H. Peter Anvin [Thu, 19 Aug 2010 23:59:35 +0000 (16:59 -0700)]
ndisasm: unify VEX handling
Unify VEX handling between the 026x and 0270 bytecodes.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Tue, 17 Aug 2010 05:13:14 +0000 (22:13 -0700)]
ndisasm: fix handing of byte codes 250-253, 324
Fix handling of byte codes 250-253 (sign-extended 32-bit immediate,
extended to 64 bits) and 324 (instruction must have osize 64).
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Mon, 16 Aug 2010 22:24:01 +0000 (15:24 -0700)]
NASM 2.09rc7
H. Peter Anvin [Mon, 16 Aug 2010 22:23:16 +0000 (15:23 -0700)]
test/avx: remove deleted instructions
Remove the deleted VPERMIL2 instructions.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 22:22:21 +0000 (15:22 -0700)]
insns.dat: permit contracted forms for VBLENDVP
Allow implicit operands for VBLENDVP, just as for other instructions,
since the semi-legacy forms now are removed.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 22:21:48 +0000 (15:21 -0700)]
assmemble.c: fix VEX.W logic
Fix the generation logic for VEX.W, which unfortunately got the wrong
constants.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 22:08:51 +0000 (15:08 -0700)]
insns.dat: fix encoding of VCVTSD2SS
Fix typo in the definition of VCVTSD2SS.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 22:05:25 +0000 (15:05 -0700)]
insns.dat: SSE encoding of VBLEND with VEX prefix is forbidden
Version 7 of the AVX spec specifically forbids (#UD) using the
66 0F 38 14/15 forms of the BLENDV instructions with a VEX prefix;
those encodings are strictly legacy SSE 4.1.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 21:57:09 +0000 (14:57 -0700)]
insns.dat: updates from AVX v7
Updates from the AVX version 7 specification: mostly tightening of the
rules for VEX.L and VEX.W, but remove the VPERMIL2 instructions.
Also encode all the full-length forms of the VCMP instructions and
prefer those for the disassembly.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 21:56:33 +0000 (14:56 -0700)]
assemble: handle vex.lig
AVX version 7 introduces the concept of .lig, meaning VEX.L is
ignored.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 16 Aug 2010 20:48:43 +0000 (13:48 -0700)]
vex: change .wx to .wig to match the latest AVX spec
Change the .wx (ignore the W field) to .wig, to match the latest
version of the AVX specification. This is not a functional change,
but just makes instruction patterns a little easier to write.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cyrill Gorcunov [Mon, 16 Aug 2010 09:53:22 +0000 (13:53 +0400)]
doc: Update -O option description
We use -Ox by default since Nasm-2.09 so "-O0" is not
longer "default" option. But first paragraph still has
the reference which confuses people. Get rid of it.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 16 Aug 2010 09:47:24 +0000 (13:47 +0400)]
doc: Fix typo in changes.src
Reported-by: Frank Kotler <fbkotler@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Keith Kanios [Sat, 14 Aug 2010 17:47:45 +0000 (12:47 -0500)]
preproc.c: revamped context-local fall-through warning message
Keith Kanios [Sat, 14 Aug 2010 17:36:49 +0000 (12:36 -0500)]
doc/nasmdoc.src: revamped context fall-through entry
Cyrill Gorcunov [Fri, 13 Aug 2010 08:32:53 +0000 (12:32 +0400)]
doc: Describe context-through lookup
Feel free to update this text as well,
and check it please for being more/less
readable.
CC: Keith Kanios <keith@kanios.net>
CC: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Fri, 13 Aug 2010 02:53:37 +0000 (19:53 -0700)]
LICENSE: update year
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cyrill Gorcunov [Thu, 12 Aug 2010 15:09:28 +0000 (19:09 +0400)]
sectalign: Implement on/off mode
And describe it in documentation. We've introduced
sectalign in 2.09rc series and have to provide an
ability to steer this feature.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 10 Aug 2010 10:19:03 +0000 (14:19 +0400)]
standart.mac: Style nitfix
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Wed, 11 Aug 2010 16:31:46 +0000 (20:31 +0400)]
preproc.c: Context-through single macros expansion is deprecated
For now we inform users about their sources need to be
updated and also since _all_ context case are legit
for single macros only we split lookup into two phases:
1) Lookup in active context, which is perfectly valid
2) Lookup in external contexts, which will be deprecated soon.
If (2) happens we yield warning.
A typical testcase is
---
%macro one 0
%push
%$a:
%assign %$b 12
%push
mov eax, %$a
mov eax, %$b ; hit -- context through
%pop
%pop
%endmacro
one
---
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Keith Kanios [Wed, 11 Aug 2010 01:12:57 +0000 (20:12 -0500)]
preproc.c: modified deprecation warning for context-local label fallthrough
Cyrill Gorcunov [Mon, 9 Aug 2010 16:17:17 +0000 (20:17 +0400)]
preproc.c: Fix error message typo
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 9 Aug 2010 14:42:57 +0000 (18:42 +0400)]
test: Add br3041451 testcase
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 9 Aug 2010 09:58:22 +0000 (13:58 +0400)]
BR3041451: Implement upper bound for %rep counter
Since %rep counter is a 64 bit signed integer we have to use some
"maximum possible value" limit (upper bound) otherwise there may be
a situation when %rep counter is 0 or even negative while user
has been passing big positive integer value.
Reported-by: nasm64developer
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Keith Kanios [Mon, 9 Aug 2010 05:55:44 +0000 (00:55 -0500)]
preproc.c: added deprecation warning for context-local label fallthrough
Cyrill Gorcunov [Fri, 6 Aug 2010 15:23:31 +0000 (19:23 +0400)]
NASM 2.09rc6
It likely to be the last -rc before final relase.
Give it a good testing. We should be in pretty good
shape.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Wed, 4 Aug 2010 13:27:17 +0000 (17:27 +0400)]
test: Add automatizing annotations to imm64.asm
H. Peter Anvin pointed
|
| Btw, test/imm64.asm needs test engine annotations.
|
Make it so.
Reported-by: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Mon, 2 Aug 2010 19:04:50 +0000 (12:04 -0700)]
insns.dat: unbreak test/imm64.bin
Unbreak the following subtest in test/imm64.bin:
mov qword [rax],dword 11223344h
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cyrill Gorcunov [Thu, 29 Jul 2010 06:12:29 +0000 (10:12 +0400)]
elf: Move stabs symbol table format into outelf.h
Get rid of code duplication
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Wed, 28 Jul 2010 19:27:11 +0000 (23:27 +0400)]
nasmlib.c: fix typo in nasm_init_malloc_error
Typo happened in
9b603082 so -DLOGALLOC gets broken.
Not that important since this is a developer oriented
feature but should be fixed anyway.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Wed, 28 Jul 2010 18:03:43 +0000 (22:03 +0400)]
NASM 2.09rc5
Cyrill Gorcunov [Wed, 28 Jul 2010 09:57:01 +0000 (13:57 +0400)]
doc: Updates on macro ranges
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Wed, 28 Jul 2010 06:18:48 +0000 (10:18 +0400)]
labels.c: cleanup
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 27 Jul 2010 11:41:03 +0000 (15:41 +0400)]
Get rid of misc/altreg.inc
It's the same as macros/altreg.mac which is already included.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Wed, 28 Jul 2010 13:59:49 +0000 (17:59 +0400)]
doc: Explain SECTALIGN
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 26 Jul 2010 21:48:27 +0000 (01:48 +0400)]
doc: Fix ..sym section name misprint
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 26 Jul 2010 19:14:40 +0000 (23:14 +0400)]
assemble.c: Style nitfix
Various tabs/space mixture cleaned and some more.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Sun, 25 Jul 2010 19:43:30 +0000 (12:43 -0700)]
Revert "Improve process_ea and introduce -OL"
This reverts commit
ac732cb6a599836bf4c988e59ac6de4498758c72.
Resolved Conflicts:
doc/nasmdoc.src
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Sun, 25 Jul 2010 07:17:12 +0000 (00:17 -0700)]
nasm.h: move "regs.h" to top; make is_register() a bool
Move the #include of regs.h to the top with the other includes, and
make is_register() -- being a boolean predicate -- return bool.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cyrill Gorcunov [Sun, 25 Jul 2010 07:09:56 +0000 (11:09 +0400)]
move is_register helper to nasm.h
H. Peter Anvin noted:
|
| Could we avoid putting static code that has no dynamic content in
| dynamically generated files... it just complicates things unnecessarily.
|
| The above can move into nasm.h for example.
|
Good idea.
CC: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Sat, 24 Jul 2010 21:16:33 +0000 (01:16 +0400)]
Use is_register helper
Save us some line of code
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Sat, 24 Jul 2010 21:10:17 +0000 (01:10 +0400)]
regs.pl: Introduce is_register() helper
Instead of opencoded check use inline helper.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Victor van den Elzen [Sat, 24 Jul 2010 20:24:15 +0000 (22:24 +0200)]
Remove redundant sentence in docs
Victor van den Elzen [Tue, 31 Mar 2009 02:59:44 +0000 (04:59 +0200)]
Improve process_ea and introduce -OL
Two fixes:
1. Optimization of [bx+0xFFFF] etc
0xFFFF is an sbyte under 16-bit semantics,
so make sure to check it right.
2. Don't optimize displacements in -O0
Displacements that fit into an sbyte or
can be removed should *not* be optimized in -O0.
Implicit zero displacements are still optimized, e.g.:
[eax] -> 0 bit displacement, [ebp] -> 8 bit displacement.
However explicit displacements are not optimized:
[eax+0] -> 32 bit displacement, [ebp+0] -> 32 bit displacement.
Because #2 breaks compatibility with 0.98,
I introduced a new optimization level: -OL, legacy.
Cyrill Gorcunov [Fri, 23 Jul 2010 14:57:40 +0000 (18:57 +0400)]
doc: Update changes.src
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Fri, 23 Jul 2010 14:51:51 +0000 (18:51 +0400)]
preproc.: Fix NULL dereference on broken %strlen argument
Under particular circumstances %strlen may cause SIGSEG. A typical
example is %strlen with nonexistent macro argument.
[ Testcase test/strlen.asm ]
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Tue, 20 Jul 2010 21:56:30 +0000 (14:56 -0700)]
preproc: allow non-identifier character in environment variables
Allow non-identifier characters in the name of environment variables,
by surrounding them with string quotes (subject to ordinary
string-quoting rules.)
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cyrill Gorcunov [Sat, 17 Jul 2010 14:24:13 +0000 (18:24 +0400)]
changes.src: Fix misprint in "instructions" word.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Fri, 16 Jul 2010 05:51:26 +0000 (22:51 -0700)]
nasmdoc: we still miss Chuck, but...
We still miss Chuck, but I don't think we can really say anymore we're
doing this particular release in his memory.
Requiescat in pace, our friend.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Fri, 16 Jul 2010 05:45:04 +0000 (22:45 -0700)]
Merge branch 'nasm-2.08.xx'
Resolved Conflicts:
version
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Fri, 16 Jul 2010 05:44:03 +0000 (22:44 -0700)]
changes.src: update with the 2.08.xx changes
Patchlevels should be documented, too...
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Fri, 16 Jul 2010 01:32:32 +0000 (18:32 -0700)]
NASM 2.09rc4
H. Peter Anvin [Fri, 16 Jul 2010 01:28:52 +0000 (18:28 -0700)]
Make -Ox the default
Make -Ox the default; it's the optimization level expected by most
users, and it is clearly still causing confusion that it has to be
specified manually.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Bryant Keller [Tue, 13 Jul 2010 19:49:39 +0000 (23:49 +0400)]
Documentation: describe %ifenv and friends
Signed-off-by: Bryant Keller <bkeller@assembly.ath.cx>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Tue, 13 Jul 2010 19:00:58 +0000 (12:00 -0700)]
preproc: add %ifenv
Add %ifenv to test for the presence of an environment variable. The
environment variable can, but does not have to be, prefixed with %!.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Tue, 13 Jul 2010 18:43:06 +0000 (11:43 -0700)]
BR3028880: Revert to nonfatal, better error message, cleanup
Revert to issuing a nonfatal error (it makes no sense to make it a
fatal error, but it probably makes sense for it to be an error instead
of a warning, especially since a lot of prior versions would crash and
apparently noone noticed.) We might have to revisit this based on
user requirements, and/or provide a method for the user to detect an
existing environment variable (%ifenv?).
Issue a better error message, indicating the nature of the failure.
Simplify the code by just updating the string in "p".
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 16:59:46 +0000 (20:59 +0400)]
Update changes.src
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 16:54:48 +0000 (20:54 +0400)]
BR3028880: Make nonexistent environment variable being fatal error
Frank suggested to just print out an error if environment
variable is not there. Agreed.
Suggested-by: Frank Kotler <fbkotler@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 16:16:50 +0000 (20:16 +0400)]
Check in test for BR3028880
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 15:08:15 +0000 (19:08 +0400)]
BR3028880: Fix NULL dereference on nonexistent environment variable
Frank reported we hit NULL dereference on nonexistent
environment variables. Fix it by leaving empty string
in text field of such token and yielding warning.
Reported-by: Frank Kotler <fbkotler@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 09:32:19 +0000 (13:32 +0400)]
tokenize: Fix wrong string index in indirect strings
At moment of calling the nasm_skip_string the string pointer
is already incremented which makes tokenize fail on correct
indirect strings.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 07:27:41 +0000 (11:27 +0400)]
preproc: Extract reading line from predefined macros from read_line
It makes read_line less complex
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 12 Jul 2010 19:15:01 +0000 (23:15 +0400)]
Documentation updates
Various small fixes.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Mon, 12 Jul 2010 11:19:17 +0000 (15:19 +0400)]
No need for \n at __OUTPUT_FORMAT__ macro end
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 13 Jul 2010 17:16:05 +0000 (21:16 +0400)]
Merge branches 'master' and 'preproc-paste-fix'
Cyrill Gorcunov [Tue, 13 Jul 2010 16:23:34 +0000 (20:23 +0400)]
BR3026808: Assign to local preprocessor variable does not work in 2.09
The commits
20a94ad7fe41c82f77fb670abb68f0de40d2b3e5
29c96651de1c43e59b7db58a4f06ff21dc854125
13dbfad76b4d3dbf27ef41761873584c6bd9fd7f
6f5f7ef417c37c154d10c2b3813808ad3fa65fd7
ddd08c3cccb4b68ecdb24d7a92eab6b2b82e8c68
seems to do the tricks we need. Eventually
get rid of commented "case".
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Sat, 10 Jul 2010 02:04:59 +0000 (19:04 -0700)]
nasmdoc: remove obsolete Sourceforge references
Remove obsolete Sourceforge references, replace with nasm.us.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Sat, 10 Jul 2010 02:02:28 +0000 (19:02 -0700)]
NASM 2.09rc3
Cyrill Gorcunov [Fri, 9 Jul 2010 22:37:49 +0000 (02:37 +0400)]
NASM 2.08.02
Cyrill Gorcunov [Fri, 9 Jul 2010 22:11:41 +0000 (02:11 +0400)]
preproc.c: Fix NULL deref on token pasting
In case if there is a whitespace before
'paste' token we may reach NULL dereference
in strlen since paste_head will point to
TOK_WHITESPACE. Fix it.
[mainstream
adabc1576b957fc7d929d2c7e749b4f45f7293aa]
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Fri, 9 Jul 2010 22:11:41 +0000 (02:11 +0400)]
preproc.c: Fix NULL deref on token pasting
In case if there is a whitespace before
'paste' token we may reach NULL dereference
in strlen since paste_head will point to
TOK_WHITESPACE. Fix it.
[test: paste.asm]
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Fri, 9 Jul 2010 22:05:32 +0000 (15:05 -0700)]
preproc: add another test case
Add another test case for preprocessor token pasting.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Thu, 8 Jul 2010 18:52:57 +0000 (11:52 -0700)]
preproc: don't paste TOK_PREPROC_ID
Trying to deal with bug reports 3005117 and 3026808: don't paste after
TOK_PREPROC_ID.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Thu, 8 Jul 2010 15:35:32 +0000 (08:35 -0700)]
Revert "BR3005117: Expland local single macro before pasting tokens"
This reverts commit
ec88c1beac003bd6660037da3cef3aebeee7af20.
Revert due to BR 3026808.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 15:34:41 +0000 (08:34 -0700)]
Revert "expand_mmac_params: Don't forget to handle TOK_OTHER"
This reverts commit
51fd86e0fed8f5162f8c1e45e4ceaf237d6e5539.
Revert due to BR 3026808.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 15:33:47 +0000 (08:33 -0700)]
Revert "expand_mmac_params: Expand local single macros unconditionally"
This reverts commit
1f6741fc78413236816c42d26b34134ab18ba4f1.
Revert due to BR 3026808.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 15:33:14 +0000 (08:33 -0700)]
Revert "expand_mmac_params: Expand local single macros unconditionally"
This reverts commit
985d880c15a5b26e59cdcec4af2eba0748ecfe1f.
Revert due to BR 3026808.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 15:30:23 +0000 (08:30 -0700)]
Check in test case from bug report br3005117
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 14:31:45 +0000 (07:31 -0700)]
br3026808: add test case
Add test case for BR 3026808 (%assign %$local).
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 00:20:19 +0000 (17:20 -0700)]
insns: add FXSAVE64/FXRSTOR64, drop np prefix
Add FXSAVE64 and FXRSTOR64; drop the np prefix on 0F AE instructions:
none of the rest of the 0F AE instructions have them, and there are no
conflicts.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Thu, 8 Jul 2010 00:17:30 +0000 (17:17 -0700)]
Merge branch 'master' of ssh://repo.or.cz/srv/git/nasm
H. Peter Anvin [Wed, 7 Jul 2010 20:48:14 +0000 (13:48 -0700)]
insns.dat: add XSAVE/XRSTOR64, XSAVEOPT, VCVTPH2PS/VCVTPS2PH
Add XSAVE64/XRSTOR64 (previously missing), XSAVEOPT/XSAVEOPT64 (per
AVX 007 spec), and VCVTPH2PS/VCVTPS2PH (per AVX 007) spec.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Wed, 7 Jul 2010 02:38:35 +0000 (19:38 -0700)]
insns.dat: remove VCVTPH2PS/VCVTPS2PH as AMD instructions
Remove VCVTPH2PS/VCVTPS2PH as AMD instructions based on version 3.04
of the AMD spec.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Tue, 6 Jul 2010 22:06:51 +0000 (15:06 -0700)]
NASM 2.09rc2
Cyrill Gorcunov [Tue, 6 Jul 2010 21:48:33 +0000 (01:48 +0400)]
doc: Update changes.src
Before soon-coming 2.09rc2.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Tue, 6 Jul 2010 16:23:24 +0000 (09:23 -0700)]
insns.dat: fix CPU flags for new instructions
FUTURE is a CPU level flag, and cannot be combined with X64 (which is
shorthand for X86_64,LONG). Also, make sure we add LONG annotations
to everything that is 64-bit mode only.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
H. Peter Anvin [Tue, 6 Jul 2010 16:15:48 +0000 (09:15 -0700)]
Add RD*SBASE, WR*SBASE, RDRAND from AVX v7
Add the RD*SBASE, WR*SBASE and RDRAND instructions from version 7 of
the AVX specification, Intel document 319433-007.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cyrill Gorcunov [Tue, 6 Jul 2010 15:40:02 +0000 (19:40 +0400)]
outbin: Cleanup bin_cleanup
No need for 'q' variable.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 6 Jul 2010 15:08:28 +0000 (19:08 +0400)]
outbin: Simplify reverse address computing
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 6 Jul 2010 14:33:29 +0000 (18:33 +0400)]
BR3025702: outbin -- Fix byte ordering
In commit
55ae12052cd1 we occasionally broke byte ordering. Fix it.
Note that current stable version 2.08.01 is not affected by this bug.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Fri, 25 Jun 2010 20:38:49 +0000 (13:38 -0700)]
doc: inslist.src is an intermediate file, move to "make clean"
inslist.src is an intermediate file, it should be in "make clean"
rather than "make spotless"; otherwise it ends up in the xdoc tarball.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cyrill Gorcunov [Fri, 25 Jun 2010 11:45:19 +0000 (15:45 +0400)]
BR 3020760: insns.dat -- confirm push imm32 on x86-64 explicitly
PUSH imm64 confuses ones who is trying to find this instruction in
processor programming manuals.
Actually it was introduced in a sake of "push `size' imm" consistency.
In other words -- to allow users to state "PUSH qword imm32" in 64bit code,
though on byte level (ie generated) code it still has a correct and valid
sign-extended "PUSH imm32" instruction.
To get rid of this ambiguie bite we make explicit "PUSH imm32"
being valid in 64bit code. This also makes "PUSH dword imm32"
valid in 64bit code as well.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Sat, 19 Jun 2010 21:52:19 +0000 (01:52 +0400)]
preproc.c: Get rid of signed/unsigned comparison warning
We need mac->nparam being explicictly int'fied otherwise
compiler issue a warning. Note that we might have been
using unsigned int but it would break an ability to pass
negative indices.
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Fri, 18 Jun 2010 23:24:46 +0000 (16:24 -0700)]
BR 3018233: handle LFS, LGS and LSS with a 64-bit register
LFS, LGS and LSS are valid in 64-bit mode and have 64-bit (REX.W)
forms.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Tue, 15 Jun 2010 17:49:09 +0000 (10:49 -0700)]
openwcom.mak: drop /f from delete commands
Classic DOS doesn't have del /f, and we don't really need it.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Tue, 15 Jun 2010 17:47:16 +0000 (10:47 -0700)]
Rename filenames to 8.3 format
Apparently some people still care about compiling native on MS-DOS,
and we don't have a significant number of files which need adjustment.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cyrill Gorcunov [Tue, 8 Jun 2010 14:45:06 +0000 (18:45 +0400)]
expand_mmac_params: Expand local single macros unconditionally
Peter proposed to expand local single macros unconditionally.
This should not hurt but give us more cleaner code in result.
Reported-by: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
Cyrill Gorcunov [Tue, 8 Jun 2010 14:39:56 +0000 (18:39 +0400)]
expand_mmac_params: Expand local single macros unconditionally
Peter proposed to expand local single macros unconditionally.
This should not hurt but give us more cleaner code in result.
Reported-by: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
H. Peter Anvin [Mon, 7 Jun 2010 18:34:28 +0000 (11:34 -0700)]
outobj: handle compilers without 64-bit switch() support
OpenWatcom, in particular, doesn't handle switch() statements with
64-bit expressions, sigh.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
H. Peter Anvin [Mon, 7 Jun 2010 18:26:11 +0000 (11:26 -0700)]
outelf32: handle compilers without 64-bit switch() support
OpenWatcom, in particular, doesn't handle switch() statements with
64-bit expressions, sigh.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>