| blob | d546b398c083c3e1cfda1c77d4f16200701bb8eb |
1 \#
2 \# NASM revision history in nasmdoc format
3 \#
5 \H{cl-2.xx} NASM 2 Series
7 The NASM 2 series support x86-64, and is the production version of NASM
8 since 2007.
11 \S{cl-2.05} Version 2.05
13 \b Fix redundant REX.W prefix on \c{JMP reg64}.
15 \b Make the behaviour of \c{-O0} match NASM 0.98 legacy behavior.
16 See \k{opt-O}.
18 \b \c{-w-user} can be used to suppress the output of \c{%warning} directives.
19 See \k{opt-w}.
21 \b Fix bug where \c{ALIGN} would issue a full alignment datum instead of
22 zero bytes.
24 \b Fix offsets in list files.
26 \b Fix \c{%include} inside multi-line macros or loops.
28 \b Fix error where NASM would generate a spurious warning on valid
29 optimizations of immediate values.
31 \b Fix arguments to a number of the \c{CVT} SSE instructions.
33 \b Fix RIP-relative offsets when the instruction carries an immediate.
35 \b Massive overhaul of the ELF64 backend for spec compliance.
37 \b Fix the Geode \c{PFRCPV} and \c{PFRSQRTV} instruction.
39 \b Fix the SSE 4.2 \c{CRC32} instruction.
42 \S{cl-2.04} Version 2.04
44 \b Sanitize macro handing in the \c{%error} directive.
46 \b New \c{%warning} directive to issue user-controlled warnings.
48 \b \c{%error} directives are now deferred to the final assembly phase.
50 \b New \c{%fatal} directive to immediately terminate assembly.
52 \b New \c{%strcat} directive to join quoted strings together.
54 \b New \c{%use} macro directive to support standard macro directives. See
55 \k{use}.
57 \b Excess default parameters to \c{%macro} now issues a warning by default.
58 See \k{mlmacro}.
60 \b Fix \c{%ifn} and \c{%elifn}.
62 \b Fix nested \c{%else} clauses.
64 \b Correct the handling of nested \c{%rep}s.
66 \b New \c{%unmacro} directive to undeclare a multi-line macro.
67 See \k{unmacro}.
69 \b Builtin macro \c{__PASS__} which expands to the current assembly pass.
70 See \k{pass_macro}.
72 \b \c{__utf16__} and \c{__utf32__} operators to generate UTF-16 and UTF-32
73 strings. See \k{unicode}.
75 \b Fix bug in case-insensitive matching when compiled on platforms that
76 don't use the \c{configure} script. Of the official release binaries,
77 that only affected the OS/2 binary.
79 \b Support for x87 packed BCD constants. See \k{bcdconst}.
81 \b Correct the \c{LTR} and \c{SLDT} instructions in 64-bit mode.
83 \b Fix unnecessary REX.W prefix on indirect jumps in 64-bit mode.
85 \b Add AVX versions of the AES instructions (\c{VAES}...).
87 \b Fix the 256-bit FMA instructions.
89 \b Add 256-bit AVX stores per the latest AVX spec.
91 \b VIA XCRYPT instructions can now be written either with or without
92 \c{REP}, apparently different versions of the VIA spec wrote them
93 differently.
95 \b Add missing 64-bit \c{MOVNTI} instruction.
97 \b Fix the operand size of \c{VMREAD} and \c{VMWRITE}.
99 \b Numerous bug fixes, especially to the AES, AVX and VTX instructions.
101 \b The optimizer now always runs until it converges. It also runs even
102 when disabled, but doesn't optimize. This allows most forward references
103 to be resolved properly.
106 \S{cl-2.03.01} Version 2.03.01
108 \b Fix buffer overflow in the listing module.
110 \b Fix the handling of hexadecimal escape codes in `...` strings.
112 \b The Postscript/PDF documentation has been reformatted.
114 \b The \c{-F} option now implies \c{-g}.
117 \S{cl-2.03} Version 2.03
119 \b Add support for Intel AVX, CLMUL and FMA instructions,
120 including YMM registers.
122 \b \c{dy}, \c{resy} and \c{yword} for 32-byte operands.
124 \b Fix some SSE5 instructions.
126 \b Intel \c{INVEPT}, \c{INVVPID} and \c{MOVBE} instructions.
128 \b Fix checking for critical expressions when the optimizer is enabled.
130 \b Support the DWARF debugging format for ELF targets.
132 \b Fix optimizations of signed bytes.
134 \b Fix operation on bigendian machines.
136 \b Fix buffer overflow in the preprocessor.
138 \b \c{SAFESEH} support for Win32, \c{IMAGEREL} for Win64 (SEH).
140 \b \c{%?} and \c{%??} to refer to the name of a macro itself. In particular,
141 \c{%idefine keyword $%?} can be used to make a keyword "disappear".
143 \b New options for dependency generation: \c{-MD}, \c{-MF},
144 \c{-MP}, \c{-MT}, \c{-MQ}.
146 \b New preprocessor directives \c{%pathsearch} and \c{%depend}; INCBIN
147 reimplemented as a macro.
149 \b \c{%include} now resolves macros in a sane manner.
151 \b \c{%substr} can now be used to get other than one-character substrings.
153 \b New type of character/string constants, using backquotes (\c{`...`}),
154 which support C-style escape sequences.
156 \b \c{%defstr} and \c{%idefstr} to stringize macro definitions before
157 creation.
159 \b Fix forward references used in \c{EQU} statements.
161 \S{cl-2.02} Version 2.02
164 \b Additional fixes for MMX operands with explicit \c{qword}, as well as
165 (hopefully) SSE operands with \c{oword}.
167 \b Fix handling of truncated strings with \c{DO}.
169 \b Fix segfaults due to memory overwrites when floating-point constants
170 were used.
172 \b Fix segfaults due to missing include files.
174 \b Fix OpenWatcom Makefiles for DOS and OS/2.
176 \b Add autogenerated instruction list back into the documentation.
178 \b ELF: Fix segfault when generating stabs, and no symbols have been
179 defined.
181 \b ELF: Experimental support for DWARF debugging information.
183 \b New compile date and time standard macros.
185 \b \c{%ifnum} now returns true for negative numbers.
187 \b New \c{%iftoken} test for a single token.
189 \b New \c{%ifempty} test for empty expansion.
191 \b Add support for the \c{XSAVE} instruction group.
193 \b Makefile for Netware/gcc.
195 \b Fix issue with some warnings getting emitted way too many times.
197 \b Autogenerated instruction list added to the documentation.
199 \S{cl-2.01} Version 2.01
202 \b Fix the handling of MMX registers with explicit \c{qword} tags on
203 memory (broken in 2.00 due to 64-bit changes.)
205 \b Fix the PREFETCH instructions.
207 \b Fix the documentation.
209 \b Fix debugging info when using \c{-f elf}
210 (backwards compatibility alias for \c{-f elf32}).
212 \b Man pages for rdoff tools (from the Debian project.)
214 \b ELF: handle large numbers of sections.
216 \b Fix corrupt output when the optimizer runs out of passes.
219 \S{cl-2.00} Version 2.00
221 \b Added c99 data-type compliance.
223 \b Added general x86-64 support.
225 \b Added win64 (x86-64 COFF) output format.
227 \b Added \c{__BITS__} standard macro.
229 \b Renamed the \c{elf} output format to \c{elf32} for clarity.
231 \b Added \c{elf64} and \c{macho} (MacOS X) output formats.
233 \b Added Numeric constants in \c{dq} directive.
235 \b Added \c{oword}, \c{do} and \c{reso} pseudo operands.
237 \b Allow underscores in numbers.
239 \b Added 8-, 16- and 128-bit floating-point formats.
241 \b Added binary, octal and hexadecimal floating-point.
243 \b Correct the generation of floating-point constants.
245 \b Added floating-point option control.
247 \b Added Infinity and NaN floating point support.
249 \b Added ELF Symbol Visibility support.
251 \b Added setting OSABI value in ELF header directive.
253 \b Added Generate Makefile Dependencies option.
255 \b Added Unlimited Optimization Passes option.
257 \b Added \c{%IFN} and \c{%ELIFN} support.
259 \b Added Logical Negation Operator.
261 \b Enhanced Stack Relative Preprocessor Directives.
263 \b Enhanced ELF Debug Formats.
265 \b Enhanced Send Errors to a File option.
267 \b Added SSSE3, SSE4.1, SSE4.2, SSE5 support.
269 \b Added a large number of additional instructions.
271 \b Significant performance improvements.
274 \H{cl-0.98.xx} NASM 0.98 Series
276 The 0.98 series was the production versions of NASM from 1999 to 2007.
279 \S{cl-0.98.39} Version 0.98.39
281 \b fix buffer overflow
283 \b fix outas86's \c{.bss} handling
285 \b "make spotless" no longer deletes config.h.in.
287 \b \c{%(el)if(n)idn} insensitivity to string quotes difference (#809300).
289 \b (nasm.c)\c{__OUTPUT_FORMAT__} changed to string value instead of symbol.
291 \S{cl-0.98.38} Version 0.98.38
294 \b Add Makefile for 16-bit DOS binaries under OpenWatcom, and modify
295 \c{mkdep.pl} to be able to generate completely pathless dependencies, as
296 required by OpenWatcom wmake (it supports path searches, but not
297 explicit paths.)
299 \b Fix the \c{STR} instruction.
301 \b Fix the ELF output format, which was broken under certain
302 circumstances due to the addition of stabs support.
304 \b Quick-fix Borland format debug-info for \c{-f obj}
306 \b Fix for \c{%rep} with no arguments (#560568)
308 \b Fix concatenation of preprocessor function call (#794686)
310 \b Fix long label causes coredump (#677841)
312 \b Use autoheader as well as autoconf to keep configure from generating
313 ridiculously long command lines.
315 \b Make sure that all of the formats which support debugging output
316 actually will suppress debugging output when \c{-g} not specified.
318 \S{cl-0.98.37} Version 0.98.37
321 \b Paths given in \c{-I} switch searched for \c{incbin}-ed as
322 well as \c{%include}-ed files.
324 \b Added stabs debugging for the ELF output format, patch from
325 Martin Wawro.
327 \b Fix \c{output/outbin.c} to allow origin > 80000000h.
329 \b Make \c{-U} switch work.
331 \b Fix the use of relative offsets with explicit prefixes, e.g.
332 \c{a32 loop foo}.
334 \b Remove \c{backslash()}.
336 \b Fix the \c{SMSW} and \c{SLDT} instructions.
338 \b \c{-O2} and \c{-O3} are no longer aliases for \c{-O10} and \c{-O15}.
339 If you mean the latter, please say so! :)
341 \S{cl-0.98.36} Version 0.98.36
344 \b Update rdoff - librarian/archiver - common rec - docs!
346 \b Fix signed/unsigned problems.
348 \b Fix \c{JMP FAR label} and \c{CALL FAR label}.
350 \b Add new multisection support - map files - fix align bug
352 \b Fix sysexit, movhps/movlps reg,reg bugs in insns.dat
354 \b \c{Q} or \c{O} suffixes indicate octal
356 \b Support Prescott new instructions (PNI).
358 \b Cyrix \c{XSTORE} instruction.
361 \S{cl-0.98.35} Version 0.98.35
363 \b Fix build failure on 16-bit DOS (Makefile.bc3 workaround for compiler bug.)
365 \b Fix dependencies and compiler warnings.
367 \b Add "const" in a number of places.
369 \b Add -X option to specify error reporting format (use -Xvc to
370 integrate with Microsoft Visual Studio.)
372 \b Minor changes for code legibility.
374 \b Drop use of tmpnam() in rdoff (security fix.)
377 \S{cl-0.98.34} Version 0.98.34
379 \b Correct additional address-size vs. operand-size confusions.
381 \b Generate dependencies for all Makefiles automatically.
383 \b Add support for unimplemented (but theoretically available)
384 registers such as tr0 and cr5. Segment registers 6 and 7 are called
385 segr6 and segr7 for the operations which they can be represented.
387 \b Correct some disassembler bugs related to redundant address-size prefixes.
388 Some work still remains in this area.
390 \b Correctly generate an error for things like "SEG eax".
392 \b Add the JMPE instruction, enabled by "CPU IA64".
394 \b Correct compilation on newer gcc/glibc platforms.
396 \b Issue an error on things like "jmp far eax".
399 \S{cl-0.98.33} Version 0.98.33
401 \b New __NASM_PATCHLEVEL__ and __NASM_VERSION_ID__ standard macros to
402 round out the version-query macros. version.pl now understands
403 X.YYplWW or X.YY.ZZplWW as a version number, equivalent to
404 X.YY.ZZ.WW (or X.YY.0.WW, as appropriate).
406 \b New keyword "strict" to disable the optimization of specific
407 operands.
409 \b Fix the handing of size overrides with JMP instructions
410 (instructions such as "jmp dword foo".)
412 \b Fix the handling of "ABSOLUTE label", where "label" points into a
413 relocatable segment.
415 \b Fix OBJ output format with lots of externs.
417 \b More documentation updates.
419 \b Add -Ov option to get verbose information about optimizations.
421 \b Undo a braindead change which broke \c{%elif} directives.
423 \b Makefile updates.
426 \S{cl-0.98.32} Version 0.98.32
428 \b Fix NASM crashing when \c{%macro} directives were left unterminated.
430 \b Lots of documentation updates.
432 \b Complete rewrite of the PostScript/PDF documentation generator.
434 \b The MS Visual C++ Makefile was updated and corrected.
436 \b Recognize .rodata as a standard section name in ELF.
438 \b Fix some obsolete Perl4-isms in Perl scripts.
440 \b Fix configure.in to work with autoconf 2.5x.
442 \b Fix a couple of "make cleaner" misses.
444 \b Make the normal "./configure && make" work with Cygwin.
447 \S{cl-0.98.31} Version 0.98.31
449 \b Correctly build in a separate object directory again.
451 \b Derive all references to the version number from the version file.
453 \b New standard macros __NASM_SUBMINOR__ and __NASM_VER__ macros.
455 \b Lots of Makefile updates and bug fixes.
457 \b New \c{%ifmacro} directive to test for multiline macros.
459 \b Documentation updates.
461 \b Fixes for 16-bit OBJ format output.
463 \b Changed the NASM environment variable to NASMENV.
466 \S{cl-0.98.30} Version 0.98.30
468 \b Changed doc files a lot: completely removed old READMExx and
469 Wishlist files, incorporating all information in CHANGES and TODO.
471 \b I waited a long time to rename zoutieee.c to (original) outieee.c
473 \b moved all output modules to output/ subdirectory.
475 \b Added 'make strip' target to strip debug info from nasm & ndisasm.
477 \b Added INSTALL file with installation instructions.
479 \b Added -v option description to nasm man.
481 \b Added dist makefile target to produce source distributions.
483 \b 16-bit support for ELF output format (GNU extension, but useful.)
486 \S{cl-0.98.28} Version 0.98.28
488 \b Fastcooked this for Debian's Woody release:
489 Frank applied the INCBIN bug patch to 0.98.25alt and called
490 it 0.98.28 to not confuse poor little apt-get.
493 \S{cl-0.98.26} Version 0.98.26
495 \b Reorganised files even better from 0.98.25alt
498 \S{cl-0.98.25alt} Version 0.98.25alt
500 \b Prettified the source tree. Moved files to more reasonable places.
502 \b Added findleak.pl script to misc/ directory.
504 \b Attempted to fix doc.
507 \S{cl-0.98.25} Version 0.98.25
509 \b Line continuation character \c{\\}.
511 \b Docs inadvertantly reverted - "dos packaging".
514 \S{cl-0.98.24p1} Version 0.98.24p1
516 \b FIXME: Someone, document this please.
519 \S{cl-0.98.24} Version 0.98.24
521 \b Documentation - Ndisasm doc added to Nasm.doc.
524 \S{cl-0.98.23} Version 0.98.23
526 \b Attempted to remove rdoff version1
528 \b Lino Mastrodomenico's patches to preproc.c (%$$ bug?).
531 \S{cl-0.98.22} Version 0.98.22
533 \b Update rdoff2 - attempt to remove v1.
536 \S{cl-0.98.21} Version 0.98.21
538 \b Optimization fixes.
541 \S{cl-0.98.20} Version 0.98.20
543 \b Optimization fixes.
546 \S{cl-0.98.19} Version 0.98.19
548 \b H. J. Lu's patch back out.
551 \S{cl-0.98.18} Version 0.98.18
553 \b Added ".rdata" to "-f win32".
556 \S{cl-0.98.17} Version 0.98.17
558 \b H. J. Lu's "bogus elf" patch. (Red Hat problem?)
561 \S{cl-0.98.16} Version 0.98.16
563 \b Fix whitespace before "[section ..." bug.
566 \S{cl-0.98.15} Version 0.98.15
568 \b Rdoff changes (?).
570 \b Fix fixes to memory leaks.
573 \S{cl-0.98.14} Version 0.98.14
575 \b Fix memory leaks.
578 \S{cl-0.98.13} Version 0.98.13
580 \b There was no 0.98.13
583 \S{cl-0.98.12} Version 0.98.12
585 \b Update optimization (new function of "-O1")
587 \b Changes to test/bintest.asm (?).
590 \S{cl-0.98.11} Version 0.98.11
592 \b Optimization changes.
594 \b Ndisasm fixed.
597 \S{cl-0.98.10} Version 0.98.10
599 \b There was no 0.98.10
602 \S{cl-0.98.09} Version 0.98.09
604 \b Add multiple sections support to "-f bin".
606 \b Changed GLOBAL_TEMP_BASE in outelf.c from 6 to 15.
608 \b Add "-v" as an alias to the "-r" switch.
610 \b Remove "#ifdef" from Tasm compatibility options.
612 \b Remove redundant size-overrides on "mov ds, ex", etc.
614 \b Fixes to SSE2, other insns.dat (?).
616 \b Enable uppercase "I" and "P" switches.
618 \b Case insinsitive "seg" and "wrt".
620 \b Update install.sh (?).
622 \b Allocate tokens in blocks.
624 \b Improve "invalid effective address" messages.
627 \S{cl-0.98.08} Version 0.98.08
629 \b Add "\c{%strlen}" and "\c{%substr}" macro operators
631 \b Fixed broken c16.mac.
633 \b Unterminated string error reported.
635 \b Fixed bugs as per 0.98bf
638 \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
640 Changes from 0.98.07 release to 98.09b as of 28-Oct-2001
642 \b More closely compatible with 0.98 when -O0 is implied
643 or specified. Not strictly identical, since backward
644 branches in range of short offsets are recognized, and signed
645 byte values with no explicit size specification will be
646 assembled as a single byte.
648 \b More forgiving with the PUSH instruction. 0.98 requires
649 a size to be specified always. 0.98.09b will imply the size
650 from the current BITS setting (16 or 32).
652 \b Changed definition of the optimization flag:
654 -O0 strict two-pass assembly, JMP and Jcc are
655 handled more like 0.98, except that back-
656 ward JMPs are short, if possible.
658 -O1 strict two-pass assembly, but forward
659 branches are assembled with code guaranteed
660 to reach; may produce larger code than
661 -O0, but will produce successful assembly
662 more often if branch offset sizes are not
663 specified.
665 -O2 multi-pass optimization, minimize branch
666 offsets; also will minimize signed immed-
667 iate bytes, overriding size specification.
669 -O3 like -O2, but more passes taken, if needed
672 \S{cl-0.98.07 released 01/28/01} Version 0.98.07 released 01/28/01
674 \b Added Stepane Denis' SSE2 instructions to a *working*
675 version of the code - some earlier versions were based on
676 broken code - sorry 'bout that. version "0.98.07"
679 01/28/01
682 \b Cosmetic modifications to nasm.c, nasm.h,
683 AUTHORS, MODIFIED
686 \S{cl-0.98.06f released 01/18/01} Version 0.98.06f released 01/18/01
689 \b - Add "metalbrain"s jecxz bug fix in insns.dat
690 - alter nasmdoc.src to match - version "0.98.06f"
693 \S{cl-0.98.06e released 01/09/01} Version 0.98.06e released 01/09/01
696 \b Removed the "outforms.h" file - it appears to be
697 someone's old backup of "outform.h". version "0.98.06e"
699 01/09/01
701 \b fbk - finally added the fix for the "multiple %includes bug",
702 known since 7/27/99 - reported originally (?) and sent to
703 us by Austin Lunnen - he reports that John Fine had a fix
704 within the day. Here it is...
706 \b Nelson Rush resigns from the group. Big thanks to Nelson for
707 his leadership and enthusiasm in getting these changes
708 incorporated into Nasm!
710 \b fbk - [list +], [list -] directives - ineptly implemented, should
711 be re-written or removed, perhaps.
713 \b Brian Raiter / fbk - "elfso bug" fix - applied to aoutb format
714 as well - testing might be desirable...
716 08/07/00
718 \b James Seter - -postfix, -prefix command line switches.
720 \b Yuri Zaporogets - rdoff utility changes.
723 \S{cl-0.98p1} Version 0.98p1
725 \b GAS-like palign (Panos Minos)
727 \b FIXME: Someone, fill this in with details
730 \S{cl-0.98bf (bug-fixed)} Version 0.98bf (bug-fixed)
732 \b Fixed - elf and aoutb bug - shared libraries
733 - multiple "%include" bug in "-f obj"
734 - jcxz, jecxz bug
735 - unrecognized option bug in ndisasm
737 \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
739 \b Added signed byte optimizations for the 0x81/0x83 class
740 of instructions: ADC, ADD, AND, CMP, OR, SBB, SUB, XOR:
741 when used as 'ADD reg16,imm' or 'ADD reg32,imm.' Also
742 optimization of signed byte form of 'PUSH imm' and 'IMUL
743 reg,imm'/'IMUL reg,reg,imm.' No size specification is needed.
745 \b Added multi-pass JMP and Jcc offset optimization. Offsets
746 on forward references will preferentially use the short form,
747 without the need to code a specific size (short or near) for
748 the branch. Added instructions for 'Jcc label' to use the
749 form 'Jnotcc $+3/JMP label', in cases where a short offset
750 is out of bounds. If compiling for a 386 or higher CPU, then
751 the 386 form of Jcc will be used instead.
753 This feature is controlled by a new command-line switch: "O",
754 (upper case letter O). "-O0" reverts the assembler to no
755 extra optimization passes, "-O1" allows up to 5 extra passes,
756 and "-O2"(default), allows up to 10 extra optimization passes.
758 \b Added a new directive: 'cpu XXX', where XXX is any of:
759 8086, 186, 286, 386, 486, 586, pentium, 686, PPro, P2, P3 or
760 Katmai. All are case insensitive. All instructions will
761 be selected only if they apply to the selected cpu or lower.
762 Corrected a couple of bugs in cpu-dependence in 'insns.dat'.
764 \b Added to 'standard.mac', the "use16" and "use32" forms of
765 the "bits 16/32" directive. This is nothing new, just conforms
766 to a lot of other assemblers. (minor)
768 \b Changed label allocation from 320/32 (10000 labels @ 200K+)
769 to 32/37 (1000 labels); makes running under DOS much easier.
770 Since additional label space is allocated dynamically, this
771 should have no effect on large programs with lots of labels.
772 The 37 is a prime, believed to be better for hashing. (minor)
775 \S{cl-0.98.03} Version 0.98.03
777 "Integrated patchfile 0.98-0.98.01. I call this version 0.98.03 for
778 historical reasons: 0.98.02 was trashed." --John Coffman
779 <johninsd@san.rr.com>, 27-Jul-2000
781 \b Kendall Bennett's SciTech MGL changes
783 \b Note that you must define "TASM_COMPAT" at compile-time
784 to get the Tasm Ideal Mode compatibility.
786 \b All changes can be compiled in and out using the TASM_COMPAT macros,
787 and when compiled without TASM_COMPAT defined we get the exact same
788 binary as the unmodified 0.98 sources.
790 \b standard.mac, macros.c: Added macros to ignore TASM directives before
791 first include
793 \b nasm.h: Added extern declaration for tasm_compatible_mode
795 \b nasm.c: Added global variable tasm_compatible_mode
797 \b Added command line switch for TASM compatible mode (-t)
799 \b Changed version command line to reflect when compiled with TASM additions
801 \b Added response file processing to allow all arguments on a single
802 line (response file is @resp rather than -@resp for NASM format).
804 \b labels.c: Changes islocal() macro to support TASM style @@local labels.
806 \b Added islocalchar() macro to support TASM style @@local labels.
808 \b parser.c: Added support for TASM style memory references (ie: mov
809 [DWORD eax],10 rather than the NASM style mov DWORD [eax],10).
811 \b preproc.c: Added new directives, \c{%arg}, \c{%local}, \c{%stacksize} to directives
812 table
814 \b Added support for TASM style directives without a leading % symbol.
816 \b Integrated a block of changes from Andrew Zabolotny <bit@eltech.ru>:
818 \b A new keyword \c{%xdefine} and its case-insensitive counterpart \c{%ixdefine}.
819 They work almost the same way as \c{%define} and \c{%idefine} but expand
820 the definition immediately, not on the invocation. Something like a cross
821 between \c{%define} and \c{%assign}. The "x" suffix stands for "eXpand", so
822 "xdefine" can be deciphered as "expand-and-define". Thus you can do
823 things like this:
825 \c %assign ofs 0
826 \c
827 \c %macro arg 1
828 \c %xdefine %1 dword [esp+ofs]
829 \c %assign ofs ofs+4
830 \c %endmacro
832 \b Changed the place where the expansion of %$name macros are expanded.
833 Now they are converted into ..@ctxnum.name form when detokenizing, so
834 there are no quirks as before when using %$name arguments to macros,
835 in macros etc. For example:
837 \c %macro abc 1
838 \c %define %1 hello
839 \c %endm
840 \c
841 \c abc %$here
842 \c %$here
844 Now last line will be expanded into "hello" as expected. This also allows
845 for lots of goodies, a good example are extended "proc" macros included
846 in this archive.
848 \b Added a check for "cstk" in smacro_defined() before calling get_ctx() -
849 this allows for things like:
851 \c %ifdef %$abc
852 \c %endif
854 to work without warnings even in no context.
856 \b Added a check for "cstk" in %if*ctx and %elif*ctx directives -
857 this allows to use \c{%ifctx} without excessive warnings. If there is
858 no active context, \c{%ifctx} goes through "false" branch.
860 \b Removed "user error: " prefix with \c{%error} directive: it just clobbers the
861 output and has absolutely no functionality. Besides, this allows to write
862 macros that does not differ from built-in functions in any way.
864 \b Added expansion of string that is output by \c{%error} directive. Now you
865 can do things like:
867 \c %define hello(x) Hello, x!
868 \c
869 \c %define %$name andy
870 \c %error "hello(%$name)"
872 Same happened with \c{%include} directive.
874 \b Now all directives that expect an identifier will try to expand and
875 concatenate everything without whitespaces in between before usage.
876 For example, with "unfixed" nasm the commands
878 \c %define %$abc hello
879 \c %define __%$abc goodbye
880 \c __%$abc
882 would produce "incorrect" output: last line will expand to
884 \c hello goodbyehello
886 Not quite what you expected, eh? :-) The answer is that preprocessor
887 treats the \c{%define} construct as if it would be
889 \c %define __ %$abc goodbye
891 (note the white space between __ and %$abc). After my "fix" it
892 will "correctly" expand into
894 \c goodbye
896 as expected. Note that I use quotes around words "correct", "incorrect"
897 etc because this is rather a feature not a bug; however current behaviour
898 is more logical (and allows more advanced macro usage :-).
900 Same change was applied to:
901 \c{%push},\c{%macro},\c{%imacro},\c{%define},\c{%idefine},\c{%xdefine},\c{%ixdefine},
902 \c{%assign},\c{%iassign},\c{%undef}
904 \b A new directive [WARNING {+|-}warning-id] have been added. It works only
905 if the assembly phase is enabled (i.e. it doesn't work with nasm -e).
907 \b A new warning type: macro-selfref. By default this warning is disabled;
908 when enabled NASM warns when a macro self-references itself; for example
909 the following source:
911 \c [WARNING macro-selfref]
912 \c
913 \c %macro push 1-*
914 \c %rep %0
915 \c push %1
916 \c %rotate 1
917 \c %endrep
918 \c %endmacro
919 \c
920 \c push eax,ebx,ecx
922 will produce a warning, but if we remove the first line we won't see it
923 anymore (which is The Right Thing To Do {tm} IMHO since C preprocessor
924 eats such constructs without warnings at all).
926 \b Added a "error" routine to preprocessor which always will set ERR_PASS1
927 bit in severity_code. This removes annoying repeated errors on first
928 and second passes from preprocessor.
930 \b Added the %+ operator in single-line macros for concatenating two
931 identifiers. Usage example:
933 \c %define _myfunc _otherfunc
934 \c %define cextern(x) _ %+ x
935 \c cextern (myfunc)
937 After first expansion, third line will become "_myfunc". After this
938 expansion is performed again so it becomes "_otherunc".
940 \b Now if preprocessor is in a non-emitting state, no warning or error
941 will be emitted. Example:
943 \c %if 1
944 \c mov eax,ebx
945 \c %else
946 \c put anything you want between these two brackets,
947 \c even macro-parameter references %1 or local
948 \c labels %$zz or macro-local labels %%zz - no
949 \c warning will be emitted.
950 \c %endif
952 \b Context-local variables on expansion as a last resort are looked up
953 in outer contexts. For example, the following piece:
955 \c %push outer
956 \c %define %$a [esp]
957 \c
958 \c %push inner
959 \c %$a
960 \c %pop
961 \c %pop
963 will expand correctly the fourth line to [esp]; if we'll define another
964 %$a inside the "inner" context, it will take precedence over outer
965 definition. However, this modification has been applied only to
966 expand_smacro and not to smacro_define: as a consequence expansion
967 looks in outer contexts, but \c{%ifdef} won't look in outer contexts.
969 This behaviour is needed because we don't want nested contexts to
970 act on already defined local macros. Example:
972 \c %define %$arg1 [esp+4]
973 \c test eax,eax
974 \c if nz
975 \c mov eax,%$arg1
976 \c endif
978 In this example the "if" mmacro enters into the "if" context, so %$arg1
979 is not valid anymore inside "if". Of course it could be worked around
980 by using explicitely %$$arg1 but this is ugly IMHO.
982 \b Fixed memory leak in \c{%undef}. The origline wasn't freed before
983 exiting on success.
985 \b Fixed trap in preprocessor when line expanded to empty set of tokens.
986 This happens, for example, in the following case:
988 \c #define SOMETHING
989 \c SOMETHING
992 \S{cl-0.98} Version 0.98
994 All changes since NASM 0.98p3 have been produced by H. Peter Anvin <hpa@zytor.com>.
996 \b The documentation comment delimiter is \# not #.
998 \b Allow EQU definitions to refer to external labels; reported by
999 Pedro Gimeno.
1001 \b Re-enable support for RDOFF v1; reported by Pedro Gimeno.
1003 \b Updated License file per OK from Simon and Julian.
1006 \S{cl-0.98p9} Version 0.98p9
1008 \b Update documentation (although the instruction set reference will
1009 have to wait; I don't want to hold up the 0.98 release for it.)
1011 \b Verified that the NASM implementation of the PEXTRW and PMOVMSKB
1012 instructions is correct. The encoding differs from what the Intel
1013 manuals document, but the Pentium III behaviour matches NASM, not
1014 the Intel manuals.
1016 \b Fix handling of implicit sizes in PSHUFW and PINSRW, reported by
1017 Stefan Hoffmeister.
1019 \b Resurrect the -s option, which was removed when changing the
1020 diagnostic output to stdout.
1023 \S{cl-0.98p8} Version 0.98p8
1025 \b Fix for "DB" when NASM is running on a bigendian machine.
1027 \b Invoke insns.pl once for each output script, making Makefile.in
1028 legal for "make -j".
1030 \b Improve the Unix configure-based makefiles to make package
1031 creation easier.
1033 \b Included an RPM .spec file for building RPM (RedHat Package Manager)
1034 packages on Linux or Unix systems.
1036 \b Fix Makefile dependency problems.
1038 \b Change src/rdsrc.pl to include sectioning information in info
1039 output; required for install-info to work.
1041 \b Updated the RDOFF distribution to version 2 from Jules; minor
1042 massaging to make it compile in my environment.
1044 \b Split doc files that can be built by anyone with a Perl interpreter off
1045 into a separate archive.
1047 \b "Dress rehearsal" release!
1050 \S{cl-0.98p7} Version 0.98p7
1052 \b Fixed opcodes with a third byte-sized immediate argument to not
1053 complain if given "byte" on the immediate.
1055 \b Allow \c{%undef} to remove single-line macros with arguments. This
1056 matches the behaviour of #undef in the C preprocessor.
1058 \b Allow -d, -u, -i and -p to be specified as -D, -U, -I and -P for
1059 compatibility with most C compilers and preprocessors. This allows
1060 Makefile options to be shared between cc and nasm, for example.
1062 \b Minor cleanups.
1064 \b Went through the list of Katmai instructions and hopefully fixed the
1065 (rather few) mistakes in it.
1067 \b (Hopefully) fixed a number of disassembler bugs related to ambiguous
1068 instructions (disambiguated by -p) and SSE instructions with REP.
1070 \b Fix for bug reported by Mark Junger: "call dword 0x12345678" should
1071 work and may add an OSP (affected CALL, JMP, Jcc).
1073 \b Fix for environments when "stderr" isn't a compile-time constant.
1076 \S{cl-0.98p6} Version 0.98p6
1079 \b Took officially over coordination of the 0.98 release; so drop
1080 the p3.x notation. Skipped p4 and p5 to avoid confusion with John
1081 Fine's J4 and J5 releases.
1083 \b Update the documentation; however, it still doesn't include
1084 documentation for the various new instructions. I somehow wonder if
1085 it makes sense to have an instruction set reference in the assembler
1086 manual when Intel et al have PDF versions of their manuals online.
1088 \b Recognize "idt" or "centaur" for the -p option to ndisasm.
1090 \b Changed error messages back to stderr where they belong, but add an
1091 -E option to redirect them elsewhere (the DOS shell cannot redirect
1092 stderr.)
1094 \b -M option to generate Makefile dependencies (based on code from Alex
1095 Verstak.)
1097 \b \c{%undef} preprocessor directive, and -u option, that undefines a
1098 single-line macro.
1100 \b OS/2 Makefile (Mkfiles/Makefile.os2) for Borland under OS/2; from
1101 Chuck Crayne.
1103 \b Various minor bugfixes (reported by):
1104 - Dangling \c{%s} in preproc.c (Martin Junker)
1106 \b THERE ARE KNOWN BUGS IN SSE AND THE OTHER KATMAI INSTRUCTIONS. I am
1107 on a trip and didn't bring the Katmai instruction reference, so I
1108 can't work on them right now.
1110 \b Updated the License file per agreement with Simon and Jules to
1111 include a GPL distribution clause.
1114 \S{cl-0.98p3.7} Version 0.98p3.7
1116 \b (Hopefully) fixed the canned Makefiles to include the outrdf2 and
1117 zoutieee modules.
1119 \b Renamed changes.asm to changed.asm.
1122 \S{cl-0.98p3.6} Version 0.98p3.6
1124 \b Fixed a bunch of instructions that were added in 0.98p3.5 which had
1125 memory operands, and the address-size prefix was missing from the
1126 instruction pattern.
1129 \S{cl-0.98p3.5} Version 0.98p3.5
1131 \b Merged in changes from John S. Fine's 0.98-J5 release. John's based
1132 0.98-J5 on my 0.98p3.3 release; this merges the changes.
1134 \b Expanded the instructions flag field to a long so we can fit more
1135 flags; mark SSE (KNI) and AMD or Katmai-specific instructions as
1136 such.
1138 \b Fix the "PRIV" flag on a bunch of instructions, and create new
1139 "PROT" flag for protected-mode-only instructions (orthogonal to if
1140 the instruction is privileged!) and new "SMM" flag for SMM-only
1141 instructions.
1143 \b Added AMD-only SYSCALL and SYSRET instructions.
1145 \b Make SSE actually work, and add new Katmai MMX instructions.
1147 \b Added a -p (preferred vendor) option to ndisasm so that it can
1148 distinguish e.g. Cyrix opcodes also used in SSE. For example:
1150 \c ndisasm -p cyrix aliased.bin
1151 \c 00000000 670F514310 paddsiw mm0,[ebx+0x10]
1152 \c 00000005 670F514320 paddsiw mm0,[ebx+0x20]
1153 \c ndisasm -p intel aliased.bin
1154 \c 00000000 670F514310 sqrtps xmm0,[ebx+0x10]
1155 \c 00000005 670F514320 sqrtps xmm0,[ebx+0x20]
1157 \b Added a bunch of Cyrix-specific instructions.
1160 \S{cl-0.98p3.4} Version 0.98p3.4
1162 \b Made at least an attempt to modify all the additional Makefiles (in
1163 the Mkfiles directory). I can't test it, but this was the best I
1164 could do.
1166 \b DOS DJGPP+"Opus Make" Makefile from John S. Fine.
1168 \b changes.asm changes from John S. Fine.
1171 \S{cl-0.98p3.3} Version 0.98p3.3
1173 \b Patch from Conan Brink to allow nesting of \c{%rep} directives.
1175 \b If we're going to allow INT01 as an alias for INT1/ICEBP (one of
1176 Jules 0.98p3 changes), then we should allow INT03 as an alias for INT3
1177 as well.
1179 \b Updated changes.asm to include the latest changes.
1181 \b Tried to clean up the <CR>s that had snuck in from a DOS/Windows
1182 environment into my Unix environment, and try to make sure than
1183 DOS/Windows users get them back.
1185 \b We would silently generate broken tools if insns.dat wasn't sorted
1186 properly. Change insns.pl so that the order doesn't matter.
1188 \b Fix bug in insns.pl (introduced by me) which would cause conditional
1189 instructions to have an extra "cc" in disassembly, e.g. "jnz"
1190 disassembled as "jccnz".
1193 \S{cl-0.98p3.2} Version 0.98p3.2
1195 \b Merged in John S. Fine's changes from his 0.98-J4 prerelease; see
1196 http://www.csoft.net/cz/johnfine/
1198 \b Changed previous "spotless" Makefile target (appropriate for distribution)
1199 to "distclean", and added "cleaner" target which is same as "clean"
1200 except deletes files generated by Perl scripts; "spotless" is union.
1202 \b Removed BASIC programs from distribution. Get a Perl interpreter
1203 instead (see below.)
1205 \b Calling this "pre-release 3.2" rather than "p3-hpa2" because of
1206 John's contributions.
1208 \b Actually link in the IEEE output format (zoutieee.c); fix a bunch of
1209 compiler warnings in that file. Note I don't know what IEEE output
1210 is supposed to look like, so these changes were made "blind".
1213 \S{cl-0.98p3-hpa} Version 0.98p3-hpa
1215 \b Merged nasm098p3.zip with nasm-0.97.tar.gz to create a fully
1216 buildable version for Unix systems (Makefile.in updates, etc.)
1218 \b Changed insns.pl to create the instruction tables in nasm.h and
1219 names.c, so that a new instruction can be added by adding it *only*
1220 to insns.dat.
1222 \b Added the following new instructions: SYSENTER, SYSEXIT, FXSAVE,
1223 FXRSTOR, UD1, UD2 (the latter two are two opcodes that Intel
1224 guarantee will never be used; one of them is documented as UD2 in
1225 Intel documentation, the other one just as "Undefined Opcode" --
1226 calling it UD1 seemed to make sense.)
1228 \b MAX_SYMBOL was defined to be 9, but LOADALL286 and LOADALL386 are 10
1229 characters long. Now MAX_SYMBOL is derived from insns.dat.
1231 \b A note on the BASIC programs included: forget them. insns.bas is
1232 already out of date. Get yourself a Perl interpreter for your
1233 platform of choice at
1234 \W{http://www.cpan.org/ports/index.html}{http://www.cpan.org/ports/index.html}.
1237 \S{cl-0.98p3} Version 0.98 pre-release 3
1239 \b added response file support, improved command line handling, new layout
1240 help screen
1242 \b fixed limit checking bug, 'OUT byte nn, reg' bug, and a couple of rdoff
1243 related bugs, updated Wishlist; 0.98 Prerelease 3.
1246 \S{cl-0.98p2} Version 0.98 pre-release 2
1248 \b fixed bug in outcoff.c to do with truncating section names longer
1249 than 8 characters, referencing beyond end of string; 0.98 pre-release 2
1252 \S{cl-0.98p1} Version 0.98 pre-release 1
1254 \b Fixed a bug whereby STRUC didn't work at all in RDF.
1256 \b Fixed a problem with group specification in PUBDEFs in OBJ.
1258 \b Improved ease of adding new output formats. Contribution due to
1259 Fox Cutter.
1261 \b Fixed a bug in relocations in the `bin' format: was showing up when
1262 a relocatable reference crossed an 8192-byte boundary in any output
1263 section.
1265 \b Fixed a bug in local labels: local-label lookups were inconsistent
1266 between passes one and two if an EQU occurred between the definition
1267 of a global label and the subsequent use of a local label local to
1268 that global.
1270 \b Fixed a seg-fault in the preprocessor (again) which happened when
1271 you use a blank line as the first line of a multi-line macro
1272 definition and then defined a label on the same line as a call to
1273 that macro.
1275 \b Fixed a stale-pointer bug in the handling of the NASM environment
1276 variable. Thanks to Thomas McWilliams.
1278 \b ELF had a hard limit on the number of sections which caused
1279 segfaults when transgressed. Fixed.
1281 \b Added ability for ndisasm to read from stdin by using `-' as the
1282 filename.
1284 \b ndisasm wasn't outputting the TO keyword. Fixed.
1286 \b Fixed error cascade on bogus expression in \c{%if} - an error in
1287 evaluation was causing the entire \c{%if} to be discarded, thus creating
1288 trouble later when the \c{%else} or \c{%endif} was encountered.
1290 \b Forward reference tracking was instruction-granular not operand-
1291 granular, which was causing 286-specific code to be generated
1292 needlessly on code of the form `shr word [forwardref],1'. Thanks to
1293 Jim Hague for sending a patch.
1295 \b All messages now appear on stdout, as sending them to stderr serves
1296 no useful purpose other than to make redirection difficult.
1298 \b Fixed the problem with EQUs pointing to an external symbol - this
1299 now generates an error message.
1301 \b Allowed multiple size prefixes to an operand, of which only the first
1302 is taken into account.
1304 \b Incorporated John Fine's changes, including fixes of a large number
1305 of preprocessor bugs, some small problems in OBJ, and a reworking of
1306 label handling to define labels before their line is assembled, rather
1307 than after.
1309 \b Reformatted a lot of the source code to be more readable. Included
1310 'coding.txt' as a guideline for how to format code for contributors.
1312 \b Stopped nested \c{%reps} causing a panic - they now cause a slightly more
1313 friendly error message instead.
1315 \b Fixed floating point constant problems (patch by Pedro Gimeno)
1317 \b Fixed the return value of insn_size() not being checked for -1, indicating
1318 an error.
1320 \b Incorporated 3Dnow! instructions.
1322 \b Fixed the 'mov eax, eax + ebx' bug.
1324 \b Fixed the GLOBAL EQU bug in ELF. Released developers release 3.
1326 \b Incorporated John Fine's command line parsing changes
1328 \b Incorporated David Lindauer's OMF debug support
1330 \b Made changes for LCC 4.0 support (\c{__NASM_CDecl__}, removed register size
1331 specification warning when sizes agree).
1334 \H{cl-0.9x} NASM 0.9 Series
1336 Revisions before 0.98.
1339 \S{cl-0.97} Version 0.97 released December 1997
1341 \b This was entirely a bug-fix release to 0.96, which seems to have got
1342 cursed. Silly me.
1344 \b Fixed stupid mistake in OBJ which caused `MOV EAX,<constant>' to
1345 fail. Caused by an error in the `MOV EAX,<segment>' support.
1347 \b ndisasm hung at EOF when compiled with lcc on Linux because lcc on
1348 Linux somehow breaks feof(). ndisasm now does not rely on feof().
1350 \b A heading in the documentation was missing due to a markup error in
1351 the indexing. Fixed.
1353 \b Fixed failure to update all pointers on realloc() within extended-
1354 operand code in parser.c. Was causing wrong behaviour and seg faults
1355 on lines such as `dd 0.0,0.0,0.0,0.0,...'
1357 \b Fixed a subtle preprocessor bug whereby invoking one multi-line
1358 macro on the first line of the expansion of another, when the second
1359 had been invoked with a label defined before it, didn't expand the
1360 inner macro.
1362 \b Added internal.doc back in to the distribution archives - it was
1363 missing in 0.96 *blush*
1365 \b Fixed bug causing 0.96 to be unable to assemble its own test files,
1366 specifically objtest.asm. *blush again*
1368 \b Fixed seg-faults and bogus error messages caused by mismatching
1369 \c{%rep} and \c{%endrep} within multi-line macro definitions.
1371 \b Fixed a problem with buffer overrun in OBJ, which was causing
1372 corruption at ends of long PUBDEF records.
1374 \b Separated DOS archives into main-program and documentation to reduce
1375 download size.
1378 \S{cl-0.96} Version 0.96 released November 1997
1380 \b Fixed a bug whereby, if `nasm sourcefile' would cause a filename
1381 collision warning and put output into `nasm.out', then `nasm
1382 sourcefile -o outputfile' still gave the warning even though the
1383 `-o' was honoured.
1384 Fixed name pollution under Digital UNIX: one of its header files
1385 defined R_SP, which broke the enum in nasm.h.
1387 \b Fixed minor instruction table problems: FUCOM and FUCOMP didn't have
1388 two-operand forms; NDISASM didn't recognise the longer register
1389 forms of PUSH and POP (eg FF F3 for PUSH BX); TEST mem,imm32 was
1390 flagged as undocumented; the 32-bit forms of CMOV had 16-bit operand
1391 size prefixes; `AAD imm' and `AAM imm' are no longer flagged as
1392 undocumented because the Intel Architecture reference documents
1393 them.
1395 \b Fixed a problem with the local-label mechanism, whereby strange
1396 types of symbol (EQUs, auto-defined OBJ segment base symbols)
1397 interfered with the `previous global label' value and screwed up
1398 local labels.
1400 \b Fixed a bug whereby the stub preprocessor didn't communicate with
1401 the listing file generator, so that the -a and -l options in
1402 conjunction would produce a useless listing file.
1404 \b Merged `os2' object file format back into `obj', after discovering
1405 that `obj' _also_ shouldn't have a link pass separator in a module
1406 containing a non-trivial MODEND. Flat segments are now declared
1407 using the FLAT attribute. `os2' is no longer a valid object format
1408 name: use `obj'.
1410 \b Removed the fixed-size temporary storage in the evaluator. Very very
1411 long expressions (like `mov ax,1+1+1+1+...' for two hundred 1s or
1412 so) should now no longer crash NASM.
1414 \b Fixed a bug involving segfaults on disassembly of MMX instructions,
1415 by changing the meaning of one of the operand-type flags in nasm.h.
1416 This may cause other apparently unrelated MMX problems; it needs to
1417 be tested thoroughly.
1419 \b Fixed some buffer overrun problems with large OBJ output files.
1420 Thanks to DJ Delorie for the bug report and fix.
1422 \b Made preprocess-only mode actually listen to the \c{%line} markers as it
1423 prints them, so that it can report errors more sanely.
1425 \b Re-designed the evaluator to keep more sensible track of expressions
1426 involving forward references: can now cope with previously-nightmare
1427 situations such as:
1429 \c mov ax,foo | bar
1430 \c foo equ 1
1431 \c bar equ 2
1433 \b Added the ALIGN and ALIGNB standard macros.
1435 \b Added PIC support in ELF: use of WRT to obtain the four extra
1436 relocation types needed.
1438 \b Added the ability for output file formats to define their own
1439 extensions to the GLOBAL, COMMON and EXTERN directives.
1441 \b Implemented common-variable alignment, and global-symbol type and
1442 size declarations, in ELF.
1444 \b Implemented NEAR and FAR keywords for common variables, plus
1445 far-common element size specification, in OBJ.
1447 \b Added a feature whereby EXTERNs and COMMONs in OBJ can be given a
1448 default WRT specification (either a segment or a group).
1450 \b Transformed the Unix NASM archive into an auto-configuring package.
1452 \b Added a sanity-check for people applying SEG to things which are
1453 already segment bases: this previously went unnoticed by the SEG
1454 processing and caused OBJ-driver panics later.
1456 \b Added the ability, in OBJ format, to deal with `MOV EAX,<segment>'
1457 type references: OBJ doesn't directly support dword-size segment
1458 base fixups, but as long as the low two bytes of the constant term
1459 are zero, a word-size fixup can be generated instead and it will
1460 work.
1462 \b Added the ability to specify sections' alignment requirements in
1463 Win32 object files and pure binary files.
1465 \b Added preprocess-time expression evaluation: the \c{%assign} (and
1466 \c{%iassign}) directive and the bare \c{%if} (and \c{%elif}) conditional. Added
1467 relational operators to the evaluator, for use only in \c{%if}
1468 constructs: the standard relationals = < > <= >= <> (and C-like
1469 synonyms == and !=) plus low-precedence logical operators &&, ^^ and
1470 ||.
1472 \b Added a preprocessor repeat construct: \c{%rep} / \c{%exitrep} / \c{%endrep}.
1474 \b Added the __FILE__ and __LINE__ standard macros.
1476 \b Added a sanity check for number constants being greater than
1477 0xFFFFFFFF. The warning can be disabled.
1479 \b Added the %0 token whereby a variadic multi-line macro can tell how
1480 many parameters it's been given in a specific invocation.
1482 \b Added \c{%rotate}, allowing multi-line macro parameters to be cycled.
1484 \b Added the `*' option for the maximum parameter count on multi-line
1485 macros, allowing them to take arbitrarily many parameters.
1487 \b Added the ability for the user-level forms of EXTERN, GLOBAL and
1488 COMMON to take more than one argument.
1490 \b Added the IMPORT and EXPORT directives in OBJ format, to deal with
1491 Windows DLLs.
1493 \b Added some more preprocessor \c{%if} constructs: \c{%ifidn} / \c{%ifidni} (exact
1494 textual identity), and \c{%ifid} / \c{%ifnum} / \c{%ifstr} (token type testing).
1496 \b Added the ability to distinguish SHL AX,1 (the 8086 version) from
1497 SHL AX,BYTE 1 (the 286-and-upwards version whose constant happens to
1498 be 1).
1500 \b Added NetBSD/FreeBSD/OpenBSD's variant of a.out format, complete
1501 with PIC shared library features.
1503 \b Changed NASM's idiosyncratic handling of FCLEX, FDISI, FENI, FINIT,
1504 FSAVE, FSTCW, FSTENV, and FSTSW to bring it into line with the
1505 otherwise accepted standard. The previous behaviour, though it was a
1506 deliberate feature, was a deliberate feature based on a
1507 misunderstanding. Apologies for the inconvenience.
1509 \b Improved the flexibility of ABSOLUTE: you can now give it an
1510 expression rather than being restricted to a constant, and it can
1511 take relocatable arguments as well.
1513 \b Added the ability for a variable to be declared as EXTERN multiple
1514 times, and the subsequent definitions are just ignored.
1516 \b We now allow instruction prefixes (CS, DS, LOCK, REPZ etc) to be
1517 alone on a line (without a following instruction).
1519 \b Improved sanity checks on whether the arguments to EXTERN, GLOBAL
1520 and COMMON are valid identifiers.
1522 \b Added misc/exebin.mac to allow direct generation of .EXE files by
1523 hacking up an EXE header using DB and DW; also added test/binexe.asm
1524 to demonstrate the use of this. Thanks to Yann Guidon for
1525 contributing the EXE header code.
1527 \b ndisasm forgot to check whether the input file had been successfully
1528 opened. Now it does. Doh!
1530 \b Added the Cyrix extensions to the MMX instruction set.
1532 \b Added a hinting mechanism to allow [EAX+EBX] and [EBX+EAX] to be
1533 assembled differently. This is important since [ESI+EBP] and
1534 [EBP+ESI] have different default base segment registers.
1536 \b Added support for the PharLap OMF extension for 4096-byte segment
1537 alignment.
1540 \S{cl-0.95 released July 1997} Version 0.95 released July 1997
1542 \b Fixed yet another ELF bug. This one manifested if the user relied on
1543 the default segment, and attempted to define global symbols without
1544 first explicitly declaring the target segment.
1546 \b Added makefiles (for NASM and the RDF tools) to build Win32 console
1547 apps under Symantec C++. Donated by Mark Junker.
1549 \b Added `macros.bas' and `insns.bas', QBasic versions of the Perl
1550 scripts that convert `standard.mac' to `macros.c' and convert
1551 `insns.dat' to `insnsa.c' and `insnsd.c'. Also thanks to Mark
1552 Junker.
1554 \b Changed the diassembled forms of the conditional instructions so
1555 that JB is now emitted as JC, and other similar changes. Suggested
1556 list by Ulrich Doewich.
1558 \b Added `@' to the list of valid characters to begin an identifier
1559 with.
1561 \b Documentary changes, notably the addition of the `Common Problems'
1562 section in nasm.doc.
1564 \b Fixed a bug relating to 32-bit PC-relative fixups in OBJ.
1566 \b Fixed a bug in perm_copy() in labels.c which was causing exceptions
1567 in cleanup_labels() on some systems.
1569 \b Positivity sanity check in TIMES argument changed from a warning to
1570 an error following a further complaint.
1572 \b Changed the acceptable limits on byte and word operands to allow
1573 things like `~10111001b' to work.
1575 \b Fixed a major problem in the preprocessor which caused seg-faults if
1576 macro definitions contained blank lines or comment-only lines.
1578 \b Fixed inadequate error checking on the commas separating the
1579 arguments to `db', `dw' etc.
1581 \b Fixed a crippling bug in the handling of macros with operand counts
1582 defined with a `+' modifier.
1584 \b Fixed a bug whereby object file formats which stored the input file
1585 name in the output file (such as OBJ and COFF) weren't doing so
1586 correctly when the output file name was specified on the command
1587 line.
1589 \b Removed [INC] and [INCLUDE] support for good, since they were
1590 obsolete anyway.
1592 \b Fixed a bug in OBJ which caused all fixups to be output in 16-bit
1593 (old-format) FIXUPP records, rather than putting the 32-bit ones in
1594 FIXUPP32 (new-format) records.
1596 \b Added, tentatively, OS/2 object file support (as a minor variant on
1597 OBJ).
1599 \b Updates to Fox Cutter's Borland C makefile, Makefile.bc2.
1601 \b Removed a spurious second fclose() on the output file.
1603 \b Added the `-s' command line option to redirect all messages which
1604 would go to stderr (errors, help text) to stdout instead.
1606 \b Added the `-w' command line option to selectively suppress some
1607 classes of assembly warning messages.
1609 \b Added the `-p' pre-include and `-d' pre-define command-line options.
1611 \b Added an include file search path: the `-i' command line option.
1613 \b Fixed a silly little preprocessor bug whereby starting a line with a
1614 `%!' environment-variable reference caused an `unknown directive'
1615 error.
1617 \b Added the long-awaited listing file support: the `-l' command line
1618 option.
1620 \b Fixed a problem with OBJ format whereby, in the absence of any
1621 explicit segment definition, non-global symbols declared in the
1622 implicit default segment generated spurious EXTDEF records in the
1623 output.
1625 \b Added the NASM environment variable.
1627 \b From this version forward, Win32 console-mode binaries will be
1628 included in the DOS distribution in addition to the 16-bit binaries.
1629 Added Makefile.vc for this purpose.
1631 \b Added `return 0;' to test/objlink.c to prevent compiler warnings.
1633 \b Added the __NASM_MAJOR__ and __NASM_MINOR__ standard defines.
1635 \b Added an alternative memory-reference syntax in which prefixing an
1636 operand with `&' is equivalent to enclosing it in square brackets,
1637 at the request of Fox Cutter.
1639 \b Errors in pass two now cause the program to return a non-zero error
1640 code, which they didn't before.
1642 \b Fixed the single-line macro cycle detection, which didn't work at
1643 all on macros with no parameters (caused an infinite loop). Also
1644 changed the behaviour of single-line macro cycle detection to work
1645 like cpp, so that macros like `extrn' as given in the documentation
1646 can be implemented.
1648 \b Fixed the implementation of WRT, which was too restrictive in that
1649 you couldn't do `mov ax,[di+abc wrt dgroup]' because (di+abc) wasn't
1650 a relocatable reference.
1653 \S{cl-0.94 released April 1997} Version 0.94 released April 1997
1656 \b Major item: added the macro processor.
1658 \b Added undocumented instructions SMI, IBTS, XBTS and LOADALL286. Also
1659 reorganised CMPXCHG instruction into early-486 and Pentium forms.
1660 Thanks to Thobias Jones for the information.
1662 \b Fixed two more stupid bugs in ELF, which were causing `ld' to
1663 continue to seg-fault in a lot of non-trivial cases.
1665 \b Fixed a seg-fault in the label manager.
1667 \b Stopped FBLD and FBSTP from _requiring_ the TWORD keyword, which is
1668 the only option for BCD loads/stores in any case.
1670 \b Ensured FLDCW, FSTCW and FSTSW can cope with the WORD keyword, if
1671 anyone bothers to provide it. Previously they complained unless no
1672 keyword at all was present.
1674 \b Some forms of FDIV/FDIVR and FSUB/FSUBR were still inverted: a
1675 vestige of a bug that I thought had been fixed in 0.92. This was
1676 fixed, hopefully for good this time...
1678 \b Another minor phase error (insofar as a phase error can _ever_ be
1679 minor) fixed, this one occurring in code of the form
1681 \c rol ax,forward_reference
1682 \c forward_reference equ 1
1684 \b The number supplied to TIMES is now sanity-checked for positivity,
1685 and also may be greater than 64K (which previously didn't work on
1686 16-bit systems).
1688 \b Added Watcom C makefiles, and misc/pmw.bat, donated by Dominik Behr.
1690 \b Added the INCBIN pseudo-opcode.
1692 \b Due to the advent of the preprocessor, the [INCLUDE] and [INC]
1693 directives have become obsolete. They are still supported in this
1694 version, with a warning, but won't be in the next.
1696 \b Fixed a bug in OBJ format, which caused incorrect object records to
1697 be output when absolute labels were made global.
1699 \b Updates to RDOFF subdirectory, and changes to outrdf.c.
1702 \S{cl-0.93 released January 1997} Version 0.93 released January 1997
1704 This release went out in a great hurry after semi-crippling bugs
1705 were found in 0.92.
1707 \b Really \e{did} fix the stack overflows this time. *blush*
1709 \b Had problems with EA instruction sizes changing between passes, when
1710 an offset contained a forward reference and so 4 bytes were
1711 allocated for the offset in pass one; by pass two the symbol had
1712 been defined and happened to be a small absolute value, so only 1
1713 byte got allocated, causing instruction size mismatch between passes
1714 and hence incorrect address calculations. Fixed.
1716 \b Stupid bug in the revised ELF section generation fixed (associated
1717 string-table section for .symtab was hard-coded as 7, even when this
1718 didn't fit with the real section table). Was causing `ld' to
1719 seg-fault under Linux.
1721 \b Included a new Borland C makefile, Makefile.bc2, donated by Fox
1722 Cutter <lmb@comtch.iea.com>.
1725 \S{cl-0.92 released January 1997} Version 0.92 released January 1997
1727 \b The FDIVP/FDIVRP and FSUBP/FSUBRP pairs had been inverted: this was
1728 fixed. This also affected the LCC driver.
1730 \b Fixed a bug regarding 32-bit effective addresses of the form
1731 \c{[other_register+ESP]}.
1733 \b Documentary changes, notably documentation of the fact that Borland
1734 Win32 compilers use `obj' rather than `win32' object format.
1736 \b Fixed the COMENT record in OBJ files, which was formatted
1737 incorrectly.
1739 \b Fixed a bug causing segfaults in large RDF files.
1741 \b OBJ format now strips initial periods from segment and group
1742 definitions, in order to avoid complications with the local label
1743 syntax.
1745 \b Fixed a bug in disassembling far calls and jumps in NDISASM.
1747 \b Added support for user-defined sections in COFF and ELF files.
1749 \b Compiled the DOS binaries with a sensible amount of stack, to
1750 prevent stack overflows on any arithmetic expression containing
1751 parentheses.
1753 \b Fixed a bug in handling of files that do not terminate in a newline.
1756 \S{cl-0.91 released November 1996} Version 0.91 released November 1996
1758 \b Loads of bug fixes.
1760 \b Support for RDF added.
1762 \b Support for DBG debugging format added.
1764 \b Support for 32-bit extensions to Microsoft OBJ format added.
1766 \b Revised for Borland C: some variable names changed, makefile added.
1768 \b LCC support revised to actually work.
1770 \b JMP/CALL NEAR/FAR notation added.
1772 \b `a16', `o16', `a32' and `o32' prefixes added.
1774 \b Range checking on short jumps implemented.
1776 \b MMX instruction support added.
1778 \b Negative floating point constant support added.
1780 \b Memory handling improved to bypass 64K barrier under DOS.
1782 \b \c{$} prefix to force treatment of reserved words as identifiers added.
1784 \b Default-size mechanism for object formats added.
1786 \b Compile-time configurability added.
1788 \b \c{#}, \c{@}, \c{~} and c\{?} are now valid characters in labels.
1790 \b \c{-e} and \c{-k} options in NDISASM added.
1793 \S{cl-0.90 released October 1996} Version 0.90 released October 1996
1795 First release version. First support for object file output. Other
1796 changes from previous version (0.3x) too numerous to document.