Skip to content

F
ffmpeg
Commit ca80f11e authored by Jason Garrett-Glaser's avatar Jason Garrett-Glaser
Browse files
Options

H.264: faster fill_decode_caches 

Aliasing avoidance and general cleanup.
parent 3b7ebeb4
Show whitespace changes
Inline Side-by-side
Showing with 96 additions and 95 deletions
libavcodec/h264.h
View file @ ca80f11e
......@@ -307,11 +307,6 @@ typedef struct H264Context{
#define LIST_NOT_USED -1 //FIXME rename?
#define PART_NOT_AVAILABLE -2
/**
* is 1 if the specific list MV&references are set to 0,0,-2.
*/
int mv_cache_clean[2];
/**
* number of neighbors (top and/or left) that used 8x8 dct
*/
......@@ -857,6 +852,8 @@ static void fill_decode_caches(H264Context *h, int mb_type){
int topleft_type, top_type, topright_type, left_type[2];
const uint8_t * left_block= h->left_block;
int i;
uint8_t *nnz;
uint8_t *nnz_cache;
topleft_xy = h->topleft_mb_xy ;
top_xy = h->top_mb_xy ;
......@@ -946,42 +943,45 @@ static void fill_decode_caches(H264Context *h, int mb_type){
5 L . .. . . . .
*/
//FIXME constraint_intra_pred & partitioning & nnz (let us hope this is just a typo in the spec)
nnz_cache = h->non_zero_count_cache;
if(top_type){
AV_COPY32(&h->non_zero_count_cache[4+8* 0], &h->non_zero_count[top_xy][4*3]);
nnz = h->non_zero_count[top_xy];
AV_COPY32(&nnz_cache[4+8* 0], &nnz[4*3]);
if(CHROMA444){
AV_COPY32(&h->non_zero_count_cache[4+8* 5], &h->non_zero_count[top_xy][4* 7]);
AV_COPY32(&h->non_zero_count_cache[4+8*10], &h->non_zero_count[top_xy][4*11]);
AV_COPY32(&nnz_cache[4+8* 5], &nnz[4* 7]);
AV_COPY32(&nnz_cache[4+8*10], &nnz[4*11]);
}else{
AV_COPY32(&h->non_zero_count_cache[4+8* 5], &h->non_zero_count[top_xy][4* 5]);
AV_COPY32(&h->non_zero_count_cache[4+8*10], &h->non_zero_count[top_xy][4* 9]);
AV_COPY32(&nnz_cache[4+8* 5], &nnz[4* 5]);
AV_COPY32(&nnz_cache[4+8*10], &nnz[4* 9]);
}
}else{
uint32_t top_empty = CABAC && !IS_INTRA(mb_type) ? 0 : 0x40404040;
AV_WN32A(&h->non_zero_count_cache[4+8* 0], top_empty);
AV_WN32A(&h->non_zero_count_cache[4+8* 5], top_empty);
AV_WN32A(&h->non_zero_count_cache[4+8*10], top_empty);
AV_WN32A(&nnz_cache[4+8* 0], top_empty);
AV_WN32A(&nnz_cache[4+8* 5], top_empty);
AV_WN32A(&nnz_cache[4+8*10], top_empty);
}
for (i=0; i<2; i++) {
if(left_type[i]){
h->non_zero_count_cache[3+8* 1 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[8+0+2*i]];
h->non_zero_count_cache[3+8* 2 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[8+1+2*i]];
nnz = h->non_zero_count[left_xy[i]];
nnz_cache[3+8* 1 + 2*8*i]= nnz[left_block[8+0+2*i]];
nnz_cache[3+8* 2 + 2*8*i]= nnz[left_block[8+1+2*i]];
if(CHROMA444){
h->non_zero_count_cache[3+8* 6 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[8+0+2*i]+4*4];
h->non_zero_count_cache[3+8* 7 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[8+1+2*i]+4*4];
h->non_zero_count_cache[3+8*11 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[8+0+2*i]+8*4];
h->non_zero_count_cache[3+8*12 + 2*8*i]= h->non_zero_count[left_xy[i]][left_block[8+1+2*i]+8*4];
nnz_cache[3+8* 6 + 2*8*i]= nnz[left_block[8+0+2*i]+4*4];
nnz_cache[3+8* 7 + 2*8*i]= nnz[left_block[8+1+2*i]+4*4];
nnz_cache[3+8*11 + 2*8*i]= nnz[left_block[8+0+2*i]+8*4];
nnz_cache[3+8*12 + 2*8*i]= nnz[left_block[8+1+2*i]+8*4];
}else{
h->non_zero_count_cache[3+8* 6 + 8*i]= h->non_zero_count[left_xy[i]][left_block[8+4+2*i]];
h->non_zero_count_cache[3+8*11 + 8*i]= h->non_zero_count[left_xy[i]][left_block[8+5+2*i]];
nnz_cache[3+8* 6 + 8*i]= nnz[left_block[8+4+2*i]];
nnz_cache[3+8*11 + 8*i]= nnz[left_block[8+5+2*i]];
}
}else{
h->non_zero_count_cache[3+8* 1 + 2*8*i]=
h->non_zero_count_cache[3+8* 2 + 2*8*i]=
h->non_zero_count_cache[3+8* 6 + 2*8*i]=
h->non_zero_count_cache[3+8* 7 + 2*8*i]=
h->non_zero_count_cache[3+8*11 + 2*8*i]=
h->non_zero_count_cache[3+8*12 + 2*8*i]= CABAC && !IS_INTRA(mb_type) ? 0 : 64;
nnz_cache[3+8* 1 + 2*8*i]=
nnz_cache[3+8* 2 + 2*8*i]=
nnz_cache[3+8* 6 + 2*8*i]=
nnz_cache[3+8* 7 + 2*8*i]=
nnz_cache[3+8*11 + 2*8*i]=
nnz_cache[3+8*12 + 2*8*i]= CABAC && !IS_INTRA(mb_type) ? 0 : 64;
}
}
......@@ -1005,141 +1005,142 @@ static void fill_decode_caches(H264Context *h, int mb_type){
if(IS_INTER(mb_type) || (IS_DIRECT(mb_type) && h->direct_spatial_mv_pred)){
int list;
int b_stride = h->b_stride;
for(list=0; list<h->list_count; list++){
int8_t *ref_cache = &h->ref_cache[list][scan8[0]];
int8_t *ref = s->current_picture.ref_index[list];
int16_t (*mv_cache)[2] = &h->mv_cache[list][scan8[0]];
int16_t (*mv)[2] = s->current_picture.motion_val[list];
if(!USES_LIST(mb_type, list)){
/*if(!h->mv_cache_clean[list]){
memset(h->mv_cache [list], 0, 8*5*2*sizeof(int16_t)); //FIXME clean only input? clean at all?
memset(h->ref_cache[list], PART_NOT_AVAILABLE, 8*5*sizeof(int8_t));
h->mv_cache_clean[list]= 1;
}*/
continue;
}
assert(!(IS_DIRECT(mb_type) && !h->direct_spatial_mv_pred));
h->mv_cache_clean[list]= 0;
if(USES_LIST(top_type, list)){
const int b_xy= h->mb2b_xy[top_xy] + 3*h->b_stride;
AV_COPY128(h->mv_cache[list][scan8[0] + 0 - 1*8], s->current_picture.motion_val[list][b_xy + 0]);
h->ref_cache[list][scan8[0] + 0 - 1*8]=
h->ref_cache[list][scan8[0] + 1 - 1*8]= s->current_picture.ref_index[list][4*top_xy + 2];
h->ref_cache[list][scan8[0] + 2 - 1*8]=
h->ref_cache[list][scan8[0] + 3 - 1*8]= s->current_picture.ref_index[list][4*top_xy + 3];
const int b_xy= h->mb2b_xy[top_xy] + 3*b_stride;
AV_COPY128(mv_cache[0 - 1*8], mv[b_xy + 0]);
ref_cache[0 - 1*8]=
ref_cache[1 - 1*8]= ref[4*top_xy + 2];
ref_cache[2 - 1*8]=
ref_cache[3 - 1*8]= ref[4*top_xy + 3];
}else{
AV_ZERO128(h->mv_cache[list][scan8[0] + 0 - 1*8]);
AV_WN32A(&h->ref_cache[list][scan8[0] + 0 - 1*8], ((top_type ? LIST_NOT_USED : PART_NOT_AVAILABLE)&0xFF)*0x01010101);
AV_ZERO128(mv_cache[0 - 1*8]);
AV_WN32A(&ref_cache[0 - 1*8], ((top_type ? LIST_NOT_USED : PART_NOT_AVAILABLE)&0xFF)*0x01010101);
}
if(mb_type & (MB_TYPE_16x8|MB_TYPE_8x8)){
for(i=0; i<2; i++){
int cache_idx = scan8[0] - 1 + i*2*8;
int cache_idx = -1 + i*2*8;
if(USES_LIST(left_type[i], list)){
const int b_xy= h->mb2b_xy[left_xy[i]] + 3;
const int b8_xy= 4*left_xy[i] + 1;
AV_COPY32(h->mv_cache[list][cache_idx ], s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[0+i*2]]);
AV_COPY32(h->mv_cache[list][cache_idx+8], s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[1+i*2]]);
h->ref_cache[list][cache_idx ]= s->current_picture.ref_index[list][b8_xy + (left_block[0+i*2]&~1)];
h->ref_cache[list][cache_idx+8]= s->current_picture.ref_index[list][b8_xy + (left_block[1+i*2]&~1)];
AV_COPY32(mv_cache[cache_idx ], mv[b_xy + b_stride*left_block[0+i*2]]);
AV_COPY32(mv_cache[cache_idx+8], mv[b_xy + b_stride*left_block[1+i*2]]);
ref_cache[cache_idx ]= ref[b8_xy + (left_block[0+i*2]&~1)];
ref_cache[cache_idx+8]= ref[b8_xy + (left_block[1+i*2]&~1)];
}else{
AV_ZERO32(h->mv_cache [list][cache_idx ]);
AV_ZERO32(h->mv_cache [list][cache_idx+8]);
h->ref_cache[list][cache_idx ]=
h->ref_cache[list][cache_idx+8]= (left_type[i]) ? LIST_NOT_USED : PART_NOT_AVAILABLE;
AV_ZERO32(mv_cache[cache_idx ]);
AV_ZERO32(mv_cache[cache_idx+8]);
ref_cache[cache_idx ]=
ref_cache[cache_idx+8]= (left_type[i]) ? LIST_NOT_USED : PART_NOT_AVAILABLE;
}
}
}else{
if(USES_LIST(left_type[0], list)){
const int b_xy= h->mb2b_xy[left_xy[0]] + 3;
const int b8_xy= 4*left_xy[0] + 1;
AV_COPY32(h->mv_cache[list][scan8[0] - 1], s->current_picture.motion_val[list][b_xy + h->b_stride*left_block[0]]);
h->ref_cache[list][scan8[0] - 1]= s->current_picture.ref_index[list][b8_xy + (left_block[0]&~1)];
AV_COPY32(mv_cache[-1], mv[b_xy + b_stride*left_block[0]]);
ref_cache[-1]= ref[b8_xy + (left_block[0]&~1)];
}else{
AV_ZERO32(h->mv_cache [list][scan8[0] - 1]);
h->ref_cache[list][scan8[0] - 1]= left_type[0] ? LIST_NOT_USED : PART_NOT_AVAILABLE;
AV_ZERO32(mv_cache[-1]);
ref_cache[-1]= left_type[0] ? LIST_NOT_USED : PART_NOT_AVAILABLE;
}
}
if(USES_LIST(topright_type, list)){
const int b_xy= h->mb2b_xy[topright_xy] + 3*h->b_stride;
AV_COPY32(h->mv_cache[list][scan8[0] + 4 - 1*8], s->current_picture.motion_val[list][b_xy]);
h->ref_cache[list][scan8[0] + 4 - 1*8]= s->current_picture.ref_index[list][4*topright_xy + 2];
const int b_xy= h->mb2b_xy[topright_xy] + 3*b_stride;
AV_COPY32(mv_cache[4 - 1*8], mv[b_xy]);
ref_cache[4 - 1*8]= ref[4*topright_xy + 2];
}else{
AV_ZERO32(h->mv_cache [list][scan8[0] + 4 - 1*8]);
h->ref_cache[list][scan8[0] + 4 - 1*8]= topright_type ? LIST_NOT_USED : PART_NOT_AVAILABLE;
AV_ZERO32(mv_cache[4 - 1*8]);
ref_cache[4 - 1*8]= topright_type ? LIST_NOT_USED : PART_NOT_AVAILABLE;
}
if(h->ref_cache[list][scan8[0] + 4 - 1*8] < 0){
if(ref_cache[4 - 1*8] < 0){
if(USES_LIST(topleft_type, list)){
const int b_xy = h->mb2b_xy [topleft_xy] + 3 + h->b_stride + (h->topleft_partition & 2*h->b_stride);
const int b_xy = h->mb2b_xy[topleft_xy] + 3 + b_stride + (h->topleft_partition & 2*b_stride);
const int b8_xy= 4*topleft_xy + 1 + (h->topleft_partition & 2);
AV_COPY32(h->mv_cache[list][scan8[0] - 1 - 1*8], s->current_picture.motion_val[list][b_xy]);
h->ref_cache[list][scan8[0] - 1 - 1*8]= s->current_picture.ref_index[list][b8_xy];
AV_COPY32(mv_cache[-1 - 1*8], mv[b_xy]);
ref_cache[-1 - 1*8]= ref[b8_xy];
}else{
AV_ZERO32(h->mv_cache[list][scan8[0] - 1 - 1*8]);
h->ref_cache[list][scan8[0] - 1 - 1*8]= topleft_type ? LIST_NOT_USED : PART_NOT_AVAILABLE;
AV_ZERO32(mv_cache[-1 - 1*8]);
ref_cache[-1 - 1*8]= topleft_type ? LIST_NOT_USED : PART_NOT_AVAILABLE;
}
}
if((mb_type&(MB_TYPE_SKIP|MB_TYPE_DIRECT2)) && !FRAME_MBAFF)
continue;
if(!(mb_type&(MB_TYPE_SKIP|MB_TYPE_DIRECT2))) {
h->ref_cache[list][scan8[4 ]] =
h->ref_cache[list][scan8[12]] = PART_NOT_AVAILABLE;
AV_ZERO32(h->mv_cache [list][scan8[4 ]]);
AV_ZERO32(h->mv_cache [list][scan8[12]]);
if(!(mb_type&(MB_TYPE_SKIP|MB_TYPE_DIRECT2))){
uint8_t (*mvd_cache)[2] = &h->mvd_cache[list][scan8[0]];
uint8_t (*mvd)[2] = h->mvd_table[list];
ref_cache[2+8*0] =
ref_cache[2+8*2] = PART_NOT_AVAILABLE;
AV_ZERO32(mv_cache[2+8*0]);
AV_ZERO32(mv_cache[2+8*2]);
if( CABAC ) {
/* XXX beurk, Load mvd */
if(USES_LIST(top_type, list)){
const int b_xy= h->mb2br_xy[top_xy];
AV_COPY64(h->mvd_cache[list][scan8[0] + 0 - 1*8], h->mvd_table[list][b_xy + 0]);
AV_COPY64(mvd_cache[0 - 1*8], mvd[b_xy + 0]);
}else{
AV_ZERO64(h->mvd_cache[list][scan8[0] + 0 - 1*8]);
AV_ZERO64(mvd_cache[0 - 1*8]);
}
if(USES_LIST(left_type[0], list)){
const int b_xy= h->mb2br_xy[left_xy[0]] + 6;
AV_COPY16(h->mvd_cache[list][scan8[0] - 1 + 0*8], h->mvd_table[list][b_xy - left_block[0]]);
AV_COPY16(h->mvd_cache[list][scan8[0] - 1 + 1*8], h->mvd_table[list][b_xy - left_block[1]]);
AV_COPY16(mvd_cache[-1 + 0*8], mvd[b_xy - left_block[0]]);
AV_COPY16(mvd_cache[-1 + 1*8], mvd[b_xy - left_block[1]]);
}else{
AV_ZERO16(h->mvd_cache [list][scan8[0] - 1 + 0*8]);
AV_ZERO16(h->mvd_cache [list][scan8[0] - 1 + 1*8]);
AV_ZERO16(mvd_cache[-1 + 0*8]);
AV_ZERO16(mvd_cache[-1 + 1*8]);
}
if(USES_LIST(left_type[1], list)){
const int b_xy= h->mb2br_xy[left_xy[1]] + 6;
AV_COPY16(h->mvd_cache[list][scan8[0] - 1 + 2*8], h->mvd_table[list][b_xy - left_block[2]]);
AV_COPY16(h->mvd_cache[list][scan8[0] - 1 + 3*8], h->mvd_table[list][b_xy - left_block[3]]);
AV_COPY16(mvd_cache[-1 + 2*8], mvd[b_xy - left_block[2]]);
AV_COPY16(mvd_cache[-1 + 3*8], mvd[b_xy - left_block[3]]);
}else{
AV_ZERO16(h->mvd_cache [list][scan8[0] - 1 + 2*8]);
AV_ZERO16(h->mvd_cache [list][scan8[0] - 1 + 3*8]);
AV_ZERO16(mvd_cache[-1 + 2*8]);
AV_ZERO16(mvd_cache[-1 + 3*8]);
}
AV_ZERO16(h->mvd_cache [list][scan8[4 ]]);
AV_ZERO16(h->mvd_cache [list][scan8[12]]);
AV_ZERO16(mvd_cache[2+8*0]);
AV_ZERO16(mvd_cache[2+8*2]);
if(h->slice_type_nos == AV_PICTURE_TYPE_B){
fill_rectangle(&h->direct_cache[scan8[0]], 4, 4, 8, MB_TYPE_16x16>>1, 1);
uint8_t *direct_cache = &h->direct_cache[scan8[0]];
uint8_t *direct_table = h->direct_table;
fill_rectangle(direct_cache, 4, 4, 8, MB_TYPE_16x16>>1, 1);
if(IS_DIRECT(top_type)){
AV_WN32A(&h->direct_cache[scan8[0] - 1*8], 0x01010101u*(MB_TYPE_DIRECT2>>1));
AV_WN32A(&direct_cache[-1*8], 0x01010101u*(MB_TYPE_DIRECT2>>1));
}else if(IS_8X8(top_type)){
int b8_xy = 4*top_xy;
h->direct_cache[scan8[0] + 0 - 1*8]= h->direct_table[b8_xy + 2];
h->direct_cache[scan8[0] + 2 - 1*8]= h->direct_table[b8_xy + 3];
direct_cache[0 - 1*8]= direct_table[b8_xy + 2];
direct_cache[2 - 1*8]= direct_table[b8_xy + 3];
}else{
AV_WN32A(&h->direct_cache[scan8[0] - 1*8], 0x01010101*(MB_TYPE_16x16>>1));
AV_WN32A(&direct_cache[-1*8], 0x01010101*(MB_TYPE_16x16>>1));
}
if(IS_DIRECT(left_type[0]))
h->direct_cache[scan8[0] - 1 + 0*8]= MB_TYPE_DIRECT2>>1;
direct_cache[-1 + 0*8]= MB_TYPE_DIRECT2>>1;
else if(IS_8X8(left_type[0]))
h->direct_cache[scan8[0] - 1 + 0*8]= h->direct_table[4*left_xy[0] + 1 + (left_block[0]&~1)];
direct_cache[-1 + 0*8]= direct_table[4*left_xy[0] + 1 + (left_block[0]&~1)];
else
h->direct_cache[scan8[0] - 1 + 0*8]= MB_TYPE_16x16>>1;
direct_cache[-1 + 0*8]= MB_TYPE_16x16>>1;
if(IS_DIRECT(left_type[1]))
h->direct_cache[scan8[0] - 1 + 2*8]= MB_TYPE_DIRECT2>>1;
direct_cache[-1 + 2*8]= MB_TYPE_DIRECT2>>1;
else if(IS_8X8(left_type[1]))
h->direct_cache[scan8[0] - 1 + 2*8]= h->direct_table[4*left_xy[1] + 1 + (left_block[2]&~1)];
direct_cache[-1 + 2*8]= direct_table[4*left_xy[1] + 1 + (left_block[2]&~1)];
else
h->direct_cache[scan8[0] - 1 + 2*8]= MB_TYPE_16x16>>1;
direct_cache[-1 + 2*8]= MB_TYPE_16x16>>1;
}
}
}
......
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