| Message ID | 20170222002745.5680-2-jdarnley@obe.tv |
|---|---|
| State | Accepted |
| Commit | 33de0fee2c33c492aae96f643ed7bbaa393043dc |
| Headers | show |
On 2017-02-22 01:27, James Darnley wrote: > --- > libavcodec/x86/h264_deblock.asm | 1 + > libavcodec/x86/h264dsp_init.c | 10 ++++++++++ > 2 files changed, 11 insertions(+) > > diff --git a/libavcodec/x86/h264_deblock.asm b/libavcodec/x86/h264_deblock.asm > index 32aa3d3..6702ae9 100644 > --- a/libavcodec/x86/h264_deblock.asm > +++ b/libavcodec/x86/h264_deblock.asm > @@ -1252,6 +1252,7 @@ RET > > %endmacro ; DEBLOCK_CHROMA_XMM > > +DEBLOCK_CHROMA_XMM sse2 > DEBLOCK_CHROMA_XMM avx > > ;----------------------------------------------------------------------------- > diff --git a/libavcodec/x86/h264dsp_init.c b/libavcodec/x86/h264dsp_init.c > index 51082e8..0643b37 100644 > --- a/libavcodec/x86/h264dsp_init.c > +++ b/libavcodec/x86/h264dsp_init.c > @@ -304,6 +304,16 @@ av_cold void ff_h264dsp_init_x86(H264DSPContext *c, const int bit_depth, > #if ARCH_X86_64 > c->h264_h_loop_filter_luma_mbaff = ff_deblock_h_luma_mbaff_8_sse2; > #endif > + > + c->h264_v_loop_filter_chroma = ff_deblock_v_chroma_8_sse2; > + c->h264_v_loop_filter_chroma_intra = ff_deblock_v_chroma_intra_8_sse2; > + if (chroma_format_idc <= 1) { > + c->h264_h_loop_filter_chroma = ff_deblock_h_chroma_8_sse2; > + c->h264_h_loop_filter_chroma_intra = ff_deblock_h_chroma_intra_8_sse2; > + } else { > + c->h264_h_loop_filter_chroma = ff_deblock_h_chroma422_8_sse2; > + c->h264_h_loop_filter_chroma_intra = ff_deblock_h_chroma422_intra_8_sse2; > + } > } > if (EXTERNAL_SSSE3(cpu_flags)) { > c->biweight_h264_pixels_tab[0] = ff_h264_biweight_16_ssse3; > Does anyone have any comments on the patch set? For example: should I merge this sse2 patch into the others?
On 2/27/17, James Darnley <jdarnley@obe.tv> wrote: > On 2017-02-22 01:27, James Darnley wrote: >> --- >> libavcodec/x86/h264_deblock.asm | 1 + >> libavcodec/x86/h264dsp_init.c | 10 ++++++++++ >> 2 files changed, 11 insertions(+) >> >> diff --git a/libavcodec/x86/h264_deblock.asm >> b/libavcodec/x86/h264_deblock.asm >> index 32aa3d3..6702ae9 100644 >> --- a/libavcodec/x86/h264_deblock.asm >> +++ b/libavcodec/x86/h264_deblock.asm >> @@ -1252,6 +1252,7 @@ RET >> >> %endmacro ; DEBLOCK_CHROMA_XMM >> >> +DEBLOCK_CHROMA_XMM sse2 >> DEBLOCK_CHROMA_XMM avx >> >> >> ;----------------------------------------------------------------------------- >> diff --git a/libavcodec/x86/h264dsp_init.c b/libavcodec/x86/h264dsp_init.c >> index 51082e8..0643b37 100644 >> --- a/libavcodec/x86/h264dsp_init.c >> +++ b/libavcodec/x86/h264dsp_init.c >> @@ -304,6 +304,16 @@ av_cold void ff_h264dsp_init_x86(H264DSPContext *c, >> const int bit_depth, >> #if ARCH_X86_64 >> c->h264_h_loop_filter_luma_mbaff = >> ff_deblock_h_luma_mbaff_8_sse2; >> #endif >> + >> + c->h264_v_loop_filter_chroma = >> ff_deblock_v_chroma_8_sse2; >> + c->h264_v_loop_filter_chroma_intra = >> ff_deblock_v_chroma_intra_8_sse2; >> + if (chroma_format_idc <= 1) { >> + c->h264_h_loop_filter_chroma = >> ff_deblock_h_chroma_8_sse2; >> + c->h264_h_loop_filter_chroma_intra = >> ff_deblock_h_chroma_intra_8_sse2; >> + } else { >> + c->h264_h_loop_filter_chroma = >> ff_deblock_h_chroma422_8_sse2; >> + c->h264_h_loop_filter_chroma_intra = >> ff_deblock_h_chroma422_intra_8_sse2; >> + } >> } >> if (EXTERNAL_SSSE3(cpu_flags)) { >> c->biweight_h264_pixels_tab[0] = ff_h264_biweight_16_ssse3; >> > > Does anyone have any comments on the patch set? For example: should I > merge this sse2 patch into the others? probably not, just commit.
On 2017-02-27 12:13, Paul B Mahol wrote: > On 2/27/17, James Darnley <jdarnley@obe.tv> wrote: >> >> Does anyone have any comments on the patch set? For example: should I >> merge this sse2 patch into the others? > > probably not, just commit. Will do. I have added to the sse2 commit message two short lines about performance increases.
2017-02-27 13:08 GMT+01:00 James Darnley <jdarnley@obe.tv>: > On 2017-02-27 12:13, Paul B Mahol wrote: >> On 2/27/17, James Darnley <jdarnley@obe.tv> wrote: >>> >>> Does anyone have any comments on the patch set? For example: should I >>> merge this sse2 patch into the others? >> >> probably not, just commit. > > Will do. I have added to the sse2 commit message two short lines about > performance increases. You created a new branch, I don't think this was intended. Carl Eugen
diff --git a/libavcodec/x86/h264_deblock.asm b/libavcodec/x86/h264_deblock.asm index 32aa3d3..6702ae9 100644 --- a/libavcodec/x86/h264_deblock.asm +++ b/libavcodec/x86/h264_deblock.asm @@ -1252,6 +1252,7 @@ RET %endmacro ; DEBLOCK_CHROMA_XMM +DEBLOCK_CHROMA_XMM sse2 DEBLOCK_CHROMA_XMM avx ;----------------------------------------------------------------------------- diff --git a/libavcodec/x86/h264dsp_init.c b/libavcodec/x86/h264dsp_init.c index 51082e8..0643b37 100644 --- a/libavcodec/x86/h264dsp_init.c +++ b/libavcodec/x86/h264dsp_init.c @@ -304,6 +304,16 @@ av_cold void ff_h264dsp_init_x86(H264DSPContext *c, const int bit_depth, #if ARCH_X86_64 c->h264_h_loop_filter_luma_mbaff = ff_deblock_h_luma_mbaff_8_sse2; #endif + + c->h264_v_loop_filter_chroma = ff_deblock_v_chroma_8_sse2; + c->h264_v_loop_filter_chroma_intra = ff_deblock_v_chroma_intra_8_sse2; + if (chroma_format_idc <= 1) { + c->h264_h_loop_filter_chroma = ff_deblock_h_chroma_8_sse2; + c->h264_h_loop_filter_chroma_intra = ff_deblock_h_chroma_intra_8_sse2; + } else { + c->h264_h_loop_filter_chroma = ff_deblock_h_chroma422_8_sse2; + c->h264_h_loop_filter_chroma_intra = ff_deblock_h_chroma422_intra_8_sse2; + } } if (EXTERNAL_SSSE3(cpu_flags)) { c->biweight_h264_pixels_tab[0] = ff_h264_biweight_16_ssse3;