2 \# NASM revision history in nasmdoc format
5 \H{cl-2.xx} NASM 2 Series
7 The NASM 2 series supports x86-64, and is the production version of NASM
10 \S{cl-2.11.02} Version 2.11.02
12 \b Add the \c{XSAVEC}, \c{XSAVES} and \c{XRSTORS} family instructions.
14 \S{cl-2.11.01} Version 2.11.01
16 \b Allow instructions which implicitly uses \c{XMM0} (\c{VBLENDVPD},
17 \c{VBLENDVPS}, \c{PBLENDVB} and \c{SHA256RNDS2}) to be specified
18 without an explicit \c{xmm0} on the assembly line. In other words,
19 the following two lines produce the same output:
21 \c vblendvpd xmm2,xmm1,xmm0 ; Last operand is fixed xmm0
22 \c vblendvpd xmm2,xmm1 ; Implicit xmm0 omitted
24 \b In the ELF backends, don't crash the assembler if \c{section align}
25 is specified without a value.
27 \S{cl-2.11} Version 2.11
29 \b Add support for the Intel AVX-512 instruction set:
31 \b 16 new, 512-bit SIMD registers. Total 32 \c{(ZMM0 ~ ZMM31)}
33 \b 8 new opmask registers \c{(K0 ~ K7)}. One of 7 registers \c{(K1 ~ K7)} can
34 be used as an opmask for conditional execution.
36 \b A new EVEX encoding prefix. EVEX is based on VEX and provides more
37 capabilities: opmasks, broadcasting, embedded rounding and compressed
41 \c VDIVPD zmm0{k1}{z}, zmm1, zmm3 ; conditional vector operation
43 \c ; {z} is for zero-masking
45 \c VDIVPS zmm4, zmm5, [rbx]{1to16} ; load single-precision float and
46 \c ; replicate it 16 times. 32 * 16 = 512
47 \c - embedded rounding
48 \c VCVTSI2SD xmm6, xmm7, {rz-sae}, rax ; round toward zero. note that it
49 \c ; is used as if a separate operand.
50 \c ; it comes after the last SIMD operand
52 \b Add support for \c{ZWORD} (512 bits), \c{DZ} and \c{RESZ}.
54 \b Add support for the MPX and SHA instruction sets.
56 \b Better handling of section redefinition.
58 \b Generate manpages when running \c{'make dist'}.
60 \b Handle all token chains in mmacro params range.
62 \b Support split [base,index] effective address:
64 \c mov eax,[eax+8,ecx*4] ; eax=base, ecx=index, 4=scale, 8=disp
66 This is expected to be most useful for the MPX instructions.
68 \b Support \c{BND} prefix for branch instructions (for MPX).
70 \b The \c{DEFAULT} directive can now take \c{BND} and \c{NOBND}
71 options to indicate whether all relevant branches should be getting
72 \c{BND} prefixes. This is expected to be the normal for use in MPX
75 \b Add \c{{evex}}, \c{{vex3}} and \c{{vex2}} instruction prefixes to
76 have NASM encode the corresponding instruction, if possible, with an EVEX,
77 3-byte VEX, or 2-byte VEX prefix, respectively.
79 \b Support for section names longer than 8 bytes in Win32/Win64 COFF.
81 \b The \c{NOSPLIT} directive by itself no longer forces a single
82 register to become an index register, unless it has an explicit
85 \c mov eax,[nosplit eax] ; eax as base register
86 \c mov eax,[nosplit eax*1] ; eax as index register
88 \S{cl-2.10.09} Version 2.10.09
90 \b Pregenerate man pages.
92 \S{cl-2.10.08} Version 2.10.08
94 \b Fix \c{VMOVNTDQA}, \c{MOVNTDQA} and \c{MOVLPD} instructions.
96 \b Fix collision for \c{VGATHERQPS}, \c{VPGATHERQD} instructions.
98 \b Fix \c{VPMOVSXBQ}, \c{VGATHERQPD}, \c{VSPLLW} instructions.
100 \b Add a bunch of AMD TBM instructions.
102 \b Fix potential stack overwrite in numbers conversion.
104 \b Allow byte size in \c{PREFETCHTx} instructions.
106 \b Make manual pages up to date.
108 \b Make \c{F3} and \c{F2} SSE prefixes to override \c{66}.
110 \b Support of AMD SVM instructions in 32 bit mode.
112 \b Fix near offsets code generation for \c{JMP}, \c{CALL} instrictions
115 \b Fix preprocessor parse regression when id is expanding to a whitespace.
117 \S{cl-2.10.07} Version 2.10.07
119 \b Fix line continuation parsing being broken in previous version.
121 \S{cl-2.10.06} Version 2.10.06
123 \b Always quote the dependency source names when using the automatic
124 dependency generation options.
126 \b If no dependency target name is specified via the \c{-MT} or
127 \c{-MQ} options, quote the default output name.
129 \b Fix assembly of shift operations in \c{CPU 8086} mode.
131 \b Fix incorrect generation of explicit immediate byte for shift by 1
132 under certain circumstances.
134 \b Fix assembly of the \c{VPCMPGTQ} instruction.
136 \b Fix RIP-relative relocations in the \c{macho64} backend.
138 \S{cl-2.10.05} Version 2.10.05
140 \b Add the \c{CLAC} and \c{STAC} instructions.
142 \S{cl-2.10.04} Version 2.10.04
144 \b Add back the inadvertently deleted 256-bit version of the \c{VORPD}
147 \b Correct disassembly of instructions starting with byte \c{82} hex.
149 \b Fix corner cases in token pasting, for example:
151 \c %define N 1e%++%+ 5
154 \S{cl-2.10.03} Version 2.10.03
156 \b Correct the assembly of the instruction:
158 \c XRELEASE MOV [absolute],AL
160 Previous versions would incorrectly generate \c{F3 A2} for this
161 instruction and issue a warning; correct behavior is to emit \c{F3 88
164 \S{cl-2.10.02} Version 2.10.02
166 \b Add the \c{ifunc} macro package with integer functions, currently
167 only integer logarithms. See \k{pkg_ifunc}.
169 \b Add the \c{RDSEED}, \c{ADCX} and \c{ADOX} instructions.
171 \S{cl-2.10.01} Version 2.10.01
173 \b Add missing VPMOVMSKB instruction with reg32, ymmreg operands.
175 \S{cl-2.10} Version 2.10
177 \b When optimization is enabled, \c{mov r64,imm} now optimizes to the
178 shortest form possible between:
180 \c mov r32,imm32 ; 5 bytes
181 \c mov r64,imm32 ; 7 bytes
182 \c mov r64,imm64 ; 10 bytes
184 To force a specific form, use the \c{STRICT} keyword, see \k{strict}.
186 \b Add support for the Intel AVX2 instruction set.
188 \b Add support for Bit Manipulation Instructions 1 and 2.
190 \b Add support for Intel Transactional Synchronization Extensions (TSX).
192 \b Add support for x32 ELF (32-bit ELF with the CPU in 64-bit mode.)
195 \b Add support for bigendian UTF-16 and UTF-32. See \k{unicode}.
197 \S{cl-2.09.10} Version 2.09.10
199 \b Fix up NSIS script to protect uninstaller against registry keys
200 absence or corruption. It brings in a few additional questions
201 to a user during deinstallation procedure but still it is better
202 than unpredictable file removal.
204 \S{cl-2.09.09} Version 2.09.09
206 \b Fix initialization of section attributes of \c{bin} output format.
208 \b Fix \c{mach64} output format bug that crashes NASM due to NULL symbols.
211 \S{cl-2.09.08} Version 2.09.08
213 \b Fix \c{__OUTPUT_FORMAT__} assignment when output driver alias
214 is used. For example when \c{-f elf} is used \c{__OUTPUT_FORMAT__}
215 must be set to \c{elf}, if \c{-f elf32} is used \c{__OUTPUT_FORMAT__}
216 must be assigned accordingly, i.e. to \c{elf32}. The rule applies to
217 all output driver aliases. See \k{ofmtm}.
220 \S{cl-2.09.07} Version 2.09.07
222 \b Fix attempts to close same file several times
223 when \c{-a} option is used.
225 \b Fixes for VEXTRACTF128, VMASKMOVPS encoding.
228 \S{cl-2.09.06} Version 2.09.06
230 \b Fix missed section attribute initialization in \c{bin} output target.
233 \S{cl-2.09.05} Version 2.09.05
235 \b Fix arguments encoding for VPEXTRW instruction.
237 \b Remove invalid form of VPEXTRW instruction.
239 \b Add \c{VLDDQU} as alias for \c{VLDQQU} to
243 \S{cl-2.09.04} Version 2.09.04
245 \b Fix incorrect labels offset for VEX intructions.
247 \b Eliminate bogus warning on implicit operand size override.
249 \b \c{%if} term could not handle 64 bit numbers.
251 \b The COFF backend was limiting relocations number to 16 bits even if
252 in real there were a way more relocations.
255 \S{cl-2.09.03} Version 2.09.03
257 \b Print \c{%macro} name inside \c{%rep} blocks on error.
259 \b Fix preprocessor expansion behaviour. It happened sometime
260 too early and sometime simply wrong. Move behaviour back to
261 the origins (down to NASM 2.05.01).
263 \b Fix unitialized data dereference on OMF output format.
265 \b Issue warning on unterminated \c{%{} construct.
267 \b Fix for documentation typo.
270 \S{cl-2.09.02} Version 2.09.02
272 \b Fix reversed tokens when \c{%deftok} produces more than one output token.
274 \b Fix segmentation fault on disassembling some VEX instructions.
276 \b Missing \c{%endif} did not always cause error.
278 \b Fix typo in documentation.
280 \b Compound context local preprocessor single line macro identifiers
281 were not expanded early enough and as result lead to unresolved
285 \S{cl-2.09.01} Version 2.09.01
287 \b Fix NULL dereference on missed %deftok second parameter.
289 \b Fix NULL dereference on invalid %substr parameters.
292 \S{cl-2.09} Version 2.09
294 \b Fixed assignment the magnitude of \c{%rep} counter. It is limited
297 \b Fixed NULL dereference if argument of \c{%strlen} resolves
298 to whitespace. For example if nonexistent macro parameter is used.
300 \b \c{%ifenv}, \c{%elifenv}, \c{%ifnenv}, and \c{%elifnenv} directives
301 introduced. See \k{ifenv}.
303 \b Fixed NULL dereference if environment variable is missed.
305 \b Updates of new AVX v7 Intel instructions.
307 \b \c{PUSH imm32} is now officially documented.
309 \b Fix for encoding the LFS, LGS and LSS in 64-bit mode.
311 \b Fixes for compatibility with OpenWatcom compiler and DOS 8.3 file
314 \b Macros parameters range expansion introduced. See \k{mlmacrange}.
316 \b Backward compatibility on expanging of local sigle macros restored.
318 \b 8 bit relocations for \c{elf} and \c{bin} output formats are introduced.
320 \b Short intersegment jumps are permitted now.
322 \b An alignment more than 64 bytes are allowed for \c{win32},
323 \c{win64} output formats.
325 \b \c{SECTALIGN} directive introduced. See \k{sectalign}.
327 \b \c{nojmp} option introduced in \c{smartalign} package. See
330 \b Short aliases \c{win}, \c{elf} and \c{macho} for output formats are
331 introduced. Each stands for \c{win32}, \c{elf32} and \c{macho32}
334 \b Faster handling of missing directives implemented.
336 \b Various small improvements in documentation.
338 \b No hang anymore if unable to open malloc.log file.
340 \b The environments without vsnprintf function are able to build nasm again.
342 \b AMD LWP instructions updated.
344 \b Tighten EA checks. We warn a user if there overflow in EA addressing.
346 \b Make \c{-Ox} the default optimization level. For the legacy
347 behavior, specify \c{-O0} explicitly. See \k{opt-O}.
349 \b Environment variables read with \c{%!} or tested with \c{%ifenv}
350 can now contain non-identifier characters if surrounded by quotes.
353 \b Add a new standard macro package \c{%use fp} for floating-point
354 convenience macros. See \k{pkg_fp}.
357 \S{cl-2.08.02} Version 2.08.02
359 \b Fix crash under certain circumstances when using the \c{%+} operator.
362 \S{cl-2.08.01} Version 2.08.01
364 \b Fix the \c{%use} statement, which was broken in 2.08.
367 \S{cl-2.08} Version 2.08
369 \b A number of enhancements/fixes in macros area.
371 \b Support for converting strings to tokens. See \k{deftok}.
373 \b Fuzzy operand size logic introduced.
375 \b Fix COFF stack overrun on too long export identifiers.
377 \b Fix Macho-O alignment bug.
379 \b Fix crashes with -fwin32 on file with many exports.
381 \b Fix stack overrun for too long [DEBUG id].
383 \b Fix incorrect sbyte usage in IMUL (hit only if optimization
386 \b Append ending token for \c{.stabs} records in the ELF output format.
388 \b New NSIS script which uses ModernUI and MultiUser approach.
390 \b Visual Studio 2008 NASM integration (rules file).
392 \b Warn a user if a constant is too long (and as result will be stripped).
394 \b The obsoleted pre-XOP AMD SSE5 instruction set which was never actualized
397 \b Fix stack overrun on too long error file name passed from the command line.
399 \b Bind symbols to the .text section by default (ie in case if SECTION
400 directive was omitted) in the ELF output format.
402 \b Fix sync points array index wrapping.
404 \b A few fixes for FMA4 and XOP instruction templates.
406 \b Add AMD Lightweight Profiling (LWP) instructions.
408 \b Fix the offset for \c{%arg} in 64-bit mode.
410 \b An undefined local macro (\c{%$}) no longer matches a global macro
413 \b Fix NULL dereference on too long local labels.
416 \S{cl-2.07} Version 2.07
418 \b NASM is now under the 2-clause BSD license. See \k{legal}.
420 \b Fix the section type for the \c{.strtab} section in the \c{elf64}
423 \b Fix the handling of \c{COMMON} directives in the \c{obj} output format.
425 \b New \c{ith} and \c{srec} output formats; these are variants of the
426 \c{bin} output format which output Intel hex and Motorola S-records,
427 respectively. See \k{ithfmt} and \k{srecfmt}.
429 \b \c{rdf2ihx} replaced with an enhanced \c{rdf2bin}, which can output
430 binary, COM, Intel hex or Motorola S-records.
432 \b The Windows installer now puts the NASM directory first in the
433 \c{PATH} of the "NASM Shell".
435 \b Revert the early expansion behavior of \c{%+} to pre-2.06 behavior:
436 \c{%+} is only expanded late.
438 \b Yet another Mach-O alignment fix.
440 \b Don't delete the list file on errors. Also, include error and
441 warning information in the list file.
443 \b Support for 64-bit Mach-O output, see \k{machofmt}.
445 \b Fix assert failure on certain operations that involve strings with
449 \S{cl-2.06} Version 2.06
451 \b This release is dedicated to the memory of Charles A. Crayne, long
452 time NASM developer as well as moderator of \c{comp.lang.asm.x86} and
453 author of the book \e{Serious Assembler}. We miss you, Chuck.
455 \b Support for indirect macro expansion (\c{%[...]}). See \k{indmacro}.
457 \b \c{%pop} can now take an argument, see \k{pushpop}.
459 \b The argument to \c{%use} is no longer macro-expanded. Use
460 \c{%[...]} if macro expansion is desired.
462 \b Support for thread-local storage in ELF32 and ELF64. See \k{elftls}.
464 \b Fix crash on \c{%ifmacro} without an argument.
466 \b Correct the arguments to the \c{POPCNT} instruction.
468 \b Fix section alignment in the Mach-O format.
470 \b Update AVX support to version 5 of the Intel specification.
472 \b Fix the handling of accesses to context-local macros from higher
473 levels in the context stack.
475 \b Treat \c{WAIT} as a prefix rather than as an instruction, thereby
476 allowing constructs like \c{O16 FSAVE} to work correctly.
478 \b Support for structures with a non-zero base offset. See \k{struc}.
480 \b Correctly handle preprocessor token concatenation (see \k{concat})
481 involving floating-point numbers.
483 \b The \c{PINSR} series of instructions have been corrected and
486 \b Removed AMD SSE5, replaced with the new XOP/FMA4/CVT16 (rev 3.03)
489 \b The ELF backends no longer automatically generate a \c{.comment} section.
491 \b Add additional "well-known" ELF sections with default attributes. See
495 \S{cl-2.05.01} Version 2.05.01
497 \b Fix the \c{-w}/\c{-W} option parsing, which was broken in NASM 2.05.
500 \S{cl-2.05} Version 2.05
502 \b Fix redundant REX.W prefix on \c{JMP reg64}.
504 \b Make the behaviour of \c{-O0} match NASM 0.98 legacy behavior.
507 \b \c{-w-user} can be used to suppress the output of \c{%warning} directives.
510 \b Fix bug where \c{ALIGN} would issue a full alignment datum instead of
513 \b Fix offsets in list files.
515 \b Fix \c{%include} inside multi-line macros or loops.
517 \b Fix error where NASM would generate a spurious warning on valid
518 optimizations of immediate values.
520 \b Fix arguments to a number of the \c{CVT} SSE instructions.
522 \b Fix RIP-relative offsets when the instruction carries an immediate.
524 \b Massive overhaul of the ELF64 backend for spec compliance.
526 \b Fix the Geode \c{PFRCPV} and \c{PFRSQRTV} instruction.
528 \b Fix the SSE 4.2 \c{CRC32} instruction.
531 \S{cl-2.04} Version 2.04
533 \b Sanitize macro handing in the \c{%error} directive.
535 \b New \c{%warning} directive to issue user-controlled warnings.
537 \b \c{%error} directives are now deferred to the final assembly phase.
539 \b New \c{%fatal} directive to immediately terminate assembly.
541 \b New \c{%strcat} directive to join quoted strings together.
543 \b New \c{%use} macro directive to support standard macro directives. See
546 \b Excess default parameters to \c{%macro} now issues a warning by default.
549 \b Fix \c{%ifn} and \c{%elifn}.
551 \b Fix nested \c{%else} clauses.
553 \b Correct the handling of nested \c{%rep}s.
555 \b New \c{%unmacro} directive to undeclare a multi-line macro.
558 \b Builtin macro \c{__PASS__} which expands to the current assembly pass.
561 \b \c{__utf16__} and \c{__utf32__} operators to generate UTF-16 and UTF-32
562 strings. See \k{unicode}.
564 \b Fix bug in case-insensitive matching when compiled on platforms that
565 don't use the \c{configure} script. Of the official release binaries,
566 that only affected the OS/2 binary.
568 \b Support for x87 packed BCD constants. See \k{bcdconst}.
570 \b Correct the \c{LTR} and \c{SLDT} instructions in 64-bit mode.
572 \b Fix unnecessary REX.W prefix on indirect jumps in 64-bit mode.
574 \b Add AVX versions of the AES instructions (\c{VAES}...).
576 \b Fix the 256-bit FMA instructions.
578 \b Add 256-bit AVX stores per the latest AVX spec.
580 \b VIA XCRYPT instructions can now be written either with or without
581 \c{REP}, apparently different versions of the VIA spec wrote them
584 \b Add missing 64-bit \c{MOVNTI} instruction.
586 \b Fix the operand size of \c{VMREAD} and \c{VMWRITE}.
588 \b Numerous bug fixes, especially to the AES, AVX and VTX instructions.
590 \b The optimizer now always runs until it converges. It also runs even
591 when disabled, but doesn't optimize. This allows most forward references
592 to be resolved properly.
594 \b \c{%push} no longer needs a context identifier; omitting the context
595 identifier results in an anonymous context.
598 \S{cl-2.03.01} Version 2.03.01
600 \b Fix buffer overflow in the listing module.
602 \b Fix the handling of hexadecimal escape codes in `...` strings.
604 \b The Postscript/PDF documentation has been reformatted.
606 \b The \c{-F} option now implies \c{-g}.
609 \S{cl-2.03} Version 2.03
611 \b Add support for Intel AVX, CLMUL and FMA instructions,
612 including YMM registers.
614 \b \c{dy}, \c{resy} and \c{yword} for 32-byte operands.
616 \b Fix some SSE5 instructions.
618 \b Intel \c{INVEPT}, \c{INVVPID} and \c{MOVBE} instructions.
620 \b Fix checking for critical expressions when the optimizer is enabled.
622 \b Support the DWARF debugging format for ELF targets.
624 \b Fix optimizations of signed bytes.
626 \b Fix operation on bigendian machines.
628 \b Fix buffer overflow in the preprocessor.
630 \b \c{SAFESEH} support for Win32, \c{IMAGEREL} for Win64 (SEH).
632 \b \c{%?} and \c{%??} to refer to the name of a macro itself. In particular,
633 \c{%idefine keyword $%?} can be used to make a keyword "disappear".
635 \b New options for dependency generation: \c{-MD}, \c{-MF},
636 \c{-MP}, \c{-MT}, \c{-MQ}.
638 \b New preprocessor directives \c{%pathsearch} and \c{%depend}; INCBIN
639 reimplemented as a macro.
641 \b \c{%include} now resolves macros in a sane manner.
643 \b \c{%substr} can now be used to get other than one-character substrings.
645 \b New type of character/string constants, using backquotes (\c{`...`}),
646 which support C-style escape sequences.
648 \b \c{%defstr} and \c{%idefstr} to stringize macro definitions before
651 \b Fix forward references used in \c{EQU} statements.
654 \S{cl-2.02} Version 2.02
656 \b Additional fixes for MMX operands with explicit \c{qword}, as well as
657 (hopefully) SSE operands with \c{oword}.
659 \b Fix handling of truncated strings with \c{DO}.
661 \b Fix segfaults due to memory overwrites when floating-point constants
664 \b Fix segfaults due to missing include files.
666 \b Fix OpenWatcom Makefiles for DOS and OS/2.
668 \b Add autogenerated instruction list back into the documentation.
670 \b ELF: Fix segfault when generating stabs, and no symbols have been
673 \b ELF: Experimental support for DWARF debugging information.
675 \b New compile date and time standard macros.
677 \b \c{%ifnum} now returns true for negative numbers.
679 \b New \c{%iftoken} test for a single token.
681 \b New \c{%ifempty} test for empty expansion.
683 \b Add support for the \c{XSAVE} instruction group.
685 \b Makefile for Netware/gcc.
687 \b Fix issue with some warnings getting emitted way too many times.
689 \b Autogenerated instruction list added to the documentation.
692 \S{cl-2.01} Version 2.01
694 \b Fix the handling of MMX registers with explicit \c{qword} tags on
695 memory (broken in 2.00 due to 64-bit changes.)
697 \b Fix the PREFETCH instructions.
699 \b Fix the documentation.
701 \b Fix debugging info when using \c{-f elf}
702 (backwards compatibility alias for \c{-f elf32}).
704 \b Man pages for rdoff tools (from the Debian project.)
706 \b ELF: handle large numbers of sections.
708 \b Fix corrupt output when the optimizer runs out of passes.
711 \S{cl-2.00} Version 2.00
713 \b Added c99 data-type compliance.
715 \b Added general x86-64 support.
717 \b Added win64 (x86-64 COFF) output format.
719 \b Added \c{__BITS__} standard macro.
721 \b Renamed the \c{elf} output format to \c{elf32} for clarity.
723 \b Added \c{elf64} and \c{macho} (MacOS X) output formats.
725 \b Added Numeric constants in \c{dq} directive.
727 \b Added \c{oword}, \c{do} and \c{reso} pseudo operands.
729 \b Allow underscores in numbers.
731 \b Added 8-, 16- and 128-bit floating-point formats.
733 \b Added binary, octal and hexadecimal floating-point.
735 \b Correct the generation of floating-point constants.
737 \b Added floating-point option control.
739 \b Added Infinity and NaN floating point support.
741 \b Added ELF Symbol Visibility support.
743 \b Added setting OSABI value in ELF header directive.
745 \b Added Generate Makefile Dependencies option.
747 \b Added Unlimited Optimization Passes option.
749 \b Added \c{%IFN} and \c{%ELIFN} support.
751 \b Added Logical Negation Operator.
753 \b Enhanced Stack Relative Preprocessor Directives.
755 \b Enhanced ELF Debug Formats.
757 \b Enhanced Send Errors to a File option.
759 \b Added SSSE3, SSE4.1, SSE4.2, SSE5 support.
761 \b Added a large number of additional instructions.
763 \b Significant performance improvements.
765 \b \c{-w+warning} and \c{-w-warning} can now be written as -Wwarning and
766 -Wno-warning, respectively. See \k{opt-w}.
768 \b Add \c{-w+error} to treat warnings as errors. See \k{opt-w}.
770 \b Add \c{-w+all} and \c{-w-all} to enable or disable all suppressible
771 warnings. See \k{opt-w}.
774 \H{cl-0.98.xx} NASM 0.98 Series
776 The 0.98 series was the production versions of NASM from 1999 to 2007.
779 \S{cl-0.98.39} Version 0.98.39
781 \b fix buffer overflow
783 \b fix outas86's \c{.bss} handling
785 \b "make spotless" no longer deletes config.h.in.
787 \b \c{%(el)if(n)idn} insensitivity to string quotes difference (#809300).
789 \b (nasm.c)\c{__OUTPUT_FORMAT__} changed to string value instead of symbol.
791 \S{cl-0.98.38} Version 0.98.38
794 \b Add Makefile for 16-bit DOS binaries under OpenWatcom, and modify
795 \c{mkdep.pl} to be able to generate completely pathless dependencies, as
796 required by OpenWatcom wmake (it supports path searches, but not
799 \b Fix the \c{STR} instruction.
801 \b Fix the ELF output format, which was broken under certain
802 circumstances due to the addition of stabs support.
804 \b Quick-fix Borland format debug-info for \c{-f obj}
806 \b Fix for \c{%rep} with no arguments (#560568)
808 \b Fix concatenation of preprocessor function call (#794686)
810 \b Fix long label causes coredump (#677841)
812 \b Use autoheader as well as autoconf to keep configure from generating
813 ridiculously long command lines.
815 \b Make sure that all of the formats which support debugging output
816 actually will suppress debugging output when \c{-g} not specified.
818 \S{cl-0.98.37} Version 0.98.37
821 \b Paths given in \c{-I} switch searched for \c{incbin}-ed as
822 well as \c{%include}-ed files.
824 \b Added stabs debugging for the ELF output format, patch from
827 \b Fix \c{output/outbin.c} to allow origin > 80000000h.
829 \b Make \c{-U} switch work.
831 \b Fix the use of relative offsets with explicit prefixes, e.g.
834 \b Remove \c{backslash()}.
836 \b Fix the \c{SMSW} and \c{SLDT} instructions.
838 \b \c{-O2} and \c{-O3} are no longer aliases for \c{-O10} and \c{-O15}.
839 If you mean the latter, please say so! :)
841 \S{cl-0.98.36} Version 0.98.36
844 \b Update rdoff - librarian/archiver - common rec - docs!
846 \b Fix signed/unsigned problems.
848 \b Fix \c{JMP FAR label} and \c{CALL FAR label}.
850 \b Add new multisection support - map files - fix align bug
852 \b Fix sysexit, movhps/movlps reg,reg bugs in insns.dat
854 \b \c{Q} or \c{O} suffixes indicate octal
856 \b Support Prescott new instructions (PNI).
858 \b Cyrix \c{XSTORE} instruction.
861 \S{cl-0.98.35} Version 0.98.35
863 \b Fix build failure on 16-bit DOS (Makefile.bc3 workaround for compiler bug.)
865 \b Fix dependencies and compiler warnings.
867 \b Add "const" in a number of places.
869 \b Add -X option to specify error reporting format (use -Xvc to
870 integrate with Microsoft Visual Studio.)
872 \b Minor changes for code legibility.
874 \b Drop use of tmpnam() in rdoff (security fix.)
877 \S{cl-0.98.34} Version 0.98.34
879 \b Correct additional address-size vs. operand-size confusions.
881 \b Generate dependencies for all Makefiles automatically.
883 \b Add support for unimplemented (but theoretically available)
884 registers such as tr0 and cr5. Segment registers 6 and 7 are called
885 segr6 and segr7 for the operations which they can be represented.
887 \b Correct some disassembler bugs related to redundant address-size prefixes.
888 Some work still remains in this area.
890 \b Correctly generate an error for things like "SEG eax".
892 \b Add the JMPE instruction, enabled by "CPU IA64".
894 \b Correct compilation on newer gcc/glibc platforms.
896 \b Issue an error on things like "jmp far eax".
899 \S{cl-0.98.33} Version 0.98.33
901 \b New __NASM_PATCHLEVEL__ and __NASM_VERSION_ID__ standard macros to
902 round out the version-query macros. version.pl now understands
903 X.YYplWW or X.YY.ZZplWW as a version number, equivalent to
904 X.YY.ZZ.WW (or X.YY.0.WW, as appropriate).
906 \b New keyword "strict" to disable the optimization of specific
909 \b Fix the handing of size overrides with JMP instructions
910 (instructions such as "jmp dword foo".)
912 \b Fix the handling of "ABSOLUTE label", where "label" points into a
915 \b Fix OBJ output format with lots of externs.
917 \b More documentation updates.
919 \b Add -Ov option to get verbose information about optimizations.
921 \b Undo a braindead change which broke \c{%elif} directives.
926 \S{cl-0.98.32} Version 0.98.32
928 \b Fix NASM crashing when \c{%macro} directives were left unterminated.
930 \b Lots of documentation updates.
932 \b Complete rewrite of the PostScript/PDF documentation generator.
934 \b The MS Visual C++ Makefile was updated and corrected.
936 \b Recognize .rodata as a standard section name in ELF.
938 \b Fix some obsolete Perl4-isms in Perl scripts.
940 \b Fix configure.in to work with autoconf 2.5x.
942 \b Fix a couple of "make cleaner" misses.
944 \b Make the normal "./configure && make" work with Cygwin.
947 \S{cl-0.98.31} Version 0.98.31
949 \b Correctly build in a separate object directory again.
951 \b Derive all references to the version number from the version file.
953 \b New standard macros __NASM_SUBMINOR__ and __NASM_VER__ macros.
955 \b Lots of Makefile updates and bug fixes.
957 \b New \c{%ifmacro} directive to test for multiline macros.
959 \b Documentation updates.
961 \b Fixes for 16-bit OBJ format output.
963 \b Changed the NASM environment variable to NASMENV.
966 \S{cl-0.98.30} Version 0.98.30
968 \b Changed doc files a lot: completely removed old READMExx and
969 Wishlist files, incorporating all information in CHANGES and TODO.
971 \b I waited a long time to rename zoutieee.c to (original) outieee.c
973 \b moved all output modules to output/ subdirectory.
975 \b Added 'make strip' target to strip debug info from nasm & ndisasm.
977 \b Added INSTALL file with installation instructions.
979 \b Added -v option description to nasm man.
981 \b Added dist makefile target to produce source distributions.
983 \b 16-bit support for ELF output format (GNU extension, but useful.)
986 \S{cl-0.98.28} Version 0.98.28
988 \b Fastcooked this for Debian's Woody release:
989 Frank applied the INCBIN bug patch to 0.98.25alt and called
990 it 0.98.28 to not confuse poor little apt-get.
993 \S{cl-0.98.26} Version 0.98.26
995 \b Reorganised files even better from 0.98.25alt
998 \S{cl-0.98.25alt} Version 0.98.25alt
1000 \b Prettified the source tree. Moved files to more reasonable places.
1002 \b Added findleak.pl script to misc/ directory.
1004 \b Attempted to fix doc.
1007 \S{cl-0.98.25} Version 0.98.25
1009 \b Line continuation character \c{\\}.
1011 \b Docs inadvertantly reverted - "dos packaging".
1014 \S{cl-0.98.24p1} Version 0.98.24p1
1016 \b FIXME: Someone, document this please.
1019 \S{cl-0.98.24} Version 0.98.24
1021 \b Documentation - Ndisasm doc added to Nasm.doc.
1024 \S{cl-0.98.23} Version 0.98.23
1026 \b Attempted to remove rdoff version1
1028 \b Lino Mastrodomenico's patches to preproc.c (%$$ bug?).
1031 \S{cl-0.98.22} Version 0.98.22
1033 \b Update rdoff2 - attempt to remove v1.
1036 \S{cl-0.98.21} Version 0.98.21
1038 \b Optimization fixes.
1041 \S{cl-0.98.20} Version 0.98.20
1043 \b Optimization fixes.
1046 \S{cl-0.98.19} Version 0.98.19
1048 \b H. J. Lu's patch back out.
1051 \S{cl-0.98.18} Version 0.98.18
1053 \b Added ".rdata" to "-f win32".
1056 \S{cl-0.98.17} Version 0.98.17
1058 \b H. J. Lu's "bogus elf" patch. (Red Hat problem?)
1061 \S{cl-0.98.16} Version 0.98.16
1063 \b Fix whitespace before "[section ..." bug.
1066 \S{cl-0.98.15} Version 0.98.15
1068 \b Rdoff changes (?).
1070 \b Fix fixes to memory leaks.
1073 \S{cl-0.98.14} Version 0.98.14
1075 \b Fix memory leaks.
1078 \S{cl-0.98.13} Version 0.98.13
1080 \b There was no 0.98.13
1083 \S{cl-0.98.12} Version 0.98.12
1085 \b Update optimization (new function of "-O1")
1087 \b Changes to test/bintest.asm (?).
1090 \S{cl-0.98.11} Version 0.98.11
1092 \b Optimization changes.
1097 \S{cl-0.98.10} Version 0.98.10
1099 \b There was no 0.98.10
1102 \S{cl-0.98.09} Version 0.98.09
1104 \b Add multiple sections support to "-f bin".
1106 \b Changed GLOBAL_TEMP_BASE in outelf.c from 6 to 15.
1108 \b Add "-v" as an alias to the "-r" switch.
1110 \b Remove "#ifdef" from Tasm compatibility options.
1112 \b Remove redundant size-overrides on "mov ds, ex", etc.
1114 \b Fixes to SSE2, other insns.dat (?).
1116 \b Enable uppercase "I" and "P" switches.
1118 \b Case insinsitive "seg" and "wrt".
1120 \b Update install.sh (?).
1122 \b Allocate tokens in blocks.
1124 \b Improve "invalid effective address" messages.
1127 \S{cl-0.98.08} Version 0.98.08
1129 \b Add "\c{%strlen}" and "\c{%substr}" macro operators
1131 \b Fixed broken c16.mac.
1133 \b Unterminated string error reported.
1135 \b Fixed bugs as per 0.98bf
1138 \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
1140 Changes from 0.98.07 release to 98.09b as of 28-Oct-2001
1142 \b More closely compatible with 0.98 when -O0 is implied
1143 or specified. Not strictly identical, since backward
1144 branches in range of short offsets are recognized, and signed
1145 byte values with no explicit size specification will be
1146 assembled as a single byte.
1148 \b More forgiving with the PUSH instruction. 0.98 requires
1149 a size to be specified always. 0.98.09b will imply the size
1150 from the current BITS setting (16 or 32).
1152 \b Changed definition of the optimization flag:
1154 -O0 strict two-pass assembly, JMP and Jcc are
1155 handled more like 0.98, except that back-
1156 ward JMPs are short, if possible.
1158 -O1 strict two-pass assembly, but forward
1159 branches are assembled with code guaranteed
1160 to reach; may produce larger code than
1161 -O0, but will produce successful assembly
1162 more often if branch offset sizes are not
1165 -O2 multi-pass optimization, minimize branch
1166 offsets; also will minimize signed immed-
1167 iate bytes, overriding size specification.
1169 -O3 like -O2, but more passes taken, if needed
1172 \S{cl-0.98.07 released 01/28/01} Version 0.98.07 released 01/28/01
1174 \b Added Stepane Denis' SSE2 instructions to a *working*
1175 version of the code - some earlier versions were based on
1176 broken code - sorry 'bout that. version "0.98.07"
1182 \b Cosmetic modifications to nasm.c, nasm.h,
1186 \S{cl-0.98.06f released 01/18/01} Version 0.98.06f released 01/18/01
1189 \b - Add "metalbrain"s jecxz bug fix in insns.dat
1190 - alter nasmdoc.src to match - version "0.98.06f"
1193 \S{cl-0.98.06e released 01/09/01} Version 0.98.06e released 01/09/01
1196 \b Removed the "outforms.h" file - it appears to be
1197 someone's old backup of "outform.h". version "0.98.06e"
1201 \b fbk - finally added the fix for the "multiple %includes bug",
1202 known since 7/27/99 - reported originally (?) and sent to
1203 us by Austin Lunnen - he reports that John Fine had a fix
1204 within the day. Here it is...
1206 \b Nelson Rush resigns from the group. Big thanks to Nelson for
1207 his leadership and enthusiasm in getting these changes
1208 incorporated into Nasm!
1210 \b fbk - [list +], [list -] directives - ineptly implemented, should
1211 be re-written or removed, perhaps.
1213 \b Brian Raiter / fbk - "elfso bug" fix - applied to aoutb format
1214 as well - testing might be desirable...
1218 \b James Seter - -postfix, -prefix command line switches.
1220 \b Yuri Zaporogets - rdoff utility changes.
1223 \S{cl-0.98p1} Version 0.98p1
1225 \b GAS-like palign (Panos Minos)
1227 \b FIXME: Someone, fill this in with details
1230 \S{cl-0.98bf (bug-fixed)} Version 0.98bf (bug-fixed)
1232 \b Fixed - elf and aoutb bug - shared libraries
1233 - multiple "%include" bug in "-f obj"
1235 - unrecognized option bug in ndisasm
1237 \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
1239 \b Added signed byte optimizations for the 0x81/0x83 class
1240 of instructions: ADC, ADD, AND, CMP, OR, SBB, SUB, XOR:
1241 when used as 'ADD reg16,imm' or 'ADD reg32,imm.' Also
1242 optimization of signed byte form of 'PUSH imm' and 'IMUL
1243 reg,imm'/'IMUL reg,reg,imm.' No size specification is needed.
1245 \b Added multi-pass JMP and Jcc offset optimization. Offsets
1246 on forward references will preferentially use the short form,
1247 without the need to code a specific size (short or near) for
1248 the branch. Added instructions for 'Jcc label' to use the
1249 form 'Jnotcc $+3/JMP label', in cases where a short offset
1250 is out of bounds. If compiling for a 386 or higher CPU, then
1251 the 386 form of Jcc will be used instead.
1253 This feature is controlled by a new command-line switch: "O",
1254 (upper case letter O). "-O0" reverts the assembler to no
1255 extra optimization passes, "-O1" allows up to 5 extra passes,
1256 and "-O2"(default), allows up to 10 extra optimization passes.
1258 \b Added a new directive: 'cpu XXX', where XXX is any of:
1259 8086, 186, 286, 386, 486, 586, pentium, 686, PPro, P2, P3 or
1260 Katmai. All are case insensitive. All instructions will
1261 be selected only if they apply to the selected cpu or lower.
1262 Corrected a couple of bugs in cpu-dependence in 'insns.dat'.
1264 \b Added to 'standard.mac', the "use16" and "use32" forms of
1265 the "bits 16/32" directive. This is nothing new, just conforms
1266 to a lot of other assemblers. (minor)
1268 \b Changed label allocation from 320/32 (10000 labels @ 200K+)
1269 to 32/37 (1000 labels); makes running under DOS much easier.
1270 Since additional label space is allocated dynamically, this
1271 should have no effect on large programs with lots of labels.
1272 The 37 is a prime, believed to be better for hashing. (minor)
1275 \S{cl-0.98.03} Version 0.98.03
1277 "Integrated patchfile 0.98-0.98.01. I call this version 0.98.03 for
1278 historical reasons: 0.98.02 was trashed." --John Coffman
1279 <johninsd@san.rr.com>, 27-Jul-2000
1281 \b Kendall Bennett's SciTech MGL changes
1283 \b Note that you must define "TASM_COMPAT" at compile-time
1284 to get the Tasm Ideal Mode compatibility.
1286 \b All changes can be compiled in and out using the TASM_COMPAT macros,
1287 and when compiled without TASM_COMPAT defined we get the exact same
1288 binary as the unmodified 0.98 sources.
1290 \b standard.mac, macros.c: Added macros to ignore TASM directives before
1293 \b nasm.h: Added extern declaration for tasm_compatible_mode
1295 \b nasm.c: Added global variable tasm_compatible_mode
1297 \b Added command line switch for TASM compatible mode (-t)
1299 \b Changed version command line to reflect when compiled with TASM additions
1301 \b Added response file processing to allow all arguments on a single
1302 line (response file is @resp rather than -@resp for NASM format).
1304 \b labels.c: Changes islocal() macro to support TASM style @@local labels.
1306 \b Added islocalchar() macro to support TASM style @@local labels.
1308 \b parser.c: Added support for TASM style memory references (ie: mov
1309 [DWORD eax],10 rather than the NASM style mov DWORD [eax],10).
1311 \b preproc.c: Added new directives, \c{%arg}, \c{%local}, \c{%stacksize} to directives
1314 \b Added support for TASM style directives without a leading % symbol.
1316 \b Integrated a block of changes from Andrew Zabolotny <bit@eltech.ru>:
1318 \b A new keyword \c{%xdefine} and its case-insensitive counterpart \c{%ixdefine}.
1319 They work almost the same way as \c{%define} and \c{%idefine} but expand
1320 the definition immediately, not on the invocation. Something like a cross
1321 between \c{%define} and \c{%assign}. The "x" suffix stands for "eXpand", so
1322 "xdefine" can be deciphered as "expand-and-define". Thus you can do
1328 \c %xdefine %1 dword [esp+ofs]
1329 \c %assign ofs ofs+4
1332 \b Changed the place where the expansion of %$name macros are expanded.
1333 Now they are converted into ..@ctxnum.name form when detokenizing, so
1334 there are no quirks as before when using %$name arguments to macros,
1335 in macros etc. For example:
1344 Now last line will be expanded into "hello" as expected. This also allows
1345 for lots of goodies, a good example are extended "proc" macros included
1348 \b Added a check for "cstk" in smacro_defined() before calling get_ctx() -
1349 this allows for things like:
1354 to work without warnings even in no context.
1356 \b Added a check for "cstk" in %if*ctx and %elif*ctx directives -
1357 this allows to use \c{%ifctx} without excessive warnings. If there is
1358 no active context, \c{%ifctx} goes through "false" branch.
1360 \b Removed "user error: " prefix with \c{%error} directive: it just clobbers the
1361 output and has absolutely no functionality. Besides, this allows to write
1362 macros that does not differ from built-in functions in any way.
1364 \b Added expansion of string that is output by \c{%error} directive. Now you
1367 \c %define hello(x) Hello, x!
1369 \c %define %$name andy
1370 \c %error "hello(%$name)"
1372 Same happened with \c{%include} directive.
1374 \b Now all directives that expect an identifier will try to expand and
1375 concatenate everything without whitespaces in between before usage.
1376 For example, with "unfixed" nasm the commands
1378 \c %define %$abc hello
1379 \c %define __%$abc goodbye
1382 would produce "incorrect" output: last line will expand to
1384 \c hello goodbyehello
1386 Not quite what you expected, eh? :-) The answer is that preprocessor
1387 treats the \c{%define} construct as if it would be
1389 \c %define __ %$abc goodbye
1391 (note the white space between __ and %$abc). After my "fix" it
1392 will "correctly" expand into
1396 as expected. Note that I use quotes around words "correct", "incorrect"
1397 etc because this is rather a feature not a bug; however current behaviour
1398 is more logical (and allows more advanced macro usage :-).
1400 Same change was applied to:
1401 \c{%push},\c{%macro},\c{%imacro},\c{%define},\c{%idefine},\c{%xdefine},\c{%ixdefine},
1402 \c{%assign},\c{%iassign},\c{%undef}
1404 \b A new directive [WARNING {+|-}warning-id] have been added. It works only
1405 if the assembly phase is enabled (i.e. it doesn't work with nasm -e).
1407 \b A new warning type: macro-selfref. By default this warning is disabled;
1408 when enabled NASM warns when a macro self-references itself; for example
1409 the following source:
1411 \c [WARNING macro-selfref]
1422 will produce a warning, but if we remove the first line we won't see it
1423 anymore (which is The Right Thing To Do {tm} IMHO since C preprocessor
1424 eats such constructs without warnings at all).
1426 \b Added a "error" routine to preprocessor which always will set ERR_PASS1
1427 bit in severity_code. This removes annoying repeated errors on first
1428 and second passes from preprocessor.
1430 \b Added the %+ operator in single-line macros for concatenating two
1431 identifiers. Usage example:
1433 \c %define _myfunc _otherfunc
1434 \c %define cextern(x) _ %+ x
1437 After first expansion, third line will become "_myfunc". After this
1438 expansion is performed again so it becomes "_otherunc".
1440 \b Now if preprocessor is in a non-emitting state, no warning or error
1441 will be emitted. Example:
1446 \c put anything you want between these two brackets,
1447 \c even macro-parameter references %1 or local
1448 \c labels %$zz or macro-local labels %%zz - no
1449 \c warning will be emitted.
1452 \b Context-local variables on expansion as a last resort are looked up
1453 in outer contexts. For example, the following piece:
1456 \c %define %$a [esp]
1463 will expand correctly the fourth line to [esp]; if we'll define another
1464 %$a inside the "inner" context, it will take precedence over outer
1465 definition. However, this modification has been applied only to
1466 expand_smacro and not to smacro_define: as a consequence expansion
1467 looks in outer contexts, but \c{%ifdef} won't look in outer contexts.
1469 This behaviour is needed because we don't want nested contexts to
1470 act on already defined local macros. Example:
1472 \c %define %$arg1 [esp+4]
1478 In this example the "if" mmacro enters into the "if" context, so %$arg1
1479 is not valid anymore inside "if". Of course it could be worked around
1480 by using explicitely %$$arg1 but this is ugly IMHO.
1482 \b Fixed memory leak in \c{%undef}. The origline wasn't freed before
1485 \b Fixed trap in preprocessor when line expanded to empty set of tokens.
1486 This happens, for example, in the following case:
1488 \c #define SOMETHING
1492 \S{cl-0.98} Version 0.98
1494 All changes since NASM 0.98p3 have been produced by H. Peter Anvin <hpa@zytor.com>.
1496 \b The documentation comment delimiter is \# not #.
1498 \b Allow EQU definitions to refer to external labels; reported by
1501 \b Re-enable support for RDOFF v1; reported by Pedro Gimeno.
1503 \b Updated License file per OK from Simon and Julian.
1506 \S{cl-0.98p9} Version 0.98p9
1508 \b Update documentation (although the instruction set reference will
1509 have to wait; I don't want to hold up the 0.98 release for it.)
1511 \b Verified that the NASM implementation of the PEXTRW and PMOVMSKB
1512 instructions is correct. The encoding differs from what the Intel
1513 manuals document, but the Pentium III behaviour matches NASM, not
1516 \b Fix handling of implicit sizes in PSHUFW and PINSRW, reported by
1519 \b Resurrect the -s option, which was removed when changing the
1520 diagnostic output to stdout.
1523 \S{cl-0.98p8} Version 0.98p8
1525 \b Fix for "DB" when NASM is running on a bigendian machine.
1527 \b Invoke insns.pl once for each output script, making Makefile.in
1528 legal for "make -j".
1530 \b Improve the Unix configure-based makefiles to make package
1533 \b Included an RPM .spec file for building RPM (RedHat Package Manager)
1534 packages on Linux or Unix systems.
1536 \b Fix Makefile dependency problems.
1538 \b Change src/rdsrc.pl to include sectioning information in info
1539 output; required for install-info to work.
1541 \b Updated the RDOFF distribution to version 2 from Jules; minor
1542 massaging to make it compile in my environment.
1544 \b Split doc files that can be built by anyone with a Perl interpreter off
1545 into a separate archive.
1547 \b "Dress rehearsal" release!
1550 \S{cl-0.98p7} Version 0.98p7
1552 \b Fixed opcodes with a third byte-sized immediate argument to not
1553 complain if given "byte" on the immediate.
1555 \b Allow \c{%undef} to remove single-line macros with arguments. This
1556 matches the behaviour of #undef in the C preprocessor.
1558 \b Allow -d, -u, -i and -p to be specified as -D, -U, -I and -P for
1559 compatibility with most C compilers and preprocessors. This allows
1560 Makefile options to be shared between cc and nasm, for example.
1564 \b Went through the list of Katmai instructions and hopefully fixed the
1565 (rather few) mistakes in it.
1567 \b (Hopefully) fixed a number of disassembler bugs related to ambiguous
1568 instructions (disambiguated by -p) and SSE instructions with REP.
1570 \b Fix for bug reported by Mark Junger: "call dword 0x12345678" should
1571 work and may add an OSP (affected CALL, JMP, Jcc).
1573 \b Fix for environments when "stderr" isn't a compile-time constant.
1576 \S{cl-0.98p6} Version 0.98p6
1579 \b Took officially over coordination of the 0.98 release; so drop
1580 the p3.x notation. Skipped p4 and p5 to avoid confusion with John
1581 Fine's J4 and J5 releases.
1583 \b Update the documentation; however, it still doesn't include
1584 documentation for the various new instructions. I somehow wonder if
1585 it makes sense to have an instruction set reference in the assembler
1586 manual when Intel et al have PDF versions of their manuals online.
1588 \b Recognize "idt" or "centaur" for the -p option to ndisasm.
1590 \b Changed error messages back to stderr where they belong, but add an
1591 -E option to redirect them elsewhere (the DOS shell cannot redirect
1594 \b -M option to generate Makefile dependencies (based on code from Alex
1597 \b \c{%undef} preprocessor directive, and -u option, that undefines a
1600 \b OS/2 Makefile (Mkfiles/Makefile.os2) for Borland under OS/2; from
1603 \b Various minor bugfixes (reported by):
1604 - Dangling \c{%s} in preproc.c (Martin Junker)
1606 \b THERE ARE KNOWN BUGS IN SSE AND THE OTHER KATMAI INSTRUCTIONS. I am
1607 on a trip and didn't bring the Katmai instruction reference, so I
1608 can't work on them right now.
1610 \b Updated the License file per agreement with Simon and Jules to
1611 include a GPL distribution clause.
1614 \S{cl-0.98p3.7} Version 0.98p3.7
1616 \b (Hopefully) fixed the canned Makefiles to include the outrdf2 and
1619 \b Renamed changes.asm to changed.asm.
1622 \S{cl-0.98p3.6} Version 0.98p3.6
1624 \b Fixed a bunch of instructions that were added in 0.98p3.5 which had
1625 memory operands, and the address-size prefix was missing from the
1626 instruction pattern.
1629 \S{cl-0.98p3.5} Version 0.98p3.5
1631 \b Merged in changes from John S. Fine's 0.98-J5 release. John's based
1632 0.98-J5 on my 0.98p3.3 release; this merges the changes.
1634 \b Expanded the instructions flag field to a long so we can fit more
1635 flags; mark SSE (KNI) and AMD or Katmai-specific instructions as
1638 \b Fix the "PRIV" flag on a bunch of instructions, and create new
1639 "PROT" flag for protected-mode-only instructions (orthogonal to if
1640 the instruction is privileged!) and new "SMM" flag for SMM-only
1643 \b Added AMD-only SYSCALL and SYSRET instructions.
1645 \b Make SSE actually work, and add new Katmai MMX instructions.
1647 \b Added a -p (preferred vendor) option to ndisasm so that it can
1648 distinguish e.g. Cyrix opcodes also used in SSE. For example:
1650 \c ndisasm -p cyrix aliased.bin
1651 \c 00000000 670F514310 paddsiw mm0,[ebx+0x10]
1652 \c 00000005 670F514320 paddsiw mm0,[ebx+0x20]
1653 \c ndisasm -p intel aliased.bin
1654 \c 00000000 670F514310 sqrtps xmm0,[ebx+0x10]
1655 \c 00000005 670F514320 sqrtps xmm0,[ebx+0x20]
1657 \b Added a bunch of Cyrix-specific instructions.
1660 \S{cl-0.98p3.4} Version 0.98p3.4
1662 \b Made at least an attempt to modify all the additional Makefiles (in
1663 the Mkfiles directory). I can't test it, but this was the best I
1666 \b DOS DJGPP+"Opus Make" Makefile from John S. Fine.
1668 \b changes.asm changes from John S. Fine.
1671 \S{cl-0.98p3.3} Version 0.98p3.3
1673 \b Patch from Conan Brink to allow nesting of \c{%rep} directives.
1675 \b If we're going to allow INT01 as an alias for INT1/ICEBP (one of
1676 Jules 0.98p3 changes), then we should allow INT03 as an alias for INT3
1679 \b Updated changes.asm to include the latest changes.
1681 \b Tried to clean up the <CR>s that had snuck in from a DOS/Windows
1682 environment into my Unix environment, and try to make sure than
1683 DOS/Windows users get them back.
1685 \b We would silently generate broken tools if insns.dat wasn't sorted
1686 properly. Change insns.pl so that the order doesn't matter.
1688 \b Fix bug in insns.pl (introduced by me) which would cause conditional
1689 instructions to have an extra "cc" in disassembly, e.g. "jnz"
1690 disassembled as "jccnz".
1693 \S{cl-0.98p3.2} Version 0.98p3.2
1695 \b Merged in John S. Fine's changes from his 0.98-J4 prerelease; see
1696 http://www.csoft.net/cz/johnfine/
1698 \b Changed previous "spotless" Makefile target (appropriate for distribution)
1699 to "distclean", and added "cleaner" target which is same as "clean"
1700 except deletes files generated by Perl scripts; "spotless" is union.
1702 \b Removed BASIC programs from distribution. Get a Perl interpreter
1703 instead (see below.)
1705 \b Calling this "pre-release 3.2" rather than "p3-hpa2" because of
1706 John's contributions.
1708 \b Actually link in the IEEE output format (zoutieee.c); fix a bunch of
1709 compiler warnings in that file. Note I don't know what IEEE output
1710 is supposed to look like, so these changes were made "blind".
1713 \S{cl-0.98p3-hpa} Version 0.98p3-hpa
1715 \b Merged nasm098p3.zip with nasm-0.97.tar.gz to create a fully
1716 buildable version for Unix systems (Makefile.in updates, etc.)
1718 \b Changed insns.pl to create the instruction tables in nasm.h and
1719 names.c, so that a new instruction can be added by adding it *only*
1722 \b Added the following new instructions: SYSENTER, SYSEXIT, FXSAVE,
1723 FXRSTOR, UD1, UD2 (the latter two are two opcodes that Intel
1724 guarantee will never be used; one of them is documented as UD2 in
1725 Intel documentation, the other one just as "Undefined Opcode" --
1726 calling it UD1 seemed to make sense.)
1728 \b MAX_SYMBOL was defined to be 9, but LOADALL286 and LOADALL386 are 10
1729 characters long. Now MAX_SYMBOL is derived from insns.dat.
1731 \b A note on the BASIC programs included: forget them. insns.bas is
1732 already out of date. Get yourself a Perl interpreter for your
1733 platform of choice at
1734 \W{http://www.cpan.org/ports/index.html}{http://www.cpan.org/ports/index.html}.
1737 \S{cl-0.98p3} Version 0.98 pre-release 3
1739 \b added response file support, improved command line handling, new layout
1742 \b fixed limit checking bug, 'OUT byte nn, reg' bug, and a couple of rdoff
1743 related bugs, updated Wishlist; 0.98 Prerelease 3.
1746 \S{cl-0.98p2} Version 0.98 pre-release 2
1748 \b fixed bug in outcoff.c to do with truncating section names longer
1749 than 8 characters, referencing beyond end of string; 0.98 pre-release 2
1752 \S{cl-0.98p1} Version 0.98 pre-release 1
1754 \b Fixed a bug whereby STRUC didn't work at all in RDF.
1756 \b Fixed a problem with group specification in PUBDEFs in OBJ.
1758 \b Improved ease of adding new output formats. Contribution due to
1761 \b Fixed a bug in relocations in the `bin' format: was showing up when
1762 a relocatable reference crossed an 8192-byte boundary in any output
1765 \b Fixed a bug in local labels: local-label lookups were inconsistent
1766 between passes one and two if an EQU occurred between the definition
1767 of a global label and the subsequent use of a local label local to
1770 \b Fixed a seg-fault in the preprocessor (again) which happened when
1771 you use a blank line as the first line of a multi-line macro
1772 definition and then defined a label on the same line as a call to
1775 \b Fixed a stale-pointer bug in the handling of the NASM environment
1776 variable. Thanks to Thomas McWilliams.
1778 \b ELF had a hard limit on the number of sections which caused
1779 segfaults when transgressed. Fixed.
1781 \b Added ability for ndisasm to read from stdin by using `-' as the
1784 \b ndisasm wasn't outputting the TO keyword. Fixed.
1786 \b Fixed error cascade on bogus expression in \c{%if} - an error in
1787 evaluation was causing the entire \c{%if} to be discarded, thus creating
1788 trouble later when the \c{%else} or \c{%endif} was encountered.
1790 \b Forward reference tracking was instruction-granular not operand-
1791 granular, which was causing 286-specific code to be generated
1792 needlessly on code of the form `shr word [forwardref],1'. Thanks to
1793 Jim Hague for sending a patch.
1795 \b All messages now appear on stdout, as sending them to stderr serves
1796 no useful purpose other than to make redirection difficult.
1798 \b Fixed the problem with EQUs pointing to an external symbol - this
1799 now generates an error message.
1801 \b Allowed multiple size prefixes to an operand, of which only the first
1802 is taken into account.
1804 \b Incorporated John Fine's changes, including fixes of a large number
1805 of preprocessor bugs, some small problems in OBJ, and a reworking of
1806 label handling to define labels before their line is assembled, rather
1809 \b Reformatted a lot of the source code to be more readable. Included
1810 'coding.txt' as a guideline for how to format code for contributors.
1812 \b Stopped nested \c{%reps} causing a panic - they now cause a slightly more
1813 friendly error message instead.
1815 \b Fixed floating point constant problems (patch by Pedro Gimeno)
1817 \b Fixed the return value of insn_size() not being checked for -1, indicating
1820 \b Incorporated 3Dnow! instructions.
1822 \b Fixed the 'mov eax, eax + ebx' bug.
1824 \b Fixed the GLOBAL EQU bug in ELF. Released developers release 3.
1826 \b Incorporated John Fine's command line parsing changes
1828 \b Incorporated David Lindauer's OMF debug support
1830 \b Made changes for LCC 4.0 support (\c{__NASM_CDecl__}, removed register size
1831 specification warning when sizes agree).
1834 \H{cl-0.9x} NASM 0.9 Series
1836 Revisions before 0.98.
1839 \S{cl-0.97} Version 0.97 released December 1997
1841 \b This was entirely a bug-fix release to 0.96, which seems to have got
1844 \b Fixed stupid mistake in OBJ which caused `MOV EAX,<constant>' to
1845 fail. Caused by an error in the `MOV EAX,<segment>' support.
1847 \b ndisasm hung at EOF when compiled with lcc on Linux because lcc on
1848 Linux somehow breaks feof(). ndisasm now does not rely on feof().
1850 \b A heading in the documentation was missing due to a markup error in
1851 the indexing. Fixed.
1853 \b Fixed failure to update all pointers on realloc() within extended-
1854 operand code in parser.c. Was causing wrong behaviour and seg faults
1855 on lines such as `dd 0.0,0.0,0.0,0.0,...'
1857 \b Fixed a subtle preprocessor bug whereby invoking one multi-line
1858 macro on the first line of the expansion of another, when the second
1859 had been invoked with a label defined before it, didn't expand the
1862 \b Added internal.doc back in to the distribution archives - it was
1863 missing in 0.96 *blush*
1865 \b Fixed bug causing 0.96 to be unable to assemble its own test files,
1866 specifically objtest.asm. *blush again*
1868 \b Fixed seg-faults and bogus error messages caused by mismatching
1869 \c{%rep} and \c{%endrep} within multi-line macro definitions.
1871 \b Fixed a problem with buffer overrun in OBJ, which was causing
1872 corruption at ends of long PUBDEF records.
1874 \b Separated DOS archives into main-program and documentation to reduce
1878 \S{cl-0.96} Version 0.96 released November 1997
1880 \b Fixed a bug whereby, if `nasm sourcefile' would cause a filename
1881 collision warning and put output into `nasm.out', then `nasm
1882 sourcefile -o outputfile' still gave the warning even though the
1884 Fixed name pollution under Digital UNIX: one of its header files
1885 defined R_SP, which broke the enum in nasm.h.
1887 \b Fixed minor instruction table problems: FUCOM and FUCOMP didn't have
1888 two-operand forms; NDISASM didn't recognise the longer register
1889 forms of PUSH and POP (eg FF F3 for PUSH BX); TEST mem,imm32 was
1890 flagged as undocumented; the 32-bit forms of CMOV had 16-bit operand
1891 size prefixes; `AAD imm' and `AAM imm' are no longer flagged as
1892 undocumented because the Intel Architecture reference documents
1895 \b Fixed a problem with the local-label mechanism, whereby strange
1896 types of symbol (EQUs, auto-defined OBJ segment base symbols)
1897 interfered with the `previous global label' value and screwed up
1900 \b Fixed a bug whereby the stub preprocessor didn't communicate with
1901 the listing file generator, so that the -a and -l options in
1902 conjunction would produce a useless listing file.
1904 \b Merged `os2' object file format back into `obj', after discovering
1905 that `obj' _also_ shouldn't have a link pass separator in a module
1906 containing a non-trivial MODEND. Flat segments are now declared
1907 using the FLAT attribute. `os2' is no longer a valid object format
1910 \b Removed the fixed-size temporary storage in the evaluator. Very very
1911 long expressions (like `mov ax,1+1+1+1+...' for two hundred 1s or
1912 so) should now no longer crash NASM.
1914 \b Fixed a bug involving segfaults on disassembly of MMX instructions,
1915 by changing the meaning of one of the operand-type flags in nasm.h.
1916 This may cause other apparently unrelated MMX problems; it needs to
1917 be tested thoroughly.
1919 \b Fixed some buffer overrun problems with large OBJ output files.
1920 Thanks to DJ Delorie for the bug report and fix.
1922 \b Made preprocess-only mode actually listen to the \c{%line} markers as it
1923 prints them, so that it can report errors more sanely.
1925 \b Re-designed the evaluator to keep more sensible track of expressions
1926 involving forward references: can now cope with previously-nightmare
1933 \b Added the ALIGN and ALIGNB standard macros.
1935 \b Added PIC support in ELF: use of WRT to obtain the four extra
1936 relocation types needed.
1938 \b Added the ability for output file formats to define their own
1939 extensions to the GLOBAL, COMMON and EXTERN directives.
1941 \b Implemented common-variable alignment, and global-symbol type and
1942 size declarations, in ELF.
1944 \b Implemented NEAR and FAR keywords for common variables, plus
1945 far-common element size specification, in OBJ.
1947 \b Added a feature whereby EXTERNs and COMMONs in OBJ can be given a
1948 default WRT specification (either a segment or a group).
1950 \b Transformed the Unix NASM archive into an auto-configuring package.
1952 \b Added a sanity-check for people applying SEG to things which are
1953 already segment bases: this previously went unnoticed by the SEG
1954 processing and caused OBJ-driver panics later.
1956 \b Added the ability, in OBJ format, to deal with `MOV EAX,<segment>'
1957 type references: OBJ doesn't directly support dword-size segment
1958 base fixups, but as long as the low two bytes of the constant term
1959 are zero, a word-size fixup can be generated instead and it will
1962 \b Added the ability to specify sections' alignment requirements in
1963 Win32 object files and pure binary files.
1965 \b Added preprocess-time expression evaluation: the \c{%assign} (and
1966 \c{%iassign}) directive and the bare \c{%if} (and \c{%elif}) conditional. Added
1967 relational operators to the evaluator, for use only in \c{%if}
1968 constructs: the standard relationals = < > <= >= <> (and C-like
1969 synonyms == and !=) plus low-precedence logical operators &&, ^^ and
1972 \b Added a preprocessor repeat construct: \c{%rep} / \c{%exitrep} / \c{%endrep}.
1974 \b Added the __FILE__ and __LINE__ standard macros.
1976 \b Added a sanity check for number constants being greater than
1977 0xFFFFFFFF. The warning can be disabled.
1979 \b Added the %0 token whereby a variadic multi-line macro can tell how
1980 many parameters it's been given in a specific invocation.
1982 \b Added \c{%rotate}, allowing multi-line macro parameters to be cycled.
1984 \b Added the `*' option for the maximum parameter count on multi-line
1985 macros, allowing them to take arbitrarily many parameters.
1987 \b Added the ability for the user-level forms of EXTERN, GLOBAL and
1988 COMMON to take more than one argument.
1990 \b Added the IMPORT and EXPORT directives in OBJ format, to deal with
1993 \b Added some more preprocessor \c{%if} constructs: \c{%ifidn} / \c{%ifidni} (exact
1994 textual identity), and \c{%ifid} / \c{%ifnum} / \c{%ifstr} (token type testing).
1996 \b Added the ability to distinguish SHL AX,1 (the 8086 version) from
1997 SHL AX,BYTE 1 (the 286-and-upwards version whose constant happens to
2000 \b Added NetBSD/FreeBSD/OpenBSD's variant of a.out format, complete
2001 with PIC shared library features.
2003 \b Changed NASM's idiosyncratic handling of FCLEX, FDISI, FENI, FINIT,
2004 FSAVE, FSTCW, FSTENV, and FSTSW to bring it into line with the
2005 otherwise accepted standard. The previous behaviour, though it was a
2006 deliberate feature, was a deliberate feature based on a
2007 misunderstanding. Apologies for the inconvenience.
2009 \b Improved the flexibility of ABSOLUTE: you can now give it an
2010 expression rather than being restricted to a constant, and it can
2011 take relocatable arguments as well.
2013 \b Added the ability for a variable to be declared as EXTERN multiple
2014 times, and the subsequent definitions are just ignored.
2016 \b We now allow instruction prefixes (CS, DS, LOCK, REPZ etc) to be
2017 alone on a line (without a following instruction).
2019 \b Improved sanity checks on whether the arguments to EXTERN, GLOBAL
2020 and COMMON are valid identifiers.
2022 \b Added misc/exebin.mac to allow direct generation of .EXE files by
2023 hacking up an EXE header using DB and DW; also added test/binexe.asm
2024 to demonstrate the use of this. Thanks to Yann Guidon for
2025 contributing the EXE header code.
2027 \b ndisasm forgot to check whether the input file had been successfully
2028 opened. Now it does. Doh!
2030 \b Added the Cyrix extensions to the MMX instruction set.
2032 \b Added a hinting mechanism to allow [EAX+EBX] and [EBX+EAX] to be
2033 assembled differently. This is important since [ESI+EBP] and
2034 [EBP+ESI] have different default base segment registers.
2036 \b Added support for the PharLap OMF extension for 4096-byte segment
2040 \S{cl-0.95 released July 1997} Version 0.95 released July 1997
2042 \b Fixed yet another ELF bug. This one manifested if the user relied on
2043 the default segment, and attempted to define global symbols without
2044 first explicitly declaring the target segment.
2046 \b Added makefiles (for NASM and the RDF tools) to build Win32 console
2047 apps under Symantec C++. Donated by Mark Junker.
2049 \b Added `macros.bas' and `insns.bas', QBasic versions of the Perl
2050 scripts that convert `standard.mac' to `macros.c' and convert
2051 `insns.dat' to `insnsa.c' and `insnsd.c'. Also thanks to Mark
2054 \b Changed the diassembled forms of the conditional instructions so
2055 that JB is now emitted as JC, and other similar changes. Suggested
2056 list by Ulrich Doewich.
2058 \b Added `@' to the list of valid characters to begin an identifier
2061 \b Documentary changes, notably the addition of the `Common Problems'
2062 section in nasm.doc.
2064 \b Fixed a bug relating to 32-bit PC-relative fixups in OBJ.
2066 \b Fixed a bug in perm_copy() in labels.c which was causing exceptions
2067 in cleanup_labels() on some systems.
2069 \b Positivity sanity check in TIMES argument changed from a warning to
2070 an error following a further complaint.
2072 \b Changed the acceptable limits on byte and word operands to allow
2073 things like `~10111001b' to work.
2075 \b Fixed a major problem in the preprocessor which caused seg-faults if
2076 macro definitions contained blank lines or comment-only lines.
2078 \b Fixed inadequate error checking on the commas separating the
2079 arguments to `db', `dw' etc.
2081 \b Fixed a crippling bug in the handling of macros with operand counts
2082 defined with a `+' modifier.
2084 \b Fixed a bug whereby object file formats which stored the input file
2085 name in the output file (such as OBJ and COFF) weren't doing so
2086 correctly when the output file name was specified on the command
2089 \b Removed [INC] and [INCLUDE] support for good, since they were
2092 \b Fixed a bug in OBJ which caused all fixups to be output in 16-bit
2093 (old-format) FIXUPP records, rather than putting the 32-bit ones in
2094 FIXUPP32 (new-format) records.
2096 \b Added, tentatively, OS/2 object file support (as a minor variant on
2099 \b Updates to Fox Cutter's Borland C makefile, Makefile.bc2.
2101 \b Removed a spurious second fclose() on the output file.
2103 \b Added the `-s' command line option to redirect all messages which
2104 would go to stderr (errors, help text) to stdout instead.
2106 \b Added the `-w' command line option to selectively suppress some
2107 classes of assembly warning messages.
2109 \b Added the `-p' pre-include and `-d' pre-define command-line options.
2111 \b Added an include file search path: the `-i' command line option.
2113 \b Fixed a silly little preprocessor bug whereby starting a line with a
2114 `%!' environment-variable reference caused an `unknown directive'
2117 \b Added the long-awaited listing file support: the `-l' command line
2120 \b Fixed a problem with OBJ format whereby, in the absence of any
2121 explicit segment definition, non-global symbols declared in the
2122 implicit default segment generated spurious EXTDEF records in the
2125 \b Added the NASM environment variable.
2127 \b From this version forward, Win32 console-mode binaries will be
2128 included in the DOS distribution in addition to the 16-bit binaries.
2129 Added Makefile.vc for this purpose.
2131 \b Added `return 0;' to test/objlink.c to prevent compiler warnings.
2133 \b Added the __NASM_MAJOR__ and __NASM_MINOR__ standard defines.
2135 \b Added an alternative memory-reference syntax in which prefixing an
2136 operand with `&' is equivalent to enclosing it in square brackets,
2137 at the request of Fox Cutter.
2139 \b Errors in pass two now cause the program to return a non-zero error
2140 code, which they didn't before.
2142 \b Fixed the single-line macro cycle detection, which didn't work at
2143 all on macros with no parameters (caused an infinite loop). Also
2144 changed the behaviour of single-line macro cycle detection to work
2145 like cpp, so that macros like `extrn' as given in the documentation
2148 \b Fixed the implementation of WRT, which was too restrictive in that
2149 you couldn't do `mov ax,[di+abc wrt dgroup]' because (di+abc) wasn't
2150 a relocatable reference.
2153 \S{cl-0.94 released April 1997} Version 0.94 released April 1997
2156 \b Major item: added the macro processor.
2158 \b Added undocumented instructions SMI, IBTS, XBTS and LOADALL286. Also
2159 reorganised CMPXCHG instruction into early-486 and Pentium forms.
2160 Thanks to Thobias Jones for the information.
2162 \b Fixed two more stupid bugs in ELF, which were causing `ld' to
2163 continue to seg-fault in a lot of non-trivial cases.
2165 \b Fixed a seg-fault in the label manager.
2167 \b Stopped FBLD and FBSTP from _requiring_ the TWORD keyword, which is
2168 the only option for BCD loads/stores in any case.
2170 \b Ensured FLDCW, FSTCW and FSTSW can cope with the WORD keyword, if
2171 anyone bothers to provide it. Previously they complained unless no
2172 keyword at all was present.
2174 \b Some forms of FDIV/FDIVR and FSUB/FSUBR were still inverted: a
2175 vestige of a bug that I thought had been fixed in 0.92. This was
2176 fixed, hopefully for good this time...
2178 \b Another minor phase error (insofar as a phase error can _ever_ be
2179 minor) fixed, this one occurring in code of the form
2181 \c rol ax,forward_reference
2182 \c forward_reference equ 1
2184 \b The number supplied to TIMES is now sanity-checked for positivity,
2185 and also may be greater than 64K (which previously didn't work on
2188 \b Added Watcom C makefiles, and misc/pmw.bat, donated by Dominik Behr.
2190 \b Added the INCBIN pseudo-opcode.
2192 \b Due to the advent of the preprocessor, the [INCLUDE] and [INC]
2193 directives have become obsolete. They are still supported in this
2194 version, with a warning, but won't be in the next.
2196 \b Fixed a bug in OBJ format, which caused incorrect object records to
2197 be output when absolute labels were made global.
2199 \b Updates to RDOFF subdirectory, and changes to outrdf.c.
2202 \S{cl-0.93 released January 1997} Version 0.93 released January 1997
2204 This release went out in a great hurry after semi-crippling bugs
2207 \b Really \e{did} fix the stack overflows this time. *blush*
2209 \b Had problems with EA instruction sizes changing between passes, when
2210 an offset contained a forward reference and so 4 bytes were
2211 allocated for the offset in pass one; by pass two the symbol had
2212 been defined and happened to be a small absolute value, so only 1
2213 byte got allocated, causing instruction size mismatch between passes
2214 and hence incorrect address calculations. Fixed.
2216 \b Stupid bug in the revised ELF section generation fixed (associated
2217 string-table section for .symtab was hard-coded as 7, even when this
2218 didn't fit with the real section table). Was causing `ld' to
2219 seg-fault under Linux.
2221 \b Included a new Borland C makefile, Makefile.bc2, donated by Fox
2222 Cutter <lmb@comtch.iea.com>.
2225 \S{cl-0.92 released January 1997} Version 0.92 released January 1997
2227 \b The FDIVP/FDIVRP and FSUBP/FSUBRP pairs had been inverted: this was
2228 fixed. This also affected the LCC driver.
2230 \b Fixed a bug regarding 32-bit effective addresses of the form
2231 \c{[other_register+ESP]}.
2233 \b Documentary changes, notably documentation of the fact that Borland
2234 Win32 compilers use `obj' rather than `win32' object format.
2236 \b Fixed the COMENT record in OBJ files, which was formatted
2239 \b Fixed a bug causing segfaults in large RDF files.
2241 \b OBJ format now strips initial periods from segment and group
2242 definitions, in order to avoid complications with the local label
2245 \b Fixed a bug in disassembling far calls and jumps in NDISASM.
2247 \b Added support for user-defined sections in COFF and ELF files.
2249 \b Compiled the DOS binaries with a sensible amount of stack, to
2250 prevent stack overflows on any arithmetic expression containing
2253 \b Fixed a bug in handling of files that do not terminate in a newline.
2256 \S{cl-0.91 released November 1996} Version 0.91 released November 1996
2258 \b Loads of bug fixes.
2260 \b Support for RDF added.
2262 \b Support for DBG debugging format added.
2264 \b Support for 32-bit extensions to Microsoft OBJ format added.
2266 \b Revised for Borland C: some variable names changed, makefile added.
2268 \b LCC support revised to actually work.
2270 \b JMP/CALL NEAR/FAR notation added.
2272 \b `a16', `o16', `a32' and `o32' prefixes added.
2274 \b Range checking on short jumps implemented.
2276 \b MMX instruction support added.
2278 \b Negative floating point constant support added.
2280 \b Memory handling improved to bypass 64K barrier under DOS.
2282 \b \c{$} prefix to force treatment of reserved words as identifiers added.
2284 \b Default-size mechanism for object formats added.
2286 \b Compile-time configurability added.
2288 \b \c{#}, \c{@}, \c{~} and c\{?} are now valid characters in labels.
2290 \b \c{-e} and \c{-k} options in NDISASM added.
2293 \S{cl-0.90 released October 1996} Version 0.90 released October 1996
2295 First release version. First support for object file output. Other
2296 changes from previous version (0.3x) too numerous to document.