Skip to content

F
ffmpeg.wasm-core
Commit 99dde603 authored by Anton Khirnov's avatar Anton Khirnov
Browse files
Options

svq3: move {ref,mv}_cache to the SVQ3Context

parent 89a13998
Hide whitespace changes
Inline Side-by-side
Showing with 32 additions and 32 deletions
libavcodec/svq3.c
View file @ 99dde603
...@@ -105,6 +105,9 @@ typedef struct SVQ3Context { ...@@ -105,6 +105,9 @@ typedef struct SVQ3Context {
unsigned int left_samples_available; unsigned int left_samples_available;
uint8_t *edge_emu_buffer; uint8_t *edge_emu_buffer;
DECLARE_ALIGNED(16, int16_t, mv_cache)[2][5 * 8][2];
DECLARE_ALIGNED(8, int8_t, ref_cache)[2][5 * 8];
} SVQ3Context; } SVQ3Context;
#define FULLPEL_MODE 1 #define FULLPEL_MODE 1
...@@ -312,18 +315,17 @@ static inline int svq3_decode_block(GetBitContext *gb, int16_t *block, ...@@ -312,18 +315,17 @@ static inline int svq3_decode_block(GetBitContext *gb, int16_t *block,
} }
static av_always_inline int static av_always_inline int
svq3_fetch_diagonal_mv(const H264Context *h, H264SliceContext *sl, svq3_fetch_diagonal_mv(const SVQ3Context *s, const int16_t **C,
const int16_t **C,
int i, int list, int part_width) int i, int list, int part_width)
{ {
const int topright_ref = sl->ref_cache[list][i - 8 + part_width]; const int topright_ref = s->ref_cache[list][i - 8 + part_width];
if (topright_ref != PART_NOT_AVAILABLE) { if (topright_ref != PART_NOT_AVAILABLE) {
*C = sl->mv_cache[list][i - 8 + part_width]; *C = s->mv_cache[list][i - 8 + part_width];
return topright_ref; return topright_ref;
} else { } else {
*C = sl->mv_cache[list][i - 8 - 1]; *C = s->mv_cache[list][i - 8 - 1];
return sl->ref_cache[list][i - 8 - 1]; return s->ref_cache[list][i - 8 - 1];
} }
} }
...@@ -334,16 +336,15 @@ svq3_fetch_diagonal_mv(const H264Context *h, H264SliceContext *sl, ...@@ -334,16 +336,15 @@ svq3_fetch_diagonal_mv(const H264Context *h, H264SliceContext *sl,
* @param mx the x component of the predicted motion vector * @param mx the x component of the predicted motion vector
* @param my the y component of the predicted motion vector * @param my the y component of the predicted motion vector
*/ */
static av_always_inline void svq3_pred_motion(const H264Context *const h, static av_always_inline void svq3_pred_motion(const SVQ3Context *s, int n,
H264SliceContext *sl, int n,
int part_width, int list, int part_width, int list,
int ref, int *const mx, int *const my) int ref, int *const mx, int *const my)
{ {
const int index8 = scan8[n]; const int index8 = scan8[n];
const int top_ref = sl->ref_cache[list][index8 - 8]; const int top_ref = s->ref_cache[list][index8 - 8];
const int left_ref = sl->ref_cache[list][index8 - 1]; const int left_ref = s->ref_cache[list][index8 - 1];
const int16_t *const A = sl->mv_cache[list][index8 - 1]; const int16_t *const A = s->mv_cache[list][index8 - 1];
const int16_t *const B = sl->mv_cache[list][index8 - 8]; const int16_t *const B = s->mv_cache[list][index8 - 8];
const int16_t *C; const int16_t *C;
int diagonal_ref, match_count; int diagonal_ref, match_count;
...@@ -355,7 +356,7 @@ static av_always_inline void svq3_pred_motion(const H264Context *const h, ...@@ -355,7 +356,7 @@ static av_always_inline void svq3_pred_motion(const H264Context *const h,
* . . . L . . . . * . . . L . . . .
*/ */
diagonal_ref = svq3_fetch_diagonal_mv(h, sl, &C, index8, list, part_width); diagonal_ref = svq3_fetch_diagonal_mv(s, &C, index8, list, part_width);
match_count = (diagonal_ref == ref) + (top_ref == ref) + (left_ref == ref); match_count = (diagonal_ref == ref) + (top_ref == ref) + (left_ref == ref);
if (match_count > 1) { //most common if (match_count > 1) { //most common
*mx = mid_pred(A[0], B[0], C[0]); *mx = mid_pred(A[0], B[0], C[0]);
...@@ -464,7 +465,6 @@ static inline int svq3_mc_dir(SVQ3Context *s, int size, int mode, ...@@ -464,7 +465,6 @@ static inline int svq3_mc_dir(SVQ3Context *s, int size, int mode,
{ {
int i, j, k, mx, my, dx, dy, x, y; int i, j, k, mx, my, dx, dy, x, y;
H264Context *h = &s->h; H264Context *h = &s->h;
H264SliceContext *sl = &h->slice_ctx[0];
const int part_width = ((size & 5) == 4) ? 4 : 16 >> (size & 1); const int part_width = ((size & 5) == 4) ? 4 : 16 >> (size & 1);
const int part_height = 16 >> ((unsigned)(size + 1) / 3); const int part_height = 16 >> ((unsigned)(size + 1) / 3);
const int extra_width = (mode == PREDICT_MODE) ? -16 * 6 : 0; const int extra_width = (mode == PREDICT_MODE) ? -16 * 6 : 0;
...@@ -482,7 +482,7 @@ static inline int svq3_mc_dir(SVQ3Context *s, int size, int mode, ...@@ -482,7 +482,7 @@ static inline int svq3_mc_dir(SVQ3Context *s, int size, int mode,
(j >> 1 & 4) + (i & 8); (j >> 1 & 4) + (i & 8);
if (mode != PREDICT_MODE) { if (mode != PREDICT_MODE) {
svq3_pred_motion(h, sl, k, part_width >> 2, dir, 1, &mx, &my); svq3_pred_motion(s, k, part_width >> 2, dir, 1, &mx, &my);
} else { } else {
mx = s->next_pic->motion_val[0][b_xy][0] << 1; mx = s->next_pic->motion_val[0][b_xy][0] << 1;
my = s->next_pic->motion_val[0][b_xy][1] << 1; my = s->next_pic->motion_val[0][b_xy][1] << 1;
...@@ -554,15 +554,15 @@ static inline int svq3_mc_dir(SVQ3Context *s, int size, int mode, ...@@ -554,15 +554,15 @@ static inline int svq3_mc_dir(SVQ3Context *s, int size, int mode,
int32_t mv = pack16to32(mx, my); int32_t mv = pack16to32(mx, my);
if (part_height == 8 && i < 8) { if (part_height == 8 && i < 8) {
AV_WN32A(sl->mv_cache[dir][scan8[k] + 1 * 8], mv); AV_WN32A(s->mv_cache[dir][scan8[k] + 1 * 8], mv);
if (part_width == 8 && j < 8) if (part_width == 8 && j < 8)
AV_WN32A(sl->mv_cache[dir][scan8[k] + 1 + 1 * 8], mv); AV_WN32A(s->mv_cache[dir][scan8[k] + 1 + 1 * 8], mv);
} }
if (part_width == 8 && j < 8) if (part_width == 8 && j < 8)
AV_WN32A(sl->mv_cache[dir][scan8[k] + 1], mv); AV_WN32A(s->mv_cache[dir][scan8[k] + 1], mv);
if (part_width == 4 || part_height == 4) if (part_width == 4 || part_height == 4)
AV_WN32A(sl->mv_cache[dir][scan8[k]], mv); AV_WN32A(s->mv_cache[dir][scan8[k]], mv);
} }
/* write back motion vectors */ /* write back motion vectors */
...@@ -741,36 +741,36 @@ static int svq3_decode_mb(SVQ3Context *s, unsigned int mb_type) ...@@ -741,36 +741,36 @@ static int svq3_decode_mb(SVQ3Context *s, unsigned int mb_type)
for (m = 0; m < 2; m++) { for (m = 0; m < 2; m++) {
if (s->mb_x > 0 && s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - 1] + 6] != -1) { if (s->mb_x > 0 && s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - 1] + 6] != -1) {
for (i = 0; i < 4; i++) for (i = 0; i < 4; i++)
AV_COPY32(sl->mv_cache[m][scan8[0] - 1 + i * 8], AV_COPY32(s->mv_cache[m][scan8[0] - 1 + i * 8],
h->cur_pic.motion_val[m][b_xy - 1 + i * h->b_stride]); h->cur_pic.motion_val[m][b_xy - 1 + i * h->b_stride]);
} else { } else {
for (i = 0; i < 4; i++) for (i = 0; i < 4; i++)
AV_ZERO32(sl->mv_cache[m][scan8[0] - 1 + i * 8]); AV_ZERO32(s->mv_cache[m][scan8[0] - 1 + i * 8]);
} }
if (s->mb_y > 0) { if (s->mb_y > 0) {
memcpy(sl->mv_cache[m][scan8[0] - 1 * 8], memcpy(s->mv_cache[m][scan8[0] - 1 * 8],
h->cur_pic.motion_val[m][b_xy - h->b_stride], h->cur_pic.motion_val[m][b_xy - h->b_stride],
4 * 2 * sizeof(int16_t)); 4 * 2 * sizeof(int16_t));
memset(&sl->ref_cache[m][scan8[0] - 1 * 8], memset(&s->ref_cache[m][scan8[0] - 1 * 8],
(s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride]] == -1) ? PART_NOT_AVAILABLE : 1, 4); (s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride]] == -1) ? PART_NOT_AVAILABLE : 1, 4);
if (s->mb_x < h->mb_width - 1) { if (s->mb_x < h->mb_width - 1) {
AV_COPY32(sl->mv_cache[m][scan8[0] + 4 - 1 * 8], AV_COPY32(s->mv_cache[m][scan8[0] + 4 - 1 * 8],
h->cur_pic.motion_val[m][b_xy - h->b_stride + 4]); h->cur_pic.motion_val[m][b_xy - h->b_stride + 4]);
sl->ref_cache[m][scan8[0] + 4 - 1 * 8] = s->ref_cache[m][scan8[0] + 4 - 1 * 8] =
(s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride + 1] + 6] == -1 || (s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride + 1] + 6] == -1 ||
s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride]] == -1) ? PART_NOT_AVAILABLE : 1; s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride]] == -1) ? PART_NOT_AVAILABLE : 1;
} else } else
sl->ref_cache[m][scan8[0] + 4 - 1 * 8] = PART_NOT_AVAILABLE; s->ref_cache[m][scan8[0] + 4 - 1 * 8] = PART_NOT_AVAILABLE;
if (s->mb_x > 0) { if (s->mb_x > 0) {
AV_COPY32(sl->mv_cache[m][scan8[0] - 1 - 1 * 8], AV_COPY32(s->mv_cache[m][scan8[0] - 1 - 1 * 8],
h->cur_pic.motion_val[m][b_xy - h->b_stride - 1]); h->cur_pic.motion_val[m][b_xy - h->b_stride - 1]);
sl->ref_cache[m][scan8[0] - 1 - 1 * 8] = s->ref_cache[m][scan8[0] - 1 - 1 * 8] =
(s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride - 1] + 3] == -1) ? PART_NOT_AVAILABLE : 1; (s->intra4x4_pred_mode[h->mb2br_xy[mb_xy - h->mb_stride - 1] + 3] == -1) ? PART_NOT_AVAILABLE : 1;
} else } else
sl->ref_cache[m][scan8[0] - 1 - 1 * 8] = PART_NOT_AVAILABLE; s->ref_cache[m][scan8[0] - 1 - 1 * 8] = PART_NOT_AVAILABLE;
} else } else
memset(&sl->ref_cache[m][scan8[0] - 1 * 8 - 1], memset(&s->ref_cache[m][scan8[0] - 1 * 8 - 1],
PART_NOT_AVAILABLE, 8); PART_NOT_AVAILABLE, 8);
if (h->pict_type != AV_PICTURE_TYPE_B) if (h->pict_type != AV_PICTURE_TYPE_B)
...@@ -1483,9 +1483,9 @@ static int svq3_decode_frame(AVCodecContext *avctx, void *data, ...@@ -1483,9 +1483,9 @@ static int svq3_decode_frame(AVCodecContext *avctx, void *data,
for (i = 0; i < 4; i++) { for (i = 0; i < 4; i++) {
int j; int j;
for (j = -1; j < 4; j++) for (j = -1; j < 4; j++)
sl->ref_cache[m][scan8[0] + 8 * i + j] = 1; s->ref_cache[m][scan8[0] + 8 * i + j] = 1;
if (i < 3) if (i < 3)
sl->ref_cache[m][scan8[0] + 8 * i + j] = PART_NOT_AVAILABLE; s->ref_cache[m][scan8[0] + 8 * i + j] = PART_NOT_AVAILABLE;
} }
} }
......
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