diff options
| author | Michael Niedermayer <michaelni@gmx.at> | 2002-04-17 04:32:12 +0000 |
|---|---|---|
| committer | Michael Niedermayer <michaelni@gmx.at> | 2002-04-17 04:32:12 +0000 |
| commit | 9dbcbd92e84b2cd033fa6688935ffaad4b67e64e (patch) | |
| tree | 4996c809abbc01abc148d63074e321dc3437e5a9 /libavcodec/i386 | |
| parent | a7c02e2589ee901e0640c8e19b898ad0213fb071 (diff) | |
| download | ffmpeg-9dbcbd92e84b2cd033fa6688935ffaad4b67e64e.tar.gz | |
fixed mpeg4 time stuff on encoding
mpeg4 b-frame enoding support
removed old, out-commented ratecontrol
reuse motion compensation code between encoding & decoding
prefix newly added global functions with ff_ to reduce namespace polution
b-frame ME (unfinished, but working)
added some comments to mpegvideo.h
do MC on encoding only once if possible
bugs? ;)
Originally committed as revision 403 to svn://svn.ffmpeg.org/ffmpeg/trunk
Diffstat (limited to 'libavcodec/i386')
| -rw-r--r-- | libavcodec/i386/dsputil_mmx.c | 30 |
1 files changed, 29 insertions, 1 deletions
diff --git a/libavcodec/i386/dsputil_mmx.c b/libavcodec/i386/dsputil_mmx.c index bf729d9638..da86eba480 100644 --- a/libavcodec/i386/dsputil_mmx.c +++ b/libavcodec/i386/dsputil_mmx.c @@ -44,7 +44,6 @@ int pix_abs8x8_x2_mmx2(UINT8 *blk1, UINT8 *blk2, int lx); int pix_abs8x8_y2_mmx2(UINT8 *blk1, UINT8 *blk2, int lx); int pix_abs8x8_xy2_mmx2(UINT8 *blk1, UINT8 *blk2, int lx); - /* external functions, from idct_mmx.c */ void ff_mmx_idct(DCTELEM *block); void ff_mmxext_idct(DCTELEM *block); @@ -136,6 +135,34 @@ static void get_pixels_mmx(DCTELEM *block, const UINT8 *pixels, int line_size) } } +static void diff_pixels_mmx(DCTELEM *block, const UINT8 *s1, const UINT8 *s2, int stride) +{ + asm volatile( + ".balign 16 \n\t" + "movl $-128, %%eax \n\t" + "1: \n\t" + "movq (%0), %%mm0 \n\t" + "movq (%1), %%mm2 \n\t" + "movq %%mm0, %%mm1 \n\t" + "movq %%mm2, %%mm3 \n\t" + "punpcklbw %%mm7, %%mm0 \n\t" + "punpckhbw %%mm7, %%mm1 \n\t" + "punpcklbw %%mm7, %%mm2 \n\t" + "punpckhbw %%mm7, %%mm3 \n\t" + "psubw %%mm2, %%mm0 \n\t" + "psubw %%mm3, %%mm1 \n\t" + "movq %%mm0, (%2, %%eax)\n\t" + "movq %%mm1, 8(%2, %%eax)\n\t" + "addl %3, %0 \n\t" + "addl %3, %1 \n\t" + "addl $16, %%eax \n\t" + "jnz 1b \n\t" + : "+r" (s1), "+r" (s2) + : "r" (block+64), "r" (stride) + : "%eax" + ); +} + static void put_pixels_clamped_mmx(const DCTELEM *block, UINT8 *pixels, int line_size) { const DCTELEM *p; @@ -1064,6 +1091,7 @@ void dsputil_init_mmx(void) if (mm_flags & MM_MMX) { get_pixels = get_pixels_mmx; + diff_pixels = diff_pixels_mmx; put_pixels_clamped = put_pixels_clamped_mmx; add_pixels_clamped = add_pixels_clamped_mmx; clear_blocks= clear_blocks_mmx; |