Port pred8x8_top_dc_mmxext (H.264 intra prediction) from x264 to FFmpeg.

Original authors: Holger Lubitz <holger lubitz org>, Jason Garrett-Glaser
<darkshikari gmail com> (approves LGPL relicensing for this code) and Loren
Merritt <lorenm at u dot washington dot edu> (approves LGPL relicensing for
this code). Patch by Daniel Kang <daniel dot d dot kang at gmail com>, as
part of Google's GCI 2010.

Originally committed as revision 26132 to svn://svn.ffmpeg.org/ffmpeg/trunk
This commit is contained in:
Daniel Kang 2010-12-29 17:42:34 +00:00 committed by Ronald S. Bultje
parent 98928c83e0
commit 725a3f9dfb
2 changed files with 39 additions and 0 deletions

View File

@ -827,6 +827,41 @@ PRED8x8_H mmx
PRED8x8_H mmxext
PRED8x8_H ssse3
;-----------------------------------------------------------------------------
; void pred8x8_top_dc_mmxext(uint8_t *src, int stride)
;-----------------------------------------------------------------------------
%ifdef CONFIG_GPL
cglobal pred8x8_top_dc_mmxext, 2,5
sub r0, r1
movq mm0, [r0]
pxor mm1, mm1
pxor mm2, mm2
lea r2, [r0+r1*2]
punpckhbw mm1, mm0
punpcklbw mm0, mm2
psadbw mm1, mm2 ; s1
lea r3, [r2+r1*2]
psadbw mm0, mm2 ; s0
psrlw mm1, 1
psrlw mm0, 1
pavgw mm1, mm2
lea r4, [r3+r1*2]
pavgw mm0, mm2
pshufw mm1, mm1, 0
pshufw mm0, mm0, 0 ; dc0 (w)
packuswb mm0, mm1 ; dc0,dc1 (b)
movq [r0+r1*1], mm0
movq [r0+r1*2], mm0
lea r0, [r3+r1*2]
movq [r2+r1*1], mm0
movq [r2+r1*2], mm0
movq [r3+r1*1], mm0
movq [r3+r1*2], mm0
movq [r0+r1*1], mm0
movq [r0+r1*2], mm0
RET
%endif
;-----------------------------------------------------------------------------
; void pred8x8_dc_rv40(uint8_t *src, int stride)
;-----------------------------------------------------------------------------

View File

@ -44,6 +44,7 @@ void ff_pred16x16_plane_svq3_ssse3 (uint8_t *src, int stride);
void ff_pred16x16_tm_vp8_mmx (uint8_t *src, int stride);
void ff_pred16x16_tm_vp8_mmxext (uint8_t *src, int stride);
void ff_pred16x16_tm_vp8_sse2 (uint8_t *src, int stride);
void ff_pred8x8_top_dc_mmxext (uint8_t *src, int stride);
void ff_pred8x8_dc_rv40_mmxext (uint8_t *src, int stride);
void ff_pred8x8_vertical_mmx (uint8_t *src, int stride);
void ff_pred8x8_horizontal_mmx (uint8_t *src, int stride);
@ -98,6 +99,9 @@ void ff_h264_pred_init_x86(H264PredContext *h, int codec_id)
#if CONFIG_GPL
if (codec_id == CODEC_ID_VP8 || codec_id == CODEC_ID_H264)
h->pred4x4 [DIAG_DOWN_LEFT_PRED ] = ff_pred4x4_down_left_mmxext;
if (codec_id == CODEC_ID_SVQ3 || codec_id == CODEC_ID_H264) {
h->pred8x8 [TOP_DC_PRED8x8 ] = ff_pred8x8_top_dc_mmxext;
}
#endif
if (codec_id == CODEC_ID_VP8) {
h->pred16x16[PLANE_PRED8x8] = ff_pred16x16_tm_vp8_mmxext;