RepoMirrors/ffmpeg
1
0
Fork 0
mirror of https://git.ffmpeg.org/ffmpeg.git synced 2025-03-03 02:58:02 +00:00
Code Issues Projects Releases Wiki Activity

vf_interlace: x86: improve asm performance

Browse Source 
4775 decicycles -> 3688 decicycles
This commit is contained in:
Michael Niedermayer 2014-11-15 04:07:08 +01:00 committed by Vittorio Giovara
parent 57ed5a64fe
commit ca5c3ff909
1 changed files with 19 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
libavfilter/x86/vf_interlace.asm
View File

@ -2,6 +2,7 @@
;* x86-optimized functions for interlace filter
;*
;* Copyright (C) 2014 Kieran Kunhya <kierank@obe.tv>
;* Copyright (c) 2014 Michael Niedermayer <michaelni@gmx.at>
;*
;* This file is part of Libav.
;*
@ -34,36 +35,27 @@ cglobal lowpass_line, 5, 5, 7
add r4, r1
neg r1
pxor m6, m6
pcmpeqb m6, m6
.loop
mova m0, [r2+r1]
punpcklbw m1, m0, m6
punpckhbw m0, m6
paddw m0, m0
paddw m1, m1
mova m0, [r3+r1]
mova m1, [r3+r1+mmsize]
pavgb m0, [r4+r1]
pavgb m1, [r4+r1+mmsize]
mova m2, [r2+r1]
mova m3, [r2+r1+mmsize]
pxor m0, m6
pxor m1, m6
pxor m2, m6, [r2+r1]
pxor m3, m6, [r2+r1+mmsize]
pavgb m0, m2
pavgb m1, m3
pxor m0, m6
pxor m1, m6
mova [r0+r1], m0
mova [r0+r1+mmsize], m1
mova m2, [r3+r1]
punpcklbw m3, m2, m6
punpckhbw m2, m6
mova m4, [r4+r1]
punpcklbw m5, m4, m6
punpckhbw m4, m6
paddw m1, m3
pavgw m1, m5
paddw m0, m2
pavgw m0, m4
psrlw m0, 1
psrlw m1, 1
packuswb m1, m0
mova [r0+r1], m1
add r1, mmsize
add r1, 2*mmsize
jl .loop
REP_RET
%endmacro