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{__utf16__} and \c{__utf32__} operators to generate UTF-16 and UTF-32
20 \b Fix bug in case-insensitive matching when compiled on platforms that
21 don't use the \c{configure} script. Of the official release binaries,
22 that only affects the OS/2 binary.
24 \b Correct the handling of nested \c{%reps}.
26 \b Support for x87 packed BCD constants.
28 \b New \c{%strcat} directive to join strings together.
30 \b Correct the \c{LTR} instruction in 64-bit mode.
32 \b Fix unnecessary REX.W prefix on indirect jumps in 64-bit mode.
35 \S{cl-2.03.01} Version 2.03.01
37 \b Fix buffer overflow in the listing module.
39 \b Fix the handling of hexadecimal escape codes in `...` strings.
41 \b The Postscript/PDF documentation has been reformatted.
43 \b The \c{-F} option now implies \c{-g}.
46 \S{cl-2.03} Version 2.03
48 \b Add support for Intel AVX, CLMUL and FMA instructions,
49 including YMM registers.
51 \b \c{dy}, \c{resy} and \c{yword} for 32-byte operands.
53 \b Fix some SSE5 instructions.
55 \b Intel \c{INVEPT}, \c{INVVPID} and \c{MOVBE} instructions.
57 \b Fix checking for critical expressions when the optimizer is enabled.
59 \b Support the DWARF debugging format for ELF targets.
61 \b Fix optimizations of signed bytes.
63 \b Fix operation on bigendian machines.
65 \b Fix buffer overflow in the preprocessor.
67 \b \c{SAFESEH} support for Win32, \c{IMAGEREL} for Win64 (SEH).
69 \b \c{%?} and \c{%??} to refer to the name of a macro itself. In particular,
70 \c{%idefine keyword $%?} can be used to make a keyword "disappear".
72 \b New options for dependency generation: \c{-MD}, \c{-MF},
73 \c{-MP}, \c{-MT}, \c{-MQ}.
75 \b New preprocessor directives \c{%pathsearch} and \c{%depend}; INCBIN
76 reimplemented as a macro.
78 \b \c{%include} now resolves macros in a sane manner.
80 \b \c{%substr} can now be used to get other than one-character substrings.
82 \b New type of character/string constants, using backquotes (\c{`...`}),
83 which support C-style escape sequences.
85 \b \c{%defstr} and \c{%idefstr} to stringize macro definitions before
88 \b Fix forward references used in \c{EQU} statements.
90 \S{cl-2.02} Version 2.02
93 \b Additional fixes for MMX operands with explicit \c{qword}, as well as
94 (hopefully) SSE operands with \c{oword}.
96 \b Fix handling of truncated strings with \c{DO}.
98 \b Fix segfaults due to memory overwrites when floating-point constants
101 \b Fix segfaults due to missing include files.
103 \b Fix OpenWatcom Makefiles for DOS and OS/2.
105 \b Add autogenerated instruction list back into the documentation.
107 \b ELF: Fix segfault when generating stabs, and no symbols have been
110 \b ELF: Experimental support for DWARF debugging information.
112 \b New compile date and time standard macros.
114 \b \c{%ifnum} now returns true for negative numbers.
116 \b New \c{%iftoken} test for a single token.
118 \b New \c{%ifempty} test for empty expansion.
120 \b Add support for the \c{XSAVE} instruction group.
122 \b Makefile for Netware/gcc.
124 \b Fix issue with some warnings getting emitted way too many times.
126 \b Autogenerated instruction list added to the documentation.
128 \S{cl-2.01} Version 2.01
131 \b Fix the handling of MMX registers with explicit \c{qword} tags on
132 memory (broken in 2.00 due to 64-bit changes.)
134 \b Fix the PREFETCH instructions.
136 \b Fix the documentation.
138 \b Fix debugging info when using \c{-f elf}
139 (backwards compatibility alias for \c{-f elf32}).
141 \b Man pages for rdoff tools (from the Debian project.)
143 \b ELF: handle large numbers of sections.
145 \b Fix corrupt output when the optimizer runs out of passes.
148 \S{cl-2.00} Version 2.00
150 \b Added c99 data-type compliance.
152 \b Added general x86-64 support.
154 \b Added win64 (x86-64 COFF) output format.
156 \b Added \c{__BITS__} standard macro.
158 \b Renamed the \c{elf} output format to \c{elf32} for clarity.
160 \b Added \c{elf64} and \c{macho} (MacOS X) output formats.
162 \b Added Numeric constants in \c{dq} directive.
164 \b Added \c{oword}, \c{do} and \c{reso} pseudo operands.
166 \b Allow underscores in numbers.
168 \b Added 8-, 16- and 128-bit floating-point formats.
170 \b Added binary, octal and hexadecimal floating-point.
172 \b Correct the generation of floating-point constants.
174 \b Added floating-point option control.
176 \b Added Infinity and NaN floating point support.
178 \b Added ELF Symbol Visibility support.
180 \b Added setting OSABI value in ELF header directive.
182 \b Added Generate Makefile Dependencies option.
184 \b Added Unlimited Optimization Passes option.
186 \b Added \c{%IFN} and \c{%ELIFN} support.
188 \b Added Logical Negation Operator.
190 \b Enhanced Stack Relative Preprocessor Directives.
192 \b Enhanced ELF Debug Formats.
194 \b Enhanced Send Errors to a File option.
196 \b Added SSSE3, SSE4.1, SSE4.2, SSE5 support.
198 \b Added a large number of additional instructions.
200 \b Significant performance improvements.
203 \H{cl-0.98.xx} NASM 0.98 Series
205 The 0.98 series was the production versions of NASM from 1999 to 2007.
208 \S{cl-0.98.39} Version 0.98.39
210 \b fix buffer overflow
212 \b fix outas86's \c{.bss} handling
214 \b "make spotless" no longer deletes config.h.in.
216 \b \c{%(el)if(n)idn} insensitivity to string quotes difference (#809300).
218 \b (nasm.c)\c{__OUTPUT_FORMAT__} changed to string value instead of symbol.
220 \S{cl-0.98.38} Version 0.98.38
223 \b Add Makefile for 16-bit DOS binaries under OpenWatcom, and modify
224 \c{mkdep.pl} to be able to generate completely pathless dependencies, as
225 required by OpenWatcom wmake (it supports path searches, but not
228 \b Fix the \c{STR} instruction.
230 \b Fix the ELF output format, which was broken under certain
231 circumstances due to the addition of stabs support.
233 \b Quick-fix Borland format debug-info for \c{-f obj}
235 \b Fix for \c{%rep} with no arguments (#560568)
237 \b Fix concatenation of preprocessor function call (#794686)
239 \b Fix long label causes coredump (#677841)
241 \b Use autoheader as well as autoconf to keep configure from generating
242 ridiculously long command lines.
244 \b Make sure that all of the formats which support debugging output
245 actually will suppress debugging output when \c{-g} not specified.
247 \S{cl-0.98.37} Version 0.98.37
250 \b Paths given in \c{-I} switch searched for \c{incbin}-ed as
251 well as \c{%include}-ed files.
253 \b Added stabs debugging for the ELF output format, patch from
256 \b Fix \c{output/outbin.c} to allow origin > 80000000h.
258 \b Make \c{-U} switch work.
260 \b Fix the use of relative offsets with explicit prefixes, e.g.
263 \b Remove \c{backslash()}.
265 \b Fix the \c{SMSW} and \c{SLDT} instructions.
267 \b \c{-O2} and \c{-O3} are no longer aliases for \c{-O10} and \c{-O15}.
268 If you mean the latter, please say so! :)
270 \S{cl-0.98.36} Version 0.98.36
273 \b Update rdoff - librarian/archiver - common rec - docs!
275 \b Fix signed/unsigned problems.
277 \b Fix \c{JMP FAR label} and \c{CALL FAR label}.
279 \b Add new multisection support - map files - fix align bug
281 \b Fix sysexit, movhps/movlps reg,reg bugs in insns.dat
283 \b \c{Q} or \c{O} suffixes indicate octal
285 \b Support Prescott new instructions (PNI).
287 \b Cyrix \c{XSTORE} instruction.
290 \S{cl-0.98.35} Version 0.98.35
292 \b Fix build failure on 16-bit DOS (Makefile.bc3 workaround for compiler bug.)
294 \b Fix dependencies and compiler warnings.
296 \b Add "const" in a number of places.
298 \b Add -X option to specify error reporting format (use -Xvc to
299 integrate with Microsoft Visual Studio.)
301 \b Minor changes for code legibility.
303 \b Drop use of tmpnam() in rdoff (security fix.)
306 \S{cl-0.98.34} Version 0.98.34
308 \b Correct additional address-size vs. operand-size confusions.
310 \b Generate dependencies for all Makefiles automatically.
312 \b Add support for unimplemented (but theoretically available)
313 registers such as tr0 and cr5. Segment registers 6 and 7 are called
314 segr6 and segr7 for the operations which they can be represented.
316 \b Correct some disassembler bugs related to redundant address-size prefixes.
317 Some work still remains in this area.
319 \b Correctly generate an error for things like "SEG eax".
321 \b Add the JMPE instruction, enabled by "CPU IA64".
323 \b Correct compilation on newer gcc/glibc platforms.
325 \b Issue an error on things like "jmp far eax".
328 \S{cl-0.98.33} Version 0.98.33
330 \b New __NASM_PATCHLEVEL__ and __NASM_VERSION_ID__ standard macros to
331 round out the version-query macros. version.pl now understands
332 X.YYplWW or X.YY.ZZplWW as a version number, equivalent to
333 X.YY.ZZ.WW (or X.YY.0.WW, as appropriate).
335 \b New keyword "strict" to disable the optimization of specific
338 \b Fix the handing of size overrides with JMP instructions
339 (instructions such as "jmp dword foo".)
341 \b Fix the handling of "ABSOLUTE label", where "label" points into a
344 \b Fix OBJ output format with lots of externs.
346 \b More documentation updates.
348 \b Add -Ov option to get verbose information about optimizations.
350 \b Undo a braindead change which broke \c{%elif} directives.
355 \S{cl-0.98.32} Version 0.98.32
357 \b Fix NASM crashing when \c{%macro} directives were left unterminated.
359 \b Lots of documentation updates.
361 \b Complete rewrite of the PostScript/PDF documentation generator.
363 \b The MS Visual C++ Makefile was updated and corrected.
365 \b Recognize .rodata as a standard section name in ELF.
367 \b Fix some obsolete Perl4-isms in Perl scripts.
369 \b Fix configure.in to work with autoconf 2.5x.
371 \b Fix a couple of "make cleaner" misses.
373 \b Make the normal "./configure && make" work with Cygwin.
376 \S{cl-0.98.31} Version 0.98.31
378 \b Correctly build in a separate object directory again.
380 \b Derive all references to the version number from the version file.
382 \b New standard macros __NASM_SUBMINOR__ and __NASM_VER__ macros.
384 \b Lots of Makefile updates and bug fixes.
386 \b New \c{%ifmacro} directive to test for multiline macros.
388 \b Documentation updates.
390 \b Fixes for 16-bit OBJ format output.
392 \b Changed the NASM environment variable to NASMENV.
395 \S{cl-0.98.30} Version 0.98.30
397 \b Changed doc files a lot: completely removed old READMExx and
398 Wishlist files, incorporating all information in CHANGES and TODO.
400 \b I waited a long time to rename zoutieee.c to (original) outieee.c
402 \b moved all output modules to output/ subdirectory.
404 \b Added 'make strip' target to strip debug info from nasm & ndisasm.
406 \b Added INSTALL file with installation instructions.
408 \b Added -v option description to nasm man.
410 \b Added dist makefile target to produce source distributions.
412 \b 16-bit support for ELF output format (GNU extension, but useful.)
415 \S{cl-0.98.28} Version 0.98.28
417 \b Fastcooked this for Debian's Woody release:
418 Frank applied the INCBIN bug patch to 0.98.25alt and called
419 it 0.98.28 to not confuse poor little apt-get.
422 \S{cl-0.98.26} Version 0.98.26
424 \b Reorganised files even better from 0.98.25alt
427 \S{cl-0.98.25alt} Version 0.98.25alt
429 \b Prettified the source tree. Moved files to more reasonable places.
431 \b Added findleak.pl script to misc/ directory.
433 \b Attempted to fix doc.
436 \S{cl-0.98.25} Version 0.98.25
438 \b Line continuation character \c{\\}.
440 \b Docs inadvertantly reverted - "dos packaging".
443 \S{cl-0.98.24p1} Version 0.98.24p1
445 \b FIXME: Someone, document this please.
448 \S{cl-0.98.24} Version 0.98.24
450 \b Documentation - Ndisasm doc added to Nasm.doc.
453 \S{cl-0.98.23} Version 0.98.23
455 \b Attempted to remove rdoff version1
457 \b Lino Mastrodomenico's patches to preproc.c (%$$ bug?).
460 \S{cl-0.98.22} Version 0.98.22
462 \b Update rdoff2 - attempt to remove v1.
465 \S{cl-0.98.21} Version 0.98.21
467 \b Optimization fixes.
470 \S{cl-0.98.20} Version 0.98.20
472 \b Optimization fixes.
475 \S{cl-0.98.19} Version 0.98.19
477 \b H. J. Lu's patch back out.
480 \S{cl-0.98.18} Version 0.98.18
482 \b Added ".rdata" to "-f win32".
485 \S{cl-0.98.17} Version 0.98.17
487 \b H. J. Lu's "bogus elf" patch. (Red Hat problem?)
490 \S{cl-0.98.16} Version 0.98.16
492 \b Fix whitespace before "[section ..." bug.
495 \S{cl-0.98.15} Version 0.98.15
497 \b Rdoff changes (?).
499 \b Fix fixes to memory leaks.
502 \S{cl-0.98.14} Version 0.98.14
507 \S{cl-0.98.13} Version 0.98.13
509 \b There was no 0.98.13
512 \S{cl-0.98.12} Version 0.98.12
514 \b Update optimization (new function of "-O1")
516 \b Changes to test/bintest.asm (?).
519 \S{cl-0.98.11} Version 0.98.11
521 \b Optimization changes.
526 \S{cl-0.98.10} Version 0.98.10
528 \b There was no 0.98.10
531 \S{cl-0.98.09} Version 0.98.09
533 \b Add multiple sections support to "-f bin".
535 \b Changed GLOBAL_TEMP_BASE in outelf.c from 6 to 15.
537 \b Add "-v" as an alias to the "-r" switch.
539 \b Remove "#ifdef" from Tasm compatibility options.
541 \b Remove redundant size-overrides on "mov ds, ex", etc.
543 \b Fixes to SSE2, other insns.dat (?).
545 \b Enable uppercase "I" and "P" switches.
547 \b Case insinsitive "seg" and "wrt".
549 \b Update install.sh (?).
551 \b Allocate tokens in blocks.
553 \b Improve "invalid effective address" messages.
556 \S{cl-0.98.08} Version 0.98.08
558 \b Add "\c{%strlen}" and "\c{%substr}" macro operators
560 \b Fixed broken c16.mac.
562 \b Unterminated string error reported.
564 \b Fixed bugs as per 0.98bf
567 \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
569 Changes from 0.98.07 release to 98.09b as of 28-Oct-2001
571 \b More closely compatible with 0.98 when -O0 is implied
572 or specified. Not strictly identical, since backward
573 branches in range of short offsets are recognized, and signed
574 byte values with no explicit size specification will be
575 assembled as a single byte.
577 \b More forgiving with the PUSH instruction. 0.98 requires
578 a size to be specified always. 0.98.09b will imply the size
579 from the current BITS setting (16 or 32).
581 \b Changed definition of the optimization flag:
583 -O0 strict two-pass assembly, JMP and Jcc are
584 handled more like 0.98, except that back-
585 ward JMPs are short, if possible.
587 -O1 strict two-pass assembly, but forward
588 branches are assembled with code guaranteed
589 to reach; may produce larger code than
590 -O0, but will produce successful assembly
591 more often if branch offset sizes are not
594 -O2 multi-pass optimization, minimize branch
595 offsets; also will minimize signed immed-
596 iate bytes, overriding size specification.
598 -O3 like -O2, but more passes taken, if needed
601 \S{cl-0.98.07 released 01/28/01} Version 0.98.07 released 01/28/01
603 \b Added Stepane Denis' SSE2 instructions to a *working*
604 version of the code - some earlier versions were based on
605 broken code - sorry 'bout that. version "0.98.07"
611 \b Cosmetic modifications to nasm.c, nasm.h,
615 \S{cl-0.98.06f released 01/18/01} Version 0.98.06f released 01/18/01
618 \b - Add "metalbrain"s jecxz bug fix in insns.dat
619 - alter nasmdoc.src to match - version "0.98.06f"
622 \S{cl-0.98.06e released 01/09/01} Version 0.98.06e released 01/09/01
625 \b Removed the "outforms.h" file - it appears to be
626 someone's old backup of "outform.h". version "0.98.06e"
630 \b fbk - finally added the fix for the "multiple %includes bug",
631 known since 7/27/99 - reported originally (?) and sent to
632 us by Austin Lunnen - he reports that John Fine had a fix
633 within the day. Here it is...
635 \b Nelson Rush resigns from the group. Big thanks to Nelson for
636 his leadership and enthusiasm in getting these changes
637 incorporated into Nasm!
639 \b fbk - [list +], [list -] directives - ineptly implemented, should
640 be re-written or removed, perhaps.
642 \b Brian Raiter / fbk - "elfso bug" fix - applied to aoutb format
643 as well - testing might be desirable...
647 \b James Seter - -postfix, -prefix command line switches.
649 \b Yuri Zaporogets - rdoff utility changes.
652 \S{cl-0.98p1} Version 0.98p1
654 \b GAS-like palign (Panos Minos)
656 \b FIXME: Someone, fill this in with details
659 \S{cl-0.98bf (bug-fixed)} Version 0.98bf (bug-fixed)
661 \b Fixed - elf and aoutb bug - shared libraries
662 - multiple "%include" bug in "-f obj"
664 - unrecognized option bug in ndisasm
666 \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
668 \b Added signed byte optimizations for the 0x81/0x83 class
669 of instructions: ADC, ADD, AND, CMP, OR, SBB, SUB, XOR:
670 when used as 'ADD reg16,imm' or 'ADD reg32,imm.' Also
671 optimization of signed byte form of 'PUSH imm' and 'IMUL
672 reg,imm'/'IMUL reg,reg,imm.' No size specification is needed.
674 \b Added multi-pass JMP and Jcc offset optimization. Offsets
675 on forward references will preferentially use the short form,
676 without the need to code a specific size (short or near) for
677 the branch. Added instructions for 'Jcc label' to use the
678 form 'Jnotcc $+3/JMP label', in cases where a short offset
679 is out of bounds. If compiling for a 386 or higher CPU, then
680 the 386 form of Jcc will be used instead.
682 This feature is controlled by a new command-line switch: "O",
683 (upper case letter O). "-O0" reverts the assembler to no
684 extra optimization passes, "-O1" allows up to 5 extra passes,
685 and "-O2"(default), allows up to 10 extra optimization passes.
687 \b Added a new directive: 'cpu XXX', where XXX is any of:
688 8086, 186, 286, 386, 486, 586, pentium, 686, PPro, P2, P3 or
689 Katmai. All are case insensitive. All instructions will
690 be selected only if they apply to the selected cpu or lower.
691 Corrected a couple of bugs in cpu-dependence in 'insns.dat'.
693 \b Added to 'standard.mac', the "use16" and "use32" forms of
694 the "bits 16/32" directive. This is nothing new, just conforms
695 to a lot of other assemblers. (minor)
697 \b Changed label allocation from 320/32 (10000 labels @ 200K+)
698 to 32/37 (1000 labels); makes running under DOS much easier.
699 Since additional label space is allocated dynamically, this
700 should have no effect on large programs with lots of labels.
701 The 37 is a prime, believed to be better for hashing. (minor)
704 \S{cl-0.98.03} Version 0.98.03
706 "Integrated patchfile 0.98-0.98.01. I call this version 0.98.03 for
707 historical reasons: 0.98.02 was trashed." --John Coffman
708 <johninsd@san.rr.com>, 27-Jul-2000
710 \b Kendall Bennett's SciTech MGL changes
712 \b Note that you must define "TASM_COMPAT" at compile-time
713 to get the Tasm Ideal Mode compatibility.
715 \b All changes can be compiled in and out using the TASM_COMPAT macros,
716 and when compiled without TASM_COMPAT defined we get the exact same
717 binary as the unmodified 0.98 sources.
719 \b standard.mac, macros.c: Added macros to ignore TASM directives before
722 \b nasm.h: Added extern declaration for tasm_compatible_mode
724 \b nasm.c: Added global variable tasm_compatible_mode
726 \b Added command line switch for TASM compatible mode (-t)
728 \b Changed version command line to reflect when compiled with TASM additions
730 \b Added response file processing to allow all arguments on a single
731 line (response file is @resp rather than -@resp for NASM format).
733 \b labels.c: Changes islocal() macro to support TASM style @@local labels.
735 \b Added islocalchar() macro to support TASM style @@local labels.
737 \b parser.c: Added support for TASM style memory references (ie: mov
738 [DWORD eax],10 rather than the NASM style mov DWORD [eax],10).
740 \b preproc.c: Added new directives, \c{%arg}, \c{%local}, \c{%stacksize} to directives
743 \b Added support for TASM style directives without a leading % symbol.
745 \b Integrated a block of changes from Andrew Zabolotny <bit@eltech.ru>:
747 \b A new keyword \c{%xdefine} and its case-insensitive counterpart \c{%ixdefine}.
748 They work almost the same way as \c{%define} and \c{%idefine} but expand
749 the definition immediately, not on the invocation. Something like a cross
750 between \c{%define} and \c{%assign}. The "x" suffix stands for "eXpand", so
751 "xdefine" can be deciphered as "expand-and-define". Thus you can do
757 \c %xdefine %1 dword [esp+ofs]
761 \b Changed the place where the expansion of %$name macros are expanded.
762 Now they are converted into ..@ctxnum.name form when detokenizing, so
763 there are no quirks as before when using %$name arguments to macros,
764 in macros etc. For example:
773 Now last line will be expanded into "hello" as expected. This also allows
774 for lots of goodies, a good example are extended "proc" macros included
777 \b Added a check for "cstk" in smacro_defined() before calling get_ctx() -
778 this allows for things like:
783 to work without warnings even in no context.
785 \b Added a check for "cstk" in %if*ctx and %elif*ctx directives -
786 this allows to use \c{%ifctx} without excessive warnings. If there is
787 no active context, \c{%ifctx} goes through "false" branch.
789 \b Removed "user error: " prefix with \c{%error} directive: it just clobbers the
790 output and has absolutely no functionality. Besides, this allows to write
791 macros that does not differ from built-in functions in any way.
793 \b Added expansion of string that is output by \c{%error} directive. Now you
796 \c %define hello(x) Hello, x!
798 \c %define %$name andy
799 \c %error "hello(%$name)"
801 Same happened with \c{%include} directive.
803 \b Now all directives that expect an identifier will try to expand and
804 concatenate everything without whitespaces in between before usage.
805 For example, with "unfixed" nasm the commands
807 \c %define %$abc hello
808 \c %define __%$abc goodbye
811 would produce "incorrect" output: last line will expand to
813 \c hello goodbyehello
815 Not quite what you expected, eh? :-) The answer is that preprocessor
816 treats the \c{%define} construct as if it would be
818 \c %define __ %$abc goodbye
820 (note the white space between __ and %$abc). After my "fix" it
821 will "correctly" expand into
825 as expected. Note that I use quotes around words "correct", "incorrect"
826 etc because this is rather a feature not a bug; however current behaviour
827 is more logical (and allows more advanced macro usage :-).
829 Same change was applied to:
830 \c{%push},\c{%macro},\c{%imacro},\c{%define},\c{%idefine},\c{%xdefine},\c{%ixdefine},
831 \c{%assign},\c{%iassign},\c{%undef}
833 \b A new directive [WARNING {+|-}warning-id] have been added. It works only
834 if the assembly phase is enabled (i.e. it doesn't work with nasm -e).
836 \b A new warning type: macro-selfref. By default this warning is disabled;
837 when enabled NASM warns when a macro self-references itself; for example
838 the following source:
840 \c [WARNING macro-selfref]
851 will produce a warning, but if we remove the first line we won't see it
852 anymore (which is The Right Thing To Do {tm} IMHO since C preprocessor
853 eats such constructs without warnings at all).
855 \b Added a "error" routine to preprocessor which always will set ERR_PASS1
856 bit in severity_code. This removes annoying repeated errors on first
857 and second passes from preprocessor.
859 \b Added the %+ operator in single-line macros for concatenating two
860 identifiers. Usage example:
862 \c %define _myfunc _otherfunc
863 \c %define cextern(x) _ %+ x
866 After first expansion, third line will become "_myfunc". After this
867 expansion is performed again so it becomes "_otherunc".
869 \b Now if preprocessor is in a non-emitting state, no warning or error
870 will be emitted. Example:
875 \c put anything you want between these two brackets,
876 \c even macro-parameter references %1 or local
877 \c labels %$zz or macro-local labels %%zz - no
878 \c warning will be emitted.
881 \b Context-local variables on expansion as a last resort are looked up
882 in outer contexts. For example, the following piece:
892 will expand correctly the fourth line to [esp]; if we'll define another
893 %$a inside the "inner" context, it will take precedence over outer
894 definition. However, this modification has been applied only to
895 expand_smacro and not to smacro_define: as a consequence expansion
896 looks in outer contexts, but \c{%ifdef} won't look in outer contexts.
898 This behaviour is needed because we don't want nested contexts to
899 act on already defined local macros. Example:
901 \c %define %$arg1 [esp+4]
907 In this example the "if" mmacro enters into the "if" context, so %$arg1
908 is not valid anymore inside "if". Of course it could be worked around
909 by using explicitely %$$arg1 but this is ugly IMHO.
911 \b Fixed memory leak in \c{%undef}. The origline wasn't freed before
914 \b Fixed trap in preprocessor when line expanded to empty set of tokens.
915 This happens, for example, in the following case:
921 \S{cl-0.98} Version 0.98
923 All changes since NASM 0.98p3 have been produced by H. Peter Anvin <hpa@zytor.com>.
925 \b The documentation comment delimiter is \# not #.
927 \b Allow EQU definitions to refer to external labels; reported by
930 \b Re-enable support for RDOFF v1; reported by Pedro Gimeno.
932 \b Updated License file per OK from Simon and Julian.
935 \S{cl-0.98p9} Version 0.98p9
937 \b Update documentation (although the instruction set reference will
938 have to wait; I don't want to hold up the 0.98 release for it.)
940 \b Verified that the NASM implementation of the PEXTRW and PMOVMSKB
941 instructions is correct. The encoding differs from what the Intel
942 manuals document, but the Pentium III behaviour matches NASM, not
945 \b Fix handling of implicit sizes in PSHUFW and PINSRW, reported by
948 \b Resurrect the -s option, which was removed when changing the
949 diagnostic output to stdout.
952 \S{cl-0.98p8} Version 0.98p8
954 \b Fix for "DB" when NASM is running on a bigendian machine.
956 \b Invoke insns.pl once for each output script, making Makefile.in
959 \b Improve the Unix configure-based makefiles to make package
962 \b Included an RPM .spec file for building RPM (RedHat Package Manager)
963 packages on Linux or Unix systems.
965 \b Fix Makefile dependency problems.
967 \b Change src/rdsrc.pl to include sectioning information in info
968 output; required for install-info to work.
970 \b Updated the RDOFF distribution to version 2 from Jules; minor
971 massaging to make it compile in my environment.
973 \b Split doc files that can be built by anyone with a Perl interpreter off
974 into a separate archive.
976 \b "Dress rehearsal" release!
979 \S{cl-0.98p7} Version 0.98p7
981 \b Fixed opcodes with a third byte-sized immediate argument to not
982 complain if given "byte" on the immediate.
984 \b Allow \c{%undef} to remove single-line macros with arguments. This
985 matches the behaviour of #undef in the C preprocessor.
987 \b Allow -d, -u, -i and -p to be specified as -D, -U, -I and -P for
988 compatibility with most C compilers and preprocessors. This allows
989 Makefile options to be shared between cc and nasm, for example.
993 \b Went through the list of Katmai instructions and hopefully fixed the
994 (rather few) mistakes in it.
996 \b (Hopefully) fixed a number of disassembler bugs related to ambiguous
997 instructions (disambiguated by -p) and SSE instructions with REP.
999 \b Fix for bug reported by Mark Junger: "call dword 0x12345678" should
1000 work and may add an OSP (affected CALL, JMP, Jcc).
1002 \b Fix for environments when "stderr" isn't a compile-time constant.
1005 \S{cl-0.98p6} Version 0.98p6
1008 \b Took officially over coordination of the 0.98 release; so drop
1009 the p3.x notation. Skipped p4 and p5 to avoid confusion with John
1010 Fine's J4 and J5 releases.
1012 \b Update the documentation; however, it still doesn't include
1013 documentation for the various new instructions. I somehow wonder if
1014 it makes sense to have an instruction set reference in the assembler
1015 manual when Intel et al have PDF versions of their manuals online.
1017 \b Recognize "idt" or "centaur" for the -p option to ndisasm.
1019 \b Changed error messages back to stderr where they belong, but add an
1020 -E option to redirect them elsewhere (the DOS shell cannot redirect
1023 \b -M option to generate Makefile dependencies (based on code from Alex
1026 \b \c{%undef} preprocessor directive, and -u option, that undefines a
1029 \b OS/2 Makefile (Mkfiles/Makefile.os2) for Borland under OS/2; from
1032 \b Various minor bugfixes (reported by):
1033 - Dangling \c{%s} in preproc.c (Martin Junker)
1035 \b THERE ARE KNOWN BUGS IN SSE AND THE OTHER KATMAI INSTRUCTIONS. I am
1036 on a trip and didn't bring the Katmai instruction reference, so I
1037 can't work on them right now.
1039 \b Updated the License file per agreement with Simon and Jules to
1040 include a GPL distribution clause.
1043 \S{cl-0.98p3.7} Version 0.98p3.7
1045 \b (Hopefully) fixed the canned Makefiles to include the outrdf2 and
1048 \b Renamed changes.asm to changed.asm.
1051 \S{cl-0.98p3.6} Version 0.98p3.6
1053 \b Fixed a bunch of instructions that were added in 0.98p3.5 which had
1054 memory operands, and the address-size prefix was missing from the
1055 instruction pattern.
1058 \S{cl-0.98p3.5} Version 0.98p3.5
1060 \b Merged in changes from John S. Fine's 0.98-J5 release. John's based
1061 0.98-J5 on my 0.98p3.3 release; this merges the changes.
1063 \b Expanded the instructions flag field to a long so we can fit more
1064 flags; mark SSE (KNI) and AMD or Katmai-specific instructions as
1067 \b Fix the "PRIV" flag on a bunch of instructions, and create new
1068 "PROT" flag for protected-mode-only instructions (orthogonal to if
1069 the instruction is privileged!) and new "SMM" flag for SMM-only
1072 \b Added AMD-only SYSCALL and SYSRET instructions.
1074 \b Make SSE actually work, and add new Katmai MMX instructions.
1076 \b Added a -p (preferred vendor) option to ndisasm so that it can
1077 distinguish e.g. Cyrix opcodes also used in SSE. For example:
1079 \c ndisasm -p cyrix aliased.bin
1080 \c 00000000 670F514310 paddsiw mm0,[ebx+0x10]
1081 \c 00000005 670F514320 paddsiw mm0,[ebx+0x20]
1082 \c ndisasm -p intel aliased.bin
1083 \c 00000000 670F514310 sqrtps xmm0,[ebx+0x10]
1084 \c 00000005 670F514320 sqrtps xmm0,[ebx+0x20]
1086 \b Added a bunch of Cyrix-specific instructions.
1089 \S{cl-0.98p3.4} Version 0.98p3.4
1091 \b Made at least an attempt to modify all the additional Makefiles (in
1092 the Mkfiles directory). I can't test it, but this was the best I
1095 \b DOS DJGPP+"Opus Make" Makefile from John S. Fine.
1097 \b changes.asm changes from John S. Fine.
1100 \S{cl-0.98p3.3} Version 0.98p3.3
1102 \b Patch from Conan Brink to allow nesting of \c{%rep} directives.
1104 \b If we're going to allow INT01 as an alias for INT1/ICEBP (one of
1105 Jules 0.98p3 changes), then we should allow INT03 as an alias for INT3
1108 \b Updated changes.asm to include the latest changes.
1110 \b Tried to clean up the <CR>s that had snuck in from a DOS/Windows
1111 environment into my Unix environment, and try to make sure than
1112 DOS/Windows users get them back.
1114 \b We would silently generate broken tools if insns.dat wasn't sorted
1115 properly. Change insns.pl so that the order doesn't matter.
1117 \b Fix bug in insns.pl (introduced by me) which would cause conditional
1118 instructions to have an extra "cc" in disassembly, e.g. "jnz"
1119 disassembled as "jccnz".
1122 \S{cl-0.98p3.2} Version 0.98p3.2
1124 \b Merged in John S. Fine's changes from his 0.98-J4 prerelease; see
1125 http://www.csoft.net/cz/johnfine/
1127 \b Changed previous "spotless" Makefile target (appropriate for distribution)
1128 to "distclean", and added "cleaner" target which is same as "clean"
1129 except deletes files generated by Perl scripts; "spotless" is union.
1131 \b Removed BASIC programs from distribution. Get a Perl interpreter
1132 instead (see below.)
1134 \b Calling this "pre-release 3.2" rather than "p3-hpa2" because of
1135 John's contributions.
1137 \b Actually link in the IEEE output format (zoutieee.c); fix a bunch of
1138 compiler warnings in that file. Note I don't know what IEEE output
1139 is supposed to look like, so these changes were made "blind".
1142 \S{cl-0.98p3-hpa} Version 0.98p3-hpa
1144 \b Merged nasm098p3.zip with nasm-0.97.tar.gz to create a fully
1145 buildable version for Unix systems (Makefile.in updates, etc.)
1147 \b Changed insns.pl to create the instruction tables in nasm.h and
1148 names.c, so that a new instruction can be added by adding it *only*
1151 \b Added the following new instructions: SYSENTER, SYSEXIT, FXSAVE,
1152 FXRSTOR, UD1, UD2 (the latter two are two opcodes that Intel
1153 guarantee will never be used; one of them is documented as UD2 in
1154 Intel documentation, the other one just as "Undefined Opcode" --
1155 calling it UD1 seemed to make sense.)
1157 \b MAX_SYMBOL was defined to be 9, but LOADALL286 and LOADALL386 are 10
1158 characters long. Now MAX_SYMBOL is derived from insns.dat.
1160 \b A note on the BASIC programs included: forget them. insns.bas is
1161 already out of date. Get yourself a Perl interpreter for your
1162 platform of choice at
1163 \W{http://www.cpan.org/ports/index.html}{http://www.cpan.org/ports/index.html}.
1166 \S{cl-0.98p3} Version 0.98 pre-release 3
1168 \b added response file support, improved command line handling, new layout
1171 \b fixed limit checking bug, 'OUT byte nn, reg' bug, and a couple of rdoff
1172 related bugs, updated Wishlist; 0.98 Prerelease 3.
1175 \S{cl-0.98p2} Version 0.98 pre-release 2
1177 \b fixed bug in outcoff.c to do with truncating section names longer
1178 than 8 characters, referencing beyond end of string; 0.98 pre-release 2
1181 \S{cl-0.98p1} Version 0.98 pre-release 1
1183 \b Fixed a bug whereby STRUC didn't work at all in RDF.
1185 \b Fixed a problem with group specification in PUBDEFs in OBJ.
1187 \b Improved ease of adding new output formats. Contribution due to
1190 \b Fixed a bug in relocations in the `bin' format: was showing up when
1191 a relocatable reference crossed an 8192-byte boundary in any output
1194 \b Fixed a bug in local labels: local-label lookups were inconsistent
1195 between passes one and two if an EQU occurred between the definition
1196 of a global label and the subsequent use of a local label local to
1199 \b Fixed a seg-fault in the preprocessor (again) which happened when
1200 you use a blank line as the first line of a multi-line macro
1201 definition and then defined a label on the same line as a call to
1204 \b Fixed a stale-pointer bug in the handling of the NASM environment
1205 variable. Thanks to Thomas McWilliams.
1207 \b ELF had a hard limit on the number of sections which caused
1208 segfaults when transgressed. Fixed.
1210 \b Added ability for ndisasm to read from stdin by using `-' as the
1213 \b ndisasm wasn't outputting the TO keyword. Fixed.
1215 \b Fixed error cascade on bogus expression in \c{%if} - an error in
1216 evaluation was causing the entire \c{%if} to be discarded, thus creating
1217 trouble later when the \c{%else} or \c{%endif} was encountered.
1219 \b Forward reference tracking was instruction-granular not operand-
1220 granular, which was causing 286-specific code to be generated
1221 needlessly on code of the form `shr word [forwardref],1'. Thanks to
1222 Jim Hague for sending a patch.
1224 \b All messages now appear on stdout, as sending them to stderr serves
1225 no useful purpose other than to make redirection difficult.
1227 \b Fixed the problem with EQUs pointing to an external symbol - this
1228 now generates an error message.
1230 \b Allowed multiple size prefixes to an operand, of which only the first
1231 is taken into account.
1233 \b Incorporated John Fine's changes, including fixes of a large number
1234 of preprocessor bugs, some small problems in OBJ, and a reworking of
1235 label handling to define labels before their line is assembled, rather
1238 \b Reformatted a lot of the source code to be more readable. Included
1239 'coding.txt' as a guideline for how to format code for contributors.
1241 \b Stopped nested \c{%reps} causing a panic - they now cause a slightly more
1242 friendly error message instead.
1244 \b Fixed floating point constant problems (patch by Pedro Gimeno)
1246 \b Fixed the return value of insn_size() not being checked for -1, indicating
1249 \b Incorporated 3Dnow! instructions.
1251 \b Fixed the 'mov eax, eax + ebx' bug.
1253 \b Fixed the GLOBAL EQU bug in ELF. Released developers release 3.
1255 \b Incorporated John Fine's command line parsing changes
1257 \b Incorporated David Lindauer's OMF debug support
1259 \b Made changes for LCC 4.0 support (\c{__NASM_CDecl__}, removed register size
1260 specification warning when sizes agree).
1263 \H{cl-0.9x} NASM 0.9 Series
1265 Revisions before 0.98.
1268 \S{cl-0.97} Version 0.97 released December 1997
1270 \b This was entirely a bug-fix release to 0.96, which seems to have got
1273 \b Fixed stupid mistake in OBJ which caused `MOV EAX,<constant>' to
1274 fail. Caused by an error in the `MOV EAX,<segment>' support.
1276 \b ndisasm hung at EOF when compiled with lcc on Linux because lcc on
1277 Linux somehow breaks feof(). ndisasm now does not rely on feof().
1279 \b A heading in the documentation was missing due to a markup error in
1280 the indexing. Fixed.
1282 \b Fixed failure to update all pointers on realloc() within extended-
1283 operand code in parser.c. Was causing wrong behaviour and seg faults
1284 on lines such as `dd 0.0,0.0,0.0,0.0,...'
1286 \b Fixed a subtle preprocessor bug whereby invoking one multi-line
1287 macro on the first line of the expansion of another, when the second
1288 had been invoked with a label defined before it, didn't expand the
1291 \b Added internal.doc back in to the distribution archives - it was
1292 missing in 0.96 *blush*
1294 \b Fixed bug causing 0.96 to be unable to assemble its own test files,
1295 specifically objtest.asm. *blush again*
1297 \b Fixed seg-faults and bogus error messages caused by mismatching
1298 \c{%rep} and \c{%endrep} within multi-line macro definitions.
1300 \b Fixed a problem with buffer overrun in OBJ, which was causing
1301 corruption at ends of long PUBDEF records.
1303 \b Separated DOS archives into main-program and documentation to reduce
1307 \S{cl-0.96} Version 0.96 released November 1997
1309 \b Fixed a bug whereby, if `nasm sourcefile' would cause a filename
1310 collision warning and put output into `nasm.out', then `nasm
1311 sourcefile -o outputfile' still gave the warning even though the
1313 Fixed name pollution under Digital UNIX: one of its header files
1314 defined R_SP, which broke the enum in nasm.h.
1316 \b Fixed minor instruction table problems: FUCOM and FUCOMP didn't have
1317 two-operand forms; NDISASM didn't recognise the longer register
1318 forms of PUSH and POP (eg FF F3 for PUSH BX); TEST mem,imm32 was
1319 flagged as undocumented; the 32-bit forms of CMOV had 16-bit operand
1320 size prefixes; `AAD imm' and `AAM imm' are no longer flagged as
1321 undocumented because the Intel Architecture reference documents
1324 \b Fixed a problem with the local-label mechanism, whereby strange
1325 types of symbol (EQUs, auto-defined OBJ segment base symbols)
1326 interfered with the `previous global label' value and screwed up
1329 \b Fixed a bug whereby the stub preprocessor didn't communicate with
1330 the listing file generator, so that the -a and -l options in
1331 conjunction would produce a useless listing file.
1333 \b Merged `os2' object file format back into `obj', after discovering
1334 that `obj' _also_ shouldn't have a link pass separator in a module
1335 containing a non-trivial MODEND. Flat segments are now declared
1336 using the FLAT attribute. `os2' is no longer a valid object format
1339 \b Removed the fixed-size temporary storage in the evaluator. Very very
1340 long expressions (like `mov ax,1+1+1+1+...' for two hundred 1s or
1341 so) should now no longer crash NASM.
1343 \b Fixed a bug involving segfaults on disassembly of MMX instructions,
1344 by changing the meaning of one of the operand-type flags in nasm.h.
1345 This may cause other apparently unrelated MMX problems; it needs to
1346 be tested thoroughly.
1348 \b Fixed some buffer overrun problems with large OBJ output files.
1349 Thanks to DJ Delorie for the bug report and fix.
1351 \b Made preprocess-only mode actually listen to the \c{%line} markers as it
1352 prints them, so that it can report errors more sanely.
1354 \b Re-designed the evaluator to keep more sensible track of expressions
1355 involving forward references: can now cope with previously-nightmare
1362 \b Added the ALIGN and ALIGNB standard macros.
1364 \b Added PIC support in ELF: use of WRT to obtain the four extra
1365 relocation types needed.
1367 \b Added the ability for output file formats to define their own
1368 extensions to the GLOBAL, COMMON and EXTERN directives.
1370 \b Implemented common-variable alignment, and global-symbol type and
1371 size declarations, in ELF.
1373 \b Implemented NEAR and FAR keywords for common variables, plus
1374 far-common element size specification, in OBJ.
1376 \b Added a feature whereby EXTERNs and COMMONs in OBJ can be given a
1377 default WRT specification (either a segment or a group).
1379 \b Transformed the Unix NASM archive into an auto-configuring package.
1381 \b Added a sanity-check for people applying SEG to things which are
1382 already segment bases: this previously went unnoticed by the SEG
1383 processing and caused OBJ-driver panics later.
1385 \b Added the ability, in OBJ format, to deal with `MOV EAX,<segment>'
1386 type references: OBJ doesn't directly support dword-size segment
1387 base fixups, but as long as the low two bytes of the constant term
1388 are zero, a word-size fixup can be generated instead and it will
1391 \b Added the ability to specify sections' alignment requirements in
1392 Win32 object files and pure binary files.
1394 \b Added preprocess-time expression evaluation: the \c{%assign} (and
1395 \c{%iassign}) directive and the bare \c{%if} (and \c{%elif}) conditional. Added
1396 relational operators to the evaluator, for use only in \c{%if}
1397 constructs: the standard relationals = < > <= >= <> (and C-like
1398 synonyms == and !=) plus low-precedence logical operators &&, ^^ and
1401 \b Added a preprocessor repeat construct: \c{%rep} / \c{%exitrep} / \c{%endrep}.
1403 \b Added the __FILE__ and __LINE__ standard macros.
1405 \b Added a sanity check for number constants being greater than
1406 0xFFFFFFFF. The warning can be disabled.
1408 \b Added the %0 token whereby a variadic multi-line macro can tell how
1409 many parameters it's been given in a specific invocation.
1411 \b Added \c{%rotate}, allowing multi-line macro parameters to be cycled.
1413 \b Added the `*' option for the maximum parameter count on multi-line
1414 macros, allowing them to take arbitrarily many parameters.
1416 \b Added the ability for the user-level forms of EXTERN, GLOBAL and
1417 COMMON to take more than one argument.
1419 \b Added the IMPORT and EXPORT directives in OBJ format, to deal with
1422 \b Added some more preprocessor \c{%if} constructs: \c{%ifidn} / \c{%ifidni} (exact
1423 textual identity), and \c{%ifid} / \c{%ifnum} / \c{%ifstr} (token type testing).
1425 \b Added the ability to distinguish SHL AX,1 (the 8086 version) from
1426 SHL AX,BYTE 1 (the 286-and-upwards version whose constant happens to
1429 \b Added NetBSD/FreeBSD/OpenBSD's variant of a.out format, complete
1430 with PIC shared library features.
1432 \b Changed NASM's idiosyncratic handling of FCLEX, FDISI, FENI, FINIT,
1433 FSAVE, FSTCW, FSTENV, and FSTSW to bring it into line with the
1434 otherwise accepted standard. The previous behaviour, though it was a
1435 deliberate feature, was a deliberate feature based on a
1436 misunderstanding. Apologies for the inconvenience.
1438 \b Improved the flexibility of ABSOLUTE: you can now give it an
1439 expression rather than being restricted to a constant, and it can
1440 take relocatable arguments as well.
1442 \b Added the ability for a variable to be declared as EXTERN multiple
1443 times, and the subsequent definitions are just ignored.
1445 \b We now allow instruction prefixes (CS, DS, LOCK, REPZ etc) to be
1446 alone on a line (without a following instruction).
1448 \b Improved sanity checks on whether the arguments to EXTERN, GLOBAL
1449 and COMMON are valid identifiers.
1451 \b Added misc/exebin.mac to allow direct generation of .EXE files by
1452 hacking up an EXE header using DB and DW; also added test/binexe.asm
1453 to demonstrate the use of this. Thanks to Yann Guidon for
1454 contributing the EXE header code.
1456 \b ndisasm forgot to check whether the input file had been successfully
1457 opened. Now it does. Doh!
1459 \b Added the Cyrix extensions to the MMX instruction set.
1461 \b Added a hinting mechanism to allow [EAX+EBX] and [EBX+EAX] to be
1462 assembled differently. This is important since [ESI+EBP] and
1463 [EBP+ESI] have different default base segment registers.
1465 \b Added support for the PharLap OMF extension for 4096-byte segment
1469 \S{cl-0.95 released July 1997} Version 0.95 released July 1997
1471 \b Fixed yet another ELF bug. This one manifested if the user relied on
1472 the default segment, and attempted to define global symbols without
1473 first explicitly declaring the target segment.
1475 \b Added makefiles (for NASM and the RDF tools) to build Win32 console
1476 apps under Symantec C++. Donated by Mark Junker.
1478 \b Added `macros.bas' and `insns.bas', QBasic versions of the Perl
1479 scripts that convert `standard.mac' to `macros.c' and convert
1480 `insns.dat' to `insnsa.c' and `insnsd.c'. Also thanks to Mark
1483 \b Changed the diassembled forms of the conditional instructions so
1484 that JB is now emitted as JC, and other similar changes. Suggested
1485 list by Ulrich Doewich.
1487 \b Added `@' to the list of valid characters to begin an identifier
1490 \b Documentary changes, notably the addition of the `Common Problems'
1491 section in nasm.doc.
1493 \b Fixed a bug relating to 32-bit PC-relative fixups in OBJ.
1495 \b Fixed a bug in perm_copy() in labels.c which was causing exceptions
1496 in cleanup_labels() on some systems.
1498 \b Positivity sanity check in TIMES argument changed from a warning to
1499 an error following a further complaint.
1501 \b Changed the acceptable limits on byte and word operands to allow
1502 things like `~10111001b' to work.
1504 \b Fixed a major problem in the preprocessor which caused seg-faults if
1505 macro definitions contained blank lines or comment-only lines.
1507 \b Fixed inadequate error checking on the commas separating the
1508 arguments to `db', `dw' etc.
1510 \b Fixed a crippling bug in the handling of macros with operand counts
1511 defined with a `+' modifier.
1513 \b Fixed a bug whereby object file formats which stored the input file
1514 name in the output file (such as OBJ and COFF) weren't doing so
1515 correctly when the output file name was specified on the command
1518 \b Removed [INC] and [INCLUDE] support for good, since they were
1521 \b Fixed a bug in OBJ which caused all fixups to be output in 16-bit
1522 (old-format) FIXUPP records, rather than putting the 32-bit ones in
1523 FIXUPP32 (new-format) records.
1525 \b Added, tentatively, OS/2 object file support (as a minor variant on
1528 \b Updates to Fox Cutter's Borland C makefile, Makefile.bc2.
1530 \b Removed a spurious second fclose() on the output file.
1532 \b Added the `-s' command line option to redirect all messages which
1533 would go to stderr (errors, help text) to stdout instead.
1535 \b Added the `-w' command line option to selectively suppress some
1536 classes of assembly warning messages.
1538 \b Added the `-p' pre-include and `-d' pre-define command-line options.
1540 \b Added an include file search path: the `-i' command line option.
1542 \b Fixed a silly little preprocessor bug whereby starting a line with a
1543 `%!' environment-variable reference caused an `unknown directive'
1546 \b Added the long-awaited listing file support: the `-l' command line
1549 \b Fixed a problem with OBJ format whereby, in the absence of any
1550 explicit segment definition, non-global symbols declared in the
1551 implicit default segment generated spurious EXTDEF records in the
1554 \b Added the NASM environment variable.
1556 \b From this version forward, Win32 console-mode binaries will be
1557 included in the DOS distribution in addition to the 16-bit binaries.
1558 Added Makefile.vc for this purpose.
1560 \b Added `return 0;' to test/objlink.c to prevent compiler warnings.
1562 \b Added the __NASM_MAJOR__ and __NASM_MINOR__ standard defines.
1564 \b Added an alternative memory-reference syntax in which prefixing an
1565 operand with `&' is equivalent to enclosing it in square brackets,
1566 at the request of Fox Cutter.
1568 \b Errors in pass two now cause the program to return a non-zero error
1569 code, which they didn't before.
1571 \b Fixed the single-line macro cycle detection, which didn't work at
1572 all on macros with no parameters (caused an infinite loop). Also
1573 changed the behaviour of single-line macro cycle detection to work
1574 like cpp, so that macros like `extrn' as given in the documentation
1577 \b Fixed the implementation of WRT, which was too restrictive in that
1578 you couldn't do `mov ax,[di+abc wrt dgroup]' because (di+abc) wasn't
1579 a relocatable reference.
1582 \S{cl-0.94 released April 1997} Version 0.94 released April 1997
1585 \b Major item: added the macro processor.
1587 \b Added undocumented instructions SMI, IBTS, XBTS and LOADALL286. Also
1588 reorganised CMPXCHG instruction into early-486 and Pentium forms.
1589 Thanks to Thobias Jones for the information.
1591 \b Fixed two more stupid bugs in ELF, which were causing `ld' to
1592 continue to seg-fault in a lot of non-trivial cases.
1594 \b Fixed a seg-fault in the label manager.
1596 \b Stopped FBLD and FBSTP from _requiring_ the TWORD keyword, which is
1597 the only option for BCD loads/stores in any case.
1599 \b Ensured FLDCW, FSTCW and FSTSW can cope with the WORD keyword, if
1600 anyone bothers to provide it. Previously they complained unless no
1601 keyword at all was present.
1603 \b Some forms of FDIV/FDIVR and FSUB/FSUBR were still inverted: a
1604 vestige of a bug that I thought had been fixed in 0.92. This was
1605 fixed, hopefully for good this time...
1607 \b Another minor phase error (insofar as a phase error can _ever_ be
1608 minor) fixed, this one occurring in code of the form
1610 \c rol ax,forward_reference
1611 \c forward_reference equ 1
1613 \b The number supplied to TIMES is now sanity-checked for positivity,
1614 and also may be greater than 64K (which previously didn't work on
1617 \b Added Watcom C makefiles, and misc/pmw.bat, donated by Dominik Behr.
1619 \b Added the INCBIN pseudo-opcode.
1621 \b Due to the advent of the preprocessor, the [INCLUDE] and [INC]
1622 directives have become obsolete. They are still supported in this
1623 version, with a warning, but won't be in the next.
1625 \b Fixed a bug in OBJ format, which caused incorrect object records to
1626 be output when absolute labels were made global.
1628 \b Updates to RDOFF subdirectory, and changes to outrdf.c.
1631 \S{cl-0.93 released January 1997} Version 0.93 released January 1997
1633 This release went out in a great hurry after semi-crippling bugs
1636 \b Really \e{did} fix the stack overflows this time. *blush*
1638 \b Had problems with EA instruction sizes changing between passes, when
1639 an offset contained a forward reference and so 4 bytes were
1640 allocated for the offset in pass one; by pass two the symbol had
1641 been defined and happened to be a small absolute value, so only 1
1642 byte got allocated, causing instruction size mismatch between passes
1643 and hence incorrect address calculations. Fixed.
1645 \b Stupid bug in the revised ELF section generation fixed (associated
1646 string-table section for .symtab was hard-coded as 7, even when this
1647 didn't fit with the real section table). Was causing `ld' to
1648 seg-fault under Linux.
1650 \b Included a new Borland C makefile, Makefile.bc2, donated by Fox
1651 Cutter <lmb@comtch.iea.com>.
1654 \S{cl-0.92 released January 1997} Version 0.92 released January 1997
1656 \b The FDIVP/FDIVRP and FSUBP/FSUBRP pairs had been inverted: this was
1657 fixed. This also affected the LCC driver.
1659 \b Fixed a bug regarding 32-bit effective addresses of the form
1660 \c{[other_register+ESP]}.
1662 \b Documentary changes, notably documentation of the fact that Borland
1663 Win32 compilers use `obj' rather than `win32' object format.
1665 \b Fixed the COMENT record in OBJ files, which was formatted
1668 \b Fixed a bug causing segfaults in large RDF files.
1670 \b OBJ format now strips initial periods from segment and group
1671 definitions, in order to avoid complications with the local label
1674 \b Fixed a bug in disassembling far calls and jumps in NDISASM.
1676 \b Added support for user-defined sections in COFF and ELF files.
1678 \b Compiled the DOS binaries with a sensible amount of stack, to
1679 prevent stack overflows on any arithmetic expression containing
1682 \b Fixed a bug in handling of files that do not terminate in a newline.
1685 \S{cl-0.91 released November 1996} Version 0.91 released November 1996
1687 \b Loads of bug fixes.
1689 \b Support for RDF added.
1691 \b Support for DBG debugging format added.
1693 \b Support for 32-bit extensions to Microsoft OBJ format added.
1695 \b Revised for Borland C: some variable names changed, makefile added.
1697 \b LCC support revised to actually work.
1699 \b JMP/CALL NEAR/FAR notation added.
1701 \b `a16', `o16', `a32' and `o32' prefixes added.
1703 \b Range checking on short jumps implemented.
1705 \b MMX instruction support added.
1707 \b Negative floating point constant support added.
1709 \b Memory handling improved to bypass 64K barrier under DOS.
1711 \b \c{$} prefix to force treatment of reserved words as identifiers added.
1713 \b Default-size mechanism for object formats added.
1715 \b Compile-time configurability added.
1717 \b \c{#}, \c{@}, \c{~} and c\{?} are now valid characters in labels.
1719 \b \c{-e} and \c{-k} options in NDISASM added.
1722 \S{cl-0.90 released October 1996} Version 0.90 released October 1996
1724 First release version. First support for object file output. Other
1725 changes from previous version (0.3x) too numerous to document.