RepoMirrors/ffmpeg
1
0
Fork 0
mirror of https://git.ffmpeg.org/ffmpeg.git synced 2024-12-28 10:22:10 +00:00
Code Issues Projects Releases Wiki Activity

Merge commit 'cc16da75c2f99d92f7a6461100f041352deb6d88'

Browse Source 
* commit 'cc16da75c2f99d92f7a6461100f041352deb6d88':
  hevc: Add coefficient limiting to speed up IDCT

Noop again as we have these changes already, only random spacing
changes.

Merged-by: Clément Bœsch <cboesch@gopro.com>
This commit is contained in:
Clément Bœsch 2017-01-31 16:02:35 +01:00
commit 05018c2cda
2 changed files with 9 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
libavcodec/hevc_cabac.c
View File

@ -1490,7 +1490,7 @@ void ff_hevc_hls_residual_coding(HEVCContext *s, int x0, int y0,
} else { } else {
int max_xy = FFMAX(last_significant_coeff_x, last_significant_coeff_y); int max_xy = FFMAX(last_significant_coeff_x, last_significant_coeff_y);
if (max_xy == 0) if (max_xy == 0)
s->hevcdsp.idct_dc[log2_trafo_size-2](coeffs); s->hevcdsp.idct_dc[log2_trafo_size - 2](coeffs);
else { else {
int col_limit = last_significant_coeff_x + last_significant_coeff_y + 4; int col_limit = last_significant_coeff_x + last_significant_coeff_y + 4;
if (max_xy < 4) if (max_xy < 4)
@ -1499,7 +1499,7 @@ void ff_hevc_hls_residual_coding(HEVCContext *s, int x0, int y0,
col_limit = FFMIN(8, col_limit); col_limit = FFMIN(8, col_limit);
else if (max_xy < 12) else if (max_xy < 12)
col_limit = FFMIN(24, col_limit); col_limit = FFMIN(24, col_limit);
s->hevcdsp.idct[log2_trafo_size-2](coeffs, col_limit); s->hevcdsp.idct[log2_trafo_size - 2](coeffs, col_limit);
} }
} }
} }

13
libavcodec/hevcdsp_template.c
View File

@ -223,7 +223,7 @@ static void FUNC(transform_4x4_luma)(int16_t *coeffs)
for (i = 0; i < 16; i++) \ for (i = 0; i < 16; i++) \
for (j = 1; j < end; j += 2) \ for (j = 1; j < end; j += 2) \
o_32[i] += transform[j][i] * src[j * sstep]; \ o_32[i] += transform[j][i] * src[j * sstep]; \
TR_16(e_32, src, 1, 2 * sstep, SET, end/2); \ TR_16(e_32, src, 1, 2 * sstep, SET, end / 2); \
\ \
for (i = 0; i < 16; i++) { \ for (i = 0; i < 16; i++) { \
assign(dst[i * dstep], e_32[i] + o_32[i]); \ assign(dst[i * dstep], e_32[i] + o_32[i]); \
@ -232,21 +232,22 @@ static void FUNC(transform_4x4_luma)(int16_t *coeffs)
} while (0) } while (0)
#define IDCT_VAR4(H) \ #define IDCT_VAR4(H) \
int limit2 = FFMIN(col_limit + 4, H) int limit2 = FFMIN(col_limit + 4, H)
#define IDCT_VAR8(H) \ #define IDCT_VAR8(H) \
int limit = FFMIN(col_limit, H); \ int limit = FFMIN(col_limit, H); \
int limit2 = FFMIN(col_limit + 4, H) int limit2 = FFMIN(col_limit + 4, H)
#define IDCT_VAR16(H) IDCT_VAR8(H) #define IDCT_VAR16(H) IDCT_VAR8(H)
#define IDCT_VAR32(H) IDCT_VAR8(H) #define IDCT_VAR32(H) IDCT_VAR8(H)
#define IDCT(H) \ #define IDCT(H) \
static void FUNC(idct_ ## H ## x ## H )(int16_t *coeffs, \ static void FUNC(idct_ ## H ## x ## H )(int16_t *coeffs, \
int col_limit) { \ int col_limit) \
{ \
int i; \ int i; \
int shift = 7; \ int shift = 7; \
int add = 1 << (shift - 1); \ int add = 1 << (shift - 1); \
int16_t *src = coeffs; \ int16_t *src = coeffs; \
IDCT_VAR ##H(H); \ IDCT_VAR ## H(H); \
\ \
for (i = 0; i < H; i++) { \ for (i = 0; i < H; i++) { \
TR_ ## H(src, src, H, H, SCALE, limit2); \ TR_ ## H(src, src, H, H, SCALE, limit2); \