]> git.sesse.net Git - x264/commit
aarch64: Faster intra_predict_4x4_h
authorJanne Grunau <janne-x264@jannau.net>
Tue, 18 Aug 2015 08:25:10 +0000 (10:25 +0200)
committerHenrik Gramner <henrik@gramner.com>
Sun, 11 Oct 2015 16:44:54 +0000 (18:44 +0200)
commitb16268ac0826d78455d0d704ea0fc8b1edc6b6bf
treea95ce0f33868cb6f8a117ab4670af8a8af94cd6a
parentf2a6be92e5e42e8ef1daf74f63dbdbc4819d2070
aarch64: Faster intra_predict_4x4_h

Use multiplication with 0x01010101 for splats.

On a cortex-a53:
                     gcc 4.9.2   llvm 3.6   neon (before)   neon (after)
intra_predict_4x4_h: 162         147        160/155         139/135
common/aarch64/predict-a.S