Skip to content
This repository was archived by the owner on Jul 29, 2024. It is now read-only.

Conflicts: Fix conflicts with SVT-AV1 #429

Merged
merged 2 commits into from
Dec 10, 2019
+2,594 −2,621
Merged

Conflicts: Fix conflicts with SVT-AV1 #429

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
898ff1a
conflicts: Fix conflicts with SVT-AV1
1480c1 Nov 15, 2019
4981843
cleanup: PascalCase
1480c1 Nov 26, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions Source/App/EbTime.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,13 +6,13 @@
#ifndef EbTime_h
#define EbTime_h

void EbStartTime(uint64_t *Startseconds, uint64_t *Startuseconds);
void EbFinishTime(uint64_t *Finishseconds, uint64_t *Finishuseconds);
void EbComputeOverallElapsedTime(uint64_t Startseconds, uint64_t Startuseconds,uint64_t Finishseconds, uint64_t Finishuseconds, double *duration);
void EbHevcStartTime(uint64_t *Startseconds, uint64_t *Startuseconds);
void EbHevcFinishTime(uint64_t *Finishseconds, uint64_t *Finishuseconds);
void EbHevcComputeOverallElapsedTime(uint64_t Startseconds, uint64_t Startuseconds,uint64_t Finishseconds, uint64_t Finishuseconds, double *duration);
void EbAppStartTime(uint64_t *Startseconds, uint64_t *Startuseconds);
void EbAppFinishTime(uint64_t *Finishseconds, uint64_t *Finishuseconds);
void EbAppComputeOverallElapsedTime(uint64_t Startseconds, uint64_t Startuseconds,uint64_t Finishseconds, uint64_t Finishuseconds, double *duration);
void EbComputeOverallElapsedTimeMs(uint64_t Startseconds, uint64_t Startuseconds, uint64_t Finishseconds, uint64_t Finishuseconds, double *duration);
void EbHevcComputeOverallElapsedTimeMs(uint64_t Startseconds, uint64_t Startuseconds, uint64_t Finishseconds, uint64_t Finishuseconds, double *duration);
void EbSleep(uint64_t milliSeconds);
void EbInjector(uint64_t processedFrameCount, uint32_t injectorFrameRate);

Expand Down
9 changes: 4 additions & 5 deletions Source/Lib/ASM_AVX2/EbMCP16bit_Intrinsic_AVX2.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,7 @@
#include "EbMcp_SSE2.h"
#include "EbDefinitions.h"

EB_EXTERN EB_ALIGN(16) const EB_S16 chromaFilterCoeffSR1_AVX[8][4] =
EB_EXTERN EB_ALIGN(16) const EB_S16 EbHevcChromaFilterCoeffSR1_AVX[8][4] =
{
{ 0, 32, 0, 0 },
{ -1, 29, 5, -1 },
Expand Down Expand Up @@ -40,7 +40,7 @@ void ChromaInterpolationFilterTwoD16bit_AVX2_INTRIN(
ChromaInterpolationFilterOneDOutRaw16bitHorizontal_AVX2_INTRIN(refPic - ((MaxChromaFilterTag - 1) >> 1)*srcStride, srcStride, firstPassIFDst, puWidth, puHeight + MaxChromaFilterTag - 1, (EB_S16 *)EB_NULL, fracPosx, 0);
#endif


//vertical filtering
ChromaInterpolationFilterTwoDInRaw16bit_SSE2_INTRIN(firstPassIFDst, dst, dstStride, puWidth, puHeight, fracPosy);
}
Expand All @@ -62,7 +62,7 @@ void ChromaInterpolationFilterTwoDOutRaw16bit_AVX2_INTRIN(
//on-the-fly scheme
ChromaInterpolationFilterOneDOutRaw16bitHorizontal_AVX2_INTRIN(refPic - ((MaxChromaFilterTag - 1) >> 1)*srcStride, srcStride, firstPassIFDst, puWidth, puHeight + MaxChromaFilterTag - 1, (EB_S16 *)EB_NULL, fracPosx, 0);
#endif

//vertical filtering
ChromaInterpolationFilterTwoDInRawOutRaw_SSE2_INTRIN(firstPassIFDst, dst, puWidth, puHeight, fracPosy);
}
Expand Down Expand Up @@ -92,7 +92,7 @@ void ChromaInterpolationFilterOneDOutRaw16bitHorizontal_AVX2_INTRIN(
refPic--;
//PrefetchBlock(refPic, srcStride, puWidth+8, puHeight);

c0 = _mm_loadl_epi64((__m128i *)chromaFilterCoeffSR1_AVX[fracPosx]);
c0 = _mm_loadl_epi64((__m128i *)EbHevcChromaFilterCoeffSR1_AVX[fracPosx]);
c0 = _mm_unpacklo_epi16(c0, c0);
c3 = _mm_shuffle_epi32(c0, 0xff);
c2 = _mm_shuffle_epi32(c0, 0xaa);
Expand Down Expand Up @@ -267,4 +267,3 @@ void ChromaInterpolationFilterOneDOutRaw16bitHorizontal_AVX2_INTRIN(
}

}

14 changes: 7 additions & 7 deletions Source/Lib/ASM_AVX2/EbNoiseExtractAVX2.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,11 +9,11 @@
#include "immintrin.h"
#include "EbUtility.h"

EB_EXTERN EB_ALIGN(16) const EB_U8 filterType[] = {
EB_EXTERN EB_ALIGN(16) const EB_U8 EbHevcFilterType[] = {
1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4, 1, 4
};

EB_EXTERN EB_ALIGN(16) const EB_U8 WeakChromafilter[2][32] = {
EB_EXTERN EB_ALIGN(16) const EB_U8 EbHevcWeakChromafilter[2][32] = {
{ 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4, 2, 4 },
{ 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2 },
};
Expand All @@ -39,7 +39,7 @@ inline void lumaWeakFilter_AVX2_INTRIN(
currPrevPermutation = _mm256_permute4x64_epi64(currPrev, 216);
currPermutation = _mm256_permute4x64_epi64(curr, 216);
currLeftMidFirstHalflo = _mm256_unpacklo_epi8(currPrevPermutation, currPermutation);
weights = _mm256_loadu_si256((__m256i*)filterType);
weights = _mm256_loadu_si256((__m256i*)EbHevcFilterType);
currLeftMidFirstHalfWeight = _mm256_maddubs_epi16(currLeftMidFirstHalflo, weights);
currNextPermutation = _mm256_permute4x64_epi64(currNext, 88);
currNextFirstHalf = _mm256_unpacklo_epi8(currNextPermutation, _mm256_setzero_si256());
Expand Down Expand Up @@ -100,7 +100,7 @@ inline void chromaWeakLumaStrongFilter_AVX2_INTRIN(
currPrevPermutation = _mm256_permute4x64_epi64(currPrev, 216);
currPermutation = _mm256_permute4x64_epi64(curr, 216);
currLeftMidFirstHalflo = _mm256_unpacklo_epi8(currPrevPermutation, currPermutation);
weights = _mm256_loadu_si256((__m256i*)WeakChromafilter[0]);
weights = _mm256_loadu_si256((__m256i*)EbHevcWeakChromafilter[0]);
currLeftMidFirstHalfWeight = _mm256_maddubs_epi16(currLeftMidFirstHalflo, weights);
currNextPermutation = _mm256_permute4x64_epi64(currNext, 88);
currNextFirstHalf = _mm256_unpacklo_epi8(currNextPermutation, _mm256_setzero_si256());
Expand All @@ -118,7 +118,7 @@ inline void chromaWeakLumaStrongFilter_AVX2_INTRIN(
topPrevPermutation = _mm256_permute4x64_epi64(topPrev, 216);
topPermutation = _mm256_permute4x64_epi64(top, 216);
topLeftMidFirstHalflo = _mm256_unpacklo_epi8(topPrevPermutation, topPermutation);
weights = _mm256_loadu_si256((__m256i*)WeakChromafilter[1]);
weights = _mm256_loadu_si256((__m256i*)EbHevcWeakChromafilter[1]);
topLeftMidFirstHalfWeight = _mm256_maddubs_epi16(topLeftMidFirstHalflo, weights);
topNextPermutation = _mm256_permute4x64_epi64(topNext, 88);
topNextFirstHalf = _mm256_unpacklo_epi8(topNextPermutation, _mm256_setzero_si256());
Expand All @@ -135,7 +135,7 @@ inline void chromaWeakLumaStrongFilter_AVX2_INTRIN(
bottomPrevPermutation = _mm256_permute4x64_epi64(bottomPrev, 216);
bottomPermutation = _mm256_permute4x64_epi64(bottom, 216);
bottomLeftMidFirstHalflo = _mm256_unpacklo_epi8(bottomPrevPermutation, bottomPermutation);
weights = _mm256_loadu_si256((__m256i*)WeakChromafilter[1]);
weights = _mm256_loadu_si256((__m256i*)EbHevcWeakChromafilter[1]);
bottomLeftMidFirstHalfWeight = _mm256_maddubs_epi16(bottomLeftMidFirstHalflo, weights);
bottomNextPermutation = _mm256_permute4x64_epi64(bottomNext, 88);
bottomNextFirstHalf = _mm256_unpacklo_epi8(bottomNextPermutation, _mm256_setzero_si256());
Expand Down Expand Up @@ -305,7 +305,7 @@ void noiseExtractLumaWeak_AVX2_INTRIN(
// p[1 + 2 * stride]) / 8;

top = curr = secondtop = secondcurr = _mm256_setzero_si256();

for (kk = 0; kk + MAX_LCU_SIZE <= picWidth; kk += MAX_LCU_SIZE)
{
for (jj = 0; jj < lcuHeight; jj++)
Expand Down
Loading