} is an optional parameter which
3785 specifies the file name of the original source file.
3787 After reading a \c{%line} preprocessor directive, NASM will report
3788 all file name and line numbers relative to the values specified
3792 \S{getenv} \i\c{%!}\c{<env>}: Read an environment variable.
3794 The \c{%!<env>} directive makes it possible to read the value of an
3795 environment variable at assembly time. This could, for example, be used
3796 to store the contents of an environment variable into a string, which
3797 could be used at some other point in your code.
3799 For example, suppose that you have an environment variable \c{FOO}, and
3800 you want the contents of \c{FOO} to be embedded in your program. You
3801 could do that as follows:
3803 \c %defstr FOO %!FOO
3805 See \k{defstr} for notes on the \c{%defstr} directive.
3807 If the name of the environment variable contains non-identifier
3808 characters, you can use string quotes to surround the name of the
3809 variable, for example:
3811 \c %defstr C_colon %!'C:'
3814 \H{comment} Comment Blocks: \i\c{%comment}
3816 The \c{%comment} and \c{%endcomment} directives are used to specify
3817 a block of commented (i.e. unprocessed) code/text. Everything between
3818 \c{%comment} and \c{%endcomment} will be ignored by the preprocessor.
3821 \c ; some code, text or data to be ignored
3825 \H{stdmac} \i{Standard Macros}
3827 NASM defines a set of standard macros, which are already defined
3828 when it starts to process any source file. If you really need a
3829 program to be assembled with no pre-defined macros, you can use the
3830 \i\c{%clear} directive to empty the preprocessor of everything but
3831 context-local preprocessor variables and single-line macros.
3833 Most \i{user-level assembler directives} (see \k{directive}) are
3834 implemented as macros which invoke primitive directives; these are
3835 described in \k{directive}. The rest of the standard macro set is
3839 \S{stdmacver} \i{NASM Version} Macros
3841 The single-line macros \i\c{__NASM_MAJOR__}, \i\c{__NASM_MINOR__},
3842 \i\c{__NASM_SUBMINOR__} and \i\c{___NASM_PATCHLEVEL__} expand to the
3843 major, minor, subminor and patch level parts of the \i{version
3844 number of NASM} being used. So, under NASM 0.98.32p1 for
3845 example, \c{__NASM_MAJOR__} would be defined to be 0, \c{__NASM_MINOR__}
3846 would be defined as 98, \c{__NASM_SUBMINOR__} would be defined to 32,
3847 and \c{___NASM_PATCHLEVEL__} would be defined as 1.
3849 Additionally, the macro \i\c{__NASM_SNAPSHOT__} is defined for
3850 automatically generated snapshot releases \e{only}.
3853 \S{stdmacverid} \i\c{__NASM_VERSION_ID__}: \i{NASM Version ID}
3855 The single-line macro \c{__NASM_VERSION_ID__} expands to a dword integer
3856 representing the full version number of the version of nasm being used.
3857 The value is the equivalent to \c{__NASM_MAJOR__}, \c{__NASM_MINOR__},
3858 \c{__NASM_SUBMINOR__} and \c{___NASM_PATCHLEVEL__} concatenated to
3859 produce a single doubleword. Hence, for 0.98.32p1, the returned number
3860 would be equivalent to:
3868 Note that the above lines are generate exactly the same code, the second
3869 line is used just to give an indication of the order that the separate
3870 values will be present in memory.
3873 \S{stdmacverstr} \i\c{__NASM_VER__}: \i{NASM Version string}
3875 The single-line macro \c{__NASM_VER__} expands to a string which defines
3876 the version number of nasm being used. So, under NASM 0.98.32 for example,
3885 \S{fileline} \i\c{__FILE__} and \i\c{__LINE__}: File Name and Line Number
3887 Like the C preprocessor, NASM allows the user to find out the file
3888 name and line number containing the current instruction. The macro
3889 \c{__FILE__} expands to a string constant giving the name of the
3890 current input file (which may change through the course of assembly
3891 if \c{%include} directives are used), and \c{__LINE__} expands to a
3892 numeric constant giving the current line number in the input file.
3894 These macros could be used, for example, to communicate debugging
3895 information to a macro, since invoking \c{__LINE__} inside a macro
3896 definition (either single-line or multi-line) will return the line
3897 number of the macro \e{call}, rather than \e{definition}. So to
3898 determine where in a piece of code a crash is occurring, for
3899 example, one could write a routine \c{stillhere}, which is passed a
3900 line number in \c{EAX} and outputs something like `line 155: still
3901 here'. You could then write a macro
3903 \c %macro notdeadyet 0
3912 and then pepper your code with calls to \c{notdeadyet} until you
3913 find the crash point.
3916 \S{bitsm} \i\c{__BITS__}: Current BITS Mode
3918 The \c{__BITS__} standard macro is updated every time that the BITS mode is
3919 set using the \c{BITS XX} or \c{[BITS XX]} directive, where XX is a valid mode
3920 number of 16, 32 or 64. \c{__BITS__} receives the specified mode number and
3921 makes it globally available. This can be very useful for those who utilize
3922 mode-dependent macros.
3924 \S{ofmtm} \i\c{__OUTPUT_FORMAT__}: Current Output Format
3926 The \c{__OUTPUT_FORMAT__} standard macro holds the current Output Format,
3927 as given by the \c{-f} option or NASM's default. Type \c{nasm -hf} for a
3930 \c %ifidn __OUTPUT_FORMAT__, win32
3931 \c %define NEWLINE 13, 10
3932 \c %elifidn __OUTPUT_FORMAT__, elf32
3933 \c %define NEWLINE 10
3937 \S{datetime} Assembly Date and Time Macros
3939 NASM provides a variety of macros that represent the timestamp of the
3942 \b The \i\c{__DATE__} and \i\c{__TIME__} macros give the assembly date and
3943 time as strings, in ISO 8601 format (\c{"YYYY-MM-DD"} and \c{"HH:MM:SS"},
3946 \b The \i\c{__DATE_NUM__} and \i\c{__TIME_NUM__} macros give the assembly
3947 date and time in numeric form; in the format \c{YYYYMMDD} and
3948 \c{HHMMSS} respectively.
3950 \b The \i\c{__UTC_DATE__} and \i\c{__UTC_TIME__} macros give the assembly
3951 date and time in universal time (UTC) as strings, in ISO 8601 format
3952 (\c{"YYYY-MM-DD"} and \c{"HH:MM:SS"}, respectively.) If the host
3953 platform doesn't provide UTC time, these macros are undefined.
3955 \b The \i\c{__UTC_DATE_NUM__} and \i\c{__UTC_TIME_NUM__} macros give the
3956 assembly date and time universal time (UTC) in numeric form; in the
3957 format \c{YYYYMMDD} and \c{HHMMSS} respectively. If the
3958 host platform doesn't provide UTC time, these macros are
3961 \b The \c{__POSIX_TIME__} macro is defined as a number containing the
3962 number of seconds since the POSIX epoch, 1 January 1970 00:00:00 UTC;
3963 excluding any leap seconds. This is computed using UTC time if
3964 available on the host platform, otherwise it is computed using the
3965 local time as if it was UTC.
3967 All instances of time and date macros in the same assembly session
3968 produce consistent output. For example, in an assembly session
3969 started at 42 seconds after midnight on January 1, 2010 in Moscow
3970 (timezone UTC+3) these macros would have the following values,
3971 assuming, of course, a properly configured environment with a correct
3974 \c __DATE__ "2010-01-01"
3975 \c __TIME__ "00:00:42"
3976 \c __DATE_NUM__ 20100101
3977 \c __TIME_NUM__ 000042
3978 \c __UTC_DATE__ "2009-12-31"
3979 \c __UTC_TIME__ "21:00:42"
3980 \c __UTC_DATE_NUM__ 20091231
3981 \c __UTC_TIME_NUM__ 210042
3982 \c __POSIX_TIME__ 1262293242
3985 \S{use_def} \I\c{__USE_*__}\c{__USE_}\e{package}\c{__}: Package
3988 When a standard macro package (see \k{macropkg}) is included with the
3989 \c{%use} directive (see \k{use}), a single-line macro of the form
3990 \c{__USE_}\e{package}\c{__} is automatically defined. This allows
3991 testing if a particular package is invoked or not.
3993 For example, if the \c{altreg} package is included (see
3994 \k{pkg_altreg}), then the macro \c{__USE_ALTREG__} is defined.
3997 \S{pass_macro} \i\c{__PASS__}: Assembly Pass
3999 The macro \c{__PASS__} is defined to be \c{1} on preparatory passes,
4000 and \c{2} on the final pass. In preprocess-only mode, it is set to
4001 \c{3}, and when running only to generate dependencies (due to the
4002 \c{-M} or \c{-MG} option, see \k{opt-M}) it is set to \c{0}.
4004 \e{Avoid using this macro if at all possible. It is tremendously easy
4005 to generate very strange errors by misusing it, and the semantics may
4006 change in future versions of NASM.}
4009 \S{struc} \i\c{STRUC} and \i\c{ENDSTRUC}: \i{Declaring Structure} Data Types
4011 The core of NASM contains no intrinsic means of defining data
4012 structures; instead, the preprocessor is sufficiently powerful that
4013 data structures can be implemented as a set of macros. The macros
4014 \c{STRUC} and \c{ENDSTRUC} are used to define a structure data type.
4016 \c{STRUC} takes one or two parameters. The first parameter is the name
4017 of the data type. The second, optional parameter is the base offset of
4018 the structure. The name of the data type is defined as a symbol with
4019 the value of the base offset, and the name of the data type with the
4020 suffix \c{_size} appended to it is defined as an \c{EQU} giving the
4021 size of the structure. Once \c{STRUC} has been issued, you are
4022 defining the structure, and should define fields using the \c{RESB}
4023 family of pseudo-instructions, and then invoke \c{ENDSTRUC} to finish
4026 For example, to define a structure called \c{mytype} containing a
4027 longword, a word, a byte and a string of bytes, you might code
4038 The above code defines six symbols: \c{mt_long} as 0 (the offset
4039 from the beginning of a \c{mytype} structure to the longword field),
4040 \c{mt_word} as 4, \c{mt_byte} as 6, \c{mt_str} as 7, \c{mytype_size}
4041 as 39, and \c{mytype} itself as zero.
4043 The reason why the structure type name is defined at zero by default
4044 is a side effect of allowing structures to work with the local label
4045 mechanism: if your structure members tend to have the same names in
4046 more than one structure, you can define the above structure like this:
4057 This defines the offsets to the structure fields as \c{mytype.long},
4058 \c{mytype.word}, \c{mytype.byte} and \c{mytype.str}.
4060 NASM, since it has no \e{intrinsic} structure support, does not
4061 support any form of period notation to refer to the elements of a
4062 structure once you have one (except the above local-label notation),
4063 so code such as \c{mov ax,[mystruc.mt_word]} is not valid.
4064 \c{mt_word} is a constant just like any other constant, so the
4065 correct syntax is \c{mov ax,[mystruc+mt_word]} or \c{mov
4066 ax,[mystruc+mytype.word]}.
4068 Sometimes you only have the address of the structure displaced by an
4069 offset. For example, consider this standard stack frame setup:
4075 In this case, you could access an element by subtracting the offset:
4077 \c mov [ebp - 40 + mytype.word], ax
4079 However, if you do not want to repeat this offset, you can use -40 as
4082 \c struc mytype, -40
4084 And access an element this way:
4086 \c mov [ebp + mytype.word], ax
4089 \S{istruc} \i\c{ISTRUC}, \i\c{AT} and \i\c{IEND}: Declaring
4090 \i{Instances of Structures}
4092 Having defined a structure type, the next thing you typically want
4093 to do is to declare instances of that structure in your data
4094 segment. NASM provides an easy way to do this in the \c{ISTRUC}
4095 mechanism. To declare a structure of type \c{mytype} in a program,
4096 you code something like this:
4101 \c at mt_long, dd 123456
4102 \c at mt_word, dw 1024
4103 \c at mt_byte, db 'x'
4104 \c at mt_str, db 'hello, world', 13, 10, 0
4108 The function of the \c{AT} macro is to make use of the \c{TIMES}
4109 prefix to advance the assembly position to the correct point for the
4110 specified structure field, and then to declare the specified data.
4111 Therefore the structure fields must be declared in the same order as
4112 they were specified in the structure definition.
4114 If the data to go in a structure field requires more than one source
4115 line to specify, the remaining source lines can easily come after
4116 the \c{AT} line. For example:
4118 \c at mt_str, db 123,134,145,156,167,178,189
4121 Depending on personal taste, you can also omit the code part of the
4122 \c{AT} line completely, and start the structure field on the next
4126 \c db 'hello, world'
4130 \S{align} \i\c{ALIGN} and \i\c{ALIGNB}: Data Alignment
4132 The \c{ALIGN} and \c{ALIGNB} macros provides a convenient way to
4133 align code or data on a word, longword, paragraph or other boundary.
4134 (Some assemblers call this directive \i\c{EVEN}.) The syntax of the
4135 \c{ALIGN} and \c{ALIGNB} macros is
4137 \c align 4 ; align on 4-byte boundary
4138 \c align 16 ; align on 16-byte boundary
4139 \c align 8,db 0 ; pad with 0s rather than NOPs
4140 \c align 4,resb 1 ; align to 4 in the BSS
4141 \c alignb 4 ; equivalent to previous line
4143 Both macros require their first argument to be a power of two; they
4144 both compute the number of additional bytes required to bring the
4145 length of the current section up to a multiple of that power of two,
4146 and then apply the \c{TIMES} prefix to their second argument to
4147 perform the alignment.
4149 If the second argument is not specified, the default for \c{ALIGN}
4150 is \c{NOP}, and the default for \c{ALIGNB} is \c{RESB 1}. So if the
4151 second argument is specified, the two macros are equivalent.
4152 Normally, you can just use \c{ALIGN} in code and data sections and
4153 \c{ALIGNB} in BSS sections, and never need the second argument
4154 except for special purposes.
4156 \c{ALIGN} and \c{ALIGNB}, being simple macros, perform no error
4157 checking: they cannot warn you if their first argument fails to be a
4158 power of two, or if their second argument generates more than one
4159 byte of code. In each of these cases they will silently do the wrong
4162 \c{ALIGNB} (or \c{ALIGN} with a second argument of \c{RESB 1}) can
4163 be used within structure definitions:
4180 This will ensure that the structure members are sensibly aligned
4181 relative to the base of the structure.
4183 A final caveat: \c{ALIGN} and \c{ALIGNB} work relative to the
4184 beginning of the \e{section}, not the beginning of the address space
4185 in the final executable. Aligning to a 16-byte boundary when the
4186 section you're in is only guaranteed to be aligned to a 4-byte
4187 boundary, for example, is a waste of effort. Again, NASM does not
4188 check that the section's alignment characteristics are sensible for
4189 the use of \c{ALIGN} or \c{ALIGNB}.
4191 Both \c{ALIGN} and \c{ALIGNB} do call \c{SECTALIGN} macro implicitly.
4192 See \k{sectalign} for details.
4194 See also the \c{smartalign} standard macro package, \k{pkg_smartalign}.
4197 \S{sectalign} \i\c{SECTALIGN}: Section Alignment
4199 The \c{SECTALIGN} macros provides a way to modify alignment attribute
4200 of output file section. Unlike the \c{align=} attribute (which is allowed
4201 at section definition only) the \c{SECTALIGN} macro may be used at any time.
4203 For example the directive
4207 sets the section alignment requirements to 16 bytes. Once increased it can
4208 not be decreased, the magnitude may grow only.
4210 Note that \c{ALIGN} (see \k{align}) calls the \c{SECTALIGN} macro implicitly
4211 so the active section alignment requirements may be updated. This is by default
4212 behaviour, if for some reason you want the \c{ALIGN} do not call \c{SECTALIGN}
4213 at all use the directive
4217 It is still possible to turn in on again by
4222 \C{macropkg} \i{Standard Macro Packages}
4224 The \i\c{%use} directive (see \k{use}) includes one of the standard
4225 macro packages included with the NASM distribution and compiled into
4226 the NASM binary. It operates like the \c{%include} directive (see
4227 \k{include}), but the included contents is provided by NASM itself.
4229 The names of standard macro packages are case insensitive, and can be
4233 \H{pkg_altreg} \i\c{altreg}: \i{Alternate Register Names}
4235 The \c{altreg} standard macro package provides alternate register
4236 names. It provides numeric register names for all registers (not just
4237 \c{R8}-\c{R15}), the Intel-defined aliases \c{R8L}-\c{R15L} for the
4238 low bytes of register (as opposed to the NASM/AMD standard names
4239 \c{R8B}-\c{R15B}), and the names \c{R0H}-\c{R3H} (by analogy with
4240 \c{R0L}-\c{R3L}) for \c{AH}, \c{CH}, \c{DH}, and \c{BH}.
4247 \c mov r0l,r3h ; mov al,bh
4253 \H{pkg_smartalign} \i\c{smartalign}\I{align, smart}: Smart \c{ALIGN} Macro
4255 The \c{smartalign} standard macro package provides for an \i\c{ALIGN}
4256 macro which is more powerful than the default (and
4257 backwards-compatible) one (see \k{align}). When the \c{smartalign}
4258 package is enabled, when \c{ALIGN} is used without a second argument,
4259 NASM will generate a sequence of instructions more efficient than a
4260 series of \c{NOP}. Furthermore, if the padding exceeds a specific
4261 threshold, then NASM will generate a jump over the entire padding
4264 The specific instructions generated can be controlled with the
4265 new \i\c{ALIGNMODE} macro. This macro takes two parameters: one mode,
4266 and an optional jump threshold override. If (for any reason) you need
4267 to turn off the jump completely just set jump threshold value to -1
4268 (or set it to \c{nojmp}). The following modes are possible:
4270 \b \c{generic}: Works on all x86 CPUs and should have reasonable
4271 performance. The default jump threshold is 8. This is the
4274 \b \c{nop}: Pad out with \c{NOP} instructions. The only difference
4275 compared to the standard \c{ALIGN} macro is that NASM can still jump
4276 over a large padding area. The default jump threshold is 16.
4278 \b \c{k7}: Optimize for the AMD K7 (Athlon/Althon XP). These
4279 instructions should still work on all x86 CPUs. The default jump
4282 \b \c{k8}: Optimize for the AMD K8 (Opteron/Althon 64). These
4283 instructions should still work on all x86 CPUs. The default jump
4286 \b \c{p6}: Optimize for Intel CPUs. This uses the long \c{NOP}
4287 instructions first introduced in Pentium Pro. This is incompatible
4288 with all CPUs of family 5 or lower, as well as some VIA CPUs and
4289 several virtualization solutions. The default jump threshold is 16.
4291 The macro \i\c{__ALIGNMODE__} is defined to contain the current
4292 alignment mode. A number of other macros beginning with \c{__ALIGN_}
4293 are used internally by this macro package.
4296 \H{pkg_fp} \i\c\{fp}: Floating-point macros
4298 This packages contains the following floating-point convenience macros:
4300 \c %define Inf __Infinity__
4301 \c %define NaN __QNaN__
4302 \c %define QNaN __QNaN__
4303 \c %define SNaN __SNaN__
4305 \c %define float8(x) __float8__(x)
4306 \c %define float16(x) __float16__(x)
4307 \c %define float32(x) __float32__(x)
4308 \c %define float64(x) __float64__(x)
4309 \c %define float80m(x) __float80m__(x)
4310 \c %define float80e(x) __float80e__(x)
4311 \c %define float128l(x) __float128l__(x)
4312 \c %define float128h(x) __float128h__(x)
4315 \H{pkg_ifunc} \i\c{ifunc}: \i{Integer functions}
4317 This package contains a set of macros which implement integer
4318 functions. These are actually implemented as special operators, but
4319 are most conveniently accessed via this macro package.
4321 The macros provided are:
4323 \S{ilog2} \i{Integer logarithms}
4325 These functions calculate the integer logarithm base 2 of their
4326 argument, considered as an unsigned integer. The only differences
4327 between the functions is their behavior if the argument provided is
4330 The function \i\c{ilog2e()} (alias \i\c{ilog2()}) generate an error if
4331 the argument is not a power of two.
4333 The function \i\c{ilog2w()} generate a warning if the argument is not
4336 The function \i\c{ilog2f()} rounds the argument down to the nearest
4337 power of two; if the argument is zero it returns zero.
4339 The function \i\c{ilog2c()} rounds the argument up to the nearest
4343 \C{directive} \i{Assembler Directives}
4345 NASM, though it attempts to avoid the bureaucracy of assemblers like
4346 MASM and TASM, is nevertheless forced to support a \e{few}
4347 directives. These are described in this chapter.
4349 NASM's directives come in two types: \I{user-level
4350 directives}\e{user-level} directives and \I{primitive
4351 directives}\e{primitive} directives. Typically, each directive has a
4352 user-level form and a primitive form. In almost all cases, we
4353 recommend that users use the user-level forms of the directives,
4354 which are implemented as macros which call the primitive forms.
4356 Primitive directives are enclosed in square brackets; user-level
4359 In addition to the universal directives described in this chapter,
4360 each object file format can optionally supply extra directives in
4361 order to control particular features of that file format. These
4362 \I{format-specific directives}\e{format-specific} directives are
4363 documented along with the formats that implement them, in \k{outfmt}.
4366 \H{bits} \i\c{BITS}: Specifying Target \i{Processor Mode}
4368 The \c{BITS} directive specifies whether NASM should generate code
4369 \I{16-bit mode, versus 32-bit mode}designed to run on a processor
4370 operating in 16-bit mode, 32-bit mode or 64-bit mode. The syntax is
4371 \c{BITS XX}, where XX is 16, 32 or 64.
4373 In most cases, you should not need to use \c{BITS} explicitly. The
4374 \c{aout}, \c{coff}, \c{elf}, \c{macho}, \c{win32} and \c{win64}
4375 object formats, which are designed for use in 32-bit or 64-bit
4376 operating systems, all cause NASM to select 32-bit or 64-bit mode,
4377 respectively, by default. The \c{obj} object format allows you
4378 to specify each segment you define as either \c{USE16} or \c{USE32},
4379 and NASM will set its operating mode accordingly, so the use of the
4380 \c{BITS} directive is once again unnecessary.
4382 The most likely reason for using the \c{BITS} directive is to write
4383 32-bit or 64-bit code in a flat binary file; this is because the \c{bin}
4384 output format defaults to 16-bit mode in anticipation of it being
4385 used most frequently to write DOS \c{.COM} programs, DOS \c{.SYS}
4386 device drivers and boot loader software.
4388 You do \e{not} need to specify \c{BITS 32} merely in order to use
4389 32-bit instructions in a 16-bit DOS program; if you do, the
4390 assembler will generate incorrect code because it will be writing
4391 code targeted at a 32-bit platform, to be run on a 16-bit one.
4393 When NASM is in \c{BITS 16} mode, instructions which use 32-bit
4394 data are prefixed with an 0x66 byte, and those referring to 32-bit
4395 addresses have an 0x67 prefix. In \c{BITS 32} mode, the reverse is
4396 true: 32-bit instructions require no prefixes, whereas instructions
4397 using 16-bit data need an 0x66 and those working on 16-bit addresses
4400 When NASM is in \c{BITS 64} mode, most instructions operate the same
4401 as they do for \c{BITS 32} mode. However, there are 8 more general and
4402 SSE registers, and 16-bit addressing is no longer supported.
4404 The default address size is 64 bits; 32-bit addressing can be selected
4405 with the 0x67 prefix. The default operand size is still 32 bits,
4406 however, and the 0x66 prefix selects 16-bit operand size. The \c{REX}
4407 prefix is used both to select 64-bit operand size, and to access the
4408 new registers. NASM automatically inserts REX prefixes when
4411 When the \c{REX} prefix is used, the processor does not know how to
4412 address the AH, BH, CH or DH (high 8-bit legacy) registers. Instead,
4413 it is possible to access the the low 8-bits of the SP, BP SI and DI
4414 registers as SPL, BPL, SIL and DIL, respectively; but only when the
4417 The \c{BITS} directive has an exactly equivalent primitive form,
4418 \c{[BITS 16]}, \c{[BITS 32]} and \c{[BITS 64]}. The user-level form is
4419 a macro which has no function other than to call the primitive form.
4421 Note that the space is neccessary, e.g. \c{BITS32} will \e{not} work!
4423 \S{USE16 & USE32} \i\c{USE16} & \i\c{USE32}: Aliases for BITS
4425 The `\c{USE16}' and `\c{USE32}' directives can be used in place of
4426 `\c{BITS 16}' and `\c{BITS 32}', for compatibility with other assemblers.
4429 \H{default} \i\c{DEFAULT}: Change the assembler defaults
4431 The \c{DEFAULT} directive changes the assembler defaults. Normally,
4432 NASM defaults to a mode where the programmer is expected to explicitly
4433 specify most features directly. However, this is occationally
4434 obnoxious, as the explicit form is pretty much the only one one wishes
4437 Currently, the only \c{DEFAULT} that is settable is whether or not
4438 registerless instructions in 64-bit mode are \c{RIP}-relative or not.
4439 By default, they are absolute unless overridden with the \i\c{REL}
4440 specifier (see \k{effaddr}). However, if \c{DEFAULT REL} is
4441 specified, \c{REL} is default, unless overridden with the \c{ABS}
4442 specifier, \e{except when used with an FS or GS segment override}.
4444 The special handling of \c{FS} and \c{GS} overrides are due to the
4445 fact that these registers are generally used as thread pointers or
4446 other special functions in 64-bit mode, and generating
4447 \c{RIP}-relative addresses would be extremely confusing.
4449 \c{DEFAULT REL} is disabled with \c{DEFAULT ABS}.
4451 \H{section} \i\c{SECTION} or \i\c{SEGMENT}: Changing and \i{Defining
4454 \I{changing sections}\I{switching between sections}The \c{SECTION}
4455 directive (\c{SEGMENT} is an exactly equivalent synonym) changes
4456 which section of the output file the code you write will be
4457 assembled into. In some object file formats, the number and names of
4458 sections are fixed; in others, the user may make up as many as they
4459 wish. Hence \c{SECTION} may sometimes give an error message, or may
4460 define a new section, if you try to switch to a section that does
4463 The Unix object formats, and the \c{bin} object format (but see
4464 \k{multisec}, all support
4465 the \i{standardized section names} \c{.text}, \c{.data} and \c{.bss}
4466 for the code, data and uninitialized-data sections. The \c{obj}
4467 format, by contrast, does not recognize these section names as being
4468 special, and indeed will strip off the leading period of any section
4472 \S{sectmac} The \i\c{__SECT__} Macro
4474 The \c{SECTION} directive is unusual in that its user-level form
4475 functions differently from its primitive form. The primitive form,
4476 \c{[SECTION xyz]}, simply switches the current target section to the
4477 one given. The user-level form, \c{SECTION xyz}, however, first
4478 defines the single-line macro \c{__SECT__} to be the primitive
4479 \c{[SECTION]} directive which it is about to issue, and then issues
4480 it. So the user-level directive
4484 expands to the two lines
4486 \c %define __SECT__ [SECTION .text]
4489 Users may find it useful to make use of this in their own macros.
4490 For example, the \c{writefile} macro defined in \k{mlmacgre} can be
4491 usefully rewritten in the following more sophisticated form:
4493 \c %macro writefile 2+
4503 \c mov cx,%%endstr-%%str
4510 This form of the macro, once passed a string to output, first
4511 switches temporarily to the data section of the file, using the
4512 primitive form of the \c{SECTION} directive so as not to modify
4513 \c{__SECT__}. It then declares its string in the data section, and
4514 then invokes \c{__SECT__} to switch back to \e{whichever} section
4515 the user was previously working in. It thus avoids the need, in the
4516 previous version of the macro, to include a \c{JMP} instruction to
4517 jump over the data, and also does not fail if, in a complicated
4518 \c{OBJ} format module, the user could potentially be assembling the
4519 code in any of several separate code sections.
4522 \H{absolute} \i\c{ABSOLUTE}: Defining Absolute Labels
4524 The \c{ABSOLUTE} directive can be thought of as an alternative form
4525 of \c{SECTION}: it causes the subsequent code to be directed at no
4526 physical section, but at the hypothetical section starting at the
4527 given absolute address. The only instructions you can use in this
4528 mode are the \c{RESB} family.
4530 \c{ABSOLUTE} is used as follows:
4538 This example describes a section of the PC BIOS data area, at
4539 segment address 0x40: the above code defines \c{kbuf_chr} to be
4540 0x1A, \c{kbuf_free} to be 0x1C, and \c{kbuf} to be 0x1E.
4542 The user-level form of \c{ABSOLUTE}, like that of \c{SECTION},
4543 redefines the \i\c{__SECT__} macro when it is invoked.
4545 \i\c{STRUC} and \i\c{ENDSTRUC} are defined as macros which use
4546 \c{ABSOLUTE} (and also \c{__SECT__}).
4548 \c{ABSOLUTE} doesn't have to take an absolute constant as an
4549 argument: it can take an expression (actually, a \i{critical
4550 expression}: see \k{crit}) and it can be a value in a segment. For
4551 example, a TSR can re-use its setup code as run-time BSS like this:
4553 \c org 100h ; it's a .COM program
4555 \c jmp setup ; setup code comes last
4557 \c ; the resident part of the TSR goes here
4559 \c ; now write the code that installs the TSR here
4563 \c runtimevar1 resw 1
4564 \c runtimevar2 resd 20
4568 This defines some variables `on top of' the setup code, so that
4569 after the setup has finished running, the space it took up can be
4570 re-used as data storage for the running TSR. The symbol `tsr_end'
4571 can be used to calculate the total size of the part of the TSR that
4572 needs to be made resident.
4575 \H{extern} \i\c{EXTERN}: \i{Importing Symbols} from Other Modules
4577 \c{EXTERN} is similar to the MASM directive \c{EXTRN} and the C
4578 keyword \c{extern}: it is used to declare a symbol which is not
4579 defined anywhere in the module being assembled, but is assumed to be
4580 defined in some other module and needs to be referred to by this
4581 one. Not every object-file format can support external variables:
4582 the \c{bin} format cannot.
4584 The \c{EXTERN} directive takes as many arguments as you like. Each
4585 argument is the name of a symbol:
4588 \c extern _sscanf,_fscanf
4590 Some object-file formats provide extra features to the \c{EXTERN}
4591 directive. In all cases, the extra features are used by suffixing a
4592 colon to the symbol name followed by object-format specific text.
4593 For example, the \c{obj} format allows you to declare that the
4594 default segment base of an external should be the group \c{dgroup}
4595 by means of the directive
4597 \c extern _variable:wrt dgroup
4599 The primitive form of \c{EXTERN} differs from the user-level form
4600 only in that it can take only one argument at a time: the support
4601 for multiple arguments is implemented at the preprocessor level.
4603 You can declare the same variable as \c{EXTERN} more than once: NASM
4604 will quietly ignore the second and later redeclarations. You can't
4605 declare a variable as \c{EXTERN} as well as something else, though.
4608 \H{global} \i\c{GLOBAL}: \i{Exporting Symbols} to Other Modules
4610 \c{GLOBAL} is the other end of \c{EXTERN}: if one module declares a
4611 symbol as \c{EXTERN} and refers to it, then in order to prevent
4612 linker errors, some other module must actually \e{define} the
4613 symbol and declare it as \c{GLOBAL}. Some assemblers use the name
4614 \i\c{PUBLIC} for this purpose.
4616 The \c{GLOBAL} directive applying to a symbol must appear \e{before}
4617 the definition of the symbol.
4619 \c{GLOBAL} uses the same syntax as \c{EXTERN}, except that it must
4620 refer to symbols which \e{are} defined in the same module as the
4621 \c{GLOBAL} directive. For example:
4627 \c{GLOBAL}, like \c{EXTERN}, allows object formats to define private
4628 extensions by means of a colon. The \c{elf} object format, for
4629 example, lets you specify whether global data items are functions or
4632 \c global hashlookup:function, hashtable:data
4634 Like \c{EXTERN}, the primitive form of \c{GLOBAL} differs from the
4635 user-level form only in that it can take only one argument at a
4639 \H{common} \i\c{COMMON}: Defining Common Data Areas
4641 The \c{COMMON} directive is used to declare \i\e{common variables}.
4642 A common variable is much like a global variable declared in the
4643 uninitialized data section, so that
4647 is similar in function to
4654 The difference is that if more than one module defines the same
4655 common variable, then at link time those variables will be
4656 \e{merged}, and references to \c{intvar} in all modules will point
4657 at the same piece of memory.
4659 Like \c{GLOBAL} and \c{EXTERN}, \c{COMMON} supports object-format
4660 specific extensions. For example, the \c{obj} format allows common
4661 variables to be NEAR or FAR, and the \c{elf} format allows you to
4662 specify the alignment requirements of a common variable:
4664 \c common commvar 4:near ; works in OBJ
4665 \c common intarray 100:4 ; works in ELF: 4 byte aligned
4667 Once again, like \c{EXTERN} and \c{GLOBAL}, the primitive form of
4668 \c{COMMON} differs from the user-level form only in that it can take
4669 only one argument at a time.
4672 \H{CPU} \i\c{CPU}: Defining CPU Dependencies
4674 The \i\c{CPU} directive restricts assembly to those instructions which
4675 are available on the specified CPU.
4679 \b\c{CPU 8086} Assemble only 8086 instruction set
4681 \b\c{CPU 186} Assemble instructions up to the 80186 instruction set
4683 \b\c{CPU 286} Assemble instructions up to the 286 instruction set
4685 \b\c{CPU 386} Assemble instructions up to the 386 instruction set
4687 \b\c{CPU 486} 486 instruction set
4689 \b\c{CPU 586} Pentium instruction set
4691 \b\c{CPU PENTIUM} Same as 586
4693 \b\c{CPU 686} P6 instruction set
4695 \b\c{CPU PPRO} Same as 686
4697 \b\c{CPU P2} Same as 686
4699 \b\c{CPU P3} Pentium III (Katmai) instruction sets
4701 \b\c{CPU KATMAI} Same as P3
4703 \b\c{CPU P4} Pentium 4 (Willamette) instruction set
4705 \b\c{CPU WILLAMETTE} Same as P4
4707 \b\c{CPU PRESCOTT} Prescott instruction set
4709 \b\c{CPU X64} x86-64 (x64/AMD64/Intel 64) instruction set
4711 \b\c{CPU IA64} IA64 CPU (in x86 mode) instruction set
4713 All options are case insensitive. All instructions will be selected
4714 only if they apply to the selected CPU or lower. By default, all
4715 instructions are available.
4718 \H{FLOAT} \i\c{FLOAT}: Handling of \I{floating-point, constants}floating-point constants
4720 By default, floating-point constants are rounded to nearest, and IEEE
4721 denormals are supported. The following options can be set to alter
4724 \b\c{FLOAT DAZ} Flush denormals to zero
4726 \b\c{FLOAT NODAZ} Do not flush denormals to zero (default)
4728 \b\c{FLOAT NEAR} Round to nearest (default)
4730 \b\c{FLOAT UP} Round up (toward +Infinity)
4732 \b\c{FLOAT DOWN} Round down (toward -Infinity)
4734 \b\c{FLOAT ZERO} Round toward zero
4736 \b\c{FLOAT DEFAULT} Restore default settings
4738 The standard macros \i\c{__FLOAT_DAZ__}, \i\c{__FLOAT_ROUND__}, and
4739 \i\c{__FLOAT__} contain the current state, as long as the programmer
4740 has avoided the use of the brackeded primitive form, (\c{[FLOAT]}).
4742 \c{__FLOAT__} contains the full set of floating-point settings; this
4743 value can be saved away and invoked later to restore the setting.
4746 \C{outfmt} \i{Output Formats}
4748 NASM is a portable assembler, designed to be able to compile on any
4749 ANSI C-supporting platform and produce output to run on a variety of
4750 Intel x86 operating systems. For this reason, it has a large number
4751 of available output formats, selected using the \i\c{-f} option on
4752 the NASM \i{command line}. Each of these formats, along with its
4753 extensions to the base NASM syntax, is detailed in this chapter.
4755 As stated in \k{opt-o}, NASM chooses a \i{default name} for your
4756 output file based on the input file name and the chosen output
4757 format. This will be generated by removing the \i{extension}
4758 (\c{.asm}, \c{.s}, or whatever you like to use) from the input file
4759 name, and substituting an extension defined by the output format.
4760 The extensions are given with each format below.
4763 \H{binfmt} \i\c{bin}: \i{Flat-Form Binary}\I{pure binary} Output
4765 The \c{bin} format does not produce object files: it generates
4766 nothing in the output file except the code you wrote. Such `pure
4767 binary' files are used by \i{MS-DOS}: \i\c{.COM} executables and
4768 \i\c{.SYS} device drivers are pure binary files. Pure binary output
4769 is also useful for \i{operating system} and \i{boot loader}
4772 The \c{bin} format supports \i{multiple section names}. For details of
4773 how NASM handles sections in the \c{bin} format, see \k{multisec}.
4775 Using the \c{bin} format puts NASM by default into 16-bit mode (see
4776 \k{bits}). In order to use \c{bin} to write 32-bit or 64-bit code,
4777 such as an OS kernel, you need to explicitly issue the \I\c{BITS}\c{BITS 32}
4778 or \I\c{BITS}\c{BITS 64} directive.
4780 \c{bin} has no default output file name extension: instead, it
4781 leaves your file name as it is once the original extension has been
4782 removed. Thus, the default is for NASM to assemble \c{binprog.asm}
4783 into a binary file called \c{binprog}.
4786 \S{org} \i\c{ORG}: Binary File \i{Program Origin}
4788 The \c{bin} format provides an additional directive to the list
4789 given in \k{directive}: \c{ORG}. The function of the \c{ORG}
4790 directive is to specify the origin address which NASM will assume
4791 the program begins at when it is loaded into memory.
4793 For example, the following code will generate the longword
4800 Unlike the \c{ORG} directive provided by MASM-compatible assemblers,
4801 which allows you to jump around in the object file and overwrite
4802 code you have already generated, NASM's \c{ORG} does exactly what
4803 the directive says: \e{origin}. Its sole function is to specify one
4804 offset which is added to all internal address references within the
4805 section; it does not permit any of the trickery that MASM's version
4806 does. See \k{proborg} for further comments.
4809 \S{binseg} \c{bin} Extensions to the \c{SECTION}
4810 Directive\I{SECTION, bin extensions to}
4812 The \c{bin} output format extends the \c{SECTION} (or \c{SEGMENT})
4813 directive to allow you to specify the alignment requirements of
4814 segments. This is done by appending the \i\c{ALIGN} qualifier to the
4815 end of the section-definition line. For example,
4817 \c section .data align=16
4819 switches to the section \c{.data} and also specifies that it must be
4820 aligned on a 16-byte boundary.
4822 The parameter to \c{ALIGN} specifies how many low bits of the
4823 section start address must be forced to zero. The alignment value
4824 given may be any power of two.\I{section alignment, in
4825 bin}\I{segment alignment, in bin}\I{alignment, in bin sections}
4828 \S{multisec} \i{Multisection}\I{bin, multisection} Support for the \c{bin} Format
4830 The \c{bin} format allows the use of multiple sections, of arbitrary names,
4831 besides the "known" \c{.text}, \c{.data}, and \c{.bss} names.
4833 \b Sections may be designated \i\c{progbits} or \i\c{nobits}. Default
4834 is \c{progbits} (except \c{.bss}, which defaults to \c{nobits},
4837 \b Sections can be aligned at a specified boundary following the previous
4838 section with \c{align=}, or at an arbitrary byte-granular position with
4841 \b Sections can be given a virtual start address, which will be used
4842 for the calculation of all memory references within that section
4845 \b Sections can be ordered using \i\c{follows=}\c{<section>} or
4846 \i\c{vfollows=}\c{<section>} as an alternative to specifying an explicit
4849 \b Arguments to \c{org}, \c{start}, \c{vstart}, and \c{align=} are
4850 critical expressions. See \k{crit}. E.g. \c{align=(1 << ALIGN_SHIFT)}
4851 - \c{ALIGN_SHIFT} must be defined before it is used here.
4853 \b Any code which comes before an explicit \c{SECTION} directive
4854 is directed by default into the \c{.text} section.
4856 \b If an \c{ORG} statement is not given, \c{ORG 0} is used
4859 \b The \c{.bss} section will be placed after the last \c{progbits}
4860 section, unless \c{start=}, \c{vstart=}, \c{follows=}, or \c{vfollows=}
4863 \b All sections are aligned on dword boundaries, unless a different
4864 alignment has been specified.
4866 \b Sections may not overlap.
4868 \b NASM creates the \c{section.<secname>.start} for each section,
4869 which may be used in your code.
4871 \S{map}\i{Map Files}
4873 Map files can be generated in \c{-f bin} format by means of the \c{[map]}
4874 option. Map types of \c{all} (default), \c{brief}, \c{sections}, \c{segments},
4875 or \c{symbols} may be specified. Output may be directed to \c{stdout}
4876 (default), \c{stderr}, or a specified file. E.g.
4877 \c{[map symbols myfile.map]}. No "user form" exists, the square
4878 brackets must be used.
4881 \H{ithfmt} \i\c{ith}: \i{Intel Hex} Output
4883 The \c{ith} file format produces Intel hex-format files. Just as the
4884 \c{bin} format, this is a flat memory image format with no support for
4885 relocation or linking. It is usually used with ROM programmers and
4888 All extensions supported by the \c{bin} file format is also supported by
4889 the \c{ith} file format.
4891 \c{ith} provides a default output file-name extension of \c{.ith}.
4894 \H{srecfmt} \i\c{srec}: \i{Motorola S-Records} Output
4896 The \c{srec} file format produces Motorola S-records files. Just as the
4897 \c{bin} format, this is a flat memory image format with no support for
4898 relocation or linking. It is usually used with ROM programmers and
4901 All extensions supported by the \c{bin} file format is also supported by
4902 the \c{srec} file format.
4904 \c{srec} provides a default output file-name extension of \c{.srec}.
4907 \H{objfmt} \i\c{obj}: \i{Microsoft OMF}\I{OMF} Object Files
4909 The \c{obj} file format (NASM calls it \c{obj} rather than \c{omf}
4910 for historical reasons) is the one produced by \i{MASM} and
4911 \i{TASM}, which is typically fed to 16-bit DOS linkers to produce
4912 \i\c{.EXE} files. It is also the format used by \i{OS/2}.
4914 \c{obj} provides a default output file-name extension of \c{.obj}.
4916 \c{obj} is not exclusively a 16-bit format, though: NASM has full
4917 support for the 32-bit extensions to the format. In particular,
4918 32-bit \c{obj} format files are used by \i{Borland's Win32
4919 compilers}, instead of using Microsoft's newer \i\c{win32} object
4922 The \c{obj} format does not define any special segment names: you
4923 can call your segments anything you like. Typical names for segments
4924 in \c{obj} format files are \c{CODE}, \c{DATA} and \c{BSS}.
4926 If your source file contains code before specifying an explicit
4927 \c{SEGMENT} directive, then NASM will invent its own segment called
4928 \i\c{__NASMDEFSEG} for you.
4930 When you define a segment in an \c{obj} file, NASM defines the
4931 segment name as a symbol as well, so that you can access the segment
4932 address of the segment. So, for example:
4941 \c mov ax,data ; get segment address of data
4942 \c mov ds,ax ; and move it into DS
4943 \c inc word [dvar] ; now this reference will work
4946 The \c{obj} format also enables the use of the \i\c{SEG} and
4947 \i\c{WRT} operators, so that you can write code which does things
4952 \c mov ax,seg foo ; get preferred segment of foo
4954 \c mov ax,data ; a different segment
4956 \c mov ax,[ds:foo] ; this accesses `foo'
4957 \c mov [es:foo wrt data],bx ; so does this
4960 \S{objseg} \c{obj} Extensions to the \c{SEGMENT}
4961 Directive\I{SEGMENT, obj extensions to}
4963 The \c{obj} output format extends the \c{SEGMENT} (or \c{SECTION})
4964 directive to allow you to specify various properties of the segment
4965 you are defining. This is done by appending extra qualifiers to the
4966 end of the segment-definition line. For example,
4968 \c segment code private align=16
4970 defines the segment \c{code}, but also declares it to be a private
4971 segment, and requires that the portion of it described in this code
4972 module must be aligned on a 16-byte boundary.
4974 The available qualifiers are:
4976 \b \i\c{PRIVATE}, \i\c{PUBLIC}, \i\c{COMMON} and \i\c{STACK} specify
4977 the combination characteristics of the segment. \c{PRIVATE} segments
4978 do not get combined with any others by the linker; \c{PUBLIC} and
4979 \c{STACK} segments get concatenated together at link time; and
4980 \c{COMMON} segments all get overlaid on top of each other rather
4981 than stuck end-to-end.
4983 \b \i\c{ALIGN} is used, as shown above, to specify how many low bits
4984 of the segment start address must be forced to zero. The alignment
4985 value given may be any power of two from 1 to 4096; in reality, the
4986 only values supported are 1, 2, 4, 16, 256 and 4096, so if 8 is
4987 specified it will be rounded up to 16, and 32, 64 and 128 will all
4988 be rounded up to 256, and so on. Note that alignment to 4096-byte
4989 boundaries is a \i{PharLap} extension to the format and may not be
4990 supported by all linkers.\I{section alignment, in OBJ}\I{segment
4991 alignment, in OBJ}\I{alignment, in OBJ sections}
4993 \b \i\c{CLASS} can be used to specify the segment class; this feature
4994 indicates to the linker that segments of the same class should be
4995 placed near each other in the output file. The class name can be any
4996 word, e.g. \c{CLASS=CODE}.
4998 \b \i\c{OVERLAY}, like \c{CLASS}, is specified with an arbitrary word
4999 as an argument, and provides overlay information to an
5000 overlay-capable linker.
5002 \b Segments can be declared as \i\c{USE16} or \i\c{USE32}, which has
5003 the effect of recording the choice in the object file and also
5004 ensuring that NASM's default assembly mode when assembling in that
5005 segment is 16-bit or 32-bit respectively.
5007 \b When writing \i{OS/2} object files, you should declare 32-bit
5008 segments as \i\c{FLAT}, which causes the default segment base for
5009 anything in the segment to be the special group \c{FLAT}, and also
5010 defines the group if it is not already defined.
5012 \b The \c{obj} file format also allows segments to be declared as
5013 having a pre-defined absolute segment address, although no linkers
5014 are currently known to make sensible use of this feature;
5015 nevertheless, NASM allows you to declare a segment such as
5016 \c{SEGMENT SCREEN ABSOLUTE=0xB800} if you need to. The \i\c{ABSOLUTE}
5017 and \c{ALIGN} keywords are mutually exclusive.
5019 NASM's default segment attributes are \c{PUBLIC}, \c{ALIGN=1}, no
5020 class, no overlay, and \c{USE16}.
5023 \S{group} \i\c{GROUP}: Defining Groups of Segments\I{segments, groups of}
5025 The \c{obj} format also allows segments to be grouped, so that a
5026 single segment register can be used to refer to all the segments in
5027 a group. NASM therefore supplies the \c{GROUP} directive, whereby
5036 \c ; some uninitialized data
5038 \c group dgroup data bss
5040 which will define a group called \c{dgroup} to contain the segments
5041 \c{data} and \c{bss}. Like \c{SEGMENT}, \c{GROUP} causes the group
5042 name to be defined as a symbol, so that you can refer to a variable
5043 \c{var} in the \c{data} segment as \c{var wrt data} or as \c{var wrt
5044 dgroup}, depending on which segment value is currently in your
5047 If you just refer to \c{var}, however, and \c{var} is declared in a
5048 segment which is part of a group, then NASM will default to giving
5049 you the offset of \c{var} from the beginning of the \e{group}, not
5050 the \e{segment}. Therefore \c{SEG var}, also, will return the group
5051 base rather than the segment base.
5053 NASM will allow a segment to be part of more than one group, but
5054 will generate a warning if you do this. Variables declared in a
5055 segment which is part of more than one group will default to being
5056 relative to the first group that was defined to contain the segment.
5058 A group does not have to contain any segments; you can still make
5059 \c{WRT} references to a group which does not contain the variable
5060 you are referring to. OS/2, for example, defines the special group
5061 \c{FLAT} with no segments in it.
5064 \S{uppercase} \i\c{UPPERCASE}: Disabling Case Sensitivity in Output
5066 Although NASM itself is \i{case sensitive}, some OMF linkers are
5067 not; therefore it can be useful for NASM to output single-case
5068 object files. The \c{UPPERCASE} format-specific directive causes all
5069 segment, group and symbol names that are written to the object file
5070 to be forced to upper case just before being written. Within a
5071 source file, NASM is still case-sensitive; but the object file can
5072 be written entirely in upper case if desired.
5074 \c{UPPERCASE} is used alone on a line; it requires no parameters.
5077 \S{import} \i\c{IMPORT}: Importing DLL Symbols\I{DLL symbols,
5078 importing}\I{symbols, importing from DLLs}
5080 The \c{IMPORT} format-specific directive defines a symbol to be
5081 imported from a DLL, for use if you are writing a DLL's \i{import
5082 library} in NASM. You still need to declare the symbol as \c{EXTERN}
5083 as well as using the \c{IMPORT} directive.
5085 The \c{IMPORT} directive takes two required parameters, separated by
5086 white space, which are (respectively) the name of the symbol you
5087 wish to import and the name of the library you wish to import it
5090 \c import WSAStartup wsock32.dll
5092 A third optional parameter gives the name by which the symbol is
5093 known in the library you are importing it from, in case this is not
5094 the same as the name you wish the symbol to be known by to your code
5095 once you have imported it. For example:
5097 \c import asyncsel wsock32.dll WSAAsyncSelect
5100 \S{export} \i\c{EXPORT}: Exporting DLL Symbols\I{DLL symbols,
5101 exporting}\I{symbols, exporting from DLLs}
5103 The \c{EXPORT} format-specific directive defines a global symbol to
5104 be exported as a DLL symbol, for use if you are writing a DLL in
5105 NASM. You still need to declare the symbol as \c{GLOBAL} as well as
5106 using the \c{EXPORT} directive.
5108 \c{EXPORT} takes one required parameter, which is the name of the
5109 symbol you wish to export, as it was defined in your source file. An
5110 optional second parameter (separated by white space from the first)
5111 gives the \e{external} name of the symbol: the name by which you
5112 wish the symbol to be known to programs using the DLL. If this name
5113 is the same as the internal name, you may leave the second parameter
5116 Further parameters can be given to define attributes of the exported
5117 symbol. These parameters, like the second, are separated by white
5118 space. If further parameters are given, the external name must also
5119 be specified, even if it is the same as the internal name. The
5120 available attributes are:
5122 \b \c{resident} indicates that the exported name is to be kept
5123 resident by the system loader. This is an optimisation for
5124 frequently used symbols imported by name.
5126 \b \c{nodata} indicates that the exported symbol is a function which
5127 does not make use of any initialized data.
5129 \b \c{parm=NNN}, where \c{NNN} is an integer, sets the number of
5130 parameter words for the case in which the symbol is a call gate
5131 between 32-bit and 16-bit segments.
5133 \b An attribute which is just a number indicates that the symbol
5134 should be exported with an identifying number (ordinal), and gives
5140 \c export myfunc TheRealMoreFormalLookingFunctionName
5141 \c export myfunc myfunc 1234 ; export by ordinal
5142 \c export myfunc myfunc resident parm=23 nodata
5145 \S{dotdotstart} \i\c{..start}: Defining the \i{Program Entry
5148 \c{OMF} linkers require exactly one of the object files being linked to
5149 define the program entry point, where execution will begin when the
5150 program is run. If the object file that defines the entry point is
5151 assembled using NASM, you specify the entry point by declaring the
5152 special symbol \c{..start} at the point where you wish execution to
5156 \S{objextern} \c{obj} Extensions to the \c{EXTERN}
5157 Directive\I{EXTERN, obj extensions to}
5159 If you declare an external symbol with the directive
5163 then references such as \c{mov ax,foo} will give you the offset of
5164 \c{foo} from its preferred segment base (as specified in whichever
5165 module \c{foo} is actually defined in). So to access the contents of
5166 \c{foo} you will usually need to do something like
5168 \c mov ax,seg foo ; get preferred segment base
5169 \c mov es,ax ; move it into ES
5170 \c mov ax,[es:foo] ; and use offset `foo' from it
5172 This is a little unwieldy, particularly if you know that an external
5173 is going to be accessible from a given segment or group, say
5174 \c{dgroup}. So if \c{DS} already contained \c{dgroup}, you could
5177 \c mov ax,[foo wrt dgroup]
5179 However, having to type this every time you want to access \c{foo}
5180 can be a pain; so NASM allows you to declare \c{foo} in the
5183 \c extern foo:wrt dgroup
5185 This form causes NASM to pretend that the preferred segment base of
5186 \c{foo} is in fact \c{dgroup}; so the expression \c{seg foo} will
5187 now return \c{dgroup}, and the expression \c{foo} is equivalent to
5190 This \I{default-WRT mechanism}default-\c{WRT} mechanism can be used
5191 to make externals appear to be relative to any group or segment in
5192 your program. It can also be applied to common variables: see
5196 \S{objcommon} \c{obj} Extensions to the \c{COMMON}
5197 Directive\I{COMMON, obj extensions to}
5199 The \c{obj} format allows common variables to be either near\I{near
5200 common variables} or far\I{far common variables}; NASM allows you to
5201 specify which your variables should be by the use of the syntax
5203 \c common nearvar 2:near ; `nearvar' is a near common
5204 \c common farvar 10:far ; and `farvar' is far
5206 Far common variables may be greater in size than 64Kb, and so the
5207 OMF specification says that they are declared as a number of
5208 \e{elements} of a given size. So a 10-byte far common variable could
5209 be declared as ten one-byte elements, five two-byte elements, two
5210 five-byte elements or one ten-byte element.
5212 Some \c{OMF} linkers require the \I{element size, in common
5213 variables}\I{common variables, element size}element size, as well as
5214 the variable size, to match when resolving common variables declared
5215 in more than one module. Therefore NASM must allow you to specify
5216 the element size on your far common variables. This is done by the
5219 \c common c_5by2 10:far 5 ; two five-byte elements
5220 \c common c_2by5 10:far 2 ; five two-byte elements
5222 If no element size is specified, the default is 1. Also, the \c{FAR}
5223 keyword is not required when an element size is specified, since
5224 only far commons may have element sizes at all. So the above
5225 declarations could equivalently be
5227 \c common c_5by2 10:5 ; two five-byte elements
5228 \c common c_2by5 10:2 ; five two-byte elements
5230 In addition to these extensions, the \c{COMMON} directive in \c{obj}
5231 also supports default-\c{WRT} specification like \c{EXTERN} does
5232 (explained in \k{objextern}). So you can also declare things like
5234 \c common foo 10:wrt dgroup
5235 \c common bar 16:far 2:wrt data
5236 \c common baz 24:wrt data:6
5239 \H{win32fmt} \i\c{win32}: Microsoft Win32 Object Files
5241 The \c{win32} output format generates Microsoft Win32 object files,
5242 suitable for passing to Microsoft linkers such as \i{Visual C++}.
5243 Note that Borland Win32 compilers do not use this format, but use
5244 \c{obj} instead (see \k{objfmt}).
5246 \c{win32} provides a default output file-name extension of \c{.obj}.
5248 Note that although Microsoft say that Win32 object files follow the
5249 \c{COFF} (Common Object File Format) standard, the object files produced
5250 by Microsoft Win32 compilers are not compatible with COFF linkers
5251 such as DJGPP's, and vice versa. This is due to a difference of
5252 opinion over the precise semantics of PC-relative relocations. To
5253 produce COFF files suitable for DJGPP, use NASM's \c{coff} output
5254 format; conversely, the \c{coff} format does not produce object
5255 files that Win32 linkers can generate correct output from.
5258 \S{win32sect} \c{win32} Extensions to the \c{SECTION}
5259 Directive\I{SECTION, win32 extensions to}
5261 Like the \c{obj} format, \c{win32} allows you to specify additional
5262 information on the \c{SECTION} directive line, to control the type
5263 and properties of sections you declare. Section types and properties
5264 are generated automatically by NASM for the \i{standard section names}
5265 \c{.text}, \c{.data} and \c{.bss}, but may still be overridden by
5268 The available qualifiers are:
5270 \b \c{code}, or equivalently \c{text}, defines the section to be a
5271 code section. This marks the section as readable and executable, but
5272 not writable, and also indicates to the linker that the type of the
5275 \b \c{data} and \c{bss} define the section to be a data section,
5276 analogously to \c{code}. Data sections are marked as readable and
5277 writable, but not executable. \c{data} declares an initialized data
5278 section, whereas \c{bss} declares an uninitialized data section.
5280 \b \c{rdata} declares an initialized data section that is readable
5281 but not writable. Microsoft compilers use this section to place
5284 \b \c{info} defines the section to be an \i{informational section},
5285 which is not included in the executable file by the linker, but may
5286 (for example) pass information \e{to} the linker. For example,
5287 declaring an \c{info}-type section called \i\c{.drectve} causes the
5288 linker to interpret the contents of the section as command-line
5291 \b \c{align=}, used with a trailing number as in \c{obj}, gives the
5292 \I{section alignment, in win32}\I{alignment, in win32
5293 sections}alignment requirements of the section. The maximum you may
5294 specify is 64: the Win32 object file format contains no means to
5295 request a greater section alignment than this. If alignment is not
5296 explicitly specified, the defaults are 16-byte alignment for code
5297 sections, 8-byte alignment for rdata sections and 4-byte alignment
5298 for data (and BSS) sections.
5299 Informational sections get a default alignment of 1 byte (no
5300 alignment), though the value does not matter.
5302 The defaults assumed by NASM if you do not specify the above
5305 \c section .text code align=16
5306 \c section .data data align=4
5307 \c section .rdata rdata align=8
5308 \c section .bss bss align=4
5310 Any other section name is treated by default like \c{.text}.
5312 \S{win32safeseh} \c{win32}: Safe Structured Exception Handling
5314 Among other improvements in Windows XP SP2 and Windows Server 2003
5315 Microsoft has introduced concept of "safe structured exception
5316 handling." General idea is to collect handlers' entry points in
5317 designated read-only table and have alleged entry point verified
5318 against this table prior exception control is passed to the handler. In
5319 order for an executable module to be equipped with such "safe exception
5320 handler table," all object modules on linker command line has to comply
5321 with certain criteria. If one single module among them does not, then
5322 the table in question is omitted and above mentioned run-time checks
5323 will not be performed for application in question. Table omission is by
5324 default silent and therefore can be easily overlooked. One can instruct
5325 linker to refuse to produce binary without such table by passing
5326 \c{/safeseh} command line option.
5328 Without regard to this run-time check merits it's natural to expect
5329 NASM to be capable of generating modules suitable for \c{/safeseh}
5330 linking. From developer's viewpoint the problem is two-fold:
5332 \b how to adapt modules not deploying exception handlers of their own;
5334 \b how to adapt/develop modules utilizing custom exception handling;
5336 Former can be easily achieved with any NASM version by adding following
5337 line to source code:
5341 As of version 2.03 NASM adds this absolute symbol automatically. If
5342 it's not already present to be precise. I.e. if for whatever reason
5343 developer would choose to assign another value in source file, it would
5344 still be perfectly possible.
5346 Registering custom exception handler on the other hand requires certain
5347 "magic." As of version 2.03 additional directive is implemented,
5348 \c{safeseh}, which instructs the assembler to produce appropriately
5349 formatted input data for above mentioned "safe exception handler
5350 table." Its typical use would be:
5353 \c extern _MessageBoxA@16
5354 \c %if __NASM_VERSION_ID__ >= 0x02030000
5355 \c safeseh handler ; register handler as "safe handler"
5358 \c push DWORD 1 ; MB_OKCANCEL
5359 \c push DWORD caption
5362 \c call _MessageBoxA@16
5363 \c sub eax,1 ; incidentally suits as return value
5364 \c ; for exception handler
5368 \c push DWORD handler
5369 \c push DWORD [fs:0]
5370 \c mov DWORD [fs:0],esp ; engage exception handler
5372 \c mov eax,DWORD[eax] ; cause exception
5373 \c pop DWORD [fs:0] ; disengage exception handler
5376 \c text: db 'OK to rethrow, CANCEL to generate core dump',0
5377 \c caption:db 'SEGV',0
5379 \c section .drectve info
5380 \c db '/defaultlib:user32.lib /defaultlib:msvcrt.lib '
5382 As you might imagine, it's perfectly possible to produce .exe binary
5383 with "safe exception handler table" and yet engage unregistered
5384 exception handler. Indeed, handler is engaged by simply manipulating
5385 \c{[fs:0]} location at run-time, something linker has no power over,
5386 run-time that is. It should be explicitly mentioned that such failure
5387 to register handler's entry point with \c{safeseh} directive has
5388 undesired side effect at run-time. If exception is raised and
5389 unregistered handler is to be executed, the application is abruptly
5390 terminated without any notification whatsoever. One can argue that
5391 system could at least have logged some kind "non-safe exception
5392 handler in x.exe at address n" message in event log, but no, literally
5393 no notification is provided and user is left with no clue on what
5394 caused application failure.
5396 Finally, all mentions of linker in this paragraph refer to Microsoft
5397 linker version 7.x and later. Presence of \c{@feat.00} symbol and input
5398 data for "safe exception handler table" causes no backward
5399 incompatibilities and "safeseh" modules generated by NASM 2.03 and
5400 later can still be linked by earlier versions or non-Microsoft linkers.
5403 \H{win64fmt} \i\c{win64}: Microsoft Win64 Object Files
5405 The \c{win64} output format generates Microsoft Win64 object files,
5406 which is nearly 100% identical to the \c{win32} object format (\k{win32fmt})
5407 with the exception that it is meant to target 64-bit code and the x86-64
5408 platform altogether. This object file is used exactly the same as the \c{win32}
5409 object format (\k{win32fmt}), in NASM, with regard to this exception.
5411 \S{win64pic} \c{win64}: Writing Position-Independent Code
5413 While \c{REL} takes good care of RIP-relative addressing, there is one
5414 aspect that is easy to overlook for a Win64 programmer: indirect
5415 references. Consider a switch dispatch table:
5417 \c jmp qword [dsptch+rax*8]
5423 Even a novice Win64 assembler programmer will soon realize that the code
5424 is not 64-bit savvy. Most notably linker will refuse to link it with
5426 \c 'ADDR32' relocation to '.text' invalid without /LARGEADDRESSAWARE:NO
5428 So [s]he will have to split jmp instruction as following:
5430 \c lea rbx,[rel dsptch]
5431 \c jmp qword [rbx+rax*8]
5433 What happens behind the scene is that effective address in \c{lea} is
5434 encoded relative to instruction pointer, or in perfectly
5435 position-independent manner. But this is only part of the problem!
5436 Trouble is that in .dll context \c{caseN} relocations will make their
5437 way to the final module and might have to be adjusted at .dll load
5438 time. To be specific when it can't be loaded at preferred address. And
5439 when this occurs, pages with such relocations will be rendered private
5440 to current process, which kind of undermines the idea of sharing .dll.
5441 But no worry, it's trivial to fix:
5443 \c lea rbx,[rel dsptch]
5444 \c add rbx,[rbx+rax*8]
5447 \c dsptch: dq case0-dsptch
5451 NASM version 2.03 and later provides another alternative, \c{wrt
5452 ..imagebase} operator, which returns offset from base address of the
5453 current image, be it .exe or .dll module, therefore the name. For those
5454 acquainted with PE-COFF format base address denotes start of
5455 \c{IMAGE_DOS_HEADER} structure. Here is how to implement switch with
5456 these image-relative references:
5458 \c lea rbx,[rel dsptch]
5459 \c mov eax,[rbx+rax*4]
5460 \c sub rbx,dsptch wrt ..imagebase
5464 \c dsptch: dd case0 wrt ..imagebase
5465 \c dd case1 wrt ..imagebase
5467 One can argue that the operator is redundant. Indeed, snippet before
5468 last works just fine with any NASM version and is not even Windows
5469 specific... The real reason for implementing \c{wrt ..imagebase} will
5470 become apparent in next paragraph.
5472 It should be noted that \c{wrt ..imagebase} is defined as 32-bit
5475 \c dd label wrt ..imagebase ; ok
5476 \c dq label wrt ..imagebase ; bad
5477 \c mov eax,label wrt ..imagebase ; ok
5478 \c mov rax,label wrt ..imagebase ; bad
5480 \S{win64seh} \c{win64}: Structured Exception Handling
5482 Structured exception handing in Win64 is completely different matter
5483 from Win32. Upon exception program counter value is noted, and
5484 linker-generated table comprising start and end addresses of all the
5485 functions [in given executable module] is traversed and compared to the
5486 saved program counter. Thus so called \c{UNWIND_INFO} structure is
5487 identified. If it's not found, then offending subroutine is assumed to
5488 be "leaf" and just mentioned lookup procedure is attempted for its
5489 caller. In Win64 leaf function is such function that does not call any
5490 other function \e{nor} modifies any Win64 non-volatile registers,
5491 including stack pointer. The latter ensures that it's possible to
5492 identify leaf function's caller by simply pulling the value from the
5495 While majority of subroutines written in assembler are not calling any
5496 other function, requirement for non-volatile registers' immutability
5497 leaves developer with not more than 7 registers and no stack frame,
5498 which is not necessarily what [s]he counted with. Customarily one would
5499 meet the requirement by saving non-volatile registers on stack and
5500 restoring them upon return, so what can go wrong? If [and only if] an
5501 exception is raised at run-time and no \c{UNWIND_INFO} structure is
5502 associated with such "leaf" function, the stack unwind procedure will
5503 expect to find caller's return address on the top of stack immediately
5504 followed by its frame. Given that developer pushed caller's
5505 non-volatile registers on stack, would the value on top point at some
5506 code segment or even addressable space? Well, developer can attempt
5507 copying caller's return address to the top of stack and this would
5508 actually work in some very specific circumstances. But unless developer
5509 can guarantee that these circumstances are always met, it's more
5510 appropriate to assume worst case scenario, i.e. stack unwind procedure
5511 going berserk. Relevant question is what happens then? Application is
5512 abruptly terminated without any notification whatsoever. Just like in
5513 Win32 case, one can argue that system could at least have logged
5514 "unwind procedure went berserk in x.exe at address n" in event log, but
5515 no, no trace of failure is left.
5517 Now, when we understand significance of the \c{UNWIND_INFO} structure,
5518 let's discuss what's in it and/or how it's processed. First of all it
5519 is checked for presence of reference to custom language-specific
5520 exception handler. If there is one, then it's invoked. Depending on the
5521 return value, execution flow is resumed (exception is said to be
5522 "handled"), \e{or} rest of \c{UNWIND_INFO} structure is processed as
5523 following. Beside optional reference to custom handler, it carries
5524 information about current callee's stack frame and where non-volatile
5525 registers are saved. Information is detailed enough to be able to
5526 reconstruct contents of caller's non-volatile registers upon call to
5527 current callee. And so caller's context is reconstructed, and then
5528 unwind procedure is repeated, i.e. another \c{UNWIND_INFO} structure is
5529 associated, this time, with caller's instruction pointer, which is then
5530 checked for presence of reference to language-specific handler, etc.
5531 The procedure is recursively repeated till exception is handled. As
5532 last resort system "handles" it by generating memory core dump and
5533 terminating the application.
5535 As for the moment of this writing NASM unfortunately does not
5536 facilitate generation of above mentioned detailed information about
5537 stack frame layout. But as of version 2.03 it implements building
5538 blocks for generating structures involved in stack unwinding. As
5539 simplest example, here is how to deploy custom exception handler for
5544 \c extern MessageBoxA
5550 \c mov r9,1 ; MB_OKCANCEL
5552 \c sub eax,1 ; incidentally suits as return value
5553 \c ; for exception handler
5559 \c mov rax,QWORD[rax] ; cause exception
5562 \c text: db 'OK to rethrow, CANCEL to generate core dump',0
5563 \c caption:db 'SEGV',0
5565 \c section .pdata rdata align=4
5566 \c dd main wrt ..imagebase
5567 \c dd main_end wrt ..imagebase
5568 \c dd xmain wrt ..imagebase
5569 \c section .xdata rdata align=8
5570 \c xmain: db 9,0,0,0
5571 \c dd handler wrt ..imagebase
5572 \c section .drectve info
5573 \c db '/defaultlib:user32.lib /defaultlib:msvcrt.lib '
5575 What you see in \c{.pdata} section is element of the "table comprising
5576 start and end addresses of function" along with reference to associated
5577 \c{UNWIND_INFO} structure. And what you see in \c{.xdata} section is
5578 \c{UNWIND_INFO} structure describing function with no frame, but with
5579 designated exception handler. References are \e{required} to be
5580 image-relative (which is the real reason for implementing \c{wrt
5581 ..imagebase} operator). It should be noted that \c{rdata align=n}, as
5582 well as \c{wrt ..imagebase}, are optional in these two segments'
5583 contexts, i.e. can be omitted. Latter means that \e{all} 32-bit
5584 references, not only above listed required ones, placed into these two
5585 segments turn out image-relative. Why is it important to understand?
5586 Developer is allowed to append handler-specific data to \c{UNWIND_INFO}
5587 structure, and if [s]he adds a 32-bit reference, then [s]he will have
5588 to remember to adjust its value to obtain the real pointer.
5590 As already mentioned, in Win64 terms leaf function is one that does not
5591 call any other function \e{nor} modifies any non-volatile register,
5592 including stack pointer. But it's not uncommon that assembler
5593 programmer plans to utilize every single register and sometimes even
5594 have variable stack frame. Is there anything one can do with bare
5595 building blocks? I.e. besides manually composing fully-fledged
5596 \c{UNWIND_INFO} structure, which would surely be considered
5597 error-prone? Yes, there is. Recall that exception handler is called
5598 first, before stack layout is analyzed. As it turned out, it's
5599 perfectly possible to manipulate current callee's context in custom
5600 handler in manner that permits further stack unwinding. General idea is
5601 that handler would not actually "handle" the exception, but instead
5602 restore callee's context, as it was at its entry point and thus mimic
5603 leaf function. In other words, handler would simply undertake part of
5604 unwinding procedure. Consider following example:
5607 \c mov rax,rsp ; copy rsp to volatile register
5608 \c push r15 ; save non-volatile registers
5611 \c mov r11,rsp ; prepare variable stack frame
5614 \c mov QWORD[r11],rax ; check for exceptions
5615 \c mov rsp,r11 ; allocate stack frame
5616 \c mov QWORD[rsp],rax ; save original rsp value
5619 \c mov r11,QWORD[rsp] ; pull original rsp value
5620 \c mov rbp,QWORD[r11-24]
5621 \c mov rbx,QWORD[r11-16]
5622 \c mov r15,QWORD[r11-8]
5623 \c mov rsp,r11 ; destroy frame
5626 The keyword is that up to \c{magic_point} original \c{rsp} value
5627 remains in chosen volatile register and no non-volatile register,
5628 except for \c{rsp}, is modified. While past \c{magic_point} \c{rsp}
5629 remains constant till the very end of the \c{function}. In this case
5630 custom language-specific exception handler would look like this:
5632 \c EXCEPTION_DISPOSITION handler (EXCEPTION_RECORD *rec,ULONG64 frame,
5633 \c CONTEXT *context,DISPATCHER_CONTEXT *disp)
5635 \c if (context->Rip<(ULONG64)magic_point)
5636 \c rsp = (ULONG64 *)context->Rax;
5638 \c { rsp = ((ULONG64 **)context->Rsp)[0];
5639 \c context->Rbp = rsp[-3];
5640 \c context->Rbx = rsp[-2];
5641 \c context->R15 = rsp[-1];
5643 \c context->Rsp = (ULONG64)rsp;
5645 \c memcpy (disp->ContextRecord,context,sizeof(CONTEXT));
5646 \c RtlVirtualUnwind(UNW_FLAG_NHANDLER,disp->ImageBase,
5647 \c dips->ControlPc,disp->FunctionEntry,disp->ContextRecord,
5648 \c &disp->HandlerData,&disp->EstablisherFrame,NULL);
5649 \c return ExceptionContinueSearch;
5652 As custom handler mimics leaf function, corresponding \c{UNWIND_INFO}
5653 structure does not have to contain any information about stack frame
5656 \H{cofffmt} \i\c{coff}: \i{Common Object File Format}
5658 The \c{coff} output type produces \c{COFF} object files suitable for
5659 linking with the \i{DJGPP} linker.
5661 \c{coff} provides a default output file-name extension of \c{.o}.
5663 The \c{coff} format supports the same extensions to the \c{SECTION}
5664 directive as \c{win32} does, except that the \c{align} qualifier and
5665 the \c{info} section type are not supported.
5667 \H{machofmt} \I{Mach-O}\i\c{macho32} and \i\c{macho64}: \i{Mach Object File Format}
5669 The \c{macho32} and \c{macho64} output formts produces \c{Mach-O}
5670 object files suitable for linking with the \i{MacOS X} linker.
5671 \i\c{macho} is a synonym for \c{macho32}.
5673 \c{macho} provides a default output file-name extension of \c{.o}.
5675 \H{elffmt} \i\c{elf32}, \i\c{elf64}, \i\c{elfx32}: \I{ELF}\I{linux, elf}\i{Executable and Linkable
5676 Format} Object Files
5678 The \c{elf32}, \c{elf64} and \c{elfx32} output formats generate
5679 \c{ELF32 and ELF64} (Executable and Linkable Format) object files, as
5680 used by Linux as well as \i{Unix System V}, including \i{Solaris x86},
5681 \i{UnixWare} and \i{SCO Unix}. \c{elf} provides a default output
5682 file-name extension of \c{.o}. \c{elf} is a synonym for \c{elf32}.
5684 The \c{elfx32} format is used for the \i{x32} ABI, which is a 32-bit
5685 ABI with the CPU in 64-bit mode.
5687 \S{abisect} ELF specific directive \i\c{osabi}
5689 The ELF header specifies the application binary interface for the target operating system (OSABI).
5690 This field can be set by using the \c{osabi} directive with the numeric value (0-255) of the target
5691 system. If this directive is not used, the default value will be "UNIX System V ABI" (0) which will work on
5692 most systems which support ELF.
5694 \S{elfsect} \c{elf} Extensions to the \c{SECTION}
5695 Directive\I{SECTION, elf extensions to}
5697 Like the \c{obj} format, \c{elf} allows you to specify additional
5698 information on the \c{SECTION} directive line, to control the type
5699 and properties of sections you declare. Section types and properties
5700 are generated automatically by NASM for the \i{standard section
5701 names}, but may still be
5702 overridden by these qualifiers.
5704 The available qualifiers are:
5706 \b \i\c{alloc} defines the section to be one which is loaded into
5707 memory when the program is run. \i\c{noalloc} defines it to be one
5708 which is not, such as an informational or comment section.
5710 \b \i\c{exec} defines the section to be one which should have execute
5711 permission when the program is run. \i\c{noexec} defines it as one
5714 \b \i\c{write} defines the section to be one which should be writable
5715 when the program is run. \i\c{nowrite} defines it as one which should
5718 \b \i\c{progbits} defines the section to be one with explicit contents
5719 stored in the object file: an ordinary code or data section, for
5720 example, \i\c{nobits} defines the section to be one with no explicit
5721 contents given, such as a BSS section.
5723 \b \c{align=}, used with a trailing number as in \c{obj}, gives the
5724 \I{section alignment, in elf}\I{alignment, in elf sections}alignment
5725 requirements of the section.
5727 \b \i\c{tls} defines the section to be one which contains
5728 thread local variables.
5730 The defaults assumed by NASM if you do not specify the above
5733 \I\c{.text} \I\c{.rodata} \I\c{.lrodata} \I\c{.data} \I\c{.ldata}
5734 \I\c{.bss} \I\c{.lbss} \I\c{.tdata} \I\c{.tbss} \I\c\{.comment}
5736 \c section .text progbits alloc exec nowrite align=16
5737 \c section .rodata progbits alloc noexec nowrite align=4
5738 \c section .lrodata progbits alloc noexec nowrite align=4
5739 \c section .data progbits alloc noexec write align=4
5740 \c section .ldata progbits alloc noexec write align=4
5741 \c section .bss nobits alloc noexec write align=4
5742 \c section .lbss nobits alloc noexec write align=4
5743 \c section .tdata progbits alloc noexec write align=4 tls
5744 \c section .tbss nobits alloc noexec write align=4 tls
5745 \c section .comment progbits noalloc noexec nowrite align=1
5746 \c section other progbits alloc noexec nowrite align=1
5748 (Any section name other than those in the above table
5749 is treated by default like \c{other} in the above table.
5750 Please note that section names are case sensitive.)
5753 \S{elfwrt} \i{Position-Independent Code}\I{PIC}: \c{elf} Special
5754 Symbols and \i\c{WRT}
5756 The \c{ELF} specification contains enough features to allow
5757 position-independent code (PIC) to be written, which makes \i{ELF
5758 shared libraries} very flexible. However, it also means NASM has to
5759 be able to generate a variety of ELF specific relocation types in ELF
5760 object files, if it is to be an assembler which can write PIC.
5762 Since \c{ELF} does not support segment-base references, the \c{WRT}
5763 operator is not used for its normal purpose; therefore NASM's
5764 \c{elf} output format makes use of \c{WRT} for a different purpose,
5765 namely the PIC-specific \I{relocations, PIC-specific}relocation
5768 \c{elf} defines five special symbols which you can use as the
5769 right-hand side of the \c{WRT} operator to obtain PIC relocation
5770 types. They are \i\c{..gotpc}, \i\c{..gotoff}, \i\c{..got},
5771 \i\c{..plt} and \i\c{..sym}. Their functions are summarized here:
5773 \b Referring to the symbol marking the global offset table base
5774 using \c{wrt ..gotpc} will end up giving the distance from the
5775 beginning of the current section to the global offset table.
5776 (\i\c{_GLOBAL_OFFSET_TABLE_} is the standard symbol name used to
5777 refer to the \i{GOT}.) So you would then need to add \i\c{$$} to the
5778 result to get the real address of the GOT.
5780 \b Referring to a location in one of your own sections using \c{wrt
5781 ..gotoff} will give the distance from the beginning of the GOT to
5782 the specified location, so that adding on the address of the GOT
5783 would give the real address of the location you wanted.
5785 \b Referring to an external or global symbol using \c{wrt ..got}
5786 causes the linker to build an entry \e{in} the GOT containing the
5787 address of the symbol, and the reference gives the distance from the
5788 beginning of the GOT to the entry; so you can add on the address of
5789 the GOT, load from the resulting address, and end up with the
5790 address of the symbol.
5792 \b Referring to a procedure name using \c{wrt ..plt} causes the
5793 linker to build a \i{procedure linkage table} entry for the symbol,
5794 and the reference gives the address of the \i{PLT} entry. You can
5795 only use this in contexts which would generate a PC-relative
5796 relocation normally (i.e. as the destination for \c{CALL} or
5797 \c{JMP}), since ELF contains no relocation type to refer to PLT
5800 \b Referring to a symbol name using \c{wrt ..sym} causes NASM to
5801 write an ordinary relocation, but instead of making the relocation
5802 relative to the start of the section and then adding on the offset
5803 to the symbol, it will write a relocation record aimed directly at
5804 the symbol in question. The distinction is a necessary one due to a
5805 peculiarity of the dynamic linker.
5807 A fuller explanation of how to use these relocation types to write
5808 shared libraries entirely in NASM is given in \k{picdll}.
5810 \S{elftls} \i{Thread Local Storage}\I{TLS}: \c{elf} Special
5811 Symbols and \i\c{WRT}
5813 \b In ELF32 mode, referring to an external or global symbol using
5814 \c{wrt ..tlsie} \I\c{..tlsie}
5815 causes the linker to build an entry \e{in} the GOT containing the
5816 offset of the symbol within the TLS block, so you can access the value
5817 of the symbol with code such as:
5819 \c mov eax,[tid wrt ..tlsie]
5823 \b In ELF64 or ELFx32 mode, referring to an external or global symbol using
5824 \c{wrt ..gottpoff} \I\c{..gottpoff}
5825 causes the linker to build an entry \e{in} the GOT containing the
5826 offset of the symbol within the TLS block, so you can access the value
5827 of the symbol with code such as:
5829 \c mov rax,[rel tid wrt ..gottpoff]
5833 \S{elfglob} \c{elf} Extensions to the \c{GLOBAL} Directive\I{GLOBAL,
5834 elf extensions to}\I{GLOBAL, aoutb extensions to}
5836 \c{ELF} object files can contain more information about a global symbol
5837 than just its address: they can contain the \I{symbol sizes,
5838 specifying}\I{size, of symbols}size of the symbol and its \I{symbol
5839 types, specifying}\I{type, of symbols}type as well. These are not
5840 merely debugger conveniences, but are actually necessary when the
5841 program being written is a \i{shared library}. NASM therefore
5842 supports some extensions to the \c{GLOBAL} directive, allowing you
5843 to specify these features.
5845 You can specify whether a global variable is a function or a data
5846 object by suffixing the name with a colon and the word
5847 \i\c{function} or \i\c{data}. (\i\c{object} is a synonym for
5848 \c{data}.) For example:
5850 \c global hashlookup:function, hashtable:data
5852 exports the global symbol \c{hashlookup} as a function and
5853 \c{hashtable} as a data object.
5855 Optionally, you can control the ELF visibility of the symbol. Just
5856 add one of the visibility keywords: \i\c{default}, \i\c{internal},
5857 \i\c{hidden}, or \i\c{protected}. The default is \i\c{default} of
5858 course. For example, to make \c{hashlookup} hidden:
5860 \c global hashlookup:function hidden
5862 You can also specify the size of the data associated with the
5863 symbol, as a numeric expression (which may involve labels, and even
5864 forward references) after the type specifier. Like this:
5866 \c global hashtable:data (hashtable.end - hashtable)
5869 \c db this,that,theother ; some data here
5872 This makes NASM automatically calculate the length of the table and
5873 place that information into the \c{ELF} symbol table.
5875 Declaring the type and size of global symbols is necessary when
5876 writing shared library code. For more information, see
5880 \S{elfcomm} \c{elf} Extensions to the \c{COMMON} Directive
5881 \I{COMMON, elf extensions to}
5883 \c{ELF} also allows you to specify alignment requirements \I{common
5884 variables, alignment in elf}\I{alignment, of elf common variables}on
5885 common variables. This is done by putting a number (which must be a
5886 power of two) after the name and size of the common variable,
5887 separated (as usual) by a colon. For example, an array of
5888 doublewords would benefit from 4-byte alignment:
5890 \c common dwordarray 128:4
5892 This declares the total size of the array to be 128 bytes, and
5893 requires that it be aligned on a 4-byte boundary.
5896 \S{elf16} 16-bit code and ELF
5897 \I{ELF, 16-bit code and}
5899 The \c{ELF32} specification doesn't provide relocations for 8- and
5900 16-bit values, but the GNU \c{ld} linker adds these as an extension.
5901 NASM can generate GNU-compatible relocations, to allow 16-bit code to
5902 be linked as ELF using GNU \c{ld}. If NASM is used with the
5903 \c{-w+gnu-elf-extensions} option, a warning is issued when one of
5904 these relocations is generated.
5906 \S{elfdbg} Debug formats and ELF
5907 \I{ELF, Debug formats and}
5909 ELF provides debug information in \c{STABS} and \c{DWARF} formats.
5910 Line number information is generated for all executable sections, but please
5911 note that only the ".text" section is executable by default.
5913 \H{aoutfmt} \i\c{aout}: Linux \I{a.out, Linux version}\I{linux, a.out}\c{a.out} Object Files
5915 The \c{aout} format generates \c{a.out} object files, in the form used
5916 by early Linux systems (current Linux systems use ELF, see
5917 \k{elffmt}.) These differ from other \c{a.out} object files in that
5918 the magic number in the first four bytes of the file is
5919 different; also, some implementations of \c{a.out}, for example
5920 NetBSD's, support position-independent code, which Linux's
5921 implementation does not.
5923 \c{a.out} provides a default output file-name extension of \c{.o}.
5925 \c{a.out} is a very simple object format. It supports no special
5926 directives, no special symbols, no use of \c{SEG} or \c{WRT}, and no
5927 extensions to any standard directives. It supports only the three
5928 \i{standard section names} \i\c{.text}, \i\c{.data} and \i\c{.bss}.
5931 \H{aoutfmt} \i\c{aoutb}: \i{NetBSD}/\i{FreeBSD}/\i{OpenBSD}
5932 \I{a.out, BSD version}\c{a.out} Object Files
5934 The \c{aoutb} format generates \c{a.out} object files, in the form
5935 used by the various free \c{BSD Unix} clones, \c{NetBSD}, \c{FreeBSD}
5936 and \c{OpenBSD}. For simple object files, this object format is exactly
5937 the same as \c{aout} except for the magic number in the first four bytes
5938 of the file. However, the \c{aoutb} format supports
5939 \I{PIC}\i{position-independent code} in the same way as the \c{elf}
5940 format, so you can use it to write \c{BSD} \i{shared libraries}.
5942 \c{aoutb} provides a default output file-name extension of \c{.o}.
5944 \c{aoutb} supports no special directives, no special symbols, and
5945 only the three \i{standard section names} \i\c{.text}, \i\c{.data}
5946 and \i\c{.bss}. However, it also supports the same use of \i\c{WRT} as
5947 \c{elf} does, to provide position-independent code relocation types.
5948 See \k{elfwrt} for full documentation of this feature.
5950 \c{aoutb} also supports the same extensions to the \c{GLOBAL}
5951 directive as \c{elf} does: see \k{elfglob} for documentation of
5955 \H{as86fmt} \c{as86}: \i{Minix}/Linux\I{linux, as86} \i\c{as86} Object Files
5957 The Minix/Linux 16-bit assembler \c{as86} has its own non-standard
5958 object file format. Although its companion linker \i\c{ld86} produces
5959 something close to ordinary \c{a.out} binaries as output, the object
5960 file format used to communicate between \c{as86} and \c{ld86} is not
5963 NASM supports this format, just in case it is useful, as \c{as86}.
5964 \c{as86} provides a default output file-name extension of \c{.o}.
5966 \c{as86} is a very simple object format (from the NASM user's point
5967 of view). It supports no special directives, no use of \c{SEG} or \c{WRT},
5968 and no extensions to any standard directives. It supports only the three
5969 \i{standard section names} \i\c{.text}, \i\c{.data} and \i\c{.bss}. The
5970 only special symbol supported is \c{..start}.
5973 \H{rdffmt} \I{RDOFF}\i\c{rdf}: \i{Relocatable Dynamic Object File
5976 The \c{rdf} output format produces \c{RDOFF} object files. \c{RDOFF}
5977 (Relocatable Dynamic Object File Format) is a home-grown object-file
5978 format, designed alongside NASM itself and reflecting in its file
5979 format the internal structure of the assembler.
5981 \c{RDOFF} is not used by any well-known operating systems. Those
5982 writing their own systems, however, may well wish to use \c{RDOFF}
5983 as their object format, on the grounds that it is designed primarily
5984 for simplicity and contains very little file-header bureaucracy.
5986 The Unix NASM archive, and the DOS archive which includes sources,
5987 both contain an \I{rdoff subdirectory}\c{rdoff} subdirectory holding
5988 a set of RDOFF utilities: an RDF linker, an \c{RDF} static-library
5989 manager, an RDF file dump utility, and a program which will load and
5990 execute an RDF executable under Linux.
5992 \c{rdf} supports only the \i{standard section names} \i\c{.text},
5993 \i\c{.data} and \i\c{.bss}.
5996 \S{rdflib} Requiring a Library: The \i\c{LIBRARY} Directive
5998 \c{RDOFF} contains a mechanism for an object file to demand a given
5999 library to be linked to the module, either at load time or run time.
6000 This is done by the \c{LIBRARY} directive, which takes one argument
6001 which is the name of the module:
6003 \c library mylib.rdl
6006 \S{rdfmod} Specifying a Module Name: The \i\c{MODULE} Directive
6008 Special \c{RDOFF} header record is used to store the name of the module.
6009 It can be used, for example, by run-time loader to perform dynamic
6010 linking. \c{MODULE} directive takes one argument which is the name
6015 Note that when you statically link modules and tell linker to strip
6016 the symbols from output file, all module names will be stripped too.
6017 To avoid it, you should start module names with \I{$, prefix}\c{$}, like:
6019 \c module $kernel.core
6022 \S{rdfglob} \c{rdf} Extensions to the \c{GLOBAL} Directive\I{GLOBAL,
6025 \c{RDOFF} global symbols can contain additional information needed by
6026 the static linker. You can mark a global symbol as exported, thus
6027 telling the linker do not strip it from target executable or library
6028 file. Like in \c{ELF}, you can also specify whether an exported symbol
6029 is a procedure (function) or data object.
6031 Suffixing the name with a colon and the word \i\c{export} you make the
6034 \c global sys_open:export
6036 To specify that exported symbol is a procedure (function), you add the
6037 word \i\c{proc} or \i\c{function} after declaration:
6039 \c global sys_open:export proc
6041 Similarly, to specify exported data object, add the word \i\c{data}
6042 or \i\c{object} to the directive:
6044 \c global kernel_ticks:export data
6047 \S{rdfimpt} \c{rdf} Extensions to the \c{EXTERN} Directive\I{EXTERN,
6050 By default the \c{EXTERN} directive in \c{RDOFF} declares a "pure external"
6051 symbol (i.e. the static linker will complain if such a symbol is not resolved).
6052 To declare an "imported" symbol, which must be resolved later during a dynamic
6053 linking phase, \c{RDOFF} offers an additional \c{import} modifier. As in
6054 \c{GLOBAL}, you can also specify whether an imported symbol is a procedure
6055 (function) or data object. For example:
6058 \c extern _open:import
6059 \c extern _printf:import proc
6060 \c extern _errno:import data
6062 Here the directive \c{LIBRARY} is also included, which gives the dynamic linker
6063 a hint as to where to find requested symbols.
6066 \H{dbgfmt} \i\c{dbg}: Debugging Format
6068 The \c{dbg} output format is not built into NASM in the default
6069 configuration. If you are building your own NASM executable from the
6070 sources, you can define \i\c{OF_DBG} in \c{output/outform.h} or on the
6071 compiler command line, and obtain the \c{dbg} output format.
6073 The \c{dbg} format does not output an object file as such; instead,
6074 it outputs a text file which contains a complete list of all the
6075 transactions between the main body of NASM and the output-format
6076 back end module. It is primarily intended to aid people who want to
6077 write their own output drivers, so that they can get a clearer idea
6078 of the various requests the main program makes of the output driver,
6079 and in what order they happen.
6081 For simple files, one can easily use the \c{dbg} format like this:
6083 \c nasm -f dbg filename.asm
6085 which will generate a diagnostic file called \c{filename.dbg}.
6086 However, this will not work well on files which were designed for a
6087 different object format, because each object format defines its own
6088 macros (usually user-level forms of directives), and those macros
6089 will not be defined in the \c{dbg} format. Therefore it can be
6090 useful to run NASM twice, in order to do the preprocessing with the
6091 native object format selected:
6093 \c nasm -e -f rdf -o rdfprog.i rdfprog.asm
6094 \c nasm -a -f dbg rdfprog.i
6096 This preprocesses \c{rdfprog.asm} into \c{rdfprog.i}, keeping the
6097 \c{rdf} object format selected in order to make sure RDF special
6098 directives are converted into primitive form correctly. Then the
6099 preprocessed source is fed through the \c{dbg} format to generate
6100 the final diagnostic output.
6102 This workaround will still typically not work for programs intended
6103 for \c{obj} format, because the \c{obj} \c{SEGMENT} and \c{GROUP}
6104 directives have side effects of defining the segment and group names
6105 as symbols; \c{dbg} will not do this, so the program will not
6106 assemble. You will have to work around that by defining the symbols
6107 yourself (using \c{EXTERN}, for example) if you really need to get a
6108 \c{dbg} trace of an \c{obj}-specific source file.
6110 \c{dbg} accepts any section name and any directives at all, and logs
6111 them all to its output file.
6114 \C{16bit} Writing 16-bit Code (DOS, Windows 3/3.1)
6116 This chapter attempts to cover some of the common issues encountered
6117 when writing 16-bit code to run under \c{MS-DOS} or \c{Windows 3.x}. It
6118 covers how to link programs to produce \c{.EXE} or \c{.COM} files,
6119 how to write \c{.SYS} device drivers, and how to interface assembly
6120 language code with 16-bit C compilers and with Borland Pascal.
6123 \H{exefiles} Producing \i\c{.EXE} Files
6125 Any large program written under DOS needs to be built as a \c{.EXE}
6126 file: only \c{.EXE} files have the necessary internal structure
6127 required to span more than one 64K segment. \i{Windows} programs,
6128 also, have to be built as \c{.EXE} files, since Windows does not
6129 support the \c{.COM} format.
6131 In general, you generate \c{.EXE} files by using the \c{obj} output
6132 format to produce one or more \i\c{.OBJ} files, and then linking
6133 them together using a linker. However, NASM also supports the direct
6134 generation of simple DOS \c{.EXE} files using the \c{bin} output
6135 format (by using \c{DB} and \c{DW} to construct the \c{.EXE} file
6136 header), and a macro package is supplied to do this. Thanks to
6137 Yann Guidon for contributing the code for this.
6139 NASM may also support \c{.EXE} natively as another output format in
6143 \S{objexe} Using the \c{obj} Format To Generate \c{.EXE} Files
6145 This section describes the usual method of generating \c{.EXE} files
6146 by linking \c{.OBJ} files together.
6148 Most 16-bit programming language packages come with a suitable
6149 linker; if you have none of these, there is a free linker called
6150 \i{VAL}\I{linker, free}, available in \c{LZH} archive format from
6151 \W{ftp://x2ftp.oulu.fi/pub/msdos/programming/lang/}\i\c{x2ftp.oulu.fi}.
6152 An LZH archiver can be found at
6153 \W{ftp://ftp.simtel.net/pub/simtelnet/msdos/arcers}\i\c{ftp.simtel.net}.
6154 There is another `free' linker (though this one doesn't come with
6155 sources) called \i{FREELINK}, available from
6156 \W{http://www.pcorner.com/tpc/old/3-101.html}\i\c{www.pcorner.com}.
6157 A third, \i\c{djlink}, written by DJ Delorie, is available at
6158 \W{http://www.delorie.com/djgpp/16bit/djlink/}\i\c{www.delorie.com}.
6159 A fourth linker, \i\c{ALINK}, written by Anthony A.J. Williams, is
6160 available at \W{http://alink.sourceforge.net}\i\c{alink.sourceforge.net}.
6162 When linking several \c{.OBJ} files into a \c{.EXE} file, you should
6163 ensure that exactly one of them has a start point defined (using the
6164 \I{program entry point}\i\c{..start} special symbol defined by the
6165 \c{obj} format: see \k{dotdotstart}). If no module defines a start
6166 point, the linker will not know what value to give the entry-point
6167 field in the output file header; if more than one defines a start
6168 point, the linker will not know \e{which} value to use.
6170 An example of a NASM source file which can be assembled to a
6171 \c{.OBJ} file and linked on its own to a \c{.EXE} is given here. It
6172 demonstrates the basic principles of defining a stack, initialising
6173 the segment registers, and declaring a start point. This file is
6174 also provided in the \I{test subdirectory}\c{test} subdirectory of
6175 the NASM archives, under the name \c{objexe.asm}.
6186 This initial piece of code sets up \c{DS} to point to the data
6187 segment, and initializes \c{SS} and \c{SP} to point to the top of
6188 the provided stack. Notice that interrupts are implicitly disabled
6189 for one instruction after a move into \c{SS}, precisely for this
6190 situation, so that there's no chance of an interrupt occurring
6191 between the loads of \c{SS} and \c{SP} and not having a stack to
6194 Note also that the special symbol \c{..start} is defined at the
6195 beginning of this code, which means that will be the entry point
6196 into the resulting executable file.
6202 The above is the main program: load \c{DS:DX} with a pointer to the
6203 greeting message (\c{hello} is implicitly relative to the segment
6204 \c{data}, which was loaded into \c{DS} in the setup code, so the
6205 full pointer is valid), and call the DOS print-string function.
6210 This terminates the program using another DOS system call.
6214 \c hello: db 'hello, world', 13, 10, '$'
6216 The data segment contains the string we want to display.
6218 \c segment stack stack
6222 The above code declares a stack segment containing 64 bytes of
6223 uninitialized stack space, and points \c{stacktop} at the top of it.
6224 The directive \c{segment stack stack} defines a segment \e{called}
6225 \c{stack}, and also of \e{type} \c{STACK}. The latter is not
6226 necessary to the correct running of the program, but linkers are
6227 likely to issue warnings or errors if your program has no segment of
6230 The above file, when assembled into a \c{.OBJ} file, will link on
6231 its own to a valid \c{.EXE} file, which when run will print `hello,
6232 world' and then exit.
6235 \S{binexe} Using the \c{bin} Format To Generate \c{.EXE} Files
6237 The \c{.EXE} file format is simple enough that it's possible to
6238 build a \c{.EXE} file by writing a pure-binary program and sticking
6239 a 32-byte header on the front. This header is simple enough that it
6240 can be generated using \c{DB} and \c{DW} commands by NASM itself, so
6241 that you can use the \c{bin} output format to directly generate
6244 Included in the NASM archives, in the \I{misc subdirectory}\c{misc}
6245 subdirectory, is a file \i\c{exebin.mac} of macros. It defines three
6246 macros: \i\c{EXE_begin}, \i\c{EXE_stack} and \i\c{EXE_end}.
6248 To produce a \c{.EXE} file using this method, you should start by
6249 using \c{%include} to load the \c{exebin.mac} macro package into
6250 your source file. You should then issue the \c{EXE_begin} macro call
6251 (which takes no arguments) to generate the file header data. Then
6252 write code as normal for the \c{bin} format - you can use all three
6253 standard sections \c{.text}, \c{.data} and \c{.bss}. At the end of
6254 the file you should call the \c{EXE_end} macro (again, no arguments),
6255 which defines some symbols to mark section sizes, and these symbols
6256 are referred to in the header code generated by \c{EXE_begin}.
6258 In this model, the code you end up writing starts at \c{0x100}, just
6259 like a \c{.COM} file - in fact, if you strip off the 32-byte header
6260 from the resulting \c{.EXE} file, you will have a valid \c{.COM}
6261 program. All the segment bases are the same, so you are limited to a
6262 64K program, again just like a \c{.COM} file. Note that an \c{ORG}
6263 directive is issued by the \c{EXE_begin} macro, so you should not
6264 explicitly issue one of your own.
6266 You can't directly refer to your segment base value, unfortunately,
6267 since this would require a relocation in the header, and things
6268 would get a lot more complicated. So you should get your segment
6269 base by copying it out of \c{CS} instead.
6271 On entry to your \c{.EXE} file, \c{SS:SP} are already set up to
6272 point to the top of a 2Kb stack. You can adjust the default stack
6273 size of 2Kb by calling the \c{EXE_stack} macro. For example, to
6274 change the stack size of your program to 64 bytes, you would call
6277 A sample program which generates a \c{.EXE} file in this way is
6278 given in the \c{test} subdirectory of the NASM archive, as
6282 \H{comfiles} Producing \i\c{.COM} Files
6284 While large DOS programs must be written as \c{.EXE} files, small
6285 ones are often better written as \c{.COM} files. \c{.COM} files are
6286 pure binary, and therefore most easily produced using the \c{bin}
6290 \S{combinfmt} Using the \c{bin} Format To Generate \c{.COM} Files
6292 \c{.COM} files expect to be loaded at offset \c{100h} into their
6293 segment (though the segment may change). Execution then begins at
6294 \I\c{ORG}\c{100h}, i.e. right at the start of the program. So to
6295 write a \c{.COM} program, you would create a source file looking
6303 \c ; put your code here
6307 \c ; put data items here
6311 \c ; put uninitialized data here
6313 The \c{bin} format puts the \c{.text} section first in the file, so
6314 you can declare data or BSS items before beginning to write code if
6315 you want to and the code will still end up at the front of the file
6318 The BSS (uninitialized data) section does not take up space in the
6319 \c{.COM} file itself: instead, addresses of BSS items are resolved
6320 to point at space beyond the end of the file, on the grounds that
6321 this will be free memory when the program is run. Therefore you
6322 should not rely on your BSS being initialized to all zeros when you
6325 To assemble the above program, you should use a command line like
6327 \c nasm myprog.asm -fbin -o myprog.com
6329 The \c{bin} format would produce a file called \c{myprog} if no
6330 explicit output file name were specified, so you have to override it
6331 and give the desired file name.
6334 \S{comobjfmt} Using the \c{obj} Format To Generate \c{.COM} Files
6336 If you are writing a \c{.COM} program as more than one module, you
6337 may wish to assemble several \c{.OBJ} files and link them together
6338 into a \c{.COM} program. You can do this, provided you have a linker
6339 capable of outputting \c{.COM} files directly (\i{TLINK} does this),
6340 or alternatively a converter program such as \i\c{EXE2BIN} to
6341 transform the \c{.EXE} file output from the linker into a \c{.COM}
6344 If you do this, you need to take care of several things:
6346 \b The first object file containing code should start its code
6347 segment with a line like \c{RESB 100h}. This is to ensure that the
6348 code begins at offset \c{100h} relative to the beginning of the code
6349 segment, so that the linker or converter program does not have to
6350 adjust address references within the file when generating the
6351 \c{.COM} file. Other assemblers use an \i\c{ORG} directive for this
6352 purpose, but \c{ORG} in NASM is a format-specific directive to the
6353 \c{bin} output format, and does not mean the same thing as it does
6354 in MASM-compatible assemblers.
6356 \b You don't need to define a stack segment.
6358 \b All your segments should be in the same group, so that every time
6359 your code or data references a symbol offset, all offsets are
6360 relative to the same segment base. This is because, when a \c{.COM}
6361 file is loaded, all the segment registers contain the same value.
6364 \H{sysfiles} Producing \i\c{.SYS} Files
6366 \i{MS-DOS device drivers} - \c{.SYS} files - are pure binary files,
6367 similar to \c{.COM} files, except that they start at origin zero
6368 rather than \c{100h}. Therefore, if you are writing a device driver
6369 using the \c{bin} format, you do not need the \c{ORG} directive,
6370 since the default origin for \c{bin} is zero. Similarly, if you are
6371 using \c{obj}, you do not need the \c{RESB 100h} at the start of
6374 \c{.SYS} files start with a header structure, containing pointers to
6375 the various routines inside the driver which do the work. This
6376 structure should be defined at the start of the code segment, even
6377 though it is not actually code.
6379 For more information on the format of \c{.SYS} files, and the data
6380 which has to go in the header structure, a list of books is given in
6381 the Frequently Asked Questions list for the newsgroup
6382 \W{news:comp.os.msdos.programmer}\i\c{comp.os.msdos.programmer}.
6385 \H{16c} Interfacing to 16-bit C Programs
6387 This section covers the basics of writing assembly routines that
6388 call, or are called from, C programs. To do this, you would
6389 typically write an assembly module as a \c{.OBJ} file, and link it
6390 with your C modules to produce a \i{mixed-language program}.
6393 \S{16cunder} External Symbol Names
6395 \I{C symbol names}\I{underscore, in C symbols}C compilers have the
6396 convention that the names of all global symbols (functions or data)
6397 they define are formed by prefixing an underscore to the name as it
6398 appears in the C program. So, for example, the function a C
6399 programmer thinks of as \c{printf} appears to an assembly language
6400 programmer as \c{_printf}. This means that in your assembly
6401 programs, you can define symbols without a leading underscore, and
6402 not have to worry about name clashes with C symbols.
6404 If you find the underscores inconvenient, you can define macros to
6405 replace the \c{GLOBAL} and \c{EXTERN} directives as follows:
6421 (These forms of the macros only take one argument at a time; a
6422 \c{%rep} construct could solve this.)
6424 If you then declare an external like this:
6428 then the macro will expand it as
6431 \c %define printf _printf
6433 Thereafter, you can reference \c{printf} as if it was a symbol, and
6434 the preprocessor will put the leading underscore on where necessary.
6436 The \c{cglobal} macro works similarly. You must use \c{cglobal}
6437 before defining the symbol in question, but you would have had to do
6438 that anyway if you used \c{GLOBAL}.
6440 Also see \k{opt-pfix}.
6442 \S{16cmodels} \i{Memory Models}
6444 NASM contains no mechanism to support the various C memory models
6445 directly; you have to keep track yourself of which one you are
6446 writing for. This means you have to keep track of the following
6449 \b In models using a single code segment (tiny, small and compact),
6450 functions are near. This means that function pointers, when stored
6451 in data segments or pushed on the stack as function arguments, are
6452 16 bits long and contain only an offset field (the \c{CS} register
6453 never changes its value, and always gives the segment part of the
6454 full function address), and that functions are called using ordinary
6455 near \c{CALL} instructions and return using \c{RETN} (which, in
6456 NASM, is synonymous with \c{RET} anyway). This means both that you
6457 should write your own routines to return with \c{RETN}, and that you
6458 should call external C routines with near \c{CALL} instructions.
6460 \b In models using more than one code segment (medium, large and
6461 huge), functions are far. This means that function pointers are 32
6462 bits long (consisting of a 16-bit offset followed by a 16-bit
6463 segment), and that functions are called using \c{CALL FAR} (or
6464 \c{CALL seg:offset}) and return using \c{RETF}. Again, you should
6465 therefore write your own routines to return with \c{RETF} and use
6466 \c{CALL FAR} to call external routines.
6468 \b In models using a single data segment (tiny, small and medium),
6469 data pointers are 16 bits long, containing only an offset field (the
6470 \c{DS} register doesn't change its value, and always gives the
6471 segment part of the full data item address).
6473 \b In models using more than one data segment (compact, large and
6474 huge), data pointers are 32 bits long, consisting of a 16-bit offset
6475 followed by a 16-bit segment. You should still be careful not to
6476 modify \c{DS} in your routines without restoring it afterwards, but
6477 \c{ES} is free for you to use to access the contents of 32-bit data
6478 pointers you are passed.
6480 \b The huge memory model allows single data items to exceed 64K in
6481 size. In all other memory models, you can access the whole of a data
6482 item just by doing arithmetic on the offset field of the pointer you
6483 are given, whether a segment field is present or not; in huge model,
6484 you have to be more careful of your pointer arithmetic.
6486 \b In most memory models, there is a \e{default} data segment, whose
6487 segment address is kept in \c{DS} throughout the program. This data
6488 segment is typically the same segment as the stack, kept in \c{SS},
6489 so that functions' local variables (which are stored on the stack)
6490 and global data items can both be accessed easily without changing
6491 \c{DS}. Particularly large data items are typically stored in other
6492 segments. However, some memory models (though not the standard
6493 ones, usually) allow the assumption that \c{SS} and \c{DS} hold the
6494 same value to be removed. Be careful about functions' local
6495 variables in this latter case.
6497 In models with a single code segment, the segment is called
6498 \i\c{_TEXT}, so your code segment must also go by this name in order
6499 to be linked into the same place as the main code segment. In models
6500 with a single data segment, or with a default data segment, it is
6504 \S{16cfunc} Function Definitions and Function Calls
6506 \I{functions, C calling convention}The \i{C calling convention} in
6507 16-bit programs is as follows. In the following description, the
6508 words \e{caller} and \e{callee} are used to denote the function
6509 doing the calling and the function which gets called.
6511 \b The caller pushes the function's parameters on the stack, one
6512 after another, in reverse order (right to left, so that the first
6513 argument specified to the function is pushed last).
6515 \b The caller then executes a \c{CALL} instruction to pass control
6516 to the callee. This \c{CALL} is either near or far depending on the
6519 \b The callee receives control, and typically (although this is not
6520 actually necessary, in functions which do not need to access their
6521 parameters) starts by saving the value of \c{SP} in \c{BP} so as to
6522 be able to use \c{BP} as a base pointer to find its parameters on
6523 the stack. However, the caller was probably doing this too, so part
6524 of the calling convention states that \c{BP} must be preserved by
6525 any C function. Hence the callee, if it is going to set up \c{BP} as
6526 a \i\e{frame pointer}, must push the previous value first.
6528 \b The callee may then access its parameters relative to \c{BP}.
6529 The word at \c{[BP]} holds the previous value of \c{BP} as it was
6530 pushed; the next word, at \c{[BP+2]}, holds the offset part of the
6531 return address, pushed implicitly by \c{CALL}. In a small-model
6532 (near) function, the parameters start after that, at \c{[BP+4]}; in
6533 a large-model (far) function, the segment part of the return address
6534 lives at \c{[BP+4]}, and the parameters begin at \c{[BP+6]}. The
6535 leftmost parameter of the function, since it was pushed last, is
6536 accessible at this offset from \c{BP}; the others follow, at
6537 successively greater offsets. Thus, in a function such as \c{printf}
6538 which takes a variable number of parameters, the pushing of the
6539 parameters in reverse order means that the function knows where to
6540 find its first parameter, which tells it the number and type of the
6543 \b The callee may also wish to decrease \c{SP} further, so as to
6544 allocate space on the stack for local variables, which will then be
6545 accessible at negative offsets from \c{BP}.
6547 \b The callee, if it wishes to return a value to the caller, should
6548 leave the value in \c{AL}, \c{AX} or \c{DX:AX} depending on the size
6549 of the value. Floating-point results are sometimes (depending on the
6550 compiler) returned in \c{ST0}.
6552 \b Once the callee has finished processing, it restores \c{SP} from
6553 \c{BP} if it had allocated local stack space, then pops the previous
6554 value of \c{BP}, and returns via \c{RETN} or \c{RETF} depending on
6557 \b When the caller regains control from the callee, the function
6558 parameters are still on the stack, so it typically adds an immediate
6559 constant to \c{SP} to remove them (instead of executing a number of
6560 slow \c{POP} instructions). Thus, if a function is accidentally
6561 called with the wrong number of parameters due to a prototype
6562 mismatch, the stack will still be returned to a sensible state since
6563 the caller, which \e{knows} how many parameters it pushed, does the
6566 It is instructive to compare this calling convention with that for
6567 Pascal programs (described in \k{16bpfunc}). Pascal has a simpler
6568 convention, since no functions have variable numbers of parameters.
6569 Therefore the callee knows how many parameters it should have been
6570 passed, and is able to deallocate them from the stack itself by
6571 passing an immediate argument to the \c{RET} or \c{RETF}
6572 instruction, so the caller does not have to do it. Also, the
6573 parameters are pushed in left-to-right order, not right-to-left,
6574 which means that a compiler can give better guarantees about
6575 sequence points without performance suffering.
6577 Thus, you would define a function in C style in the following way.
6578 The following example is for small model:
6585 \c sub sp,0x40 ; 64 bytes of local stack space
6586 \c mov bx,[bp+4] ; first parameter to function
6590 \c mov sp,bp ; undo "sub sp,0x40" above
6594 For a large-model function, you would replace \c{RET} by \c{RETF},
6595 and look for the first parameter at \c{[BP+6]} instead of
6596 \c{[BP+4]}. Of course, if one of the parameters is a pointer, then
6597 the offsets of \e{subsequent} parameters will change depending on
6598 the memory model as well: far pointers take up four bytes on the
6599 stack when passed as a parameter, whereas near pointers take up two.
6601 At the other end of the process, to call a C function from your
6602 assembly code, you would do something like this:
6606 \c ; and then, further down...
6608 \c push word [myint] ; one of my integer variables
6609 \c push word mystring ; pointer into my data segment
6611 \c add sp,byte 4 ; `byte' saves space
6613 \c ; then those data items...
6618 \c mystring db 'This number -> %d <- should be 1234',10,0
6620 This piece of code is the small-model assembly equivalent of the C
6623 \c int myint = 1234;
6624 \c printf("This number -> %d <- should be 1234\n", myint);
6626 In large model, the function-call code might look more like this. In
6627 this example, it is assumed that \c{DS} already holds the segment
6628 base of the segment \c{_DATA}. If not, you would have to initialize
6631 \c push word [myint]
6632 \c push word seg mystring ; Now push the segment, and...
6633 \c push word mystring ; ... offset of "mystring"
6637 The integer value still takes up one word on the stack, since large
6638 model does not affect the size of the \c{int} data type. The first
6639 argument (pushed last) to \c{printf}, however, is a data pointer,
6640 and therefore has to contain a segment and offset part. The segment
6641 should be stored second in memory, and therefore must be pushed
6642 first. (Of course, \c{PUSH DS} would have been a shorter instruction
6643 than \c{PUSH WORD SEG mystring}, if \c{DS} was set up as the above
6644 example assumed.) Then the actual call becomes a far call, since
6645 functions expect far calls in large model; and \c{SP} has to be
6646 increased by 6 rather than 4 afterwards to make up for the extra
6650 \S{16cdata} Accessing Data Items
6652 To get at the contents of C variables, or to declare variables which
6653 C can access, you need only declare the names as \c{GLOBAL} or
6654 \c{EXTERN}. (Again, the names require leading underscores, as stated
6655 in \k{16cunder}.) Thus, a C variable declared as \c{int i} can be
6656 accessed from assembler as
6662 And to declare your own integer variable which C programs can access
6663 as \c{extern int j}, you do this (making sure you are assembling in
6664 the \c{_DATA} segment, if necessary):
6670 To access a C array, you need to know the size of the components of
6671 the array. For example, \c{int} variables are two bytes long, so if
6672 a C program declares an array as \c{int a[10]}, you can access
6673 \c{a[3]} by coding \c{mov ax,[_a+6]}. (The byte offset 6 is obtained
6674 by multiplying the desired array index, 3, by the size of the array
6675 element, 2.) The sizes of the C base types in 16-bit compilers are:
6676 1 for \c{char}, 2 for \c{short} and \c{int}, 4 for \c{long} and
6677 \c{float}, and 8 for \c{double}.
6679 To access a C \i{data structure}, you need to know the offset from
6680 the base of the structure to the field you are interested in. You
6681 can either do this by converting the C structure definition into a
6682 NASM structure definition (using \i\c{STRUC}), or by calculating the
6683 one offset and using just that.
6685 To do either of these, you should read your C compiler's manual to
6686 find out how it organizes data structures. NASM gives no special
6687 alignment to structure members in its own \c{STRUC} macro, so you
6688 have to specify alignment yourself if the C compiler generates it.
6689 Typically, you might find that a structure like
6696 might be four bytes long rather than three, since the \c{int} field
6697 would be aligned to a two-byte boundary. However, this sort of
6698 feature tends to be a configurable option in the C compiler, either
6699 using command-line options or \c{#pragma} lines, so you have to find
6700 out how your own compiler does it.
6703 \S{16cmacro} \i\c{c16.mac}: Helper Macros for the 16-bit C Interface
6705 Included in the NASM archives, in the \I{misc subdirectory}\c{misc}
6706 directory, is a file \c{c16.mac} of macros. It defines three macros:
6707 \i\c{proc}, \i\c{arg} and \i\c{endproc}. These are intended to be
6708 used for C-style procedure definitions, and they automate a lot of
6709 the work involved in keeping track of the calling convention.
6711 (An alternative, TASM compatible form of \c{arg} is also now built
6712 into NASM's preprocessor. See \k{stackrel} for details.)
6714 An example of an assembly function using the macro set is given
6721 \c mov ax,[bp + %$i]
6722 \c mov bx,[bp + %$j]
6727 This defines \c{_nearproc} to be a procedure taking two arguments,
6728 the first (\c{i}) an integer and the second (\c{j}) a pointer to an
6729 integer. It returns \c{i + *j}.
6731 Note that the \c{arg} macro has an \c{EQU} as the first line of its
6732 expansion, and since the label before the macro call gets prepended
6733 to the first line of the expanded macro, the \c{EQU} works, defining
6734 \c{%$i} to be an offset from \c{BP}. A context-local variable is
6735 used, local to the context pushed by the \c{proc} macro and popped
6736 by the \c{endproc} macro, so that the same argument name can be used
6737 in later procedures. Of course, you don't \e{have} to do that.
6739 The macro set produces code for near functions (tiny, small and
6740 compact-model code) by default. You can have it generate far
6741 functions (medium, large and huge-model code) by means of coding
6742 \I\c{FARCODE}\c{%define FARCODE}. This changes the kind of return
6743 instruction generated by \c{endproc}, and also changes the starting
6744 point for the argument offsets. The macro set contains no intrinsic
6745 dependency on whether data pointers are far or not.
6747 \c{arg} can take an optional parameter, giving the size of the
6748 argument. If no size is given, 2 is assumed, since it is likely that
6749 many function parameters will be of type \c{int}.
6751 The large-model equivalent of the above function would look like this:
6759 \c mov ax,[bp + %$i]
6760 \c mov bx,[bp + %$j]
6761 \c mov es,[bp + %$j + 2]
6766 This makes use of the argument to the \c{arg} macro to define a
6767 parameter of size 4, because \c{j} is now a far pointer. When we
6768 load from \c{j}, we must load a segment and an offset.
6771 \H{16bp} Interfacing to \i{Borland Pascal} Programs
6773 Interfacing to Borland Pascal programs is similar in concept to
6774 interfacing to 16-bit C programs. The differences are:
6776 \b The leading underscore required for interfacing to C programs is
6777 not required for Pascal.
6779 \b The memory model is always large: functions are far, data
6780 pointers are far, and no data item can be more than 64K long.
6781 (Actually, some functions are near, but only those functions that
6782 are local to a Pascal unit and never called from outside it. All
6783 assembly functions that Pascal calls, and all Pascal functions that
6784 assembly routines are able to call, are far.) However, all static
6785 data declared in a Pascal program goes into the default data
6786 segment, which is the one whose segment address will be in \c{DS}
6787 when control is passed to your assembly code. The only things that
6788 do not live in the default data segment are local variables (they
6789 live in the stack segment) and dynamically allocated variables. All
6790 data \e{pointers}, however, are far.
6792 \b The function calling convention is different - described below.
6794 \b Some data types, such as strings, are stored differently.
6796 \b There are restrictions on the segment names you are allowed to
6797 use - Borland Pascal will ignore code or data declared in a segment
6798 it doesn't like the name of. The restrictions are described below.
6801 \S{16bpfunc} The Pascal Calling Convention
6803 \I{functions, Pascal calling convention}\I{Pascal calling
6804 convention}The 16-bit Pascal calling convention is as follows. In
6805 the following description, the words \e{caller} and \e{callee} are
6806 used to denote the function doing the calling and the function which
6809 \b The caller pushes the function's parameters on the stack, one
6810 after another, in normal order (left to right, so that the first
6811 argument specified to the function is pushed first).
6813 \b The caller then executes a far \c{CALL} instruction to pass
6814 control to the callee.
6816 \b The callee receives control, and typically (although this is not
6817 actually necessary, in functions which do not need to access their
6818 parameters) starts by saving the value of \c{SP} in \c{BP} so as to
6819 be able to use \c{BP} as a base pointer to find its parameters on
6820 the stack. However, the caller was probably doing this too, so part
6821 of the calling convention states that \c{BP} must be preserved by
6822 any function. Hence the callee, if it is going to set up \c{BP} as a
6823 \i{frame pointer}, must push the previous value first.
6825 \b The callee may then access its parameters relative to \c{BP}.
6826 The word at \c{[BP]} holds the previous value of \c{BP} as it was
6827 pushed. The next word, at \c{[BP+2]}, holds the offset part of the
6828 return address, and the next one at \c{[BP+4]} the segment part. The
6829 parameters begin at \c{[BP+6]}. The rightmost parameter of the
6830 function, since it was pushed last, is accessible at this offset
6831 from \c{BP}; the others follow, at successively greater offsets.
6833 \b The callee may also wish to decrease \c{SP} further, so as to
6834 allocate space on the stack for local variables, which will then be
6835 accessible at negative offsets from \c{BP}.
6837 \b The callee, if it wishes to return a value to the caller, should
6838 leave the value in \c{AL}, \c{AX} or \c{DX:AX} depending on the size
6839 of the value. Floating-point results are returned in \c{ST0}.
6840 Results of type \c{Real} (Borland's own custom floating-point data
6841 type, not handled directly by the FPU) are returned in \c{DX:BX:AX}.
6842 To return a result of type \c{String}, the caller pushes a pointer
6843 to a temporary string before pushing the parameters, and the callee
6844 places the returned string value at that location. The pointer is
6845 not a parameter, and should not be removed from the stack by the
6846 \c{RETF} instruction.
6848 \b Once the callee has finished processing, it restores \c{SP} from
6849 \c{BP} if it had allocated local stack space, then pops the previous
6850 value of \c{BP}, and returns via \c{RETF}. It uses the form of
6851 \c{RETF} with an immediate parameter, giving the number of bytes
6852 taken up by the parameters on the stack. This causes the parameters
6853 to be removed from the stack as a side effect of the return
6856 \b When the caller regains control from the callee, the function
6857 parameters have already been removed from the stack, so it needs to
6860 Thus, you would define a function in Pascal style, taking two
6861 \c{Integer}-type parameters, in the following way:
6867 \c sub sp,0x40 ; 64 bytes of local stack space
6868 \c mov bx,[bp+8] ; first parameter to function
6869 \c mov bx,[bp+6] ; second parameter to function
6873 \c mov sp,bp ; undo "sub sp,0x40" above
6875 \c retf 4 ; total size of params is 4
6877 At the other end of the process, to call a Pascal function from your
6878 assembly code, you would do something like this:
6882 \c ; and then, further down...
6884 \c push word seg mystring ; Now push the segment, and...
6885 \c push word mystring ; ... offset of "mystring"
6886 \c push word [myint] ; one of my variables
6887 \c call far SomeFunc
6889 This is equivalent to the Pascal code
6891 \c procedure SomeFunc(String: PChar; Int: Integer);
6892 \c SomeFunc(@mystring, myint);
6895 \S{16bpseg} Borland Pascal \I{segment names, Borland Pascal}Segment
6898 Since Borland Pascal's internal unit file format is completely
6899 different from \c{OBJ}, it only makes a very sketchy job of actually
6900 reading and understanding the various information contained in a
6901 real \c{OBJ} file when it links that in. Therefore an object file
6902 intended to be linked to a Pascal program must obey a number of
6905 \b Procedures and functions must be in a segment whose name is
6906 either \c{CODE}, \c{CSEG}, or something ending in \c{_TEXT}.
6908 \b initialized data must be in a segment whose name is either
6909 \c{CONST} or something ending in \c{_DATA}.
6911 \b Uninitialized data must be in a segment whose name is either
6912 \c{DATA}, \c{DSEG}, or something ending in \c{_BSS}.
6914 \b Any other segments in the object file are completely ignored.
6915 \c{GROUP} directives and segment attributes are also ignored.
6918 \S{16bpmacro} Using \i\c{c16.mac} With Pascal Programs
6920 The \c{c16.mac} macro package, described in \k{16cmacro}, can also
6921 be used to simplify writing functions to be called from Pascal
6922 programs, if you code \I\c{PASCAL}\c{%define PASCAL}. This
6923 definition ensures that functions are far (it implies
6924 \i\c{FARCODE}), and also causes procedure return instructions to be
6925 generated with an operand.
6927 Defining \c{PASCAL} does not change the code which calculates the
6928 argument offsets; you must declare your function's arguments in
6929 reverse order. For example:
6937 \c mov ax,[bp + %$i]
6938 \c mov bx,[bp + %$j]
6939 \c mov es,[bp + %$j + 2]
6944 This defines the same routine, conceptually, as the example in
6945 \k{16cmacro}: it defines a function taking two arguments, an integer
6946 and a pointer to an integer, which returns the sum of the integer
6947 and the contents of the pointer. The only difference between this
6948 code and the large-model C version is that \c{PASCAL} is defined
6949 instead of \c{FARCODE}, and that the arguments are declared in
6953 \C{32bit} Writing 32-bit Code (Unix, Win32, DJGPP)
6955 This chapter attempts to cover some of the common issues involved
6956 when writing 32-bit code, to run under \i{Win32} or Unix, or to be
6957 linked with C code generated by a Unix-style C compiler such as
6958 \i{DJGPP}. It covers how to write assembly code to interface with
6959 32-bit C routines, and how to write position-independent code for
6962 Almost all 32-bit code, and in particular all code running under
6963 \c{Win32}, \c{DJGPP} or any of the PC Unix variants, runs in \I{flat
6964 memory model}\e{flat} memory model. This means that the segment registers
6965 and paging have already been set up to give you the same 32-bit 4Gb
6966 address space no matter what segment you work relative to, and that
6967 you should ignore all segment registers completely. When writing
6968 flat-model application code, you never need to use a segment
6969 override or modify any segment register, and the code-section
6970 addresses you pass to \c{CALL} and \c{JMP} live in the same address
6971 space as the data-section addresses you access your variables by and
6972 the stack-section addresses you access local variables and procedure
6973 parameters by. Every address is 32 bits long and contains only an
6977 \H{32c} Interfacing to 32-bit C Programs
6979 A lot of the discussion in \k{16c}, about interfacing to 16-bit C
6980 programs, still applies when working in 32 bits. The absence of
6981 memory models or segmentation worries simplifies things a lot.
6984 \S{32cunder} External Symbol Names
6986 Most 32-bit C compilers share the convention used by 16-bit
6987 compilers, that the names of all global symbols (functions or data)
6988 they define are formed by prefixing an underscore to the name as it
6989 appears in the C program. However, not all of them do: the \c{ELF}
6990 specification states that C symbols do \e{not} have a leading
6991 underscore on their assembly-language names.
6993 The older Linux \c{a.out} C compiler, all \c{Win32} compilers,
6994 \c{DJGPP}, and \c{NetBSD} and \c{FreeBSD}, all use the leading
6995 underscore; for these compilers, the macros \c{cextern} and
6996 \c{cglobal}, as given in \k{16cunder}, will still work. For \c{ELF},
6997 though, the leading underscore should not be used.
6999 See also \k{opt-pfix}.
7001 \S{32cfunc} Function Definitions and Function Calls
7003 \I{functions, C calling convention}The \i{C calling convention}
7004 in 32-bit programs is as follows. In the following description,
7005 the words \e{caller} and \e{callee} are used to denote
7006 the function doing the calling and the function which gets called.
7008 \b The caller pushes the function's parameters on the stack, one
7009 after another, in reverse order (right to left, so that the first
7010 argument specified to the function is pushed last).
7012 \b The caller then executes a near \c{CALL} instruction to pass
7013 control to the callee.
7015 \b The callee receives control, and typically (although this is not
7016 actually necessary, in functions which do not need to access their
7017 parameters) starts by saving the value of \c{ESP} in \c{EBP} so as
7018 to be able to use \c{EBP} as a base pointer to find its parameters
7019 on the stack. However, the caller was probably doing this too, so
7020 part of the calling convention states that \c{EBP} must be preserved
7021 by any C function. Hence the callee, if it is going to set up
7022 \c{EBP} as a \i{frame pointer}, must push the previous value first.
7024 \b The callee may then access its parameters relative to \c{EBP}.
7025 The doubleword at \c{[EBP]} holds the previous value of \c{EBP} as
7026 it was pushed; the next doubleword, at \c{[EBP+4]}, holds the return
7027 address, pushed implicitly by \c{CALL}. The parameters start after
7028 that, at \c{[EBP+8]}. The leftmost parameter of the function, since
7029 it was pushed last, is accessible at this offset from \c{EBP}; the
7030 others follow, at successively greater offsets. Thus, in a function
7031 such as \c{printf} which takes a variable number of parameters, the
7032 pushing of the parameters in reverse order means that the function
7033 knows where to find its first parameter, which tells it the number
7034 and type of the remaining ones.
7036 \b The callee may also wish to decrease \c{ESP} further, so as to
7037 allocate space on the stack for local variables, which will then be
7038 accessible at negative offsets from \c{EBP}.
7040 \b The callee, if it wishes to return a value to the caller, should
7041 leave the value in \c{AL}, \c{AX} or \c{EAX} depending on the size
7042 of the value. Floating-point results are typically returned in
7045 \b Once the callee has finished processing, it restores \c{ESP} from
7046 \c{EBP} if it had allocated local stack space, then pops the previous
7047 value of \c{EBP}, and returns via \c{RET} (equivalently, \c{RETN}).
7049 \b When the caller regains control from the callee, the function
7050 parameters are still on the stack, so it typically adds an immediate
7051 constant to \c{ESP} to remove them (instead of executing a number of
7052 slow \c{POP} instructions). Thus, if a function is accidentally
7053 called with the wrong number of parameters due to a prototype
7054 mismatch, the stack will still be returned to a sensible state since
7055 the caller, which \e{knows} how many parameters it pushed, does the
7058 There is an alternative calling convention used by Win32 programs
7059 for Windows API calls, and also for functions called \e{by} the
7060 Windows API such as window procedures: they follow what Microsoft
7061 calls the \c{__stdcall} convention. This is slightly closer to the
7062 Pascal convention, in that the callee clears the stack by passing a
7063 parameter to the \c{RET} instruction. However, the parameters are
7064 still pushed in right-to-left order.
7066 Thus, you would define a function in C style in the following way:
7073 \c sub esp,0x40 ; 64 bytes of local stack space
7074 \c mov ebx,[ebp+8] ; first parameter to function
7078 \c leave ; mov esp,ebp / pop ebp
7081 At the other end of the process, to call a C function from your
7082 assembly code, you would do something like this:
7086 \c ; and then, further down...
7088 \c push dword [myint] ; one of my integer variables
7089 \c push dword mystring ; pointer into my data segment
7091 \c add esp,byte 8 ; `byte' saves space
7093 \c ; then those data items...
7098 \c mystring db 'This number -> %d <- should be 1234',10,0
7100 This piece of code is the assembly equivalent of the C code
7102 \c int myint = 1234;
7103 \c printf("This number -> %d <- should be 1234\n", myint);
7106 \S{32cdata} Accessing Data Items
7108 To get at the contents of C variables, or to declare variables which
7109 C can access, you need only declare the names as \c{GLOBAL} or
7110 \c{EXTERN}. (Again, the names require leading underscores, as stated
7111 in \k{32cunder}.) Thus, a C variable declared as \c{int i} can be
7112 accessed from assembler as
7117 And to declare your own integer variable which C programs can access
7118 as \c{extern int j}, you do this (making sure you are assembling in
7119 the \c{_DATA} segment, if necessary):
7124 To access a C array, you need to know the size of the components of
7125 the array. For example, \c{int} variables are four bytes long, so if
7126 a C program declares an array as \c{int a[10]}, you can access
7127 \c{a[3]} by coding \c{mov ax,[_a+12]}. (The byte offset 12 is obtained
7128 by multiplying the desired array index, 3, by the size of the array
7129 element, 4.) The sizes of the C base types in 32-bit compilers are:
7130 1 for \c{char}, 2 for \c{short}, 4 for \c{int}, \c{long} and
7131 \c{float}, and 8 for \c{double}. Pointers, being 32-bit addresses,
7132 are also 4 bytes long.
7134 To access a C \i{data structure}, you need to know the offset from
7135 the base of the structure to the field you are interested in. You
7136 can either do this by converting the C structure definition into a
7137 NASM structure definition (using \c{STRUC}), or by calculating the
7138 one offset and using just that.
7140 To do either of these, you should read your C compiler's manual to
7141 find out how it organizes data structures. NASM gives no special
7142 alignment to structure members in its own \i\c{STRUC} macro, so you
7143 have to specify alignment yourself if the C compiler generates it.
7144 Typically, you might find that a structure like
7151 might be eight bytes long rather than five, since the \c{int} field
7152 would be aligned to a four-byte boundary. However, this sort of
7153 feature is sometimes a configurable option in the C compiler, either
7154 using command-line options or \c{#pragma} lines, so you have to find
7155 out how your own compiler does it.
7158 \S{32cmacro} \i\c{c32.mac}: Helper Macros for the 32-bit C Interface
7160 Included in the NASM archives, in the \I{misc directory}\c{misc}
7161 directory, is a file \c{c32.mac} of macros. It defines three macros:
7162 \i\c{proc}, \i\c{arg} and \i\c{endproc}. These are intended to be
7163 used for C-style procedure definitions, and they automate a lot of
7164 the work involved in keeping track of the calling convention.
7166 An example of an assembly function using the macro set is given
7173 \c mov eax,[ebp + %$i]
7174 \c mov ebx,[ebp + %$j]
7179 This defines \c{_proc32} to be a procedure taking two arguments, the
7180 first (\c{i}) an integer and the second (\c{j}) a pointer to an
7181 integer. It returns \c{i + *j}.
7183 Note that the \c{arg} macro has an \c{EQU} as the first line of its
7184 expansion, and since the label before the macro call gets prepended
7185 to the first line of the expanded macro, the \c{EQU} works, defining
7186 \c{%$i} to be an offset from \c{BP}. A context-local variable is
7187 used, local to the context pushed by the \c{proc} macro and popped
7188 by the \c{endproc} macro, so that the same argument name can be used
7189 in later procedures. Of course, you don't \e{have} to do that.
7191 \c{arg} can take an optional parameter, giving the size of the
7192 argument. If no size is given, 4 is assumed, since it is likely that
7193 many function parameters will be of type \c{int} or pointers.
7196 \H{picdll} Writing NetBSD/FreeBSD/OpenBSD and Linux/ELF \i{Shared
7199 \c{ELF} replaced the older \c{a.out} object file format under Linux
7200 because it contains support for \i{position-independent code}
7201 (\i{PIC}), which makes writing shared libraries much easier. NASM
7202 supports the \c{ELF} position-independent code features, so you can
7203 write Linux \c{ELF} shared libraries in NASM.
7205 \i{NetBSD}, and its close cousins \i{FreeBSD} and \i{OpenBSD}, take
7206 a different approach by hacking PIC support into the \c{a.out}
7207 format. NASM supports this as the \i\c{aoutb} output format, so you
7208 can write \i{BSD} shared libraries in NASM too.
7210 The operating system loads a PIC shared library by memory-mapping
7211 the library file at an arbitrarily chosen point in the address space
7212 of the running process. The contents of the library's code section
7213 must therefore not depend on where it is loaded in memory.
7215 Therefore, you cannot get at your variables by writing code like
7218 \c mov eax,[myvar] ; WRONG
7220 Instead, the linker provides an area of memory called the
7221 \i\e{global offset table}, or \i{GOT}; the GOT is situated at a
7222 constant distance from your library's code, so if you can find out
7223 where your library is loaded (which is typically done using a
7224 \c{CALL} and \c{POP} combination), you can obtain the address of the
7225 GOT, and you can then load the addresses of your variables out of
7226 linker-generated entries in the GOT.
7228 The \e{data} section of a PIC shared library does not have these
7229 restrictions: since the data section is writable, it has to be
7230 copied into memory anyway rather than just paged in from the library
7231 file, so as long as it's being copied it can be relocated too. So
7232 you can put ordinary types of relocation in the data section without
7233 too much worry (but see \k{picglobal} for a caveat).
7236 \S{picgot} Obtaining the Address of the GOT
7238 Each code module in your shared library should define the GOT as an
7241 \c extern _GLOBAL_OFFSET_TABLE_ ; in ELF
7242 \c extern __GLOBAL_OFFSET_TABLE_ ; in BSD a.out
7244 At the beginning of any function in your shared library which plans
7245 to access your data or BSS sections, you must first calculate the
7246 address of the GOT. This is typically done by writing the function
7255 \c add ebx,_GLOBAL_OFFSET_TABLE_+$$-.get_GOT wrt ..gotpc
7257 \c ; the function body comes here
7264 (For BSD, again, the symbol \c{_GLOBAL_OFFSET_TABLE} requires a
7265 second leading underscore.)
7267 The first two lines of this function are simply the standard C
7268 prologue to set up a stack frame, and the last three lines are
7269 standard C function epilogue. The third line, and the fourth to last
7270 line, save and restore the \c{EBX} register, because PIC shared
7271 libraries use this register to store the address of the GOT.
7273 The interesting bit is the \c{CALL} instruction and the following
7274 two lines. The \c{CALL} and \c{POP} combination obtains the address
7275 of the label \c{.get_GOT}, without having to know in advance where
7276 the program was loaded (since the \c{CALL} instruction is encoded
7277 relative to the current position). The \c{ADD} instruction makes use
7278 of one of the special PIC relocation types: \i{GOTPC relocation}.
7279 With the \i\c{WRT ..gotpc} qualifier specified, the symbol
7280 referenced (here \c{_GLOBAL_OFFSET_TABLE_}, the special symbol
7281 assigned to the GOT) is given as an offset from the beginning of the
7282 section. (Actually, \c{ELF} encodes it as the offset from the operand
7283 field of the \c{ADD} instruction, but NASM simplifies this
7284 deliberately, so you do things the same way for both \c{ELF} and
7285 \c{BSD}.) So the instruction then \e{adds} the beginning of the section,
7286 to get the real address of the GOT, and subtracts the value of
7287 \c{.get_GOT} which it knows is in \c{EBX}. Therefore, by the time
7288 that instruction has finished, \c{EBX} contains the address of the GOT.
7290 If you didn't follow that, don't worry: it's never necessary to
7291 obtain the address of the GOT by any other means, so you can put
7292 those three instructions into a macro and safely ignore them:
7299 \c add ebx,_GLOBAL_OFFSET_TABLE_+$$-%%getgot wrt ..gotpc
7303 \S{piclocal} Finding Your Local Data Items
7305 Having got the GOT, you can then use it to obtain the addresses of
7306 your data items. Most variables will reside in the sections you have
7307 declared; they can be accessed using the \I{GOTOFF
7308 relocation}\c{..gotoff} special \I\c{WRT ..gotoff}\c{WRT} type. The
7309 way this works is like this:
7311 \c lea eax,[ebx+myvar wrt ..gotoff]
7313 The expression \c{myvar wrt ..gotoff} is calculated, when the shared
7314 library is linked, to be the offset to the local variable \c{myvar}
7315 from the beginning of the GOT. Therefore, adding it to \c{EBX} as
7316 above will place the real address of \c{myvar} in \c{EAX}.
7318 If you declare variables as \c{GLOBAL} without specifying a size for
7319 them, they are shared between code modules in the library, but do
7320 not get exported from the library to the program that loaded it.
7321 They will still be in your ordinary data and BSS sections, so you
7322 can access them in the same way as local variables, using the above
7323 \c{..gotoff} mechanism.
7325 Note that due to a peculiarity of the way BSD \c{a.out} format
7326 handles this relocation type, there must be at least one non-local
7327 symbol in the same section as the address you're trying to access.
7330 \S{picextern} Finding External and Common Data Items
7332 If your library needs to get at an external variable (external to
7333 the \e{library}, not just to one of the modules within it), you must
7334 use the \I{GOT relocations}\I\c{WRT ..got}\c{..got} type to get at
7335 it. The \c{..got} type, instead of giving you the offset from the
7336 GOT base to the variable, gives you the offset from the GOT base to
7337 a GOT \e{entry} containing the address of the variable. The linker
7338 will set up this GOT entry when it builds the library, and the
7339 dynamic linker will place the correct address in it at load time. So
7340 to obtain the address of an external variable \c{extvar} in \c{EAX},
7343 \c mov eax,[ebx+extvar wrt ..got]
7345 This loads the address of \c{extvar} out of an entry in the GOT. The
7346 linker, when it builds the shared library, collects together every
7347 relocation of type \c{..got}, and builds the GOT so as to ensure it
7348 has every necessary entry present.
7350 Common variables must also be accessed in this way.
7353 \S{picglobal} Exporting Symbols to the Library User
7355 If you want to export symbols to the user of the library, you have
7356 to declare whether they are functions or data, and if they are data,
7357 you have to give the size of the data item. This is because the
7358 dynamic linker has to build \I{PLT}\i{procedure linkage table}
7359 entries for any exported functions, and also moves exported data
7360 items away from the library's data section in which they were
7363 So to export a function to users of the library, you must use
7365 \c global func:function ; declare it as a function
7371 And to export a data item such as an array, you would have to code
7373 \c global array:data array.end-array ; give the size too
7378 Be careful: If you export a variable to the library user, by
7379 declaring it as \c{GLOBAL} and supplying a size, the variable will
7380 end up living in the data section of the main program, rather than
7381 in your library's data section, where you declared it. So you will
7382 have to access your own global variable with the \c{..got} mechanism
7383 rather than \c{..gotoff}, as if it were external (which,
7384 effectively, it has become).
7386 Equally, if you need to store the address of an exported global in
7387 one of your data sections, you can't do it by means of the standard
7390 \c dataptr: dd global_data_item ; WRONG
7392 NASM will interpret this code as an ordinary relocation, in which
7393 \c{global_data_item} is merely an offset from the beginning of the
7394 \c{.data} section (or whatever); so this reference will end up
7395 pointing at your data section instead of at the exported global
7396 which resides elsewhere.
7398 Instead of the above code, then, you must write
7400 \c dataptr: dd global_data_item wrt ..sym
7402 which makes use of the special \c{WRT} type \I\c{WRT ..sym}\c{..sym}
7403 to instruct NASM to search the symbol table for a particular symbol
7404 at that address, rather than just relocating by section base.
7406 Either method will work for functions: referring to one of your
7407 functions by means of
7409 \c funcptr: dd my_function
7411 will give the user the address of the code you wrote, whereas
7413 \c funcptr: dd my_function wrt ..sym
7415 will give the address of the procedure linkage table for the
7416 function, which is where the calling program will \e{believe} the
7417 function lives. Either address is a valid way to call the function.
7420 \S{picproc} Calling Procedures Outside the Library
7422 Calling procedures outside your shared library has to be done by
7423 means of a \i\e{procedure linkage table}, or \i{PLT}. The PLT is
7424 placed at a known offset from where the library is loaded, so the
7425 library code can make calls to the PLT in a position-independent
7426 way. Within the PLT there is code to jump to offsets contained in
7427 the GOT, so function calls to other shared libraries or to routines
7428 in the main program can be transparently passed off to their real
7431 To call an external routine, you must use another special PIC
7432 relocation type, \I{PLT relocations}\i\c{WRT ..plt}. This is much
7433 easier than the GOT-based ones: you simply replace calls such as
7434 \c{CALL printf} with the PLT-relative version \c{CALL printf WRT
7438 \S{link} Generating the Library File
7440 Having written some code modules and assembled them to \c{.o} files,
7441 you then generate your shared library with a command such as
7443 \c ld -shared -o library.so module1.o module2.o # for ELF
7444 \c ld -Bshareable -o library.so module1.o module2.o # for BSD
7446 For ELF, if your shared library is going to reside in system
7447 directories such as \c{/usr/lib} or \c{/lib}, it is usually worth
7448 using the \i\c{-soname} flag to the linker, to store the final
7449 library file name, with a version number, into the library:
7451 \c ld -shared -soname library.so.1 -o library.so.1.2 *.o
7453 You would then copy \c{library.so.1.2} into the library directory,
7454 and create \c{library.so.1} as a symbolic link to it.
7457 \C{mixsize} Mixing 16 and 32 Bit Code
7459 This chapter tries to cover some of the issues, largely related to
7460 unusual forms of addressing and jump instructions, encountered when
7461 writing operating system code such as protected-mode initialisation
7462 routines, which require code that operates in mixed segment sizes,
7463 such as code in a 16-bit segment trying to modify data in a 32-bit
7464 one, or jumps between different-size segments.
7467 \H{mixjump} Mixed-Size Jumps\I{jumps, mixed-size}
7469 \I{operating system, writing}\I{writing operating systems}The most
7470 common form of \i{mixed-size instruction} is the one used when
7471 writing a 32-bit OS: having done your setup in 16-bit mode, such as
7472 loading the kernel, you then have to boot it by switching into
7473 protected mode and jumping to the 32-bit kernel start address. In a
7474 fully 32-bit OS, this tends to be the \e{only} mixed-size
7475 instruction you need, since everything before it can be done in pure
7476 16-bit code, and everything after it can be pure 32-bit.
7478 This jump must specify a 48-bit far address, since the target
7479 segment is a 32-bit one. However, it must be assembled in a 16-bit
7480 segment, so just coding, for example,
7482 \c jmp 0x1234:0x56789ABC ; wrong!
7484 will not work, since the offset part of the address will be
7485 truncated to \c{0x9ABC} and the jump will be an ordinary 16-bit far
7488 The Linux kernel setup code gets round the inability of \c{as86} to
7489 generate the required instruction by coding it manually, using
7490 \c{DB} instructions. NASM can go one better than that, by actually
7491 generating the right instruction itself. Here's how to do it right:
7493 \c jmp dword 0x1234:0x56789ABC ; right
7495 \I\c{JMP DWORD}The \c{DWORD} prefix (strictly speaking, it should
7496 come \e{after} the colon, since it is declaring the \e{offset} field
7497 to be a doubleword; but NASM will accept either form, since both are
7498 unambiguous) forces the offset part to be treated as far, in the
7499 assumption that you are deliberately writing a jump from a 16-bit
7500 segment to a 32-bit one.
7502 You can do the reverse operation, jumping from a 32-bit segment to a
7503 16-bit one, by means of the \c{WORD} prefix:
7505 \c jmp word 0x8765:0x4321 ; 32 to 16 bit
7507 If the \c{WORD} prefix is specified in 16-bit mode, or the \c{DWORD}
7508 prefix in 32-bit mode, they will be ignored, since each is
7509 explicitly forcing NASM into a mode it was in anyway.
7512 \H{mixaddr} Addressing Between Different-Size Segments\I{addressing,
7513 mixed-size}\I{mixed-size addressing}
7515 If your OS is mixed 16 and 32-bit, or if you are writing a DOS
7516 extender, you are likely to have to deal with some 16-bit segments
7517 and some 32-bit ones. At some point, you will probably end up
7518 writing code in a 16-bit segment which has to access data in a
7519 32-bit segment, or vice versa.
7521 If the data you are trying to access in a 32-bit segment lies within
7522 the first 64K of the segment, you may be able to get away with using
7523 an ordinary 16-bit addressing operation for the purpose; but sooner
7524 or later, you will want to do 32-bit addressing from 16-bit mode.
7526 The easiest way to do this is to make sure you use a register for
7527 the address, since any effective address containing a 32-bit
7528 register is forced to be a 32-bit address. So you can do
7530 \c mov eax,offset_into_32_bit_segment_specified_by_fs
7531 \c mov dword [fs:eax],0x11223344
7533 This is fine, but slightly cumbersome (since it wastes an
7534 instruction and a register) if you already know the precise offset
7535 you are aiming at. The x86 architecture does allow 32-bit effective
7536 addresses to specify nothing but a 4-byte offset, so why shouldn't
7537 NASM be able to generate the best instruction for the purpose?
7539 It can. As in \k{mixjump}, you need only prefix the address with the
7540 \c{DWORD} keyword, and it will be forced to be a 32-bit address:
7542 \c mov dword [fs:dword my_offset],0x11223344
7544 Also as in \k{mixjump}, NASM is not fussy about whether the
7545 \c{DWORD} prefix comes before or after the segment override, so
7546 arguably a nicer-looking way to code the above instruction is
7548 \c mov dword [dword fs:my_offset],0x11223344
7550 Don't confuse the \c{DWORD} prefix \e{outside} the square brackets,
7551 which controls the size of the data stored at the address, with the
7552 one \c{inside} the square brackets which controls the length of the
7553 address itself. The two can quite easily be different:
7555 \c mov word [dword 0x12345678],0x9ABC
7557 This moves 16 bits of data to an address specified by a 32-bit
7560 You can also specify \c{WORD} or \c{DWORD} prefixes along with the
7561 \c{FAR} prefix to indirect far jumps or calls. For example:
7563 \c call dword far [fs:word 0x4321]
7565 This instruction contains an address specified by a 16-bit offset;
7566 it loads a 48-bit far pointer from that (16-bit segment and 32-bit
7567 offset), and calls that address.
7570 \H{mixother} Other Mixed-Size Instructions
7572 The other way you might want to access data might be using the
7573 string instructions (\c{LODSx}, \c{STOSx} and so on) or the
7574 \c{XLATB} instruction. These instructions, since they take no
7575 parameters, might seem to have no easy way to make them perform
7576 32-bit addressing when assembled in a 16-bit segment.
7578 This is the purpose of NASM's \i\c{a16}, \i\c{a32} and \i\c{a64} prefixes. If
7579 you are coding \c{LODSB} in a 16-bit segment but it is supposed to
7580 be accessing a string in a 32-bit segment, you should load the
7581 desired address into \c{ESI} and then code
7585 The prefix forces the addressing size to 32 bits, meaning that
7586 \c{LODSB} loads from \c{[DS:ESI]} instead of \c{[DS:SI]}. To access
7587 a string in a 16-bit segment when coding in a 32-bit one, the
7588 corresponding \c{a16} prefix can be used.
7590 The \c{a16}, \c{a32} and \c{a64} prefixes can be applied to any instruction
7591 in NASM's instruction table, but most of them can generate all the
7592 useful forms without them. The prefixes are necessary only for
7593 instructions with implicit addressing:
7594 \# \c{CMPSx} (\k{insCMPSB}),
7595 \# \c{SCASx} (\k{insSCASB}), \c{LODSx} (\k{insLODSB}), \c{STOSx}
7596 \# (\k{insSTOSB}), \c{MOVSx} (\k{insMOVSB}), \c{INSx} (\k{insINSB}),
7597 \# \c{OUTSx} (\k{insOUTSB}), and \c{XLATB} (\k{insXLATB}).
7598 \c{CMPSx}, \c{SCASx}, \c{LODSx}, \c{STOSx}, \c{MOVSx}, \c{INSx},
7599 \c{OUTSx}, and \c{XLATB}.
7601 various push and pop instructions (\c{PUSHA} and \c{POPF} as well as
7602 the more usual \c{PUSH} and \c{POP}) can accept \c{a16}, \c{a32} or \c{a64}
7603 prefixes to force a particular one of \c{SP}, \c{ESP} or \c{RSP} to be used
7604 as a stack pointer, in case the stack segment in use is a different
7605 size from the code segment.
7607 \c{PUSH} and \c{POP}, when applied to segment registers in 32-bit
7608 mode, also have the slightly odd behaviour that they push and pop 4
7609 bytes at a time, of which the top two are ignored and the bottom two
7610 give the value of the segment register being manipulated. To force
7611 the 16-bit behaviour of segment-register push and pop instructions,
7612 you can use the operand-size prefix \i\c{o16}:
7617 This code saves a doubleword of stack space by fitting two segment
7618 registers into the space which would normally be consumed by pushing
7621 (You can also use the \i\c{o32} prefix to force the 32-bit behaviour
7622 when in 16-bit mode, but this seems less useful.)
7625 \C{64bit} Writing 64-bit Code (Unix, Win64)
7627 This chapter attempts to cover some of the common issues involved when
7628 writing 64-bit code, to run under \i{Win64} or Unix. It covers how to
7629 write assembly code to interface with 64-bit C routines, and how to
7630 write position-independent code for shared libraries.
7632 All 64-bit code uses a flat memory model, since segmentation is not
7633 available in 64-bit mode. The one exception is the \c{FS} and \c{GS}
7634 registers, which still add their bases.
7636 Position independence in 64-bit mode is significantly simpler, since
7637 the processor supports \c{RIP}-relative addressing directly; see the
7638 \c{REL} keyword (\k{effaddr}). On most 64-bit platforms, it is
7639 probably desirable to make that the default, using the directive
7640 \c{DEFAULT REL} (\k{default}).
7642 64-bit programming is relatively similar to 32-bit programming, but
7643 of course pointers are 64 bits long; additionally, all existing
7644 platforms pass arguments in registers rather than on the stack.
7645 Furthermore, 64-bit platforms use SSE2 by default for floating point.
7646 Please see the ABI documentation for your platform.
7648 64-bit platforms differ in the sizes of the fundamental datatypes, not
7649 just from 32-bit platforms but from each other. If a specific size
7650 data type is desired, it is probably best to use the types defined in
7651 the Standard C header \c{<inttypes.h>}.
7653 In 64-bit mode, the default instruction size is still 32 bits. When
7654 loading a value into a 32-bit register (but not an 8- or 16-bit
7655 register), the upper 32 bits of the corresponding 64-bit register are
7658 \H{reg64} Register Names in 64-bit Mode
7660 NASM uses the following names for general-purpose registers in 64-bit
7661 mode, for 8-, 16-, 32- and 64-bit references, respectively:
7663 \c AL/AH, CL/CH, DL/DH, BL/BH, SPL, BPL, SIL, DIL, R8B-R15B
7664 \c AX, CX, DX, BX, SP, BP, SI, DI, R8W-R15W
7665 \c EAX, ECX, EDX, EBX, ESP, EBP, ESI, EDI, R8D-R15D
7666 \c RAX, RCX, RDX, RBX, RSP, RBP, RSI, RDI, R8-R15
7668 This is consistent with the AMD documentation and most other
7669 assemblers. The Intel documentation, however, uses the names
7670 \c{R8L-R15L} for 8-bit references to the higher registers. It is
7671 possible to use those names by definiting them as macros; similarly,
7672 if one wants to use numeric names for the low 8 registers, define them
7673 as macros. The standard macro package \c{altreg} (see \k{pkg_altreg})
7674 can be used for this purpose.
7676 \H{id64} Immediates and Displacements in 64-bit Mode
7678 In 64-bit mode, immediates and displacements are generally only 32
7679 bits wide. NASM will therefore truncate most displacements and
7680 immediates to 32 bits.
7682 The only instruction which takes a full \i{64-bit immediate} is:
7686 NASM will produce this instruction whenever the programmer uses
7687 \c{MOV} with an immediate into a 64-bit register. If this is not
7688 desirable, simply specify the equivalent 32-bit register, which will
7689 be automatically zero-extended by the processor, or specify the
7690 immediate as \c{DWORD}:
7692 \c mov rax,foo ; 64-bit immediate
7693 \c mov rax,qword foo ; (identical)
7694 \c mov eax,foo ; 32-bit immediate, zero-extended
7695 \c mov rax,dword foo ; 32-bit immediate, sign-extended
7697 The length of these instructions are 10, 5 and 7 bytes, respectively.
7699 The only instructions which take a full \I{64-bit displacement}64-bit
7700 \e{displacement} is loading or storing, using \c{MOV}, \c{AL}, \c{AX},
7701 \c{EAX} or \c{RAX} (but no other registers) to an absolute 64-bit address.
7702 Since this is a relatively rarely used instruction (64-bit code generally uses
7703 relative addressing), the programmer has to explicitly declare the
7704 displacement size as \c{QWORD}:
7708 \c mov eax,[foo] ; 32-bit absolute disp, sign-extended
7709 \c mov eax,[a32 foo] ; 32-bit absolute disp, zero-extended
7710 \c mov eax,[qword foo] ; 64-bit absolute disp
7714 \c mov eax,[foo] ; 32-bit relative disp
7715 \c mov eax,[a32 foo] ; d:o, address truncated to 32 bits(!)
7716 \c mov eax,[qword foo] ; error
7717 \c mov eax,[abs qword foo] ; 64-bit absolute disp
7719 A sign-extended absolute displacement can access from -2 GB to +2 GB;
7720 a zero-extended absolute displacement can access from 0 to 4 GB.
7722 \H{unix64} Interfacing to 64-bit C Programs (Unix)
7724 On Unix, the 64-bit ABI is defined by the document:
7726 \W{http://www.nasm.us/links/unix64abi}\c{http://www.nasm.us/links/unix64abi}
7728 Although written for AT&T-syntax assembly, the concepts apply equally
7729 well for NASM-style assembly. What follows is a simplified summary.
7731 The first six integer arguments (from the left) are passed in \c{RDI},
7732 \c{RSI}, \c{RDX}, \c{RCX}, \c{R8}, and \c{R9}, in that order.
7733 Additional integer arguments are passed on the stack. These
7734 registers, plus \c{RAX}, \c{R10} and \c{R11} are destroyed by function
7735 calls, and thus are available for use by the function without saving.
7737 Integer return values are passed in \c{RAX} and \c{RDX}, in that order.
7739 Floating point is done using SSE registers, except for \c{long
7740 double}. Floating-point arguments are passed in \c{XMM0} to \c{XMM7};
7741 return is \c{XMM0} and \c{XMM1}. \c{long double} are passed on the
7742 stack, and returned in \c{ST0} and \c{ST1}.
7744 All SSE and x87 registers are destroyed by function calls.
7746 On 64-bit Unix, \c{long} is 64 bits.
7748 Integer and SSE register arguments are counted separately, so for the case of
7750 \c void foo(long a, double b, int c)
7752 \c{a} is passed in \c{RDI}, \c{b} in \c{XMM0}, and \c{c} in \c{ESI}.
7754 \H{win64} Interfacing to 64-bit C Programs (Win64)
7756 The Win64 ABI is described at:
7758 \W{http://www.nasm.us/links/win64abi}\c{http://www.nasm.us/links/win64abi}
7760 What follows is a simplified summary.
7762 The first four integer arguments are passed in \c{RCX}, \c{RDX},
7763 \c{R8} and \c{R9}, in that order. Additional integer arguments are
7764 passed on the stack. These registers, plus \c{RAX}, \c{R10} and
7765 \c{R11} are destroyed by function calls, and thus are available for
7766 use by the function without saving.
7768 Integer return values are passed in \c{RAX} only.
7770 Floating point is done using SSE registers, except for \c{long
7771 double}. Floating-point arguments are passed in \c{XMM0} to \c{XMM3};
7772 return is \c{XMM0} only.
7774 On Win64, \c{long} is 32 bits; \c{long long} or \c{_int64} is 64 bits.
7776 Integer and SSE register arguments are counted together, so for the case of
7778 \c void foo(long long a, double b, int c)
7780 \c{a} is passed in \c{RCX}, \c{b} in \c{XMM1}, and \c{c} in \c{R8D}.
7782 \C{trouble} Troubleshooting
7784 This chapter describes some of the common problems that users have
7785 been known to encounter with NASM, and answers them. It also gives
7786 instructions for reporting bugs in NASM if you find a difficulty
7787 that isn't listed here.
7790 \H{problems} Common Problems
7792 \S{inefficient} NASM Generates \i{Inefficient Code}
7794 We sometimes get `bug' reports about NASM generating inefficient, or
7795 even `wrong', code on instructions such as \c{ADD ESP,8}. This is a
7796 deliberate design feature, connected to predictability of output:
7797 NASM, on seeing \c{ADD ESP,8}, will generate the form of the
7798 instruction which leaves room for a 32-bit offset. You need to code
7799 \I\c{BYTE}\c{ADD ESP,BYTE 8} if you want the space-efficient form of
7800 the instruction. This isn't a bug, it's user error: if you prefer to
7801 have NASM produce the more efficient code automatically enable
7802 optimization with the \c{-O} option (see \k{opt-O}).
7805 \S{jmprange} My Jumps are Out of Range\I{out of range, jumps}
7807 Similarly, people complain that when they issue \i{conditional
7808 jumps} (which are \c{SHORT} by default) that try to jump too far,
7809 NASM reports `short jump out of range' instead of making the jumps
7812 This, again, is partly a predictability issue, but in fact has a
7813 more practical reason as well. NASM has no means of being told what
7814 type of processor the code it is generating will be run on; so it
7815 cannot decide for itself that it should generate \i\c{Jcc NEAR} type
7816 instructions, because it doesn't know that it's working for a 386 or
7817 above. Alternatively, it could replace the out-of-range short
7818 \c{JNE} instruction with a very short \c{JE} instruction that jumps
7819 over a \c{JMP NEAR}; this is a sensible solution for processors
7820 below a 386, but hardly efficient on processors which have good
7821 branch prediction \e{and} could have used \c{JNE NEAR} instead. So,
7822 once again, it's up to the user, not the assembler, to decide what
7823 instructions should be generated. See \k{opt-O}.
7826 \S{proborg} \i\c{ORG} Doesn't Work
7828 People writing \i{boot sector} programs in the \c{bin} format often
7829 complain that \c{ORG} doesn't work the way they'd like: in order to
7830 place the \c{0xAA55} signature word at the end of a 512-byte boot
7831 sector, people who are used to MASM tend to code
7835 \c ; some boot sector code
7840 This is not the intended use of the \c{ORG} directive in NASM, and
7841 will not work. The correct way to solve this problem in NASM is to
7842 use the \i\c{TIMES} directive, like this:
7846 \c ; some boot sector code
7848 \c TIMES 510-($-$$) DB 0
7851 The \c{TIMES} directive will insert exactly enough zero bytes into
7852 the output to move the assembly point up to 510. This method also
7853 has the advantage that if you accidentally fill your boot sector too
7854 full, NASM will catch the problem at assembly time and report it, so
7855 you won't end up with a boot sector that you have to disassemble to
7856 find out what's wrong with it.
7859 \S{probtimes} \i\c{TIMES} Doesn't Work
7861 The other common problem with the above code is people who write the
7866 by reasoning that \c{$} should be a pure number, just like 510, so
7867 the difference between them is also a pure number and can happily be
7870 NASM is a \e{modular} assembler: the various component parts are
7871 designed to be easily separable for re-use, so they don't exchange
7872 information unnecessarily. In consequence, the \c{bin} output
7873 format, even though it has been told by the \c{ORG} directive that
7874 the \c{.text} section should start at 0, does not pass that
7875 information back to the expression evaluator. So from the
7876 evaluator's point of view, \c{$} isn't a pure number: it's an offset
7877 from a section base. Therefore the difference between \c{$} and 510
7878 is also not a pure number, but involves a section base. Values
7879 involving section bases cannot be passed as arguments to \c{TIMES}.
7881 The solution, as in the previous section, is to code the \c{TIMES}
7884 \c TIMES 510-($-$$) DB 0
7886 in which \c{$} and \c{$$} are offsets from the same section base,
7887 and so their difference is a pure number. This will solve the
7888 problem and generate sensible code.
7891 \H{bugs} \i{Bugs}\I{reporting bugs}
7893 We have never yet released a version of NASM with any \e{known}
7894 bugs. That doesn't usually stop there being plenty we didn't know
7895 about, though. Any that you find should be reported firstly via the
7897 \W{http://www.nasm.us/}\c{http://www.nasm.us/}
7898 (click on "Bug Tracker"), or if that fails then through one of the
7899 contacts in \k{contact}.
7901 Please read \k{qstart} first, and don't report the bug if it's
7902 listed in there as a deliberate feature. (If you think the feature
7903 is badly thought out, feel free to send us reasons why you think it
7904 should be changed, but don't just send us mail saying `This is a
7905 bug' if the documentation says we did it on purpose.) Then read
7906 \k{problems}, and don't bother reporting the bug if it's listed
7909 If you do report a bug, \e{please} give us all of the following
7912 \b What operating system you're running NASM under. DOS, Linux,
7913 NetBSD, Win16, Win32, VMS (I'd be impressed), whatever.
7915 \b If you're running NASM under DOS or Win32, tell us whether you've
7916 compiled your own executable from the DOS source archive, or whether
7917 you were using the standard distribution binaries out of the
7918 archive. If you were using a locally built executable, try to
7919 reproduce the problem using one of the standard binaries, as this
7920 will make it easier for us to reproduce your problem prior to fixing
7923 \b Which version of NASM you're using, and exactly how you invoked
7924 it. Give us the precise command line, and the contents of the
7925 \c{NASMENV} environment variable if any.
7927 \b Which versions of any supplementary programs you're using, and
7928 how you invoked them. If the problem only becomes visible at link
7929 time, tell us what linker you're using, what version of it you've
7930 got, and the exact linker command line. If the problem involves
7931 linking against object files generated by a compiler, tell us what
7932 compiler, what version, and what command line or options you used.
7933 (If you're compiling in an IDE, please try to reproduce the problem
7934 with the command-line version of the compiler.)
7936 \b If at all possible, send us a NASM source file which exhibits the
7937 problem. If this causes copyright problems (e.g. you can only
7938 reproduce the bug in restricted-distribution code) then bear in mind
7939 the following two points: firstly, we guarantee that any source code
7940 sent to us for the purposes of debugging NASM will be used \e{only}
7941 for the purposes of debugging NASM, and that we will delete all our
7942 copies of it as soon as we have found and fixed the bug or bugs in
7943 question; and secondly, we would prefer \e{not} to be mailed large
7944 chunks of code anyway. The smaller the file, the better. A
7945 three-line sample file that does nothing useful \e{except}
7946 demonstrate the problem is much easier to work with than a
7947 fully fledged ten-thousand-line program. (Of course, some errors
7948 \e{do} only crop up in large files, so this may not be possible.)
7950 \b A description of what the problem actually \e{is}. `It doesn't
7951 work' is \e{not} a helpful description! Please describe exactly what
7952 is happening that shouldn't be, or what isn't happening that should.
7953 Examples might be: `NASM generates an error message saying Line 3
7954 for an error that's actually on Line 5'; `NASM generates an error
7955 message that I believe it shouldn't be generating at all'; `NASM
7956 fails to generate an error message that I believe it \e{should} be
7957 generating'; `the object file produced from this source code crashes
7958 my linker'; `the ninth byte of the output file is 66 and I think it
7959 should be 77 instead'.
7961 \b If you believe the output file from NASM to be faulty, send it to
7962 us. That allows us to determine whether our own copy of NASM
7963 generates the same file, or whether the problem is related to
7964 portability issues between our development platforms and yours. We
7965 can handle binary files mailed to us as MIME attachments, uuencoded,
7966 and even BinHex. Alternatively, we may be able to provide an FTP
7967 site you can upload the suspect files to; but mailing them is easier
7970 \b Any other information or data files that might be helpful. If,
7971 for example, the problem involves NASM failing to generate an object
7972 file while TASM can generate an equivalent file without trouble,
7973 then send us \e{both} object files, so we can see what TASM is doing
7974 differently from us.
7977 \A{ndisasm} \i{Ndisasm}
7979 The Netwide Disassembler, NDISASM
7981 \H{ndisintro} Introduction
7984 The Netwide Disassembler is a small companion program to the Netwide
7985 Assembler, NASM. It seemed a shame to have an x86 assembler,
7986 complete with a full instruction table, and not make as much use of
7987 it as possible, so here's a disassembler which shares the
7988 instruction table (and some other bits of code) with NASM.
7990 The Netwide Disassembler does nothing except to produce
7991 disassemblies of \e{binary} source files. NDISASM does not have any
7992 understanding of object file formats, like \c{objdump}, and it will
7993 not understand \c{DOS .EXE} files like \c{debug} will. It just
7997 \H{ndisstart} Getting Started: Installation
7999 See \k{install} for installation instructions. NDISASM, like NASM,
8000 has a \c{man page} which you may want to put somewhere useful, if you
8001 are on a Unix system.
8004 \H{ndisrun} Running NDISASM
8006 To disassemble a file, you will typically use a command of the form
8008 \c ndisasm -b {16|32|64} filename
8010 NDISASM can disassemble 16-, 32- or 64-bit code equally easily,
8011 provided of course that you remember to specify which it is to work
8012 with. If no \i\c{-b} switch is present, NDISASM works in 16-bit mode
8013 by default. The \i\c{-u} switch (for USE32) also invokes 32-bit mode.
8015 Two more command line options are \i\c{-r} which reports the version
8016 number of NDISASM you are running, and \i\c{-h} which gives a short
8017 summary of command line options.
8020 \S{ndiscom} COM Files: Specifying an Origin
8022 To disassemble a \c{DOS .COM} file correctly, a disassembler must assume
8023 that the first instruction in the file is loaded at address \c{0x100},
8024 rather than at zero. NDISASM, which assumes by default that any file
8025 you give it is loaded at zero, will therefore need to be informed of
8028 The \i\c{-o} option allows you to declare a different origin for the
8029 file you are disassembling. Its argument may be expressed in any of
8030 the NASM numeric formats: decimal by default, if it begins with `\c{$}'
8031 or `\c{0x}' or ends in `\c{H}' it's \c{hex}, if it ends in `\c{Q}' it's
8032 \c{octal}, and if it ends in `\c{B}' it's \c{binary}.
8034 Hence, to disassemble a \c{.COM} file:
8036 \c ndisasm -o100h filename.com
8041 \S{ndissync} Code Following Data: Synchronisation
8043 Suppose you are disassembling a file which contains some data which
8044 isn't machine code, and \e{then} contains some machine code. NDISASM
8045 will faithfully plough through the data section, producing machine
8046 instructions wherever it can (although most of them will look
8047 bizarre, and some may have unusual prefixes, e.g. `\c{FS OR AX,0x240A}'),
8048 and generating `DB' instructions ever so often if it's totally stumped.
8049 Then it will reach the code section.
8051 Supposing NDISASM has just finished generating a strange machine
8052 instruction from part of the data section, and its file position is
8053 now one byte \e{before} the beginning of the code section. It's
8054 entirely possible that another spurious instruction will get
8055 generated, starting with the final byte of the data section, and
8056 then the correct first instruction in the code section will not be
8057 seen because the starting point skipped over it. This isn't really
8060 To avoid this, you can specify a `\i\c{synchronisation}' point, or indeed
8061 as many synchronisation points as you like (although NDISASM can
8062 only handle 2147483647 sync points internally). The definition of a sync
8063 point is this: NDISASM guarantees to hit sync points exactly during
8064 disassembly. If it is thinking about generating an instruction which
8065 would cause it to jump over a sync point, it will discard that
8066 instruction and output a `\c{db}' instead. So it \e{will} start
8067 disassembly exactly from the sync point, and so you \e{will} see all
8068 the instructions in your code section.
8070 Sync points are specified using the \i\c{-s} option: they are measured
8071 in terms of the program origin, not the file position. So if you
8072 want to synchronize after 32 bytes of a \c{.COM} file, you would have to
8075 \c ndisasm -o100h -s120h file.com
8079 \c ndisasm -o100h -s20h file.com
8081 As stated above, you can specify multiple sync markers if you need
8082 to, just by repeating the \c{-s} option.
8085 \S{ndisisync} Mixed Code and Data: Automatic (Intelligent) Synchronisation
8088 Suppose you are disassembling the boot sector of a \c{DOS} floppy (maybe
8089 it has a virus, and you need to understand the virus so that you
8090 know what kinds of damage it might have done you). Typically, this
8091 will contain a \c{JMP} instruction, then some data, then the rest of the
8092 code. So there is a very good chance of NDISASM being \e{misaligned}
8093 when the data ends and the code begins. Hence a sync point is
8096 On the other hand, why should you have to specify the sync point
8097 manually? What you'd do in order to find where the sync point would
8098 be, surely, would be to read the \c{JMP} instruction, and then to use
8099 its target address as a sync point. So can NDISASM do that for you?
8101 The answer, of course, is yes: using either of the synonymous
8102 switches \i\c{-a} (for automatic sync) or \i\c{-i} (for intelligent
8103 sync) will enable \c{auto-sync} mode. Auto-sync mode automatically
8104 generates a sync point for any forward-referring PC-relative jump or
8105 call instruction that NDISASM encounters. (Since NDISASM is one-pass,
8106 if it encounters a PC-relative jump whose target has already been
8107 processed, there isn't much it can do about it...)
8109 Only PC-relative jumps are processed, since an absolute jump is
8110 either through a register (in which case NDISASM doesn't know what
8111 the register contains) or involves a segment address (in which case
8112 the target code isn't in the same segment that NDISASM is working
8113 in, and so the sync point can't be placed anywhere useful).
8115 For some kinds of file, this mechanism will automatically put sync
8116 points in all the right places, and save you from having to place
8117 any sync points manually. However, it should be stressed that
8118 auto-sync mode is \e{not} guaranteed to catch all the sync points, and
8119 you may still have to place some manually.
8121 Auto-sync mode doesn't prevent you from declaring manual sync
8122 points: it just adds automatically generated ones to the ones you
8123 provide. It's perfectly feasible to specify \c{-i} \e{and} some \c{-s}
8126 Another caveat with auto-sync mode is that if, by some unpleasant
8127 fluke, something in your data section should disassemble to a
8128 PC-relative call or jump instruction, NDISASM may obediently place a
8129 sync point in a totally random place, for example in the middle of
8130 one of the instructions in your code section. So you may end up with
8131 a wrong disassembly even if you use auto-sync. Again, there isn't
8132 much I can do about this. If you have problems, you'll have to use
8133 manual sync points, or use the \c{-k} option (documented below) to
8134 suppress disassembly of the data area.
8137 \S{ndisother} Other Options
8139 The \i\c{-e} option skips a header on the file, by ignoring the first N
8140 bytes. This means that the header is \e{not} counted towards the
8141 disassembly offset: if you give \c{-e10 -o10}, disassembly will start
8142 at byte 10 in the file, and this will be given offset 10, not 20.
8144 The \i\c{-k} option is provided with two comma-separated numeric
8145 arguments, the first of which is an assembly offset and the second
8146 is a number of bytes to skip. This \e{will} count the skipped bytes
8147 towards the assembly offset: its use is to suppress disassembly of a
8148 data section which wouldn't contain anything you wanted to see
8152 \H{ndisbugs} Bugs and Improvements
8154 There are no known bugs. However, any you find, with patches if
8155 possible, should be sent to
8156 \W{mailto:nasm-bugs@lists.sourceforge.net}\c{nasm-bugs@lists.sourceforge.net}, or to the
8158 \W{http://www.nasm.us/}\c{http://www.nasm.us/}
8159 and we'll try to fix them. Feel free to send contributions and
8160 new features as well.
8162 \A{inslist} \i{Instruction List}
8164 \H{inslistintro} Introduction
8166 The following sections show the instructions which NASM currently supports. For each
8167 instruction, there is a separate entry for each supported addressing mode. The third
8168 column shows the processor type in which the instruction was introduced and,
8169 when appropriate, one or more usage flags.
8173 \A{changelog} \i{NASM Version History}