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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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>
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.
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>
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>
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>
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>
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>
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>
