[PATCH v2 0/2] x86/h264_intrapred: AVX2 for 10-bit intra 16x16 prediction

From: Zane Hambly <[email protected]>

Two AVX2 kernels for 10-bit H.264 intra 16x16 prediction.

Patch 1 (dc, top/left/128 dc, vertical, horizontal) was first sent in
January and got no review. Patch 2 adds plane prediction, which had no
10-bit SIMD at all; it depends on patch 1 (shares the AVX2 dispatch block).
Original patch 1:
https://patchwork.ffmpeg.org/project/ffmpeg/patch/[email protected]/

checkasm passes for the series.

v2:
 - add Signed-off-by (missing in v1)
 - pred16x16_plane: declare 5 xmm regs instead of 8, avoiding a Win64
   xmm6/xmm7 save/restore (same reduction mkver is applying to the 8-bit
   plane functions).
 - add the plane benchmark to its commit message

Zane Hambly (2):
  avcodec/x86/h264_intrapred: add AVX2 for 10-bit pred16x16 functions
  avcodec/x86/h264_intrapred: add AVX2 for 10-bit pred16x16 plane

 libavcodec/x86/h264_intrapred_10bit.asm | 285 ++++++++++++++++++++++++
 libavcodec/x86/h264_intrapred_init.c    |  16 ++
 2 files changed, 301 insertions(+)

-- 
2.51.0.windows.2

_______________________________________________
ffmpeg-devel mailing list -- [email protected]
To unsubscribe send an email to [email protected]