Re: [PATCH] Add XOP support for upcoming AMD Orochi processor

[Quentin Neill <quentin dot neill dot gnu at gmail dot com>]

Hi,

We found another issue related to PR
http://sourceware.org/bugzilla/show_bug.cgi?id=10973 when decoding
operands with 4 byte displacements.   The attached patch fixes this
bug like this:

        gas/testsuite/
	* gas/testsuite/gas/i386/fma4.d: Add test cases.
	* gas/testsuite/gas/i386/fma4.s: Add test cases.
	* gas/testsuite/gas/i386/x86-64-fma4.d: Add test cases.
	* gas/testsuite/gas/i386/x86-64-fma4.s: Add test cases.

        opcodes/
	* opcodes/i386-dis.c (get_vex_imm8): Extend logic to
	apply in all cases, to avoid fetching ahead for the immediate
	bytes when OP_E_memory has already been called.  Fix indentation.

Tested with make -k check.
Ok to commit?

Thanks,
-- 
Quentin Neill


On Fri, Nov 20, 2009 at 10:34 PM, Sebastian Pop <sebpop@gmail.com> wrote:
> Hi,
>
> The attached patch fixes one more bug again in the decoding operands
> like this:
>
> 2009-11-17 ?Quentin Neill ?<quentin.neill@amd.com>
> ? ? ? ? ? ?Sebastian Pop ?<sebastian.pop@amd.com>
>
> ? ? ? ?gas/testsuite/
> ? ? ? ?* gas/i386/x86-64-fma4.d: Add new patterns.
> ? ? ? ?* gas/i386/x86-64-fma4.s: Same.
> ? ? ? ?* gas/i386/x86-64-xop.d: Adjusted.
>
> ? ? ? ?opcodes/
> ? ? ? ?* i386-dis.c (get_vex_imm8): Increase bytes_before_imm when
> ? ? ? ?decoding the second source operand from the immediate byte.
> ? ? ? ?(OP_EX_VexW): Pass an extra integer to identify the second
> ? ? ? ?and third source arguments.
>
> Tested with make -k check and on simnow.
> Ok to commit?
>
> Thanks,
> Sebastian Pop
> --
> AMD / Open Source Compiler Engineering / GNU Tools
>

Attachment: 7003_fma4_src3_fix.diff
Description: Binary data