some MMX optimizations for the CAVS decoder

Originally committed as revision 5846 to svn://svn.ffmpeg.org/ffmpeg/trunk

6 files changed, 551 insertions, 3 deletions

diff --git a/libavcodec/Makefile b/libavcodec/Makefile
index 76d996d161..1f6078458e 100644
--- a/libavcodec/Makefile
+++ b/libavcodec/Makefile
@@ -288,6 +288,9 @@ OBJS += i386/fdct_mmx.o i386/cputest.o \
 ifeq ($(CONFIG_GPL),yes)
 OBJS += i386/idct_mmx_xvid.o
 endif
+ifeq ($(CONFIG_CAVS_DECODER),yes)
+OBJS += i386/cavsdsp_mmx.o
+endif
 ifeq ($(TARGET_BUILTIN_VECTOR),yes)
 i386/fft_sse.o: CFLAGS+= -msse
 depend: CFLAGS+= -msse
diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index 1c0d1ea66b..6a0347e1c3 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -1194,6 +1194,7 @@ typedef struct AVCodecContext {
 #define FF_IDCT_VP3          12
 #define FF_IDCT_IPP          13
 #define FF_IDCT_XVIDMMX      14
+#define FF_IDCT_CAVS         15
 
     /**
      * slice count.
diff --git a/libavcodec/cavs.c b/libavcodec/cavs.c
index 220a731529..5200892685 100644
--- a/libavcodec/cavs.c
+++ b/libavcodec/cavs.c
@@ -78,6 +78,7 @@ typedef struct {
     int qp;
     int qp_fixed;
     int cbp;
+    ScanTable scantable;
 
     /** intra prediction is done with un-deblocked samples
      they are saved here before deblocking the MB  */
@@ -97,6 +98,7 @@ typedef struct {
     int scale_den[2];  ///< for scaling neighbouring MVs
 
     int got_keyframe;
+    DCTELEM *block;
 } AVSContext;
 
 /*****************************************************************************
@@ -649,10 +651,9 @@ static int decode_residual_block(AVSContext *h, GetBitContext *gb,
     int dqm = dequant_mul[qp];
     int dqs = dequant_shift[qp];
     int dqa = 1 << (dqs - 1);
-    const uint8_t *scantab = ff_zigzag_direct;
-    DCTELEM block[64];
+    const uint8_t *scantab = h->scantable.permutated;
+    DCTELEM *block = h->block;
 
-    memset(block,0,64*sizeof(DCTELEM));
     for(i=0;i<65;i++) {
         level_code = get_ue_code(gb,r->golomb_order);
         if(level_code >= ESCAPE_CODE) {
@@ -1135,8 +1136,10 @@ static int decode_pic(AVSContext *h) {
     enum mb_t mb_type;
 
     if (!s->context_initialized) {
+        s->avctx->idct_algo = FF_IDCT_CAVS;
         if (MPV_common_init(s) < 0)
             return -1;
+        ff_init_scantable(s->dsp.idct_permutation,&h->scantable,ff_zigzag_direct);
     }
     get_bits(&s->gb,16);//bbv_dwlay
     if(h->stc == PIC_PB_START_CODE) {
@@ -1281,6 +1284,7 @@ static void init_top_lines(AVSContext *h) {
     /* alloc space for co-located MVs and types */
     h->col_mv       = av_malloc( h->mb_width*h->mb_height*4*sizeof(vector_t));
     h->col_type_base = av_malloc(h->mb_width*h->mb_height);
+    h->block        = av_mallocz(64*sizeof(DCTELEM));
 }
 
 static int decode_seq_header(AVSContext *h) {
@@ -1478,6 +1482,7 @@ static int cavs_decode_end(AVCodecContext * avctx) {
     av_free(h->top_border_v);
     av_free(h->col_mv);
     av_free(h->col_type_base);
+    av_free(h->block);
     return 0;
 }
 
diff --git a/libavcodec/cavsdsp.c b/libavcodec/cavsdsp.c
index 27586b60dc..d2517d9e9d 100644
--- a/libavcodec/cavsdsp.c
+++ b/libavcodec/cavsdsp.c
@@ -246,6 +246,7 @@ static void cavs_idct8_add_c(uint8_t *dst, DCTELEM *block, int stride) {
         dst[i + 6*stride] = cm[ dst[i + 6*stride] + ((b1 - b5) >> 7)];
         dst[i + 7*stride] = cm[ dst[i + 7*stride] + ((b0 - b4) >> 7)];
     }
+    memset(block,0,64*sizeof(DCTELEM));
 }
 
 /*****************************************************************************
diff --git a/libavcodec/i386/cavsdsp_mmx.c b/libavcodec/i386/cavsdsp_mmx.c
new file mode 100644
index 0000000000..71cf72d62e
--- /dev/null
+++ b/libavcodec/i386/cavsdsp_mmx.c
@@ -0,0 +1,516 @@
+/*
+ * Chinese AVS video (AVS1-P2, JiZhun profile) decoder.
+ * Copyright (c) 2006  Stefan Gehrer <stefan.gehrer@gmx.de>
+ *
+ * MMX optimised DSP functions, based on H.264 optimisations by
+ * Michael Niedermayer and Loren Merritt
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+
+#include "../dsputil.h"
+#include "common.h"
+
+DECLARE_ALIGNED_8(static const uint64_t,ff_pw_4 ) = 0x0004000400040004ULL;
+DECLARE_ALIGNED_8(static const uint64_t,ff_pw_5 ) = 0x0005000500050005ULL;
+DECLARE_ALIGNED_8(static const uint64_t,ff_pw_7 ) = 0x0007000700070007ULL;
+DECLARE_ALIGNED_8(static const uint64_t,ff_pw_42) = 0x002A002A002A002AULL;
+DECLARE_ALIGNED_8(static const uint64_t,ff_pw_64) = 0x0040004000400040ULL;
+DECLARE_ALIGNED_8(static const uint64_t,ff_pw_96) = 0x0060006000600060ULL;
+
+/*****************************************************************************
+ *
+ * inverse transform
+ *
+ ****************************************************************************/
+
+#define SUMSUB_BA( a, b ) \
+    "paddw "#b", "#a" \n\t"\
+    "paddw "#b", "#b" \n\t"\
+    "psubw "#a", "#b" \n\t"
+
+#define SBUTTERFLY(a,b,t,n)\
+    "movq " #a ", " #t "              \n\t" /* abcd */\
+    "punpckl" #n " " #b ", " #a "     \n\t" /* aebf */\
+    "punpckh" #n " " #b ", " #t "     \n\t" /* cgdh */
+
+#define TRANSPOSE4(a,b,c,d,t)\
+    SBUTTERFLY(a,b,t,wd) /* a=aebf t=cgdh */\
+    SBUTTERFLY(c,d,b,wd) /* c=imjn b=kolp */\
+    SBUTTERFLY(a,c,d,dq) /* a=aeim d=bfjn */\
+    SBUTTERFLY(t,b,c,dq) /* t=cgko c=dhlp */
+
+static inline void cavs_idct8_1d(int16_t *block, uint64_t bias)
+{
+    asm volatile(
+        "movq 112(%0), %%mm4  \n\t" /* mm4 = src7 */
+        "movq  16(%0), %%mm5  \n\t" /* mm5 = src1 */
+        "movq  80(%0), %%mm2  \n\t" /* mm2 = src5 */
+        "movq  48(%0), %%mm7  \n\t" /* mm7 = src3 */
+        "movq   %%mm4, %%mm0  \n\t"
+        "movq   %%mm5, %%mm3  \n\t"
+        "movq   %%mm2, %%mm6  \n\t"
+        "movq   %%mm7, %%mm1  \n\t"
+
+        "paddw  %%mm4, %%mm4  \n\t" /* mm4 = 2*src7 */
+        "paddw  %%mm3, %%mm3  \n\t" /* mm3 = 2*src1 */
+        "paddw  %%mm6, %%mm6  \n\t" /* mm6 = 2*src5 */
+        "paddw  %%mm1, %%mm1  \n\t" /* mm1 = 2*src3 */
+        "paddw  %%mm4, %%mm0  \n\t" /* mm0 = 3*src7 */
+        "paddw  %%mm3, %%mm5  \n\t" /* mm5 = 3*src1 */
+        "paddw  %%mm6, %%mm2  \n\t" /* mm2 = 3*src5 */
+        "paddw  %%mm1, %%mm7  \n\t" /* mm7 = 3*src3 */
+        "psubw  %%mm4, %%mm5  \n\t" /* mm5 = 3*src1 - 2*src7 = a0 */
+        "paddw  %%mm6, %%mm7  \n\t" /* mm7 = 3*src3 + 2*src5 = a1 */
+        "psubw  %%mm2, %%mm1  \n\t" /* mm1 = 2*src3 - 3*src5 = a2 */
+        "paddw  %%mm0, %%mm3  \n\t" /* mm3 = 2*src1 + 3*src7 = a3 */
+
+        "movq   %%mm5, %%mm4  \n\t"
+        "movq   %%mm7, %%mm6  \n\t"
+        "movq   %%mm3, %%mm0  \n\t"
+        "movq   %%mm1, %%mm2  \n\t"
+        SUMSUB_BA( %%mm7, %%mm5 )   /* mm7 = a0 + a1  mm5 = a0 - a1 */
+        "paddw  %%mm3, %%mm7  \n\t" /* mm7 = a0 + a1 + a3 */
+        "paddw  %%mm1, %%mm5  \n\t" /* mm5 = a0 - a1 + a2 */
+        "paddw  %%mm7, %%mm7  \n\t"
+        "paddw  %%mm5, %%mm5  \n\t"
+        "paddw  %%mm6, %%mm7  \n\t" /* mm7 = b4 */
+        "paddw  %%mm4, %%mm5  \n\t" /* mm5 = b5 */
+
+        SUMSUB_BA( %%mm1, %%mm3 )   /* mm1 = a3 + a2  mm3 = a3 - a2 */
+        "psubw  %%mm1, %%mm4  \n\t" /* mm4 = a0 - a2 - a3 */
+        "movq   %%mm4, %%mm1  \n\t" /* mm1 = a0 - a2 - a3 */
+        "psubw  %%mm6, %%mm3  \n\t" /* mm3 = a3 - a2 - a1 */
+        "paddw  %%mm1, %%mm1  \n\t"
+        "paddw  %%mm3, %%mm3  \n\t"
+        "psubw  %%mm2, %%mm1  \n\t" /* mm1 = b7 */
+        "paddw  %%mm0, %%mm3  \n\t" /* mm3 = b6 */
+
+        "movq  32(%0), %%mm2  \n\t" /* mm2 = src2 */
+        "movq  96(%0), %%mm6  \n\t" /* mm6 = src6 */
+        "movq   %%mm2, %%mm4  \n\t"
+        "movq   %%mm6, %%mm0  \n\t"
+        "psllw  $2,    %%mm4  \n\t" /* mm4 = 4*src2 */
+        "psllw  $2,    %%mm6  \n\t" /* mm6 = 4*src6 */
+        "paddw  %%mm4, %%mm2  \n\t" /* mm2 = 5*src2 */
+        "paddw  %%mm6, %%mm0  \n\t" /* mm0 = 5*src6 */
+        "paddw  %%mm2, %%mm2  \n\t"
+        "paddw  %%mm0, %%mm0  \n\t"
+        "psubw  %%mm0, %%mm4  \n\t" /* mm4 = 4*src2 - 10*src6 = a7 */
+        "paddw  %%mm2, %%mm6  \n\t" /* mm6 = 4*src6 + 10*src2 = a6 */
+
+        "movq    (%0), %%mm2  \n\t" /* mm2 = src0 */
+        "movq  64(%0), %%mm0  \n\t" /* mm0 = src4 */
+        SUMSUB_BA( %%mm0, %%mm2 )   /* mm0 = src0+src4  mm2 = src0-src4 */
+        "psllw  $3,    %%mm0  \n\t"
+        "psllw  $3,    %%mm2  \n\t"
+        "paddw  %1,    %%mm0  \n\t" /* add rounding bias */
+        "paddw  %1,    %%mm2  \n\t" /* add rounding bias */
+
+        SUMSUB_BA( %%mm6, %%mm0 )   /* mm6 = a4 + a6  mm0 = a4 - a6 */
+        SUMSUB_BA( %%mm4, %%mm2 )   /* mm4 = a5 + a7  mm2 = a5 - a7 */
+        SUMSUB_BA( %%mm7, %%mm6 )   /* mm7 = dst0  mm6 = dst7 */
+        SUMSUB_BA( %%mm5, %%mm4 )   /* mm5 = dst1  mm4 = dst6 */
+        SUMSUB_BA( %%mm3, %%mm2 )   /* mm3 = dst2  mm2 = dst5 */
+        SUMSUB_BA( %%mm1, %%mm0 )   /* mm1 = dst3  mm0 = dst4 */
+        :: "r"(block), "m"(bias)
+    );
+}
+
+static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
+{
+    int i;
+    DECLARE_ALIGNED_8(int16_t, b2[64]);
+
+    for(i=0; i<2; i++){
+        DECLARE_ALIGNED_8(uint64_t, tmp);
+
+        cavs_idct8_1d(block+4*i, ff_pw_4);
+
+        asm volatile(
+            "psraw     $3, %%mm7  \n\t"
+            "psraw     $3, %%mm6  \n\t"
+            "psraw     $3, %%mm5  \n\t"
+            "psraw     $3, %%mm4  \n\t"
+            "psraw     $3, %%mm3  \n\t"
+            "psraw     $3, %%mm2  \n\t"
+            "psraw     $3, %%mm1  \n\t"
+            "psraw     $3, %%mm0  \n\t"
+            "movq   %%mm7,    %0   \n\t"
+            TRANSPOSE4( %%mm0, %%mm2, %%mm4, %%mm6, %%mm7 )
+            "movq   %%mm0,  8(%1)  \n\t"
+            "movq   %%mm6, 24(%1)  \n\t"
+            "movq   %%mm7, 40(%1)  \n\t"
+            "movq   %%mm4, 56(%1)  \n\t"
+            "movq    %0,    %%mm7  \n\t"
+            TRANSPOSE4( %%mm7, %%mm5, %%mm3, %%mm1, %%mm0 )
+            "movq   %%mm7,   (%1)  \n\t"
+            "movq   %%mm1, 16(%1)  \n\t"
+            "movq   %%mm0, 32(%1)  \n\t"
+            "movq   %%mm3, 48(%1)  \n\t"
+            : "=m"(tmp)
+            : "r"(b2+32*i)
+            : "memory"
+        );
+    }
+
+    for(i=0; i<2; i++){
+        cavs_idct8_1d(b2+4*i, ff_pw_64);
+
+        asm volatile(
+            "psraw     $7, %%mm7  \n\t"
+            "psraw     $7, %%mm6  \n\t"
+            "psraw     $7, %%mm5  \n\t"
+            "psraw     $7, %%mm4  \n\t"
+            "psraw     $7, %%mm3  \n\t"
+            "psraw     $7, %%mm2  \n\t"
+            "psraw     $7, %%mm1  \n\t"
+            "psraw     $7, %%mm0  \n\t"
+            "movq   %%mm7,    (%0)  \n\t"
+            "movq   %%mm5,  16(%0)  \n\t"
+            "movq   %%mm3,  32(%0)  \n\t"
+            "movq   %%mm1,  48(%0)  \n\t"
+            "movq   %%mm0,  64(%0)  \n\t"
+            "movq   %%mm2,  80(%0)  \n\t"
+            "movq   %%mm4,  96(%0)  \n\t"
+            "movq   %%mm6, 112(%0)  \n\t"
+            :: "r"(b2+4*i)
+            : "memory"
+        );
+    }
+
+    add_pixels_clamped_mmx(b2, dst, stride);
+
+    /* clear block */
+    asm volatile(
+            "pxor %%mm7, %%mm7   \n\t"
+            "movq %%mm7, (%0)    \n\t"
+            "movq %%mm7, 8(%0)   \n\t"
+            "movq %%mm7, 16(%0)  \n\t"
+            "movq %%mm7, 24(%0)  \n\t"
+            "movq %%mm7, 32(%0)  \n\t"
+            "movq %%mm7, 40(%0)  \n\t"
+            "movq %%mm7, 48(%0)  \n\t"
+            "movq %%mm7, 56(%0)  \n\t"
+            "movq %%mm7, 64(%0)  \n\t"
+            "movq %%mm7, 72(%0)  \n\t"
+            "movq %%mm7, 80(%0)  \n\t"
+            "movq %%mm7, 88(%0)  \n\t"
+            "movq %%mm7, 96(%0)  \n\t"
+            "movq %%mm7, 104(%0) \n\t"
+            "movq %%mm7, 112(%0) \n\t"
+            "movq %%mm7, 120(%0) \n\t"
+            :: "r" (block)
+    );
+}
+
+/*****************************************************************************
+ *
+ * motion compensation
+ *
+ ****************************************************************************/
+
+/* vertical filter [-1 -2 96 42 -7  0]  */
+#define QPEL_CAVSV1(A,B,C,D,E,F,OP)      \
+        "movd (%0), "#F"            \n\t"\
+        "movq "#C", %%mm6           \n\t"\
+        "pmullw %5, %%mm6           \n\t"\
+        "movq "#D", %%mm7           \n\t"\
+        "pmullw %6, %%mm7           \n\t"\
+        "psllw $3, "#E"             \n\t"\
+        "psubw "#E", %%mm6          \n\t"\
+        "psraw $3, "#E"             \n\t"\
+        "paddw %%mm7, %%mm6         \n\t"\
+        "paddw "#E", %%mm6          \n\t"\
+        "paddw "#B", "#B"           \n\t"\
+        "pxor %%mm7, %%mm7          \n\t"\
+        "add %2, %0                 \n\t"\
+        "punpcklbw %%mm7, "#F"      \n\t"\
+        "psubw "#B", %%mm6          \n\t"\
+        "psraw $1, "#B"             \n\t"\
+        "psubw "#A", %%mm6          \n\t"\
+        "paddw %4, %%mm6            \n\t"\
+        "psraw $7, %%mm6            \n\t"\
+        "packuswb %%mm6, %%mm6      \n\t"\
+        OP(%%mm6, (%1), A, d)            \
+        "add %3, %1                 \n\t"
+
+/* vertical filter [ 0 -1  5  5 -1  0]  */
+#define QPEL_CAVSV2(A,B,C,D,E,F,OP)      \
+        "movd (%0), "#F"            \n\t"\
+        "movq "#C", %%mm6           \n\t"\
+        "paddw "#D", %%mm6          \n\t"\
+        "pmullw %5, %%mm6           \n\t"\
+        "add %2, %0                 \n\t"\
+        "punpcklbw %%mm7, "#F"      \n\t"\
+        "psubw "#B", %%mm6          \n\t"\
+        "psubw "#E", %%mm6          \n\t"\
+        "paddw %4, %%mm6            \n\t"\
+        "psraw $3, %%mm6            \n\t"\
+        "packuswb %%mm6, %%mm6      \n\t"\
+        OP(%%mm6, (%1), A, d)            \
+        "add %3, %1                 \n\t"
+
+/* vertical filter [ 0 -7 42 96 -2 -1]  */
+#define QPEL_CAVSV3(A,B,C,D,E,F,OP)      \
+        "movd (%0), "#F"            \n\t"\
+        "movq "#C", %%mm6           \n\t"\
+        "pmullw %6, %%mm6           \n\t"\
+        "movq "#D", %%mm7           \n\t"\
+        "pmullw %5, %%mm7           \n\t"\
+        "psllw $3, "#B"             \n\t"\
+        "psubw "#B", %%mm6          \n\t"\
+        "psraw $3, "#B"             \n\t"\
+        "paddw %%mm7, %%mm6         \n\t"\
+        "paddw "#B", %%mm6          \n\t"\
+        "paddw "#E", "#E"           \n\t"\
+        "pxor %%mm7, %%mm7          \n\t"\
+        "add %2, %0                 \n\t"\
+        "punpcklbw %%mm7, "#F"      \n\t"\
+        "psubw "#E", %%mm6          \n\t"\
+        "psraw $1, "#E"             \n\t"\
+        "psubw "#F", %%mm6          \n\t"\
+        "paddw %4, %%mm6            \n\t"\
+        "psraw $7, %%mm6            \n\t"\
+        "packuswb %%mm6, %%mm6      \n\t"\
+        OP(%%mm6, (%1), A, d)            \
+        "add %3, %1                 \n\t"
+
+
+#define QPEL_CAVSVNUM(VOP,OP,ADD,MUL1,MUL2)\
+    int w= 2;\
+    src -= 2*srcStride;\
+    \
+    while(w--){\
+      asm volatile(\
+        "pxor %%mm7, %%mm7          \n\t"\
+        "movd (%0), %%mm0           \n\t"\
+        "add %2, %0                 \n\t"\
+        "movd (%0), %%mm1           \n\t"\
+        "add %2, %0                 \n\t"\
+        "movd (%0), %%mm2           \n\t"\
+        "add %2, %0                 \n\t"\
+        "movd (%0), %%mm3           \n\t"\
+        "add %2, %0                 \n\t"\
+        "movd (%0), %%mm4           \n\t"\
+        "add %2, %0                 \n\t"\
+        "punpcklbw %%mm7, %%mm0     \n\t"\
+        "punpcklbw %%mm7, %%mm1     \n\t"\
+        "punpcklbw %%mm7, %%mm2     \n\t"\
+        "punpcklbw %%mm7, %%mm3     \n\t"\
+        "punpcklbw %%mm7, %%mm4     \n\t"\
+        VOP(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\
+        VOP(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\
+        VOP(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\
+        VOP(%%mm3, %%mm4, %%mm5, %%mm0, %%mm1, %%mm2, OP)\
+        VOP(%%mm4, %%mm5, %%mm0, %%mm1, %%mm2, %%mm3, OP)\
+        VOP(%%mm5, %%mm0, %%mm1, %%mm2, %%mm3, %%mm4, OP)\
+        VOP(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\
+        VOP(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\
+        \
+        : "+a"(src), "+c"(dst)\
+        : "S"((long)srcStride), "D"((long)dstStride), "m"(ADD), "m"(MUL1), "m"(MUL2)\
+        : "memory"\
+     );\
+     if(h==16){\
+        asm volatile(\
+            VOP(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\
+            VOP(%%mm3, %%mm4, %%mm5, %%mm0, %%mm1, %%mm2, OP)\
+            VOP(%%mm4, %%mm5, %%mm0, %%mm1, %%mm2, %%mm3, OP)\
+            VOP(%%mm5, %%mm0, %%mm1, %%mm2, %%mm3, %%mm4, OP)\
+            VOP(%%mm0, %%mm1, %%mm2, %%mm3, %%mm4, %%mm5, OP)\
+            VOP(%%mm1, %%mm2, %%mm3, %%mm4, %%mm5, %%mm0, OP)\
+            VOP(%%mm2, %%mm3, %%mm4, %%mm5, %%mm0, %%mm1, OP)\
+            VOP(%%mm3, %%mm4, %%mm5, %%mm0, %%mm1, %%mm2, OP)\
+            \
+           : "+a"(src), "+c"(dst)\
+           : "S"((long)srcStride), "D"((long)dstStride), "m"(ADD),  "m"(MUL1), "m"(MUL2)\
+           : "memory"\
+        );\
+     }\
+     src += 4-(h+5)*srcStride;\
+     dst += 4-h*dstStride;\
+   }
+
+#define QPEL_CAVS(OPNAME, OP, MMX)\
+static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    int h=8;\
+    asm volatile(\
+        "pxor %%mm7, %%mm7          \n\t"\
+        "movq %5, %%mm6             \n\t"\
+        "1:                         \n\t"\
+        "movq    (%0), %%mm0        \n\t"\
+        "movq   1(%0), %%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"\
+        "paddw %%mm2, %%mm0         \n\t"\
+        "paddw %%mm3, %%mm1         \n\t"\
+        "pmullw %%mm6, %%mm0        \n\t"\
+        "pmullw %%mm6, %%mm1        \n\t"\
+        "movq   -1(%0), %%mm2       \n\t"\
+        "movq    2(%0), %%mm4       \n\t"\
+        "movq %%mm2, %%mm3          \n\t"\
+        "movq %%mm4, %%mm5          \n\t"\
+        "punpcklbw %%mm7, %%mm2     \n\t"\
+        "punpckhbw %%mm7, %%mm3     \n\t"\
+        "punpcklbw %%mm7, %%mm4     \n\t"\
+        "punpckhbw %%mm7, %%mm5     \n\t"\
+        "paddw %%mm4, %%mm2         \n\t"\
+        "paddw %%mm3, %%mm5         \n\t"\
+        "psubw %%mm2, %%mm0         \n\t"\
+        "psubw %%mm5, %%mm1         \n\t"\
+        "movq %6, %%mm5             \n\t"\
+        "paddw %%mm5, %%mm0         \n\t"\
+        "paddw %%mm5, %%mm1         \n\t"\
+        "psraw $3, %%mm0            \n\t"\
+        "psraw $3, %%mm1            \n\t"\
+        "packuswb %%mm1, %%mm0      \n\t"\
+        OP(%%mm0, (%1),%%mm5, q)         \
+        "add %3, %0                 \n\t"\
+        "add %4, %1                 \n\t"\
+        "decl %2                    \n\t"\
+        " jnz 1b                    \n\t"\
+        : "+a"(src), "+c"(dst), "+m"(h)\
+        : "d"((long)srcStride), "S"((long)dstStride), "m"(ff_pw_5), "m"(ff_pw_4)\
+        : "memory"\
+    );\
+}\
+\
+static inline void OPNAME ## cavs_qpel8or16_v1_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
+  QPEL_CAVSVNUM(QPEL_CAVSV1,OP,ff_pw_64,ff_pw_96,ff_pw_42)      \
+}\
+\
+static inline void OPNAME ## cavs_qpel8or16_v2_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
+  QPEL_CAVSVNUM(QPEL_CAVSV2,OP,ff_pw_4,ff_pw_5,ff_pw_5)         \
+}\
+\
+static inline void OPNAME ## cavs_qpel8or16_v3_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
+  QPEL_CAVSVNUM(QPEL_CAVSV3,OP,ff_pw_64,ff_pw_96,ff_pw_42)      \
+}\
+\
+static void OPNAME ## cavs_qpel8_v1_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst  , src  , dstStride, srcStride, 8);\
+}\
+static void OPNAME ## cavs_qpel16_v1_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst  , src  , dstStride, srcStride, 16);\
+    OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
+}\
+\
+static void OPNAME ## cavs_qpel8_v2_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst  , src  , dstStride, srcStride, 8);\
+}\
+static void OPNAME ## cavs_qpel16_v2_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst  , src  , dstStride, srcStride, 16);\
+    OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
+}\
+\
+static void OPNAME ## cavs_qpel8_v3_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst  , src  , dstStride, srcStride, 8);\
+}\
+static void OPNAME ## cavs_qpel16_v3_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst  , src  , dstStride, srcStride, 16);\
+    OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
+}\
+\
+static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
+    OPNAME ## cavs_qpel8_h_ ## MMX(dst  , src  , dstStride, srcStride);\
+    OPNAME ## cavs_qpel8_h_ ## MMX(dst+8, src+8, dstStride, srcStride);\
+    src += 8*srcStride;\
+    dst += 8*dstStride;\
+    OPNAME ## cavs_qpel8_h_ ## MMX(dst  , src  , dstStride, srcStride);\
+    OPNAME ## cavs_qpel8_h_ ## MMX(dst+8, src+8, dstStride, srcStride);\
+}\
+
+#define CAVS_MC(OPNAME, SIZE, MMX) \
+static void ff_ ## OPNAME ## cavs_qpel ## SIZE ## _mc20_ ## MMX(uint8_t *dst, uint8_t *src, int stride){\
+    OPNAME ## cavs_qpel ## SIZE ## _h_ ## MMX(dst, src, stride, stride);\
+}\
+\
+static void ff_ ## OPNAME ## cavs_qpel ## SIZE ## _mc01_ ## MMX(uint8_t *dst, uint8_t *src, int stride){\
+    OPNAME ## cavs_qpel ## SIZE ## _v1_ ## MMX(dst, src, stride, stride);\
+}\
+\
+static void ff_ ## OPNAME ## cavs_qpel ## SIZE ## _mc02_ ## MMX(uint8_t *dst, uint8_t *src, int stride){\
+    OPNAME ## cavs_qpel ## SIZE ## _v2_ ## MMX(dst, src, stride, stride);\
+}\
+\
+static void ff_ ## OPNAME ## cavs_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t *src, int stride){\
+    OPNAME ## cavs_qpel ## SIZE ## _v3_ ## MMX(dst, src, stride, stride);\
+}\
+
+#define PUT_OP(a,b,temp, size) "mov" #size " " #a ", " #b "    \n\t"
+#define AVG_3DNOW_OP(a,b,temp, size) \
+"mov" #size " " #b ", " #temp "   \n\t"\
+"pavgusb " #temp ", " #a "        \n\t"\
+"mov" #size " " #a ", " #b "      \n\t"
+#define AVG_MMX2_OP(a,b,temp, size) \
+"mov" #size " " #b ", " #temp "   \n\t"\
+"pavgb " #temp ", " #a "          \n\t"\
+"mov" #size " " #a ", " #b "      \n\t"
+
+QPEL_CAVS(put_,       PUT_OP, 3dnow)
+QPEL_CAVS(avg_, AVG_3DNOW_OP, 3dnow)
+QPEL_CAVS(put_,       PUT_OP, mmx2)
+QPEL_CAVS(avg_,  AVG_MMX2_OP, mmx2)
+
+CAVS_MC(put_, 8, 3dnow)
+CAVS_MC(put_, 16,3dnow)
+CAVS_MC(avg_, 8, 3dnow)
+CAVS_MC(avg_, 16,3dnow)
+CAVS_MC(put_, 8, mmx2)
+CAVS_MC(put_, 16,mmx2)
+CAVS_MC(avg_, 8, mmx2)
+CAVS_MC(avg_, 16,mmx2)
+
+void ff_put_cavs_qpel8_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride);
+void ff_avg_cavs_qpel8_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride);
+void ff_put_cavs_qpel16_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride);
+void ff_avg_cavs_qpel16_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride);
+
+void ff_cavsdsp_init_mmx2(DSPContext* c, AVCodecContext *avctx) {
+#define dspfunc(PFX, IDX, NUM) \
+    c->PFX ## _pixels_tab[IDX][ 0] = ff_ ## PFX ## NUM ## _mc00_mmx2; \
+    c->PFX ## _pixels_tab[IDX][ 2] = ff_ ## PFX ## NUM ## _mc20_mmx2; \
+    c->PFX ## _pixels_tab[IDX][ 4] = ff_ ## PFX ## NUM ## _mc01_mmx2; \
+    c->PFX ## _pixels_tab[IDX][ 8] = ff_ ## PFX ## NUM ## _mc02_mmx2; \
+    c->PFX ## _pixels_tab[IDX][12] = ff_ ## PFX ## NUM ## _mc03_mmx2; \
+
+    dspfunc(put_cavs_qpel, 0, 16);
+    dspfunc(put_cavs_qpel, 1, 8);
+    dspfunc(avg_cavs_qpel, 0, 16);
+    dspfunc(avg_cavs_qpel, 1, 8);
+#undef dspfunc
+    c->cavs_idct8_add = cavs_idct8_add_mmx;
+}
+
+void ff_cavsdsp_init_3dnow(DSPContext* c, AVCodecContext *avctx) {
+#define dspfunc(PFX, IDX, NUM) \
+    c->PFX ## _pixels_tab[IDX][ 0] = ff_ ## PFX ## NUM ## _mc00_mmx2; \
+    c->PFX ## _pixels_tab[IDX][ 2] = ff_ ## PFX ## NUM ## _mc20_3dnow; \
+    c->PFX ## _pixels_tab[IDX][ 4] = ff_ ## PFX ## NUM ## _mc01_3dnow; \
+    c->PFX ## _pixels_tab[IDX][ 8] = ff_ ## PFX ## NUM ## _mc02_3dnow; \
+    c->PFX ## _pixels_tab[IDX][12] = ff_ ## PFX ## NUM ## _mc03_3dnow; \
+
+    dspfunc(put_cavs_qpel, 0, 16);
+    dspfunc(put_cavs_qpel, 1, 8);
+    dspfunc(avg_cavs_qpel, 0, 16);
+    dspfunc(avg_cavs_qpel, 1, 8);
+#undef dspfunc
+    c->cavs_idct8_add = cavs_idct8_add_mmx;
+}
diff --git a/libavcodec/i386/dsputil_mmx.c b/libavcodec/i386/dsputil_mmx.c
index c8d3ee8501..ec6b2ad1a7 100644
--- a/libavcodec/i386/dsputil_mmx.c
+++ b/libavcodec/i386/dsputil_mmx.c
@@ -2622,6 +2622,22 @@ PREFETCH(prefetch_3dnow, prefetch)
 
 #include "h264dsp_mmx.c"
 
+/* AVS specific */
+void ff_cavsdsp_init_mmx2(DSPContext* c, AVCodecContext *avctx);
+
+void ff_put_cavs_qpel8_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride) {
+    put_pixels8_mmx(dst, src, stride, 8);
+}
+void ff_avg_cavs_qpel8_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride) {
+    avg_pixels8_mmx(dst, src, stride, 8);
+}
+void ff_put_cavs_qpel16_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride) {
+    put_pixels16_mmx(dst, src, stride, 16);
+}
+void ff_avg_cavs_qpel16_mc00_mmx2(uint8_t *dst, uint8_t *src, int stride) {
+    avg_pixels16_mmx(dst, src, stride, 16);
+}
+
 /* external functions, from idct_mmx.c */
 void ff_mmx_idct(DCTELEM *block);
 void ff_mmxext_idct(DCTELEM *block);
@@ -2779,6 +2795,8 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
                     c->idct_permutation_type= FF_PARTTRANS_IDCT_PERM;
                 }
 #endif
+            }else if(idct_algo==FF_IDCT_CAVS){
+                    c->idct_permutation_type= FF_TRANSPOSE_IDCT_PERM;
 #ifdef CONFIG_GPL
             }else if(idct_algo==FF_IDCT_XVIDMMX){
                 if(mm_flags & MM_MMXEXT){
@@ -3012,6 +3030,10 @@ void dsputil_init_mmx(DSPContext* c, AVCodecContext *avctx)
             c->biweight_h264_pixels_tab[6]= ff_h264_biweight_4x4_mmx2;
             c->biweight_h264_pixels_tab[7]= ff_h264_biweight_4x2_mmx2;
 
+#ifdef CONFIG_CAVS_DECODER
+            ff_cavsdsp_init_mmx2(c, avctx);
+#endif
+
 #ifdef CONFIG_ENCODERS
             c->sub_hfyu_median_prediction= sub_hfyu_median_prediction_mmx2;
 #endif //CONFIG_ENCODERS