diff options
author | Wu Jianhua <jianhua.wu@intel.com> | 2021-08-04 10:06:15 +0800 |
---|---|---|
committer | Paul B Mahol <onemda@gmail.com> | 2021-08-29 19:58:33 +0200 |
commit | 4041c1029b93162faacda9e3f3cd083d1fbca7ce (patch) | |
tree | 0c7693b822eb51ccbef214df0be4e91a56f734f5 /libavfilter/vf_gblur.c | |
parent | 0c54ab20c254bf26c33a5cceb83862d3a59b3db7 (diff) | |
download | ffmpeg-4041c1029b93162faacda9e3f3cd083d1fbca7ce.tar.gz |
libavfilter/x86/vf_gblur: add localbuf and ff_horiz_slice_avx2/512()
We introduced a ff_horiz_slice_avx2/512() implemented on a new algorithm.
In a nutshell, the new algorithm does three things, gathering data from
8/16 rows, blurring data, and scattering data back to the image buffer.
Here we used a customized transpose 8x8/16x16 to avoid the huge overhead
brought by gather and scatter instructions, which is dependent on the
temporary buffer called localbuf added newly.
Performance data:
ff_horiz_slice_avx2(old): 109.89
ff_horiz_slice_avx2(new): 666.67
ff_horiz_slice_avx512: 1000
Co-authored-by: Cheng Yanfei <yanfei.cheng@intel.com>
Co-authored-by: Jin Jun <jun.i.jin@intel.com>
Signed-off-by: Wu Jianhua <jianhua.wu@intel.com>
Diffstat (limited to 'libavfilter/vf_gblur.c')
-rw-r--r-- | libavfilter/vf_gblur.c | 11 |
1 files changed, 9 insertions, 2 deletions
diff --git a/libavfilter/vf_gblur.c b/libavfilter/vf_gblur.c index a2c410c07b..ad48eae672 100644 --- a/libavfilter/vf_gblur.c +++ b/libavfilter/vf_gblur.c @@ -64,7 +64,7 @@ static void postscale_c(float *buffer, int length, } static void horiz_slice_c(float *buffer, int width, int height, int steps, - float nu, float bscale) + float nu, float bscale, float *localbuf) { int step, x, y; float *ptr; @@ -97,9 +97,13 @@ static int filter_horizontally(AVFilterContext *ctx, void *arg, int jobnr, int n const int steps = s->steps; const float nu = s->nu; float *buffer = s->buffer; + float *localbuf = NULL; + + if (s->localbuf) + localbuf = s->localbuf + s->stride * width * slice_start; s->horiz_slice(buffer + width * slice_start, width, slice_end - slice_start, - steps, nu, boundaryscale); + steps, nu, boundaryscale, localbuf); emms_c(); return 0; } @@ -242,6 +246,7 @@ static int query_formats(AVFilterContext *ctx) void ff_gblur_init(GBlurContext *s) { + s->localbuf = NULL; s->horiz_slice = horiz_slice_c; s->verti_slice = verti_slice_c; s->postscale_slice = postscale_c; @@ -384,6 +389,8 @@ static av_cold void uninit(AVFilterContext *ctx) GBlurContext *s = ctx->priv; av_freep(&s->buffer); + if (s->localbuf) + av_free(s->localbuf); } static const AVFilterPad gblur_inputs[] = { |