virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2025-04-14 00:58:38 +02:00
Code Issues Releases Activity

swscale/x86/range_convert: reduce amount of xmm regs clobbered in luma functions

Browse Source 
Signed-off-by: James Almer <jamrial@gmail.com>
This commit is contained in:
James Almer 2024-06-15 21:00:17 -03:00
parent f6859cade3
commit 8a4c9d6bd3
1 changed files with 10 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
libswscale/x86/range_convert.asm
View File

@ -52,21 +52,21 @@ SECTION .text
;----------------------------------------------------------------------------- ;-----------------------------------------------------------------------------
%macro LUMCONVERTRANGE 4 %macro LUMCONVERTRANGE 4
cglobal %1, 2, 2, 7, dst, width cglobal %1, 2, 2, 5, dst, width
shl widthd, 1 shl widthd, 1
VBROADCASTI128 m4, [%2] VBROADCASTI128 m2, [%2]
VBROADCASTI128 m5, [%3] VBROADCASTI128 m3, [%3]
pxor m6, m6 pxor m4, m4
add dstq, widthq add dstq, widthq
neg widthq neg widthq
.loop: .loop:
movu m0, [dstq+widthq] movu m0, [dstq+widthq]
punpckhwd m1, m0, m6 punpckhwd m1, m0, m4
punpcklwd m0, m6 punpcklwd m0, m4
pmaddwd m0, m4 pmaddwd m0, m2
pmaddwd m1, m4 pmaddwd m1, m2
paddd m0, m5 paddd m0, m3
paddd m1, m5 paddd m1, m3
psrad m0, %4 psrad m0, %4
psrad m1, %4 psrad m1, %4
packssdw m0, m1 packssdw m0, m1