Files
FFmpeg/libavcodec
Andreas Rheinhardt 1a7979a2f8 avcodec/x86/h26x/h2656_inter: Simplify splatting coefficients
For pre-AVX2, vpbroadcastw is emulated via a load, followed
by two shuffles. Yet given that one always wants to splat
multiple pairs of coefficients which are adjacent in memory,
one can do better than that: Load all of them at once, perform
a punpcklwd with itself and use one pshufd per register.
In case one has to sign-extend the coefficients, too,
one can replace the punpcklwd with one pmovsxbw (instead of one
per register) and use pshufd directly afterwards.

This saved 4816B of .text here.

Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
2026-03-29 01:05:23 +01:00
..
2026-03-28 11:25:38 +01:00
2026-03-10 13:52:18 +01:00
2026-03-10 13:52:19 +01:00
2026-03-10 19:31:00 +01:00
2026-03-10 13:52:19 +01:00
2026-02-23 23:58:11 +01:00
2026-03-13 16:09:40 +00:00
2026-03-13 02:49:59 +01:00
2026-03-10 13:52:19 +01:00
2026-03-07 19:22:40 -03:00
2026-02-11 20:35:20 +00:00
2026-03-10 13:52:19 +01:00
2026-02-19 22:39:35 +00:00
2026-03-28 11:25:38 +01:00
2026-03-28 11:25:38 +01:00
2026-03-28 11:25:38 +01:00
2026-03-02 19:01:46 +01:00
2026-03-02 19:01:46 +01:00
2026-03-10 13:52:18 +01:00
2026-03-03 02:41:05 +01:00
2026-03-10 13:52:19 +01:00
2026-02-23 15:29:27 +00:00
2026-03-10 13:52:18 +01:00
2026-03-16 10:24:33 +00:00