diff options
author | Michael Niedermayer <michaelni@gmx.at> | 2013-01-20 13:57:10 +0100 |
---|---|---|
committer | Michael Niedermayer <michaelni@gmx.at> | 2013-01-20 13:57:10 +0100 |
commit | cf061a9c3b861a048dd5b67ded5265c6f53805e5 (patch) | |
tree | 7b939285a34abfe719f38e594afa105c41574014 /libavcodec/x86/vp3dsp.asm | |
parent | b3b456b2588f67a152f3035cbc80d349898534c1 (diff) | |
parent | aeaf268e52fc11c1f64914a319e0edddf1346d6a (diff) | |
download | ffmpeg-cf061a9c3b861a048dd5b67ded5265c6f53805e5.tar.gz |
Merge commit 'aeaf268e52fc11c1f64914a319e0edddf1346d6a'
* commit 'aeaf268e52fc11c1f64914a319e0edddf1346d6a':
vp3: integrate clear_blocks with idct of previous block.
mpegvideo: fix loop condition in draw_line()
dvdsubdec: parse the size from the extradata
Conflicts:
libavcodec/dvdsubdec.c
libavcodec/mpegvideo.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Diffstat (limited to 'libavcodec/x86/vp3dsp.asm')
-rw-r--r-- | libavcodec/x86/vp3dsp.asm | 27 |
1 files changed, 20 insertions, 7 deletions
diff --git a/libavcodec/x86/vp3dsp.asm b/libavcodec/x86/vp3dsp.asm index 3a5ed03c64..be5f21b2d8 100644 --- a/libavcodec/x86/vp3dsp.asm +++ b/libavcodec/x86/vp3dsp.asm @@ -562,6 +562,13 @@ cglobal vp3_idct_put, 3, 4, 9 %endif %assign %%i %%i+64 %endrep + + pxor m0, m0 +%assign %%offset 0 +%rep 128/mmsize + mova [r2+%%offset], m0 +%assign %%offset %%offset+mmsize +%endrep RET cglobal vp3_idct_add, 3, 4, 9 @@ -600,6 +607,11 @@ cglobal vp3_idct_add, 3, 4, 9 movhps [r0+r1], m0 %endif lea r0, [r0+r1*2] +%assign %%offset 0 +%rep 32/mmsize + mova [r2+%%offset], m4 +%assign %%offset %%offset+mmsize +%endrep add r2, 32 dec r3 jg .loop @@ -620,7 +632,7 @@ vp3_idct_funcs paddusb m2, m0 movq m4, [r0+r1*2] paddusb m3, m0 - movq m5, [r0+r3 ] + movq m5, [r0+r2 ] paddusb m4, m0 paddusb m5, m0 psubusb m2, m1 @@ -630,7 +642,7 @@ vp3_idct_funcs movq [r0+r1 ], m3 psubusb m5, m1 movq [r0+r1*2], m4 - movq [r0+r3 ], m5 + movq [r0+r2 ], m5 %endmacro INIT_MMX mmxext @@ -638,11 +650,12 @@ cglobal vp3_idct_dc_add, 3, 4 %if ARCH_X86_64 movsxd r1, r1d %endif - lea r3, [r1*3] - movsx r2, word [r2] - add r2, 15 - sar r2, 5 - movd m0, r2d + movsx r3, word [r2] + mov word [r2], 0 + lea r2, [r1*3] + add r3, 15 + sar r3, 5 + movd m0, r3d pshufw m0, m0, 0x0 pxor m1, m1 psubw m1, m0 |