Chiebot-Mirror
/
FFmpeg
mirror of https://github.com/FFmpeg/FFmpeg.git
8
0
Fork 0
Code Issues Projects Releases Wiki Activity

vp9: split decode_coeff_b loop inside txsz branch.

Browse Source 
The advantage of this is that the is32x32 division branch in
decode_coeffs_b is removed from the inner loop to outside the block
coef decoding loop in decode_coeffs. Also, it allows us to merge the
txsz branches from the block coef decoding loop, the context merge
and the context split.
pull/56/merge
Ronald S. Bultje 11 years ago committed by Clément Bœsch
parent fdb093c4e4
commit 21a0451167
1 changed files with 99 additions and 56 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Unified View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 155
      libavcodec/vp9.c

155
libavcodec/vp9.c
Unescape View File

@ -2011,12 +2011,13 @@ static void decode_mode(AVCodecContext *ctx)
} }
} }
// FIXME remove tx argument, and merge cnt/eob arguments? // FIXME merge cnt/eob arguments?
static int decode_coeffs_b(VP56RangeCoder *c, int16_t *coef, int n_coeffs, static av_always_inline int
enum TxfmMode tx, unsigned (*cnt)[6][3], decode_coeffs_b_generic(VP56RangeCoder *c, int16_t *coef, int n_coeffs,
unsigned (*eob)[6][2], uint8_t (*p)[6][11], int is_tx32x32, unsigned (*cnt)[6][3],
int nnz, const int16_t *scan, const int16_t (*nb)[2], unsigned (*eob)[6][2], uint8_t (*p)[6][11],
const int16_t *band_counts, const int16_t *qmul) int nnz, const int16_t *scan, const int16_t (*nb)[2],
const int16_t *band_counts, const int16_t *qmul)
{ {
int i = 0, band = 0, band_left = band_counts[band]; int i = 0, band = 0, band_left = band_counts[band];
uint8_t *tp = p[0][nnz]; uint8_t *tp = p[0][nnz];
@ -2108,7 +2109,7 @@ static int decode_coeffs_b(VP56RangeCoder *c, int16_t *coef, int n_coeffs,
} }
if (!--band_left) if (!--band_left)
band_left = band_counts[++band]; band_left = band_counts[++band];
if (tx == TX_32X32) // FIXME slow if (is_tx32x32)
coef[rc] = ((vp8_rac_get(c) ? -val : val) * qmul[!!i]) / 2; coef[rc] = ((vp8_rac_get(c) ? -val : val) * qmul[!!i]) / 2;
else else
coef[rc] = (vp8_rac_get(c) ? -val : val) * qmul[!!i]; coef[rc] = (vp8_rac_get(c) ? -val : val) * qmul[!!i];
@ -2119,6 +2120,26 @@ static int decode_coeffs_b(VP56RangeCoder *c, int16_t *coef, int n_coeffs,
return i; return i;
} }
static int decode_coeffs_b(VP56RangeCoder *c, int16_t *coef, int n_coeffs,
unsigned (*cnt)[6][3], unsigned (*eob)[6][2],
uint8_t (*p)[6][11], int nnz, const int16_t *scan,
const int16_t (*nb)[2], const int16_t *band_counts,
const int16_t *qmul)
{
return decode_coeffs_b_generic(c, coef, n_coeffs, 0, cnt, eob, p,
nnz, scan, nb, band_counts, qmul);
}
static int decode_coeffs_b32(VP56RangeCoder *c, int16_t *coef, int n_coeffs,
unsigned (*cnt)[6][3], unsigned (*eob)[6][2],
uint8_t (*p)[6][11], int nnz, const int16_t *scan,
const int16_t (*nb)[2], const int16_t *band_counts,
const int16_t *qmul)
{
return decode_coeffs_b_generic(c, coef, n_coeffs, 1, cnt, eob, p,
nnz, scan, nb, band_counts, qmul);
}
static void decode_coeffs(AVCodecContext *ctx) static void decode_coeffs(AVCodecContext *ctx)
{ {
VP9Context *s = ctx->priv_data; VP9Context *s = ctx->priv_data;
@ -2130,8 +2151,7 @@ static void decode_coeffs(AVCodecContext *ctx)
int w4 = bwh_tab[1][b->bs][0] << 1, h4 = bwh_tab[1][b->bs][1] << 1; int w4 = bwh_tab[1][b->bs][0] << 1, h4 = bwh_tab[1][b->bs][1] << 1;
int end_x = FFMIN(2 * (s->cols - col), w4); int end_x = FFMIN(2 * (s->cols - col), w4);
int end_y = FFMIN(2 * (s->rows - row), h4); int end_y = FFMIN(2 * (s->rows - row), h4);
int n, pl, x, y, step1d = 1 << b->tx, step = 1 << (b->tx * 2); int n, pl, x, y, res;
int uvstep1d = 1 << b->uvtx, uvstep = 1 << (b->uvtx * 2), res;
int16_t (*qmul)[2] = s->segmentation.feat[b->seg_id].qmul; int16_t (*qmul)[2] = s->segmentation.feat[b->seg_id].qmul;
int tx = 4 * s->lossless + b->tx; int tx = 4 * s->lossless + b->tx;
const int16_t * const *yscans = vp9_scans[tx]; const int16_t * const *yscans = vp9_scans[tx];
@ -2158,29 +2178,22 @@ static void decode_coeffs(AVCodecContext *ctx)
MERGE(a, end_x, step, rd); \ MERGE(a, end_x, step, rd); \
} while (0) } while (0)
/* y tokens */ #define DECODE_Y_COEF_LOOP(step, mode_index, v) \
switch (b->tx) { for (n = 0, y = 0; y < end_y; y += step) { \
case TX_8X8: MERGE_CTX(2, AV_RN16A); break; for (x = 0; x < end_x; x += step, n += step * step) { \
case TX_16X16: MERGE_CTX(4, AV_RN32A); break; enum TxfmType txtp = vp9_intra_txfm_type[b->mode[mode_index]]; \
case TX_32X32: MERGE_CTX(8, AV_RN64A); break; res = decode_coeffs_b##v(&s->c, s->block + 16 * n, 16 * step * step, \
} c, e, p, a[x] + l[y], yscans[txtp], \
for (n = 0, y = 0; y < end_y; y += step1d) { ynbs[txtp], y_band_counts, qmul[0]); \
for (x = 0; x < end_x; x += step1d, n += step) { a[x] = l[y] = !!res; \
enum TxfmType txtp = vp9_intra_txfm_type[b->mode[b->tx == TX_4X4 && if (step >= 4) { \
b->bs > BS_8x8 ? AV_WN16A(&s->eob[n], res); \
n : 0]]; } else { \
int nnz = a[x] + l[y]; s->eob[n] = res; \
res = decode_coeffs_b(&s->c, s->block + 16 * n, 16 * step, } \
b->tx, c, e, p, nnz, yscans[txtp], } \
ynbs[txtp], y_band_counts, qmul[0]);
a[x] = l[y] = !!res;
if (b->tx > TX_8X8) {
AV_WN16A(&s->eob[n], res);
} else {
s->eob[n] = res;
}
}
} }
#define SPLAT(la, end, step, cond) \ #define SPLAT(la, end, step, cond) \
if (step == 2) { \ if (step == 2) { \
for (n = 1; n < end; n += step) \ for (n = 1; n < end; n += step) \
@ -2215,10 +2228,38 @@ static void decode_coeffs(AVCodecContext *ctx)
SPLAT(a, end_x, step, end_x == w4); \ SPLAT(a, end_x, step, end_x == w4); \
SPLAT(l, end_y, step, end_y == h4); \ SPLAT(l, end_y, step, end_y == h4); \
} while (0) } while (0)
/* y tokens */
switch (b->tx) { switch (b->tx) {
case TX_8X8: SPLAT_CTX(2); break; case TX_4X4:
case TX_16X16: SPLAT_CTX(4); break; DECODE_Y_COEF_LOOP(1, b->bs > BS_8x8 ? n : 0,);
case TX_32X32: SPLAT_CTX(8); break; break;
case TX_8X8:
MERGE_CTX(2, AV_RN16A);
DECODE_Y_COEF_LOOP(2, 0,);
SPLAT_CTX(2);
break;
case TX_16X16:
MERGE_CTX(4, AV_RN32A);
DECODE_Y_COEF_LOOP(4, 0,);
SPLAT_CTX(4);
break;
case TX_32X32:
MERGE_CTX(8, AV_RN64A);
DECODE_Y_COEF_LOOP(8, 0, 32);
SPLAT_CTX(8);
break;
}
#define DECODE_UV_COEF_LOOP(step) \
for (n = 0, y = 0; y < end_y; y += step) { \
for (x = 0; x < end_x; x += step, n += step * step) { \
res = decode_coeffs_b(&s->c, s->uvblock[pl] + 16 * n, \
16 * step * step, c, e, p, a[x] + l[y], \
uvscan, uvnb, uv_band_counts, qmul[1]); \
a[x] = l[y] = !!res; \
s->uveob[pl][n] = res; \
} \
} }
p = s->prob.coef[b->uvtx][1 /* uv */][!b->intra]; p = s->prob.coef[b->uvtx][1 /* uv */][!b->intra];
@ -2232,28 +2273,30 @@ static void decode_coeffs(AVCodecContext *ctx)
a = &s->above_uv_nnz_ctx[pl][col]; a = &s->above_uv_nnz_ctx[pl][col];
l = &s->left_uv_nnz_ctx[pl][row & 7]; l = &s->left_uv_nnz_ctx[pl][row & 7];
switch (b->uvtx) { switch (b->uvtx) {
case TX_8X8: MERGE_CTX(2, AV_RN16A); break; case TX_4X4:
case TX_16X16: MERGE_CTX(4, AV_RN32A); break; DECODE_UV_COEF_LOOP(1);
case TX_32X32: MERGE_CTX(8, AV_RN64A); break; break;
} case TX_8X8:
for (n = 0, y = 0; y < end_y; y += uvstep1d) { MERGE_CTX(2, AV_RN16A);
for (x = 0; x < end_x; x += uvstep1d, n += uvstep) { DECODE_UV_COEF_LOOP(2);
int nnz = a[x] + l[y]; SPLAT_CTX(2);
res = decode_coeffs_b(&s->c, s->uvblock[pl] + 16 * n, break;
16 * uvstep, b->uvtx, c, e, p, nnz, case TX_16X16:
uvscan, uvnb, uv_band_counts, qmul[1]); MERGE_CTX(4, AV_RN32A);
a[x] = l[y] = !!res; DECODE_UV_COEF_LOOP(4);
if (b->uvtx > TX_8X8) { SPLAT_CTX(4);
AV_WN16A(&s->uveob[pl][n], res); break;
} else { case TX_32X32:
s->uveob[pl][n] = res; MERGE_CTX(8, AV_RN64A);
} // a 64x64 (max) uv block can ever only contain 1 tx32x32 block
} // so there is no need to loop
} res = decode_coeffs_b32(&s->c, s->uvblock[pl],
switch (b->uvtx) { 1024, c, e, p, a[0] + l[0],
case TX_8X8: SPLAT_CTX(2); break; uvscan, uvnb, uv_band_counts, qmul[1]);
case TX_16X16: SPLAT_CTX(4); break; a[0] = l[0] = !!res;
case TX_32X32: SPLAT_CTX(8); break; AV_WN16A(&s->uveob[pl][0], res);
SPLAT_CTX(8);
break;
} }
} }
} }

Write Preview
Loading…
Cancel
Save