2 Commits

Author	SHA1	Message	Date
Lynne	bbe95f7353	x86: replace explicit REP_RETs with RETs ... From x86inc: > On AMD cpus <=K10, an ordinary ret is slow if it immediately follows either > a branch or a branch target. So switch to a 2-byte form of ret in that case. > We can automatically detect "follows a branch", but not a branch target. > (SSSE3 is a sufficient condition to know that your cpu doesn't have this problem.) x86inc can automatically determine whether to use REP_RET rather than REP in most of these cases, so impact is minimal. Additionally, a few REP_RETs were used unnecessary, despite the return being nowhere near a branch. The only CPUs affected were AMD K10s, made between 2007 and 2011, 16 years ago and 12 years ago, respectively. In the future, everyone involved with x86inc should consider dropping REP_RETs altogether.	2023-02-01 04:23:55 +01:00
Marton Balint	4d95c6d5d7	avfilter/vf_framerate: add SIMD functions for frame blending ... Blend function speedups on x86_64 Core i5 4460: ffmpeg -f lavfi -i allyuv -vf framerate=60:threads=1 -f null none C: 447548411 decicycles in Blend, 2048 runs, 0 skips SSSE3: 130020087 decicycles in Blend, 2048 runs, 0 skips AVX2: 128508221 decicycles in Blend, 2048 runs, 0 skips ffmpeg -f lavfi -i allyuv -vf format=yuv420p12,framerate=60:threads=1 -f null none C: 228932745 decicycles in Blend, 2048 runs, 0 skips SSE4: 123357781 decicycles in Blend, 2048 runs, 0 skips AVX2: 121215353 decicycles in Blend, 2048 runs, 0 skips Signed-off-by: Marton Balint <cus@passwd.hu>	2018-01-28 18:50:52 +01:00