2 \# NASM revision history in nasmdoc format
5 \H{cl-2.xx} NASM 2 Series
7 The NASM 2 series support x86-64, and is the production version of NASM
11 \S{cl-2.04} Version 2.04
13 \b Sanitize macro handing in the \c{%error} directive.
15 \b New \c{%warning} directive to issue user-controlled warnings.
17 \b \c{%error} directives are now deferred to the final assembly phase.
19 \b New \c{%fatal} directive to immediately terminate assembly.
21 \b New \c{%strcat} directive to join quoted strings together.
23 \b New \c{%use} macro directive to support standard macro directives. See
26 \b Excess default parameters to \c{%macro} now issues a warning by default.
29 \b Fix \c{%ifn} and \c{%elifn}.
31 \b Fix nested \c{%else} clauses.
33 \b Correct the handling of nested \c{%rep}s.
35 \b New \c{%unmacro} directive to undeclare a multi-line macro.
38 \b Builtin macro \c{__PASS__} which expands to the current assembly pass.
41 \b \c{__utf16__} and \c{__utf32__} operators to generate UTF-16 and UTF-32
42 strings. See \k{unicode}.
44 \b Fix bug in case-insensitive matching when compiled on platforms that
45 don't use the \c{configure} script. Of the official release binaries,
46 that only affected the OS/2 binary.
48 \b Support for x87 packed BCD constants. See \k{bcdconst}.
50 \b Correct the \c{LTR} and \c{SLDT} instructions in 64-bit mode.
52 \b Fix unnecessary REX.W prefix on indirect jumps in 64-bit mode.
54 \b Add AVX versions of the AES instructions (\c{VAES}...).
56 \b Fix the 256-bit FMA instructions.
58 \b Add 256-bit AVX stores per the latest AVX spec.
60 \b VIA XCRYPT instructions can now be written either with or without
61 \c{REP}, apparently different versions of the VIA spec wrote them
64 \b Add missing 64-bit \c{MOVNTI} instruction.
66 \b Fix the operand size of \c{VMREAD} and \c{VMWRITE}.
68 \b Numerous bug fixes, especially to the AES, AVX and VTX instructions.
70 \b The optimizer now always runs until it converges. It also runs even
71 when disabled, but doesn't optimize. This allows most forward references
72 to be resolved properly.
75 \S{cl-2.03.01} Version 2.03.01
77 \b Fix buffer overflow in the listing module.
79 \b Fix the handling of hexadecimal escape codes in `...` strings.
81 \b The Postscript/PDF documentation has been reformatted.
83 \b The \c{-F} option now implies \c{-g}.
86 \S{cl-2.03} Version 2.03
88 \b Add support for Intel AVX, CLMUL and FMA instructions,
89 including YMM registers.
91 \b \c{dy}, \c{resy} and \c{yword} for 32-byte operands.
93 \b Fix some SSE5 instructions.
95 \b Intel \c{INVEPT}, \c{INVVPID} and \c{MOVBE} instructions.
97 \b Fix checking for critical expressions when the optimizer is enabled.
99 \b Support the DWARF debugging format for ELF targets.
101 \b Fix optimizations of signed bytes.
103 \b Fix operation on bigendian machines.
105 \b Fix buffer overflow in the preprocessor.
107 \b \c{SAFESEH} support for Win32, \c{IMAGEREL} for Win64 (SEH).
109 \b \c{%?} and \c{%??} to refer to the name of a macro itself. In particular,
110 \c{%idefine keyword $%?} can be used to make a keyword "disappear".
112 \b New options for dependency generation: \c{-MD}, \c{-MF},
113 \c{-MP}, \c{-MT}, \c{-MQ}.
115 \b New preprocessor directives \c{%pathsearch} and \c{%depend}; INCBIN
116 reimplemented as a macro.
118 \b \c{%include} now resolves macros in a sane manner.
120 \b \c{%substr} can now be used to get other than one-character substrings.
122 \b New type of character/string constants, using backquotes (\c{`...`}),
123 which support C-style escape sequences.
125 \b \c{%defstr} and \c{%idefstr} to stringize macro definitions before
128 \b Fix forward references used in \c{EQU} statements.
130 \S{cl-2.02} Version 2.02
133 \b Additional fixes for MMX operands with explicit \c{qword}, as well as
134 (hopefully) SSE operands with \c{oword}.
136 \b Fix handling of truncated strings with \c{DO}.
138 \b Fix segfaults due to memory overwrites when floating-point constants
141 \b Fix segfaults due to missing include files.
143 \b Fix OpenWatcom Makefiles for DOS and OS/2.
145 \b Add autogenerated instruction list back into the documentation.
147 \b ELF: Fix segfault when generating stabs, and no symbols have been
150 \b ELF: Experimental support for DWARF debugging information.
152 \b New compile date and time standard macros.
154 \b \c{%ifnum} now returns true for negative numbers.
156 \b New \c{%iftoken} test for a single token.
158 \b New \c{%ifempty} test for empty expansion.
160 \b Add support for the \c{XSAVE} instruction group.
162 \b Makefile for Netware/gcc.
164 \b Fix issue with some warnings getting emitted way too many times.
166 \b Autogenerated instruction list added to the documentation.
168 \S{cl-2.01} Version 2.01
171 \b Fix the handling of MMX registers with explicit \c{qword} tags on
172 memory (broken in 2.00 due to 64-bit changes.)
174 \b Fix the PREFETCH instructions.
176 \b Fix the documentation.
178 \b Fix debugging info when using \c{-f elf}
179 (backwards compatibility alias for \c{-f elf32}).
181 \b Man pages for rdoff tools (from the Debian project.)
183 \b ELF: handle large numbers of sections.
185 \b Fix corrupt output when the optimizer runs out of passes.
188 \S{cl-2.00} Version 2.00
190 \b Added c99 data-type compliance.
192 \b Added general x86-64 support.
194 \b Added win64 (x86-64 COFF) output format.
196 \b Added \c{__BITS__} standard macro.
198 \b Renamed the \c{elf} output format to \c{elf32} for clarity.
200 \b Added \c{elf64} and \c{macho} (MacOS X) output formats.
202 \b Added Numeric constants in \c{dq} directive.
204 \b Added \c{oword}, \c{do} and \c{reso} pseudo operands.
206 \b Allow underscores in numbers.
208 \b Added 8-, 16- and 128-bit floating-point formats.
210 \b Added binary, octal and hexadecimal floating-point.
212 \b Correct the generation of floating-point constants.
214 \b Added floating-point option control.
216 \b Added Infinity and NaN floating point support.
218 \b Added ELF Symbol Visibility support.
220 \b Added setting OSABI value in ELF header directive.
222 \b Added Generate Makefile Dependencies option.
224 \b Added Unlimited Optimization Passes option.
226 \b Added \c{%IFN} and \c{%ELIFN} support.
228 \b Added Logical Negation Operator.
230 \b Enhanced Stack Relative Preprocessor Directives.
232 \b Enhanced ELF Debug Formats.
234 \b Enhanced Send Errors to a File option.
236 \b Added SSSE3, SSE4.1, SSE4.2, SSE5 support.
238 \b Added a large number of additional instructions.
240 \b Significant performance improvements.
243 \H{cl-0.98.xx} NASM 0.98 Series
245 The 0.98 series was the production versions of NASM from 1999 to 2007.
248 \S{cl-0.98.39} Version 0.98.39
250 \b fix buffer overflow
252 \b fix outas86's \c{.bss} handling
254 \b "make spotless" no longer deletes config.h.in.
256 \b \c{%(el)if(n)idn} insensitivity to string quotes difference (#809300).
258 \b (nasm.c)\c{__OUTPUT_FORMAT__} changed to string value instead of symbol.
260 \S{cl-0.98.38} Version 0.98.38
263 \b Add Makefile for 16-bit DOS binaries under OpenWatcom, and modify
264 \c{mkdep.pl} to be able to generate completely pathless dependencies, as
265 required by OpenWatcom wmake (it supports path searches, but not
268 \b Fix the \c{STR} instruction.
270 \b Fix the ELF output format, which was broken under certain
271 circumstances due to the addition of stabs support.
273 \b Quick-fix Borland format debug-info for \c{-f obj}
275 \b Fix for \c{%rep} with no arguments (#560568)
277 \b Fix concatenation of preprocessor function call (#794686)
279 \b Fix long label causes coredump (#677841)
281 \b Use autoheader as well as autoconf to keep configure from generating
282 ridiculously long command lines.
284 \b Make sure that all of the formats which support debugging output
285 actually will suppress debugging output when \c{-g} not specified.
287 \S{cl-0.98.37} Version 0.98.37
290 \b Paths given in \c{-I} switch searched for \c{incbin}-ed as
291 well as \c{%include}-ed files.
293 \b Added stabs debugging for the ELF output format, patch from
296 \b Fix \c{output/outbin.c} to allow origin > 80000000h.
298 \b Make \c{-U} switch work.
300 \b Fix the use of relative offsets with explicit prefixes, e.g.
303 \b Remove \c{backslash()}.
305 \b Fix the \c{SMSW} and \c{SLDT} instructions.
307 \b \c{-O2} and \c{-O3} are no longer aliases for \c{-O10} and \c{-O15}.
308 If you mean the latter, please say so! :)
310 \S{cl-0.98.36} Version 0.98.36
313 \b Update rdoff - librarian/archiver - common rec - docs!
315 \b Fix signed/unsigned problems.
317 \b Fix \c{JMP FAR label} and \c{CALL FAR label}.
319 \b Add new multisection support - map files - fix align bug
321 \b Fix sysexit, movhps/movlps reg,reg bugs in insns.dat
323 \b \c{Q} or \c{O} suffixes indicate octal
325 \b Support Prescott new instructions (PNI).
327 \b Cyrix \c{XSTORE} instruction.
330 \S{cl-0.98.35} Version 0.98.35
332 \b Fix build failure on 16-bit DOS (Makefile.bc3 workaround for compiler bug.)
334 \b Fix dependencies and compiler warnings.
336 \b Add "const" in a number of places.
338 \b Add -X option to specify error reporting format (use -Xvc to
339 integrate with Microsoft Visual Studio.)
341 \b Minor changes for code legibility.
343 \b Drop use of tmpnam() in rdoff (security fix.)
346 \S{cl-0.98.34} Version 0.98.34
348 \b Correct additional address-size vs. operand-size confusions.
350 \b Generate dependencies for all Makefiles automatically.
352 \b Add support for unimplemented (but theoretically available)
353 registers such as tr0 and cr5. Segment registers 6 and 7 are called
354 segr6 and segr7 for the operations which they can be represented.
356 \b Correct some disassembler bugs related to redundant address-size prefixes.
357 Some work still remains in this area.
359 \b Correctly generate an error for things like "SEG eax".
361 \b Add the JMPE instruction, enabled by "CPU IA64".
363 \b Correct compilation on newer gcc/glibc platforms.
365 \b Issue an error on things like "jmp far eax".
368 \S{cl-0.98.33} Version 0.98.33
370 \b New __NASM_PATCHLEVEL__ and __NASM_VERSION_ID__ standard macros to
371 round out the version-query macros. version.pl now understands
372 X.YYplWW or X.YY.ZZplWW as a version number, equivalent to
373 X.YY.ZZ.WW (or X.YY.0.WW, as appropriate).
375 \b New keyword "strict" to disable the optimization of specific
378 \b Fix the handing of size overrides with JMP instructions
379 (instructions such as "jmp dword foo".)
381 \b Fix the handling of "ABSOLUTE label", where "label" points into a
384 \b Fix OBJ output format with lots of externs.
386 \b More documentation updates.
388 \b Add -Ov option to get verbose information about optimizations.
390 \b Undo a braindead change which broke \c{%elif} directives.
395 \S{cl-0.98.32} Version 0.98.32
397 \b Fix NASM crashing when \c{%macro} directives were left unterminated.
399 \b Lots of documentation updates.
401 \b Complete rewrite of the PostScript/PDF documentation generator.
403 \b The MS Visual C++ Makefile was updated and corrected.
405 \b Recognize .rodata as a standard section name in ELF.
407 \b Fix some obsolete Perl4-isms in Perl scripts.
409 \b Fix configure.in to work with autoconf 2.5x.
411 \b Fix a couple of "make cleaner" misses.
413 \b Make the normal "./configure && make" work with Cygwin.
416 \S{cl-0.98.31} Version 0.98.31
418 \b Correctly build in a separate object directory again.
420 \b Derive all references to the version number from the version file.
422 \b New standard macros __NASM_SUBMINOR__ and __NASM_VER__ macros.
424 \b Lots of Makefile updates and bug fixes.
426 \b New \c{%ifmacro} directive to test for multiline macros.
428 \b Documentation updates.
430 \b Fixes for 16-bit OBJ format output.
432 \b Changed the NASM environment variable to NASMENV.
435 \S{cl-0.98.30} Version 0.98.30
437 \b Changed doc files a lot: completely removed old READMExx and
438 Wishlist files, incorporating all information in CHANGES and TODO.
440 \b I waited a long time to rename zoutieee.c to (original) outieee.c
442 \b moved all output modules to output/ subdirectory.
444 \b Added 'make strip' target to strip debug info from nasm & ndisasm.
446 \b Added INSTALL file with installation instructions.
448 \b Added -v option description to nasm man.
450 \b Added dist makefile target to produce source distributions.
452 \b 16-bit support for ELF output format (GNU extension, but useful.)
455 \S{cl-0.98.28} Version 0.98.28
457 \b Fastcooked this for Debian's Woody release:
458 Frank applied the INCBIN bug patch to 0.98.25alt and called
459 it 0.98.28 to not confuse poor little apt-get.
462 \S{cl-0.98.26} Version 0.98.26
464 \b Reorganised files even better from 0.98.25alt
467 \S{cl-0.98.25alt} Version 0.98.25alt
469 \b Prettified the source tree. Moved files to more reasonable places.
471 \b Added findleak.pl script to misc/ directory.
473 \b Attempted to fix doc.
476 \S{cl-0.98.25} Version 0.98.25
478 \b Line continuation character \c{\\}.
480 \b Docs inadvertantly reverted - "dos packaging".
483 \S{cl-0.98.24p1} Version 0.98.24p1
485 \b FIXME: Someone, document this please.
488 \S{cl-0.98.24} Version 0.98.24
490 \b Documentation - Ndisasm doc added to Nasm.doc.
493 \S{cl-0.98.23} Version 0.98.23
495 \b Attempted to remove rdoff version1
497 \b Lino Mastrodomenico's patches to preproc.c (%$$ bug?).
500 \S{cl-0.98.22} Version 0.98.22
502 \b Update rdoff2 - attempt to remove v1.
505 \S{cl-0.98.21} Version 0.98.21
507 \b Optimization fixes.
510 \S{cl-0.98.20} Version 0.98.20
512 \b Optimization fixes.
515 \S{cl-0.98.19} Version 0.98.19
517 \b H. J. Lu's patch back out.
520 \S{cl-0.98.18} Version 0.98.18
522 \b Added ".rdata" to "-f win32".
525 \S{cl-0.98.17} Version 0.98.17
527 \b H. J. Lu's "bogus elf" patch. (Red Hat problem?)
530 \S{cl-0.98.16} Version 0.98.16
532 \b Fix whitespace before "[section ..." bug.
535 \S{cl-0.98.15} Version 0.98.15
537 \b Rdoff changes (?).
539 \b Fix fixes to memory leaks.
542 \S{cl-0.98.14} Version 0.98.14
547 \S{cl-0.98.13} Version 0.98.13
549 \b There was no 0.98.13
552 \S{cl-0.98.12} Version 0.98.12
554 \b Update optimization (new function of "-O1")
556 \b Changes to test/bintest.asm (?).
559 \S{cl-0.98.11} Version 0.98.11
561 \b Optimization changes.
566 \S{cl-0.98.10} Version 0.98.10
568 \b There was no 0.98.10
571 \S{cl-0.98.09} Version 0.98.09
573 \b Add multiple sections support to "-f bin".
575 \b Changed GLOBAL_TEMP_BASE in outelf.c from 6 to 15.
577 \b Add "-v" as an alias to the "-r" switch.
579 \b Remove "#ifdef" from Tasm compatibility options.
581 \b Remove redundant size-overrides on "mov ds, ex", etc.
583 \b Fixes to SSE2, other insns.dat (?).
585 \b Enable uppercase "I" and "P" switches.
587 \b Case insinsitive "seg" and "wrt".
589 \b Update install.sh (?).
591 \b Allocate tokens in blocks.
593 \b Improve "invalid effective address" messages.
596 \S{cl-0.98.08} Version 0.98.08
598 \b Add "\c{%strlen}" and "\c{%substr}" macro operators
600 \b Fixed broken c16.mac.
602 \b Unterminated string error reported.
604 \b Fixed bugs as per 0.98bf
607 \S{cl-0.98.09b with John Coffman patches released 28-Oct-2001} Version 0.98.09b with John Coffman patches released 28-Oct-2001
609 Changes from 0.98.07 release to 98.09b as of 28-Oct-2001
611 \b More closely compatible with 0.98 when -O0 is implied
612 or specified. Not strictly identical, since backward
613 branches in range of short offsets are recognized, and signed
614 byte values with no explicit size specification will be
615 assembled as a single byte.
617 \b More forgiving with the PUSH instruction. 0.98 requires
618 a size to be specified always. 0.98.09b will imply the size
619 from the current BITS setting (16 or 32).
621 \b Changed definition of the optimization flag:
623 -O0 strict two-pass assembly, JMP and Jcc are
624 handled more like 0.98, except that back-
625 ward JMPs are short, if possible.
627 -O1 strict two-pass assembly, but forward
628 branches are assembled with code guaranteed
629 to reach; may produce larger code than
630 -O0, but will produce successful assembly
631 more often if branch offset sizes are not
634 -O2 multi-pass optimization, minimize branch
635 offsets; also will minimize signed immed-
636 iate bytes, overriding size specification.
638 -O3 like -O2, but more passes taken, if needed
641 \S{cl-0.98.07 released 01/28/01} Version 0.98.07 released 01/28/01
643 \b Added Stepane Denis' SSE2 instructions to a *working*
644 version of the code - some earlier versions were based on
645 broken code - sorry 'bout that. version "0.98.07"
651 \b Cosmetic modifications to nasm.c, nasm.h,
655 \S{cl-0.98.06f released 01/18/01} Version 0.98.06f released 01/18/01
658 \b - Add "metalbrain"s jecxz bug fix in insns.dat
659 - alter nasmdoc.src to match - version "0.98.06f"
662 \S{cl-0.98.06e released 01/09/01} Version 0.98.06e released 01/09/01
665 \b Removed the "outforms.h" file - it appears to be
666 someone's old backup of "outform.h". version "0.98.06e"
670 \b fbk - finally added the fix for the "multiple %includes bug",
671 known since 7/27/99 - reported originally (?) and sent to
672 us by Austin Lunnen - he reports that John Fine had a fix
673 within the day. Here it is...
675 \b Nelson Rush resigns from the group. Big thanks to Nelson for
676 his leadership and enthusiasm in getting these changes
677 incorporated into Nasm!
679 \b fbk - [list +], [list -] directives - ineptly implemented, should
680 be re-written or removed, perhaps.
682 \b Brian Raiter / fbk - "elfso bug" fix - applied to aoutb format
683 as well - testing might be desirable...
687 \b James Seter - -postfix, -prefix command line switches.
689 \b Yuri Zaporogets - rdoff utility changes.
692 \S{cl-0.98p1} Version 0.98p1
694 \b GAS-like palign (Panos Minos)
696 \b FIXME: Someone, fill this in with details
699 \S{cl-0.98bf (bug-fixed)} Version 0.98bf (bug-fixed)
701 \b Fixed - elf and aoutb bug - shared libraries
702 - multiple "%include" bug in "-f obj"
704 - unrecognized option bug in ndisasm
706 \S{cl-0.98.03 with John Coffman's changes released 27-Jul-2000} Version 0.98.03 with John Coffman's changes released 27-Jul-2000
708 \b Added signed byte optimizations for the 0x81/0x83 class
709 of instructions: ADC, ADD, AND, CMP, OR, SBB, SUB, XOR:
710 when used as 'ADD reg16,imm' or 'ADD reg32,imm.' Also
711 optimization of signed byte form of 'PUSH imm' and 'IMUL
712 reg,imm'/'IMUL reg,reg,imm.' No size specification is needed.
714 \b Added multi-pass JMP and Jcc offset optimization. Offsets
715 on forward references will preferentially use the short form,
716 without the need to code a specific size (short or near) for
717 the branch. Added instructions for 'Jcc label' to use the
718 form 'Jnotcc $+3/JMP label', in cases where a short offset
719 is out of bounds. If compiling for a 386 or higher CPU, then
720 the 386 form of Jcc will be used instead.
722 This feature is controlled by a new command-line switch: "O",
723 (upper case letter O). "-O0" reverts the assembler to no
724 extra optimization passes, "-O1" allows up to 5 extra passes,
725 and "-O2"(default), allows up to 10 extra optimization passes.
727 \b Added a new directive: 'cpu XXX', where XXX is any of:
728 8086, 186, 286, 386, 486, 586, pentium, 686, PPro, P2, P3 or
729 Katmai. All are case insensitive. All instructions will
730 be selected only if they apply to the selected cpu or lower.
731 Corrected a couple of bugs in cpu-dependence in 'insns.dat'.
733 \b Added to 'standard.mac', the "use16" and "use32" forms of
734 the "bits 16/32" directive. This is nothing new, just conforms
735 to a lot of other assemblers. (minor)
737 \b Changed label allocation from 320/32 (10000 labels @ 200K+)
738 to 32/37 (1000 labels); makes running under DOS much easier.
739 Since additional label space is allocated dynamically, this
740 should have no effect on large programs with lots of labels.
741 The 37 is a prime, believed to be better for hashing. (minor)
744 \S{cl-0.98.03} Version 0.98.03
746 "Integrated patchfile 0.98-0.98.01. I call this version 0.98.03 for
747 historical reasons: 0.98.02 was trashed." --John Coffman
748 <johninsd@san.rr.com>, 27-Jul-2000
750 \b Kendall Bennett's SciTech MGL changes
752 \b Note that you must define "TASM_COMPAT" at compile-time
753 to get the Tasm Ideal Mode compatibility.
755 \b All changes can be compiled in and out using the TASM_COMPAT macros,
756 and when compiled without TASM_COMPAT defined we get the exact same
757 binary as the unmodified 0.98 sources.
759 \b standard.mac, macros.c: Added macros to ignore TASM directives before
762 \b nasm.h: Added extern declaration for tasm_compatible_mode
764 \b nasm.c: Added global variable tasm_compatible_mode
766 \b Added command line switch for TASM compatible mode (-t)
768 \b Changed version command line to reflect when compiled with TASM additions
770 \b Added response file processing to allow all arguments on a single
771 line (response file is @resp rather than -@resp for NASM format).
773 \b labels.c: Changes islocal() macro to support TASM style @@local labels.
775 \b Added islocalchar() macro to support TASM style @@local labels.
777 \b parser.c: Added support for TASM style memory references (ie: mov
778 [DWORD eax],10 rather than the NASM style mov DWORD [eax],10).
780 \b preproc.c: Added new directives, \c{%arg}, \c{%local}, \c{%stacksize} to directives
783 \b Added support for TASM style directives without a leading % symbol.
785 \b Integrated a block of changes from Andrew Zabolotny <bit@eltech.ru>:
787 \b A new keyword \c{%xdefine} and its case-insensitive counterpart \c{%ixdefine}.
788 They work almost the same way as \c{%define} and \c{%idefine} but expand
789 the definition immediately, not on the invocation. Something like a cross
790 between \c{%define} and \c{%assign}. The "x" suffix stands for "eXpand", so
791 "xdefine" can be deciphered as "expand-and-define". Thus you can do
797 \c %xdefine %1 dword [esp+ofs]
801 \b Changed the place where the expansion of %$name macros are expanded.
802 Now they are converted into ..@ctxnum.name form when detokenizing, so
803 there are no quirks as before when using %$name arguments to macros,
804 in macros etc. For example:
813 Now last line will be expanded into "hello" as expected. This also allows
814 for lots of goodies, a good example are extended "proc" macros included
817 \b Added a check for "cstk" in smacro_defined() before calling get_ctx() -
818 this allows for things like:
823 to work without warnings even in no context.
825 \b Added a check for "cstk" in %if*ctx and %elif*ctx directives -
826 this allows to use \c{%ifctx} without excessive warnings. If there is
827 no active context, \c{%ifctx} goes through "false" branch.
829 \b Removed "user error: " prefix with \c{%error} directive: it just clobbers the
830 output and has absolutely no functionality. Besides, this allows to write
831 macros that does not differ from built-in functions in any way.
833 \b Added expansion of string that is output by \c{%error} directive. Now you
836 \c %define hello(x) Hello, x!
838 \c %define %$name andy
839 \c %error "hello(%$name)"
841 Same happened with \c{%include} directive.
843 \b Now all directives that expect an identifier will try to expand and
844 concatenate everything without whitespaces in between before usage.
845 For example, with "unfixed" nasm the commands
847 \c %define %$abc hello
848 \c %define __%$abc goodbye
851 would produce "incorrect" output: last line will expand to
853 \c hello goodbyehello
855 Not quite what you expected, eh? :-) The answer is that preprocessor
856 treats the \c{%define} construct as if it would be
858 \c %define __ %$abc goodbye
860 (note the white space between __ and %$abc). After my "fix" it
861 will "correctly" expand into
865 as expected. Note that I use quotes around words "correct", "incorrect"
866 etc because this is rather a feature not a bug; however current behaviour
867 is more logical (and allows more advanced macro usage :-).
869 Same change was applied to:
870 \c{%push},\c{%macro},\c{%imacro},\c{%define},\c{%idefine},\c{%xdefine},\c{%ixdefine},
871 \c{%assign},\c{%iassign},\c{%undef}
873 \b A new directive [WARNING {+|-}warning-id] have been added. It works only
874 if the assembly phase is enabled (i.e. it doesn't work with nasm -e).
876 \b A new warning type: macro-selfref. By default this warning is disabled;
877 when enabled NASM warns when a macro self-references itself; for example
878 the following source:
880 \c [WARNING macro-selfref]
891 will produce a warning, but if we remove the first line we won't see it
892 anymore (which is The Right Thing To Do {tm} IMHO since C preprocessor
893 eats such constructs without warnings at all).
895 \b Added a "error" routine to preprocessor which always will set ERR_PASS1
896 bit in severity_code. This removes annoying repeated errors on first
897 and second passes from preprocessor.
899 \b Added the %+ operator in single-line macros for concatenating two
900 identifiers. Usage example:
902 \c %define _myfunc _otherfunc
903 \c %define cextern(x) _ %+ x
906 After first expansion, third line will become "_myfunc". After this
907 expansion is performed again so it becomes "_otherunc".
909 \b Now if preprocessor is in a non-emitting state, no warning or error
910 will be emitted. Example:
915 \c put anything you want between these two brackets,
916 \c even macro-parameter references %1 or local
917 \c labels %$zz or macro-local labels %%zz - no
918 \c warning will be emitted.
921 \b Context-local variables on expansion as a last resort are looked up
922 in outer contexts. For example, the following piece:
932 will expand correctly the fourth line to [esp]; if we'll define another
933 %$a inside the "inner" context, it will take precedence over outer
934 definition. However, this modification has been applied only to
935 expand_smacro and not to smacro_define: as a consequence expansion
936 looks in outer contexts, but \c{%ifdef} won't look in outer contexts.
938 This behaviour is needed because we don't want nested contexts to
939 act on already defined local macros. Example:
941 \c %define %$arg1 [esp+4]
947 In this example the "if" mmacro enters into the "if" context, so %$arg1
948 is not valid anymore inside "if". Of course it could be worked around
949 by using explicitely %$$arg1 but this is ugly IMHO.
951 \b Fixed memory leak in \c{%undef}. The origline wasn't freed before
954 \b Fixed trap in preprocessor when line expanded to empty set of tokens.
955 This happens, for example, in the following case:
961 \S{cl-0.98} Version 0.98
963 All changes since NASM 0.98p3 have been produced by H. Peter Anvin <hpa@zytor.com>.
965 \b The documentation comment delimiter is \# not #.
967 \b Allow EQU definitions to refer to external labels; reported by
970 \b Re-enable support for RDOFF v1; reported by Pedro Gimeno.
972 \b Updated License file per OK from Simon and Julian.
975 \S{cl-0.98p9} Version 0.98p9
977 \b Update documentation (although the instruction set reference will
978 have to wait; I don't want to hold up the 0.98 release for it.)
980 \b Verified that the NASM implementation of the PEXTRW and PMOVMSKB
981 instructions is correct. The encoding differs from what the Intel
982 manuals document, but the Pentium III behaviour matches NASM, not
985 \b Fix handling of implicit sizes in PSHUFW and PINSRW, reported by
988 \b Resurrect the -s option, which was removed when changing the
989 diagnostic output to stdout.
992 \S{cl-0.98p8} Version 0.98p8
994 \b Fix for "DB" when NASM is running on a bigendian machine.
996 \b Invoke insns.pl once for each output script, making Makefile.in
999 \b Improve the Unix configure-based makefiles to make package
1002 \b Included an RPM .spec file for building RPM (RedHat Package Manager)
1003 packages on Linux or Unix systems.
1005 \b Fix Makefile dependency problems.
1007 \b Change src/rdsrc.pl to include sectioning information in info
1008 output; required for install-info to work.
1010 \b Updated the RDOFF distribution to version 2 from Jules; minor
1011 massaging to make it compile in my environment.
1013 \b Split doc files that can be built by anyone with a Perl interpreter off
1014 into a separate archive.
1016 \b "Dress rehearsal" release!
1019 \S{cl-0.98p7} Version 0.98p7
1021 \b Fixed opcodes with a third byte-sized immediate argument to not
1022 complain if given "byte" on the immediate.
1024 \b Allow \c{%undef} to remove single-line macros with arguments. This
1025 matches the behaviour of #undef in the C preprocessor.
1027 \b Allow -d, -u, -i and -p to be specified as -D, -U, -I and -P for
1028 compatibility with most C compilers and preprocessors. This allows
1029 Makefile options to be shared between cc and nasm, for example.
1033 \b Went through the list of Katmai instructions and hopefully fixed the
1034 (rather few) mistakes in it.
1036 \b (Hopefully) fixed a number of disassembler bugs related to ambiguous
1037 instructions (disambiguated by -p) and SSE instructions with REP.
1039 \b Fix for bug reported by Mark Junger: "call dword 0x12345678" should
1040 work and may add an OSP (affected CALL, JMP, Jcc).
1042 \b Fix for environments when "stderr" isn't a compile-time constant.
1045 \S{cl-0.98p6} Version 0.98p6
1048 \b Took officially over coordination of the 0.98 release; so drop
1049 the p3.x notation. Skipped p4 and p5 to avoid confusion with John
1050 Fine's J4 and J5 releases.
1052 \b Update the documentation; however, it still doesn't include
1053 documentation for the various new instructions. I somehow wonder if
1054 it makes sense to have an instruction set reference in the assembler
1055 manual when Intel et al have PDF versions of their manuals online.
1057 \b Recognize "idt" or "centaur" for the -p option to ndisasm.
1059 \b Changed error messages back to stderr where they belong, but add an
1060 -E option to redirect them elsewhere (the DOS shell cannot redirect
1063 \b -M option to generate Makefile dependencies (based on code from Alex
1066 \b \c{%undef} preprocessor directive, and -u option, that undefines a
1069 \b OS/2 Makefile (Mkfiles/Makefile.os2) for Borland under OS/2; from
1072 \b Various minor bugfixes (reported by):
1073 - Dangling \c{%s} in preproc.c (Martin Junker)
1075 \b THERE ARE KNOWN BUGS IN SSE AND THE OTHER KATMAI INSTRUCTIONS. I am
1076 on a trip and didn't bring the Katmai instruction reference, so I
1077 can't work on them right now.
1079 \b Updated the License file per agreement with Simon and Jules to
1080 include a GPL distribution clause.
1083 \S{cl-0.98p3.7} Version 0.98p3.7
1085 \b (Hopefully) fixed the canned Makefiles to include the outrdf2 and
1088 \b Renamed changes.asm to changed.asm.
1091 \S{cl-0.98p3.6} Version 0.98p3.6
1093 \b Fixed a bunch of instructions that were added in 0.98p3.5 which had
1094 memory operands, and the address-size prefix was missing from the
1095 instruction pattern.
1098 \S{cl-0.98p3.5} Version 0.98p3.5
1100 \b Merged in changes from John S. Fine's 0.98-J5 release. John's based
1101 0.98-J5 on my 0.98p3.3 release; this merges the changes.
1103 \b Expanded the instructions flag field to a long so we can fit more
1104 flags; mark SSE (KNI) and AMD or Katmai-specific instructions as
1107 \b Fix the "PRIV" flag on a bunch of instructions, and create new
1108 "PROT" flag for protected-mode-only instructions (orthogonal to if
1109 the instruction is privileged!) and new "SMM" flag for SMM-only
1112 \b Added AMD-only SYSCALL and SYSRET instructions.
1114 \b Make SSE actually work, and add new Katmai MMX instructions.
1116 \b Added a -p (preferred vendor) option to ndisasm so that it can
1117 distinguish e.g. Cyrix opcodes also used in SSE. For example:
1119 \c ndisasm -p cyrix aliased.bin
1120 \c 00000000 670F514310 paddsiw mm0,[ebx+0x10]
1121 \c 00000005 670F514320 paddsiw mm0,[ebx+0x20]
1122 \c ndisasm -p intel aliased.bin
1123 \c 00000000 670F514310 sqrtps xmm0,[ebx+0x10]
1124 \c 00000005 670F514320 sqrtps xmm0,[ebx+0x20]
1126 \b Added a bunch of Cyrix-specific instructions.
1129 \S{cl-0.98p3.4} Version 0.98p3.4
1131 \b Made at least an attempt to modify all the additional Makefiles (in
1132 the Mkfiles directory). I can't test it, but this was the best I
1135 \b DOS DJGPP+"Opus Make" Makefile from John S. Fine.
1137 \b changes.asm changes from John S. Fine.
1140 \S{cl-0.98p3.3} Version 0.98p3.3
1142 \b Patch from Conan Brink to allow nesting of \c{%rep} directives.
1144 \b If we're going to allow INT01 as an alias for INT1/ICEBP (one of
1145 Jules 0.98p3 changes), then we should allow INT03 as an alias for INT3
1148 \b Updated changes.asm to include the latest changes.
1150 \b Tried to clean up the <CR>s that had snuck in from a DOS/Windows
1151 environment into my Unix environment, and try to make sure than
1152 DOS/Windows users get them back.
1154 \b We would silently generate broken tools if insns.dat wasn't sorted
1155 properly. Change insns.pl so that the order doesn't matter.
1157 \b Fix bug in insns.pl (introduced by me) which would cause conditional
1158 instructions to have an extra "cc" in disassembly, e.g. "jnz"
1159 disassembled as "jccnz".
1162 \S{cl-0.98p3.2} Version 0.98p3.2
1164 \b Merged in John S. Fine's changes from his 0.98-J4 prerelease; see
1165 http://www.csoft.net/cz/johnfine/
1167 \b Changed previous "spotless" Makefile target (appropriate for distribution)
1168 to "distclean", and added "cleaner" target which is same as "clean"
1169 except deletes files generated by Perl scripts; "spotless" is union.
1171 \b Removed BASIC programs from distribution. Get a Perl interpreter
1172 instead (see below.)
1174 \b Calling this "pre-release 3.2" rather than "p3-hpa2" because of
1175 John's contributions.
1177 \b Actually link in the IEEE output format (zoutieee.c); fix a bunch of
1178 compiler warnings in that file. Note I don't know what IEEE output
1179 is supposed to look like, so these changes were made "blind".
1182 \S{cl-0.98p3-hpa} Version 0.98p3-hpa
1184 \b Merged nasm098p3.zip with nasm-0.97.tar.gz to create a fully
1185 buildable version for Unix systems (Makefile.in updates, etc.)
1187 \b Changed insns.pl to create the instruction tables in nasm.h and
1188 names.c, so that a new instruction can be added by adding it *only*
1191 \b Added the following new instructions: SYSENTER, SYSEXIT, FXSAVE,
1192 FXRSTOR, UD1, UD2 (the latter two are two opcodes that Intel
1193 guarantee will never be used; one of them is documented as UD2 in
1194 Intel documentation, the other one just as "Undefined Opcode" --
1195 calling it UD1 seemed to make sense.)
1197 \b MAX_SYMBOL was defined to be 9, but LOADALL286 and LOADALL386 are 10
1198 characters long. Now MAX_SYMBOL is derived from insns.dat.
1200 \b A note on the BASIC programs included: forget them. insns.bas is
1201 already out of date. Get yourself a Perl interpreter for your
1202 platform of choice at
1203 \W{http://www.cpan.org/ports/index.html}{http://www.cpan.org/ports/index.html}.
1206 \S{cl-0.98p3} Version 0.98 pre-release 3
1208 \b added response file support, improved command line handling, new layout
1211 \b fixed limit checking bug, 'OUT byte nn, reg' bug, and a couple of rdoff
1212 related bugs, updated Wishlist; 0.98 Prerelease 3.
1215 \S{cl-0.98p2} Version 0.98 pre-release 2
1217 \b fixed bug in outcoff.c to do with truncating section names longer
1218 than 8 characters, referencing beyond end of string; 0.98 pre-release 2
1221 \S{cl-0.98p1} Version 0.98 pre-release 1
1223 \b Fixed a bug whereby STRUC didn't work at all in RDF.
1225 \b Fixed a problem with group specification in PUBDEFs in OBJ.
1227 \b Improved ease of adding new output formats. Contribution due to
1230 \b Fixed a bug in relocations in the `bin' format: was showing up when
1231 a relocatable reference crossed an 8192-byte boundary in any output
1234 \b Fixed a bug in local labels: local-label lookups were inconsistent
1235 between passes one and two if an EQU occurred between the definition
1236 of a global label and the subsequent use of a local label local to
1239 \b Fixed a seg-fault in the preprocessor (again) which happened when
1240 you use a blank line as the first line of a multi-line macro
1241 definition and then defined a label on the same line as a call to
1244 \b Fixed a stale-pointer bug in the handling of the NASM environment
1245 variable. Thanks to Thomas McWilliams.
1247 \b ELF had a hard limit on the number of sections which caused
1248 segfaults when transgressed. Fixed.
1250 \b Added ability for ndisasm to read from stdin by using `-' as the
1253 \b ndisasm wasn't outputting the TO keyword. Fixed.
1255 \b Fixed error cascade on bogus expression in \c{%if} - an error in
1256 evaluation was causing the entire \c{%if} to be discarded, thus creating
1257 trouble later when the \c{%else} or \c{%endif} was encountered.
1259 \b Forward reference tracking was instruction-granular not operand-
1260 granular, which was causing 286-specific code to be generated
1261 needlessly on code of the form `shr word [forwardref],1'. Thanks to
1262 Jim Hague for sending a patch.
1264 \b All messages now appear on stdout, as sending them to stderr serves
1265 no useful purpose other than to make redirection difficult.
1267 \b Fixed the problem with EQUs pointing to an external symbol - this
1268 now generates an error message.
1270 \b Allowed multiple size prefixes to an operand, of which only the first
1271 is taken into account.
1273 \b Incorporated John Fine's changes, including fixes of a large number
1274 of preprocessor bugs, some small problems in OBJ, and a reworking of
1275 label handling to define labels before their line is assembled, rather
1278 \b Reformatted a lot of the source code to be more readable. Included
1279 'coding.txt' as a guideline for how to format code for contributors.
1281 \b Stopped nested \c{%reps} causing a panic - they now cause a slightly more
1282 friendly error message instead.
1284 \b Fixed floating point constant problems (patch by Pedro Gimeno)
1286 \b Fixed the return value of insn_size() not being checked for -1, indicating
1289 \b Incorporated 3Dnow! instructions.
1291 \b Fixed the 'mov eax, eax + ebx' bug.
1293 \b Fixed the GLOBAL EQU bug in ELF. Released developers release 3.
1295 \b Incorporated John Fine's command line parsing changes
1297 \b Incorporated David Lindauer's OMF debug support
1299 \b Made changes for LCC 4.0 support (\c{__NASM_CDecl__}, removed register size
1300 specification warning when sizes agree).
1303 \H{cl-0.9x} NASM 0.9 Series
1305 Revisions before 0.98.
1308 \S{cl-0.97} Version 0.97 released December 1997
1310 \b This was entirely a bug-fix release to 0.96, which seems to have got
1313 \b Fixed stupid mistake in OBJ which caused `MOV EAX,<constant>' to
1314 fail. Caused by an error in the `MOV EAX,<segment>' support.
1316 \b ndisasm hung at EOF when compiled with lcc on Linux because lcc on
1317 Linux somehow breaks feof(). ndisasm now does not rely on feof().
1319 \b A heading in the documentation was missing due to a markup error in
1320 the indexing. Fixed.
1322 \b Fixed failure to update all pointers on realloc() within extended-
1323 operand code in parser.c. Was causing wrong behaviour and seg faults
1324 on lines such as `dd 0.0,0.0,0.0,0.0,...'
1326 \b Fixed a subtle preprocessor bug whereby invoking one multi-line
1327 macro on the first line of the expansion of another, when the second
1328 had been invoked with a label defined before it, didn't expand the
1331 \b Added internal.doc back in to the distribution archives - it was
1332 missing in 0.96 *blush*
1334 \b Fixed bug causing 0.96 to be unable to assemble its own test files,
1335 specifically objtest.asm. *blush again*
1337 \b Fixed seg-faults and bogus error messages caused by mismatching
1338 \c{%rep} and \c{%endrep} within multi-line macro definitions.
1340 \b Fixed a problem with buffer overrun in OBJ, which was causing
1341 corruption at ends of long PUBDEF records.
1343 \b Separated DOS archives into main-program and documentation to reduce
1347 \S{cl-0.96} Version 0.96 released November 1997
1349 \b Fixed a bug whereby, if `nasm sourcefile' would cause a filename
1350 collision warning and put output into `nasm.out', then `nasm
1351 sourcefile -o outputfile' still gave the warning even though the
1353 Fixed name pollution under Digital UNIX: one of its header files
1354 defined R_SP, which broke the enum in nasm.h.
1356 \b Fixed minor instruction table problems: FUCOM and FUCOMP didn't have
1357 two-operand forms; NDISASM didn't recognise the longer register
1358 forms of PUSH and POP (eg FF F3 for PUSH BX); TEST mem,imm32 was
1359 flagged as undocumented; the 32-bit forms of CMOV had 16-bit operand
1360 size prefixes; `AAD imm' and `AAM imm' are no longer flagged as
1361 undocumented because the Intel Architecture reference documents
1364 \b Fixed a problem with the local-label mechanism, whereby strange
1365 types of symbol (EQUs, auto-defined OBJ segment base symbols)
1366 interfered with the `previous global label' value and screwed up
1369 \b Fixed a bug whereby the stub preprocessor didn't communicate with
1370 the listing file generator, so that the -a and -l options in
1371 conjunction would produce a useless listing file.
1373 \b Merged `os2' object file format back into `obj', after discovering
1374 that `obj' _also_ shouldn't have a link pass separator in a module
1375 containing a non-trivial MODEND. Flat segments are now declared
1376 using the FLAT attribute. `os2' is no longer a valid object format
1379 \b Removed the fixed-size temporary storage in the evaluator. Very very
1380 long expressions (like `mov ax,1+1+1+1+...' for two hundred 1s or
1381 so) should now no longer crash NASM.
1383 \b Fixed a bug involving segfaults on disassembly of MMX instructions,
1384 by changing the meaning of one of the operand-type flags in nasm.h.
1385 This may cause other apparently unrelated MMX problems; it needs to
1386 be tested thoroughly.
1388 \b Fixed some buffer overrun problems with large OBJ output files.
1389 Thanks to DJ Delorie for the bug report and fix.
1391 \b Made preprocess-only mode actually listen to the \c{%line} markers as it
1392 prints them, so that it can report errors more sanely.
1394 \b Re-designed the evaluator to keep more sensible track of expressions
1395 involving forward references: can now cope with previously-nightmare
1402 \b Added the ALIGN and ALIGNB standard macros.
1404 \b Added PIC support in ELF: use of WRT to obtain the four extra
1405 relocation types needed.
1407 \b Added the ability for output file formats to define their own
1408 extensions to the GLOBAL, COMMON and EXTERN directives.
1410 \b Implemented common-variable alignment, and global-symbol type and
1411 size declarations, in ELF.
1413 \b Implemented NEAR and FAR keywords for common variables, plus
1414 far-common element size specification, in OBJ.
1416 \b Added a feature whereby EXTERNs and COMMONs in OBJ can be given a
1417 default WRT specification (either a segment or a group).
1419 \b Transformed the Unix NASM archive into an auto-configuring package.
1421 \b Added a sanity-check for people applying SEG to things which are
1422 already segment bases: this previously went unnoticed by the SEG
1423 processing and caused OBJ-driver panics later.
1425 \b Added the ability, in OBJ format, to deal with `MOV EAX,<segment>'
1426 type references: OBJ doesn't directly support dword-size segment
1427 base fixups, but as long as the low two bytes of the constant term
1428 are zero, a word-size fixup can be generated instead and it will
1431 \b Added the ability to specify sections' alignment requirements in
1432 Win32 object files and pure binary files.
1434 \b Added preprocess-time expression evaluation: the \c{%assign} (and
1435 \c{%iassign}) directive and the bare \c{%if} (and \c{%elif}) conditional. Added
1436 relational operators to the evaluator, for use only in \c{%if}
1437 constructs: the standard relationals = < > <= >= <> (and C-like
1438 synonyms == and !=) plus low-precedence logical operators &&, ^^ and
1441 \b Added a preprocessor repeat construct: \c{%rep} / \c{%exitrep} / \c{%endrep}.
1443 \b Added the __FILE__ and __LINE__ standard macros.
1445 \b Added a sanity check for number constants being greater than
1446 0xFFFFFFFF. The warning can be disabled.
1448 \b Added the %0 token whereby a variadic multi-line macro can tell how
1449 many parameters it's been given in a specific invocation.
1451 \b Added \c{%rotate}, allowing multi-line macro parameters to be cycled.
1453 \b Added the `*' option for the maximum parameter count on multi-line
1454 macros, allowing them to take arbitrarily many parameters.
1456 \b Added the ability for the user-level forms of EXTERN, GLOBAL and
1457 COMMON to take more than one argument.
1459 \b Added the IMPORT and EXPORT directives in OBJ format, to deal with
1462 \b Added some more preprocessor \c{%if} constructs: \c{%ifidn} / \c{%ifidni} (exact
1463 textual identity), and \c{%ifid} / \c{%ifnum} / \c{%ifstr} (token type testing).
1465 \b Added the ability to distinguish SHL AX,1 (the 8086 version) from
1466 SHL AX,BYTE 1 (the 286-and-upwards version whose constant happens to
1469 \b Added NetBSD/FreeBSD/OpenBSD's variant of a.out format, complete
1470 with PIC shared library features.
1472 \b Changed NASM's idiosyncratic handling of FCLEX, FDISI, FENI, FINIT,
1473 FSAVE, FSTCW, FSTENV, and FSTSW to bring it into line with the
1474 otherwise accepted standard. The previous behaviour, though it was a
1475 deliberate feature, was a deliberate feature based on a
1476 misunderstanding. Apologies for the inconvenience.
1478 \b Improved the flexibility of ABSOLUTE: you can now give it an
1479 expression rather than being restricted to a constant, and it can
1480 take relocatable arguments as well.
1482 \b Added the ability for a variable to be declared as EXTERN multiple
1483 times, and the subsequent definitions are just ignored.
1485 \b We now allow instruction prefixes (CS, DS, LOCK, REPZ etc) to be
1486 alone on a line (without a following instruction).
1488 \b Improved sanity checks on whether the arguments to EXTERN, GLOBAL
1489 and COMMON are valid identifiers.
1491 \b Added misc/exebin.mac to allow direct generation of .EXE files by
1492 hacking up an EXE header using DB and DW; also added test/binexe.asm
1493 to demonstrate the use of this. Thanks to Yann Guidon for
1494 contributing the EXE header code.
1496 \b ndisasm forgot to check whether the input file had been successfully
1497 opened. Now it does. Doh!
1499 \b Added the Cyrix extensions to the MMX instruction set.
1501 \b Added a hinting mechanism to allow [EAX+EBX] and [EBX+EAX] to be
1502 assembled differently. This is important since [ESI+EBP] and
1503 [EBP+ESI] have different default base segment registers.
1505 \b Added support for the PharLap OMF extension for 4096-byte segment
1509 \S{cl-0.95 released July 1997} Version 0.95 released July 1997
1511 \b Fixed yet another ELF bug. This one manifested if the user relied on
1512 the default segment, and attempted to define global symbols without
1513 first explicitly declaring the target segment.
1515 \b Added makefiles (for NASM and the RDF tools) to build Win32 console
1516 apps under Symantec C++. Donated by Mark Junker.
1518 \b Added `macros.bas' and `insns.bas', QBasic versions of the Perl
1519 scripts that convert `standard.mac' to `macros.c' and convert
1520 `insns.dat' to `insnsa.c' and `insnsd.c'. Also thanks to Mark
1523 \b Changed the diassembled forms of the conditional instructions so
1524 that JB is now emitted as JC, and other similar changes. Suggested
1525 list by Ulrich Doewich.
1527 \b Added `@' to the list of valid characters to begin an identifier
1530 \b Documentary changes, notably the addition of the `Common Problems'
1531 section in nasm.doc.
1533 \b Fixed a bug relating to 32-bit PC-relative fixups in OBJ.
1535 \b Fixed a bug in perm_copy() in labels.c which was causing exceptions
1536 in cleanup_labels() on some systems.
1538 \b Positivity sanity check in TIMES argument changed from a warning to
1539 an error following a further complaint.
1541 \b Changed the acceptable limits on byte and word operands to allow
1542 things like `~10111001b' to work.
1544 \b Fixed a major problem in the preprocessor which caused seg-faults if
1545 macro definitions contained blank lines or comment-only lines.
1547 \b Fixed inadequate error checking on the commas separating the
1548 arguments to `db', `dw' etc.
1550 \b Fixed a crippling bug in the handling of macros with operand counts
1551 defined with a `+' modifier.
1553 \b Fixed a bug whereby object file formats which stored the input file
1554 name in the output file (such as OBJ and COFF) weren't doing so
1555 correctly when the output file name was specified on the command
1558 \b Removed [INC] and [INCLUDE] support for good, since they were
1561 \b Fixed a bug in OBJ which caused all fixups to be output in 16-bit
1562 (old-format) FIXUPP records, rather than putting the 32-bit ones in
1563 FIXUPP32 (new-format) records.
1565 \b Added, tentatively, OS/2 object file support (as a minor variant on
1568 \b Updates to Fox Cutter's Borland C makefile, Makefile.bc2.
1570 \b Removed a spurious second fclose() on the output file.
1572 \b Added the `-s' command line option to redirect all messages which
1573 would go to stderr (errors, help text) to stdout instead.
1575 \b Added the `-w' command line option to selectively suppress some
1576 classes of assembly warning messages.
1578 \b Added the `-p' pre-include and `-d' pre-define command-line options.
1580 \b Added an include file search path: the `-i' command line option.
1582 \b Fixed a silly little preprocessor bug whereby starting a line with a
1583 `%!' environment-variable reference caused an `unknown directive'
1586 \b Added the long-awaited listing file support: the `-l' command line
1589 \b Fixed a problem with OBJ format whereby, in the absence of any
1590 explicit segment definition, non-global symbols declared in the
1591 implicit default segment generated spurious EXTDEF records in the
1594 \b Added the NASM environment variable.
1596 \b From this version forward, Win32 console-mode binaries will be
1597 included in the DOS distribution in addition to the 16-bit binaries.
1598 Added Makefile.vc for this purpose.
1600 \b Added `return 0;' to test/objlink.c to prevent compiler warnings.
1602 \b Added the __NASM_MAJOR__ and __NASM_MINOR__ standard defines.
1604 \b Added an alternative memory-reference syntax in which prefixing an
1605 operand with `&' is equivalent to enclosing it in square brackets,
1606 at the request of Fox Cutter.
1608 \b Errors in pass two now cause the program to return a non-zero error
1609 code, which they didn't before.
1611 \b Fixed the single-line macro cycle detection, which didn't work at
1612 all on macros with no parameters (caused an infinite loop). Also
1613 changed the behaviour of single-line macro cycle detection to work
1614 like cpp, so that macros like `extrn' as given in the documentation
1617 \b Fixed the implementation of WRT, which was too restrictive in that
1618 you couldn't do `mov ax,[di+abc wrt dgroup]' because (di+abc) wasn't
1619 a relocatable reference.
1622 \S{cl-0.94 released April 1997} Version 0.94 released April 1997
1625 \b Major item: added the macro processor.
1627 \b Added undocumented instructions SMI, IBTS, XBTS and LOADALL286. Also
1628 reorganised CMPXCHG instruction into early-486 and Pentium forms.
1629 Thanks to Thobias Jones for the information.
1631 \b Fixed two more stupid bugs in ELF, which were causing `ld' to
1632 continue to seg-fault in a lot of non-trivial cases.
1634 \b Fixed a seg-fault in the label manager.
1636 \b Stopped FBLD and FBSTP from _requiring_ the TWORD keyword, which is
1637 the only option for BCD loads/stores in any case.
1639 \b Ensured FLDCW, FSTCW and FSTSW can cope with the WORD keyword, if
1640 anyone bothers to provide it. Previously they complained unless no
1641 keyword at all was present.
1643 \b Some forms of FDIV/FDIVR and FSUB/FSUBR were still inverted: a
1644 vestige of a bug that I thought had been fixed in 0.92. This was
1645 fixed, hopefully for good this time...
1647 \b Another minor phase error (insofar as a phase error can _ever_ be
1648 minor) fixed, this one occurring in code of the form
1650 \c rol ax,forward_reference
1651 \c forward_reference equ 1
1653 \b The number supplied to TIMES is now sanity-checked for positivity,
1654 and also may be greater than 64K (which previously didn't work on
1657 \b Added Watcom C makefiles, and misc/pmw.bat, donated by Dominik Behr.
1659 \b Added the INCBIN pseudo-opcode.
1661 \b Due to the advent of the preprocessor, the [INCLUDE] and [INC]
1662 directives have become obsolete. They are still supported in this
1663 version, with a warning, but won't be in the next.
1665 \b Fixed a bug in OBJ format, which caused incorrect object records to
1666 be output when absolute labels were made global.
1668 \b Updates to RDOFF subdirectory, and changes to outrdf.c.
1671 \S{cl-0.93 released January 1997} Version 0.93 released January 1997
1673 This release went out in a great hurry after semi-crippling bugs
1676 \b Really \e{did} fix the stack overflows this time. *blush*
1678 \b Had problems with EA instruction sizes changing between passes, when
1679 an offset contained a forward reference and so 4 bytes were
1680 allocated for the offset in pass one; by pass two the symbol had
1681 been defined and happened to be a small absolute value, so only 1
1682 byte got allocated, causing instruction size mismatch between passes
1683 and hence incorrect address calculations. Fixed.
1685 \b Stupid bug in the revised ELF section generation fixed (associated
1686 string-table section for .symtab was hard-coded as 7, even when this
1687 didn't fit with the real section table). Was causing `ld' to
1688 seg-fault under Linux.
1690 \b Included a new Borland C makefile, Makefile.bc2, donated by Fox
1691 Cutter <lmb@comtch.iea.com>.
1694 \S{cl-0.92 released January 1997} Version 0.92 released January 1997
1696 \b The FDIVP/FDIVRP and FSUBP/FSUBRP pairs had been inverted: this was
1697 fixed. This also affected the LCC driver.
1699 \b Fixed a bug regarding 32-bit effective addresses of the form
1700 \c{[other_register+ESP]}.
1702 \b Documentary changes, notably documentation of the fact that Borland
1703 Win32 compilers use `obj' rather than `win32' object format.
1705 \b Fixed the COMENT record in OBJ files, which was formatted
1708 \b Fixed a bug causing segfaults in large RDF files.
1710 \b OBJ format now strips initial periods from segment and group
1711 definitions, in order to avoid complications with the local label
1714 \b Fixed a bug in disassembling far calls and jumps in NDISASM.
1716 \b Added support for user-defined sections in COFF and ELF files.
1718 \b Compiled the DOS binaries with a sensible amount of stack, to
1719 prevent stack overflows on any arithmetic expression containing
1722 \b Fixed a bug in handling of files that do not terminate in a newline.
1725 \S{cl-0.91 released November 1996} Version 0.91 released November 1996
1727 \b Loads of bug fixes.
1729 \b Support for RDF added.
1731 \b Support for DBG debugging format added.
1733 \b Support for 32-bit extensions to Microsoft OBJ format added.
1735 \b Revised for Borland C: some variable names changed, makefile added.
1737 \b LCC support revised to actually work.
1739 \b JMP/CALL NEAR/FAR notation added.
1741 \b `a16', `o16', `a32' and `o32' prefixes added.
1743 \b Range checking on short jumps implemented.
1745 \b MMX instruction support added.
1747 \b Negative floating point constant support added.
1749 \b Memory handling improved to bypass 64K barrier under DOS.
1751 \b \c{$} prefix to force treatment of reserved words as identifiers added.
1753 \b Default-size mechanism for object formats added.
1755 \b Compile-time configurability added.
1757 \b \c{#}, \c{@}, \c{~} and c\{?} are now valid characters in labels.
1759 \b \c{-e} and \c{-k} options in NDISASM added.
1762 \S{cl-0.90 released October 1996} Version 0.90 released October 1996
1764 First release version. First support for object file output. Other
1765 changes from previous version (0.3x) too numerous to document.