1 \# --------------------------------------------------------------------------
3 \# Copyright 1996-2010 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}
52 \IR{-MD} \c{-MD} option
53 \IR{-MF} \c{-MF} option
54 \IR{-MG} \c{-MG} option
55 \IR{-MP} \c{-MP} option
56 \IR{-MQ} \c{-MQ} option
57 \IR{-MT} \c{-MT} option
78 \IR{!=} \c{!=} operator
79 \IR{$, here} \c{$}, Here token
80 \IR{$, prefix} \c{$}, prefix
83 \IR{%%} \c{%%} operator
84 \IR{%+1} \c{%+1} and \c{%-1} syntax
86 \IR{%0} \c{%0} parameter count
88 \IR{&&} \c{&&} operator
90 \IR{..@} \c{..@} symbol prefix
92 \IR{//} \c{//} operator
94 \IR{<<} \c{<<} operator
95 \IR{<=} \c{<=} operator
96 \IR{<>} \c{<>} operator
98 \IR{==} \c{==} operator
100 \IR{>=} \c{>=} operator
101 \IR{>>} \c{>>} operator
102 \IR{?} \c{?} MASM syntax
103 \IR{^} \c{^} operator
104 \IR{^^} \c{^^} operator
105 \IR{|} \c{|} operator
106 \IR{||} \c{||} operator
107 \IR{~} \c{~} operator
108 \IR{%$} \c{%$} and \c{%$$} prefixes
110 \IR{+ opaddition} \c{+} operator, binary
111 \IR{+ opunary} \c{+} operator, unary
112 \IR{+ modifier} \c{+} modifier
113 \IR{- opsubtraction} \c{-} operator, binary
114 \IR{- opunary} \c{-} operator, unary
115 \IR{! opunary} \c{!} operator, unary
116 \IR{alignment, in bin sections} alignment, in \c{bin} sections
117 \IR{alignment, in elf sections} alignment, in \c{elf} sections
118 \IR{alignment, in win32 sections} alignment, in \c{win32} sections
119 \IR{alignment, of elf common variables} alignment, of \c{elf} common
121 \IR{alignment, in obj sections} alignment, in \c{obj} sections
122 \IR{a.out, bsd version} \c{a.out}, BSD version
123 \IR{a.out, linux version} \c{a.out}, Linux version
124 \IR{autoconf} Autoconf
126 \IR{bitwise and} bitwise AND
127 \IR{bitwise or} bitwise OR
128 \IR{bitwise xor} bitwise XOR
129 \IR{block ifs} block IFs
130 \IR{borland pascal} Borland, Pascal
131 \IR{borland's win32 compilers} Borland, Win32 compilers
132 \IR{braces, after % sign} braces, after \c{%} sign
134 \IR{c calling convention} C calling convention
135 \IR{c symbol names} C symbol names
136 \IA{critical expressions}{critical expression}
137 \IA{command line}{command-line}
138 \IA{case sensitivity}{case sensitive}
139 \IA{case-sensitive}{case sensitive}
140 \IA{case-insensitive}{case sensitive}
141 \IA{character constants}{character constant}
142 \IR{common object file format} Common Object File Format
143 \IR{common variables, alignment in elf} common variables, alignment
145 \IR{common, elf extensions to} \c{COMMON}, \c{elf} extensions to
146 \IR{common, obj extensions to} \c{COMMON}, \c{obj} extensions to
147 \IR{declaring structure} declaring structures
148 \IR{default-wrt mechanism} default-\c{WRT} mechanism
151 \IR{dll symbols, exporting} DLL symbols, exporting
152 \IR{dll symbols, importing} DLL symbols, importing
154 \IR{dos archive} DOS archive
155 \IR{dos source archive} DOS source archive
156 \IA{effective address}{effective addresses}
157 \IA{effective-address}{effective addresses}
159 \IR{elf, 16-bit code and} ELF, 16-bit code and
160 \IR{elf shared libraries} ELF, shared libraries
163 \IR{executable and linkable format} Executable and Linkable Format
164 \IR{extern, obj extensions to} \c{EXTERN}, \c{obj} extensions to
165 \IR{extern, rdf extensions to} \c{EXTERN}, \c{rdf} extensions to
166 \IR{floating-point, constants} floating-point, constants
167 \IR{floating-point, packed bcd constants} floating-point, packed BCD constants
169 \IR{freelink} FreeLink
170 \IR{functions, c calling convention} functions, C calling convention
171 \IR{functions, pascal calling convention} functions, Pascal calling
173 \IR{global, aoutb extensions to} \c{GLOBAL}, \c{aoutb} extensions to
174 \IR{global, elf extensions to} \c{GLOBAL}, \c{elf} extensions to
175 \IR{global, rdf extensions to} \c{GLOBAL}, \c{rdf} extensions to
177 \IR{got relocations} \c{GOT} relocations
178 \IR{gotoff relocation} \c{GOTOFF} relocations
179 \IR{gotpc relocation} \c{GOTPC} relocations
180 \IR{intel number formats} Intel number formats
181 \IR{linux, elf} Linux, ELF
182 \IR{linux, a.out} Linux, \c{a.out}
183 \IR{linux, as86} Linux, \c{as86}
184 \IR{logical and} logical AND
185 \IR{logical or} logical OR
186 \IR{logical xor} logical XOR
187 \IR{mach object file format} Mach, object file format
189 \IR{macho32} \c{macho32}
190 \IR{macho64} \c{macho64}
193 \IA{memory reference}{memory references}
195 \IA{misc directory}{misc subdirectory}
196 \IR{misc subdirectory} \c{misc} subdirectory
197 \IR{microsoft omf} Microsoft OMF
198 \IR{mmx registers} MMX registers
199 \IA{modr/m}{modr/m byte}
200 \IR{modr/m byte} ModR/M byte
202 \IR{ms-dos device drivers} MS-DOS device drivers
203 \IR{multipush} \c{multipush} macro
205 \IR{nasm version} NASM version
209 \IR{operating system} operating system
211 \IR{pascal calling convention}Pascal calling convention
212 \IR{passes} passes, assembly
217 \IR{plt} \c{PLT} relocations
218 \IA{pre-defining macros}{pre-define}
219 \IA{preprocessor expressions}{preprocessor, expressions}
220 \IA{preprocessor loops}{preprocessor, loops}
221 \IA{preprocessor variables}{preprocessor, variables}
222 \IA{rdoff subdirectory}{rdoff}
223 \IR{rdoff} \c{rdoff} subdirectory
224 \IR{relocatable dynamic object file format} Relocatable Dynamic
226 \IR{relocations, pic-specific} relocations, PIC-specific
227 \IA{repeating}{repeating code}
228 \IR{section alignment, in elf} section alignment, in \c{elf}
229 \IR{section alignment, in bin} section alignment, in \c{bin}
230 \IR{section alignment, in obj} section alignment, in \c{obj}
231 \IR{section alignment, in win32} section alignment, in \c{win32}
232 \IR{section, elf extensions to} \c{SECTION}, \c{elf} extensions to
233 \IR{section, win32 extensions to} \c{SECTION}, \c{win32} extensions to
234 \IR{segment alignment, in bin} segment alignment, in \c{bin}
235 \IR{segment alignment, in obj} segment alignment, in \c{obj}
236 \IR{segment, obj extensions to} \c{SEGMENT}, \c{elf} extensions to
237 \IR{segment names, borland pascal} segment names, Borland Pascal
238 \IR{shift command} \c{shift} command
240 \IR{sib byte} SIB byte
241 \IR{align, smart} \c{ALIGN}, smart
242 \IR{solaris x86} Solaris x86
243 \IA{standard section names}{standardized section names}
244 \IR{symbols, exporting from dlls} symbols, exporting from DLLs
245 \IR{symbols, importing from dlls} symbols, importing from DLLs
246 \IR{test subdirectory} \c{test} subdirectory
248 \IR{underscore, in c symbols} underscore, in C symbols
254 \IA{sco unix}{unix, sco}
255 \IR{unix, sco} Unix, SCO
256 \IA{unix source archive}{unix, source archive}
257 \IR{unix, source archive} Unix, source archive
258 \IA{unix system v}{unix, system v}
259 \IR{unix, system v} Unix, System V
260 \IR{unixware} UnixWare
262 \IR{version number of nasm} version number of NASM
263 \IR{visual c++} Visual C++
264 \IR{www page} WWW page
268 \IR{windows 95} Windows 95
269 \IR{windows nt} Windows NT
270 \# \IC{program entry point}{entry point, program}
271 \# \IC{program entry point}{start point, program}
272 \# \IC{MS-DOS device drivers}{device drivers, MS-DOS}
273 \# \IC{16-bit mode, versus 32-bit mode}{32-bit mode, versus 16-bit mode}
274 \# \IC{c symbol names}{symbol names, in C}
277 \C{intro} Introduction
279 \H{whatsnasm} What Is NASM?
281 The Netwide Assembler, NASM, is an 80x86 and x86-64 assembler designed
282 for portability and modularity. It supports a range of object file
283 formats, including Linux and \c{*BSD} \c{a.out}, \c{ELF}, \c{COFF},
284 \c{Mach-O}, Microsoft 16-bit \c{OBJ}, \c{Win32} and \c{Win64}. It will
285 also output plain binary files. Its syntax is designed to be simple
286 and easy to understand, similar to Intel's but less complex. It
287 supports all currently known x86 architectural extensions, and has
288 strong support for macros.
291 \S{yaasm} Why Yet Another Assembler?
293 The Netwide Assembler grew out of an idea on \i\c{comp.lang.asm.x86}
294 (or possibly \i\c{alt.lang.asm} - I forget which), which was
295 essentially that there didn't seem to be a good \e{free} x86-series
296 assembler around, and that maybe someone ought to write one.
298 \b \i\c{a86} is good, but not free, and in particular you don't get any
299 32-bit capability until you pay. It's DOS only, too.
301 \b \i\c{gas} is free, and ports over to DOS and Unix, but it's not
302 very good, since it's designed to be a back end to \i\c{gcc}, which
303 always feeds it correct code. So its error checking is minimal. Also,
304 its syntax is horrible, from the point of view of anyone trying to
305 actually \e{write} anything in it. Plus you can't write 16-bit code in
308 \b \i\c{as86} is specific to Minix and Linux, and (my version at least)
309 doesn't seem to have much (or any) documentation.
311 \b \i\c{MASM} isn't very good, and it's (was) expensive, and it runs only under
314 \b \i\c{TASM} is better, but still strives for MASM compatibility,
315 which means millions of directives and tons of red tape. And its syntax
316 is essentially MASM's, with the contradictions and quirks that
317 entails (although it sorts out some of those by means of Ideal mode.)
318 It's expensive too. And it's DOS-only.
320 So here, for your coding pleasure, is NASM. At present it's
321 still in prototype stage - we don't promise that it can outperform
322 any of these assemblers. But please, \e{please} send us bug reports,
323 fixes, helpful information, and anything else you can get your hands
324 on (and thanks to the many people who've done this already! You all
325 know who you are), and we'll improve it out of all recognition.
329 \S{legal} \i{License} Conditions
331 Please see the file \c{LICENSE}, supplied as part of any NASM
332 distribution archive, for the license conditions under which you may
333 use NASM. NASM is now under the so-called 2-clause BSD license, also
334 known as the simplified BSD license.
336 Copyright 1996-2010 the NASM Authors - All rights reserved.
338 Redistribution and use in source and binary forms, with or without
339 modification, are permitted provided that the following conditions are
342 \b Redistributions of source code must retain the above copyright
343 notice, this list of conditions and the following disclaimer.
345 \b Redistributions in binary form must reproduce the above copyright
346 notice, this list of conditions and the following disclaimer in the
347 documentation and/or other materials provided with the distribution.
349 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
350 CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
351 INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
352 MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
353 DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
354 CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
355 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
356 NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
357 LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
358 HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
359 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
360 OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
361 EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
364 \H{contact} Contact Information
366 The current version of NASM (since about 0.98.08) is maintained by a
367 team of developers, accessible through the \c{nasm-devel} mailing list
368 (see below for the link).
369 If you want to report a bug, please read \k{bugs} first.
371 NASM has a \i{website} at
372 \W{http://www.nasm.us/}\c{http://www.nasm.us/}. If it's not there,
375 \i{New releases}, \i{release candidates}, and \I{snapshots, daily
376 development}\i{daily development snapshots} of NASM are available from
377 the official web site.
379 Announcements are posted to
380 \W{news:comp.lang.asm.x86}\i\c{comp.lang.asm.x86},
382 \W{http://www.freshmeat.net/}\c{http://www.freshmeat.net/}.
384 If you want information about the current development status, please
385 subscribe to the \i\c{nasm-devel} email list; see link from the
389 \H{install} Installation
391 \S{instdos} \i{Installing} NASM under MS-\i{DOS} or Windows
393 Once you've obtained the appropriate archive for NASM,
394 \i\c{nasm-XXX-dos.zip} or \i\c{nasm-XXX-win32.zip} (where \c{XXX}
395 denotes the version number of NASM contained in the archive), unpack
396 it into its own directory (for example \c{c:\\nasm}).
398 The archive will contain a set of executable files: the NASM
399 executable file \i\c{nasm.exe}, the NDISASM executable file
400 \i\c{ndisasm.exe}, and possibly additional utilities to handle the
403 The only file NASM needs to run is its own executable, so copy
404 \c{nasm.exe} to a directory on your PATH, or alternatively edit
405 \i\c{autoexec.bat} to add the \c{nasm} directory to your
406 \i\c{PATH} (to do that under Windows XP, go to Start > Control Panel >
407 System > Advanced > Environment Variables; these instructions may work
408 under other versions of Windows as well.)
410 That's it - NASM is installed. You don't need the nasm directory
411 to be present to run NASM (unless you've added it to your \c{PATH}),
412 so you can delete it if you need to save space; however, you may
413 want to keep the documentation or test programs.
415 If you've downloaded the \i{DOS source archive}, \i\c{nasm-XXX.zip},
416 the \c{nasm} directory will also contain the full NASM \i{source
417 code}, and a selection of \i{Makefiles} you can (hopefully) use to
418 rebuild your copy of NASM from scratch. See the file \c{INSTALL} in
421 Note that a number of files are generated from other files by Perl
422 scripts. Although the NASM source distribution includes these
423 generated files, you will need to rebuild them (and hence, will need a
424 Perl interpreter) if you change insns.dat, standard.mac or the
425 documentation. It is possible future source distributions may not
426 include these files at all. Ports of \i{Perl} for a variety of
427 platforms, including DOS and Windows, are available from
428 \W{http://www.cpan.org/ports/}\i{www.cpan.org}.
431 \S{instdos} Installing NASM under \i{Unix}
433 Once you've obtained the \i{Unix source archive} for NASM,
434 \i\c{nasm-XXX.tar.gz} (where \c{XXX} denotes the version number of
435 NASM contained in the archive), unpack it into a directory such
436 as \c{/usr/local/src}. The archive, when unpacked, will create its
437 own subdirectory \c{nasm-XXX}.
439 NASM is an \I{Autoconf}\I\c{configure}auto-configuring package: once
440 you've unpacked it, \c{cd} to the directory it's been unpacked into
441 and type \c{./configure}. This shell script will find the best C
442 compiler to use for building NASM and set up \i{Makefiles}
445 Once NASM has auto-configured, you can type \i\c{make} to build the
446 \c{nasm} and \c{ndisasm} binaries, and then \c{make install} to
447 install them in \c{/usr/local/bin} and install the \i{man pages}
448 \i\c{nasm.1} and \i\c{ndisasm.1} in \c{/usr/local/man/man1}.
449 Alternatively, you can give options such as \c{--prefix} to the
450 configure script (see the file \i\c{INSTALL} for more details), or
451 install the programs yourself.
453 NASM also comes with a set of utilities for handling the \c{RDOFF}
454 custom object-file format, which are in the \i\c{rdoff} subdirectory
455 of the NASM archive. You can build these with \c{make rdf} and
456 install them with \c{make rdf_install}, if you want them.
459 \C{running} Running NASM
461 \H{syntax} NASM \i{Command-Line} Syntax
463 To assemble a file, you issue a command of the form
465 \c nasm -f <format> <filename> [-o <output>]
469 \c nasm -f elf myfile.asm
471 will assemble \c{myfile.asm} into an \c{ELF} object file \c{myfile.o}. And
473 \c nasm -f bin myfile.asm -o myfile.com
475 will assemble \c{myfile.asm} into a raw binary file \c{myfile.com}.
477 To produce a listing file, with the hex codes output from NASM
478 displayed on the left of the original sources, use the \c{-l} option
479 to give a listing file name, for example:
481 \c nasm -f coff myfile.asm -l myfile.lst
483 To get further usage instructions from NASM, try typing
487 As \c{-hf}, this will also list the available output file formats, and what they
490 If you use Linux but aren't sure whether your system is \c{a.out}
495 (in the directory in which you put the NASM binary when you
496 installed it). If it says something like
498 \c nasm: ELF 32-bit LSB executable i386 (386 and up) Version 1
500 then your system is \c{ELF}, and you should use the option \c{-f elf}
501 when you want NASM to produce Linux object files. If it says
503 \c nasm: Linux/i386 demand-paged executable (QMAGIC)
505 or something similar, your system is \c{a.out}, and you should use
506 \c{-f aout} instead (Linux \c{a.out} systems have long been obsolete,
507 and are rare these days.)
509 Like Unix compilers and assemblers, NASM is silent unless it
510 goes wrong: you won't see any output at all, unless it gives error
514 \S{opt-o} The \i\c{-o} Option: Specifying the Output File Name
516 NASM will normally choose the name of your output file for you;
517 precisely how it does this is dependent on the object file format.
518 For Microsoft object file formats (\c{obj}, \c{win32} and \c{win64}),
519 it will remove the \c{.asm} \i{extension} (or whatever extension you
520 like to use - NASM doesn't care) from your source file name and
521 substitute \c{.obj}. For Unix object file formats (\c{aout}, \c{as86},
522 \c{coff}, \c{elf32}, \c{elf64}, \c{ieee}, \c{macho32} and \c{macho64})
523 it will substitute \c{.o}. For \c{dbg}, \c{rdf}, \c{ith} and \c{srec},
524 it will use \c{.dbg}, \c{.rdf}, \c{.ith} and \c{.srec}, respectively,
525 and for the \c{bin} format it will simply remove the extension, so
526 that \c{myfile.asm} produces the output file \c{myfile}.
528 If the output file already exists, NASM will overwrite it, unless it
529 has the same name as the input file, in which case it will give a
530 warning and use \i\c{nasm.out} as the output file name instead.
532 For situations in which this behaviour is unacceptable, NASM
533 provides the \c{-o} command-line option, which allows you to specify
534 your desired output file name. You invoke \c{-o} by following it
535 with the name you wish for the output file, either with or without
536 an intervening space. For example:
538 \c nasm -f bin program.asm -o program.com
539 \c nasm -f bin driver.asm -odriver.sys
541 Note that this is a small o, and is different from a capital O , which
542 is used to specify the optimization level. See \k{opt-O}.
545 \S{opt-f} The \i\c{-f} Option: Specifying the \i{Output File Format}
547 If you do not supply the \c{-f} option to NASM, it will choose an
548 output file format for you itself. In the distribution versions of
549 NASM, the default is always \i\c{bin}; if you've compiled your own
550 copy of NASM, you can redefine \i\c{OF_DEFAULT} at compile time and
551 choose what you want the default to be.
553 Like \c{-o}, the intervening space between \c{-f} and the output
554 file format is optional; so \c{-f elf} and \c{-felf} are both valid.
556 A complete list of the available output file formats can be given by
557 issuing the command \i\c{nasm -hf}.
560 \S{opt-l} The \i\c{-l} Option: Generating a \i{Listing File}
562 If you supply the \c{-l} option to NASM, followed (with the usual
563 optional space) by a file name, NASM will generate a
564 \i{source-listing file} for you, in which addresses and generated
565 code are listed on the left, and the actual source code, with
566 expansions of multi-line macros (except those which specifically
567 request no expansion in source listings: see \k{nolist}) on the
570 \c nasm -f elf myfile.asm -l myfile.lst
572 If a list file is selected, you may turn off listing for a
573 section of your source with \c{[list -]}, and turn it back on
574 with \c{[list +]}, (the default, obviously). There is no "user
575 form" (without the brackets). This can be used to list only
576 sections of interest, avoiding excessively long listings.
579 \S{opt-M} The \i\c{-M} Option: Generate \i{Makefile Dependencies}
581 This option can be used to generate makefile dependencies on stdout.
582 This can be redirected to a file for further processing. For example:
584 \c nasm -M myfile.asm > myfile.dep
587 \S{opt-MG} The \i\c{-MG} Option: Generate \i{Makefile Dependencies}
589 This option can be used to generate makefile dependencies on stdout.
590 This differs from the \c{-M} option in that if a nonexisting file is
591 encountered, it is assumed to be a generated file and is added to the
592 dependency list without a prefix.
595 \S{opt-MF} The \i\c\{-MF} Option: Set Makefile Dependency File
597 This option can be used with the \c{-M} or \c{-MG} options to send the
598 output to a file, rather than to stdout. For example:
600 \c nasm -M -MF myfile.dep myfile.asm
603 \S{opt-MD} The \i\c{-MD} Option: Assemble and Generate Dependencies
605 The \c{-MD} option acts as the combination of the \c{-M} and \c{-MF}
606 options (i.e. a filename has to be specified.) However, unlike the
607 \c{-M} or \c{-MG} options, \c{-MD} does \e{not} inhibit the normal
608 operation of the assembler. Use this to automatically generate
609 updated dependencies with every assembly session. For example:
611 \c nasm -f elf -o myfile.o -MD myfile.dep myfile.asm
614 \S{opt-MT} The \i\c{-MT} Option: Dependency Target Name
616 The \c{-MT} option can be used to override the default name of the
617 dependency target. This is normally the same as the output filename,
618 specified by the \c{-o} option.
621 \S{opt-MQ} The \i\c{-MQ} Option: Dependency Target Name (Quoted)
623 The \c{-MQ} option acts as the \c{-MT} option, except it tries to
624 quote characters that have special meaning in Makefile syntax. This
625 is not foolproof, as not all characters with special meaning are
629 \S{opt-MP} The \i\c{-MP} Option: Emit phony targets
631 When used with any of the dependency generation options, the \c{-MP}
632 option causes NASM to emit a phony target without dependencies for
633 each header file. This prevents Make from complaining if a header
634 file has been removed.
637 \S{opt-F} The \i\c{-F} Option: Selecting a \i{Debug Information Format}
639 This option is used to select the format of the debug information
640 emitted into the output file, to be used by a debugger (or \e{will}
641 be). Prior to version 2.03.01, the use of this switch did \e{not} enable
642 output of the selected debug info format. Use \c{-g}, see \k{opt-g},
643 to enable output. Versions 2.03.01 and later automatically enable \c{-g}
644 if \c{-F} is specified.
646 A complete list of the available debug file formats for an output
647 format can be seen by issuing the command \c{nasm -f <format> -y}. Not
648 all output formats currently support debugging output. See \k{opt-y}.
650 This should not be confused with the \c{-f dbg} output format option which
651 is not built into NASM by default. For information on how
652 to enable it when building from the sources, see \k{dbgfmt}.
655 \S{opt-g} The \i\c{-g} Option: Enabling \i{Debug Information}.
657 This option can be used to generate debugging information in the specified
658 format. See \k{opt-F}. Using \c{-g} without \c{-F} results in emitting
659 debug info in the default format, if any, for the selected output format.
660 If no debug information is currently implemented in the selected output
661 format, \c{-g} is \e{silently ignored}.
664 \S{opt-X} The \i\c{-X} Option: Selecting an \i{Error Reporting Format}
666 This option can be used to select an error reporting format for any
667 error messages that might be produced by NASM.
669 Currently, two error reporting formats may be selected. They are
670 the \c{-Xvc} option and the \c{-Xgnu} option. The GNU format is
671 the default and looks like this:
673 \c filename.asm:65: error: specific error message
675 where \c{filename.asm} is the name of the source file in which the
676 error was detected, \c{65} is the source file line number on which
677 the error was detected, \c{error} is the severity of the error (this
678 could be \c{warning}), and \c{specific error message} is a more
679 detailed text message which should help pinpoint the exact problem.
681 The other format, specified by \c{-Xvc} is the style used by Microsoft
682 Visual C++ and some other programs. It looks like this:
684 \c filename.asm(65) : error: specific error message
686 where the only difference is that the line number is in parentheses
687 instead of being delimited by colons.
689 See also the \c{Visual C++} output format, \k{win32fmt}.
691 \S{opt-Z} The \i\c{-Z} Option: Send Errors to a File
693 Under \I{DOS}\c{MS-DOS} it can be difficult (though there are ways) to
694 redirect the standard-error output of a program to a file. Since
695 NASM usually produces its warning and \i{error messages} on
696 \i\c{stderr}, this can make it hard to capture the errors if (for
697 example) you want to load them into an editor.
699 NASM therefore provides the \c{-Z} option, taking a filename argument
700 which causes errors to be sent to the specified files rather than
701 standard error. Therefore you can \I{redirecting errors}redirect
702 the errors into a file by typing
704 \c nasm -Z myfile.err -f obj myfile.asm
706 In earlier versions of NASM, this option was called \c{-E}, but it was
707 changed since \c{-E} is an option conventionally used for
708 preprocessing only, with disastrous results. See \k{opt-E}.
710 \S{opt-s} The \i\c{-s} Option: Send Errors to \i\c{stdout}
712 The \c{-s} option redirects \i{error messages} to \c{stdout} rather
713 than \c{stderr}, so it can be redirected under \I{DOS}\c{MS-DOS}. To
714 assemble the file \c{myfile.asm} and pipe its output to the \c{more}
715 program, you can type:
717 \c nasm -s -f obj myfile.asm | more
719 See also the \c{-Z} option, \k{opt-Z}.
722 \S{opt-i} The \i\c{-i}\I\c{-I} Option: Include File Search Directories
724 When NASM sees the \i\c{%include} or \i\c{%pathsearch} directive in a
725 source file (see \k{include}, \k{pathsearch} or \k{incbin}), it will
726 search for the given file not only in the current directory, but also
727 in any directories specified on the command line by the use of the
728 \c{-i} option. Therefore you can include files from a \i{macro
729 library}, for example, by typing
731 \c nasm -ic:\macrolib\ -f obj myfile.asm
733 (As usual, a space between \c{-i} and the path name is allowed, and
736 NASM, in the interests of complete source-code portability, does not
737 understand the file naming conventions of the OS it is running on;
738 the string you provide as an argument to the \c{-i} option will be
739 prepended exactly as written to the name of the include file.
740 Therefore the trailing backslash in the above example is necessary.
741 Under Unix, a trailing forward slash is similarly necessary.
743 (You can use this to your advantage, if you're really \i{perverse},
744 by noting that the option \c{-ifoo} will cause \c{%include "bar.i"}
745 to search for the file \c{foobar.i}...)
747 If you want to define a \e{standard} \i{include search path},
748 similar to \c{/usr/include} on Unix systems, you should place one or
749 more \c{-i} directives in the \c{NASMENV} environment variable (see
752 For Makefile compatibility with many C compilers, this option can also
753 be specified as \c{-I}.
756 \S{opt-p} The \i\c{-p}\I\c{-P} Option: \I{pre-including files}Pre-Include a File
758 \I\c{%include}NASM allows you to specify files to be
759 \e{pre-included} into your source file, by the use of the \c{-p}
762 \c nasm myfile.asm -p myinc.inc
764 is equivalent to running \c{nasm myfile.asm} and placing the
765 directive \c{%include "myinc.inc"} at the start of the file.
767 For consistency with the \c{-I}, \c{-D} and \c{-U} options, this
768 option can also be specified as \c{-P}.
771 \S{opt-d} The \i\c{-d}\I\c{-D} Option: \I{pre-defining macros}Pre-Define a Macro
773 \I\c{%define}Just as the \c{-p} option gives an alternative to placing
774 \c{%include} directives at the start of a source file, the \c{-d}
775 option gives an alternative to placing a \c{%define} directive. You
778 \c nasm myfile.asm -dFOO=100
780 as an alternative to placing the directive
784 at the start of the file. You can miss off the macro value, as well:
785 the option \c{-dFOO} is equivalent to coding \c{%define FOO}. This
786 form of the directive may be useful for selecting \i{assembly-time
787 options} which are then tested using \c{%ifdef}, for example
790 For Makefile compatibility with many C compilers, this option can also
791 be specified as \c{-D}.
794 \S{opt-u} The \i\c{-u}\I\c{-U} Option: \I{Undefining macros}Undefine a Macro
796 \I\c{%undef}The \c{-u} option undefines a macro that would otherwise
797 have been pre-defined, either automatically or by a \c{-p} or \c{-d}
798 option specified earlier on the command lines.
800 For example, the following command line:
802 \c nasm myfile.asm -dFOO=100 -uFOO
804 would result in \c{FOO} \e{not} being a predefined macro in the
805 program. This is useful to override options specified at a different
808 For Makefile compatibility with many C compilers, this option can also
809 be specified as \c{-U}.
812 \S{opt-E} The \i\c{-E}\I{-e} Option: Preprocess Only
814 NASM allows the \i{preprocessor} to be run on its own, up to a
815 point. Using the \c{-E} option (which requires no arguments) will
816 cause NASM to preprocess its input file, expand all the macro
817 references, remove all the comments and preprocessor directives, and
818 print the resulting file on standard output (or save it to a file,
819 if the \c{-o} option is also used).
821 This option cannot be applied to programs which require the
822 preprocessor to evaluate \I{preprocessor expressions}\i{expressions}
823 which depend on the values of symbols: so code such as
825 \c %assign tablesize ($-tablestart)
827 will cause an error in \i{preprocess-only mode}.
829 For compatiblity with older version of NASM, this option can also be
830 written \c{-e}. \c{-E} in older versions of NASM was the equivalent
831 of the current \c{-Z} option, \k{opt-Z}.
833 \S{opt-a} The \i\c{-a} Option: Don't Preprocess At All
835 If NASM is being used as the back end to a compiler, it might be
836 desirable to \I{suppressing preprocessing}suppress preprocessing
837 completely and assume the compiler has already done it, to save time
838 and increase compilation speeds. The \c{-a} option, requiring no
839 argument, instructs NASM to replace its powerful \i{preprocessor}
840 with a \i{stub preprocessor} which does nothing.
843 \S{opt-O} The \i\c{-O} Option: Specifying \i{Code Size Optimization}
845 Using the \c{-O} option, you can tell NASM to carry out different
846 levels of optimization. The syntax is:
848 \b \c{-O0}: No optimization. All operands take their long forms,
849 if a short form is not specified, except conditional jumps.
851 \b \c{-OL}: Legacy optimization. Like \c{-O0}, but displacements
852 which will fit in a signed byte and do not contain references
853 to labels later in the file are optimized, unless the long form
855 This is intended to match NASM 0.98 behavior.
857 \b \c{-O1}: Minimal optimization. Like \c{-O0}, but immediate operands
858 which will fit in a signed byte are optimized,
859 unless the long form is specified. Conditional jumps default
860 to the long form unless otherwise specified.
862 \b \c{-Ox} (where \c{x} is the actual letter \c{x}): Full optimization.
863 Minimize branch offsets, displacements and immediates,
864 overriding size specification unless the \c{strict} keyword
865 has been used (see \k{strict}). For compatibility with earlier
866 releases, the letter \c{x} may also be any number greater than
867 one. This number has no effect on the actual number of passes.
869 The \c{-Ox} mode is recommended for most uses, and is the default
872 Note that this is a capital \c{O}, and is different from a small \c{o}, which
873 is used to specify the output file name. See \k{opt-o}.
876 \S{opt-t} The \i\c{-t} Option: Enable TASM Compatibility Mode
878 NASM includes a limited form of compatibility with Borland's \i\c{TASM}.
879 When NASM's \c{-t} option is used, the following changes are made:
881 \b local labels may be prefixed with \c{@@} instead of \c{.}
883 \b size override is supported within brackets. In TASM compatible mode,
884 a size override inside square brackets changes the size of the operand,
885 and not the address type of the operand as it does in NASM syntax. E.g.
886 \c{mov eax,[DWORD val]} is valid syntax in TASM compatibility mode.
887 Note that you lose the ability to override the default address type for
890 \b unprefixed forms of some directives supported (\c{arg}, \c{elif},
891 \c{else}, \c{endif}, \c{if}, \c{ifdef}, \c{ifdifi}, \c{ifndef},
892 \c{include}, \c{local})
894 \S{opt-w} The \i\c{-w} and \i\c{-W} Options: Enable or Disable Assembly \i{Warnings}
896 NASM can observe many conditions during the course of assembly which
897 are worth mentioning to the user, but not a sufficiently severe
898 error to justify NASM refusing to generate an output file. These
899 conditions are reported like errors, but come up with the word
900 `warning' before the message. Warnings do not prevent NASM from
901 generating an output file and returning a success status to the
904 Some conditions are even less severe than that: they are only
905 sometimes worth mentioning to the user. Therefore NASM supports the
906 \c{-w} command-line option, which enables or disables certain
907 classes of assembly warning. Such warning classes are described by a
908 name, for example \c{orphan-labels}; you can enable warnings of
909 this class by the command-line option \c{-w+orphan-labels} and
910 disable it by \c{-w-orphan-labels}.
912 The \i{suppressible warning} classes are:
914 \b \i\c{macro-params} covers warnings about \i{multi-line macros}
915 being invoked with the wrong number of parameters. This warning
916 class is enabled by default; see \k{mlmacover} for an example of why
917 you might want to disable it.
919 \b \i\c{macro-selfref} warns if a macro references itself. This
920 warning class is disabled by default.
922 \b\i\c{macro-defaults} warns when a macro has more default
923 parameters than optional parameters. This warning class
924 is enabled by default; see \k{mlmacdef} for why you might want to disable it.
926 \b \i\c{orphan-labels} covers warnings about source lines which
927 contain no instruction but define a label without a trailing colon.
928 NASM warns about this somewhat obscure condition by default;
929 see \k{syntax} for more information.
931 \b \i\c{number-overflow} covers warnings about numeric constants which
932 don't fit in 64 bits. This warning class is enabled by default.
934 \b \i\c{gnu-elf-extensions} warns if 8-bit or 16-bit relocations
935 are used in \c{-f elf} format. The GNU extensions allow this.
936 This warning class is disabled by default.
938 \b \i\c{float-overflow} warns about floating point overflow.
941 \b \i\c{float-denorm} warns about floating point denormals.
944 \b \i\c{float-underflow} warns about floating point underflow.
947 \b \i\c{float-toolong} warns about too many digits in floating-point numbers.
950 \b \i\c{user} controls \c{%warning} directives (see \k{pperror}).
953 \b \i\c{error} causes warnings to be treated as errors. Disabled by
956 \b \i\c{all} is an alias for \e{all} suppressible warning classes (not
957 including \c{error}). Thus, \c{-w+all} enables all available warnings.
959 In addition, you can set warning classes across sections.
960 Warning classes may be enabled with \i\c{[warning +warning-name]},
961 disabled with \i\c{[warning -warning-name]} or reset to their
962 original value with \i\c{[warning *warning-name]}. No "user form"
963 (without the brackets) exists.
965 Since version 2.00, NASM has also supported the gcc-like syntax
966 \c{-Wwarning} and \c{-Wno-warning} instead of \c{-w+warning} and
967 \c{-w-warning}, respectively.
970 \S{opt-v} The \i\c{-v} Option: Display \i{Version} Info
972 Typing \c{NASM -v} will display the version of NASM which you are using,
973 and the date on which it was compiled.
975 You will need the version number if you report a bug.
977 \S{opt-y} The \i\c{-y} Option: Display Available Debug Info Formats
979 Typing \c{nasm -f <option> -y} will display a list of the available
980 debug info formats for the given output format. The default format
981 is indicated by an asterisk. For example:
985 \c valid debug formats for 'elf32' output format are
986 \c ('*' denotes default):
987 \c * stabs ELF32 (i386) stabs debug format for Linux
988 \c dwarf elf32 (i386) dwarf debug format for Linux
991 \S{opt-pfix} The \i\c{--prefix} and \i\c{--postfix} Options.
993 The \c{--prefix} and \c{--postfix} options prepend or append
994 (respectively) the given argument to all \c{global} or
995 \c{extern} variables. E.g. \c{--prefix _} will prepend the
996 underscore to all global and external variables, as C sometimes
997 (but not always) likes it.
1000 \S{nasmenv} The \i\c{NASMENV} \i{Environment} Variable
1002 If you define an environment variable called \c{NASMENV}, the program
1003 will interpret it as a list of extra command-line options, which are
1004 processed before the real command line. You can use this to define
1005 standard search directories for include files, by putting \c{-i}
1006 options in the \c{NASMENV} variable.
1008 The value of the variable is split up at white space, so that the
1009 value \c{-s -ic:\\nasmlib\\} will be treated as two separate options.
1010 However, that means that the value \c{-dNAME="my name"} won't do
1011 what you might want, because it will be split at the space and the
1012 NASM command-line processing will get confused by the two
1013 nonsensical words \c{-dNAME="my} and \c{name"}.
1015 To get round this, NASM provides a feature whereby, if you begin the
1016 \c{NASMENV} environment variable with some character that isn't a minus
1017 sign, then NASM will treat this character as the \i{separator
1018 character} for options. So setting the \c{NASMENV} variable to the
1019 value \c{!-s!-ic:\\nasmlib\\} is equivalent to setting it to \c{-s
1020 -ic:\\nasmlib\\}, but \c{!-dNAME="my name"} will work.
1022 This environment variable was previously called \c{NASM}. This was
1023 changed with version 0.98.31.
1026 \H{qstart} \i{Quick Start} for \i{MASM} Users
1028 If you're used to writing programs with MASM, or with \i{TASM} in
1029 MASM-compatible (non-Ideal) mode, or with \i\c{a86}, this section
1030 attempts to outline the major differences between MASM's syntax and
1031 NASM's. If you're not already used to MASM, it's probably worth
1032 skipping this section.
1035 \S{qscs} NASM Is \I{case sensitivity}Case-Sensitive
1037 One simple difference is that NASM is case-sensitive. It makes a
1038 difference whether you call your label \c{foo}, \c{Foo} or \c{FOO}.
1039 If you're assembling to \c{DOS} or \c{OS/2} \c{.OBJ} files, you can
1040 invoke the \i\c{UPPERCASE} directive (documented in \k{objfmt}) to
1041 ensure that all symbols exported to other code modules are forced
1042 to be upper case; but even then, \e{within} a single module, NASM
1043 will distinguish between labels differing only in case.
1046 \S{qsbrackets} NASM Requires \i{Square Brackets} For \i{Memory References}
1048 NASM was designed with simplicity of syntax in mind. One of the
1049 \i{design goals} of NASM is that it should be possible, as far as is
1050 practical, for the user to look at a single line of NASM code
1051 and tell what opcode is generated by it. You can't do this in MASM:
1052 if you declare, for example,
1057 then the two lines of code
1062 generate completely different opcodes, despite having
1063 identical-looking syntaxes.
1065 NASM avoids this undesirable situation by having a much simpler
1066 syntax for memory references. The rule is simply that any access to
1067 the \e{contents} of a memory location requires square brackets
1068 around the address, and any access to the \e{address} of a variable
1069 doesn't. So an instruction of the form \c{mov ax,foo} will
1070 \e{always} refer to a compile-time constant, whether it's an \c{EQU}
1071 or the address of a variable; and to access the \e{contents} of the
1072 variable \c{bar}, you must code \c{mov ax,[bar]}.
1074 This also means that NASM has no need for MASM's \i\c{OFFSET}
1075 keyword, since the MASM code \c{mov ax,offset bar} means exactly the
1076 same thing as NASM's \c{mov ax,bar}. If you're trying to get
1077 large amounts of MASM code to assemble sensibly under NASM, you
1078 can always code \c{%idefine offset} to make the preprocessor treat
1079 the \c{OFFSET} keyword as a no-op.
1081 This issue is even more confusing in \i\c{a86}, where declaring a
1082 label with a trailing colon defines it to be a `label' as opposed to
1083 a `variable' and causes \c{a86} to adopt NASM-style semantics; so in
1084 \c{a86}, \c{mov ax,var} has different behaviour depending on whether
1085 \c{var} was declared as \c{var: dw 0} (a label) or \c{var dw 0} (a
1086 word-size variable). NASM is very simple by comparison:
1087 \e{everything} is a label.
1089 NASM, in the interests of simplicity, also does not support the
1090 \i{hybrid syntaxes} supported by MASM and its clones, such as
1091 \c{mov ax,table[bx]}, where a memory reference is denoted by one
1092 portion outside square brackets and another portion inside. The
1093 correct syntax for the above is \c{mov ax,[table+bx]}. Likewise,
1094 \c{mov ax,es:[di]} is wrong and \c{mov ax,[es:di]} is right.
1097 \S{qstypes} NASM Doesn't Store \i{Variable Types}
1099 NASM, by design, chooses not to remember the types of variables you
1100 declare. Whereas MASM will remember, on seeing \c{var dw 0}, that
1101 you declared \c{var} as a word-size variable, and will then be able
1102 to fill in the \i{ambiguity} in the size of the instruction \c{mov
1103 var,2}, NASM will deliberately remember nothing about the symbol
1104 \c{var} except where it begins, and so you must explicitly code
1105 \c{mov word [var],2}.
1107 For this reason, NASM doesn't support the \c{LODS}, \c{MOVS},
1108 \c{STOS}, \c{SCAS}, \c{CMPS}, \c{INS}, or \c{OUTS} instructions,
1109 but only supports the forms such as \c{LODSB}, \c{MOVSW}, and
1110 \c{SCASD}, which explicitly specify the size of the components of
1111 the strings being manipulated.
1114 \S{qsassume} NASM Doesn't \i\c{ASSUME}
1116 As part of NASM's drive for simplicity, it also does not support the
1117 \c{ASSUME} directive. NASM will not keep track of what values you
1118 choose to put in your segment registers, and will never
1119 \e{automatically} generate a \i{segment override} prefix.
1122 \S{qsmodel} NASM Doesn't Support \i{Memory Models}
1124 NASM also does not have any directives to support different 16-bit
1125 memory models. The programmer has to keep track of which functions
1126 are supposed to be called with a \i{far call} and which with a
1127 \i{near call}, and is responsible for putting the correct form of
1128 \c{RET} instruction (\c{RETN} or \c{RETF}; NASM accepts \c{RET}
1129 itself as an alternate form for \c{RETN}); in addition, the
1130 programmer is responsible for coding CALL FAR instructions where
1131 necessary when calling \e{external} functions, and must also keep
1132 track of which external variable definitions are far and which are
1136 \S{qsfpu} \i{Floating-Point} Differences
1138 NASM uses different names to refer to floating-point registers from
1139 MASM: where MASM would call them \c{ST(0)}, \c{ST(1)} and so on, and
1140 \i\c{a86} would call them simply \c{0}, \c{1} and so on, NASM
1141 chooses to call them \c{st0}, \c{st1} etc.
1143 As of version 0.96, NASM now treats the instructions with
1144 \i{`nowait'} forms in the same way as MASM-compatible assemblers.
1145 The idiosyncratic treatment employed by 0.95 and earlier was based
1146 on a misunderstanding by the authors.
1149 \S{qsother} Other Differences
1151 For historical reasons, NASM uses the keyword \i\c{TWORD} where MASM
1152 and compatible assemblers use \i\c{TBYTE}.
1154 NASM does not declare \i{uninitialized storage} in the same way as
1155 MASM: where a MASM programmer might use \c{stack db 64 dup (?)},
1156 NASM requires \c{stack resb 64}, intended to be read as `reserve 64
1157 bytes'. For a limited amount of compatibility, since NASM treats
1158 \c{?} as a valid character in symbol names, you can code \c{? equ 0}
1159 and then writing \c{dw ?} will at least do something vaguely useful.
1160 \I\c{RESB}\i\c{DUP} is still not a supported syntax, however.
1162 In addition to all of this, macros and directives work completely
1163 differently to MASM. See \k{preproc} and \k{directive} for further
1167 \C{lang} The NASM Language
1169 \H{syntax} Layout of a NASM Source Line
1171 Like most assemblers, each NASM source line contains (unless it
1172 is a macro, a preprocessor directive or an assembler directive: see
1173 \k{preproc} and \k{directive}) some combination of the four fields
1175 \c label: instruction operands ; comment
1177 As usual, most of these fields are optional; the presence or absence
1178 of any combination of a label, an instruction and a comment is allowed.
1179 Of course, the operand field is either required or forbidden by the
1180 presence and nature of the instruction field.
1182 NASM uses backslash (\\) as the line continuation character; if a line
1183 ends with backslash, the next line is considered to be a part of the
1184 backslash-ended line.
1186 NASM places no restrictions on white space within a line: labels may
1187 have white space before them, or instructions may have no space
1188 before them, or anything. The \i{colon} after a label is also
1189 optional. (Note that this means that if you intend to code \c{lodsb}
1190 alone on a line, and type \c{lodab} by accident, then that's still a
1191 valid source line which does nothing but define a label. Running
1192 NASM with the command-line option
1193 \I{orphan-labels}\c{-w+orphan-labels} will cause it to warn you if
1194 you define a label alone on a line without a \i{trailing colon}.)
1196 \i{Valid characters} in labels are letters, numbers, \c{_}, \c{$},
1197 \c{#}, \c{@}, \c{~}, \c{.}, and \c{?}. The only characters which may
1198 be used as the \e{first} character of an identifier are letters,
1199 \c{.} (with special meaning: see \k{locallab}), \c{_} and \c{?}.
1200 An identifier may also be prefixed with a \I{$, prefix}\c{$} to
1201 indicate that it is intended to be read as an identifier and not a
1202 reserved word; thus, if some other module you are linking with
1203 defines a symbol called \c{eax}, you can refer to \c{$eax} in NASM
1204 code to distinguish the symbol from the register. Maximum length of
1205 an identifier is 4095 characters.
1207 The instruction field may contain any machine instruction: Pentium
1208 and P6 instructions, FPU instructions, MMX instructions and even
1209 undocumented instructions are all supported. The instruction may be
1210 prefixed by \c{LOCK}, \c{REP}, \c{REPE}/\c{REPZ} or
1211 \c{REPNE}/\c{REPNZ}, in the usual way. Explicit \I{address-size
1212 prefixes}address-size and \i{operand-size prefixes} \i\c{A16},
1213 \i\c{A32}, \i\c{A64}, \i\c{O16} and \i\c{O32}, \i\c{O64} are provided - one example of their use
1214 is given in \k{mixsize}. You can also use the name of a \I{segment
1215 override}segment register as an instruction prefix: coding
1216 \c{es mov [bx],ax} is equivalent to coding \c{mov [es:bx],ax}. We
1217 recommend the latter syntax, since it is consistent with other
1218 syntactic features of the language, but for instructions such as
1219 \c{LODSB}, which has no operands and yet can require a segment
1220 override, there is no clean syntactic way to proceed apart from
1223 An instruction is not required to use a prefix: prefixes such as
1224 \c{CS}, \c{A32}, \c{LOCK} or \c{REPE} can appear on a line by
1225 themselves, and NASM will just generate the prefix bytes.
1227 In addition to actual machine instructions, NASM also supports a
1228 number of pseudo-instructions, described in \k{pseudop}.
1230 Instruction \i{operands} may take a number of forms: they can be
1231 registers, described simply by the register name (e.g. \c{ax},
1232 \c{bp}, \c{ebx}, \c{cr0}: NASM does not use the \c{gas}-style
1233 syntax in which register names must be prefixed by a \c{%} sign), or
1234 they can be \i{effective addresses} (see \k{effaddr}), constants
1235 (\k{const}) or expressions (\k{expr}).
1237 For x87 \i{floating-point} instructions, NASM accepts a wide range of
1238 syntaxes: you can use two-operand forms like MASM supports, or you
1239 can use NASM's native single-operand forms in most cases.
1241 \# all forms of each supported instruction are given in
1243 For example, you can code:
1245 \c fadd st1 ; this sets st0 := st0 + st1
1246 \c fadd st0,st1 ; so does this
1248 \c fadd st1,st0 ; this sets st1 := st1 + st0
1249 \c fadd to st1 ; so does this
1251 Almost any x87 floating-point instruction that references memory must
1252 use one of the prefixes \i\c{DWORD}, \i\c{QWORD} or \i\c{TWORD} to
1253 indicate what size of \i{memory operand} it refers to.
1256 \H{pseudop} \i{Pseudo-Instructions}
1258 Pseudo-instructions are things which, though not real x86 machine
1259 instructions, are used in the instruction field anyway because that's
1260 the most convenient place to put them. The current pseudo-instructions
1261 are \i\c{DB}, \i\c{DW}, \i\c{DD}, \i\c{DQ}, \i\c{DT}, \i\c{DO} and
1262 \i\c{DY}; their \i{uninitialized} counterparts \i\c{RESB}, \i\c{RESW},
1263 \i\c{RESD}, \i\c{RESQ}, \i\c{REST}, \i\c{RESO} and \i\c{RESY}; the
1264 \i\c{INCBIN} command, the \i\c{EQU} command, and the \i\c{TIMES}
1268 \S{db} \c{DB} and Friends: Declaring Initialized Data
1270 \i\c{DB}, \i\c{DW}, \i\c{DD}, \i\c{DQ}, \i\c{DT}, \i\c{DO} and
1271 \i\c{DY} are used, much as in MASM, to declare initialized data in the
1272 output file. They can be invoked in a wide range of ways:
1273 \I{floating-point}\I{character constant}\I{string constant}
1275 \c db 0x55 ; just the byte 0x55
1276 \c db 0x55,0x56,0x57 ; three bytes in succession
1277 \c db 'a',0x55 ; character constants are OK
1278 \c db 'hello',13,10,'$' ; so are string constants
1279 \c dw 0x1234 ; 0x34 0x12
1280 \c dw 'a' ; 0x61 0x00 (it's just a number)
1281 \c dw 'ab' ; 0x61 0x62 (character constant)
1282 \c dw 'abc' ; 0x61 0x62 0x63 0x00 (string)
1283 \c dd 0x12345678 ; 0x78 0x56 0x34 0x12
1284 \c dd 1.234567e20 ; floating-point constant
1285 \c dq 0x123456789abcdef0 ; eight byte constant
1286 \c dq 1.234567e20 ; double-precision float
1287 \c dt 1.234567e20 ; extended-precision float
1289 \c{DT}, \c{DO} and \c{DY} do not accept \i{numeric constants} as operands.
1292 \S{resb} \c{RESB} and Friends: Declaring \i{Uninitialized} Data
1294 \i\c{RESB}, \i\c{RESW}, \i\c{RESD}, \i\c{RESQ}, \i\c{REST}, \i\c{RESO}
1295 and \i\c{RESY} are designed to be used in the BSS section of a module:
1296 they declare \e{uninitialized} storage space. Each takes a single
1297 operand, which is the number of bytes, words, doublewords or whatever
1298 to reserve. As stated in \k{qsother}, NASM does not support the
1299 MASM/TASM syntax of reserving uninitialized space by writing
1300 \I\c{?}\c{DW ?} or similar things: this is what it does instead. The
1301 operand to a \c{RESB}-type pseudo-instruction is a \i\e{critical
1302 expression}: see \k{crit}.
1306 \c buffer: resb 64 ; reserve 64 bytes
1307 \c wordvar: resw 1 ; reserve a word
1308 \c realarray resq 10 ; array of ten reals
1309 \c ymmval: resy 1 ; one YMM register
1311 \S{incbin} \i\c{INCBIN}: Including External \i{Binary Files}
1313 \c{INCBIN} is borrowed from the old Amiga assembler \i{DevPac}: it
1314 includes a binary file verbatim into the output file. This can be
1315 handy for (for example) including \i{graphics} and \i{sound} data
1316 directly into a game executable file. It can be called in one of
1319 \c incbin "file.dat" ; include the whole file
1320 \c incbin "file.dat",1024 ; skip the first 1024 bytes
1321 \c incbin "file.dat",1024,512 ; skip the first 1024, and
1322 \c ; actually include at most 512
1324 \c{INCBIN} is both a directive and a standard macro; the standard
1325 macro version searches for the file in the include file search path
1326 and adds the file to the dependency lists. This macro can be
1327 overridden if desired.
1330 \S{equ} \i\c{EQU}: Defining Constants
1332 \c{EQU} defines a symbol to a given constant value: when \c{EQU} is
1333 used, the source line must contain a label. The action of \c{EQU} is
1334 to define the given label name to the value of its (only) operand.
1335 This definition is absolute, and cannot change later. So, for
1338 \c message db 'hello, world'
1339 \c msglen equ $-message
1341 defines \c{msglen} to be the constant 12. \c{msglen} may not then be
1342 redefined later. This is not a \i{preprocessor} definition either:
1343 the value of \c{msglen} is evaluated \e{once}, using the value of
1344 \c{$} (see \k{expr} for an explanation of \c{$}) at the point of
1345 definition, rather than being evaluated wherever it is referenced
1346 and using the value of \c{$} at the point of reference.
1349 \S{times} \i\c{TIMES}: \i{Repeating} Instructions or Data
1351 The \c{TIMES} prefix causes the instruction to be assembled multiple
1352 times. This is partly present as NASM's equivalent of the \i\c{DUP}
1353 syntax supported by \i{MASM}-compatible assemblers, in that you can
1356 \c zerobuf: times 64 db 0
1358 or similar things; but \c{TIMES} is more versatile than that. The
1359 argument to \c{TIMES} is not just a numeric constant, but a numeric
1360 \e{expression}, so you can do things like
1362 \c buffer: db 'hello, world'
1363 \c times 64-$+buffer db ' '
1365 which will store exactly enough spaces to make the total length of
1366 \c{buffer} up to 64. Finally, \c{TIMES} can be applied to ordinary
1367 instructions, so you can code trivial \i{unrolled loops} in it:
1371 Note that there is no effective difference between \c{times 100 resb
1372 1} and \c{resb 100}, except that the latter will be assembled about
1373 100 times faster due to the internal structure of the assembler.
1375 The operand to \c{TIMES} is a critical expression (\k{crit}).
1377 Note also that \c{TIMES} can't be applied to \i{macros}: the reason
1378 for this is that \c{TIMES} is processed after the macro phase, which
1379 allows the argument to \c{TIMES} to contain expressions such as
1380 \c{64-$+buffer} as above. To repeat more than one line of code, or a
1381 complex macro, use the preprocessor \i\c{%rep} directive.
1384 \H{effaddr} Effective Addresses
1386 An \i{effective address} is any operand to an instruction which
1387 \I{memory reference}references memory. Effective addresses, in NASM,
1388 have a very simple syntax: they consist of an expression evaluating
1389 to the desired address, enclosed in \i{square brackets}. For
1394 \c mov ax,[wordvar+1]
1395 \c mov ax,[es:wordvar+bx]
1397 Anything not conforming to this simple system is not a valid memory
1398 reference in NASM, for example \c{es:wordvar[bx]}.
1400 More complicated effective addresses, such as those involving more
1401 than one register, work in exactly the same way:
1403 \c mov eax,[ebx*2+ecx+offset]
1406 NASM is capable of doing \i{algebra} on these effective addresses,
1407 so that things which don't necessarily \e{look} legal are perfectly
1410 \c mov eax,[ebx*5] ; assembles as [ebx*4+ebx]
1411 \c mov eax,[label1*2-label2] ; ie [label1+(label1-label2)]
1413 Some forms of effective address have more than one assembled form;
1414 in most such cases NASM will generate the smallest form it can. For
1415 example, there are distinct assembled forms for the 32-bit effective
1416 addresses \c{[eax*2+0]} and \c{[eax+eax]}, and NASM will generally
1417 generate the latter on the grounds that the former requires four
1418 bytes to store a zero offset.
1420 NASM has a hinting mechanism which will cause \c{[eax+ebx]} and
1421 \c{[ebx+eax]} to generate different opcodes; this is occasionally
1422 useful because \c{[esi+ebp]} and \c{[ebp+esi]} have different
1423 default segment registers.
1425 However, you can force NASM to generate an effective address in a
1426 particular form by the use of the keywords \c{BYTE}, \c{WORD},
1427 \c{DWORD} and \c{NOSPLIT}. If you need \c{[eax+3]} to be assembled
1428 using a double-word offset field instead of the one byte NASM will
1429 normally generate, you can code \c{[dword eax+3]}. Similarly, you
1430 can force NASM to use a byte offset for a small value which it
1431 hasn't seen on the first pass (see \k{crit} for an example of such a
1432 code fragment) by using \c{[byte eax+offset]}. As special cases,
1433 \c{[byte eax]} will code \c{[eax+0]} with a byte offset of zero, and
1434 \c{[dword eax]} will code it with a double-word offset of zero. The
1435 normal form, \c{[eax]}, will be coded with no offset field.
1437 The form described in the previous paragraph is also useful if you
1438 are trying to access data in a 32-bit segment from within 16 bit code.
1439 For more information on this see the section on mixed-size addressing
1440 (\k{mixaddr}). In particular, if you need to access data with a known
1441 offset that is larger than will fit in a 16-bit value, if you don't
1442 specify that it is a dword offset, nasm will cause the high word of
1443 the offset to be lost.
1445 Similarly, NASM will split \c{[eax*2]} into \c{[eax+eax]} because
1446 that allows the offset field to be absent and space to be saved; in
1447 fact, it will also split \c{[eax*2+offset]} into
1448 \c{[eax+eax+offset]}. You can combat this behaviour by the use of
1449 the \c{NOSPLIT} keyword: \c{[nosplit eax*2]} will force
1450 \c{[eax*2+0]} to be generated literally.
1452 In 64-bit mode, NASM will by default generate absolute addresses. The
1453 \i\c{REL} keyword makes it produce \c{RIP}-relative addresses. Since
1454 this is frequently the normally desired behaviour, see the \c{DEFAULT}
1455 directive (\k{default}). The keyword \i\c{ABS} overrides \i\c{REL}.
1458 \H{const} \i{Constants}
1460 NASM understands four different types of constant: numeric,
1461 character, string and floating-point.
1464 \S{numconst} \i{Numeric Constants}
1466 A numeric constant is simply a number. NASM allows you to specify
1467 numbers in a variety of number bases, in a variety of ways: you can
1468 suffix \c{H} or \c{X}, \c{D} or \c{T}, \c{Q} or \c{O}, and \c{B} or
1469 \c{Y} for \i{hexadecimal}, \i{decimal}, \i{octal} and \i{binary}
1470 respectively, or you can prefix \c{0x}, for hexadecimal in the style
1471 of C, or you can prefix \c{$} for hexadecimal in the style of Borland
1472 Pascal or Motorola Assemblers. Note, though, that the \I{$,
1473 prefix}\c{$} prefix does double duty as a prefix on identifiers (see
1474 \k{syntax}), so a hex number prefixed with a \c{$} sign must have a
1475 digit after the \c{$} rather than a letter. In addition, current
1476 versions of NASM accept the prefix \c{0h} for hexadecimal, \c{0d} or
1477 \c{0t} for decimal, \c{0o} or \c{0q} for octal, and \c{0b} or \c{0y}
1478 for binary. Please note that unlike C, a \c{0} prefix by itself does
1479 \e{not} imply an octal constant!
1481 Numeric constants can have underscores (\c{_}) interspersed to break
1484 Some examples (all producing exactly the same code):
1486 \c mov ax,200 ; decimal
1487 \c mov ax,0200 ; still decimal
1488 \c mov ax,0200d ; explicitly decimal
1489 \c mov ax,0d200 ; also decimal
1490 \c mov ax,0c8h ; hex
1491 \c mov ax,$0c8 ; hex again: the 0 is required
1492 \c mov ax,0xc8 ; hex yet again
1493 \c mov ax,0hc8 ; still hex
1494 \c mov ax,310q ; octal
1495 \c mov ax,310o ; octal again
1496 \c mov ax,0o310 ; octal yet again
1497 \c mov ax,0q310 ; hex yet again
1498 \c mov ax,11001000b ; binary
1499 \c mov ax,1100_1000b ; same binary constant
1500 \c mov ax,1100_1000y ; same binary constant once more
1501 \c mov ax,0b1100_1000 ; same binary constant yet again
1502 \c mov ax,0y1100_1000 ; same binary constant yet again
1504 \S{strings} \I{Strings}\i{Character Strings}
1506 A character string consists of up to eight characters enclosed in
1507 either single quotes (\c{'...'}), double quotes (\c{"..."}) or
1508 backquotes (\c{`...`}). Single or double quotes are equivalent to
1509 NASM (except of course that surrounding the constant with single
1510 quotes allows double quotes to appear within it and vice versa); the
1511 contents of those are represented verbatim. Strings enclosed in
1512 backquotes support C-style \c{\\}-escapes for special characters.
1515 The following \i{escape sequences} are recognized by backquoted strings:
1517 \c \' single quote (')
1518 \c \" double quote (")
1520 \c \\\ backslash (\)
1521 \c \? question mark (?)
1529 \c \e ESC (ASCII 27)
1530 \c \377 Up to 3 octal digits - literal byte
1531 \c \xFF Up to 2 hexadecimal digits - literal byte
1532 \c \u1234 4 hexadecimal digits - Unicode character
1533 \c \U12345678 8 hexadecimal digits - Unicode character
1535 All other escape sequences are reserved. Note that \c{\\0}, meaning a
1536 \c{NUL} character (ASCII 0), is a special case of the octal escape
1539 \i{Unicode} characters specified with \c{\\u} or \c{\\U} are converted to
1540 \i{UTF-8}. For example, the following lines are all equivalent:
1542 \c db `\u263a` ; UTF-8 smiley face
1543 \c db `\xe2\x98\xba` ; UTF-8 smiley face
1544 \c db 0E2h, 098h, 0BAh ; UTF-8 smiley face
1547 \S{chrconst} \i{Character Constants}
1549 A character constant consists of a string up to eight bytes long, used
1550 in an expression context. It is treated as if it was an integer.
1552 A character constant with more than one byte will be arranged
1553 with \i{little-endian} order in mind: if you code
1557 then the constant generated is not \c{0x61626364}, but
1558 \c{0x64636261}, so that if you were then to store the value into
1559 memory, it would read \c{abcd} rather than \c{dcba}. This is also
1560 the sense of character constants understood by the Pentium's
1561 \i\c{CPUID} instruction.
1564 \S{strconst} \i{String Constants}
1566 String constants are character strings used in the context of some
1567 pseudo-instructions, namely the
1568 \I\c{DW}\I\c{DD}\I\c{DQ}\I\c{DT}\I\c{DO}\I\c{DY}\i\c{DB} family and
1569 \i\c{INCBIN} (where it represents a filename.) They are also used in
1570 certain preprocessor directives.
1572 A string constant looks like a character constant, only longer. It
1573 is treated as a concatenation of maximum-size character constants
1574 for the conditions. So the following are equivalent:
1576 \c db 'hello' ; string constant
1577 \c db 'h','e','l','l','o' ; equivalent character constants
1579 And the following are also equivalent:
1581 \c dd 'ninechars' ; doubleword string constant
1582 \c dd 'nine','char','s' ; becomes three doublewords
1583 \c db 'ninechars',0,0,0 ; and really looks like this
1585 Note that when used in a string-supporting context, quoted strings are
1586 treated as a string constants even if they are short enough to be a
1587 character constant, because otherwise \c{db 'ab'} would have the same
1588 effect as \c{db 'a'}, which would be silly. Similarly, three-character
1589 or four-character constants are treated as strings when they are
1590 operands to \c{DW}, and so forth.
1592 \S{unicode} \I{UTF-16}\I{UTF-32}\i{Unicode} Strings
1594 The special operators \i\c{__utf16__} and \i\c{__utf32__} allows
1595 definition of Unicode strings. They take a string in UTF-8 format and
1596 converts it to (littleendian) UTF-16 or UTF-32, respectively.
1600 \c %define u(x) __utf16__(x)
1601 \c %define w(x) __utf32__(x)
1603 \c dw u('C:\WINDOWS'), 0 ; Pathname in UTF-16
1604 \c dd w(`A + B = \u206a`), 0 ; String in UTF-32
1606 \c{__utf16__} and \c{__utf32__} can be applied either to strings
1607 passed to the \c{DB} family instructions, or to character constants in
1608 an expression context.
1610 \S{fltconst} \I{floating-point, constants}Floating-Point Constants
1612 \i{Floating-point} constants are acceptable only as arguments to
1613 \i\c{DB}, \i\c{DW}, \i\c{DD}, \i\c{DQ}, \i\c{DT}, and \i\c{DO}, or as
1614 arguments to the special operators \i\c{__float8__},
1615 \i\c{__float16__}, \i\c{__float32__}, \i\c{__float64__},
1616 \i\c{__float80m__}, \i\c{__float80e__}, \i\c{__float128l__}, and
1617 \i\c{__float128h__}.
1619 Floating-point constants are expressed in the traditional form:
1620 digits, then a period, then optionally more digits, then optionally an
1621 \c{E} followed by an exponent. The period is mandatory, so that NASM
1622 can distinguish between \c{dd 1}, which declares an integer constant,
1623 and \c{dd 1.0} which declares a floating-point constant.
1625 NASM also support C99-style hexadecimal floating-point: \c{0x},
1626 hexadecimal digits, period, optionally more hexadeximal digits, then
1627 optionally a \c{P} followed by a \e{binary} (not hexadecimal) exponent
1628 in decimal notation. As an extension, NASM additionally supports the
1629 \c{0h} and \c{$} prefixes for hexadecimal, as well binary and octal
1630 floating-point, using the \c{0b} or \c{0y} and \c{0o} or \c{0q}
1631 prefixes, respectively.
1633 Underscores to break up groups of digits are permitted in
1634 floating-point constants as well.
1638 \c db -0.2 ; "Quarter precision"
1639 \c dw -0.5 ; IEEE 754r/SSE5 half precision
1640 \c dd 1.2 ; an easy one
1641 \c dd 1.222_222_222 ; underscores are permitted
1642 \c dd 0x1p+2 ; 1.0x2^2 = 4.0
1643 \c dq 0x1p+32 ; 1.0x2^32 = 4 294 967 296.0
1644 \c dq 1.e10 ; 10 000 000 000.0
1645 \c dq 1.e+10 ; synonymous with 1.e10
1646 \c dq 1.e-10 ; 0.000 000 000 1
1647 \c dt 3.141592653589793238462 ; pi
1648 \c do 1.e+4000 ; IEEE 754r quad precision
1650 The 8-bit "quarter-precision" floating-point format is
1651 sign:exponent:mantissa = 1:4:3 with an exponent bias of 7. This
1652 appears to be the most frequently used 8-bit floating-point format,
1653 although it is not covered by any formal standard. This is sometimes
1654 called a "\i{minifloat}."
1656 The special operators are used to produce floating-point numbers in
1657 other contexts. They produce the binary representation of a specific
1658 floating-point number as an integer, and can use anywhere integer
1659 constants are used in an expression. \c{__float80m__} and
1660 \c{__float80e__} produce the 64-bit mantissa and 16-bit exponent of an
1661 80-bit floating-point number, and \c{__float128l__} and
1662 \c{__float128h__} produce the lower and upper 64-bit halves of a 128-bit
1663 floating-point number, respectively.
1667 \c mov rax,__float64__(3.141592653589793238462)
1669 ... would assign the binary representation of pi as a 64-bit floating
1670 point number into \c{RAX}. This is exactly equivalent to:
1672 \c mov rax,0x400921fb54442d18
1674 NASM cannot do compile-time arithmetic on floating-point constants.
1675 This is because NASM is designed to be portable - although it always
1676 generates code to run on x86 processors, the assembler itself can
1677 run on any system with an ANSI C compiler. Therefore, the assembler
1678 cannot guarantee the presence of a floating-point unit capable of
1679 handling the \i{Intel number formats}, and so for NASM to be able to
1680 do floating arithmetic it would have to include its own complete set
1681 of floating-point routines, which would significantly increase the
1682 size of the assembler for very little benefit.
1684 The special tokens \i\c{__Infinity__}, \i\c{__QNaN__} (or
1685 \i\c{__NaN__}) and \i\c{__SNaN__} can be used to generate
1686 \I{infinity}infinities, quiet \i{NaN}s, and signalling NaNs,
1687 respectively. These are normally used as macros:
1689 \c %define Inf __Infinity__
1690 \c %define NaN __QNaN__
1692 \c dq +1.5, -Inf, NaN ; Double-precision constants
1694 \S{bcdconst} \I{floating-point, packed BCD constants}Packed BCD Constants
1696 x87-style packed BCD constants can be used in the same contexts as
1697 80-bit floating-point numbers. They are suffixed with \c{p} or
1698 prefixed with \c{0p}, and can include up to 18 decimal digits.
1700 As with other numeric constants, underscores can be used to separate
1705 \c dt 12_345_678_901_245_678p
1706 \c dt -12_345_678_901_245_678p
1711 \H{expr} \i{Expressions}
1713 Expressions in NASM are similar in syntax to those in C. Expressions
1714 are evaluated as 64-bit integers which are then adjusted to the
1717 NASM supports two special tokens in expressions, allowing
1718 calculations to involve the current assembly position: the
1719 \I{$, here}\c{$} and \i\c{$$} tokens. \c{$} evaluates to the assembly
1720 position at the beginning of the line containing the expression; so
1721 you can code an \i{infinite loop} using \c{JMP $}. \c{$$} evaluates
1722 to the beginning of the current section; so you can tell how far
1723 into the section you are by using \c{($-$$)}.
1725 The arithmetic \i{operators} provided by NASM are listed here, in
1726 increasing order of \i{precedence}.
1729 \S{expor} \i\c{|}: \i{Bitwise OR} Operator
1731 The \c{|} operator gives a bitwise OR, exactly as performed by the
1732 \c{OR} machine instruction. Bitwise OR is the lowest-priority
1733 arithmetic operator supported by NASM.
1736 \S{expxor} \i\c{^}: \i{Bitwise XOR} Operator
1738 \c{^} provides the bitwise XOR operation.
1741 \S{expand} \i\c{&}: \i{Bitwise AND} Operator
1743 \c{&} provides the bitwise AND operation.
1746 \S{expshift} \i\c{<<} and \i\c{>>}: \i{Bit Shift} Operators
1748 \c{<<} gives a bit-shift to the left, just as it does in C. So \c{5<<3}
1749 evaluates to 5 times 8, or 40. \c{>>} gives a bit-shift to the
1750 right; in NASM, such a shift is \e{always} unsigned, so that
1751 the bits shifted in from the left-hand end are filled with zero
1752 rather than a sign-extension of the previous highest bit.
1755 \S{expplmi} \I{+ opaddition}\c{+} and \I{- opsubtraction}\c{-}:
1756 \i{Addition} and \i{Subtraction} Operators
1758 The \c{+} and \c{-} operators do perfectly ordinary addition and
1762 \S{expmul} \i\c{*}, \i\c{/}, \i\c{//}, \i\c{%} and \i\c{%%}:
1763 \i{Multiplication} and \i{Division}
1765 \c{*} is the multiplication operator. \c{/} and \c{//} are both
1766 division operators: \c{/} is \i{unsigned division} and \c{//} is
1767 \i{signed division}. Similarly, \c{%} and \c{%%} provide \I{unsigned
1768 modulo}\I{modulo operators}unsigned and
1769 \i{signed modulo} operators respectively.
1771 NASM, like ANSI C, provides no guarantees about the sensible
1772 operation of the signed modulo operator.
1774 Since the \c{%} character is used extensively by the macro
1775 \i{preprocessor}, you should ensure that both the signed and unsigned
1776 modulo operators are followed by white space wherever they appear.
1779 \S{expmul} \i{Unary Operators}: \I{+ opunary}\c{+}, \I{- opunary}\c{-},
1780 \i\c{~}, \I{! opunary}\c{!} and \i\c{SEG}
1782 The highest-priority operators in NASM's expression grammar are
1783 those which only apply to one argument. \c{-} negates its operand,
1784 \c{+} does nothing (it's provided for symmetry with \c{-}), \c{~}
1785 computes the \i{one's complement} of its operand, \c{!} is the
1786 \i{logical negation} operator, and \c{SEG} provides the \i{segment address}
1787 of its operand (explained in more detail in \k{segwrt}).
1790 \H{segwrt} \i\c{SEG} and \i\c{WRT}
1792 When writing large 16-bit programs, which must be split into
1793 multiple \i{segments}, it is often necessary to be able to refer to
1794 the \I{segment address}segment part of the address of a symbol. NASM
1795 supports the \c{SEG} operator to perform this function.
1797 The \c{SEG} operator returns the \i\e{preferred} segment base of a
1798 symbol, defined as the segment base relative to which the offset of
1799 the symbol makes sense. So the code
1801 \c mov ax,seg symbol
1805 will load \c{ES:BX} with a valid pointer to the symbol \c{symbol}.
1807 Things can be more complex than this: since 16-bit segments and
1808 \i{groups} may \I{overlapping segments}overlap, you might occasionally
1809 want to refer to some symbol using a different segment base from the
1810 preferred one. NASM lets you do this, by the use of the \c{WRT}
1811 (With Reference To) keyword. So you can do things like
1813 \c mov ax,weird_seg ; weird_seg is a segment base
1815 \c mov bx,symbol wrt weird_seg
1817 to load \c{ES:BX} with a different, but functionally equivalent,
1818 pointer to the symbol \c{symbol}.
1820 NASM supports far (inter-segment) calls and jumps by means of the
1821 syntax \c{call segment:offset}, where \c{segment} and \c{offset}
1822 both represent immediate values. So to call a far procedure, you
1823 could code either of
1825 \c call (seg procedure):procedure
1826 \c call weird_seg:(procedure wrt weird_seg)
1828 (The parentheses are included for clarity, to show the intended
1829 parsing of the above instructions. They are not necessary in
1832 NASM supports the syntax \I\c{CALL FAR}\c{call far procedure} as a
1833 synonym for the first of the above usages. \c{JMP} works identically
1834 to \c{CALL} in these examples.
1836 To declare a \i{far pointer} to a data item in a data segment, you
1839 \c dw symbol, seg symbol
1841 NASM supports no convenient synonym for this, though you can always
1842 invent one using the macro processor.
1845 \H{strict} \i\c{STRICT}: Inhibiting Optimization
1847 When assembling with the optimizer set to full (see \k{opt-O}),
1848 NASM will use size specifiers (\c{BYTE}, \c{WORD},
1849 \c{DWORD}, \c{QWORD}, \c{TWORD}, \c{OWORD} or \c{YWORD}), but will
1850 give them the smallest possible size. The keyword \c{STRICT} can be
1851 used to inhibit optimization and force a particular operand to be
1852 emitted in the specified size. For example, with the optimizer on, and
1853 in \c{BITS 16} mode,
1857 is encoded in three bytes \c{66 6A 21}, whereas
1859 \c push strict dword 33
1861 is encoded in six bytes, with a full dword immediate operand \c{66 68
1864 With the optimizer off, the same code (six bytes) is generated whether
1865 the \c{STRICT} keyword was used or not.
1868 \H{crit} \i{Critical Expressions}
1870 Although NASM has an optional multi-pass optimizer, there are some
1871 expressions which must be resolvable on the first pass. These are
1872 called \e{Critical Expressions}.
1874 The first pass is used to determine the size of all the assembled
1875 code and data, so that the second pass, when generating all the
1876 code, knows all the symbol addresses the code refers to. So one
1877 thing NASM can't handle is code whose size depends on the value of a
1878 symbol declared after the code in question. For example,
1880 \c times (label-$) db 0
1881 \c label: db 'Where am I?'
1883 The argument to \i\c{TIMES} in this case could equally legally
1884 evaluate to anything at all; NASM will reject this example because
1885 it cannot tell the size of the \c{TIMES} line when it first sees it.
1886 It will just as firmly reject the slightly \I{paradox}paradoxical
1889 \c times (label-$+1) db 0
1890 \c label: db 'NOW where am I?'
1892 in which \e{any} value for the \c{TIMES} argument is by definition
1895 NASM rejects these examples by means of a concept called a
1896 \e{critical expression}, which is defined to be an expression whose
1897 value is required to be computable in the first pass, and which must
1898 therefore depend only on symbols defined before it. The argument to
1899 the \c{TIMES} prefix is a critical expression.
1901 \H{locallab} \i{Local Labels}
1903 NASM gives special treatment to symbols beginning with a \i{period}.
1904 A label beginning with a single period is treated as a \e{local}
1905 label, which means that it is associated with the previous non-local
1906 label. So, for example:
1908 \c label1 ; some code
1916 \c label2 ; some code
1924 In the above code fragment, each \c{JNE} instruction jumps to the
1925 line immediately before it, because the two definitions of \c{.loop}
1926 are kept separate by virtue of each being associated with the
1927 previous non-local label.
1929 This form of local label handling is borrowed from the old Amiga
1930 assembler \i{DevPac}; however, NASM goes one step further, in
1931 allowing access to local labels from other parts of the code. This
1932 is achieved by means of \e{defining} a local label in terms of the
1933 previous non-local label: the first definition of \c{.loop} above is
1934 really defining a symbol called \c{label1.loop}, and the second
1935 defines a symbol called \c{label2.loop}. So, if you really needed
1938 \c label3 ; some more code
1943 Sometimes it is useful - in a macro, for instance - to be able to
1944 define a label which can be referenced from anywhere but which
1945 doesn't interfere with the normal local-label mechanism. Such a
1946 label can't be non-local because it would interfere with subsequent
1947 definitions of, and references to, local labels; and it can't be
1948 local because the macro that defined it wouldn't know the label's
1949 full name. NASM therefore introduces a third type of label, which is
1950 probably only useful in macro definitions: if a label begins with
1951 the \I{label prefix}special prefix \i\c{..@}, then it does nothing
1952 to the local label mechanism. So you could code
1954 \c label1: ; a non-local label
1955 \c .local: ; this is really label1.local
1956 \c ..@foo: ; this is a special symbol
1957 \c label2: ; another non-local label
1958 \c .local: ; this is really label2.local
1960 \c jmp ..@foo ; this will jump three lines up
1962 NASM has the capacity to define other special symbols beginning with
1963 a double period: for example, \c{..start} is used to specify the
1964 entry point in the \c{obj} output format (see \k{dotdotstart}),
1965 \c{..imagebase} is used to find out the offset from a base address
1966 of the current image in the \c{win64} output format (see \k{win64pic}).
1967 So just keep in mind that symbols beginning with a double period are
1971 \C{preproc} The NASM \i{Preprocessor}
1973 NASM contains a powerful \i{macro processor}, which supports
1974 conditional assembly, multi-level file inclusion, two forms of macro
1975 (single-line and multi-line), and a `context stack' mechanism for
1976 extra macro power. Preprocessor directives all begin with a \c{%}
1979 The preprocessor collapses all lines which end with a backslash (\\)
1980 character into a single line. Thus:
1982 \c %define THIS_VERY_LONG_MACRO_NAME_IS_DEFINED_TO \\
1985 will work like a single-line macro without the backslash-newline
1988 \H{slmacro} \i{Single-Line Macros}
1990 \S{define} The Normal Way: \I\c{%idefine}\i\c{%define}
1992 Single-line macros are defined using the \c{%define} preprocessor
1993 directive. The definitions work in a similar way to C; so you can do
1996 \c %define ctrl 0x1F &
1997 \c %define param(a,b) ((a)+(a)*(b))
1999 \c mov byte [param(2,ebx)], ctrl 'D'
2001 which will expand to
2003 \c mov byte [(2)+(2)*(ebx)], 0x1F & 'D'
2005 When the expansion of a single-line macro contains tokens which
2006 invoke another macro, the expansion is performed at invocation time,
2007 not at definition time. Thus the code
2009 \c %define a(x) 1+b(x)
2014 will evaluate in the expected way to \c{mov ax,1+2*8}, even though
2015 the macro \c{b} wasn't defined at the time of definition of \c{a}.
2017 Macros defined with \c{%define} are \i{case sensitive}: after
2018 \c{%define foo bar}, only \c{foo} will expand to \c{bar}: \c{Foo} or
2019 \c{FOO} will not. By using \c{%idefine} instead of \c{%define} (the
2020 `i' stands for `insensitive') you can define all the case variants
2021 of a macro at once, so that \c{%idefine foo bar} would cause
2022 \c{foo}, \c{Foo}, \c{FOO}, \c{fOO} and so on all to expand to
2025 There is a mechanism which detects when a macro call has occurred as
2026 a result of a previous expansion of the same macro, to guard against
2027 \i{circular references} and infinite loops. If this happens, the
2028 preprocessor will only expand the first occurrence of the macro.
2031 \c %define a(x) 1+a(x)
2035 the macro \c{a(3)} will expand once, becoming \c{1+a(3)}, and will
2036 then expand no further. This behaviour can be useful: see \k{32c}
2037 for an example of its use.
2039 You can \I{overloading, single-line macros}overload single-line
2040 macros: if you write
2042 \c %define foo(x) 1+x
2043 \c %define foo(x,y) 1+x*y
2045 the preprocessor will be able to handle both types of macro call,
2046 by counting the parameters you pass; so \c{foo(3)} will become
2047 \c{1+3} whereas \c{foo(ebx,2)} will become \c{1+ebx*2}. However, if
2052 then no other definition of \c{foo} will be accepted: a macro with
2053 no parameters prohibits the definition of the same name as a macro
2054 \e{with} parameters, and vice versa.
2056 This doesn't prevent single-line macros being \e{redefined}: you can
2057 perfectly well define a macro with
2061 and then re-define it later in the same source file with
2065 Then everywhere the macro \c{foo} is invoked, it will be expanded
2066 according to the most recent definition. This is particularly useful
2067 when defining single-line macros with \c{%assign} (see \k{assign}).
2069 You can \i{pre-define} single-line macros using the `-d' option on
2070 the NASM command line: see \k{opt-d}.
2073 \S{xdefine} Resolving \c{%define}: \I\c{%ixdefine}\i\c{%xdefine}
2075 To have a reference to an embedded single-line macro resolved at the
2076 time that the embedding macro is \e{defined}, as opposed to when the
2077 embedding macro is \e{expanded}, you need a different mechanism to the
2078 one offered by \c{%define}. The solution is to use \c{%xdefine}, or
2079 it's \I{case sensitive}case-insensitive counterpart \c{%ixdefine}.
2081 Suppose you have the following code:
2084 \c %define isFalse isTrue
2093 In this case, \c{val1} is equal to 0, and \c{val2} is equal to 1.
2094 This is because, when a single-line macro is defined using
2095 \c{%define}, it is expanded only when it is called. As \c{isFalse}
2096 expands to \c{isTrue}, the expansion will be the current value of
2097 \c{isTrue}. The first time it is called that is 0, and the second
2100 If you wanted \c{isFalse} to expand to the value assigned to the
2101 embedded macro \c{isTrue} at the time that \c{isFalse} was defined,
2102 you need to change the above code to use \c{%xdefine}.
2104 \c %xdefine isTrue 1
2105 \c %xdefine isFalse isTrue
2106 \c %xdefine isTrue 0
2110 \c %xdefine isTrue 1
2114 Now, each time that \c{isFalse} is called, it expands to 1,
2115 as that is what the embedded macro \c{isTrue} expanded to at
2116 the time that \c{isFalse} was defined.
2119 \S{indmacro} \i{Macro Indirection}: \I\c{%[}\c{%[...]}
2121 The \c{%[...]} construct can be used to expand macros in contexts
2122 where macro expansion would otherwise not occur, including in the
2123 names other macros. For example, if you have a set of macros named
2124 \c{Foo16}, \c{Foo32} and \c{Foo64}, you could write:
2126 \c mov ax,Foo%[__BITS__] ; The Foo value
2128 to use the builtin macro \c{__BITS__} (see \k{bitsm}) to automatically
2129 select between them. Similarly, the two statements:
2131 \c %xdefine Bar Quux ; Expands due to %xdefine
2132 \c %define Bar %[Quux] ; Expands due to %[...]
2134 have, in fact, exactly the same effect.
2136 \c{%[...]} concatenates to adjacent tokens in the same way that
2137 multi-line macro parameters do, see \k{concat} for details.
2140 \S{concat%+} Concatenating Single Line Macro Tokens: \i\c{%+}
2142 Individual tokens in single line macros can be concatenated, to produce
2143 longer tokens for later processing. This can be useful if there are
2144 several similar macros that perform similar functions.
2146 Please note that a space is required after \c{%+}, in order to
2147 disambiguate it from the syntax \c{%+1} used in multiline macros.
2149 As an example, consider the following:
2151 \c %define BDASTART 400h ; Start of BIOS data area
2153 \c struc tBIOSDA ; its structure
2159 Now, if we need to access the elements of tBIOSDA in different places,
2162 \c mov ax,BDASTART + tBIOSDA.COM1addr
2163 \c mov bx,BDASTART + tBIOSDA.COM2addr
2165 This will become pretty ugly (and tedious) if used in many places, and
2166 can be reduced in size significantly by using the following macro:
2168 \c ; Macro to access BIOS variables by their names (from tBDA):
2170 \c %define BDA(x) BDASTART + tBIOSDA. %+ x
2172 Now the above code can be written as:
2174 \c mov ax,BDA(COM1addr)
2175 \c mov bx,BDA(COM2addr)
2177 Using this feature, we can simplify references to a lot of macros (and,
2178 in turn, reduce typing errors).
2181 \S{selfref%?} The Macro Name Itself: \i\c{%?} and \i\c{%??}
2183 The special symbols \c{%?} and \c{%??} can be used to reference the
2184 macro name itself inside a macro expansion, this is supported for both
2185 single-and multi-line macros. \c{%?} refers to the macro name as
2186 \e{invoked}, whereas \c{%??} refers to the macro name as
2187 \e{declared}. The two are always the same for case-sensitive
2188 macros, but for case-insensitive macros, they can differ.
2192 \c %idefine Foo mov %?,%??
2204 \c %idefine keyword $%?
2206 can be used to make a keyword "disappear", for example in case a new
2207 instruction has been used as a label in older code. For example:
2209 \c %idefine pause $%? ; Hide the PAUSE instruction
2212 \S{undef} Undefining Single-Line Macros: \i\c{%undef}
2214 Single-line macros can be removed with the \c{%undef} directive. For
2215 example, the following sequence:
2222 will expand to the instruction \c{mov eax, foo}, since after
2223 \c{%undef} the macro \c{foo} is no longer defined.
2225 Macros that would otherwise be pre-defined can be undefined on the
2226 command-line using the `-u' option on the NASM command line: see
2230 \S{assign} \i{Preprocessor Variables}: \i\c{%assign}
2232 An alternative way to define single-line macros is by means of the
2233 \c{%assign} command (and its \I{case sensitive}case-insensitive
2234 counterpart \i\c{%iassign}, which differs from \c{%assign} in
2235 exactly the same way that \c{%idefine} differs from \c{%define}).
2237 \c{%assign} is used to define single-line macros which take no
2238 parameters and have a numeric value. This value can be specified in
2239 the form of an expression, and it will be evaluated once, when the
2240 \c{%assign} directive is processed.
2242 Like \c{%define}, macros defined using \c{%assign} can be re-defined
2243 later, so you can do things like
2247 to increment the numeric value of a macro.
2249 \c{%assign} is useful for controlling the termination of \c{%rep}
2250 preprocessor loops: see \k{rep} for an example of this. Another
2251 use for \c{%assign} is given in \k{16c} and \k{32c}.
2253 The expression passed to \c{%assign} is a \i{critical expression}
2254 (see \k{crit}), and must also evaluate to a pure number (rather than
2255 a relocatable reference such as a code or data address, or anything
2256 involving a register).
2259 \S{defstr} Defining Strings: \I\c{%idefstr}\i\c{%defstr}
2261 \c{%defstr}, and its case-insensitive counterpart \c{%idefstr}, define
2262 or redefine a single-line macro without parameters but converts the
2263 entire right-hand side, after macro expansion, to a quoted string
2268 \c %defstr test TEST
2272 \c %define test 'TEST'
2274 This can be used, for example, with the \c{%!} construct (see
2277 \c %defstr PATH %!PATH ; The operating system PATH variable
2280 \S{deftok} Defining Tokens: \I\c{%ideftok}\i\c{%deftok}
2282 \c{%deftok}, and its case-insensitive counterpart \c{%ideftok}, define
2283 or redefine a single-line macro without parameters but converts the
2284 second parameter, after string conversion, to a sequence of tokens.
2288 \c %deftok test 'TEST'
2292 \c %define test TEST
2295 \H{strlen} \i{String Manipulation in Macros}
2297 It's often useful to be able to handle strings in macros. NASM
2298 supports a few simple string handling macro operators from which
2299 more complex operations can be constructed.
2301 All the string operators define or redefine a value (either a string
2302 or a numeric value) to a single-line macro. When producing a string
2303 value, it may change the style of quoting of the input string or
2304 strings, and possibly use \c{\\}-escapes inside \c{`}-quoted strings.
2306 \S{strcat} \i{Concatenating Strings}: \i\c{%strcat}
2308 The \c{%strcat} operator concatenates quoted strings and assign them to
2309 a single-line macro.
2313 \c %strcat alpha "Alpha: ", '12" screen'
2315 ... would assign the value \c{'Alpha: 12" screen'} to \c{alpha}.
2318 \c %strcat beta '"foo"\', "'bar'"
2320 ... would assign the value \c{`"foo"\\\\'bar'`} to \c{beta}.
2322 The use of commas to separate strings is permitted but optional.
2325 \S{strlen} \i{String Length}: \i\c{%strlen}
2327 The \c{%strlen} operator assigns the length of a string to a macro.
2330 \c %strlen charcnt 'my string'
2332 In this example, \c{charcnt} would receive the value 9, just as
2333 if an \c{%assign} had been used. In this example, \c{'my string'}
2334 was a literal string but it could also have been a single-line
2335 macro that expands to a string, as in the following example:
2337 \c %define sometext 'my string'
2338 \c %strlen charcnt sometext
2340 As in the first case, this would result in \c{charcnt} being
2341 assigned the value of 9.
2344 \S{substr} \i{Extracting Substrings}: \i\c{%substr}
2346 Individual letters or substrings in strings can be extracted using the
2347 \c{%substr} operator. An example of its use is probably more useful
2348 than the description:
2350 \c %substr mychar 'xyzw' 1 ; equivalent to %define mychar 'x'
2351 \c %substr mychar 'xyzw' 2 ; equivalent to %define mychar 'y'
2352 \c %substr mychar 'xyzw' 3 ; equivalent to %define mychar 'z'
2353 \c %substr mychar 'xyzw' 2,2 ; equivalent to %define mychar 'yz'
2354 \c %substr mychar 'xyzw' 2,-1 ; equivalent to %define mychar 'yzw'
2355 \c %substr mychar 'xyzw' 2,-2 ; equivalent to %define mychar 'yz'
2357 As with \c{%strlen} (see \k{strlen}), the first parameter is the
2358 single-line macro to be created and the second is the string. The
2359 third parameter specifies the first character to be selected, and the
2360 optional fourth parameter preceeded by comma) is the length. Note
2361 that the first index is 1, not 0 and the last index is equal to the
2362 value that \c{%strlen} would assign given the same string. Index
2363 values out of range result in an empty string. A negative length
2364 means "until N-1 characters before the end of string", i.e. \c{-1}
2365 means until end of string, \c{-2} until one character before, etc.
2368 \H{mlmacro} \i{Multi-Line Macros}: \I\c{%imacro}\i\c{%macro}
2370 Multi-line macros are much more like the type of macro seen in MASM
2371 and TASM: a multi-line macro definition in NASM looks something like
2374 \c %macro prologue 1
2382 This defines a C-like function prologue as a macro: so you would
2383 invoke the macro with a call such as
2385 \c myfunc: prologue 12
2387 which would expand to the three lines of code
2393 The number \c{1} after the macro name in the \c{%macro} line defines
2394 the number of parameters the macro \c{prologue} expects to receive.
2395 The use of \c{%1} inside the macro definition refers to the first
2396 parameter to the macro call. With a macro taking more than one
2397 parameter, subsequent parameters would be referred to as \c{%2},
2400 Multi-line macros, like single-line macros, are \i{case-sensitive},
2401 unless you define them using the alternative directive \c{%imacro}.
2403 If you need to pass a comma as \e{part} of a parameter to a
2404 multi-line macro, you can do that by enclosing the entire parameter
2405 in \I{braces, around macro parameters}braces. So you could code
2414 \c silly 'a', letter_a ; letter_a: db 'a'
2415 \c silly 'ab', string_ab ; string_ab: db 'ab'
2416 \c silly {13,10}, crlf ; crlf: db 13,10
2419 \#\S{mlrmacro} \i{Recursive Multi-Line Macros}: \I\c{%irmacro}\i\c{%rmacro}
2421 \#A multi-line macro cannot be referenced within itself, in order to
2422 \#prevent accidental infinite recursion.
2424 \#Recursive multi-line macros allow for self-referencing, with the
2425 \#caveat that the user is aware of the existence, use and purpose of
2426 \#recursive multi-line macros. There is also a generous, but sane, upper
2427 \#limit to the number of recursions, in order to prevent run-away memory
2428 \#consumption in case of accidental infinite recursion.
2430 \#As with non-recursive multi-line macros, recursive multi-line macros are
2431 \#\i{case-sensitive}, unless you define them using the alternative
2432 \#directive \c{%irmacro}.
2434 \S{mlmacover} Overloading Multi-Line Macros\I{overloading, multi-line macros}
2436 As with single-line macros, multi-line macros can be overloaded by
2437 defining the same macro name several times with different numbers of
2438 parameters. This time, no exception is made for macros with no
2439 parameters at all. So you could define
2441 \c %macro prologue 0
2448 to define an alternative form of the function prologue which
2449 allocates no local stack space.
2451 Sometimes, however, you might want to `overload' a machine
2452 instruction; for example, you might want to define
2461 so that you could code
2463 \c push ebx ; this line is not a macro call
2464 \c push eax,ecx ; but this one is
2466 Ordinarily, NASM will give a warning for the first of the above two
2467 lines, since \c{push} is now defined to be a macro, and is being
2468 invoked with a number of parameters for which no definition has been
2469 given. The correct code will still be generated, but the assembler
2470 will give a warning. This warning can be disabled by the use of the
2471 \c{-w-macro-params} command-line option (see \k{opt-w}).
2474 \S{maclocal} \i{Macro-Local Labels}
2476 NASM allows you to define labels within a multi-line macro
2477 definition in such a way as to make them local to the macro call: so
2478 calling the same macro multiple times will use a different label
2479 each time. You do this by prefixing \i\c{%%} to the label name. So
2480 you can invent an instruction which executes a \c{RET} if the \c{Z}
2481 flag is set by doing this:
2491 You can call this macro as many times as you want, and every time
2492 you call it NASM will make up a different `real' name to substitute
2493 for the label \c{%%skip}. The names NASM invents are of the form
2494 \c{..@2345.skip}, where the number 2345 changes with every macro
2495 call. The \i\c{..@} prefix prevents macro-local labels from
2496 interfering with the local label mechanism, as described in
2497 \k{locallab}. You should avoid defining your own labels in this form
2498 (the \c{..@} prefix, then a number, then another period) in case
2499 they interfere with macro-local labels.
2502 \S{mlmacgre} \i{Greedy Macro Parameters}
2504 Occasionally it is useful to define a macro which lumps its entire
2505 command line into one parameter definition, possibly after
2506 extracting one or two smaller parameters from the front. An example
2507 might be a macro to write a text string to a file in MS-DOS, where
2508 you might want to be able to write
2510 \c writefile [filehandle],"hello, world",13,10
2512 NASM allows you to define the last parameter of a macro to be
2513 \e{greedy}, meaning that if you invoke the macro with more
2514 parameters than it expects, all the spare parameters get lumped into
2515 the last defined one along with the separating commas. So if you
2518 \c %macro writefile 2+
2524 \c mov cx,%%endstr-%%str
2531 then the example call to \c{writefile} above will work as expected:
2532 the text before the first comma, \c{[filehandle]}, is used as the
2533 first macro parameter and expanded when \c{%1} is referred to, and
2534 all the subsequent text is lumped into \c{%2} and placed after the
2537 The greedy nature of the macro is indicated to NASM by the use of
2538 the \I{+ modifier}\c{+} sign after the parameter count on the
2541 If you define a greedy macro, you are effectively telling NASM how
2542 it should expand the macro given \e{any} number of parameters from
2543 the actual number specified up to infinity; in this case, for
2544 example, NASM now knows what to do when it sees a call to
2545 \c{writefile} with 2, 3, 4 or more parameters. NASM will take this
2546 into account when overloading macros, and will not allow you to
2547 define another form of \c{writefile} taking 4 parameters (for
2550 Of course, the above macro could have been implemented as a
2551 non-greedy macro, in which case the call to it would have had to
2554 \c writefile [filehandle], {"hello, world",13,10}
2556 NASM provides both mechanisms for putting \i{commas in macro
2557 parameters}, and you choose which one you prefer for each macro
2560 See \k{sectmac} for a better way to write the above macro.
2562 \S{mlmacrange} \i{Macro Parameters Range}
2564 NASM also allows you to expand parameters via special construction \c{%\{x:y\}}
2565 where \c{x} is the first parameter index and \c{y} is the last. Any index can
2566 be either negative or positive. Though the indices must never be zero.
2576 expands to \c{3,4,5} range.
2578 Even more, the parameters can be reversed so that
2586 expands to \c{5,4,3} range.
2588 But even this is not the last. The parameters can be addressed via negative
2589 indices so NASM will count them reversed. The ones who know Python may see
2598 expands to \c{6,5,4} range.
2600 Note that NASM uses \i{comma} to separate parameters being expanded.
2602 By the way, here is a trick - you might use the index \c{%{-1:-1}} gives
2603 you the \i{last} argument passed to a macro.
2605 \S{mlmacdef} \i{Default Macro Parameters}
2607 NASM also allows you to define a multi-line macro with a \e{range}
2608 of allowable parameter counts. If you do this, you can specify
2609 defaults for \i{omitted parameters}. So, for example:
2611 \c %macro die 0-1 "Painful program death has occurred."
2619 This macro (which makes use of the \c{writefile} macro defined in
2620 \k{mlmacgre}) can be called with an explicit error message, which it
2621 will display on the error output stream before exiting, or it can be
2622 called with no parameters, in which case it will use the default
2623 error message supplied in the macro definition.
2625 In general, you supply a minimum and maximum number of parameters
2626 for a macro of this type; the minimum number of parameters are then
2627 required in the macro call, and then you provide defaults for the
2628 optional ones. So if a macro definition began with the line
2630 \c %macro foobar 1-3 eax,[ebx+2]
2632 then it could be called with between one and three parameters, and
2633 \c{%1} would always be taken from the macro call. \c{%2}, if not
2634 specified by the macro call, would default to \c{eax}, and \c{%3} if
2635 not specified would default to \c{[ebx+2]}.
2637 You can provide extra information to a macro by providing
2638 too many default parameters:
2640 \c %macro quux 1 something
2642 This will trigger a warning by default; see \k{opt-w} for
2644 When \c{quux} is invoked, it receives not one but two parameters.
2645 \c{something} can be referred to as \c{%2}. The difference
2646 between passing \c{something} this way and writing \c{something}
2647 in the macro body is that with this way \c{something} is evaluated
2648 when the macro is defined, not when it is expanded.
2650 You may omit parameter defaults from the macro definition, in which
2651 case the parameter default is taken to be blank. This can be useful
2652 for macros which can take a variable number of parameters, since the
2653 \i\c{%0} token (see \k{percent0}) allows you to determine how many
2654 parameters were really passed to the macro call.
2656 This defaulting mechanism can be combined with the greedy-parameter
2657 mechanism; so the \c{die} macro above could be made more powerful,
2658 and more useful, by changing the first line of the definition to
2660 \c %macro die 0-1+ "Painful program death has occurred.",13,10
2662 The maximum parameter count can be infinite, denoted by \c{*}. In
2663 this case, of course, it is impossible to provide a \e{full} set of
2664 default parameters. Examples of this usage are shown in \k{rotate}.
2667 \S{percent0} \i\c{%0}: \I{counting macro parameters}Macro Parameter Counter
2669 The parameter reference \c{%0} will return a numeric constant giving the
2670 number of parameters received, that is, if \c{%0} is n then \c{%}n is the
2671 last parameter. \c{%0} is mostly useful for macros that can take a variable
2672 number of parameters. It can be used as an argument to \c{%rep}
2673 (see \k{rep}) in order to iterate through all the parameters of a macro.
2674 Examples are given in \k{rotate}.
2677 \S{rotate} \i\c{%rotate}: \i{Rotating Macro Parameters}
2679 Unix shell programmers will be familiar with the \I{shift
2680 command}\c{shift} shell command, which allows the arguments passed
2681 to a shell script (referenced as \c{$1}, \c{$2} and so on) to be
2682 moved left by one place, so that the argument previously referenced
2683 as \c{$2} becomes available as \c{$1}, and the argument previously
2684 referenced as \c{$1} is no longer available at all.
2686 NASM provides a similar mechanism, in the form of \c{%rotate}. As
2687 its name suggests, it differs from the Unix \c{shift} in that no
2688 parameters are lost: parameters rotated off the left end of the
2689 argument list reappear on the right, and vice versa.
2691 \c{%rotate} is invoked with a single numeric argument (which may be
2692 an expression). The macro parameters are rotated to the left by that
2693 many places. If the argument to \c{%rotate} is negative, the macro
2694 parameters are rotated to the right.
2696 \I{iterating over macro parameters}So a pair of macros to save and
2697 restore a set of registers might work as follows:
2699 \c %macro multipush 1-*
2708 This macro invokes the \c{PUSH} instruction on each of its arguments
2709 in turn, from left to right. It begins by pushing its first
2710 argument, \c{%1}, then invokes \c{%rotate} to move all the arguments
2711 one place to the left, so that the original second argument is now
2712 available as \c{%1}. Repeating this procedure as many times as there
2713 were arguments (achieved by supplying \c{%0} as the argument to
2714 \c{%rep}) causes each argument in turn to be pushed.
2716 Note also the use of \c{*} as the maximum parameter count,
2717 indicating that there is no upper limit on the number of parameters
2718 you may supply to the \i\c{multipush} macro.
2720 It would be convenient, when using this macro, to have a \c{POP}
2721 equivalent, which \e{didn't} require the arguments to be given in
2722 reverse order. Ideally, you would write the \c{multipush} macro
2723 call, then cut-and-paste the line to where the pop needed to be
2724 done, and change the name of the called macro to \c{multipop}, and
2725 the macro would take care of popping the registers in the opposite
2726 order from the one in which they were pushed.
2728 This can be done by the following definition:
2730 \c %macro multipop 1-*
2739 This macro begins by rotating its arguments one place to the
2740 \e{right}, so that the original \e{last} argument appears as \c{%1}.
2741 This is then popped, and the arguments are rotated right again, so
2742 the second-to-last argument becomes \c{%1}. Thus the arguments are
2743 iterated through in reverse order.
2746 \S{concat} \i{Concatenating Macro Parameters}
2748 NASM can concatenate macro parameters and macro indirection constructs
2749 on to other text surrounding them. This allows you to declare a family
2750 of symbols, for example, in a macro definition. If, for example, you
2751 wanted to generate a table of key codes along with offsets into the
2752 table, you could code something like
2754 \c %macro keytab_entry 2
2756 \c keypos%1 equ $-keytab
2762 \c keytab_entry F1,128+1
2763 \c keytab_entry F2,128+2
2764 \c keytab_entry Return,13
2766 which would expand to
2769 \c keyposF1 equ $-keytab
2771 \c keyposF2 equ $-keytab
2773 \c keyposReturn equ $-keytab
2776 You can just as easily concatenate text on to the other end of a
2777 macro parameter, by writing \c{%1foo}.
2779 If you need to append a \e{digit} to a macro parameter, for example
2780 defining labels \c{foo1} and \c{foo2} when passed the parameter
2781 \c{foo}, you can't code \c{%11} because that would be taken as the
2782 eleventh macro parameter. Instead, you must code
2783 \I{braces, after % sign}\c{%\{1\}1}, which will separate the first
2784 \c{1} (giving the number of the macro parameter) from the second
2785 (literal text to be concatenated to the parameter).
2787 This concatenation can also be applied to other preprocessor in-line
2788 objects, such as macro-local labels (\k{maclocal}) and context-local
2789 labels (\k{ctxlocal}). In all cases, ambiguities in syntax can be
2790 resolved by enclosing everything after the \c{%} sign and before the
2791 literal text in braces: so \c{%\{%foo\}bar} concatenates the text
2792 \c{bar} to the end of the real name of the macro-local label
2793 \c{%%foo}. (This is unnecessary, since the form NASM uses for the
2794 real names of macro-local labels means that the two usages
2795 \c{%\{%foo\}bar} and \c{%%foobar} would both expand to the same
2796 thing anyway; nevertheless, the capability is there.)
2798 The single-line macro indirection construct, \c{%[...]}
2799 (\k{indmacro}), behaves the same way as macro parameters for the
2800 purpose of concatenation.
2802 See also the \c{%+} operator, \k{concat%+}.
2805 \S{mlmaccc} \i{Condition Codes as Macro Parameters}
2807 NASM can give special treatment to a macro parameter which contains
2808 a condition code. For a start, you can refer to the macro parameter
2809 \c{%1} by means of the alternative syntax \i\c{%+1}, which informs
2810 NASM that this macro parameter is supposed to contain a condition
2811 code, and will cause the preprocessor to report an error message if
2812 the macro is called with a parameter which is \e{not} a valid
2815 Far more usefully, though, you can refer to the macro parameter by
2816 means of \i\c{%-1}, which NASM will expand as the \e{inverse}
2817 condition code. So the \c{retz} macro defined in \k{maclocal} can be
2818 replaced by a general \i{conditional-return macro} like this:
2828 This macro can now be invoked using calls like \c{retc ne}, which
2829 will cause the conditional-jump instruction in the macro expansion
2830 to come out as \c{JE}, or \c{retc po} which will make the jump a
2833 The \c{%+1} macro-parameter reference is quite happy to interpret
2834 the arguments \c{CXZ} and \c{ECXZ} as valid condition codes;
2835 however, \c{%-1} will report an error if passed either of these,
2836 because no inverse condition code exists.
2839 \S{nolist} \i{Disabling Listing Expansion}\I\c{.nolist}
2841 When NASM is generating a listing file from your program, it will
2842 generally expand multi-line macros by means of writing the macro
2843 call and then listing each line of the expansion. This allows you to
2844 see which instructions in the macro expansion are generating what
2845 code; however, for some macros this clutters the listing up
2848 NASM therefore provides the \c{.nolist} qualifier, which you can
2849 include in a macro definition to inhibit the expansion of the macro
2850 in the listing file. The \c{.nolist} qualifier comes directly after
2851 the number of parameters, like this:
2853 \c %macro foo 1.nolist
2857 \c %macro bar 1-5+.nolist a,b,c,d,e,f,g,h
2859 \S{unmacro} Undefining Multi-Line Macros: \i\c{%unmacro}
2861 Multi-line macros can be removed with the \c{%unmacro} directive.
2862 Unlike the \c{%undef} directive, however, \c{%unmacro} takes an
2863 argument specification, and will only remove \i{exact matches} with
2864 that argument specification.
2873 removes the previously defined macro \c{foo}, but
2880 does \e{not} remove the macro \c{bar}, since the argument
2881 specification does not match exactly.
2884 \#\S{exitmacro} Exiting Multi-Line Macros: \i\c{%exitmacro}
2886 \#Multi-line macro expansions can be arbitrarily terminated with
2887 \#the \c{%exitmacro} directive.
2899 \H{condasm} \i{Conditional Assembly}\I\c{%if}
2901 Similarly to the C preprocessor, NASM allows sections of a source
2902 file to be assembled only if certain conditions are met. The general
2903 syntax of this feature looks like this:
2906 \c ; some code which only appears if <condition> is met
2907 \c %elif<condition2>
2908 \c ; only appears if <condition> is not met but <condition2> is
2910 \c ; this appears if neither <condition> nor <condition2> was met
2913 The inverse forms \i\c{%ifn} and \i\c{%elifn} are also supported.
2915 The \i\c{%else} clause is optional, as is the \i\c{%elif} clause.
2916 You can have more than one \c{%elif} clause as well.
2918 There are a number of variants of the \c{%if} directive. Each has its
2919 corresponding \c{%elif}, \c{%ifn}, and \c{%elifn} directives; for
2920 example, the equivalents to the \c{%ifdef} directive are \c{%elifdef},
2921 \c{%ifndef}, and \c{%elifndef}.
2923 \S{ifdef} \i\c{%ifdef}: Testing Single-Line Macro Existence\I{testing,
2924 single-line macro existence}
2926 Beginning a conditional-assembly block with the line \c{%ifdef
2927 MACRO} will assemble the subsequent code if, and only if, a
2928 single-line macro called \c{MACRO} is defined. If not, then the
2929 \c{%elif} and \c{%else} blocks (if any) will be processed instead.
2931 For example, when debugging a program, you might want to write code
2934 \c ; perform some function
2936 \c writefile 2,"Function performed successfully",13,10
2938 \c ; go and do something else
2940 Then you could use the command-line option \c{-dDEBUG} to create a
2941 version of the program which produced debugging messages, and remove
2942 the option to generate the final release version of the program.
2944 You can test for a macro \e{not} being defined by using
2945 \i\c{%ifndef} instead of \c{%ifdef}. You can also test for macro
2946 definitions in \c{%elif} blocks by using \i\c{%elifdef} and
2950 \S{ifmacro} \i\c{%ifmacro}: Testing Multi-Line Macro
2951 Existence\I{testing, multi-line macro existence}
2953 The \c{%ifmacro} directive operates in the same way as the \c{%ifdef}
2954 directive, except that it checks for the existence of a multi-line macro.
2956 For example, you may be working with a large project and not have control
2957 over the macros in a library. You may want to create a macro with one
2958 name if it doesn't already exist, and another name if one with that name
2961 The \c{%ifmacro} is considered true if defining a macro with the given name
2962 and number of arguments would cause a definitions conflict. For example:
2964 \c %ifmacro MyMacro 1-3
2966 \c %error "MyMacro 1-3" causes a conflict with an existing macro.
2970 \c %macro MyMacro 1-3
2972 \c ; insert code to define the macro
2978 This will create the macro "MyMacro 1-3" if no macro already exists which
2979 would conflict with it, and emits a warning if there would be a definition
2982 You can test for the macro not existing by using the \i\c{%ifnmacro} instead
2983 of \c{%ifmacro}. Additional tests can be performed in \c{%elif} blocks by using
2984 \i\c{%elifmacro} and \i\c{%elifnmacro}.
2987 \S{ifctx} \i\c{%ifctx}: Testing the Context Stack\I{testing, context
2990 The conditional-assembly construct \c{%ifctx} will cause the
2991 subsequent code to be assembled if and only if the top context on
2992 the preprocessor's context stack has the same name as one of the arguments.
2993 As with \c{%ifdef}, the inverse and \c{%elif} forms \i\c{%ifnctx},
2994 \i\c{%elifctx} and \i\c{%elifnctx} are also supported.
2996 For more details of the context stack, see \k{ctxstack}. For a
2997 sample use of \c{%ifctx}, see \k{blockif}.
3000 \S{if} \i\c{%if}: Testing Arbitrary Numeric Expressions\I{testing,
3001 arbitrary numeric expressions}
3003 The conditional-assembly construct \c{%if expr} will cause the
3004 subsequent code to be assembled if and only if the value of the
3005 numeric expression \c{expr} is non-zero. An example of the use of
3006 this feature is in deciding when to break out of a \c{%rep}
3007 preprocessor loop: see \k{rep} for a detailed example.
3009 The expression given to \c{%if}, and its counterpart \i\c{%elif}, is
3010 a critical expression (see \k{crit}).
3012 \c{%if} extends the normal NASM expression syntax, by providing a
3013 set of \i{relational operators} which are not normally available in
3014 expressions. The operators \i\c{=}, \i\c{<}, \i\c{>}, \i\c{<=},
3015 \i\c{>=} and \i\c{<>} test equality, less-than, greater-than,
3016 less-or-equal, greater-or-equal and not-equal respectively. The
3017 C-like forms \i\c{==} and \i\c{!=} are supported as alternative
3018 forms of \c{=} and \c{<>}. In addition, low-priority logical
3019 operators \i\c{&&}, \i\c{^^} and \i\c{||} are provided, supplying
3020 \i{logical AND}, \i{logical XOR} and \i{logical OR}. These work like
3021 the C logical operators (although C has no logical XOR), in that
3022 they always return either 0 or 1, and treat any non-zero input as 1
3023 (so that \c{^^}, for example, returns 1 if exactly one of its inputs
3024 is zero, and 0 otherwise). The relational operators also return 1
3025 for true and 0 for false.
3027 Like other \c{%if} constructs, \c{%if} has a counterpart
3028 \i\c{%elif}, and negative forms \i\c{%ifn} and \i\c{%elifn}.
3030 \S{ifidn} \i\c{%ifidn} and \i\c{%ifidni}: Testing Exact Text
3031 Identity\I{testing, exact text identity}
3033 The construct \c{%ifidn text1,text2} will cause the subsequent code
3034 to be assembled if and only if \c{text1} and \c{text2}, after
3035 expanding single-line macros, are identical pieces of text.
3036 Differences in white space are not counted.
3038 \c{%ifidni} is similar to \c{%ifidn}, but is \i{case-insensitive}.
3040 For example, the following macro pushes a register or number on the
3041 stack, and allows you to treat \c{IP} as a real register:
3043 \c %macro pushparam 1
3054 Like other \c{%if} constructs, \c{%ifidn} has a counterpart
3055 \i\c{%elifidn}, and negative forms \i\c{%ifnidn} and \i\c{%elifnidn}.
3056 Similarly, \c{%ifidni} has counterparts \i\c{%elifidni},
3057 \i\c{%ifnidni} and \i\c{%elifnidni}.
3059 \S{iftyp} \i\c{%ifid}, \i\c{%ifnum}, \i\c{%ifstr}: Testing Token
3060 Types\I{testing, token types}
3062 Some macros will want to perform different tasks depending on
3063 whether they are passed a number, a string, or an identifier. For
3064 example, a string output macro might want to be able to cope with
3065 being passed either a string constant or a pointer to an existing
3068 The conditional assembly construct \c{%ifid}, taking one parameter
3069 (which may be blank), assembles the subsequent code if and only if
3070 the first token in the parameter exists and is an identifier.
3071 \c{%ifnum} works similarly, but tests for the token being a numeric
3072 constant; \c{%ifstr} tests for it being a string.
3074 For example, the \c{writefile} macro defined in \k{mlmacgre} can be
3075 extended to take advantage of \c{%ifstr} in the following fashion:
3077 \c %macro writefile 2-3+
3086 \c %%endstr: mov dx,%%str
3087 \c mov cx,%%endstr-%%str
3098 Then the \c{writefile} macro can cope with being called in either of
3099 the following two ways:
3101 \c writefile [file], strpointer, length
3102 \c writefile [file], "hello", 13, 10
3104 In the first, \c{strpointer} is used as the address of an
3105 already-declared string, and \c{length} is used as its length; in
3106 the second, a string is given to the macro, which therefore declares
3107 it itself and works out the address and length for itself.
3109 Note the use of \c{%if} inside the \c{%ifstr}: this is to detect
3110 whether the macro was passed two arguments (so the string would be a
3111 single string constant, and \c{db %2} would be adequate) or more (in
3112 which case, all but the first two would be lumped together into
3113 \c{%3}, and \c{db %2,%3} would be required).
3115 The usual \I\c{%elifid}\I\c{%elifnum}\I\c{%elifstr}\c{%elif}...,
3116 \I\c{%ifnid}\I\c{%ifnnum}\I\c{%ifnstr}\c{%ifn}..., and
3117 \I\c{%elifnid}\I\c{%elifnnum}\I\c{%elifnstr}\c{%elifn}... versions
3118 exist for each of \c{%ifid}, \c{%ifnum} and \c{%ifstr}.
3120 \S{iftoken} \i\c{%iftoken}: Test for a Single Token
3122 Some macros will want to do different things depending on if it is
3123 passed a single token (e.g. paste it to something else using \c{%+})
3124 versus a multi-token sequence.
3126 The conditional assembly construct \c{%iftoken} assembles the
3127 subsequent code if and only if the expanded parameters consist of
3128 exactly one token, possibly surrounded by whitespace.
3134 will assemble the subsequent code, but
3138 will not, since \c{-1} contains two tokens: the unary minus operator
3139 \c{-}, and the number \c{1}.
3141 The usual \i\c{%eliftoken}, \i\c\{%ifntoken}, and \i\c{%elifntoken}
3142 variants are also provided.
3144 \S{ifempty} \i\c{%ifempty}: Test for Empty Expansion
3146 The conditional assembly construct \c{%ifempty} assembles the
3147 subsequent code if and only if the expanded parameters do not contain
3148 any tokens at all, whitespace excepted.
3150 The usual \i\c{%elifempty}, \i\c\{%ifnempty}, and \i\c{%elifnempty}
3151 variants are also provided.
3153 \S{ifenv} \i\c{%ifenv}: Test If Environment Variable Exists
3155 The conditional assembly construct \c{%ifenv} assembles the
3156 subsequent code if and only if the environment variable referenced by
3157 the \c{%!<env>} directive exists.
3159 The usual \i\c{%elifenv}, \i\c\{%ifnenv}, and \i\c{%elifnenv}
3160 variants are also provided.
3162 Just as for \c{%!<env>} the argument should be written as a string if
3163 it contains characters that would not be legal in an identifier. See
3166 \H{rep} \i{Preprocessor Loops}\I{repeating code}: \i\c{%rep}
3168 NASM's \c{TIMES} prefix, though useful, cannot be used to invoke a
3169 multi-line macro multiple times, because it is processed by NASM
3170 after macros have already been expanded. Therefore NASM provides
3171 another form of loop, this time at the preprocessor level: \c{%rep}.
3173 The directives \c{%rep} and \i\c{%endrep} (\c{%rep} takes a numeric
3174 argument, which can be an expression; \c{%endrep} takes no
3175 arguments) can be used to enclose a chunk of code, which is then
3176 replicated as many times as specified by the preprocessor:
3180 \c inc word [table+2*i]
3184 This will generate a sequence of 64 \c{INC} instructions,
3185 incrementing every word of memory from \c{[table]} to
3188 For more complex termination conditions, or to break out of a repeat
3189 loop part way along, you can use the \i\c{%exitrep} directive to
3190 terminate the loop, like this:
3205 \c fib_number equ ($-fibonacci)/2
3207 This produces a list of all the Fibonacci numbers that will fit in
3208 16 bits. Note that a maximum repeat count must still be given to
3209 \c{%rep}. This is to prevent the possibility of NASM getting into an
3210 infinite loop in the preprocessor, which (on multitasking or
3211 multi-user systems) would typically cause all the system memory to
3212 be gradually used up and other applications to start crashing.
3215 \H{files} Source Files and Dependencies
3217 These commands allow you to split your sources into multiple files.
3219 \S{include} \i\c{%include}: \i{Including Other Files}
3221 Using, once again, a very similar syntax to the C preprocessor,
3222 NASM's preprocessor lets you include other source files into your
3223 code. This is done by the use of the \i\c{%include} directive:
3225 \c %include "macros.mac"
3227 will include the contents of the file \c{macros.mac} into the source
3228 file containing the \c{%include} directive.
3230 Include files are \I{searching for include files}searched for in the
3231 current directory (the directory you're in when you run NASM, as
3232 opposed to the location of the NASM executable or the location of
3233 the source file), plus any directories specified on the NASM command
3234 line using the \c{-i} option.
3236 The standard C idiom for preventing a file being included more than
3237 once is just as applicable in NASM: if the file \c{macros.mac} has
3240 \c %ifndef MACROS_MAC
3241 \c %define MACROS_MAC
3242 \c ; now define some macros
3245 then including the file more than once will not cause errors,
3246 because the second time the file is included nothing will happen
3247 because the macro \c{MACROS_MAC} will already be defined.
3249 You can force a file to be included even if there is no \c{%include}
3250 directive that explicitly includes it, by using the \i\c{-p} option
3251 on the NASM command line (see \k{opt-p}).
3254 \S{pathsearch} \i\c{%pathsearch}: Search the Include Path
3256 The \c{%pathsearch} directive takes a single-line macro name and a
3257 filename, and declare or redefines the specified single-line macro to
3258 be the include-path-resolved version of the filename, if the file
3259 exists (otherwise, it is passed unchanged.)
3263 \c %pathsearch MyFoo "foo.bin"
3265 ... with \c{-Ibins/} in the include path may end up defining the macro
3266 \c{MyFoo} to be \c{"bins/foo.bin"}.
3269 \S{depend} \i\c{%depend}: Add Dependent Files
3271 The \c{%depend} directive takes a filename and adds it to the list of
3272 files to be emitted as dependency generation when the \c{-M} options
3273 and its relatives (see \k{opt-M}) are used. It produces no output.
3275 This is generally used in conjunction with \c{%pathsearch}. For
3276 example, a simplified version of the standard macro wrapper for the
3277 \c{INCBIN} directive looks like:
3279 \c %imacro incbin 1-2+ 0
3280 \c %pathsearch dep %1
3285 This first resolves the location of the file into the macro \c{dep},
3286 then adds it to the dependency lists, and finally issues the
3287 assembler-level \c{INCBIN} directive.
3290 \S{use} \i\c{%use}: Include Standard Macro Package
3292 The \c{%use} directive is similar to \c{%include}, but rather than
3293 including the contents of a file, it includes a named standard macro
3294 package. The standard macro packages are part of NASM, and are
3295 described in \k{macropkg}.
3297 Unlike the \c{%include} directive, package names for the \c{%use}
3298 directive do not require quotes, but quotes are permitted. In NASM
3299 2.04 and 2.05 the unquoted form would be macro-expanded; this is no
3300 longer true. Thus, the following lines are equivalent:
3305 Standard macro packages are protected from multiple inclusion. When a
3306 standard macro package is used, a testable single-line macro of the
3307 form \c{__USE_}\e{package}\c{__} is also defined, see \k{use_def}.
3309 \H{ctxstack} The \i{Context Stack}
3311 Having labels that are local to a macro definition is sometimes not
3312 quite powerful enough: sometimes you want to be able to share labels
3313 between several macro calls. An example might be a \c{REPEAT} ...
3314 \c{UNTIL} loop, in which the expansion of the \c{REPEAT} macro
3315 would need to be able to refer to a label which the \c{UNTIL} macro
3316 had defined. However, for such a macro you would also want to be
3317 able to nest these loops.
3319 NASM provides this level of power by means of a \e{context stack}.
3320 The preprocessor maintains a stack of \e{contexts}, each of which is
3321 characterized by a name. You add a new context to the stack using
3322 the \i\c{%push} directive, and remove one using \i\c{%pop}. You can
3323 define labels that are local to a particular context on the stack.
3326 \S{pushpop} \i\c{%push} and \i\c{%pop}: \I{creating
3327 contexts}\I{removing contexts}Creating and Removing Contexts
3329 The \c{%push} directive is used to create a new context and place it
3330 on the top of the context stack. \c{%push} takes an optional argument,
3331 which is the name of the context. For example:
3335 This pushes a new context called \c{foobar} on the stack. You can have
3336 several contexts on the stack with the same name: they can still be
3337 distinguished. If no name is given, the context is unnamed (this is
3338 normally used when both the \c{%push} and the \c{%pop} are inside a
3339 single macro definition.)
3341 The directive \c{%pop}, taking one optional argument, removes the top
3342 context from the context stack and destroys it, along with any
3343 labels associated with it. If an argument is given, it must match the
3344 name of the current context, otherwise it will issue an error.
3347 \S{ctxlocal} \i{Context-Local Labels}
3349 Just as the usage \c{%%foo} defines a label which is local to the
3350 particular macro call in which it is used, the usage \I{%$}\c{%$foo}
3351 is used to define a label which is local to the context on the top
3352 of the context stack. So the \c{REPEAT} and \c{UNTIL} example given
3353 above could be implemented by means of:
3369 and invoked by means of, for example,
3377 which would scan every fourth byte of a string in search of the byte
3380 If you need to define, or access, labels local to the context
3381 \e{below} the top one on the stack, you can use \I{%$$}\c{%$$foo}, or
3382 \c{%$$$foo} for the context below that, and so on.
3385 \S{ctxdefine} \i{Context-Local Single-Line Macros}
3387 NASM also allows you to define single-line macros which are local to
3388 a particular context, in just the same way:
3390 \c %define %$localmac 3
3392 will define the single-line macro \c{%$localmac} to be local to the
3393 top context on the stack. Of course, after a subsequent \c{%push},
3394 it can then still be accessed by the name \c{%$$localmac}.
3397 \S{ctxrepl} \i\c{%repl}: \I{renaming contexts}Renaming a Context
3399 If you need to change the name of the top context on the stack (in
3400 order, for example, to have it respond differently to \c{%ifctx}),
3401 you can execute a \c{%pop} followed by a \c{%push}; but this will
3402 have the side effect of destroying all context-local labels and
3403 macros associated with the context that was just popped.
3405 NASM provides the directive \c{%repl}, which \e{replaces} a context
3406 with a different name, without touching the associated macros and
3407 labels. So you could replace the destructive code
3412 with the non-destructive version \c{%repl newname}.
3415 \S{blockif} Example Use of the \i{Context Stack}: \i{Block IFs}
3417 This example makes use of almost all the context-stack features,
3418 including the conditional-assembly construct \i\c{%ifctx}, to
3419 implement a block IF statement as a set of macros.
3435 \c %error "expected `if' before `else'"
3449 \c %error "expected `if' or `else' before `endif'"
3454 This code is more robust than the \c{REPEAT} and \c{UNTIL} macros
3455 given in \k{ctxlocal}, because it uses conditional assembly to check
3456 that the macros are issued in the right order (for example, not
3457 calling \c{endif} before \c{if}) and issues a \c{%error} if they're
3460 In addition, the \c{endif} macro has to be able to cope with the two
3461 distinct cases of either directly following an \c{if}, or following
3462 an \c{else}. It achieves this, again, by using conditional assembly
3463 to do different things depending on whether the context on top of
3464 the stack is \c{if} or \c{else}.
3466 The \c{else} macro has to preserve the context on the stack, in
3467 order to have the \c{%$ifnot} referred to by the \c{if} macro be the
3468 same as the one defined by the \c{endif} macro, but has to change
3469 the context's name so that \c{endif} will know there was an
3470 intervening \c{else}. It does this by the use of \c{%repl}.
3472 A sample usage of these macros might look like:
3494 The block-\c{IF} macros handle nesting quite happily, by means of
3495 pushing another context, describing the inner \c{if}, on top of the
3496 one describing the outer \c{if}; thus \c{else} and \c{endif} always
3497 refer to the last unmatched \c{if} or \c{else}.
3500 \H{stackrel} \i{Stack Relative Preprocessor Directives}
3502 The following preprocessor directives provide a way to use
3503 labels to refer to local variables allocated on the stack.
3505 \b\c{%arg} (see \k{arg})
3507 \b\c{%stacksize} (see \k{stacksize})
3509 \b\c{%local} (see \k{local})
3512 \S{arg} \i\c{%arg} Directive
3514 The \c{%arg} directive is used to simplify the handling of
3515 parameters passed on the stack. Stack based parameter passing
3516 is used by many high level languages, including C, C++ and Pascal.
3518 While NASM has macros which attempt to duplicate this
3519 functionality (see \k{16cmacro}), the syntax is not particularly
3520 convenient to use and is not TASM compatible. Here is an example
3521 which shows the use of \c{%arg} without any external macros:
3525 \c %push mycontext ; save the current context
3526 \c %stacksize large ; tell NASM to use bp
3527 \c %arg i:word, j_ptr:word
3534 \c %pop ; restore original context
3536 This is similar to the procedure defined in \k{16cmacro} and adds
3537 the value in i to the value pointed to by j_ptr and returns the
3538 sum in the ax register. See \k{pushpop} for an explanation of
3539 \c{push} and \c{pop} and the use of context stacks.
3542 \S{stacksize} \i\c{%stacksize} Directive
3544 The \c{%stacksize} directive is used in conjunction with the
3545 \c{%arg} (see \k{arg}) and the \c{%local} (see \k{local}) directives.
3546 It tells NASM the default size to use for subsequent \c{%arg} and
3547 \c{%local} directives. The \c{%stacksize} directive takes one
3548 required argument which is one of \c{flat}, \c{flat64}, \c{large} or \c{small}.
3552 This form causes NASM to use stack-based parameter addressing
3553 relative to \c{ebp} and it assumes that a near form of call was used
3554 to get to this label (i.e. that \c{eip} is on the stack).
3556 \c %stacksize flat64
3558 This form causes NASM to use stack-based parameter addressing
3559 relative to \c{rbp} and it assumes that a near form of call was used
3560 to get to this label (i.e. that \c{rip} is on the stack).
3564 This form uses \c{bp} to do stack-based parameter addressing and
3565 assumes that a far form of call was used to get to this address
3566 (i.e. that \c{ip} and \c{cs} are on the stack).
3570 This form also uses \c{bp} to address stack parameters, but it is
3571 different from \c{large} because it also assumes that the old value
3572 of bp is pushed onto the stack (i.e. it expects an \c{ENTER}
3573 instruction). In other words, it expects that \c{bp}, \c{ip} and
3574 \c{cs} are on the top of the stack, underneath any local space which
3575 may have been allocated by \c{ENTER}. This form is probably most
3576 useful when used in combination with the \c{%local} directive
3580 \S{local} \i\c{%local} Directive
3582 The \c{%local} directive is used to simplify the use of local
3583 temporary stack variables allocated in a stack frame. Automatic
3584 local variables in C are an example of this kind of variable. The
3585 \c{%local} directive is most useful when used with the \c{%stacksize}
3586 (see \k{stacksize} and is also compatible with the \c{%arg} directive
3587 (see \k{arg}). It allows simplified reference to variables on the
3588 stack which have been allocated typically by using the \c{ENTER}
3590 \# (see \k{insENTER} for a description of that instruction).
3591 An example of its use is the following:
3595 \c %push mycontext ; save the current context
3596 \c %stacksize small ; tell NASM to use bp
3597 \c %assign %$localsize 0 ; see text for explanation
3598 \c %local old_ax:word, old_dx:word
3600 \c enter %$localsize,0 ; see text for explanation
3601 \c mov [old_ax],ax ; swap ax & bx
3602 \c mov [old_dx],dx ; and swap dx & cx
3607 \c leave ; restore old bp
3610 \c %pop ; restore original context
3612 The \c{%$localsize} variable is used internally by the
3613 \c{%local} directive and \e{must} be defined within the
3614 current context before the \c{%local} directive may be used.
3615 Failure to do so will result in one expression syntax error for
3616 each \c{%local} variable declared. It then may be used in
3617 the construction of an appropriately sized ENTER instruction
3618 as shown in the example.
3621 \H{pperror} Reporting \i{User-Defined Errors}: \i\c{%error}, \i\c{%warning}, \i\c{%fatal}
3623 The preprocessor directive \c{%error} will cause NASM to report an
3624 error if it occurs in assembled code. So if other users are going to
3625 try to assemble your source files, you can ensure that they define the
3626 right macros by means of code like this:
3631 \c ; do some different setup
3633 \c %error "Neither F1 nor F2 was defined."
3636 Then any user who fails to understand the way your code is supposed
3637 to be assembled will be quickly warned of their mistake, rather than
3638 having to wait until the program crashes on being run and then not
3639 knowing what went wrong.
3641 Similarly, \c{%warning} issues a warning, but allows assembly to continue:
3646 \c ; do some different setup
3648 \c %warning "Neither F1 nor F2 was defined, assuming F1."
3652 \c{%error} and \c{%warning} are issued only on the final assembly
3653 pass. This makes them safe to use in conjunction with tests that
3654 depend on symbol values.
3656 \c{%fatal} terminates assembly immediately, regardless of pass. This
3657 is useful when there is no point in continuing the assembly further,
3658 and doing so is likely just going to cause a spew of confusing error
3661 It is optional for the message string after \c{%error}, \c{%warning}
3662 or \c{%fatal} to be quoted. If it is \e{not}, then single-line macros
3663 are expanded in it, which can be used to display more information to
3664 the user. For example:
3667 \c %assign foo_over foo-64
3668 \c %error foo is foo_over bytes too large
3672 \H{otherpreproc} \i{Other Preprocessor Directives}
3674 NASM also has preprocessor directives which allow access to
3675 information from external sources. Currently they include:
3677 \b\c{%line} enables NASM to correctly handle the output of another
3678 preprocessor (see \k{line}).
3680 \b\c{%!} enables NASM to read in the value of an environment variable,
3681 which can then be used in your program (see \k{getenv}).
3683 \S{line} \i\c{%line} Directive
3685 The \c{%line} directive is used to notify NASM that the input line
3686 corresponds to a specific line number in another file. Typically
3687 this other file would be an original source file, with the current
3688 NASM input being the output of a pre-processor. The \c{%line}
3689 directive allows NASM to output messages which indicate the line
3690 number of the original source file, instead of the file that is being
3693 This preprocessor directive is not generally of use to programmers,
3694 by may be of interest to preprocessor authors. The usage of the
3695 \c{%line} preprocessor directive is as follows:
3697 \c %line nnn[+mmm] [filename]
3699 In this directive, \c{nnn} identifies the line of the original source
3700 file which this line corresponds to. \c{mmm} is an optional parameter
3701 which specifies a line increment value; each line of the input file
3702 read in is considered to correspond to \c{mmm} lines of the original
3703 source file. Finally, \c{filename} is an optional parameter which
3704 specifies the file name of the original source file.
3706 After reading a \c{%line} preprocessor directive, NASM will report
3707 all file name and line numbers relative to the values specified
3711 \S{getenv} \i\c{%!}\c{<env>}: Read an environment variable.
3713 The \c{%!<env>} directive makes it possible to read the value of an
3714 environment variable at assembly time. This could, for example, be used
3715 to store the contents of an environment variable into a string, which
3716 could be used at some other point in your code.
3718 For example, suppose that you have an environment variable \c{FOO}, and
3719 you want the contents of \c{FOO} to be embedded in your program. You
3720 could do that as follows:
3722 \c %defstr FOO %!FOO
3724 See \k{defstr} for notes on the \c{%defstr} directive.
3726 If the name of the environment variable contains non-identifier
3727 characters, you can use string quotes to surround the name of the
3728 variable, for example:
3730 \c %defstr C_colon %!'C:'
3733 \H{stdmac} \i{Standard Macros}
3735 NASM defines a set of standard macros, which are already defined
3736 when it starts to process any source file. If you really need a
3737 program to be assembled with no pre-defined macros, you can use the
3738 \i\c{%clear} directive to empty the preprocessor of everything but
3739 context-local preprocessor variables and single-line macros.
3741 Most \i{user-level assembler directives} (see \k{directive}) are
3742 implemented as macros which invoke primitive directives; these are
3743 described in \k{directive}. The rest of the standard macro set is
3747 \S{stdmacver} \i{NASM Version} Macros
3749 The single-line macros \i\c{__NASM_MAJOR__}, \i\c{__NASM_MINOR__},
3750 \i\c{__NASM_SUBMINOR__} and \i\c{___NASM_PATCHLEVEL__} expand to the
3751 major, minor, subminor and patch level parts of the \i{version
3752 number of NASM} being used. So, under NASM 0.98.32p1 for
3753 example, \c{__NASM_MAJOR__} would be defined to be 0, \c{__NASM_MINOR__}
3754 would be defined as 98, \c{__NASM_SUBMINOR__} would be defined to 32,
3755 and \c{___NASM_PATCHLEVEL__} would be defined as 1.
3757 Additionally, the macro \i\c{__NASM_SNAPSHOT__} is defined for
3758 automatically generated snapshot releases \e{only}.
3761 \S{stdmacverid} \i\c{__NASM_VERSION_ID__}: \i{NASM Version ID}
3763 The single-line macro \c{__NASM_VERSION_ID__} expands to a dword integer
3764 representing the full version number of the version of nasm being used.
3765 The value is the equivalent to \c{__NASM_MAJOR__}, \c{__NASM_MINOR__},
3766 \c{__NASM_SUBMINOR__} and \c{___NASM_PATCHLEVEL__} concatenated to
3767 produce a single doubleword. Hence, for 0.98.32p1, the returned number
3768 would be equivalent to:
3776 Note that the above lines are generate exactly the same code, the second
3777 line is used just to give an indication of the order that the separate
3778 values will be present in memory.
3781 \S{stdmacverstr} \i\c{__NASM_VER__}: \i{NASM Version string}
3783 The single-line macro \c{__NASM_VER__} expands to a string which defines
3784 the version number of nasm being used. So, under NASM 0.98.32 for example,
3793 \S{fileline} \i\c{__FILE__} and \i\c{__LINE__}: File Name and Line Number
3795 Like the C preprocessor, NASM allows the user to find out the file
3796 name and line number containing the current instruction. The macro
3797 \c{__FILE__} expands to a string constant giving the name of the
3798 current input file (which may change through the course of assembly
3799 if \c{%include} directives are used), and \c{__LINE__} expands to a
3800 numeric constant giving the current line number in the input file.
3802 These macros could be used, for example, to communicate debugging
3803 information to a macro, since invoking \c{__LINE__} inside a macro
3804 definition (either single-line or multi-line) will return the line
3805 number of the macro \e{call}, rather than \e{definition}. So to
3806 determine where in a piece of code a crash is occurring, for
3807 example, one could write a routine \c{stillhere}, which is passed a
3808 line number in \c{EAX} and outputs something like `line 155: still
3809 here'. You could then write a macro
3811 \c %macro notdeadyet 0
3820 and then pepper your code with calls to \c{notdeadyet} until you
3821 find the crash point.
3824 \S{bitsm} \i\c{__BITS__}: Current BITS Mode
3826 The \c{__BITS__} standard macro is updated every time that the BITS mode is
3827 set using the \c{BITS XX} or \c{[BITS XX]} directive, where XX is a valid mode
3828 number of 16, 32 or 64. \c{__BITS__} receives the specified mode number and
3829 makes it globally available. This can be very useful for those who utilize
3830 mode-dependent macros.
3832 \S{ofmtm} \i\c{__OUTPUT_FORMAT__}: Current Output Format
3834 The \c{__OUTPUT_FORMAT__} standard macro holds the current Output Format,
3835 as given by the \c{-f} option or NASM's default. Type \c{nasm -hf} for a
3838 \c %ifidn __OUTPUT_FORMAT__, win32
3839 \c %define NEWLINE 13, 10
3840 \c %elifidn __OUTPUT_FORMAT__, elf32
3841 \c %define NEWLINE 10
3845 \S{datetime} Assembly Date and Time Macros
3847 NASM provides a variety of macros that represent the timestamp of the
3850 \b The \i\c{__DATE__} and \i\c{__TIME__} macros give the assembly date and
3851 time as strings, in ISO 8601 format (\c{"YYYY-MM-DD"} and \c{"HH:MM:SS"},
3854 \b The \i\c{__DATE_NUM__} and \i\c{__TIME_NUM__} macros give the assembly
3855 date and time in numeric form; in the format \c{YYYYMMDD} and
3856 \c{HHMMSS} respectively.
3858 \b The \i\c{__UTC_DATE__} and \i\c{__UTC_TIME__} macros give the assembly
3859 date and time in universal time (UTC) as strings, in ISO 8601 format
3860 (\c{"YYYY-MM-DD"} and \c{"HH:MM:SS"}, respectively.) If the host
3861 platform doesn't provide UTC time, these macros are undefined.
3863 \b The \i\c{__UTC_DATE_NUM__} and \i\c{__UTC_TIME_NUM__} macros give the
3864 assembly date and time universal time (UTC) in numeric form; in the
3865 format \c{YYYYMMDD} and \c{HHMMSS} respectively. If the
3866 host platform doesn't provide UTC time, these macros are
3869 \b The \c{__POSIX_TIME__} macro is defined as a number containing the
3870 number of seconds since the POSIX epoch, 1 January 1970 00:00:00 UTC;
3871 excluding any leap seconds. This is computed using UTC time if
3872 available on the host platform, otherwise it is computed using the
3873 local time as if it was UTC.
3875 All instances of time and date macros in the same assembly session
3876 produce consistent output. For example, in an assembly session
3877 started at 42 seconds after midnight on January 1, 2010 in Moscow
3878 (timezone UTC+3) these macros would have the following values,
3879 assuming, of course, a properly configured environment with a correct
3882 \c __DATE__ "2010-01-01"
3883 \c __TIME__ "00:00:42"
3884 \c __DATE_NUM__ 20100101
3885 \c __TIME_NUM__ 000042
3886 \c __UTC_DATE__ "2009-12-31"
3887 \c __UTC_TIME__ "21:00:42"
3888 \c __UTC_DATE_NUM__ 20091231
3889 \c __UTC_TIME_NUM__ 210042
3890 \c __POSIX_TIME__ 1262293242
3893 \S{use_def} \I\c{__USE_*__}\c{__USE_}\e{package}\c{__}: Package
3896 When a standard macro package (see \k{macropkg}) is included with the
3897 \c{%use} directive (see \k{use}), a single-line macro of the form
3898 \c{__USE_}\e{package}\c{__} is automatically defined. This allows
3899 testing if a particular package is invoked or not.
3901 For example, if the \c{altreg} package is included (see
3902 \k{pkg_altreg}), then the macro \c{__USE_ALTREG__} is defined.
3905 \S{pass_macro} \i\c{__PASS__}: Assembly Pass
3907 The macro \c{__PASS__} is defined to be \c{1} on preparatory passes,
3908 and \c{2} on the final pass. In preprocess-only mode, it is set to
3909 \c{3}, and when running only to generate dependencies (due to the
3910 \c{-M} or \c{-MG} option, see \k{opt-M}) it is set to \c{0}.
3912 \e{Avoid using this macro if at all possible. It is tremendously easy
3913 to generate very strange errors by misusing it, and the semantics may
3914 change in future versions of NASM.}
3917 \S{struc} \i\c{STRUC} and \i\c{ENDSTRUC}: \i{Declaring Structure} Data Types
3919 The core of NASM contains no intrinsic means of defining data
3920 structures; instead, the preprocessor is sufficiently powerful that
3921 data structures can be implemented as a set of macros. The macros
3922 \c{STRUC} and \c{ENDSTRUC} are used to define a structure data type.
3924 \c{STRUC} takes one or two parameters. The first parameter is the name
3925 of the data type. The second, optional parameter is the base offset of
3926 the structure. The name of the data type is defined as a symbol with
3927 the value of the base offset, and the name of the data type with the
3928 suffix \c{_size} appended to it is defined as an \c{EQU} giving the
3929 size of the structure. Once \c{STRUC} has been issued, you are
3930 defining the structure, and should define fields using the \c{RESB}
3931 family of pseudo-instructions, and then invoke \c{ENDSTRUC} to finish
3934 For example, to define a structure called \c{mytype} containing a
3935 longword, a word, a byte and a string of bytes, you might code
3946 The above code defines six symbols: \c{mt_long} as 0 (the offset
3947 from the beginning of a \c{mytype} structure to the longword field),
3948 \c{mt_word} as 4, \c{mt_byte} as 6, \c{mt_str} as 7, \c{mytype_size}
3949 as 39, and \c{mytype} itself as zero.
3951 The reason why the structure type name is defined at zero by default
3952 is a side effect of allowing structures to work with the local label
3953 mechanism: if your structure members tend to have the same names in
3954 more than one structure, you can define the above structure like this:
3965 This defines the offsets to the structure fields as \c{mytype.long},
3966 \c{mytype.word}, \c{mytype.byte} and \c{mytype.str}.
3968 NASM, since it has no \e{intrinsic} structure support, does not
3969 support any form of period notation to refer to the elements of a
3970 structure once you have one (except the above local-label notation),
3971 so code such as \c{mov ax,[mystruc.mt_word]} is not valid.
3972 \c{mt_word} is a constant just like any other constant, so the
3973 correct syntax is \c{mov ax,[mystruc+mt_word]} or \c{mov
3974 ax,[mystruc+mytype.word]}.
3976 Sometimes you only have the address of the structure displaced by an
3977 offset. For example, consider this standard stack frame setup:
3983 In this case, you could access an element by subtracting the offset:
3985 \c mov [ebp - 40 + mytype.word], ax
3987 However, if you do not want to repeat this offset, you can use -40 as
3990 \c struc mytype, -40
3992 And access an element this way:
3994 \c mov [ebp + mytype.word], ax
3997 \S{istruc} \i\c{ISTRUC}, \i\c{AT} and \i\c{IEND}: Declaring
3998 \i{Instances of Structures}
4000 Having defined a structure type, the next thing you typically want
4001 to do is to declare instances of that structure in your data
4002 segment. NASM provides an easy way to do this in the \c{ISTRUC}
4003 mechanism. To declare a structure of type \c{mytype} in a program,
4004 you code something like this:
4009 \c at mt_long, dd 123456
4010 \c at mt_word, dw 1024
4011 \c at mt_byte, db 'x'
4012 \c at mt_str, db 'hello, world', 13, 10, 0
4016 The function of the \c{AT} macro is to make use of the \c{TIMES}
4017 prefix to advance the assembly position to the correct point for the
4018 specified structure field, and then to declare the specified data.
4019 Therefore the structure fields must be declared in the same order as
4020 they were specified in the structure definition.
4022 If the data to go in a structure field requires more than one source
4023 line to specify, the remaining source lines can easily come after
4024 the \c{AT} line. For example:
4026 \c at mt_str, db 123,134,145,156,167,178,189
4029 Depending on personal taste, you can also omit the code part of the
4030 \c{AT} line completely, and start the structure field on the next
4034 \c db 'hello, world'
4038 \S{align} \i\c{ALIGN} and \i\c{ALIGNB}: Data Alignment
4040 The \c{ALIGN} and \c{ALIGNB} macros provides a convenient way to
4041 align code or data on a word, longword, paragraph or other boundary.
4042 (Some assemblers call this directive \i\c{EVEN}.) The syntax of the
4043 \c{ALIGN} and \c{ALIGNB} macros is
4045 \c align 4 ; align on 4-byte boundary
4046 \c align 16 ; align on 16-byte boundary
4047 \c align 8,db 0 ; pad with 0s rather than NOPs
4048 \c align 4,resb 1 ; align to 4 in the BSS
4049 \c alignb 4 ; equivalent to previous line
4051 Both macros require their first argument to be a power of two; they
4052 both compute the number of additional bytes required to bring the
4053 length of the current section up to a multiple of that power of two,
4054 and then apply the \c{TIMES} prefix to their second argument to
4055 perform the alignment.
4057 If the second argument is not specified, the default for \c{ALIGN}
4058 is \c{NOP}, and the default for \c{ALIGNB} is \c{RESB 1}. So if the
4059 second argument is specified, the two macros are equivalent.
4060 Normally, you can just use \c{ALIGN} in code and data sections and
4061 \c{ALIGNB} in BSS sections, and never need the second argument
4062 except for special purposes.
4064 \c{ALIGN} and \c{ALIGNB}, being simple macros, perform no error
4065 checking: they cannot warn you if their first argument fails to be a
4066 power of two, or if their second argument generates more than one
4067 byte of code. In each of these cases they will silently do the wrong
4070 \c{ALIGNB} (or \c{ALIGN} with a second argument of \c{RESB 1}) can
4071 be used within structure definitions:
4088 This will ensure that the structure members are sensibly aligned
4089 relative to the base of the structure.
4091 A final caveat: \c{ALIGN} and \c{ALIGNB} work relative to the
4092 beginning of the \e{section}, not the beginning of the address space
4093 in the final executable. Aligning to a 16-byte boundary when the
4094 section you're in is only guaranteed to be aligned to a 4-byte
4095 boundary, for example, is a waste of effort. Again, NASM does not
4096 check that the section's alignment characteristics are sensible for
4097 the use of \c{ALIGN} or \c{ALIGNB}.
4099 See also the \c{smartalign} standard macro package, \k{pkg_smartalign}.
4102 \C{macropkg} \i{Standard Macro Packages}
4104 The \i\c{%use} directive (see \k{use}) includes one of the standard
4105 macro packages included with the NASM distribution and compiled into
4106 the NASM binary. It operates like the \c{%include} directive (see
4107 \k{include}), but the included contents is provided by NASM itself.
4109 The names of standard macro packages are case insensitive, and can be
4113 \H{pkg_altreg} \i\c{altreg}: \i{Alternate Register Names}
4115 The \c{altreg} standard macro package provides alternate register
4116 names. It provides numeric register names for all registers (not just
4117 \c{R8}-\c{R15}), the Intel-defined aliases \c{R8L}-\c{R15L} for the
4118 low bytes of register (as opposed to the NASM/AMD standard names
4119 \c{R8B}-\c{R15B}), and the names \c{R0H}-\c{R3H} (by analogy with
4120 \c{R0L}-\c{R3L}) for \c{AH}, \c{CH}, \c{DH}, and \c{BH}.
4127 \c mov r0l,r3h ; mov al,bh
4133 \H{pkg_smartalign} \i\c{smartalign}\I{align, smart}: Smart \c{ALIGN} Macro
4135 The \c{smartalign} standard macro package provides for an \i\c{ALIGN}
4136 macro which is more powerful than the default (and
4137 backwards-compatible) one (see \k{align}). When the \c{smartalign}
4138 package is enabled, when \c{ALIGN} is used without a second argument,
4139 NASM will generate a sequence of instructions more efficient than a
4140 series of \c{NOP}. Furthermore, if the padding exceeds a specific
4141 threshold, then NASM will generate a jump over the entire padding
4144 The specific instructions generated can be controlled with the
4145 new \i\c{ALIGNMODE} macro. This macro takes two parameters: one mode,
4146 and an optional jump threshold override. If (for any reason) you need
4147 to turn off the jump completely just set jump threshold value to -1
4148 (or set it to \c{nojmp}). The following modes are possible:
4150 \b \c{generic}: Works on all x86 CPUs and should have reasonable
4151 performance. The default jump threshold is 8. This is the
4154 \b \c{nop}: Pad out with \c{NOP} instructions. The only difference
4155 compared to the standard \c{ALIGN} macro is that NASM can still jump
4156 over a large padding area. The default jump threshold is 16.
4158 \b \c{k7}: Optimize for the AMD K7 (Athlon/Althon XP). These
4159 instructions should still work on all x86 CPUs. The default jump
4162 \b \c{k8}: Optimize for the AMD K8 (Opteron/Althon 64). These
4163 instructions should still work on all x86 CPUs. The default jump
4166 \b \c{p6}: Optimize for Intel CPUs. This uses the long \c{NOP}
4167 instructions first introduced in Pentium Pro. This is incompatible
4168 with all CPUs of family 5 or lower, as well as some VIA CPUs and
4169 several virtualization solutions. The default jump threshold is 16.
4171 The macro \i\c{__ALIGNMODE__} is defined to contain the current
4172 alignment mode. A number of other macros beginning with \c{__ALIGN_}
4173 are used internally by this macro package.
4176 \C{directive} \i{Assembler Directives}
4178 NASM, though it attempts to avoid the bureaucracy of assemblers like
4179 MASM and TASM, is nevertheless forced to support a \e{few}
4180 directives. These are described in this chapter.
4182 NASM's directives come in two types: \I{user-level
4183 directives}\e{user-level} directives and \I{primitive
4184 directives}\e{primitive} directives. Typically, each directive has a
4185 user-level form and a primitive form. In almost all cases, we
4186 recommend that users use the user-level forms of the directives,
4187 which are implemented as macros which call the primitive forms.
4189 Primitive directives are enclosed in square brackets; user-level
4192 In addition to the universal directives described in this chapter,
4193 each object file format can optionally supply extra directives in
4194 order to control particular features of that file format. These
4195 \I{format-specific directives}\e{format-specific} directives are
4196 documented along with the formats that implement them, in \k{outfmt}.
4199 \H{bits} \i\c{BITS}: Specifying Target \i{Processor Mode}
4201 The \c{BITS} directive specifies whether NASM should generate code
4202 \I{16-bit mode, versus 32-bit mode}designed to run on a processor
4203 operating in 16-bit mode, 32-bit mode or 64-bit mode. The syntax is
4204 \c{BITS XX}, where XX is 16, 32 or 64.
4206 In most cases, you should not need to use \c{BITS} explicitly. The
4207 \c{aout}, \c{coff}, \c{elf}, \c{macho}, \c{win32} and \c{win64}
4208 object formats, which are designed for use in 32-bit or 64-bit
4209 operating systems, all cause NASM to select 32-bit or 64-bit mode,
4210 respectively, by default. The \c{obj} object format allows you
4211 to specify each segment you define as either \c{USE16} or \c{USE32},
4212 and NASM will set its operating mode accordingly, so the use of the
4213 \c{BITS} directive is once again unnecessary.
4215 The most likely reason for using the \c{BITS} directive is to write
4216 32-bit or 64-bit code in a flat binary file; this is because the \c{bin}
4217 output format defaults to 16-bit mode in anticipation of it being
4218 used most frequently to write DOS \c{.COM} programs, DOS \c{.SYS}
4219 device drivers and boot loader software.
4221 You do \e{not} need to specify \c{BITS 32} merely in order to use
4222 32-bit instructions in a 16-bit DOS program; if you do, the
4223 assembler will generate incorrect code because it will be writing
4224 code targeted at a 32-bit platform, to be run on a 16-bit one.
4226 When NASM is in \c{BITS 16} mode, instructions which use 32-bit
4227 data are prefixed with an 0x66 byte, and those referring to 32-bit
4228 addresses have an 0x67 prefix. In \c{BITS 32} mode, the reverse is
4229 true: 32-bit instructions require no prefixes, whereas instructions
4230 using 16-bit data need an 0x66 and those working on 16-bit addresses
4233 When NASM is in \c{BITS 64} mode, most instructions operate the same
4234 as they do for \c{BITS 32} mode. However, there are 8 more general and
4235 SSE registers, and 16-bit addressing is no longer supported.
4237 The default address size is 64 bits; 32-bit addressing can be selected
4238 with the 0x67 prefix. The default operand size is still 32 bits,
4239 however, and the 0x66 prefix selects 16-bit operand size. The \c{REX}
4240 prefix is used both to select 64-bit operand size, and to access the
4241 new registers. NASM automatically inserts REX prefixes when
4244 When the \c{REX} prefix is used, the processor does not know how to
4245 address the AH, BH, CH or DH (high 8-bit legacy) registers. Instead,
4246 it is possible to access the the low 8-bits of the SP, BP SI and DI
4247 registers as SPL, BPL, SIL and DIL, respectively; but only when the
4250 The \c{BITS} directive has an exactly equivalent primitive form,
4251 \c{[BITS 16]}, \c{[BITS 32]} and \c{[BITS 64]}. The user-level form is
4252 a macro which has no function other than to call the primitive form.
4254 Note that the space is neccessary, e.g. \c{BITS32} will \e{not} work!
4256 \S{USE16 & USE32} \i\c{USE16} & \i\c{USE32}: Aliases for BITS
4258 The `\c{USE16}' and `\c{USE32}' directives can be used in place of
4259 `\c{BITS 16}' and `\c{BITS 32}', for compatibility with other assemblers.
4262 \H{default} \i\c{DEFAULT}: Change the assembler defaults
4264 The \c{DEFAULT} directive changes the assembler defaults. Normally,
4265 NASM defaults to a mode where the programmer is expected to explicitly
4266 specify most features directly. However, this is occationally
4267 obnoxious, as the explicit form is pretty much the only one one wishes
4270 Currently, the only \c{DEFAULT} that is settable is whether or not
4271 registerless instructions in 64-bit mode are \c{RIP}-relative or not.
4272 By default, they are absolute unless overridden with the \i\c{REL}
4273 specifier (see \k{effaddr}). However, if \c{DEFAULT REL} is
4274 specified, \c{REL} is default, unless overridden with the \c{ABS}
4275 specifier, \e{except when used with an FS or GS segment override}.
4277 The special handling of \c{FS} and \c{GS} overrides are due to the
4278 fact that these registers are generally used as thread pointers or
4279 other special functions in 64-bit mode, and generating
4280 \c{RIP}-relative addresses would be extremely confusing.
4282 \c{DEFAULT REL} is disabled with \c{DEFAULT ABS}.
4284 \H{section} \i\c{SECTION} or \i\c{SEGMENT}: Changing and \i{Defining
4287 \I{changing sections}\I{switching between sections}The \c{SECTION}
4288 directive (\c{SEGMENT} is an exactly equivalent synonym) changes
4289 which section of the output file the code you write will be
4290 assembled into. In some object file formats, the number and names of
4291 sections are fixed; in others, the user may make up as many as they
4292 wish. Hence \c{SECTION} may sometimes give an error message, or may
4293 define a new section, if you try to switch to a section that does
4296 The Unix object formats, and the \c{bin} object format (but see
4297 \k{multisec}, all support
4298 the \i{standardized section names} \c{.text}, \c{.data} and \c{.bss}
4299 for the code, data and uninitialized-data sections. The \c{obj}
4300 format, by contrast, does not recognize these section names as being
4301 special, and indeed will strip off the leading period of any section
4305 \S{sectmac} The \i\c{__SECT__} Macro
4307 The \c{SECTION} directive is unusual in that its user-level form
4308 functions differently from its primitive form. The primitive form,
4309 \c{[SECTION xyz]}, simply switches the current target section to the
4310 one given. The user-level form, \c{SECTION xyz}, however, first
4311 defines the single-line macro \c{__SECT__} to be the primitive
4312 \c{[SECTION]} directive which it is about to issue, and then issues
4313 it. So the user-level directive
4317 expands to the two lines
4319 \c %define __SECT__ [SECTION .text]
4322 Users may find it useful to make use of this in their own macros.
4323 For example, the \c{writefile} macro defined in \k{mlmacgre} can be
4324 usefully rewritten in the following more sophisticated form:
4326 \c %macro writefile 2+
4336 \c mov cx,%%endstr-%%str
4343 This form of the macro, once passed a string to output, first
4344 switches temporarily to the data section of the file, using the
4345 primitive form of the \c{SECTION} directive so as not to modify
4346 \c{__SECT__}. It then declares its string in the data section, and
4347 then invokes \c{__SECT__} to switch back to \e{whichever} section
4348 the user was previously working in. It thus avoids the need, in the
4349 previous version of the macro, to include a \c{JMP} instruction to
4350 jump over the data, and also does not fail if, in a complicated
4351 \c{OBJ} format module, the user could potentially be assembling the
4352 code in any of several separate code sections.
4355 \H{absolute} \i\c{ABSOLUTE}: Defining Absolute Labels
4357 The \c{ABSOLUTE} directive can be thought of as an alternative form
4358 of \c{SECTION}: it causes the subsequent code to be directed at no
4359 physical section, but at the hypothetical section starting at the
4360 given absolute address. The only instructions you can use in this
4361 mode are the \c{RESB} family.
4363 \c{ABSOLUTE} is used as follows:
4371 This example describes a section of the PC BIOS data area, at
4372 segment address 0x40: the above code defines \c{kbuf_chr} to be
4373 0x1A, \c{kbuf_free} to be 0x1C, and \c{kbuf} to be 0x1E.
4375 The user-level form of \c{ABSOLUTE}, like that of \c{SECTION},
4376 redefines the \i\c{__SECT__} macro when it is invoked.
4378 \i\c{STRUC} and \i\c{ENDSTRUC} are defined as macros which use
4379 \c{ABSOLUTE} (and also \c{__SECT__}).
4381 \c{ABSOLUTE} doesn't have to take an absolute constant as an
4382 argument: it can take an expression (actually, a \i{critical
4383 expression}: see \k{crit}) and it can be a value in a segment. For
4384 example, a TSR can re-use its setup code as run-time BSS like this:
4386 \c org 100h ; it's a .COM program
4388 \c jmp setup ; setup code comes last
4390 \c ; the resident part of the TSR goes here
4392 \c ; now write the code that installs the TSR here
4396 \c runtimevar1 resw 1
4397 \c runtimevar2 resd 20
4401 This defines some variables `on top of' the setup code, so that
4402 after the setup has finished running, the space it took up can be
4403 re-used as data storage for the running TSR. The symbol `tsr_end'
4404 can be used to calculate the total size of the part of the TSR that
4405 needs to be made resident.
4408 \H{extern} \i\c{EXTERN}: \i{Importing Symbols} from Other Modules
4410 \c{EXTERN} is similar to the MASM directive \c{EXTRN} and the C
4411 keyword \c{extern}: it is used to declare a symbol which is not
4412 defined anywhere in the module being assembled, but is assumed to be
4413 defined in some other module and needs to be referred to by this
4414 one. Not every object-file format can support external variables:
4415 the \c{bin} format cannot.
4417 The \c{EXTERN} directive takes as many arguments as you like. Each
4418 argument is the name of a symbol:
4421 \c extern _sscanf,_fscanf
4423 Some object-file formats provide extra features to the \c{EXTERN}
4424 directive. In all cases, the extra features are used by suffixing a
4425 colon to the symbol name followed by object-format specific text.
4426 For example, the \c{obj} format allows you to declare that the
4427 default segment base of an external should be the group \c{dgroup}
4428 by means of the directive
4430 \c extern _variable:wrt dgroup
4432 The primitive form of \c{EXTERN} differs from the user-level form
4433 only in that it can take only one argument at a time: the support
4434 for multiple arguments is implemented at the preprocessor level.
4436 You can declare the same variable as \c{EXTERN} more than once: NASM
4437 will quietly ignore the second and later redeclarations. You can't
4438 declare a variable as \c{EXTERN} as well as something else, though.
4441 \H{global} \i\c{GLOBAL}: \i{Exporting Symbols} to Other Modules
4443 \c{GLOBAL} is the other end of \c{EXTERN}: if one module declares a
4444 symbol as \c{EXTERN} and refers to it, then in order to prevent
4445 linker errors, some other module must actually \e{define} the
4446 symbol and declare it as \c{GLOBAL}. Some assemblers use the name
4447 \i\c{PUBLIC} for this purpose.
4449 The \c{GLOBAL} directive applying to a symbol must appear \e{before}
4450 the definition of the symbol.
4452 \c{GLOBAL} uses the same syntax as \c{EXTERN}, except that it must
4453 refer to symbols which \e{are} defined in the same module as the
4454 \c{GLOBAL} directive. For example:
4460 \c{GLOBAL}, like \c{EXTERN}, allows object formats to define private
4461 extensions by means of a colon. The \c{elf} object format, for
4462 example, lets you specify whether global data items are functions or
4465 \c global hashlookup:function, hashtable:data
4467 Like \c{EXTERN}, the primitive form of \c{GLOBAL} differs from the
4468 user-level form only in that it can take only one argument at a
4472 \H{common} \i\c{COMMON}: Defining Common Data Areas
4474 The \c{COMMON} directive is used to declare \i\e{common variables}.
4475 A common variable is much like a global variable declared in the
4476 uninitialized data section, so that
4480 is similar in function to
4487 The difference is that if more than one module defines the same
4488 common variable, then at link time those variables will be
4489 \e{merged}, and references to \c{intvar} in all modules will point
4490 at the same piece of memory.
4492 Like \c{GLOBAL} and \c{EXTERN}, \c{COMMON} supports object-format
4493 specific extensions. For example, the \c{obj} format allows common
4494 variables to be NEAR or FAR, and the \c{elf} format allows you to
4495 specify the alignment requirements of a common variable:
4497 \c common commvar 4:near ; works in OBJ
4498 \c common intarray 100:4 ; works in ELF: 4 byte aligned
4500 Once again, like \c{EXTERN} and \c{GLOBAL}, the primitive form of
4501 \c{COMMON} differs from the user-level form only in that it can take
4502 only one argument at a time.
4505 \H{CPU} \i\c{CPU}: Defining CPU Dependencies
4507 The \i\c{CPU} directive restricts assembly to those instructions which
4508 are available on the specified CPU.
4512 \b\c{CPU 8086} Assemble only 8086 instruction set
4514 \b\c{CPU 186} Assemble instructions up to the 80186 instruction set
4516 \b\c{CPU 286} Assemble instructions up to the 286 instruction set
4518 \b\c{CPU 386} Assemble instructions up to the 386 instruction set
4520 \b\c{CPU 486} 486 instruction set
4522 \b\c{CPU 586} Pentium instruction set
4524 \b\c{CPU PENTIUM} Same as 586
4526 \b\c{CPU 686} P6 instruction set
4528 \b\c{CPU PPRO} Same as 686
4530 \b\c{CPU P2} Same as 686
4532 \b\c{CPU P3} Pentium III (Katmai) instruction sets
4534 \b\c{CPU KATMAI} Same as P3
4536 \b\c{CPU P4} Pentium 4 (Willamette) instruction set
4538 \b\c{CPU WILLAMETTE} Same as P4
4540 \b\c{CPU PRESCOTT} Prescott instruction set
4542 \b\c{CPU X64} x86-64 (x64/AMD64/Intel 64) instruction set
4544 \b\c{CPU IA64} IA64 CPU (in x86 mode) instruction set
4546 All options are case insensitive. All instructions will be selected
4547 only if they apply to the selected CPU or lower. By default, all
4548 instructions are available.
4551 \H{FLOAT} \i\c{FLOAT}: Handling of \I{floating-point, constants}floating-point constants
4553 By default, floating-point constants are rounded to nearest, and IEEE
4554 denormals are supported. The following options can be set to alter
4557 \b\c{FLOAT DAZ} Flush denormals to zero
4559 \b\c{FLOAT NODAZ} Do not flush denormals to zero (default)
4561 \b\c{FLOAT NEAR} Round to nearest (default)
4563 \b\c{FLOAT UP} Round up (toward +Infinity)
4565 \b\c{FLOAT DOWN} Round down (toward -Infinity)
4567 \b\c{FLOAT ZERO} Round toward zero
4569 \b\c{FLOAT DEFAULT} Restore default settings
4571 The standard macros \i\c{__FLOAT_DAZ__}, \i\c{__FLOAT_ROUND__}, and
4572 \i\c{__FLOAT__} contain the current state, as long as the programmer
4573 has avoided the use of the brackeded primitive form, (\c{[FLOAT]}).
4575 \c{__FLOAT__} contains the full set of floating-point settings; this
4576 value can be saved away and invoked later to restore the setting.
4579 \C{outfmt} \i{Output Formats}
4581 NASM is a portable assembler, designed to be able to compile on any
4582 ANSI C-supporting platform and produce output to run on a variety of
4583 Intel x86 operating systems. For this reason, it has a large number
4584 of available output formats, selected using the \i\c{-f} option on
4585 the NASM \i{command line}. Each of these formats, along with its
4586 extensions to the base NASM syntax, is detailed in this chapter.
4588 As stated in \k{opt-o}, NASM chooses a \i{default name} for your
4589 output file based on the input file name and the chosen output
4590 format. This will be generated by removing the \i{extension}
4591 (\c{.asm}, \c{.s}, or whatever you like to use) from the input file
4592 name, and substituting an extension defined by the output format.
4593 The extensions are given with each format below.
4596 \H{binfmt} \i\c{bin}: \i{Flat-Form Binary}\I{pure binary} Output
4598 The \c{bin} format does not produce object files: it generates
4599 nothing in the output file except the code you wrote. Such `pure
4600 binary' files are used by \i{MS-DOS}: \i\c{.COM} executables and
4601 \i\c{.SYS} device drivers are pure binary files. Pure binary output
4602 is also useful for \i{operating system} and \i{boot loader}
4605 The \c{bin} format supports \i{multiple section names}. For details of
4606 how NASM handles sections in the \c{bin} format, see \k{multisec}.
4608 Using the \c{bin} format puts NASM by default into 16-bit mode (see
4609 \k{bits}). In order to use \c{bin} to write 32-bit or 64-bit code,
4610 such as an OS kernel, you need to explicitly issue the \I\c{BITS}\c{BITS 32}
4611 or \I\c{BITS}\c{BITS 64} directive.
4613 \c{bin} has no default output file name extension: instead, it
4614 leaves your file name as it is once the original extension has been
4615 removed. Thus, the default is for NASM to assemble \c{binprog.asm}
4616 into a binary file called \c{binprog}.
4619 \S{org} \i\c{ORG}: Binary File \i{Program Origin}
4621 The \c{bin} format provides an additional directive to the list
4622 given in \k{directive}: \c{ORG}. The function of the \c{ORG}
4623 directive is to specify the origin address which NASM will assume
4624 the program begins at when it is loaded into memory.
4626 For example, the following code will generate the longword
4633 Unlike the \c{ORG} directive provided by MASM-compatible assemblers,
4634 which allows you to jump around in the object file and overwrite
4635 code you have already generated, NASM's \c{ORG} does exactly what
4636 the directive says: \e{origin}. Its sole function is to specify one
4637 offset which is added to all internal address references within the
4638 section; it does not permit any of the trickery that MASM's version
4639 does. See \k{proborg} for further comments.
4642 \S{binseg} \c{bin} Extensions to the \c{SECTION}
4643 Directive\I{SECTION, bin extensions to}
4645 The \c{bin} output format extends the \c{SECTION} (or \c{SEGMENT})
4646 directive to allow you to specify the alignment requirements of
4647 segments. This is done by appending the \i\c{ALIGN} qualifier to the
4648 end of the section-definition line. For example,
4650 \c section .data align=16
4652 switches to the section \c{.data} and also specifies that it must be
4653 aligned on a 16-byte boundary.
4655 The parameter to \c{ALIGN} specifies how many low bits of the
4656 section start address must be forced to zero. The alignment value
4657 given may be any power of two.\I{section alignment, in
4658 bin}\I{segment alignment, in bin}\I{alignment, in bin sections}
4661 \S{multisec} \i{Multisection}\I{bin, multisection} Support for the \c{bin} Format
4663 The \c{bin} format allows the use of multiple sections, of arbitrary names,
4664 besides the "known" \c{.text}, \c{.data}, and \c{.bss} names.
4666 \b Sections may be designated \i\c{progbits} or \i\c{nobits}. Default
4667 is \c{progbits} (except \c{.bss}, which defaults to \c{nobits},
4670 \b Sections can be aligned at a specified boundary following the previous
4671 section with \c{align=}, or at an arbitrary byte-granular position with
4674 \b Sections can be given a virtual start address, which will be used
4675 for the calculation of all memory references within that section
4678 \b Sections can be ordered using \i\c{follows=}\c{<section>} or
4679 \i\c{vfollows=}\c{<section>} as an alternative to specifying an explicit
4682 \b Arguments to \c{org}, \c{start}, \c{vstart}, and \c{align=} are
4683 critical expressions. See \k{crit}. E.g. \c{align=(1 << ALIGN_SHIFT)}
4684 - \c{ALIGN_SHIFT} must be defined before it is used here.
4686 \b Any code which comes before an explicit \c{SECTION} directive
4687 is directed by default into the \c{.text} section.
4689 \b If an \c{ORG} statement is not given, \c{ORG 0} is used
4692 \b The \c{.bss} section will be placed after the last \c{progbits}
4693 section, unless \c{start=}, \c{vstart=}, \c{follows=}, or \c{vfollows=}
4696 \b All sections are aligned on dword boundaries, unless a different
4697 alignment has been specified.
4699 \b Sections may not overlap.
4701 \b NASM creates the \c{section.<secname>.start} for each section,
4702 which may be used in your code.
4704 \S{map}\i{Map Files}
4706 Map files can be generated in \c{-f bin} format by means of the \c{[map]}
4707 option. Map types of \c{all} (default), \c{brief}, \c{sections}, \c{segments},
4708 or \c{symbols} may be specified. Output may be directed to \c{stdout}
4709 (default), \c{stderr}, or a specified file. E.g.
4710 \c{[map symbols myfile.map]}. No "user form" exists, the square
4711 brackets must be used.
4714 \H{ithfmt} \i\c{ith}: \i{Intel Hex} Output
4716 The \c{ith} file format produces Intel hex-format files. Just as the
4717 \c{bin} format, this is a flat memory image format with no support for
4718 relocation or linking. It is usually used with ROM programmers and
4721 All extensions supported by the \c{bin} file format is also supported by
4722 the \c{ith} file format.
4724 \c{ith} provides a default output file-name extension of \c{.ith}.
4727 \H{srecfmt} \i\c{srec}: \i{Motorola S-Records} Output
4729 The \c{srec} file format produces Motorola S-records files. Just as the
4730 \c{bin} format, this is a flat memory image format with no support for
4731 relocation or linking. It is usually used with ROM programmers and
4734 All extensions supported by the \c{bin} file format is also supported by
4735 the \c{srec} file format.
4737 \c{srec} provides a default output file-name extension of \c{.srec}.
4740 \H{objfmt} \i\c{obj}: \i{Microsoft OMF}\I{OMF} Object Files
4742 The \c{obj} file format (NASM calls it \c{obj} rather than \c{omf}
4743 for historical reasons) is the one produced by \i{MASM} and
4744 \i{TASM}, which is typically fed to 16-bit DOS linkers to produce
4745 \i\c{.EXE} files. It is also the format used by \i{OS/2}.
4747 \c{obj} provides a default output file-name extension of \c{.obj}.
4749 \c{obj} is not exclusively a 16-bit format, though: NASM has full
4750 support for the 32-bit extensions to the format. In particular,
4751 32-bit \c{obj} format files are used by \i{Borland's Win32
4752 compilers}, instead of using Microsoft's newer \i\c{win32} object
4755 The \c{obj} format does not define any special segment names: you
4756 can call your segments anything you like. Typical names for segments
4757 in \c{obj} format files are \c{CODE}, \c{DATA} and \c{BSS}.
4759 If your source file contains code before specifying an explicit
4760 \c{SEGMENT} directive, then NASM will invent its own segment called
4761 \i\c{__NASMDEFSEG} for you.
4763 When you define a segment in an \c{obj} file, NASM defines the
4764 segment name as a symbol as well, so that you can access the segment
4765 address of the segment. So, for example:
4774 \c mov ax,data ; get segment address of data
4775 \c mov ds,ax ; and move it into DS
4776 \c inc word [dvar] ; now this reference will work
4779 The \c{obj} format also enables the use of the \i\c{SEG} and
4780 \i\c{WRT} operators, so that you can write code which does things
4785 \c mov ax,seg foo ; get preferred segment of foo
4787 \c mov ax,data ; a different segment
4789 \c mov ax,[ds:foo] ; this accesses `foo'
4790 \c mov [es:foo wrt data],bx ; so does this
4793 \S{objseg} \c{obj} Extensions to the \c{SEGMENT}
4794 Directive\I{SEGMENT, obj extensions to}
4796 The \c{obj} output format extends the \c{SEGMENT} (or \c{SECTION})
4797 directive to allow you to specify various properties of the segment
4798 you are defining. This is done by appending extra qualifiers to the
4799 end of the segment-definition line. For example,
4801 \c segment code private align=16
4803 defines the segment \c{code}, but also declares it to be a private
4804 segment, and requires that the portion of it described in this code
4805 module must be aligned on a 16-byte boundary.
4807 The available qualifiers are:
4809 \b \i\c{PRIVATE}, \i\c{PUBLIC}, \i\c{COMMON} and \i\c{STACK} specify
4810 the combination characteristics of the segment. \c{PRIVATE} segments
4811 do not get combined with any others by the linker; \c{PUBLIC} and
4812 \c{STACK} segments get concatenated together at link time; and
4813 \c{COMMON} segments all get overlaid on top of each other rather
4814 than stuck end-to-end.
4816 \b \i\c{ALIGN} is used, as shown above, to specify how many low bits
4817 of the segment start address must be forced to zero. The alignment
4818 value given may be any power of two from 1 to 4096; in reality, the
4819 only values supported are 1, 2, 4, 16, 256 and 4096, so if 8 is
4820 specified it will be rounded up to 16, and 32, 64 and 128 will all
4821 be rounded up to 256, and so on. Note that alignment to 4096-byte
4822 boundaries is a \i{PharLap} extension to the format and may not be
4823 supported by all linkers.\I{section alignment, in OBJ}\I{segment
4824 alignment, in OBJ}\I{alignment, in OBJ sections}
4826 \b \i\c{CLASS} can be used to specify the segment class; this feature
4827 indicates to the linker that segments of the same class should be
4828 placed near each other in the output file. The class name can be any
4829 word, e.g. \c{CLASS=CODE}.
4831 \b \i\c{OVERLAY}, like \c{CLASS}, is specified with an arbitrary word
4832 as an argument, and provides overlay information to an
4833 overlay-capable linker.
4835 \b Segments can be declared as \i\c{USE16} or \i\c{USE32}, which has
4836 the effect of recording the choice in the object file and also
4837 ensuring that NASM's default assembly mode when assembling in that
4838 segment is 16-bit or 32-bit respectively.
4840 \b When writing \i{OS/2} object files, you should declare 32-bit
4841 segments as \i\c{FLAT}, which causes the default segment base for
4842 anything in the segment to be the special group \c{FLAT}, and also
4843 defines the group if it is not already defined.
4845 \b The \c{obj} file format also allows segments to be declared as
4846 having a pre-defined absolute segment address, although no linkers
4847 are currently known to make sensible use of this feature;
4848 nevertheless, NASM allows you to declare a segment such as
4849 \c{SEGMENT SCREEN ABSOLUTE=0xB800} if you need to. The \i\c{ABSOLUTE}
4850 and \c{ALIGN} keywords are mutually exclusive.
4852 NASM's default segment attributes are \c{PUBLIC}, \c{ALIGN=1}, no
4853 class, no overlay, and \c{USE16}.
4856 \S{group} \i\c{GROUP}: Defining Groups of Segments\I{segments, groups of}
4858 The \c{obj} format also allows segments to be grouped, so that a
4859 single segment register can be used to refer to all the segments in
4860 a group. NASM therefore supplies the \c{GROUP} directive, whereby
4869 \c ; some uninitialized data
4871 \c group dgroup data bss
4873 which will define a group called \c{dgroup} to contain the segments
4874 \c{data} and \c{bss}. Like \c{SEGMENT}, \c{GROUP} causes the group
4875 name to be defined as a symbol, so that you can refer to a variable
4876 \c{var} in the \c{data} segment as \c{var wrt data} or as \c{var wrt
4877 dgroup}, depending on which segment value is currently in your
4880 If you just refer to \c{var}, however, and \c{var} is declared in a
4881 segment which is part of a group, then NASM will default to giving
4882 you the offset of \c{var} from the beginning of the \e{group}, not
4883 the \e{segment}. Therefore \c{SEG var}, also, will return the group
4884 base rather than the segment base.
4886 NASM will allow a segment to be part of more than one group, but
4887 will generate a warning if you do this. Variables declared in a
4888 segment which is part of more than one group will default to being
4889 relative to the first group that was defined to contain the segment.
4891 A group does not have to contain any segments; you can still make
4892 \c{WRT} references to a group which does not contain the variable
4893 you are referring to. OS/2, for example, defines the special group
4894 \c{FLAT} with no segments in it.
4897 \S{uppercase} \i\c{UPPERCASE}: Disabling Case Sensitivity in Output
4899 Although NASM itself is \i{case sensitive}, some OMF linkers are
4900 not; therefore it can be useful for NASM to output single-case
4901 object files. The \c{UPPERCASE} format-specific directive causes all
4902 segment, group and symbol names that are written to the object file
4903 to be forced to upper case just before being written. Within a
4904 source file, NASM is still case-sensitive; but the object file can
4905 be written entirely in upper case if desired.
4907 \c{UPPERCASE} is used alone on a line; it requires no parameters.
4910 \S{import} \i\c{IMPORT}: Importing DLL Symbols\I{DLL symbols,
4911 importing}\I{symbols, importing from DLLs}
4913 The \c{IMPORT} format-specific directive defines a symbol to be
4914 imported from a DLL, for use if you are writing a DLL's \i{import
4915 library} in NASM. You still need to declare the symbol as \c{EXTERN}
4916 as well as using the \c{IMPORT} directive.
4918 The \c{IMPORT} directive takes two required parameters, separated by
4919 white space, which are (respectively) the name of the symbol you
4920 wish to import and the name of the library you wish to import it
4923 \c import WSAStartup wsock32.dll
4925 A third optional parameter gives the name by which the symbol is
4926 known in the library you are importing it from, in case this is not
4927 the same as the name you wish the symbol to be known by to your code
4928 once you have imported it. For example:
4930 \c import asyncsel wsock32.dll WSAAsyncSelect
4933 \S{export} \i\c{EXPORT}: Exporting DLL Symbols\I{DLL symbols,
4934 exporting}\I{symbols, exporting from DLLs}
4936 The \c{EXPORT} format-specific directive defines a global symbol to
4937 be exported as a DLL symbol, for use if you are writing a DLL in
4938 NASM. You still need to declare the symbol as \c{GLOBAL} as well as
4939 using the \c{EXPORT} directive.
4941 \c{EXPORT} takes one required parameter, which is the name of the
4942 symbol you wish to export, as it was defined in your source file. An
4943 optional second parameter (separated by white space from the first)
4944 gives the \e{external} name of the symbol: the name by which you
4945 wish the symbol to be known to programs using the DLL. If this name
4946 is the same as the internal name, you may leave the second parameter
4949 Further parameters can be given to define attributes of the exported
4950 symbol. These parameters, like the second, are separated by white
4951 space. If further parameters are given, the external name must also
4952 be specified, even if it is the same as the internal name. The
4953 available attributes are:
4955 \b \c{resident} indicates that the exported name is to be kept
4956 resident by the system loader. This is an optimization for
4957 frequently used symbols imported by name.
4959 \b \c{nodata} indicates that the exported symbol is a function which
4960 does not make use of any initialized data.
4962 \b \c{parm=NNN}, where \c{NNN} is an integer, sets the number of
4963 parameter words for the case in which the symbol is a call gate
4964 between 32-bit and 16-bit segments.
4966 \b An attribute which is just a number indicates that the symbol
4967 should be exported with an identifying number (ordinal), and gives
4973 \c export myfunc TheRealMoreFormalLookingFunctionName
4974 \c export myfunc myfunc 1234 ; export by ordinal
4975 \c export myfunc myfunc resident parm=23 nodata
4978 \S{dotdotstart} \i\c{..start}: Defining the \i{Program Entry
4981 \c{OMF} linkers require exactly one of the object files being linked to
4982 define the program entry point, where execution will begin when the
4983 program is run. If the object file that defines the entry point is
4984 assembled using NASM, you specify the entry point by declaring the
4985 special symbol \c{..start} at the point where you wish execution to
4989 \S{objextern} \c{obj} Extensions to the \c{EXTERN}
4990 Directive\I{EXTERN, obj extensions to}
4992 If you declare an external symbol with the directive
4996 then references such as \c{mov ax,foo} will give you the offset of
4997 \c{foo} from its preferred segment base (as specified in whichever
4998 module \c{foo} is actually defined in). So to access the contents of
4999 \c{foo} you will usually need to do something like
5001 \c mov ax,seg foo ; get preferred segment base
5002 \c mov es,ax ; move it into ES
5003 \c mov ax,[es:foo] ; and use offset `foo' from it
5005 This is a little unwieldy, particularly if you know that an external
5006 is going to be accessible from a given segment or group, say
5007 \c{dgroup}. So if \c{DS} already contained \c{dgroup}, you could
5010 \c mov ax,[foo wrt dgroup]
5012 However, having to type this every time you want to access \c{foo}
5013 can be a pain; so NASM allows you to declare \c{foo} in the
5016 \c extern foo:wrt dgroup
5018 This form causes NASM to pretend that the preferred segment base of
5019 \c{foo} is in fact \c{dgroup}; so the expression \c{seg foo} will
5020 now return \c{dgroup}, and the expression \c{foo} is equivalent to
5023 This \I{default-WRT mechanism}default-\c{WRT} mechanism can be used
5024 to make externals appear to be relative to any group or segment in
5025 your program. It can also be applied to common variables: see
5029 \S{objcommon} \c{obj} Extensions to the \c{COMMON}
5030 Directive\I{COMMON, obj extensions to}
5032 The \c{obj} format allows common variables to be either near\I{near
5033 common variables} or far\I{far common variables}; NASM allows you to
5034 specify which your variables should be by the use of the syntax
5036 \c common nearvar 2:near ; `nearvar' is a near common
5037 \c common farvar 10:far ; and `farvar' is far
5039 Far common variables may be greater in size than 64Kb, and so the
5040 OMF specification says that they are declared as a number of
5041 \e{elements} of a given size. So a 10-byte far common variable could
5042 be declared as ten one-byte elements, five two-byte elements, two
5043 five-byte elements or one ten-byte element.
5045 Some \c{OMF} linkers require the \I{element size, in common
5046 variables}\I{common variables, element size}element size, as well as
5047 the variable size, to match when resolving common variables declared
5048 in more than one module. Therefore NASM must allow you to specify
5049 the element size on your far common variables. This is done by the
5052 \c common c_5by2 10:far 5 ; two five-byte elements
5053 \c common c_2by5 10:far 2 ; five two-byte elements
5055 If no element size is specified, the default is 1. Also, the \c{FAR}
5056 keyword is not required when an element size is specified, since
5057 only far commons may have element sizes at all. So the above
5058 declarations could equivalently be
5060 \c common c_5by2 10:5 ; two five-byte elements
5061 \c common c_2by5 10:2 ; five two-byte elements
5063 In addition to these extensions, the \c{COMMON} directive in \c{obj}
5064 also supports default-\c{WRT} specification like \c{EXTERN} does
5065 (explained in \k{objextern}). So you can also declare things like
5067 \c common foo 10:wrt dgroup
5068 \c common bar 16:far 2:wrt data
5069 \c common baz 24:wrt data:6
5072 \H{win32fmt} \i\c{win32}: Microsoft Win32 Object Files
5074 The \c{win32} output format generates Microsoft Win32 object files,
5075 suitable for passing to Microsoft linkers such as \i{Visual C++}.
5076 Note that Borland Win32 compilers do not use this format, but use
5077 \c{obj} instead (see \k{objfmt}).
5079 \c{win32} provides a default output file-name extension of \c{.obj}.
5081 Note that although Microsoft say that Win32 object files follow the
5082 \c{COFF} (Common Object File Format) standard, the object files produced
5083 by Microsoft Win32 compilers are not compatible with COFF linkers
5084 such as DJGPP's, and vice versa. This is due to a difference of
5085 opinion over the precise semantics of PC-relative relocations. To
5086 produce COFF files suitable for DJGPP, use NASM's \c{coff} output
5087 format; conversely, the \c{coff} format does not produce object
5088 files that Win32 linkers can generate correct output from.
5091 \S{win32sect} \c{win32} Extensions to the \c{SECTION}
5092 Directive\I{SECTION, win32 extensions to}
5094 Like the \c{obj} format, \c{win32} allows you to specify additional
5095 information on the \c{SECTION} directive line, to control the type
5096 and properties of sections you declare. Section types and properties
5097 are generated automatically by NASM for the \i{standard section names}
5098 \c{.text}, \c{.data} and \c{.bss}, but may still be overridden by
5101 The available qualifiers are:
5103 \b \c{code}, or equivalently \c{text}, defines the section to be a
5104 code section. This marks the section as readable and executable, but
5105 not writable, and also indicates to the linker that the type of the
5108 \b \c{data} and \c{bss} define the section to be a data section,
5109 analogously to \c{code}. Data sections are marked as readable and
5110 writable, but not executable. \c{data} declares an initialized data
5111 section, whereas \c{bss} declares an uninitialized data section.
5113 \b \c{rdata} declares an initialized data section that is readable
5114 but not writable. Microsoft compilers use this section to place
5117 \b \c{info} defines the section to be an \i{informational section},
5118 which is not included in the executable file by the linker, but may
5119 (for example) pass information \e{to} the linker. For example,
5120 declaring an \c{info}-type section called \i\c{.drectve} causes the
5121 linker to interpret the contents of the section as command-line
5124 \b \c{align=}, used with a trailing number as in \c{obj}, gives the
5125 \I{section alignment, in win32}\I{alignment, in win32
5126 sections}alignment requirements of the section. The maximum you may
5127 specify is 64: the Win32 object file format contains no means to
5128 request a greater section alignment than this. If alignment is not
5129 explicitly specified, the defaults are 16-byte alignment for code
5130 sections, 8-byte alignment for rdata sections and 4-byte alignment
5131 for data (and BSS) sections.
5132 Informational sections get a default alignment of 1 byte (no
5133 alignment), though the value does not matter.
5135 The defaults assumed by NASM if you do not specify the above
5138 \c section .text code align=16
5139 \c section .data data align=4
5140 \c section .rdata rdata align=8
5141 \c section .bss bss align=4
5143 Any other section name is treated by default like \c{.text}.
5145 \S{win32safeseh} \c{win32}: Safe Structured Exception Handling
5147 Among other improvements in Windows XP SP2 and Windows Server 2003
5148 Microsoft has introduced concept of "safe structured exception
5149 handling." General idea is to collect handlers' entry points in
5150 designated read-only table and have alleged entry point verified
5151 against this table prior exception control is passed to the handler. In
5152 order for an executable module to be equipped with such "safe exception
5153 handler table," all object modules on linker command line has to comply
5154 with certain criteria. If one single module among them does not, then
5155 the table in question is omitted and above mentioned run-time checks
5156 will not be performed for application in question. Table omission is by
5157 default silent and therefore can be easily overlooked. One can instruct
5158 linker to refuse to produce binary without such table by passing
5159 \c{/safeseh} command line option.
5161 Without regard to this run-time check merits it's natural to expect
5162 NASM to be capable of generating modules suitable for \c{/safeseh}
5163 linking. From developer's viewpoint the problem is two-fold:
5165 \b how to adapt modules not deploying exception handlers of their own;
5167 \b how to adapt/develop modules utilizing custom exception handling;
5169 Former can be easily achieved with any NASM version by adding following
5170 line to source code:
5174 As of version 2.03 NASM adds this absolute symbol automatically. If
5175 it's not already present to be precise. I.e. if for whatever reason
5176 developer would choose to assign another value in source file, it would
5177 still be perfectly possible.
5179 Registering custom exception handler on the other hand requires certain
5180 "magic." As of version 2.03 additional directive is implemented,
5181 \c{safeseh}, which instructs the assembler to produce appropriately
5182 formatted input data for above mentioned "safe exception handler
5183 table." Its typical use would be:
5186 \c extern _MessageBoxA@16
5187 \c %if __NASM_VERSION_ID__ >= 0x02030000
5188 \c safeseh handler ; register handler as "safe handler"
5191 \c push DWORD 1 ; MB_OKCANCEL
5192 \c push DWORD caption
5195 \c call _MessageBoxA@16
5196 \c sub eax,1 ; incidentally suits as return value
5197 \c ; for exception handler
5201 \c push DWORD handler
5202 \c push DWORD [fs:0]
5203 \c mov DWORD [fs:0],esp ; engage exception handler
5205 \c mov eax,DWORD[eax] ; cause exception
5206 \c pop DWORD [fs:0] ; disengage exception handler
5209 \c text: db 'OK to rethrow, CANCEL to generate core dump',0
5210 \c caption:db 'SEGV',0
5212 \c section .drectve info
5213 \c db '/defaultlib:user32.lib /defaultlib:msvcrt.lib '
5215 As you might imagine, it's perfectly possible to produce .exe binary
5216 with "safe exception handler table" and yet engage unregistered
5217 exception handler. Indeed, handler is engaged by simply manipulating
5218 \c{[fs:0]} location at run-time, something linker has no power over,
5219 run-time that is. It should be explicitly mentioned that such failure
5220 to register handler's entry point with \c{safeseh} directive has
5221 undesired side effect at run-time. If exception is raised and
5222 unregistered handler is to be executed, the application is abruptly
5223 terminated without any notification whatsoever. One can argue that
5224 system could at least have logged some kind "non-safe exception
5225 handler in x.exe at address n" message in event log, but no, literally
5226 no notification is provided and user is left with no clue on what
5227 caused application failure.
5229 Finally, all mentions of linker in this paragraph refer to Microsoft
5230 linker version 7.x and later. Presence of \c{@feat.00} symbol and input
5231 data for "safe exception handler table" causes no backward
5232 incompatibilities and "safeseh" modules generated by NASM 2.03 and
5233 later can still be linked by earlier versions or non-Microsoft linkers.
5236 \H{win64fmt} \i\c{win64}: Microsoft Win64 Object Files
5238 The \c{win64} output format generates Microsoft Win64 object files,
5239 which is nearly 100% identical to the \c{win32} object format (\k{win32fmt})
5240 with the exception that it is meant to target 64-bit code and the x86-64
5241 platform altogether. This object file is used exactly the same as the \c{win32}
5242 object format (\k{win32fmt}), in NASM, with regard to this exception.
5244 \S{win64pic} \c{win64}: Writing Position-Independent Code
5246 While \c{REL} takes good care of RIP-relative addressing, there is one
5247 aspect that is easy to overlook for a Win64 programmer: indirect
5248 references. Consider a switch dispatch table:
5250 \c jmp QWORD[dsptch+rax*8]
5256 Even novice Win64 assembler programmer will soon realize that the code
5257 is not 64-bit savvy. Most notably linker will refuse to link it with
5258 "\c{'ADDR32' relocation to '.text' invalid without
5259 /LARGEADDRESSAWARE:NO}". So [s]he will have to split jmp instruction as
5262 \c lea rbx,[rel dsptch]
5263 \c jmp QWORD[rbx+rax*8]
5265 What happens behind the scene is that effective address in \c{lea} is
5266 encoded relative to instruction pointer, or in perfectly
5267 position-independent manner. But this is only part of the problem!
5268 Trouble is that in .dll context \c{caseN} relocations will make their
5269 way to the final module and might have to be adjusted at .dll load
5270 time. To be specific when it can't be loaded at preferred address. And
5271 when this occurs, pages with such relocations will be rendered private
5272 to current process, which kind of undermines the idea of sharing .dll.
5273 But no worry, it's trivial to fix:
5275 \c lea rbx,[rel dsptch]
5276 \c add rbx,QWORD[rbx+rax*8]
5279 \c dsptch: dq case0-dsptch
5283 NASM version 2.03 and later provides another alternative, \c{wrt
5284 ..imagebase} operator, which returns offset from base address of the
5285 current image, be it .exe or .dll module, therefore the name. For those
5286 acquainted with PE-COFF format base address denotes start of
5287 \c{IMAGE_DOS_HEADER} structure. Here is how to implement switch with
5288 these image-relative references:
5290 \c lea rbx,[rel dsptch]
5291 \c mov eax,DWORD[rbx+rax*4]
5292 \c sub rbx,dsptch wrt ..imagebase
5296 \c dsptch: dd case0 wrt ..imagebase
5297 \c dd case1 wrt ..imagebase
5299 One can argue that the operator is redundant. Indeed, snippet before
5300 last works just fine with any NASM version and is not even Windows
5301 specific... The real reason for implementing \c{wrt ..imagebase} will
5302 become apparent in next paragraph.
5304 It should be noted that \c{wrt ..imagebase} is defined as 32-bit
5307 \c dd label wrt ..imagebase ; ok
5308 \c dq label wrt ..imagebase ; bad
5309 \c mov eax,label wrt ..imagebase ; ok
5310 \c mov rax,label wrt ..imagebase ; bad
5312 \S{win64seh} \c{win64}: Structured Exception Handling
5314 Structured exception handing in Win64 is completely different matter
5315 from Win32. Upon exception program counter value is noted, and
5316 linker-generated table comprising start and end addresses of all the
5317 functions [in given executable module] is traversed and compared to the
5318 saved program counter. Thus so called \c{UNWIND_INFO} structure is
5319 identified. If it's not found, then offending subroutine is assumed to
5320 be "leaf" and just mentioned lookup procedure is attempted for its
5321 caller. In Win64 leaf function is such function that does not call any
5322 other function \e{nor} modifies any Win64 non-volatile registers,
5323 including stack pointer. The latter ensures that it's possible to
5324 identify leaf function's caller by simply pulling the value from the
5327 While majority of subroutines written in assembler are not calling any
5328 other function, requirement for non-volatile registers' immutability
5329 leaves developer with not more than 7 registers and no stack frame,
5330 which is not necessarily what [s]he counted with. Customarily one would
5331 meet the requirement by saving non-volatile registers on stack and
5332 restoring them upon return, so what can go wrong? If [and only if] an
5333 exception is raised at run-time and no \c{UNWIND_INFO} structure is
5334 associated with such "leaf" function, the stack unwind procedure will
5335 expect to find caller's return address on the top of stack immediately
5336 followed by its frame. Given that developer pushed caller's
5337 non-volatile registers on stack, would the value on top point at some
5338 code segment or even addressable space? Well, developer can attempt
5339 copying caller's return address to the top of stack and this would
5340 actually work in some very specific circumstances. But unless developer
5341 can guarantee that these circumstances are always met, it's more
5342 appropriate to assume worst case scenario, i.e. stack unwind procedure
5343 going berserk. Relevant question is what happens then? Application is
5344 abruptly terminated without any notification whatsoever. Just like in
5345 Win32 case, one can argue that system could at least have logged
5346 "unwind procedure went berserk in x.exe at address n" in event log, but
5347 no, no trace of failure is left.
5349 Now, when we understand significance of the \c{UNWIND_INFO} structure,
5350 let's discuss what's in it and/or how it's processed. First of all it
5351 is checked for presence of reference to custom language-specific
5352 exception handler. If there is one, then it's invoked. Depending on the
5353 return value, execution flow is resumed (exception is said to be
5354 "handled"), \e{or} rest of \c{UNWIND_INFO} structure is processed as
5355 following. Beside optional reference to custom handler, it carries
5356 information about current callee's stack frame and where non-volatile
5357 registers are saved. Information is detailed enough to be able to
5358 reconstruct contents of caller's non-volatile registers upon call to
5359 current callee. And so caller's context is reconstructed, and then
5360 unwind procedure is repeated, i.e. another \c{UNWIND_INFO} structure is
5361 associated, this time, with caller's instruction pointer, which is then
5362 checked for presence of reference to language-specific handler, etc.
5363 The procedure is recursively repeated till exception is handled. As
5364 last resort system "handles" it by generating memory core dump and
5365 terminating the application.
5367 As for the moment of this writing NASM unfortunately does not
5368 facilitate generation of above mentioned detailed information about
5369 stack frame layout. But as of version 2.03 it implements building
5370 blocks for generating structures involved in stack unwinding. As
5371 simplest example, here is how to deploy custom exception handler for
5376 \c extern MessageBoxA
5382 \c mov r9,1 ; MB_OKCANCEL
5384 \c sub eax,1 ; incidentally suits as return value
5385 \c ; for exception handler
5391 \c mov rax,QWORD[rax] ; cause exception
5394 \c text: db 'OK to rethrow, CANCEL to generate core dump',0
5395 \c caption:db 'SEGV',0
5397 \c section .pdata rdata align=4
5398 \c dd main wrt ..imagebase
5399 \c dd main_end wrt ..imagebase
5400 \c dd xmain wrt ..imagebase
5401 \c section .xdata rdata align=8
5402 \c xmain: db 9,0,0,0
5403 \c dd handler wrt ..imagebase
5404 \c section .drectve info
5405 \c db '/defaultlib:user32.lib /defaultlib:msvcrt.lib '
5407 What you see in \c{.pdata} section is element of the "table comprising
5408 start and end addresses of function" along with reference to associated
5409 \c{UNWIND_INFO} structure. And what you see in \c{.xdata} section is
5410 \c{UNWIND_INFO} structure describing function with no frame, but with
5411 designated exception handler. References are \e{required} to be
5412 image-relative (which is the real reason for implementing \c{wrt
5413 ..imagebase} operator). It should be noted that \c{rdata align=n}, as
5414 well as \c{wrt ..imagebase}, are optional in these two segments'
5415 contexts, i.e. can be omitted. Latter means that \e{all} 32-bit
5416 references, not only above listed required ones, placed into these two
5417 segments turn out image-relative. Why is it important to understand?
5418 Developer is allowed to append handler-specific data to \c{UNWIND_INFO}
5419 structure, and if [s]he adds a 32-bit reference, then [s]he will have
5420 to remember to adjust its value to obtain the real pointer.
5422 As already mentioned, in Win64 terms leaf function is one that does not
5423 call any other function \e{nor} modifies any non-volatile register,
5424 including stack pointer. But it's not uncommon that assembler
5425 programmer plans to utilize every single register and sometimes even
5426 have variable stack frame. Is there anything one can do with bare
5427 building blocks? I.e. besides manually composing fully-fledged
5428 \c{UNWIND_INFO} structure, which would surely be considered
5429 error-prone? Yes, there is. Recall that exception handler is called
5430 first, before stack layout is analyzed. As it turned out, it's
5431 perfectly possible to manipulate current callee's context in custom
5432 handler in manner that permits further stack unwinding. General idea is
5433 that handler would not actually "handle" the exception, but instead
5434 restore callee's context, as it was at its entry point and thus mimic
5435 leaf function. In other words, handler would simply undertake part of
5436 unwinding procedure. Consider following example:
5439 \c mov rax,rsp ; copy rsp to volatile register
5440 \c push r15 ; save non-volatile registers
5443 \c mov r11,rsp ; prepare variable stack frame
5446 \c mov QWORD[r11],rax ; check for exceptions
5447 \c mov rsp,r11 ; allocate stack frame
5448 \c mov QWORD[rsp],rax ; save original rsp value
5451 \c mov r11,QWORD[rsp] ; pull original rsp value
5452 \c mov rbp,QWORD[r11-24]
5453 \c mov rbx,QWORD[r11-16]
5454 \c mov r15,QWORD[r11-8]
5455 \c mov rsp,r11 ; destroy frame
5458 The keyword is that up to \c{magic_point} original \c{rsp} value
5459 remains in chosen volatile register and no non-volatile register,
5460 except for \c{rsp}, is modified. While past \c{magic_point} \c{rsp}
5461 remains constant till the very end of the \c{function}. In this case
5462 custom language-specific exception handler would look like this:
5464 \c EXCEPTION_DISPOSITION handler (EXCEPTION_RECORD *rec,ULONG64 frame,
5465 \c CONTEXT *context,DISPATCHER_CONTEXT *disp)
5467 \c if (context->Rip<(ULONG64)magic_point)
5468 \c rsp = (ULONG64 *)context->Rax;
5470 \c { rsp = ((ULONG64 **)context->Rsp)[0];
5471 \c context->Rbp = rsp[-3];
5472 \c context->Rbx = rsp[-2];
5473 \c context->R15 = rsp[-1];
5475 \c context->Rsp = (ULONG64)rsp;
5477 \c memcpy (disp->ContextRecord,context,sizeof(CONTEXT));
5478 \c RtlVirtualUnwind(UNW_FLAG_NHANDLER,disp->ImageBase,
5479 \c dips->ControlPc,disp->FunctionEntry,disp->ContextRecord,
5480 \c &disp->HandlerData,&disp->EstablisherFrame,NULL);
5481 \c return ExceptionContinueSearch;
5484 As custom handler mimics leaf function, corresponding \c{UNWIND_INFO}
5485 structure does not have to contain any information about stack frame
5488 \H{cofffmt} \i\c{coff}: \i{Common Object File Format}
5490 The \c{coff} output type produces \c{COFF} object files suitable for
5491 linking with the \i{DJGPP} linker.
5493 \c{coff} provides a default output file-name extension of \c{.o}.
5495 The \c{coff} format supports the same extensions to the \c{SECTION}
5496 directive as \c{win32} does, except that the \c{align} qualifier and
5497 the \c{info} section type are not supported.
5499 \H{machofmt} \I{Mach-O}\i\c{macho32} and \i\c{macho64}: \i{Mach Object File Format}
5501 The \c{macho32} and \c{macho64} output formts produces \c{Mach-O}
5502 object files suitable for linking with the \i{MacOS X} linker.
5503 \i\c{macho} is a synonym for \c{macho32}.
5505 \c{macho} provides a default output file-name extension of \c{.o}.
5507 \H{elffmt} \i\c{elf32} and \i\c{elf64}: \I{ELF}\I{linux, elf}\i{Executable and Linkable
5508 Format} Object Files
5510 The \c{elf32} and \c{elf64} output formats generate \c{ELF32 and ELF64} (Executable and Linkable Format) object files, as used by Linux as well as \i{Unix System V},
5511 including \i{Solaris x86}, \i{UnixWare} and \i{SCO Unix}. \c{elf}
5512 provides a default output file-name extension of \c{.o}.
5513 \c{elf} is a synonym for \c{elf32}.
5515 \S{abisect} ELF specific directive \i\c{osabi}
5517 The ELF header specifies the application binary interface for the target operating system (OSABI).
5518 This field can be set by using the \c{osabi} directive with the numeric value (0-255) of the target
5519 system. If this directive is not used, the default value will be "UNIX System V ABI" (0) which will work on
5520 most systems which support ELF.
5522 \S{elfsect} \c{elf} Extensions to the \c{SECTION}
5523 Directive\I{SECTION, elf extensions to}
5525 Like the \c{obj} format, \c{elf} allows you to specify additional
5526 information on the \c{SECTION} directive line, to control the type
5527 and properties of sections you declare. Section types and properties
5528 are generated automatically by NASM for the \i{standard section
5529 names}, but may still be
5530 overridden by these qualifiers.
5532 The available qualifiers are:
5534 \b \i\c{alloc} defines the section to be one which is loaded into
5535 memory when the program is run. \i\c{noalloc} defines it to be one
5536 which is not, such as an informational or comment section.
5538 \b \i\c{exec} defines the section to be one which should have execute
5539 permission when the program is run. \i\c{noexec} defines it as one
5542 \b \i\c{write} defines the section to be one which should be writable
5543 when the program is run. \i\c{nowrite} defines it as one which should
5546 \b \i\c{progbits} defines the section to be one with explicit contents
5547 stored in the object file: an ordinary code or data section, for
5548 example, \i\c{nobits} defines the section to be one with no explicit
5549 contents given, such as a BSS section.
5551 \b \c{align=}, used with a trailing number as in \c{obj}, gives the
5552 \I{section alignment, in elf}\I{alignment, in elf sections}alignment
5553 requirements of the section.
5555 \b \i\c{tls} defines the section to be one which contains
5556 thread local variables.
5558 The defaults assumed by NASM if you do not specify the above
5561 \I\c{.text} \I\c{.rodata} \I\c{.lrodata} \I\c{.data} \I\c{.ldata}
5562 \I\c{.bss} \I\c{.lbss} \I\c{.tdata} \I\c{.tbss} \I\c\{.comment}
5564 \c section .text progbits alloc exec nowrite align=16
5565 \c section .rodata progbits alloc noexec nowrite align=4
5566 \c section .lrodata progbits alloc noexec nowrite align=4
5567 \c section .data progbits alloc noexec write align=4
5568 \c section .ldata progbits alloc noexec write align=4
5569 \c section .bss nobits alloc noexec write align=4
5570 \c section .lbss nobits alloc noexec write align=4
5571 \c section .tdata progbits alloc noexec write align=4 tls
5572 \c section .tbss nobits alloc noexec write align=4 tls
5573 \c section .comment progbits noalloc noexec nowrite align=1
5574 \c section other progbits alloc noexec nowrite align=1
5576 (Any section name other than those in the above table
5577 is treated by default like \c{other} in the above table.
5578 Please note that section names are case sensitive.)
5581 \S{elfwrt} \i{Position-Independent Code}\I{PIC}: \c{elf} Special
5582 Symbols and \i\c{WRT}
5584 The \c{ELF} specification contains enough features to allow
5585 position-independent code (PIC) to be written, which makes \i{ELF
5586 shared libraries} very flexible. However, it also means NASM has to
5587 be able to generate a variety of ELF specific relocation types in ELF
5588 object files, if it is to be an assembler which can write PIC.
5590 Since \c{ELF} does not support segment-base references, the \c{WRT}
5591 operator is not used for its normal purpose; therefore NASM's
5592 \c{elf} output format makes use of \c{WRT} for a different purpose,
5593 namely the PIC-specific \I{relocations, PIC-specific}relocation
5596 \c{elf} defines five special symbols which you can use as the
5597 right-hand side of the \c{WRT} operator to obtain PIC relocation
5598 types. They are \i\c{..gotpc}, \i\c{..gotoff}, \i\c{..got},
5599 \i\c{..plt} and \i\c{..sym}. Their functions are summarized here:
5601 \b Referring to the symbol marking the global offset table base
5602 using \c{wrt ..gotpc} will end up giving the distance from the
5603 beginning of the current section to the global offset table.
5604 (\i\c{_GLOBAL_OFFSET_TABLE_} is the standard symbol name used to
5605 refer to the \i{GOT}.) So you would then need to add \i\c{$$} to the
5606 result to get the real address of the GOT.
5608 \b Referring to a location in one of your own sections using \c{wrt
5609 ..gotoff} will give the distance from the beginning of the GOT to
5610 the specified location, so that adding on the address of the GOT
5611 would give the real address of the location you wanted.
5613 \b Referring to an external or global symbol using \c{wrt ..got}
5614 causes the linker to build an entry \e{in} the GOT containing the
5615 address of the symbol, and the reference gives the distance from the
5616 beginning of the GOT to the entry; so you can add on the address of
5617 the GOT, load from the resulting address, and end up with the
5618 address of the symbol.
5620 \b Referring to a procedure name using \c{wrt ..plt} causes the
5621 linker to build a \i{procedure linkage table} entry for the symbol,
5622 and the reference gives the address of the \i{PLT} entry. You can
5623 only use this in contexts which would generate a PC-relative
5624 relocation normally (i.e. as the destination for \c{CALL} or
5625 \c{JMP}), since ELF contains no relocation type to refer to PLT
5628 \b Referring to a symbol name using \c{wrt ..sym} causes NASM to
5629 write an ordinary relocation, but instead of making the relocation
5630 relative to the start of the section and then adding on the offset
5631 to the symbol, it will write a relocation record aimed directly at
5632 the symbol in question. The distinction is a necessary one due to a
5633 peculiarity of the dynamic linker.
5635 A fuller explanation of how to use these relocation types to write
5636 shared libraries entirely in NASM is given in \k{picdll}.
5638 \S{elftls} \i{Thread Local Storage}\I{TLS}: \c{elf} Special
5639 Symbols and \i\c{WRT}
5641 \b In ELF32 mode, referring to an external or global symbol using
5642 \c{wrt ..tlsie} \I\c{..tlsie}
5643 causes the linker to build an entry \e{in} the GOT containing the
5644 offset of the symbol within the TLS block, so you can access the value
5645 of the symbol with code such as:
5647 \c mov eax,[tid wrt ..tlsie]
5651 \b In ELF64 mode, referring to an external or global symbol using
5652 \c{wrt ..gottpoff} \I\c{..gottpoff}
5653 causes the linker to build an entry \e{in} the GOT containing the
5654 offset of the symbol within the TLS block, so you can access the value
5655 of the symbol with code such as:
5657 \c mov rax,[rel tid wrt ..gottpoff]
5661 \S{elfglob} \c{elf} Extensions to the \c{GLOBAL} Directive\I{GLOBAL,
5662 elf extensions to}\I{GLOBAL, aoutb extensions to}
5664 \c{ELF} object files can contain more information about a global symbol
5665 than just its address: they can contain the \I{symbol sizes,
5666 specifying}\I{size, of symbols}size of the symbol and its \I{symbol
5667 types, specifying}\I{type, of symbols}type as well. These are not
5668 merely debugger conveniences, but are actually necessary when the
5669 program being written is a \i{shared library}. NASM therefore
5670 supports some extensions to the \c{GLOBAL} directive, allowing you
5671 to specify these features.
5673 You can specify whether a global variable is a function or a data
5674 object by suffixing the name with a colon and the word
5675 \i\c{function} or \i\c{data}. (\i\c{object} is a synonym for
5676 \c{data}.) For example:
5678 \c global hashlookup:function, hashtable:data
5680 exports the global symbol \c{hashlookup} as a function and
5681 \c{hashtable} as a data object.
5683 Optionally, you can control the ELF visibility of the symbol. Just
5684 add one of the visibility keywords: \i\c{default}, \i\c{internal},
5685 \i\c{hidden}, or \i\c{protected}. The default is \i\c{default} of
5686 course. For example, to make \c{hashlookup} hidden:
5688 \c global hashlookup:function hidden
5690 You can also specify the size of the data associated with the
5691 symbol, as a numeric expression (which may involve labels, and even
5692 forward references) after the type specifier. Like this:
5694 \c global hashtable:data (hashtable.end - hashtable)
5697 \c db this,that,theother ; some data here
5700 This makes NASM automatically calculate the length of the table and
5701 place that information into the \c{ELF} symbol table.
5703 Declaring the type and size of global symbols is necessary when
5704 writing shared library code. For more information, see
5708 \S{elfcomm} \c{elf} Extensions to the \c{COMMON} Directive
5709 \I{COMMON, elf extensions to}
5711 \c{ELF} also allows you to specify alignment requirements \I{common
5712 variables, alignment in elf}\I{alignment, of elf common variables}on
5713 common variables. This is done by putting a number (which must be a
5714 power of two) after the name and size of the common variable,
5715 separated (as usual) by a colon. For example, an array of
5716 doublewords would benefit from 4-byte alignment:
5718 \c common dwordarray 128:4
5720 This declares the total size of the array to be 128 bytes, and
5721 requires that it be aligned on a 4-byte boundary.
5724 \S{elf16} 16-bit code and ELF
5725 \I{ELF, 16-bit code and}
5727 The \c{ELF32} specification doesn't provide relocations for 8- and
5728 16-bit values, but the GNU \c{ld} linker adds these as an extension.
5729 NASM can generate GNU-compatible relocations, to allow 16-bit code to
5730 be linked as ELF using GNU \c{ld}. If NASM is used with the
5731 \c{-w+gnu-elf-extensions} option, a warning is issued when one of
5732 these relocations is generated.
5734 \S{elfdbg} Debug formats and ELF
5735 \I{ELF, Debug formats and}
5737 \c{ELF32} and \c{ELF64} provide debug information in \c{STABS} and \c{DWARF} formats.
5738 Line number information is generated for all executable sections, but please
5739 note that only the ".text" section is executable by default.
5741 \H{aoutfmt} \i\c{aout}: Linux \I{a.out, Linux version}\I{linux, a.out}\c{a.out} Object Files
5743 The \c{aout} format generates \c{a.out} object files, in the form used
5744 by early Linux systems (current Linux systems use ELF, see
5745 \k{elffmt}.) These differ from other \c{a.out} object files in that
5746 the magic number in the first four bytes of the file is
5747 different; also, some implementations of \c{a.out}, for example
5748 NetBSD's, support position-independent code, which Linux's
5749 implementation does not.
5751 \c{a.out} provides a default output file-name extension of \c{.o}.
5753 \c{a.out} is a very simple object format. It supports no special
5754 directives, no special symbols, no use of \c{SEG} or \c{WRT}, and no
5755 extensions to any standard directives. It supports only the three
5756 \i{standard section names} \i\c{.text}, \i\c{.data} and \i\c{.bss}.
5759 \H{aoutfmt} \i\c{aoutb}: \i{NetBSD}/\i{FreeBSD}/\i{OpenBSD}
5760 \I{a.out, BSD version}\c{a.out} Object Files
5762 The \c{aoutb} format generates \c{a.out} object files, in the form
5763 used by the various free \c{BSD Unix} clones, \c{NetBSD}, \c{FreeBSD}
5764 and \c{OpenBSD}. For simple object files, this object format is exactly
5765 the same as \c{aout} except for the magic number in the first four bytes
5766 of the file. However, the \c{aoutb} format supports
5767 \I{PIC}\i{position-independent code} in the same way as the \c{elf}
5768 format, so you can use it to write \c{BSD} \i{shared libraries}.
5770 \c{aoutb} provides a default output file-name extension of \c{.o}.
5772 \c{aoutb} supports no special directives, no special symbols, and
5773 only the three \i{standard section names} \i\c{.text}, \i\c{.data}
5774 and \i\c{.bss}. However, it also supports the same use of \i\c{WRT} as
5775 \c{elf} does, to provide position-independent code relocation types.
5776 See \k{elfwrt} for full documentation of this feature.
5778 \c{aoutb} also supports the same extensions to the \c{GLOBAL}
5779 directive as \c{elf} does: see \k{elfglob} for documentation of
5783 \H{as86fmt} \c{as86}: \i{Minix}/Linux\I{linux, as86} \i\c{as86} Object Files
5785 The Minix/Linux 16-bit assembler \c{as86} has its own non-standard
5786 object file format. Although its companion linker \i\c{ld86} produces
5787 something close to ordinary \c{a.out} binaries as output, the object
5788 file format used to communicate between \c{as86} and \c{ld86} is not
5791 NASM supports this format, just in case it is useful, as \c{as86}.
5792 \c{as86} provides a default output file-name extension of \c{.o}.
5794 \c{as86} is a very simple object format (from the NASM user's point
5795 of view). It supports no special directives, no use of \c{SEG} or \c{WRT},
5796 and no extensions to any standard directives. It supports only the three
5797 \i{standard section names} \i\c{.text}, \i\c{.data} and \i\c{.bss}. The
5798 only special symbol supported is \c{..start}.
5801 \H{rdffmt} \I{RDOFF}\i\c{rdf}: \i{Relocatable Dynamic Object File
5804 The \c{rdf} output format produces \c{RDOFF} object files. \c{RDOFF}
5805 (Relocatable Dynamic Object File Format) is a home-grown object-file
5806 format, designed alongside NASM itself and reflecting in its file
5807 format the internal structure of the assembler.
5809 \c{RDOFF} is not used by any well-known operating systems. Those
5810 writing their own systems, however, may well wish to use \c{RDOFF}
5811 as their object format, on the grounds that it is designed primarily
5812 for simplicity and contains very little file-header bureaucracy.
5814 The Unix NASM archive, and the DOS archive which includes sources,
5815 both contain an \I{rdoff subdirectory}\c{rdoff} subdirectory holding
5816 a set of RDOFF utilities: an RDF linker, an \c{RDF} static-library
5817 manager, an RDF file dump utility, and a program which will load and
5818 execute an RDF executable under Linux.
5820 \c{rdf} supports only the \i{standard section names} \i\c{.text},
5821 \i\c{.data} and \i\c{.bss}.
5824 \S{rdflib} Requiring a Library: The \i\c{LIBRARY} Directive
5826 \c{RDOFF} contains a mechanism for an object file to demand a given
5827 library to be linked to the module, either at load time or run time.
5828 This is done by the \c{LIBRARY} directive, which takes one argument
5829 which is the name of the module:
5831 \c library mylib.rdl
5834 \S{rdfmod} Specifying a Module Name: The \i\c{MODULE} Directive
5836 Special \c{RDOFF} header record is used to store the name of the module.
5837 It can be used, for example, by run-time loader to perform dynamic
5838 linking. \c{MODULE} directive takes one argument which is the name
5843 Note that when you statically link modules and tell linker to strip
5844 the symbols from output file, all module names will be stripped too.
5845 To avoid it, you should start module names with \I{$, prefix}\c{$}, like:
5847 \c module $kernel.core
5850 \S{rdfglob} \c{rdf} Extensions to the \c{GLOBAL} Directive\I{GLOBAL,
5853 \c{RDOFF} global symbols can contain additional information needed by
5854 the static linker. You can mark a global symbol as exported, thus
5855 telling the linker do not strip it from target executable or library
5856 file. Like in \c{ELF}, you can also specify whether an exported symbol
5857 is a procedure (function) or data object.
5859 Suffixing the name with a colon and the word \i\c{export} you make the
5862 \c global sys_open:export
5864 To specify that exported symbol is a procedure (function), you add the
5865 word \i\c{proc} or \i\c{function} after declaration:
5867 \c global sys_open:export proc
5869 Similarly, to specify exported data object, add the word \i\c{data}
5870 or \i\c{object} to the directive:
5872 \c global kernel_ticks:export data
5875 \S{rdfimpt} \c{rdf} Extensions to the \c{EXTERN} Directive\I{EXTERN,
5878 By default the \c{EXTERN} directive in \c{RDOFF} declares a "pure external"
5879 symbol (i.e. the static linker will complain if such a symbol is not resolved).
5880 To declare an "imported" symbol, which must be resolved later during a dynamic
5881 linking phase, \c{RDOFF} offers an additional \c{import} modifier. As in
5882 \c{GLOBAL}, you can also specify whether an imported symbol is a procedure
5883 (function) or data object. For example:
5886 \c extern _open:import
5887 \c extern _printf:import proc
5888 \c extern _errno:import data
5890 Here the directive \c{LIBRARY} is also included, which gives the dynamic linker
5891 a hint as to where to find requested symbols.
5894 \H{dbgfmt} \i\c{dbg}: Debugging Format
5896 The \c{dbg} output format is not built into NASM in the default
5897 configuration. If you are building your own NASM executable from the
5898 sources, you can define \i\c{OF_DBG} in \c{output/outform.h} or on the
5899 compiler command line, and obtain the \c{dbg} output format.
5901 The \c{dbg} format does not output an object file as such; instead,
5902 it outputs a text file which contains a complete list of all the
5903 transactions between the main body of NASM and the output-format
5904 back end module. It is primarily intended to aid people who want to
5905 write their own output drivers, so that they can get a clearer idea
5906 of the various requests the main program makes of the output driver,
5907 and in what order they happen.
5909 For simple files, one can easily use the \c{dbg} format like this:
5911 \c nasm -f dbg filename.asm
5913 which will generate a diagnostic file called \c{filename.dbg}.
5914 However, this will not work well on files which were designed for a
5915 different object format, because each object format defines its own
5916 macros (usually user-level forms of directives), and those macros
5917 will not be defined in the \c{dbg} format. Therefore it can be
5918 useful to run NASM twice, in order to do the preprocessing with the
5919 native object format selected:
5921 \c nasm -e -f rdf -o rdfprog.i rdfprog.asm
5922 \c nasm -a -f dbg rdfprog.i
5924 This preprocesses \c{rdfprog.asm} into \c{rdfprog.i}, keeping the
5925 \c{rdf} object format selected in order to make sure RDF special
5926 directives are converted into primitive form correctly. Then the
5927 preprocessed source is fed through the \c{dbg} format to generate
5928 the final diagnostic output.
5930 This workaround will still typically not work for programs intended
5931 for \c{obj} format, because the \c{obj} \c{SEGMENT} and \c{GROUP}
5932 directives have side effects of defining the segment and group names
5933 as symbols; \c{dbg} will not do this, so the program will not
5934 assemble. You will have to work around that by defining the symbols
5935 yourself (using \c{EXTERN}, for example) if you really need to get a
5936 \c{dbg} trace of an \c{obj}-specific source file.
5938 \c{dbg} accepts any section name and any directives at all, and logs
5939 them all to its output file.
5942 \C{16bit} Writing 16-bit Code (DOS, Windows 3/3.1)
5944 This chapter attempts to cover some of the common issues encountered
5945 when writing 16-bit code to run under \c{MS-DOS} or \c{Windows 3.x}. It
5946 covers how to link programs to produce \c{.EXE} or \c{.COM} files,
5947 how to write \c{.SYS} device drivers, and how to interface assembly
5948 language code with 16-bit C compilers and with Borland Pascal.
5951 \H{exefiles} Producing \i\c{.EXE} Files
5953 Any large program written under DOS needs to be built as a \c{.EXE}
5954 file: only \c{.EXE} files have the necessary internal structure
5955 required to span more than one 64K segment. \i{Windows} programs,
5956 also, have to be built as \c{.EXE} files, since Windows does not
5957 support the \c{.COM} format.
5959 In general, you generate \c{.EXE} files by using the \c{obj} output
5960 format to produce one or more \i\c{.OBJ} files, and then linking
5961 them together using a linker. However, NASM also supports the direct
5962 generation of simple DOS \c{.EXE} files using the \c{bin} output
5963 format (by using \c{DB} and \c{DW} to construct the \c{.EXE} file
5964 header), and a macro package is supplied to do this. Thanks to
5965 Yann Guidon for contributing the code for this.
5967 NASM may also support \c{.EXE} natively as another output format in
5971 \S{objexe} Using the \c{obj} Format To Generate \c{.EXE} Files
5973 This section describes the usual method of generating \c{.EXE} files
5974 by linking \c{.OBJ} files together.
5976 Most 16-bit programming language packages come with a suitable
5977 linker; if you have none of these, there is a free linker called
5978 \i{VAL}\I{linker, free}, available in \c{LZH} archive format from
5979 \W{ftp://x2ftp.oulu.fi/pub/msdos/programming/lang/}\i\c{x2ftp.oulu.fi}.
5980 An LZH archiver can be found at
5981 \W{ftp://ftp.simtel.net/pub/simtelnet/msdos/arcers}\i\c{ftp.simtel.net}.
5982 There is another `free' linker (though this one doesn't come with
5983 sources) called \i{FREELINK}, available from
5984 \W{http://www.pcorner.com/tpc/old/3-101.html}\i\c{www.pcorner.com}.
5985 A third, \i\c{djlink}, written by DJ Delorie, is available at
5986 \W{http://www.delorie.com/djgpp/16bit/djlink/}\i\c{www.delorie.com}.
5987 A fourth linker, \i\c{ALINK}, written by Anthony A.J. Williams, is
5988 available at \W{http://alink.sourceforge.net}\i\c{alink.sourceforge.net}.
5990 When linking several \c{.OBJ} files into a \c{.EXE} file, you should
5991 ensure that exactly one of them has a start point defined (using the
5992 \I{program entry point}\i\c{..start} special symbol defined by the
5993 \c{obj} format: see \k{dotdotstart}). If no module defines a start
5994 point, the linker will not know what value to give the entry-point
5995 field in the output file header; if more than one defines a start
5996 point, the linker will not know \e{which} value to use.
5998 An example of a NASM source file which can be assembled to a
5999 \c{.OBJ} file and linked on its own to a \c{.EXE} is given here. It
6000 demonstrates the basic principles of defining a stack, initialising
6001 the segment registers, and declaring a start point. This file is
6002 also provided in the \I{test subdirectory}\c{test} subdirectory of
6003 the NASM archives, under the name \c{objexe.asm}.
6014 This initial piece of code sets up \c{DS} to point to the data
6015 segment, and initializes \c{SS} and \c{SP} to point to the top of
6016 the provided stack. Notice that interrupts are implicitly disabled
6017 for one instruction after a move into \c{SS}, precisely for this
6018 situation, so that there's no chance of an interrupt occurring
6019 between the loads of \c{SS} and \c{SP} and not having a stack to
6022 Note also that the special symbol \c{..start} is defined at the
6023 beginning of this code, which means that will be the entry point
6024 into the resulting executable file.
6030 The above is the main program: load \c{DS:DX} with a pointer to the
6031 greeting message (\c{hello} is implicitly relative to the segment
6032 \c{data}, which was loaded into \c{DS} in the setup code, so the
6033 full pointer is valid), and call the DOS print-string function.
6038 This terminates the program using another DOS system call.
6042 \c hello: db 'hello, world', 13, 10, '$'
6044 The data segment contains the string we want to display.
6046 \c segment stack stack
6050 The above code declares a stack segment containing 64 bytes of
6051 uninitialized stack space, and points \c{stacktop} at the top of it.
6052 The directive \c{segment stack stack} defines a segment \e{called}
6053 \c{stack}, and also of \e{type} \c{STACK}. The latter is not
6054 necessary to the correct running of the program, but linkers are
6055 likely to issue warnings or errors if your program has no segment of
6058 The above file, when assembled into a \c{.OBJ} file, will link on
6059 its own to a valid \c{.EXE} file, which when run will print `hello,
6060 world' and then exit.
6063 \S{binexe} Using the \c{bin} Format To Generate \c{.EXE} Files
6065 The \c{.EXE} file format is simple enough that it's possible to
6066 build a \c{.EXE} file by writing a pure-binary program and sticking
6067 a 32-byte header on the front. This header is simple enough that it
6068 can be generated using \c{DB} and \c{DW} commands by NASM itself, so
6069 that you can use the \c{bin} output format to directly generate
6072 Included in the NASM archives, in the \I{misc subdirectory}\c{misc}
6073 subdirectory, is a file \i\c{exebin.mac} of macros. It defines three
6074 macros: \i\c{EXE_begin}, \i\c{EXE_stack} and \i\c{EXE_end}.
6076 To produce a \c{.EXE} file using this method, you should start by
6077 using \c{%include} to load the \c{exebin.mac} macro package into
6078 your source file. You should then issue the \c{EXE_begin} macro call
6079 (which takes no arguments) to generate the file header data. Then
6080 write code as normal for the \c{bin} format - you can use all three
6081 standard sections \c{.text}, \c{.data} and \c{.bss}. At the end of
6082 the file you should call the \c{EXE_end} macro (again, no arguments),
6083 which defines some symbols to mark section sizes, and these symbols
6084 are referred to in the header code generated by \c{EXE_begin}.
6086 In this model, the code you end up writing starts at \c{0x100}, just
6087 like a \c{.COM} file - in fact, if you strip off the 32-byte header
6088 from the resulting \c{.EXE} file, you will have a valid \c{.COM}
6089 program. All the segment bases are the same, so you are limited to a
6090 64K program, again just like a \c{.COM} file. Note that an \c{ORG}
6091 directive is issued by the \c{EXE_begin} macro, so you should not
6092 explicitly issue one of your own.
6094 You can't directly refer to your segment base value, unfortunately,
6095 since this would require a relocation in the header, and things
6096 would get a lot more complicated. So you should get your segment
6097 base by copying it out of \c{CS} instead.
6099 On entry to your \c{.EXE} file, \c{SS:SP} are already set up to
6100 point to the top of a 2Kb stack. You can adjust the default stack
6101 size of 2Kb by calling the \c{EXE_stack} macro. For example, to
6102 change the stack size of your program to 64 bytes, you would call
6105 A sample program which generates a \c{.EXE} file in this way is
6106 given in the \c{test} subdirectory of the NASM archive, as
6110 \H{comfiles} Producing \i\c{.COM} Files
6112 While large DOS programs must be written as \c{.EXE} files, small
6113 ones are often better written as \c{.COM} files. \c{.COM} files are
6114 pure binary, and therefore most easily produced using the \c{bin}
6118 \S{combinfmt} Using the \c{bin} Format To Generate \c{.COM} Files
6120 \c{.COM} files expect to be loaded at offset \c{100h} into their
6121 segment (though the segment may change). Execution then begins at
6122 \I\c{ORG}\c{100h}, i.e. right at the start of the program. So to
6123 write a \c{.COM} program, you would create a source file looking
6131 \c ; put your code here
6135 \c ; put data items here
6139 \c ; put uninitialized data here
6141 The \c{bin} format puts the \c{.text} section first in the file, so
6142 you can declare data or BSS items before beginning to write code if
6143 you want to and the code will still end up at the front of the file
6146 The BSS (uninitialized data) section does not take up space in the
6147 \c{.COM} file itself: instead, addresses of BSS items are resolved
6148 to point at space beyond the end of the file, on the grounds that
6149 this will be free memory when the program is run. Therefore you
6150 should not rely on your BSS being initialized to all zeros when you
6153 To assemble the above program, you should use a command line like
6155 \c nasm myprog.asm -fbin -o myprog.com
6157 The \c{bin} format would produce a file called \c{myprog} if no
6158 explicit output file name were specified, so you have to override it
6159 and give the desired file name.
6162 \S{comobjfmt} Using the \c{obj} Format To Generate \c{.COM} Files
6164 If you are writing a \c{.COM} program as more than one module, you
6165 may wish to assemble several \c{.OBJ} files and link them together
6166 into a \c{.COM} program. You can do this, provided you have a linker
6167 capable of outputting \c{.COM} files directly (\i{TLINK} does this),
6168 or alternatively a converter program such as \i\c{EXE2BIN} to
6169 transform the \c{.EXE} file output from the linker into a \c{.COM}
6172 If you do this, you need to take care of several things:
6174 \b The first object file containing code should start its code
6175 segment with a line like \c{RESB 100h}. This is to ensure that the
6176 code begins at offset \c{100h} relative to the beginning of the code
6177 segment, so that the linker or converter program does not have to
6178 adjust address references within the file when generating the
6179 \c{.COM} file. Other assemblers use an \i\c{ORG} directive for this
6180 purpose, but \c{ORG} in NASM is a format-specific directive to the
6181 \c{bin} output format, and does not mean the same thing as it does
6182 in MASM-compatible assemblers.
6184 \b You don't need to define a stack segment.
6186 \b All your segments should be in the same group, so that every time
6187 your code or data references a symbol offset, all offsets are
6188 relative to the same segment base. This is because, when a \c{.COM}
6189 file is loaded, all the segment registers contain the same value.
6192 \H{sysfiles} Producing \i\c{.SYS} Files
6194 \i{MS-DOS device drivers} - \c{.SYS} files - are pure binary files,
6195 similar to \c{.COM} files, except that they start at origin zero
6196 rather than \c{100h}. Therefore, if you are writing a device driver
6197 using the \c{bin} format, you do not need the \c{ORG} directive,
6198 since the default origin for \c{bin} is zero. Similarly, if you are
6199 using \c{obj}, you do not need the \c{RESB 100h} at the start of
6202 \c{.SYS} files start with a header structure, containing pointers to
6203 the various routines inside the driver which do the work. This
6204 structure should be defined at the start of the code segment, even
6205 though it is not actually code.
6207 For more information on the format of \c{.SYS} files, and the data
6208 which has to go in the header structure, a list of books is given in
6209 the Frequently Asked Questions list for the newsgroup
6210 \W{news:comp.os.msdos.programmer}\i\c{comp.os.msdos.programmer}.
6213 \H{16c} Interfacing to 16-bit C Programs
6215 This section covers the basics of writing assembly routines that
6216 call, or are called from, C programs. To do this, you would
6217 typically write an assembly module as a \c{.OBJ} file, and link it
6218 with your C modules to produce a \i{mixed-language program}.
6221 \S{16cunder} External Symbol Names
6223 \I{C symbol names}\I{underscore, in C symbols}C compilers have the
6224 convention that the names of all global symbols (functions or data)
6225 they define are formed by prefixing an underscore to the name as it
6226 appears in the C program. So, for example, the function a C
6227 programmer thinks of as \c{printf} appears to an assembly language
6228 programmer as \c{_printf}. This means that in your assembly
6229 programs, you can define symbols without a leading underscore, and
6230 not have to worry about name clashes with C symbols.
6232 If you find the underscores inconvenient, you can define macros to
6233 replace the \c{GLOBAL} and \c{EXTERN} directives as follows:
6249 (These forms of the macros only take one argument at a time; a
6250 \c{%rep} construct could solve this.)
6252 If you then declare an external like this:
6256 then the macro will expand it as
6259 \c %define printf _printf
6261 Thereafter, you can reference \c{printf} as if it was a symbol, and
6262 the preprocessor will put the leading underscore on where necessary.
6264 The \c{cglobal} macro works similarly. You must use \c{cglobal}
6265 before defining the symbol in question, but you would have had to do
6266 that anyway if you used \c{GLOBAL}.
6268 Also see \k{opt-pfix}.
6270 \S{16cmodels} \i{Memory Models}
6272 NASM contains no mechanism to support the various C memory models
6273 directly; you have to keep track yourself of which one you are
6274 writing for. This means you have to keep track of the following
6277 \b In models using a single code segment (tiny, small and compact),
6278 functions are near. This means that function pointers, when stored
6279 in data segments or pushed on the stack as function arguments, are
6280 16 bits long and contain only an offset field (the \c{CS} register
6281 never changes its value, and always gives the segment part of the
6282 full function address), and that functions are called using ordinary
6283 near \c{CALL} instructions and return using \c{RETN} (which, in
6284 NASM, is synonymous with \c{RET} anyway). This means both that you
6285 should write your own routines to return with \c{RETN}, and that you
6286 should call external C routines with near \c{CALL} instructions.
6288 \b In models using more than one code segment (medium, large and
6289 huge), functions are far. This means that function pointers are 32
6290 bits long (consisting of a 16-bit offset followed by a 16-bit
6291 segment), and that functions are called using \c{CALL FAR} (or
6292 \c{CALL seg:offset}) and return using \c{RETF}. Again, you should
6293 therefore write your own routines to return with \c{RETF} and use
6294 \c{CALL FAR} to call external routines.
6296 \b In models using a single data segment (tiny, small and medium),
6297 data pointers are 16 bits long, containing only an offset field (the
6298 \c{DS} register doesn't change its value, and always gives the
6299 segment part of the full data item address).
6301 \b In models using more than one data segment (compact, large and
6302 huge), data pointers are 32 bits long, consisting of a 16-bit offset
6303 followed by a 16-bit segment. You should still be careful not to
6304 modify \c{DS} in your routines without restoring it afterwards, but
6305 \c{ES} is free for you to use to access the contents of 32-bit data
6306 pointers you are passed.
6308 \b The huge memory model allows single data items to exceed 64K in
6309 size. In all other memory models, you can access the whole of a data
6310 item just by doing arithmetic on the offset field of the pointer you
6311 are given, whether a segment field is present or not; in huge model,
6312 you have to be more careful of your pointer arithmetic.
6314 \b In most memory models, there is a \e{default} data segment, whose
6315 segment address is kept in \c{DS} throughout the program. This data
6316 segment is typically the same segment as the stack, kept in \c{SS},
6317 so that functions' local variables (which are stored on the stack)
6318 and global data items can both be accessed easily without changing
6319 \c{DS}. Particularly large data items are typically stored in other
6320 segments. However, some memory models (though not the standard
6321 ones, usually) allow the assumption that \c{SS} and \c{DS} hold the
6322 same value to be removed. Be careful about functions' local
6323 variables in this latter case.
6325 In models with a single code segment, the segment is called
6326 \i\c{_TEXT}, so your code segment must also go by this name in order
6327 to be linked into the same place as the main code segment. In models
6328 with a single data segment, or with a default data segment, it is
6332 \S{16cfunc} Function Definitions and Function Calls
6334 \I{functions, C calling convention}The \i{C calling convention} in
6335 16-bit programs is as follows. In the following description, the
6336 words \e{caller} and \e{callee} are used to denote the function
6337 doing the calling and the function which gets called.
6339 \b The caller pushes the function's parameters on the stack, one
6340 after another, in reverse order (right to left, so that the first
6341 argument specified to the function is pushed last).
6343 \b The caller then executes a \c{CALL} instruction to pass control
6344 to the callee. This \c{CALL} is either near or far depending on the
6347 \b The callee receives control, and typically (although this is not
6348 actually necessary, in functions which do not need to access their
6349 parameters) starts by saving the value of \c{SP} in \c{BP} so as to
6350 be able to use \c{BP} as a base pointer to find its parameters on
6351 the stack. However, the caller was probably doing this too, so part
6352 of the calling convention states that \c{BP} must be preserved by
6353 any C function. Hence the callee, if it is going to set up \c{BP} as
6354 a \i\e{frame pointer}, must push the previous value first.
6356 \b The callee may then access its parameters relative to \c{BP}.
6357 The word at \c{[BP]} holds the previous value of \c{BP} as it was
6358 pushed; the next word, at \c{[BP+2]}, holds the offset part of the
6359 return address, pushed implicitly by \c{CALL}. In a small-model
6360 (near) function, the parameters start after that, at \c{[BP+4]}; in
6361 a large-model (far) function, the segment part of the return address
6362 lives at \c{[BP+4]}, and the parameters begin at \c{[BP+6]}. The
6363 leftmost parameter of the function, since it was pushed last, is
6364 accessible at this offset from \c{BP}; the others follow, at
6365 successively greater offsets. Thus, in a function such as \c{printf}
6366 which takes a variable number of parameters, the pushing of the
6367 parameters in reverse order means that the function knows where to
6368 find its first parameter, which tells it the number and type of the
6371 \b The callee may also wish to decrease \c{SP} further, so as to
6372 allocate space on the stack for local variables, which will then be
6373 accessible at negative offsets from \c{BP}.
6375 \b The callee, if it wishes to return a value to the caller, should
6376 leave the value in \c{AL}, \c{AX} or \c{DX:AX} depending on the size
6377 of the value. Floating-point results are sometimes (depending on the
6378 compiler) returned in \c{ST0}.
6380 \b Once the callee has finished processing, it restores \c{SP} from
6381 \c{BP} if it had allocated local stack space, then pops the previous
6382 value of \c{BP}, and returns via \c{RETN} or \c{RETF} depending on
6385 \b When the caller regains control from the callee, the function
6386 parameters are still on the stack, so it typically adds an immediate
6387 constant to \c{SP} to remove them (instead of executing a number of
6388 slow \c{POP} instructions). Thus, if a function is accidentally
6389 called with the wrong number of parameters due to a prototype
6390 mismatch, the stack will still be returned to a sensible state since
6391 the caller, which \e{knows} how many parameters it pushed, does the
6394 It is instructive to compare this calling convention with that for
6395 Pascal programs (described in \k{16bpfunc}). Pascal has a simpler
6396 convention, since no functions have variable numbers of parameters.
6397 Therefore the callee knows how many parameters it should have been
6398 passed, and is able to deallocate them from the stack itself by
6399 passing an immediate argument to the \c{RET} or \c{RETF}
6400 instruction, so the caller does not have to do it. Also, the
6401 parameters are pushed in left-to-right order, not right-to-left,
6402 which means that a compiler can give better guarantees about
6403 sequence points without performance suffering.
6405 Thus, you would define a function in C style in the following way.
6406 The following example is for small model:
6413 \c sub sp,0x40 ; 64 bytes of local stack space
6414 \c mov bx,[bp+4] ; first parameter to function
6418 \c mov sp,bp ; undo "sub sp,0x40" above
6422 For a large-model function, you would replace \c{RET} by \c{RETF},
6423 and look for the first parameter at \c{[BP+6]} instead of
6424 \c{[BP+4]}. Of course, if one of the parameters is a pointer, then
6425 the offsets of \e{subsequent} parameters will change depending on
6426 the memory model as well: far pointers take up four bytes on the
6427 stack when passed as a parameter, whereas near pointers take up two.
6429 At the other end of the process, to call a C function from your
6430 assembly code, you would do something like this:
6434 \c ; and then, further down...
6436 \c push word [myint] ; one of my integer variables
6437 \c push word mystring ; pointer into my data segment
6439 \c add sp,byte 4 ; `byte' saves space
6441 \c ; then those data items...
6446 \c mystring db 'This number -> %d <- should be 1234',10,0
6448 This piece of code is the small-model assembly equivalent of the C
6451 \c int myint = 1234;
6452 \c printf("This number -> %d <- should be 1234\n", myint);
6454 In large model, the function-call code might look more like this. In
6455 this example, it is assumed that \c{DS} already holds the segment
6456 base of the segment \c{_DATA}. If not, you would have to initialize
6459 \c push word [myint]
6460 \c push word seg mystring ; Now push the segment, and...
6461 \c push word mystring ; ... offset of "mystring"
6465 The integer value still takes up one word on the stack, since large
6466 model does not affect the size of the \c{int} data type. The first
6467 argument (pushed last) to \c{printf}, however, is a data pointer,
6468 and therefore has to contain a segment and offset part. The segment
6469 should be stored second in memory, and therefore must be pushed
6470 first. (Of course, \c{PUSH DS} would have been a shorter instruction
6471 than \c{PUSH WORD SEG mystring}, if \c{DS} was set up as the above
6472 example assumed.) Then the actual call becomes a far call, since
6473 functions expect far calls in large model; and \c{SP} has to be
6474 increased by 6 rather than 4 afterwards to make up for the extra
6478 \S{16cdata} Accessing Data Items
6480 To get at the contents of C variables, or to declare variables which
6481 C can access, you need only declare the names as \c{GLOBAL} or
6482 \c{EXTERN}. (Again, the names require leading underscores, as stated
6483 in \k{16cunder}.) Thus, a C variable declared as \c{int i} can be
6484 accessed from assembler as
6490 And to declare your own integer variable which C programs can access
6491 as \c{extern int j}, you do this (making sure you are assembling in
6492 the \c{_DATA} segment, if necessary):
6498 To access a C array, you need to know the size of the components of
6499 the array. For example, \c{int} variables are two bytes long, so if
6500 a C program declares an array as \c{int a[10]}, you can access
6501 \c{a[3]} by coding \c{mov ax,[_a+6]}. (The byte offset 6 is obtained
6502 by multiplying the desired array index, 3, by the size of the array
6503 element, 2.) The sizes of the C base types in 16-bit compilers are:
6504 1 for \c{char}, 2 for \c{short} and \c{int}, 4 for \c{long} and
6505 \c{float}, and 8 for \c{double}.
6507 To access a C \i{data structure}, you need to know the offset from
6508 the base of the structure to the field you are interested in. You
6509 can either do this by converting the C structure definition into a
6510 NASM structure definition (using \i\c{STRUC}), or by calculating the
6511 one offset and using just that.
6513 To do either of these, you should read your C compiler's manual to
6514 find out how it organizes data structures. NASM gives no special
6515 alignment to structure members in its own \c{STRUC} macro, so you
6516 have to specify alignment yourself if the C compiler generates it.
6517 Typically, you might find that a structure like
6524 might be four bytes long rather than three, since the \c{int} field
6525 would be aligned to a two-byte boundary. However, this sort of
6526 feature tends to be a configurable option in the C compiler, either
6527 using command-line options or \c{#pragma} lines, so you have to find
6528 out how your own compiler does it.
6531 \S{16cmacro} \i\c{c16.mac}: Helper Macros for the 16-bit C Interface
6533 Included in the NASM archives, in the \I{misc subdirectory}\c{misc}
6534 directory, is a file \c{c16.mac} of macros. It defines three macros:
6535 \i\c{proc}, \i\c{arg} and \i\c{endproc}. These are intended to be
6536 used for C-style procedure definitions, and they automate a lot of
6537 the work involved in keeping track of the calling convention.
6539 (An alternative, TASM compatible form of \c{arg} is also now built
6540 into NASM's preprocessor. See \k{stackrel} for details.)
6542 An example of an assembly function using the macro set is given
6549 \c mov ax,[bp + %$i]
6550 \c mov bx,[bp + %$j]
6555 This defines \c{_nearproc} to be a procedure taking two arguments,
6556 the first (\c{i}) an integer and the second (\c{j}) a pointer to an
6557 integer. It returns \c{i + *j}.
6559 Note that the \c{arg} macro has an \c{EQU} as the first line of its
6560 expansion, and since the label before the macro call gets prepended
6561 to the first line of the expanded macro, the \c{EQU} works, defining
6562 \c{%$i} to be an offset from \c{BP}. A context-local variable is
6563 used, local to the context pushed by the \c{proc} macro and popped
6564 by the \c{endproc} macro, so that the same argument name can be used
6565 in later procedures. Of course, you don't \e{have} to do that.
6567 The macro set produces code for near functions (tiny, small and
6568 compact-model code) by default. You can have it generate far
6569 functions (medium, large and huge-model code) by means of coding
6570 \I\c{FARCODE}\c{%define FARCODE}. This changes the kind of return
6571 instruction generated by \c{endproc}, and also changes the starting
6572 point for the argument offsets. The macro set contains no intrinsic
6573 dependency on whether data pointers are far or not.
6575 \c{arg} can take an optional parameter, giving the size of the
6576 argument. If no size is given, 2 is assumed, since it is likely that
6577 many function parameters will be of type \c{int}.
6579 The large-model equivalent of the above function would look like this:
6587 \c mov ax,[bp + %$i]
6588 \c mov bx,[bp + %$j]
6589 \c mov es,[bp + %$j + 2]
6594 This makes use of the argument to the \c{arg} macro to define a
6595 parameter of size 4, because \c{j} is now a far pointer. When we
6596 load from \c{j}, we must load a segment and an offset.
6599 \H{16bp} Interfacing to \i{Borland Pascal} Programs
6601 Interfacing to Borland Pascal programs is similar in concept to
6602 interfacing to 16-bit C programs. The differences are:
6604 \b The leading underscore required for interfacing to C programs is
6605 not required for Pascal.
6607 \b The memory model is always large: functions are far, data
6608 pointers are far, and no data item can be more than 64K long.
6609 (Actually, some functions are near, but only those functions that
6610 are local to a Pascal unit and never called from outside it. All
6611 assembly functions that Pascal calls, and all Pascal functions that
6612 assembly routines are able to call, are far.) However, all static
6613 data declared in a Pascal program goes into the default data
6614 segment, which is the one whose segment address will be in \c{DS}
6615 when control is passed to your assembly code. The only things that
6616 do not live in the default data segment are local variables (they
6617 live in the stack segment) and dynamically allocated variables. All
6618 data \e{pointers}, however, are far.
6620 \b The function calling convention is different - described below.
6622 \b Some data types, such as strings, are stored differently.
6624 \b There are restrictions on the segment names you are allowed to
6625 use - Borland Pascal will ignore code or data declared in a segment
6626 it doesn't like the name of. The restrictions are described below.
6629 \S{16bpfunc} The Pascal Calling Convention
6631 \I{functions, Pascal calling convention}\I{Pascal calling
6632 convention}The 16-bit Pascal calling convention is as follows. In
6633 the following description, the words \e{caller} and \e{callee} are
6634 used to denote the function doing the calling and the function which
6637 \b The caller pushes the function's parameters on the stack, one
6638 after another, in normal order (left to right, so that the first
6639 argument specified to the function is pushed first).
6641 \b The caller then executes a far \c{CALL} instruction to pass
6642 control to the callee.
6644 \b The callee receives control, and typically (although this is not
6645 actually necessary, in functions which do not need to access their
6646 parameters) starts by saving the value of \c{SP} in \c{BP} so as to
6647 be able to use \c{BP} as a base pointer to find its parameters on
6648 the stack. However, the caller was probably doing this too, so part
6649 of the calling convention states that \c{BP} must be preserved by
6650 any function. Hence the callee, if it is going to set up \c{BP} as a
6651 \i{frame pointer}, must push the previous value first.
6653 \b The callee may then access its parameters relative to \c{BP}.
6654 The word at \c{[BP]} holds the previous value of \c{BP} as it was
6655 pushed. The next word, at \c{[BP+2]}, holds the offset part of the
6656 return address, and the next one at \c{[BP+4]} the segment part. The
6657 parameters begin at \c{[BP+6]}. The rightmost parameter of the
6658 function, since it was pushed last, is accessible at this offset
6659 from \c{BP}; the others follow, at successively greater offsets.
6661 \b The callee may also wish to decrease \c{SP} further, so as to
6662 allocate space on the stack for local variables, which will then be
6663 accessible at negative offsets from \c{BP}.
6665 \b The callee, if it wishes to return a value to the caller, should
6666 leave the value in \c{AL}, \c{AX} or \c{DX:AX} depending on the size
6667 of the value. Floating-point results are returned in \c{ST0}.
6668 Results of type \c{Real} (Borland's own custom floating-point data
6669 type, not handled directly by the FPU) are returned in \c{DX:BX:AX}.
6670 To return a result of type \c{String}, the caller pushes a pointer
6671 to a temporary string before pushing the parameters, and the callee
6672 places the returned string value at that location. The pointer is
6673 not a parameter, and should not be removed from the stack by the
6674 \c{RETF} instruction.
6676 \b Once the callee has finished processing, it restores \c{SP} from
6677 \c{BP} if it had allocated local stack space, then pops the previous
6678 value of \c{BP}, and returns via \c{RETF}. It uses the form of
6679 \c{RETF} with an immediate parameter, giving the number of bytes
6680 taken up by the parameters on the stack. This causes the parameters
6681 to be removed from the stack as a side effect of the return
6684 \b When the caller regains control from the callee, the function
6685 parameters have already been removed from the stack, so it needs to
6688 Thus, you would define a function in Pascal style, taking two
6689 \c{Integer}-type parameters, in the following way:
6695 \c sub sp,0x40 ; 64 bytes of local stack space
6696 \c mov bx,[bp+8] ; first parameter to function
6697 \c mov bx,[bp+6] ; second parameter to function
6701 \c mov sp,bp ; undo "sub sp,0x40" above
6703 \c retf 4 ; total size of params is 4
6705 At the other end of the process, to call a Pascal function from your
6706 assembly code, you would do something like this:
6710 \c ; and then, further down...
6712 \c push word seg mystring ; Now push the segment, and...
6713 \c push word mystring ; ... offset of "mystring"
6714 \c push word [myint] ; one of my variables
6715 \c call far SomeFunc
6717 This is equivalent to the Pascal code
6719 \c procedure SomeFunc(String: PChar; Int: Integer);
6720 \c SomeFunc(@mystring, myint);
6723 \S{16bpseg} Borland Pascal \I{segment names, Borland Pascal}Segment
6726 Since Borland Pascal's internal unit file format is completely
6727 different from \c{OBJ}, it only makes a very sketchy job of actually
6728 reading and understanding the various information contained in a
6729 real \c{OBJ} file when it links that in. Therefore an object file
6730 intended to be linked to a Pascal program must obey a number of
6733 \b Procedures and functions must be in a segment whose name is
6734 either \c{CODE}, \c{CSEG}, or something ending in \c{_TEXT}.
6736 \b initialized data must be in a segment whose name is either
6737 \c{CONST} or something ending in \c{_DATA}.
6739 \b Uninitialized data must be in a segment whose name is either
6740 \c{DATA}, \c{DSEG}, or something ending in \c{_BSS}.
6742 \b Any other segments in the object file are completely ignored.
6743 \c{GROUP} directives and segment attributes are also ignored.
6746 \S{16bpmacro} Using \i\c{c16.mac} With Pascal Programs
6748 The \c{c16.mac} macro package, described in \k{16cmacro}, can also
6749 be used to simplify writing functions to be called from Pascal
6750 programs, if you code \I\c{PASCAL}\c{%define PASCAL}. This
6751 definition ensures that functions are far (it implies
6752 \i\c{FARCODE}), and also causes procedure return instructions to be
6753 generated with an operand.
6755 Defining \c{PASCAL} does not change the code which calculates the
6756 argument offsets; you must declare your function's arguments in
6757 reverse order. For example:
6765 \c mov ax,[bp + %$i]
6766 \c mov bx,[bp + %$j]
6767 \c mov es,[bp + %$j + 2]
6772 This defines the same routine, conceptually, as the example in
6773 \k{16cmacro}: it defines a function taking two arguments, an integer
6774 and a pointer to an integer, which returns the sum of the integer
6775 and the contents of the pointer. The only difference between this
6776 code and the large-model C version is that \c{PASCAL} is defined
6777 instead of \c{FARCODE}, and that the arguments are declared in
6781 \C{32bit} Writing 32-bit Code (Unix, Win32, DJGPP)
6783 This chapter attempts to cover some of the common issues involved
6784 when writing 32-bit code, to run under \i{Win32} or Unix, or to be
6785 linked with C code generated by a Unix-style C compiler such as
6786 \i{DJGPP}. It covers how to write assembly code to interface with
6787 32-bit C routines, and how to write position-independent code for
6790 Almost all 32-bit code, and in particular all code running under
6791 \c{Win32}, \c{DJGPP} or any of the PC Unix variants, runs in \I{flat
6792 memory model}\e{flat} memory model. This means that the segment registers
6793 and paging have already been set up to give you the same 32-bit 4Gb
6794 address space no matter what segment you work relative to, and that
6795 you should ignore all segment registers completely. When writing
6796 flat-model application code, you never need to use a segment
6797 override or modify any segment register, and the code-section
6798 addresses you pass to \c{CALL} and \c{JMP} live in the same address
6799 space as the data-section addresses you access your variables by and
6800 the stack-section addresses you access local variables and procedure
6801 parameters by. Every address is 32 bits long and contains only an
6805 \H{32c} Interfacing to 32-bit C Programs
6807 A lot of the discussion in \k{16c}, about interfacing to 16-bit C
6808 programs, still applies when working in 32 bits. The absence of
6809 memory models or segmentation worries simplifies things a lot.
6812 \S{32cunder} External Symbol Names
6814 Most 32-bit C compilers share the convention used by 16-bit
6815 compilers, that the names of all global symbols (functions or data)
6816 they define are formed by prefixing an underscore to the name as it
6817 appears in the C program. However, not all of them do: the \c{ELF}
6818 specification states that C symbols do \e{not} have a leading
6819 underscore on their assembly-language names.
6821 The older Linux \c{a.out} C compiler, all \c{Win32} compilers,
6822 \c{DJGPP}, and \c{NetBSD} and \c{FreeBSD}, all use the leading
6823 underscore; for these compilers, the macros \c{cextern} and
6824 \c{cglobal}, as given in \k{16cunder}, will still work. For \c{ELF},
6825 though, the leading underscore should not be used.
6827 See also \k{opt-pfix}.
6829 \S{32cfunc} Function Definitions and Function Calls
6831 \I{functions, C calling convention}The \i{C calling convention}
6832 in 32-bit programs is as follows. In the following description,
6833 the words \e{caller} and \e{callee} are used to denote
6834 the function doing the calling and the function which gets called.
6836 \b The caller pushes the function's parameters on the stack, one
6837 after another, in reverse order (right to left, so that the first
6838 argument specified to the function is pushed last).
6840 \b The caller then executes a near \c{CALL} instruction to pass
6841 control to the callee.
6843 \b The callee receives control, and typically (although this is not
6844 actually necessary, in functions which do not need to access their
6845 parameters) starts by saving the value of \c{ESP} in \c{EBP} so as
6846 to be able to use \c{EBP} as a base pointer to find its parameters
6847 on the stack. However, the caller was probably doing this too, so
6848 part of the calling convention states that \c{EBP} must be preserved
6849 by any C function. Hence the callee, if it is going to set up
6850 \c{EBP} as a \i{frame pointer}, must push the previous value first.
6852 \b The callee may then access its parameters relative to \c{EBP}.
6853 The doubleword at \c{[EBP]} holds the previous value of \c{EBP} as
6854 it was pushed; the next doubleword, at \c{[EBP+4]}, holds the return
6855 address, pushed implicitly by \c{CALL}. The parameters start after
6856 that, at \c{[EBP+8]}. The leftmost parameter of the function, since
6857 it was pushed last, is accessible at this offset from \c{EBP}; the
6858 others follow, at successively greater offsets. Thus, in a function
6859 such as \c{printf} which takes a variable number of parameters, the
6860 pushing of the parameters in reverse order means that the function
6861 knows where to find its first parameter, which tells it the number
6862 and type of the remaining ones.
6864 \b The callee may also wish to decrease \c{ESP} further, so as to
6865 allocate space on the stack for local variables, which will then be
6866 accessible at negative offsets from \c{EBP}.
6868 \b The callee, if it wishes to return a value to the caller, should
6869 leave the value in \c{AL}, \c{AX} or \c{EAX} depending on the size
6870 of the value. Floating-point results are typically returned in
6873 \b Once the callee has finished processing, it restores \c{ESP} from
6874 \c{EBP} if it had allocated local stack space, then pops the previous
6875 value of \c{EBP}, and returns via \c{RET} (equivalently, \c{RETN}).
6877 \b When the caller regains control from the callee, the function
6878 parameters are still on the stack, so it typically adds an immediate
6879 constant to \c{ESP} to remove them (instead of executing a number of
6880 slow \c{POP} instructions). Thus, if a function is accidentally
6881 called with the wrong number of parameters due to a prototype
6882 mismatch, the stack will still be returned to a sensible state since
6883 the caller, which \e{knows} how many parameters it pushed, does the
6886 There is an alternative calling convention used by Win32 programs
6887 for Windows API calls, and also for functions called \e{by} the
6888 Windows API such as window procedures: they follow what Microsoft
6889 calls the \c{__stdcall} convention. This is slightly closer to the
6890 Pascal convention, in that the callee clears the stack by passing a
6891 parameter to the \c{RET} instruction. However, the parameters are
6892 still pushed in right-to-left order.
6894 Thus, you would define a function in C style in the following way:
6901 \c sub esp,0x40 ; 64 bytes of local stack space
6902 \c mov ebx,[ebp+8] ; first parameter to function
6906 \c leave ; mov esp,ebp / pop ebp
6909 At the other end of the process, to call a C function from your
6910 assembly code, you would do something like this:
6914 \c ; and then, further down...
6916 \c push dword [myint] ; one of my integer variables
6917 \c push dword mystring ; pointer into my data segment
6919 \c add esp,byte 8 ; `byte' saves space
6921 \c ; then those data items...
6926 \c mystring db 'This number -> %d <- should be 1234',10,0
6928 This piece of code is the assembly equivalent of the C code
6930 \c int myint = 1234;
6931 \c printf("This number -> %d <- should be 1234\n", myint);
6934 \S{32cdata} Accessing Data Items
6936 To get at the contents of C variables, or to declare variables which
6937 C can access, you need only declare the names as \c{GLOBAL} or
6938 \c{EXTERN}. (Again, the names require leading underscores, as stated
6939 in \k{32cunder}.) Thus, a C variable declared as \c{int i} can be
6940 accessed from assembler as
6945 And to declare your own integer variable which C programs can access
6946 as \c{extern int j}, you do this (making sure you are assembling in
6947 the \c{_DATA} segment, if necessary):
6952 To access a C array, you need to know the size of the components of
6953 the array. For example, \c{int} variables are four bytes long, so if
6954 a C program declares an array as \c{int a[10]}, you can access
6955 \c{a[3]} by coding \c{mov ax,[_a+12]}. (The byte offset 12 is obtained
6956 by multiplying the desired array index, 3, by the size of the array
6957 element, 4.) The sizes of the C base types in 32-bit compilers are:
6958 1 for \c{char}, 2 for \c{short}, 4 for \c{int}, \c{long} and
6959 \c{float}, and 8 for \c{double}. Pointers, being 32-bit addresses,
6960 are also 4 bytes long.
6962 To access a C \i{data structure}, you need to know the offset from
6963 the base of the structure to the field you are interested in. You
6964 can either do this by converting the C structure definition into a
6965 NASM structure definition (using \c{STRUC}), or by calculating the
6966 one offset and using just that.
6968 To do either of these, you should read your C compiler's manual to
6969 find out how it organizes data structures. NASM gives no special
6970 alignment to structure members in its own \i\c{STRUC} macro, so you
6971 have to specify alignment yourself if the C compiler generates it.
6972 Typically, you might find that a structure like
6979 might be eight bytes long rather than five, since the \c{int} field
6980 would be aligned to a four-byte boundary. However, this sort of
6981 feature is sometimes a configurable option in the C compiler, either
6982 using command-line options or \c{#pragma} lines, so you have to find
6983 out how your own compiler does it.
6986 \S{32cmacro} \i\c{c32.mac}: Helper Macros for the 32-bit C Interface
6988 Included in the NASM archives, in the \I{misc directory}\c{misc}
6989 directory, is a file \c{c32.mac} of macros. It defines three macros:
6990 \i\c{proc}, \i\c{arg} and \i\c{endproc}. These are intended to be
6991 used for C-style procedure definitions, and they automate a lot of
6992 the work involved in keeping track of the calling convention.
6994 An example of an assembly function using the macro set is given
7001 \c mov eax,[ebp + %$i]
7002 \c mov ebx,[ebp + %$j]
7007 This defines \c{_proc32} to be a procedure taking two arguments, the
7008 first (\c{i}) an integer and the second (\c{j}) a pointer to an
7009 integer. It returns \c{i + *j}.
7011 Note that the \c{arg} macro has an \c{EQU} as the first line of its
7012 expansion, and since the label before the macro call gets prepended
7013 to the first line of the expanded macro, the \c{EQU} works, defining
7014 \c{%$i} to be an offset from \c{BP}. A context-local variable is
7015 used, local to the context pushed by the \c{proc} macro and popped
7016 by the \c{endproc} macro, so that the same argument name can be used
7017 in later procedures. Of course, you don't \e{have} to do that.
7019 \c{arg} can take an optional parameter, giving the size of the
7020 argument. If no size is given, 4 is assumed, since it is likely that
7021 many function parameters will be of type \c{int} or pointers.
7024 \H{picdll} Writing NetBSD/FreeBSD/OpenBSD and Linux/ELF \i{Shared
7027 \c{ELF} replaced the older \c{a.out} object file format under Linux
7028 because it contains support for \i{position-independent code}
7029 (\i{PIC}), which makes writing shared libraries much easier. NASM
7030 supports the \c{ELF} position-independent code features, so you can
7031 write Linux \c{ELF} shared libraries in NASM.
7033 \i{NetBSD}, and its close cousins \i{FreeBSD} and \i{OpenBSD}, take
7034 a different approach by hacking PIC support into the \c{a.out}
7035 format. NASM supports this as the \i\c{aoutb} output format, so you
7036 can write \i{BSD} shared libraries in NASM too.
7038 The operating system loads a PIC shared library by memory-mapping
7039 the library file at an arbitrarily chosen point in the address space
7040 of the running process. The contents of the library's code section
7041 must therefore not depend on where it is loaded in memory.
7043 Therefore, you cannot get at your variables by writing code like
7046 \c mov eax,[myvar] ; WRONG
7048 Instead, the linker provides an area of memory called the
7049 \i\e{global offset table}, or \i{GOT}; the GOT is situated at a
7050 constant distance from your library's code, so if you can find out
7051 where your library is loaded (which is typically done using a
7052 \c{CALL} and \c{POP} combination), you can obtain the address of the
7053 GOT, and you can then load the addresses of your variables out of
7054 linker-generated entries in the GOT.
7056 The \e{data} section of a PIC shared library does not have these
7057 restrictions: since the data section is writable, it has to be
7058 copied into memory anyway rather than just paged in from the library
7059 file, so as long as it's being copied it can be relocated too. So
7060 you can put ordinary types of relocation in the data section without
7061 too much worry (but see \k{picglobal} for a caveat).
7064 \S{picgot} Obtaining the Address of the GOT
7066 Each code module in your shared library should define the GOT as an
7069 \c extern _GLOBAL_OFFSET_TABLE_ ; in ELF
7070 \c extern __GLOBAL_OFFSET_TABLE_ ; in BSD a.out
7072 At the beginning of any function in your shared library which plans
7073 to access your data or BSS sections, you must first calculate the
7074 address of the GOT. This is typically done by writing the function
7083 \c add ebx,_GLOBAL_OFFSET_TABLE_+$$-.get_GOT wrt ..gotpc
7085 \c ; the function body comes here
7092 (For BSD, again, the symbol \c{_GLOBAL_OFFSET_TABLE} requires a
7093 second leading underscore.)
7095 The first two lines of this function are simply the standard C
7096 prologue to set up a stack frame, and the last three lines are
7097 standard C function epilogue. The third line, and the fourth to last
7098 line, save and restore the \c{EBX} register, because PIC shared
7099 libraries use this register to store the address of the GOT.
7101 The interesting bit is the \c{CALL} instruction and the following
7102 two lines. The \c{CALL} and \c{POP} combination obtains the address
7103 of the label \c{.get_GOT}, without having to know in advance where
7104 the program was loaded (since the \c{CALL} instruction is encoded
7105 relative to the current position). The \c{ADD} instruction makes use
7106 of one of the special PIC relocation types: \i{GOTPC relocation}.
7107 With the \i\c{WRT ..gotpc} qualifier specified, the symbol
7108 referenced (here \c{_GLOBAL_OFFSET_TABLE_}, the special symbol
7109 assigned to the GOT) is given as an offset from the beginning of the
7110 section. (Actually, \c{ELF} encodes it as the offset from the operand
7111 field of the \c{ADD} instruction, but NASM simplifies this
7112 deliberately, so you do things the same way for both \c{ELF} and
7113 \c{BSD}.) So the instruction then \e{adds} the beginning of the section,
7114 to get the real address of the GOT, and subtracts the value of
7115 \c{.get_GOT} which it knows is in \c{EBX}. Therefore, by the time
7116 that instruction has finished, \c{EBX} contains the address of the GOT.
7118 If you didn't follow that, don't worry: it's never necessary to
7119 obtain the address of the GOT by any other means, so you can put
7120 those three instructions into a macro and safely ignore them:
7127 \c add ebx,_GLOBAL_OFFSET_TABLE_+$$-%%getgot wrt ..gotpc
7131 \S{piclocal} Finding Your Local Data Items
7133 Having got the GOT, you can then use it to obtain the addresses of
7134 your data items. Most variables will reside in the sections you have
7135 declared; they can be accessed using the \I{GOTOFF
7136 relocation}\c{..gotoff} special \I\c{WRT ..gotoff}\c{WRT} type. The
7137 way this works is like this:
7139 \c lea eax,[ebx+myvar wrt ..gotoff]
7141 The expression \c{myvar wrt ..gotoff} is calculated, when the shared
7142 library is linked, to be the offset to the local variable \c{myvar}
7143 from the beginning of the GOT. Therefore, adding it to \c{EBX} as
7144 above will place the real address of \c{myvar} in \c{EAX}.
7146 If you declare variables as \c{GLOBAL} without specifying a size for
7147 them, they are shared between code modules in the library, but do
7148 not get exported from the library to the program that loaded it.
7149 They will still be in your ordinary data and BSS sections, so you
7150 can access them in the same way as local variables, using the above
7151 \c{..gotoff} mechanism.
7153 Note that due to a peculiarity of the way BSD \c{a.out} format
7154 handles this relocation type, there must be at least one non-local
7155 symbol in the same section as the address you're trying to access.
7158 \S{picextern} Finding External and Common Data Items
7160 If your library needs to get at an external variable (external to
7161 the \e{library}, not just to one of the modules within it), you must
7162 use the \I{GOT relocations}\I\c{WRT ..got}\c{..got} type to get at
7163 it. The \c{..got} type, instead of giving you the offset from the
7164 GOT base to the variable, gives you the offset from the GOT base to
7165 a GOT \e{entry} containing the address of the variable. The linker
7166 will set up this GOT entry when it builds the library, and the
7167 dynamic linker will place the correct address in it at load time. So
7168 to obtain the address of an external variable \c{extvar} in \c{EAX},
7171 \c mov eax,[ebx+extvar wrt ..got]
7173 This loads the address of \c{extvar} out of an entry in the GOT. The
7174 linker, when it builds the shared library, collects together every
7175 relocation of type \c{..got}, and builds the GOT so as to ensure it
7176 has every necessary entry present.
7178 Common variables must also be accessed in this way.
7181 \S{picglobal} Exporting Symbols to the Library User
7183 If you want to export symbols to the user of the library, you have
7184 to declare whether they are functions or data, and if they are data,
7185 you have to give the size of the data item. This is because the
7186 dynamic linker has to build \I{PLT}\i{procedure linkage table}
7187 entries for any exported functions, and also moves exported data
7188 items away from the library's data section in which they were
7191 So to export a function to users of the library, you must use
7193 \c global func:function ; declare it as a function
7199 And to export a data item such as an array, you would have to code
7201 \c global array:data array.end-array ; give the size too
7206 Be careful: If you export a variable to the library user, by
7207 declaring it as \c{GLOBAL} and supplying a size, the variable will
7208 end up living in the data section of the main program, rather than
7209 in your library's data section, where you declared it. So you will
7210 have to access your own global variable with the \c{..got} mechanism
7211 rather than \c{..gotoff}, as if it were external (which,
7212 effectively, it has become).
7214 Equally, if you need to store the address of an exported global in
7215 one of your data sections, you can't do it by means of the standard
7218 \c dataptr: dd global_data_item ; WRONG
7220 NASM will interpret this code as an ordinary relocation, in which
7221 \c{global_data_item} is merely an offset from the beginning of the
7222 \c{.data} section (or whatever); so this reference will end up
7223 pointing at your data section instead of at the exported global
7224 which resides elsewhere.
7226 Instead of the above code, then, you must write
7228 \c dataptr: dd global_data_item wrt ..sym
7230 which makes use of the special \c{WRT} type \I\c{WRT ..sym}\c{..sym}
7231 to instruct NASM to search the symbol table for a particular symbol
7232 at that address, rather than just relocating by section base.
7234 Either method will work for functions: referring to one of your
7235 functions by means of
7237 \c funcptr: dd my_function
7239 will give the user the address of the code you wrote, whereas
7241 \c funcptr: dd my_function wrt .sym
7243 will give the address of the procedure linkage table for the
7244 function, which is where the calling program will \e{believe} the
7245 function lives. Either address is a valid way to call the function.
7248 \S{picproc} Calling Procedures Outside the Library
7250 Calling procedures outside your shared library has to be done by
7251 means of a \i\e{procedure linkage table}, or \i{PLT}. The PLT is
7252 placed at a known offset from where the library is loaded, so the
7253 library code can make calls to the PLT in a position-independent
7254 way. Within the PLT there is code to jump to offsets contained in
7255 the GOT, so function calls to other shared libraries or to routines
7256 in the main program can be transparently passed off to their real
7259 To call an external routine, you must use another special PIC
7260 relocation type, \I{PLT relocations}\i\c{WRT ..plt}. This is much
7261 easier than the GOT-based ones: you simply replace calls such as
7262 \c{CALL printf} with the PLT-relative version \c{CALL printf WRT
7266 \S{link} Generating the Library File
7268 Having written some code modules and assembled them to \c{.o} files,
7269 you then generate your shared library with a command such as
7271 \c ld -shared -o library.so module1.o module2.o # for ELF
7272 \c ld -Bshareable -o library.so module1.o module2.o # for BSD
7274 For ELF, if your shared library is going to reside in system
7275 directories such as \c{/usr/lib} or \c{/lib}, it is usually worth
7276 using the \i\c{-soname} flag to the linker, to store the final
7277 library file name, with a version number, into the library:
7279 \c ld -shared -soname library.so.1 -o library.so.1.2 *.o
7281 You would then copy \c{library.so.1.2} into the library directory,
7282 and create \c{library.so.1} as a symbolic link to it.
7285 \C{mixsize} Mixing 16 and 32 Bit Code
7287 This chapter tries to cover some of the issues, largely related to
7288 unusual forms of addressing and jump instructions, encountered when
7289 writing operating system code such as protected-mode initialisation
7290 routines, which require code that operates in mixed segment sizes,
7291 such as code in a 16-bit segment trying to modify data in a 32-bit
7292 one, or jumps between different-size segments.
7295 \H{mixjump} Mixed-Size Jumps\I{jumps, mixed-size}
7297 \I{operating system, writing}\I{writing operating systems}The most
7298 common form of \i{mixed-size instruction} is the one used when
7299 writing a 32-bit OS: having done your setup in 16-bit mode, such as
7300 loading the kernel, you then have to boot it by switching into
7301 protected mode and jumping to the 32-bit kernel start address. In a
7302 fully 32-bit OS, this tends to be the \e{only} mixed-size
7303 instruction you need, since everything before it can be done in pure
7304 16-bit code, and everything after it can be pure 32-bit.
7306 This jump must specify a 48-bit far address, since the target
7307 segment is a 32-bit one. However, it must be assembled in a 16-bit
7308 segment, so just coding, for example,
7310 \c jmp 0x1234:0x56789ABC ; wrong!
7312 will not work, since the offset part of the address will be
7313 truncated to \c{0x9ABC} and the jump will be an ordinary 16-bit far
7316 The Linux kernel setup code gets round the inability of \c{as86} to
7317 generate the required instruction by coding it manually, using
7318 \c{DB} instructions. NASM can go one better than that, by actually
7319 generating the right instruction itself. Here's how to do it right:
7321 \c jmp dword 0x1234:0x56789ABC ; right
7323 \I\c{JMP DWORD}The \c{DWORD} prefix (strictly speaking, it should
7324 come \e{after} the colon, since it is declaring the \e{offset} field
7325 to be a doubleword; but NASM will accept either form, since both are
7326 unambiguous) forces the offset part to be treated as far, in the
7327 assumption that you are deliberately writing a jump from a 16-bit
7328 segment to a 32-bit one.
7330 You can do the reverse operation, jumping from a 32-bit segment to a
7331 16-bit one, by means of the \c{WORD} prefix:
7333 \c jmp word 0x8765:0x4321 ; 32 to 16 bit
7335 If the \c{WORD} prefix is specified in 16-bit mode, or the \c{DWORD}
7336 prefix in 32-bit mode, they will be ignored, since each is
7337 explicitly forcing NASM into a mode it was in anyway.
7340 \H{mixaddr} Addressing Between Different-Size Segments\I{addressing,
7341 mixed-size}\I{mixed-size addressing}
7343 If your OS is mixed 16 and 32-bit, or if you are writing a DOS
7344 extender, you are likely to have to deal with some 16-bit segments
7345 and some 32-bit ones. At some point, you will probably end up
7346 writing code in a 16-bit segment which has to access data in a
7347 32-bit segment, or vice versa.
7349 If the data you are trying to access in a 32-bit segment lies within
7350 the first 64K of the segment, you may be able to get away with using
7351 an ordinary 16-bit addressing operation for the purpose; but sooner
7352 or later, you will want to do 32-bit addressing from 16-bit mode.
7354 The easiest way to do this is to make sure you use a register for
7355 the address, since any effective address containing a 32-bit
7356 register is forced to be a 32-bit address. So you can do
7358 \c mov eax,offset_into_32_bit_segment_specified_by_fs
7359 \c mov dword [fs:eax],0x11223344
7361 This is fine, but slightly cumbersome (since it wastes an
7362 instruction and a register) if you already know the precise offset
7363 you are aiming at. The x86 architecture does allow 32-bit effective
7364 addresses to specify nothing but a 4-byte offset, so why shouldn't
7365 NASM be able to generate the best instruction for the purpose?
7367 It can. As in \k{mixjump}, you need only prefix the address with the
7368 \c{DWORD} keyword, and it will be forced to be a 32-bit address:
7370 \c mov dword [fs:dword my_offset],0x11223344
7372 Also as in \k{mixjump}, NASM is not fussy about whether the
7373 \c{DWORD} prefix comes before or after the segment override, so
7374 arguably a nicer-looking way to code the above instruction is
7376 \c mov dword [dword fs:my_offset],0x11223344
7378 Don't confuse the \c{DWORD} prefix \e{outside} the square brackets,
7379 which controls the size of the data stored at the address, with the
7380 one \c{inside} the square brackets which controls the length of the
7381 address itself. The two can quite easily be different:
7383 \c mov word [dword 0x12345678],0x9ABC
7385 This moves 16 bits of data to an address specified by a 32-bit
7388 You can also specify \c{WORD} or \c{DWORD} prefixes along with the
7389 \c{FAR} prefix to indirect far jumps or calls. For example:
7391 \c call dword far [fs:word 0x4321]
7393 This instruction contains an address specified by a 16-bit offset;
7394 it loads a 48-bit far pointer from that (16-bit segment and 32-bit
7395 offset), and calls that address.
7398 \H{mixother} Other Mixed-Size Instructions
7400 The other way you might want to access data might be using the
7401 string instructions (\c{LODSx}, \c{STOSx} and so on) or the
7402 \c{XLATB} instruction. These instructions, since they take no
7403 parameters, might seem to have no easy way to make them perform
7404 32-bit addressing when assembled in a 16-bit segment.
7406 This is the purpose of NASM's \i\c{a16}, \i\c{a32} and \i\c{a64} prefixes. If
7407 you are coding \c{LODSB} in a 16-bit segment but it is supposed to
7408 be accessing a string in a 32-bit segment, you should load the
7409 desired address into \c{ESI} and then code
7413 The prefix forces the addressing size to 32 bits, meaning that
7414 \c{LODSB} loads from \c{[DS:ESI]} instead of \c{[DS:SI]}. To access
7415 a string in a 16-bit segment when coding in a 32-bit one, the
7416 corresponding \c{a16} prefix can be used.
7418 The \c{a16}, \c{a32} and \c{a64} prefixes can be applied to any instruction
7419 in NASM's instruction table, but most of them can generate all the
7420 useful forms without them. The prefixes are necessary only for
7421 instructions with implicit addressing:
7422 \# \c{CMPSx} (\k{insCMPSB}),
7423 \# \c{SCASx} (\k{insSCASB}), \c{LODSx} (\k{insLODSB}), \c{STOSx}
7424 \# (\k{insSTOSB}), \c{MOVSx} (\k{insMOVSB}), \c{INSx} (\k{insINSB}),
7425 \# \c{OUTSx} (\k{insOUTSB}), and \c{XLATB} (\k{insXLATB}).
7426 \c{CMPSx}, \c{SCASx}, \c{LODSx}, \c{STOSx}, \c{MOVSx}, \c{INSx},
7427 \c{OUTSx}, and \c{XLATB}.
7429 various push and pop instructions (\c{PUSHA} and \c{POPF} as well as
7430 the more usual \c{PUSH} and \c{POP}) can accept \c{a16}, \c{a32} or \c{a64}
7431 prefixes to force a particular one of \c{SP}, \c{ESP} or \c{RSP} to be used
7432 as a stack pointer, in case the stack segment in use is a different
7433 size from the code segment.
7435 \c{PUSH} and \c{POP}, when applied to segment registers in 32-bit
7436 mode, also have the slightly odd behaviour that they push and pop 4
7437 bytes at a time, of which the top two are ignored and the bottom two
7438 give the value of the segment register being manipulated. To force
7439 the 16-bit behaviour of segment-register push and pop instructions,
7440 you can use the operand-size prefix \i\c{o16}:
7445 This code saves a doubleword of stack space by fitting two segment
7446 registers into the space which would normally be consumed by pushing
7449 (You can also use the \i\c{o32} prefix to force the 32-bit behaviour
7450 when in 16-bit mode, but this seems less useful.)
7453 \C{64bit} Writing 64-bit Code (Unix, Win64)
7455 This chapter attempts to cover some of the common issues involved when
7456 writing 64-bit code, to run under \i{Win64} or Unix. It covers how to
7457 write assembly code to interface with 64-bit C routines, and how to
7458 write position-independent code for shared libraries.
7460 All 64-bit code uses a flat memory model, since segmentation is not
7461 available in 64-bit mode. The one exception is the \c{FS} and \c{GS}
7462 registers, which still add their bases.
7464 Position independence in 64-bit mode is significantly simpler, since
7465 the processor supports \c{RIP}-relative addressing directly; see the
7466 \c{REL} keyword (\k{effaddr}). On most 64-bit platforms, it is
7467 probably desirable to make that the default, using the directive
7468 \c{DEFAULT REL} (\k{default}).
7470 64-bit programming is relatively similar to 32-bit programming, but
7471 of course pointers are 64 bits long; additionally, all existing
7472 platforms pass arguments in registers rather than on the stack.
7473 Furthermore, 64-bit platforms use SSE2 by default for floating point.
7474 Please see the ABI documentation for your platform.
7476 64-bit platforms differ in the sizes of the fundamental datatypes, not
7477 just from 32-bit platforms but from each other. If a specific size
7478 data type is desired, it is probably best to use the types defined in
7479 the Standard C header \c{<inttypes.h>}.
7481 In 64-bit mode, the default instruction size is still 32 bits. When
7482 loading a value into a 32-bit register (but not an 8- or 16-bit
7483 register), the upper 32 bits of the corresponding 64-bit register are
7486 \H{reg64} Register Names in 64-bit Mode
7488 NASM uses the following names for general-purpose registers in 64-bit
7489 mode, for 8-, 16-, 32- and 64-bit references, respecitively:
7491 \c AL/AH, CL/CH, DL/DH, BL/BH, SPL, BPL, SIL, DIL, R8B-R15B
7492 \c AX, CX, DX, BX, SP, BP, SI, DI, R8W-R15W
7493 \c EAX, ECX, EDX, EBX, ESP, EBP, ESI, EDI, R8D-R15D
7494 \c RAX, RCX, RDX, RBX, RSP, RBP, RSI, RDI, R8-R15
7496 This is consistent with the AMD documentation and most other
7497 assemblers. The Intel documentation, however, uses the names
7498 \c{R8L-R15L} for 8-bit references to the higher registers. It is
7499 possible to use those names by definiting them as macros; similarly,
7500 if one wants to use numeric names for the low 8 registers, define them
7501 as macros. The standard macro package \c{altreg} (see \k{pkg_altreg})
7502 can be used for this purpose.
7504 \H{id64} Immediates and Displacements in 64-bit Mode
7506 In 64-bit mode, immediates and displacements are generally only 32
7507 bits wide. NASM will therefore truncate most displacements and
7508 immediates to 32 bits.
7510 The only instruction which takes a full \i{64-bit immediate} is:
7514 NASM will produce this instruction whenever the programmer uses
7515 \c{MOV} with an immediate into a 64-bit register. If this is not
7516 desirable, simply specify the equivalent 32-bit register, which will
7517 be automatically zero-extended by the processor, or specify the
7518 immediate as \c{DWORD}:
7520 \c mov rax,foo ; 64-bit immediate
7521 \c mov rax,qword foo ; (identical)
7522 \c mov eax,foo ; 32-bit immediate, zero-extended
7523 \c mov rax,dword foo ; 32-bit immediate, sign-extended
7525 The length of these instructions are 10, 5 and 7 bytes, respectively.
7527 The only instructions which take a full \I{64-bit displacement}64-bit
7528 \e{displacement} is loading or storing, using \c{MOV}, \c{AL}, \c{AX},
7529 \c{EAX} or \c{RAX} (but no other registers) to an absolute 64-bit address.
7530 Since this is a relatively rarely used instruction (64-bit code generally uses
7531 relative addressing), the programmer has to explicitly declare the
7532 displacement size as \c{QWORD}:
7536 \c mov eax,[foo] ; 32-bit absolute disp, sign-extended
7537 \c mov eax,[a32 foo] ; 32-bit absolute disp, zero-extended
7538 \c mov eax,[qword foo] ; 64-bit absolute disp
7542 \c mov eax,[foo] ; 32-bit relative disp
7543 \c mov eax,[a32 foo] ; d:o, address truncated to 32 bits(!)
7544 \c mov eax,[qword foo] ; error
7545 \c mov eax,[abs qword foo] ; 64-bit absolute disp
7547 A sign-extended absolute displacement can access from -2 GB to +2 GB;
7548 a zero-extended absolute displacement can access from 0 to 4 GB.
7550 \H{unix64} Interfacing to 64-bit C Programs (Unix)
7552 On Unix, the 64-bit ABI is defined by the document:
7554 \W{http://www.nasm.us/links/unix64abi}\c{http://www.nasm.us/links/unix64abi}
7556 Although written for AT&T-syntax assembly, the concepts apply equally
7557 well for NASM-style assembly. What follows is a simplified summary.
7559 The first six integer arguments (from the left) are passed in \c{RDI},
7560 \c{RSI}, \c{RDX}, \c{RCX}, \c{R8}, and \c{R9}, in that order.
7561 Additional integer arguments are passed on the stack. These
7562 registers, plus \c{RAX}, \c{R10} and \c{R11} are destroyed by function
7563 calls, and thus are available for use by the function without saving.
7565 Integer return values are passed in \c{RAX} and \c{RDX}, in that order.
7567 Floating point is done using SSE registers, except for \c{long
7568 double}. Floating-point arguments are passed in \c{XMM0} to \c{XMM7};
7569 return is \c{XMM0} and \c{XMM1}. \c{long double} are passed on the
7570 stack, and returned in \c{ST0} and \c{ST1}.
7572 All SSE and x87 registers are destroyed by function calls.
7574 On 64-bit Unix, \c{long} is 64 bits.
7576 Integer and SSE register arguments are counted separately, so for the case of
7578 \c void foo(long a, double b, int c)
7580 \c{a} is passed in \c{RDI}, \c{b} in \c{XMM0}, and \c{c} in \c{ESI}.
7582 \H{win64} Interfacing to 64-bit C Programs (Win64)
7584 The Win64 ABI is described at:
7586 \W{http://www.nasm.us/links/win64abi}\c{http://www.nasm.us/links/win64abi}
7588 What follows is a simplified summary.
7590 The first four integer arguments are passed in \c{RCX}, \c{RDX},
7591 \c{R8} and \c{R9}, in that order. Additional integer arguments are
7592 passed on the stack. These registers, plus \c{RAX}, \c{R10} and
7593 \c{R11} are destroyed by function calls, and thus are available for
7594 use by the function without saving.
7596 Integer return values are passed in \c{RAX} only.
7598 Floating point is done using SSE registers, except for \c{long
7599 double}. Floating-point arguments are passed in \c{XMM0} to \c{XMM3};
7600 return is \c{XMM0} only.
7602 On Win64, \c{long} is 32 bits; \c{long long} or \c{_int64} is 64 bits.
7604 Integer and SSE register arguments are counted together, so for the case of
7606 \c void foo(long long a, double b, int c)
7608 \c{a} is passed in \c{RCX}, \c{b} in \c{XMM1}, and \c{c} in \c{R8D}.
7610 \C{trouble} Troubleshooting
7612 This chapter describes some of the common problems that users have
7613 been known to encounter with NASM, and answers them. It also gives
7614 instructions for reporting bugs in NASM if you find a difficulty
7615 that isn't listed here.
7618 \H{problems} Common Problems
7620 \S{inefficient} NASM Generates \i{Inefficient Code}
7622 We sometimes get `bug' reports about NASM generating inefficient, or
7623 even `wrong', code on instructions such as \c{ADD ESP,8}. This is a
7624 deliberate design feature, connected to predictability of output:
7625 NASM, on seeing \c{ADD ESP,8}, will generate the form of the
7626 instruction which leaves room for a 32-bit offset. You need to code
7627 \I\c{BYTE}\c{ADD ESP,BYTE 8} if you want the space-efficient form of
7628 the instruction. This isn't a bug, it's user error: if you prefer to
7629 have NASM produce the more efficient code automatically enable
7630 optimization with the \c{-O} option (see \k{opt-O}).
7633 \S{jmprange} My Jumps are Out of Range\I{out of range, jumps}
7635 Similarly, people complain that when they issue \i{conditional
7636 jumps} (which are \c{SHORT} by default) that try to jump too far,
7637 NASM reports `short jump out of range' instead of making the jumps
7640 This, again, is partly a predictability issue, but in fact has a
7641 more practical reason as well. NASM has no means of being told what
7642 type of processor the code it is generating will be run on; so it
7643 cannot decide for itself that it should generate \i\c{Jcc NEAR} type
7644 instructions, because it doesn't know that it's working for a 386 or
7645 above. Alternatively, it could replace the out-of-range short
7646 \c{JNE} instruction with a very short \c{JE} instruction that jumps
7647 over a \c{JMP NEAR}; this is a sensible solution for processors
7648 below a 386, but hardly efficient on processors which have good
7649 branch prediction \e{and} could have used \c{JNE NEAR} instead. So,
7650 once again, it's up to the user, not the assembler, to decide what
7651 instructions should be generated. See \k{opt-O}.
7654 \S{proborg} \i\c{ORG} Doesn't Work
7656 People writing \i{boot sector} programs in the \c{bin} format often
7657 complain that \c{ORG} doesn't work the way they'd like: in order to
7658 place the \c{0xAA55} signature word at the end of a 512-byte boot
7659 sector, people who are used to MASM tend to code
7663 \c ; some boot sector code
7668 This is not the intended use of the \c{ORG} directive in NASM, and
7669 will not work. The correct way to solve this problem in NASM is to
7670 use the \i\c{TIMES} directive, like this:
7674 \c ; some boot sector code
7676 \c TIMES 510-($-$$) DB 0
7679 The \c{TIMES} directive will insert exactly enough zero bytes into
7680 the output to move the assembly point up to 510. This method also
7681 has the advantage that if you accidentally fill your boot sector too
7682 full, NASM will catch the problem at assembly time and report it, so
7683 you won't end up with a boot sector that you have to disassemble to
7684 find out what's wrong with it.
7687 \S{probtimes} \i\c{TIMES} Doesn't Work
7689 The other common problem with the above code is people who write the
7694 by reasoning that \c{$} should be a pure number, just like 510, so
7695 the difference between them is also a pure number and can happily be
7698 NASM is a \e{modular} assembler: the various component parts are
7699 designed to be easily separable for re-use, so they don't exchange
7700 information unnecessarily. In consequence, the \c{bin} output
7701 format, even though it has been told by the \c{ORG} directive that
7702 the \c{.text} section should start at 0, does not pass that
7703 information back to the expression evaluator. So from the
7704 evaluator's point of view, \c{$} isn't a pure number: it's an offset
7705 from a section base. Therefore the difference between \c{$} and 510
7706 is also not a pure number, but involves a section base. Values
7707 involving section bases cannot be passed as arguments to \c{TIMES}.
7709 The solution, as in the previous section, is to code the \c{TIMES}
7712 \c TIMES 510-($-$$) DB 0
7714 in which \c{$} and \c{$$} are offsets from the same section base,
7715 and so their difference is a pure number. This will solve the
7716 problem and generate sensible code.
7719 \H{bugs} \i{Bugs}\I{reporting bugs}
7721 We have never yet released a version of NASM with any \e{known}
7722 bugs. That doesn't usually stop there being plenty we didn't know
7723 about, though. Any that you find should be reported firstly via the
7725 \W{http://www.nasm.us/}\c{http://www.nasm.us/}
7726 (click on "Bug Tracker"), or if that fails then through one of the
7727 contacts in \k{contact}.
7729 Please read \k{qstart} first, and don't report the bug if it's
7730 listed in there as a deliberate feature. (If you think the feature
7731 is badly thought out, feel free to send us reasons why you think it
7732 should be changed, but don't just send us mail saying `This is a
7733 bug' if the documentation says we did it on purpose.) Then read
7734 \k{problems}, and don't bother reporting the bug if it's listed
7737 If you do report a bug, \e{please} give us all of the following
7740 \b What operating system you're running NASM under. DOS, Linux,
7741 NetBSD, Win16, Win32, VMS (I'd be impressed), whatever.
7743 \b If you're running NASM under DOS or Win32, tell us whether you've
7744 compiled your own executable from the DOS source archive, or whether
7745 you were using the standard distribution binaries out of the
7746 archive. If you were using a locally built executable, try to
7747 reproduce the problem using one of the standard binaries, as this
7748 will make it easier for us to reproduce your problem prior to fixing
7751 \b Which version of NASM you're using, and exactly how you invoked
7752 it. Give us the precise command line, and the contents of the
7753 \c{NASMENV} environment variable if any.
7755 \b Which versions of any supplementary programs you're using, and
7756 how you invoked them. If the problem only becomes visible at link
7757 time, tell us what linker you're using, what version of it you've
7758 got, and the exact linker command line. If the problem involves
7759 linking against object files generated by a compiler, tell us what
7760 compiler, what version, and what command line or options you used.
7761 (If you're compiling in an IDE, please try to reproduce the problem
7762 with the command-line version of the compiler.)
7764 \b If at all possible, send us a NASM source file which exhibits the
7765 problem. If this causes copyright problems (e.g. you can only
7766 reproduce the bug in restricted-distribution code) then bear in mind
7767 the following two points: firstly, we guarantee that any source code
7768 sent to us for the purposes of debugging NASM will be used \e{only}
7769 for the purposes of debugging NASM, and that we will delete all our
7770 copies of it as soon as we have found and fixed the bug or bugs in
7771 question; and secondly, we would prefer \e{not} to be mailed large
7772 chunks of code anyway. The smaller the file, the better. A
7773 three-line sample file that does nothing useful \e{except}
7774 demonstrate the problem is much easier to work with than a
7775 fully fledged ten-thousand-line program. (Of course, some errors
7776 \e{do} only crop up in large files, so this may not be possible.)
7778 \b A description of what the problem actually \e{is}. `It doesn't
7779 work' is \e{not} a helpful description! Please describe exactly what
7780 is happening that shouldn't be, or what isn't happening that should.
7781 Examples might be: `NASM generates an error message saying Line 3
7782 for an error that's actually on Line 5'; `NASM generates an error
7783 message that I believe it shouldn't be generating at all'; `NASM
7784 fails to generate an error message that I believe it \e{should} be
7785 generating'; `the object file produced from this source code crashes
7786 my linker'; `the ninth byte of the output file is 66 and I think it
7787 should be 77 instead'.
7789 \b If you believe the output file from NASM to be faulty, send it to
7790 us. That allows us to determine whether our own copy of NASM
7791 generates the same file, or whether the problem is related to
7792 portability issues between our development platforms and yours. We
7793 can handle binary files mailed to us as MIME attachments, uuencoded,
7794 and even BinHex. Alternatively, we may be able to provide an FTP
7795 site you can upload the suspect files to; but mailing them is easier
7798 \b Any other information or data files that might be helpful. If,
7799 for example, the problem involves NASM failing to generate an object
7800 file while TASM can generate an equivalent file without trouble,
7801 then send us \e{both} object files, so we can see what TASM is doing
7802 differently from us.
7805 \A{ndisasm} \i{Ndisasm}
7807 The Netwide Disassembler, NDISASM
7809 \H{ndisintro} Introduction
7812 The Netwide Disassembler is a small companion program to the Netwide
7813 Assembler, NASM. It seemed a shame to have an x86 assembler,
7814 complete with a full instruction table, and not make as much use of
7815 it as possible, so here's a disassembler which shares the
7816 instruction table (and some other bits of code) with NASM.
7818 The Netwide Disassembler does nothing except to produce
7819 disassemblies of \e{binary} source files. NDISASM does not have any
7820 understanding of object file formats, like \c{objdump}, and it will
7821 not understand \c{DOS .EXE} files like \c{debug} will. It just
7825 \H{ndisstart} Getting Started: Installation
7827 See \k{install} for installation instructions. NDISASM, like NASM,
7828 has a \c{man page} which you may want to put somewhere useful, if you
7829 are on a Unix system.
7832 \H{ndisrun} Running NDISASM
7834 To disassemble a file, you will typically use a command of the form
7836 \c ndisasm -b {16|32|64} filename
7838 NDISASM can disassemble 16-, 32- or 64-bit code equally easily,
7839 provided of course that you remember to specify which it is to work
7840 with. If no \i\c{-b} switch is present, NDISASM works in 16-bit mode
7841 by default. The \i\c{-u} switch (for USE32) also invokes 32-bit mode.
7843 Two more command line options are \i\c{-r} which reports the version
7844 number of NDISASM you are running, and \i\c{-h} which gives a short
7845 summary of command line options.
7848 \S{ndiscom} COM Files: Specifying an Origin
7850 To disassemble a \c{DOS .COM} file correctly, a disassembler must assume
7851 that the first instruction in the file is loaded at address \c{0x100},
7852 rather than at zero. NDISASM, which assumes by default that any file
7853 you give it is loaded at zero, will therefore need to be informed of
7856 The \i\c{-o} option allows you to declare a different origin for the
7857 file you are disassembling. Its argument may be expressed in any of
7858 the NASM numeric formats: decimal by default, if it begins with `\c{$}'
7859 or `\c{0x}' or ends in `\c{H}' it's \c{hex}, if it ends in `\c{Q}' it's
7860 \c{octal}, and if it ends in `\c{B}' it's \c{binary}.
7862 Hence, to disassemble a \c{.COM} file:
7864 \c ndisasm -o100h filename.com
7869 \S{ndissync} Code Following Data: Synchronisation
7871 Suppose you are disassembling a file which contains some data which
7872 isn't machine code, and \e{then} contains some machine code. NDISASM
7873 will faithfully plough through the data section, producing machine
7874 instructions wherever it can (although most of them will look
7875 bizarre, and some may have unusual prefixes, e.g. `\c{FS OR AX,0x240A}'),
7876 and generating `DB' instructions ever so often if it's totally stumped.
7877 Then it will reach the code section.
7879 Supposing NDISASM has just finished generating a strange machine
7880 instruction from part of the data section, and its file position is
7881 now one byte \e{before} the beginning of the code section. It's
7882 entirely possible that another spurious instruction will get
7883 generated, starting with the final byte of the data section, and
7884 then the correct first instruction in the code section will not be
7885 seen because the starting point skipped over it. This isn't really
7888 To avoid this, you can specify a `\i\c{synchronisation}' point, or indeed
7889 as many synchronisation points as you like (although NDISASM can
7890 only handle 2147483647 sync points internally). The definition of a sync
7891 point is this: NDISASM guarantees to hit sync points exactly during
7892 disassembly. If it is thinking about generating an instruction which
7893 would cause it to jump over a sync point, it will discard that
7894 instruction and output a `\c{db}' instead. So it \e{will} start
7895 disassembly exactly from the sync point, and so you \e{will} see all
7896 the instructions in your code section.
7898 Sync points are specified using the \i\c{-s} option: they are measured
7899 in terms of the program origin, not the file position. So if you
7900 want to synchronize after 32 bytes of a \c{.COM} file, you would have to
7903 \c ndisasm -o100h -s120h file.com
7907 \c ndisasm -o100h -s20h file.com
7909 As stated above, you can specify multiple sync markers if you need
7910 to, just by repeating the \c{-s} option.
7913 \S{ndisisync} Mixed Code and Data: Automatic (Intelligent) Synchronisation
7916 Suppose you are disassembling the boot sector of a \c{DOS} floppy (maybe
7917 it has a virus, and you need to understand the virus so that you
7918 know what kinds of damage it might have done you). Typically, this
7919 will contain a \c{JMP} instruction, then some data, then the rest of the
7920 code. So there is a very good chance of NDISASM being \e{misaligned}
7921 when the data ends and the code begins. Hence a sync point is
7924 On the other hand, why should you have to specify the sync point
7925 manually? What you'd do in order to find where the sync point would
7926 be, surely, would be to read the \c{JMP} instruction, and then to use
7927 its target address as a sync point. So can NDISASM do that for you?
7929 The answer, of course, is yes: using either of the synonymous
7930 switches \i\c{-a} (for automatic sync) or \i\c{-i} (for intelligent
7931 sync) will enable \c{auto-sync} mode. Auto-sync mode automatically
7932 generates a sync point for any forward-referring PC-relative jump or
7933 call instruction that NDISASM encounters. (Since NDISASM is one-pass,
7934 if it encounters a PC-relative jump whose target has already been
7935 processed, there isn't much it can do about it...)
7937 Only PC-relative jumps are processed, since an absolute jump is
7938 either through a register (in which case NDISASM doesn't know what
7939 the register contains) or involves a segment address (in which case
7940 the target code isn't in the same segment that NDISASM is working
7941 in, and so the sync point can't be placed anywhere useful).
7943 For some kinds of file, this mechanism will automatically put sync
7944 points in all the right places, and save you from having to place
7945 any sync points manually. However, it should be stressed that
7946 auto-sync mode is \e{not} guaranteed to catch all the sync points, and
7947 you may still have to place some manually.
7949 Auto-sync mode doesn't prevent you from declaring manual sync
7950 points: it just adds automatically generated ones to the ones you
7951 provide. It's perfectly feasible to specify \c{-i} \e{and} some \c{-s}
7954 Another caveat with auto-sync mode is that if, by some unpleasant
7955 fluke, something in your data section should disassemble to a
7956 PC-relative call or jump instruction, NDISASM may obediently place a
7957 sync point in a totally random place, for example in the middle of
7958 one of the instructions in your code section. So you may end up with
7959 a wrong disassembly even if you use auto-sync. Again, there isn't
7960 much I can do about this. If you have problems, you'll have to use
7961 manual sync points, or use the \c{-k} option (documented below) to
7962 suppress disassembly of the data area.
7965 \S{ndisother} Other Options
7967 The \i\c{-e} option skips a header on the file, by ignoring the first N
7968 bytes. This means that the header is \e{not} counted towards the
7969 disassembly offset: if you give \c{-e10 -o10}, disassembly will start
7970 at byte 10 in the file, and this will be given offset 10, not 20.
7972 The \i\c{-k} option is provided with two comma-separated numeric
7973 arguments, the first of which is an assembly offset and the second
7974 is a number of bytes to skip. This \e{will} count the skipped bytes
7975 towards the assembly offset: its use is to suppress disassembly of a
7976 data section which wouldn't contain anything you wanted to see
7980 \H{ndisbugs} Bugs and Improvements
7982 There are no known bugs. However, any you find, with patches if
7983 possible, should be sent to
7984 \W{mailto:nasm-bugs@lists.sourceforge.net}\c{nasm-bugs@lists.sourceforge.net}, or to the
7986 \W{http://www.nasm.us/}\c{http://www.nasm.us/}
7987 and we'll try to fix them. Feel free to send contributions and
7988 new features as well.
7990 \A{inslist} \i{Instruction List}
7992 \H{inslistintro} Introduction
7994 The following sections show the instructions which NASM currently supports. For each
7995 instruction, there is a separate entry for each supported addressing mode. The third
7996 column shows the processor type in which the instruction was introduced and,
7997 when appropriate, one or more usage flags.
8001 \A{changelog} \i{NASM Version History}