Commit 14268254 authored by David Conrad's avatar David Conrad

vp3: Split out motion vectors to their own array

1.5% faster overall decode on my penryn

Originally committed as revision 22504 to svn://svn.ffmpeg.org/ffmpeg/trunk
parent 99de2b59
...@@ -48,8 +48,6 @@ static av_cold int vp3_decode_end(AVCodecContext *avctx); ...@@ -48,8 +48,6 @@ static av_cold int vp3_decode_end(AVCodecContext *avctx);
typedef struct Vp3Fragment { typedef struct Vp3Fragment {
int16_t dc; int16_t dc;
uint8_t coding_method; uint8_t coding_method;
int8_t motion_x;
int8_t motion_y;
uint8_t qpi; uint8_t qpi;
} Vp3Fragment; } Vp3Fragment;
...@@ -166,6 +164,8 @@ typedef struct Vp3DecodeContext { ...@@ -166,6 +164,8 @@ typedef struct Vp3DecodeContext {
int fragment_start[3]; int fragment_start[3];
int data_offset[3]; int data_offset[3];
int8_t (*motion_val[2])[2];
ScanTable scantable; ScanTable scantable;
/* tables */ /* tables */
...@@ -624,7 +624,7 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb) ...@@ -624,7 +624,7 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb)
int prior_last_motion_y = 0; int prior_last_motion_y = 0;
int current_macroblock; int current_macroblock;
int current_fragment; int current_fragment;
Vp3Fragment *frag; int frag;
if (s->keyframe) if (s->keyframe)
return 0; return 0;
...@@ -731,20 +731,14 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb) ...@@ -731,20 +731,14 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb)
current_fragment = current_fragment =
BLOCK_Y*s->fragment_width[0] + BLOCK_X; BLOCK_Y*s->fragment_width[0] + BLOCK_X;
if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) { if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) {
s->all_fragments[current_fragment].motion_x = motion_x[k]; s->motion_val[0][current_fragment][0] = motion_x[k];
s->all_fragments[current_fragment].motion_y = motion_y[k]; s->motion_val[0][current_fragment][1] = motion_y[k];
} else { } else {
s->all_fragments[current_fragment].motion_x = motion_x[0]; s->motion_val[0][current_fragment][0] = motion_x[0];
s->all_fragments[current_fragment].motion_y = motion_y[0]; s->motion_val[0][current_fragment][1] = motion_y[0];
} }
} }
#define SET_CHROMA_MV(mx, my) \
frag[s->fragment_start[1]].motion_x = mx; \
frag[s->fragment_start[1]].motion_y = my; \
frag[s->fragment_start[2]].motion_x = mx; \
frag[s->fragment_start[2]].motion_y = my
if (s->chroma_y_shift) { if (s->chroma_y_shift) {
if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) { if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) {
motion_x[0] = RSHIFT(motion_x[0] + motion_x[1] + motion_x[2] + motion_x[3], 2); motion_x[0] = RSHIFT(motion_x[0] + motion_x[1] + motion_x[2] + motion_x[3], 2);
...@@ -752,8 +746,9 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb) ...@@ -752,8 +746,9 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb)
} }
motion_x[0] = (motion_x[0]>>1) | (motion_x[0]&1); motion_x[0] = (motion_x[0]>>1) | (motion_x[0]&1);
motion_y[0] = (motion_y[0]>>1) | (motion_y[0]&1); motion_y[0] = (motion_y[0]>>1) | (motion_y[0]&1);
frag = s->all_fragments + mb_y*s->fragment_width[1] + mb_x; frag = mb_y*s->fragment_width[1] + mb_x;
SET_CHROMA_MV(motion_x[0], motion_y[0]); s->motion_val[1][frag][0] = motion_x[0];
s->motion_val[1][frag][1] = motion_y[0];
} else if (s->chroma_x_shift) { } else if (s->chroma_x_shift) {
if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) { if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) {
motion_x[0] = RSHIFT(motion_x[0] + motion_x[1], 1); motion_x[0] = RSHIFT(motion_x[0] + motion_x[1], 1);
...@@ -767,18 +762,21 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb) ...@@ -767,18 +762,21 @@ static int unpack_vectors(Vp3DecodeContext *s, GetBitContext *gb)
motion_x[0] = (motion_x[0]>>1) | (motion_x[0]&1); motion_x[0] = (motion_x[0]>>1) | (motion_x[0]&1);
motion_x[1] = (motion_x[1]>>1) | (motion_x[1]&1); motion_x[1] = (motion_x[1]>>1) | (motion_x[1]&1);
frag = s->all_fragments + 2*mb_y*s->fragment_width[1] + mb_x; frag = 2*mb_y*s->fragment_width[1] + mb_x;
for (k = 0; k < 2; k++) { for (k = 0; k < 2; k++) {
SET_CHROMA_MV(motion_x[k], motion_y[k]); s->motion_val[1][frag][0] = motion_x[k];
s->motion_val[1][frag][1] = motion_y[k];
frag += s->fragment_width[1]; frag += s->fragment_width[1];
} }
} else { } else {
for (k = 0; k < 4; k++) { for (k = 0; k < 4; k++) {
frag = s->all_fragments + BLOCK_Y*s->fragment_width[1] + BLOCK_X; frag = BLOCK_Y*s->fragment_width[1] + BLOCK_X;
if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) { if (s->macroblock_coding[current_macroblock] == MODE_INTER_FOURMV) {
SET_CHROMA_MV(motion_x[k], motion_y[k]); s->motion_val[1][frag][0] = motion_x[k];
s->motion_val[1][frag][1] = motion_y[k];
} else { } else {
SET_CHROMA_MV(motion_x[0], motion_y[0]); s->motion_val[1][frag][0] = motion_x[0];
s->motion_val[1][frag][1] = motion_y[0];
} }
} }
} }
...@@ -1354,6 +1352,7 @@ static void render_slice(Vp3DecodeContext *s, int slice) ...@@ -1354,6 +1352,7 @@ static void render_slice(Vp3DecodeContext *s, int slice)
int stride = s->current_frame.linesize[plane]; int stride = s->current_frame.linesize[plane];
int plane_width = s->width >> (plane && s->chroma_x_shift); int plane_width = s->width >> (plane && s->chroma_x_shift);
int plane_height = s->height >> (plane && s->chroma_y_shift); int plane_height = s->height >> (plane && s->chroma_y_shift);
int8_t (*motion_val)[2] = s->motion_val[!!plane];
int sb_x, sb_y = slice << (!plane && s->chroma_y_shift); int sb_x, sb_y = slice << (!plane && s->chroma_y_shift);
int slice_height = sb_y + 1 + (!plane && s->chroma_y_shift); int slice_height = sb_y + 1 + (!plane && s->chroma_y_shift);
...@@ -1408,8 +1407,8 @@ static void render_slice(Vp3DecodeContext *s, int slice) ...@@ -1408,8 +1407,8 @@ static void render_slice(Vp3DecodeContext *s, int slice)
if ((s->all_fragments[i].coding_method > MODE_INTRA) && if ((s->all_fragments[i].coding_method > MODE_INTRA) &&
(s->all_fragments[i].coding_method != MODE_USING_GOLDEN)) { (s->all_fragments[i].coding_method != MODE_USING_GOLDEN)) {
int src_x, src_y; int src_x, src_y;
motion_x = s->all_fragments[i].motion_x; motion_x = motion_val[y*fragment_width + x][0];
motion_y = s->all_fragments[i].motion_y; motion_y = motion_val[y*fragment_width + x][1];
src_x= (motion_x>>1) + 8*x; src_x= (motion_x>>1) + 8*x;
src_y= (motion_y>>1) + 8*y; src_y= (motion_y>>1) + 8*y;
...@@ -1568,8 +1567,11 @@ static av_cold int vp3_decode_init(AVCodecContext *avctx) ...@@ -1568,8 +1567,11 @@ static av_cold int vp3_decode_init(AVCodecContext *avctx)
s->all_fragments = av_malloc(s->fragment_count * sizeof(Vp3Fragment)); s->all_fragments = av_malloc(s->fragment_count * sizeof(Vp3Fragment));
s->coded_fragment_list[0] = av_malloc(s->fragment_count * sizeof(int)); s->coded_fragment_list[0] = av_malloc(s->fragment_count * sizeof(int));
s->dct_tokens_base = av_malloc(64*s->fragment_count * sizeof(*s->dct_tokens_base)); s->dct_tokens_base = av_malloc(64*s->fragment_count * sizeof(*s->dct_tokens_base));
s->motion_val[0] = av_malloc(y_fragment_count * sizeof(*s->motion_val[0]));
s->motion_val[1] = av_malloc(c_fragment_count * sizeof(*s->motion_val[1]));
if (!s->superblock_coding || !s->all_fragments || !s->dct_tokens_base || if (!s->superblock_coding || !s->all_fragments || !s->dct_tokens_base ||
!s->coded_fragment_list[0]) { !s->coded_fragment_list[0] || !s->motion_val[0] || !s->motion_val[1]) {
vp3_decode_end(avctx); vp3_decode_end(avctx);
return -1; return -1;
} }
...@@ -1874,6 +1876,8 @@ static av_cold int vp3_decode_end(AVCodecContext *avctx) ...@@ -1874,6 +1876,8 @@ static av_cold int vp3_decode_end(AVCodecContext *avctx)
av_free(s->dct_tokens_base); av_free(s->dct_tokens_base);
av_free(s->superblock_fragments); av_free(s->superblock_fragments);
av_free(s->macroblock_coding); av_free(s->macroblock_coding);
av_free(s->motion_val[0]);
av_free(s->motion_val[1]);
for (i = 0; i < 16; i++) { for (i = 0; i < 16; i++) {
free_vlc(&s->dc_vlc[i]); free_vlc(&s->dc_vlc[i]);
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment