Commit e84d14df authored by Ronald S. Bultje's avatar Ronald S. Bultje

vp9/x86: idct_32x32_add_ssse3.

Sub-IDCTs will follow later. ped1080.webm goes from 9.295s to 8.191s
(13.5% faster). The IDCT itself goes from 4372 (intra) or 4337 (inter)
to 403 (intra) or 329 (inter) cycles for the DC-only form, 23755 (intra)
or 23723 (inter) to 3497 (intra) or 3607 (inter) cycles for the no-DC
form, which averages from 23393 (intra) or 16612 (inter) to 3449 (intra)
or 2392 (inter) for all 32x32s together, i.e. about ~7x faster (all
tests done on ped1080p.webm).
parent b0517467
...@@ -159,6 +159,7 @@ filters_8tap_1d_fn3(avg) ...@@ -159,6 +159,7 @@ filters_8tap_1d_fn3(avg)
void ff_vp9_idct_idct_4x4_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob); void ff_vp9_idct_idct_4x4_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob);
void ff_vp9_idct_idct_8x8_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob); void ff_vp9_idct_idct_8x8_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob);
void ff_vp9_idct_idct_16x16_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob); void ff_vp9_idct_idct_16x16_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob);
void ff_vp9_idct_idct_32x32_add_ssse3(uint8_t *dst, ptrdiff_t stride, int16_t *block, int eob);
#endif /* HAVE_YASM */ #endif /* HAVE_YASM */
...@@ -217,6 +218,10 @@ av_cold void ff_vp9dsp_init_x86(VP9DSPContext *dsp) ...@@ -217,6 +218,10 @@ av_cold void ff_vp9dsp_init_x86(VP9DSPContext *dsp)
if (ARCH_X86_64) { if (ARCH_X86_64) {
dsp->itxfm_add[TX_8X8][DCT_DCT] = ff_vp9_idct_idct_8x8_add_ssse3; dsp->itxfm_add[TX_8X8][DCT_DCT] = ff_vp9_idct_idct_8x8_add_ssse3;
dsp->itxfm_add[TX_16X16][DCT_DCT] = ff_vp9_idct_idct_16x16_add_ssse3; dsp->itxfm_add[TX_16X16][DCT_DCT] = ff_vp9_idct_idct_16x16_add_ssse3;
dsp->itxfm_add[TX_32X32][ADST_ADST] =
dsp->itxfm_add[TX_32X32][ADST_DCT] =
dsp->itxfm_add[TX_32X32][DCT_ADST] =
dsp->itxfm_add[TX_32X32][DCT_DCT] = ff_vp9_idct_idct_32x32_add_ssse3;
} }
} }
......
This diff is collapsed.
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