aarch64: vp9itxfm: Optimize 16x16 and 32x32 idct dc by unrolling

Multimedia / Libav - Martin Storsjö [martin.st] - 10 February 2017 17:31 EST

This work is sponsored by, and copyright, Google.

Before: Cortex A53
vp9_inv_dct_dct_16x16_sub1_add_neon: 235.3
vp9_inv_dct_dct_32x32_sub1_add_neon: 555.1 After:
vp9_inv_dct_dct_16x16_sub1_add_neon: 180.2
vp9_inv_dct_dct_32x32_sub1_add_neon: 475.3

3fcf788 aarch64: vp9itxfm: Optimize 16x16 and 32x32 idct dc by unrolling
libavcodec/aarch64/vp9itxfm_neon.S | 54 +++++++++++++++++++++++++-------------
1 file changed, 36 insertions(+), 18 deletions(-)

Upstream: git.libav.org


  • Share