Skip to content

GitLab

Commit 4ef1d35a authored by Jingning Han's avatar Jingning Han Committed by Gerrit Code Review
Browse files

Merge "Skip inverse transform when eob is zero"

parents c3871f8f 0359ad7f
Hide whitespace changes
Inline Side-by-side
Showing with 6 additions and 4 deletions
vp9/encoder/vp9_encodemb.c
View file @ 4ef1d35a
...@@ -533,6 +533,8 @@ static void encode_block(int plane, int block, BLOCK_SIZE_TYPE bsize, ...@@ -533,6 +533,8 @@ static void encode_block(int plane, int block, BLOCK_SIZE_TYPE bsize,
if (x->skip_encode) if (x->skip_encode)
return; return;
if (pd->eobs[block] == 0)
return;
switch (ss_txfrm_size / 2) { switch (ss_txfrm_size / 2) {
case TX_32X32: case TX_32X32:
...@@ -657,7 +659,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize, ...@@ -657,7 +659,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize,
vp9_quantize_b_32x32(coeff, 1024, x->skip_block, p->zbin, p->round, vp9_quantize_b_32x32(coeff, 1024, x->skip_block, p->zbin, p->round,
p->quant, p->quant_shift, qcoeff, dqcoeff, p->quant, p->quant_shift, qcoeff, dqcoeff,
pd->dequant, p->zbin_extra, eob, scan, iscan); pd->dequant, p->zbin_extra, eob, scan, iscan);
if (!x->skip_encode) if (!x->skip_encode && *eob)
vp9_short_idct32x32_add(dqcoeff, dst, pd->dst.stride); vp9_short_idct32x32_add(dqcoeff, dst, pd->dst.stride);
break; break;
case TX_16X16: case TX_16X16:
...@@ -682,7 +684,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize, ...@@ -682,7 +684,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize,
vp9_quantize_b(coeff, 256, x->skip_block, p->zbin, p->round, p->quant, vp9_quantize_b(coeff, 256, x->skip_block, p->zbin, p->round, p->quant,
p->quant_shift, qcoeff, dqcoeff, p->quant_shift, qcoeff, dqcoeff,
pd->dequant, p->zbin_extra, eob, scan, iscan); pd->dequant, p->zbin_extra, eob, scan, iscan);
if (!x->skip_encode) { if (!x->skip_encode && *eob) {
if (tx_type == DCT_DCT) if (tx_type == DCT_DCT)
vp9_short_idct16x16_add(dqcoeff, dst, pd->dst.stride); vp9_short_idct16x16_add(dqcoeff, dst, pd->dst.stride);
else else
...@@ -711,7 +713,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize, ...@@ -711,7 +713,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize,
vp9_quantize_b(coeff, 64, x->skip_block, p->zbin, p->round, p->quant, vp9_quantize_b(coeff, 64, x->skip_block, p->zbin, p->round, p->quant,
p->quant_shift, qcoeff, dqcoeff, p->quant_shift, qcoeff, dqcoeff,
pd->dequant, p->zbin_extra, eob, scan, iscan); pd->dequant, p->zbin_extra, eob, scan, iscan);
if (!x->skip_encode) { if (!x->skip_encode && *eob) {
if (tx_type == DCT_DCT) if (tx_type == DCT_DCT)
vp9_short_idct8x8_add(dqcoeff, dst, pd->dst.stride); vp9_short_idct8x8_add(dqcoeff, dst, pd->dst.stride);
else else
...@@ -743,7 +745,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize, ...@@ -743,7 +745,7 @@ void encode_block_intra(int plane, int block, BLOCK_SIZE_TYPE bsize,
vp9_quantize_b(coeff, 16, x->skip_block, p->zbin, p->round, p->quant, vp9_quantize_b(coeff, 16, x->skip_block, p->zbin, p->round, p->quant,
p->quant_shift, qcoeff, dqcoeff, p->quant_shift, qcoeff, dqcoeff,
pd->dequant, p->zbin_extra, eob, scan, iscan); pd->dequant, p->zbin_extra, eob, scan, iscan);
if (!x->skip_encode) { if (!x->skip_encode && *eob) {
if (tx_type == DCT_DCT) if (tx_type == DCT_DCT)
// this is like vp9_short_idct4x4 but has a special case around eob<=1 // this is like vp9_short_idct4x4 but has a special case around eob<=1
// which is significant (not just an optimization) for the lossless // which is significant (not just an optimization) for the lossless
......
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