diff options
| author | Alan Kelly <alankelly@google.com> | 2021-12-15 10:35:02 +0100 |
|---|---|---|
| committer | James Almer <jamrial@gmail.com> | 2021-12-15 20:04:59 -0300 |
| commit | f900a19fa94b1a55b660ec2e5c13419d59754bc0 (patch) | |
| tree | 7fedf4f6b884fd7172a666f1706653b950fc6a90 /libswscale/x86/swscale.c | |
| parent | e9ba40c5c9a49bc97d16d66c46ff993fa84a6c31 (diff) | |
| download | ffmpeg-f900a19fa94b1a55b660ec2e5c13419d59754bc0.tar.gz | |
libswscale: Adds ff_hscale8to15_4_avx2 and ff_hscale8to15_X4_avx2 for all filter sizes.
Fixes so that fate under 64 bit Windows passes.
These functions replace all ff_hscale8to15_*_ssse3 when avx2 is available.
Signed-off-by: James Almer <jamrial@gmail.com>
Diffstat (limited to 'libswscale/x86/swscale.c')
| -rw-r--r-- | libswscale/x86/swscale.c | 19 |
1 files changed, 19 insertions, 0 deletions
diff --git a/libswscale/x86/swscale.c b/libswscale/x86/swscale.c index 0848a31461..164b06d6ba 100644 --- a/libswscale/x86/swscale.c +++ b/libswscale/x86/swscale.c @@ -276,6 +276,9 @@ SCALE_FUNCS_SSE(sse2); SCALE_FUNCS_SSE(ssse3); SCALE_FUNCS_SSE(sse4); +SCALE_FUNC(4, 8, 15, avx2); +SCALE_FUNC(X4, 8, 15, avx2); + #define VSCALEX_FUNC(size, opt) \ void ff_yuv2planeX_ ## size ## _ ## opt(const int16_t *filter, int filterSize, \ const int16_t **src, uint8_t *dest, int dstW, \ @@ -568,6 +571,22 @@ switch(c->dstBpc){ \ } #if ARCH_X86_64 +#define ASSIGN_AVX2_SCALE_FUNC(hscalefn, filtersize) \ + switch (filtersize) { \ + case 4: hscalefn = ff_hscale8to15_4_avx2; break; \ + default: hscalefn = ff_hscale8to15_X4_avx2; break; \ + break; \ + } + + if (EXTERNAL_AVX2_FAST(cpu_flags)){ + if ((c->srcBpc == 8) && (c->dstBpc <= 14)){ + if(c->chrDstW % 16 == 0) + ASSIGN_AVX2_SCALE_FUNC(c->hcScale, c->hChrFilterSize); + if(c->dstW % 16 == 0) + ASSIGN_AVX2_SCALE_FUNC(c->hyScale, c->hLumFilterSize); + } + } + if (EXTERNAL_AVX2_FAST(cpu_flags)) { switch (c->dstFormat) { case AV_PIX_FMT_NV12: |