diff options
| author | James Almer <jamrial@gmail.com> | 2014-03-17 17:07:34 -0300 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2014-03-18 15:00:50 +0100 |
| commit | 7c8bf09eddf4fe738eb05bec68479024a66bbde9 (patch) | |
| tree | 593c8e43f2d4045af13976ec7bd206a69a9f4195 /libswresample/resample_template.c | |
| parent | c56d25c4764fee4b7b0e94212ff279f34c8ba0c8 (diff) | |
| download | ffmpeg-7c8bf09eddf4fe738eb05bec68479024a66bbde9.tar.gz | |
swresample: change COMMON_CORE_INT16 asm from SSSE3 to SSE2
pshuf+paddd is slightly faster than phaddd.
The real gain is in pre-ssse3 processors like AMD K8 and K10, which get
a big boost in performance compared to the mmxext version
Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'libswresample/resample_template.c')
| -rw-r--r-- | libswresample/resample_template.c | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/libswresample/resample_template.c b/libswresample/resample_template.c index 5bc12bcb71..bdb038a56f 100644 --- a/libswresample/resample_template.c +++ b/libswresample/resample_template.c @@ -57,7 +57,7 @@ #elif defined(TEMPLATE_RESAMPLE_S16) \ || defined(TEMPLATE_RESAMPLE_S16_MMX2) \ - || defined(TEMPLATE_RESAMPLE_S16_SSSE3) + || defined(TEMPLATE_RESAMPLE_S16_SSE2) # define FILTER_SHIFT 15 # define DELEM int16_t @@ -74,9 +74,9 @@ # elif defined(TEMPLATE_RESAMPLE_S16_MMX2) # define COMMON_CORE COMMON_CORE_INT16_MMX2 # define RENAME(N) N ## _int16_mmx2 -# elif defined(TEMPLATE_RESAMPLE_S16_SSSE3) -# define COMMON_CORE COMMON_CORE_INT16_SSSE3 -# define RENAME(N) N ## _int16_ssse3 +# elif defined(TEMPLATE_RESAMPLE_S16_SSE2) +# define COMMON_CORE COMMON_CORE_INT16_SSE2 +# define RENAME(N) N ## _int16_sse2 # endif #endif |