1 .TH NASM 1 "The Netwide Assembler Project"
3 nasm \- the Netwide Assembler, a portable 80x86 assembler
22 command assembles the file
24 and directs output to the file
30 will derive a default output file name from the name of its input
31 file, usually by appending `.o' or `.obj', or by removing all
32 extensions for a raw binary file. Failing that, the output file name
39 to exit immediately, after giving a summary of its invocation
40 options, and listing all its supported output file formats.
45 to assemble the given input file without first applying the macro
51 to preprocess the given input file, and write the output to
53 (or the specified output file name), and not actually assemble
59 to output Makefile-style dependencies to stdout; normal output is
65 to redirect error messages to
67 This option exists to support operating systems on which stderr is not
73 to exit immediately, after displaying its version number.
79 to exit immediately, after displaying its version number.
82 Specifies the output file format. Formats include
84 to produce flat-form binary files, and
88 to produce Linux a.out and ELF object files, respectively.
91 Specifies a precise name for the output file, overriding
93 default means of determining it.
96 Causes an assembly listing to be directed to the given file, in
97 which the original source is displayed on the right hand side (plus
98 the source for included files and the expansions of multi-line
99 macros) and the generated code is shown in hex on the left.
104 to send its error messages and/or help text to
112 to enable or disable certain classes of warning messages, for
117 to, respectively, enable warnings about labels alone on lines or
118 disable warnings about incorrect numbers of parameters in macro
122 Adds a directory to the search path for include files. The directory
123 specification must include the trailing slash, as it will be
124 directly prepended to the name of the include file.
132 Specifies a file to be pre-included, before the main source file
133 starts to be processed.
140 .BI \-D " macro[=value]"
141 Pre-defines a single-line macro.
143 .BI \-d " macro[=value]"
149 Undefines a single-line macro.
158 This man page does not fully describe the syntax of
160 assembly language, but does give a summary of the differences from
164 have no leading `%' sign, unlike
166 and floating-point stack registers are referred to as
171 .I Floating-point instructions
172 may use either the single-operand form or the double. A
174 keyword is provided; thus, one could either write
182 or one could use the alternative single-operand forms
190 .I Uninitialised storage
191 is reserved using the
199 pseudo-opcodes, each taking one parameter which gives the number of
200 bytes, words, doublewords, quadwords or ten-byte words to reserve.
203 of data items is not done by the
205 keyword as seen in DOS assemblers, but by the use of the
211 message: times 3 db 'abc'
214 times 64-$+message db 0
216 which defines the string `abcabcabc', followed by the right number
217 of zero bytes to make the total length up to 64 bytes.
220 are always understood to be immediate (i.e. the address of the
221 symbol), unless square brackets are used, in which case the contents
222 of the memory location are used. Thus:
227 loads AX with the address of the variable `wordvar', whereas
236 mov ax,[es:wordvar+bx]
240 of memory locations. The syntaxes
243 mov ax,es:wordvar[bx]
248 are not legal at all, although the use of a segment register name as
249 an instruction prefix is valid, and can be used with instructions
252 which can't be overridden any other way.
255 may be expressed numerically in most formats: a trailing H, Q or B
256 denotes hex, octal or binary respectively, and a leading `0x' or `$'
257 denotes hex as well. Leading zeros are not treated specially at all.
258 Character constants may be enclosed in single or double quotes;
259 there is no escape character. The ordering is little-endian
260 (reversed), so that the character constant
262 denotes 0x64636261 and not 0x61626364.
265 begin with a period, and their `locality' is granted by the
266 assembler prepending the name of the previous non-local symbol. Thus
267 declaring a label `.loop' after a label `label' has actually defined
268 a symbol called `label.loop'.
275 to direct all following code to the named section. Section names
276 vary with output file format, although most formats support the
282 (The exception is the
284 format, in which all segments are user-definable.)
289 to position its notional assembly point at an absolute address: so
290 no code or data may be generated, but you can use
295 to move the assembly point further on, and you can define labels. So
296 this directive may be used to define data structures. When you have
297 finished doing absolute assembly, you must issue another
299 directive to return to normal assembly.
305 switches the default processor mode for which
307 is generating code: it is equivalent to
316 import and export symbol definitions, respectively, from and to
317 other modules. Note that the
319 directive must appear before the definition of the symbol it refers
325 when used to bracket a number of
328 or similar instructions, define a data structure. In addition to
329 defining the offsets of the structure members, the construct also
330 defines a symbol for the size of the structure, which is simply the
333 tacked on to the end.
334 .SS FORMAT-SPECIFIC DIRECTIVES
338 flat-form binary output format, and specifies the address at which
339 the output code will eventually be loaded.
341 .I GROUP grpname seg1 seg2...
344 (Microsoft 16-bit) output format, and defines segment groups. This
347 which directs that all segment, group and symbol names output to the
348 object file should be in uppercase. Note that the actual assembly is
349 still case sensitive.
354 output format, and causes a dependency record to be written to the
355 output file which indicates that the program requires a certain
356 library in order to run.
357 .SS MACRO PREPROCESSOR
358 Single-line macros are defined using the
362 commands, in a similar fashion to the C preprocessor. They can be
363 overloaded with respect to number of parameters, although defining a
364 macro with no parameters prevents the definition of any macro with
365 the same name taking parameters, and vice versa.
367 defines macros whose names match case-sensitively, whereas
369 defines case-insensitive macros.
371 Multi-line macros are defined using
375 (the distinction is the same as that between
379 whose syntax is as follows:
384 .IR minprm [- maxprm "][+][.nolist] [" defaults ]
387 <some lines of macro expansion text>
392 Again, these macros may be overloaded. The trailing plus sign
393 indicates that any parameters after the last one get subsumed, with
394 their separating commas, into the last parameter. The
396 part can be used to specify defaults for unspecified macro
400 is a valid synonym for
403 To refer to the macro parameters within a macro expansion, you use
406 and so on. You can also enforce that a macro parameter should
407 contain a condition code by using
409 and you can invert the condition code by using
411 You can also define a label specific to a macro invocation by
412 prefixing it with a double % sign.
414 Files can be included using the
416 directive, which works like C.
418 The preprocessor has a `context stack', which may be used by one
419 macro to store information that a later one will retrieve. You can
420 push a context on the stack using
424 and change the name of the top context (without disturbing any
425 associated definitions) using
429 macros specific to the top context may be defined by prefixing their
430 names with %$, and things specific to the next context down with
433 Conditional assembly is done by means of
439 as in C. (Except that
441 can accept several putative macro names, and will evaluate TRUE if
442 any of them is defined.) In addition, the directives
446 can be used to condition on the name of the top context on the
447 context stack. The obvious set of `else-if' directives,
455 There is a reported seg-fault on some (Linux) systems with some
456 large source files. This appears to be very hard to reproduce. All
459 bugs have been fixed...
461 There is no support for listing files, symbol maps, or debugging
462 object-file records. The advanced features of the ELF and Win32
463 object file formats are not supported, and there is no means for
464 warning the programmer against using an instruction beyond the
465 capability of the target processor.