[FFmpeg-devel] [PATCH] Fix SSE code to not use SSE2.

Paul B Mahol onemda at gmail.com
Wed Mar 7 17:17:07 CET 2012


On 3/7/12, Jason Garrett-Glaser <jason at x264.com> wrote:
> On Tue, Mar 6, 2012 at 11:27 PM, Reimar Doeffinger
> <Reimar.Doeffinger at gmx.de> wrote:
>> On 6 Mar 2012, at 22:49, Jason Garrett-Glaser <jason at x264.com> wrote:
>>> On Tue, Mar 6, 2012 at 1:11 PM, Reimar Doeffinger
>>> <Reimar.Doeffinger at gmx.de> wrote:
>>>> movq from SSE register _to_ memory is an SSE2 instruction.
>>>> Use the SSE movlps function instead that does the same thing.
>>>>
>>>> Signed-off-by: Reimar Doeffinger <Reimar.Doeffinger at gmx.de>
>>>> ---
>>>>  libavcodec/x86/sbrdsp.asm |    2 +-
>>>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>>>
>>>> diff --git a/libavcodec/x86/sbrdsp.asm b/libavcodec/x86/sbrdsp.asm
>>>> index c165c52..c3b559b 100644
>>>> --- a/libavcodec/x86/sbrdsp.asm
>>>> +++ b/libavcodec/x86/sbrdsp.asm
>>>> @@ -104,7 +104,7 @@ cglobal sbr_hf_g_filt, 5, 6, 5
>>>>     movq        m2, [r1]
>>>>     punpckldq   m0, m0
>>>
>>> These look pretty SSE2 to me, too.
>>
>> Unfortunately that depends on the specific opcode chosen, they all have
>> SSE equivalents after all.
>
> What are you talking about?  punpckldq does not have an "SSE equivalent".

Just make this SSE2 only and be done with it.


More information about the ffmpeg-devel mailing list