RepoMirrors
/
ffmpeg
mirror of https://git.ffmpeg.org/ffmpeg.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

slightly faster scalarproduct_and_madd_int16_ssse3 on penryn, no change on conroe 

Originally committed as revision 20743 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
Loren Merritt 2009-12-05 17:53:11 +00:00
parent 91e644ff77
commit a4605efdf5
1 changed files with 13 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
libavcodec/x86/dsputil_yasm.asm
View File

@ -202,12 +202,20 @@ align 16
mova m2, [v3q + orderq]
mova m3, [v3q + orderq + mmsize]
%endif
pmaddwd m0, [v1q + orderq]
pmaddwd m1, [v1q + orderq + mmsize]
%define t0 [v1q + orderq]
%define t1 [v1q + orderq + mmsize]
%ifdef ARCH_X86_64
mova m8, t0
mova m9, t1
%define t0 m8
%define t1 m9
%endif
pmaddwd m0, t0
pmaddwd m1, t1
pmullw m2, m7
pmullw m3, m7
paddw m2, [v1q + orderq]
paddw m3, [v1q + orderq + mmsize]
paddw m2, t0
paddw m3, t1
paddd m6, m0
paddd m6, m1
mova [v1q + orderq], m2
@ -219,7 +227,7 @@ align 16
%endmacro
; int scalarproduct_and_madd_int16(int16_t *v1, int16_t *v2, int16_t *v3, int order, int mul)
cglobal scalarproduct_and_madd_int16_ssse3, 4,5,8, v1, v2, v3, order, mul
cglobal scalarproduct_and_madd_int16_ssse3, 4,5,10, v1, v2, v3, order, mul
shl orderq, 1
movd m7, mulm
pshuflw m7, m7, 0