generic-library/openh264
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Convert tabs to spaces in code that looks like tables

Browse Source 
Also fix the alignment in some related cases, even though they
didn't use tabs originally.
This commit is contained in:
Martin Storsjö
2015-05-15 11:08:40 +03:00
parent 8c60f84f7b
commit b052a9580e
8 changed files with 74 additions and 78 deletions
Download Patch File Download Diff File Expand all files Collapse all files

36
codec/common/src/copy_mb.cpp
View File

@@ -51,8 +51,8 @@ void WelsCopy4x4_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStr
const int32_t kiDstStride2 = iStrideD << 1; const int32_t kiDstStride2 = iStrideD << 1;
const int32_t kiDstStride3 = iStrideD + kiDstStride2; const int32_t kiDstStride3 = iStrideD + kiDstStride2;
ST32 (pDst, LD32 (pSrc)); ST32 (pDst, LD32 (pSrc));
ST32 (pDst + iStrideD, LD32 (pSrc + iStrideS)); ST32 (pDst + iStrideD, LD32 (pSrc + iStrideS));
ST32 (pDst + kiDstStride2, LD32 (pSrc + kiSrcStride2)); ST32 (pDst + kiDstStride2, LD32 (pSrc + kiSrcStride2));
ST32 (pDst + kiDstStride3, LD32 (pSrc + kiSrcStride3)); ST32 (pDst + kiDstStride3, LD32 (pSrc + kiSrcStride3));
} }
@@ -67,10 +67,10 @@ void WelsCopy4x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStr
void WelsCopy8x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) { void WelsCopy8x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) {
int32_t i; int32_t i;
for (i = 0; i < 4; i++) { for (i = 0; i < 4; i++) {
ST32 (pDst, LD32 (pSrc)); ST32 (pDst, LD32 (pSrc));
ST32 (pDst + 4 , LD32 (pSrc + 4)); ST32 (pDst + 4 , LD32 (pSrc + 4));
ST32 (pDst + iStrideD, LD32 (pSrc + iStrideS)); ST32 (pDst + iStrideD, LD32 (pSrc + iStrideS));
ST32 (pDst + iStrideD + 4 , LD32 (pSrc + iStrideS + 4)); ST32 (pDst + iStrideD + 4 , LD32 (pSrc + iStrideS + 4));
pDst += iStrideD << 1; pDst += iStrideD << 1;
pSrc += iStrideS << 1; pSrc += iStrideS << 1;
} }
@@ -78,10 +78,10 @@ void WelsCopy8x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStr
void WelsCopy8x16_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) { void WelsCopy8x16_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) {
int32_t i; int32_t i;
for (i = 0; i < 8; ++i) { for (i = 0; i < 8; ++i) {
ST32 (pDst, LD32 (pSrc)); ST32 (pDst, LD32 (pSrc));
ST32 (pDst + 4 , LD32 (pSrc + 4)); ST32 (pDst + 4 , LD32 (pSrc + 4));
ST32 (pDst + iStrideD, LD32 (pSrc + iStrideS)); ST32 (pDst + iStrideD, LD32 (pSrc + iStrideS));
ST32 (pDst + iStrideD + 4 , LD32 (pSrc + iStrideS + 4)); ST32 (pDst + iStrideD + 4 , LD32 (pSrc + iStrideS + 4));
pDst += iStrideD << 1; pDst += iStrideD << 1;
pSrc += iStrideS << 1; pSrc += iStrideS << 1;
} }
@@ -89,10 +89,10 @@ void WelsCopy8x16_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iSt
void WelsCopy16x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) { void WelsCopy16x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) {
int32_t i; int32_t i;
for (i = 0; i < 8; i++) { for (i = 0; i < 8; i++) {
ST32 (pDst, LD32 (pSrc)); ST32 (pDst, LD32 (pSrc));
ST32 (pDst + 4 , LD32 (pSrc + 4)); ST32 (pDst + 4 , LD32 (pSrc + 4));
ST32 (pDst + 8 , LD32 (pSrc + 8)); ST32 (pDst + 8 , LD32 (pSrc + 8));
ST32 (pDst + 12 , LD32 (pSrc + 12)); ST32 (pDst + 12 , LD32 (pSrc + 12));
pDst += iStrideD ; pDst += iStrideD ;
pSrc += iStrideS; pSrc += iStrideS;
} }
@@ -100,10 +100,10 @@ void WelsCopy16x8_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iSt
void WelsCopy16x16_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) { void WelsCopy16x16_c (uint8_t* pDst, int32_t iStrideD, uint8_t* pSrc, int32_t iStrideS) {
int32_t i; int32_t i;
for (i = 0; i < 16; i++) { for (i = 0; i < 16; i++) {
ST32 (pDst, LD32 (pSrc)); ST32 (pDst, LD32 (pSrc));
ST32 (pDst + 4 , LD32 (pSrc + 4)); ST32 (pDst + 4 , LD32 (pSrc + 4));
ST32 (pDst + 8 , LD32 (pSrc + 8)); ST32 (pDst + 8 , LD32 (pSrc + 8));
ST32 (pDst + 12 , LD32 (pSrc + 12)); ST32 (pDst + 12 , LD32 (pSrc + 12));
pDst += iStrideD ; pDst += iStrideD ;
pSrc += iStrideS; pSrc += iStrideS;
} }

38
codec/decoder/core/src/get_intra_predictor.cpp
View File

@@ -54,10 +54,10 @@ namespace WelsDec {
void WelsI4x4LumaPredV_c (uint8_t* pPred, const int32_t kiStride) { void WelsI4x4LumaPredV_c (uint8_t* pPred, const int32_t kiStride) {
const uint32_t kuiVal = LD32A4 (pPred - kiStride); const uint32_t kuiVal = LD32A4 (pPred - kiStride);
ST32A4 (pPred , kuiVal); ST32A4 (pPred , kuiVal);
ST32A4 (pPred + kiStride , kuiVal); ST32A4 (pPred + kiStride , kuiVal);
ST32A4 (pPred + (kiStride << 1) , kuiVal); ST32A4 (pPred + (kiStride << 1) , kuiVal);
ST32A4 (pPred + (kiStride << 1) + kiStride , kuiVal); ST32A4 (pPred + (kiStride << 1) + kiStride , kuiVal);
} }
void WelsI4x4LumaPredH_c (uint8_t* pPred, const int32_t kiStride) { void WelsI4x4LumaPredH_c (uint8_t* pPred, const int32_t kiStride) {
@@ -68,7 +68,7 @@ void WelsI4x4LumaPredH_c (uint8_t* pPred, const int32_t kiStride) {
const uint32_t kuiL2 = 0x01010101U * pPred[-1 + kiStride2]; const uint32_t kuiL2 = 0x01010101U * pPred[-1 + kiStride2];
const uint32_t kuiL3 = 0x01010101U * pPred[-1 + kiStride3]; const uint32_t kuiL3 = 0x01010101U * pPred[-1 + kiStride3];
ST32A4 (pPred , kuiL0); ST32A4 (pPred , kuiL0);
ST32A4 (pPred + kiStride , kuiL1); ST32A4 (pPred + kiStride , kuiL1);
ST32A4 (pPred + kiStride2, kuiL2); ST32A4 (pPred + kiStride2, kuiL2);
ST32A4 (pPred + kiStride3, kuiL3); ST32A4 (pPred + kiStride3, kuiL3);
@@ -81,7 +81,7 @@ void WelsI4x4LumaPredDc_c (uint8_t* pPred, const int32_t kiStride) {
pPred[-kiStride] + pPred[-kiStride + 1] + pPred[-kiStride + 2] + pPred[-kiStride + 3] + 4) >> 3; pPred[-kiStride] + pPred[-kiStride + 1] + pPred[-kiStride + 2] + pPred[-kiStride + 3] + 4) >> 3;
const uint32_t kuiMean32 = 0x01010101U * kuiMean; const uint32_t kuiMean32 = 0x01010101U * kuiMean;
ST32A4 (pPred , kuiMean32); ST32A4 (pPred , kuiMean32);
ST32A4 (pPred + kiStride , kuiMean32); ST32A4 (pPred + kiStride , kuiMean32);
ST32A4 (pPred + kiStride2, kuiMean32); ST32A4 (pPred + kiStride2, kuiMean32);
ST32A4 (pPred + kiStride3, kuiMean32); ST32A4 (pPred + kiStride3, kuiMean32);
@@ -93,7 +93,7 @@ void WelsI4x4LumaPredDcLeft_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiMean = (pPred[-1] + pPred[-1 + kiStride] + pPred[-1 + kiStride2] + pPred[-1 + kiStride3] + 2) >> 2; const uint8_t kuiMean = (pPred[-1] + pPred[-1 + kiStride] + pPred[-1 + kiStride2] + pPred[-1 + kiStride3] + 2) >> 2;
const uint32_t kuiMean32 = 0x01010101U * kuiMean; const uint32_t kuiMean32 = 0x01010101U * kuiMean;
ST32A4 (pPred , kuiMean32); ST32A4 (pPred , kuiMean32);
ST32A4 (pPred + kiStride , kuiMean32); ST32A4 (pPred + kiStride , kuiMean32);
ST32A4 (pPred + kiStride2, kuiMean32); ST32A4 (pPred + kiStride2, kuiMean32);
ST32A4 (pPred + kiStride3, kuiMean32); ST32A4 (pPred + kiStride3, kuiMean32);
@@ -106,7 +106,7 @@ void WelsI4x4LumaPredDcTop_c (uint8_t* pPred, const int32_t kiStride) {
2; 2;
const uint32_t kuiMean32 = 0x01010101U * kuiMean; const uint32_t kuiMean32 = 0x01010101U * kuiMean;
ST32A4 (pPred , kuiMean32); ST32A4 (pPred , kuiMean32);
ST32A4 (pPred + kiStride , kuiMean32); ST32A4 (pPred + kiStride , kuiMean32);
ST32A4 (pPred + kiStride2, kuiMean32); ST32A4 (pPred + kiStride2, kuiMean32);
ST32A4 (pPred + kiStride3, kuiMean32); ST32A4 (pPred + kiStride3, kuiMean32);
@@ -115,9 +115,9 @@ void WelsI4x4LumaPredDcTop_c (uint8_t* pPred, const int32_t kiStride) {
void WelsI4x4LumaPredDcNA_c (uint8_t* pPred, const int32_t kiStride) { void WelsI4x4LumaPredDcNA_c (uint8_t* pPred, const int32_t kiStride) {
const uint32_t kuiDC32 = 0x80808080U; const uint32_t kuiDC32 = 0x80808080U;
ST32A4 (pPred , kuiDC32); ST32A4 (pPred , kuiDC32);
ST32A4 (pPred + kiStride , kuiDC32); ST32A4 (pPred + kiStride , kuiDC32);
ST32A4 (pPred + (kiStride << 1) , kuiDC32); ST32A4 (pPred + (kiStride << 1) , kuiDC32);
ST32A4 (pPred + (kiStride << 1) + kiStride, kuiDC32); ST32A4 (pPred + (kiStride << 1) + kiStride, kuiDC32);
} }
@@ -144,7 +144,7 @@ void WelsI4x4LumaPredDDL_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiDDL6 = (2 + kuiT6 + kuiT7 + (kuiT7 << 1)) >> 2; // kDDL6 const uint8_t kuiDDL6 = (2 + kuiT6 + kuiT7 + (kuiT7 << 1)) >> 2; // kDDL6
const uint8_t kuiList[8] = { kuiDDL0, kuiDDL1, kuiDDL2, kuiDDL3, kuiDDL4, kuiDDL5, kuiDDL6, 0 }; const uint8_t kuiList[8] = { kuiDDL0, kuiDDL1, kuiDDL2, kuiDDL3, kuiDDL4, kuiDDL5, kuiDDL6, 0 };
ST32A4 (pPred , LD32 (kuiList)); ST32A4 (pPred , LD32 (kuiList));
ST32A4 (pPred + kiStride , LD32 (kuiList + 1)); ST32A4 (pPred + kiStride , LD32 (kuiList + 1));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 2)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 2));
ST32A4 (pPred + kiStride3, LD32 (kuiList + 3)); ST32A4 (pPred + kiStride3, LD32 (kuiList + 3));
@@ -170,7 +170,7 @@ void WelsI4x4LumaPredDDLTop_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiDLT3 = kuiT33 >> 1; // kDLT3 const uint8_t kuiDLT3 = kuiT33 >> 1; // kDLT3
const uint8_t kuiList[8] = { kuiDLT0, kuiDLT1, kuiDLT2, kuiDLT3, kuiDLT3, kuiDLT3, kuiDLT3 , kuiDLT3 }; const uint8_t kuiList[8] = { kuiDLT0, kuiDLT1, kuiDLT2, kuiDLT3, kuiDLT3, kuiDLT3, kuiDLT3 , kuiDLT3 };
ST32A4 (pPred, LD32 (kuiList)); ST32A4 (pPred, LD32 (kuiList));
ST32A4 (pPred + kiStride, LD32 (kuiList + 1)); ST32A4 (pPred + kiStride, LD32 (kuiList + 1));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 2)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 2));
ST32A4 (pPred + kiStride3, LD32 (kuiList + 3)); ST32A4 (pPred + kiStride3, LD32 (kuiList + 3));
@@ -210,7 +210,7 @@ void WelsI4x4LumaPredDDR_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiDDR6 = (kuiL12 + kuiL23) >> 2; // kuiDDR6 const uint8_t kuiDDR6 = (kuiL12 + kuiL23) >> 2; // kuiDDR6
const uint8_t kuiList[8] = { kuiDDR6, kuiDDR5, kuiDDR4, kuiDDR0, kuiDDR1, kuiDDR2, kuiDDR3, 0 }; const uint8_t kuiList[8] = { kuiDDR6, kuiDDR5, kuiDDR4, kuiDDR0, kuiDDR1, kuiDDR2, kuiDDR3, 0 };
ST32A4 (pPred , LD32 (kuiList + 3)); ST32A4 (pPred , LD32 (kuiList + 3));
ST32A4 (pPred + kiStride , LD32 (kuiList + 2)); ST32A4 (pPred + kiStride , LD32 (kuiList + 2));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 1)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 1));
ST32A4 (pPred + kiStride3, LD32 (kuiList)); ST32A4 (pPred + kiStride3, LD32 (kuiList));
@@ -248,7 +248,7 @@ void WelsI4x4LumaPredVL_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiVL9 = (kuiT45 + kuiT56) >> 2; // kuiVL9 const uint8_t kuiVL9 = (kuiT45 + kuiT56) >> 2; // kuiVL9
const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL8, kuiVL9 }; const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL8, kuiVL9 };
ST32A4 (pPred, LD32 (kuiList)); ST32A4 (pPred, LD32 (kuiList));
ST32A4 (pPred + kiStride, LD32 (kuiList + 5)); ST32A4 (pPred + kiStride, LD32 (kuiList + 5));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 1)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 1));
ST32A4 (pPred + kiStride3, LD32 (kuiList + 6)); ST32A4 (pPred + kiStride3, LD32 (kuiList + 6));
@@ -278,7 +278,7 @@ void WelsI4x4LumaPredVLTop_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiVL7 = kuiVL3; const uint8_t kuiVL7 = kuiVL3;
const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL7 }; const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL7 };
ST32A4 (pPred , LD32 (kuiList)); ST32A4 (pPred , LD32 (kuiList));
ST32A4 (pPred + kiStride , LD32 (kuiList + 5)); ST32A4 (pPred + kiStride , LD32 (kuiList + 5));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 1)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 1));
ST32A4 (pPred + kiStride3, LD32 (kuiList + 6)); ST32A4 (pPred + kiStride3, LD32 (kuiList + 6));
@@ -310,7 +310,7 @@ void WelsI4x4LumaPredVR_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiVR9 = (2 + kuiL0 + (kuiL1 << 1) + kuiL2) >> 2; // kuiVR9 const uint8_t kuiVR9 = (2 + kuiL0 + (kuiL1 << 1) + kuiL2) >> 2; // kuiVR9
const uint8_t kuiList[10] = { kuiVR8, kuiVR0, kuiVR1, kuiVR2, kuiVR3, kuiVR9, kuiVR4, kuiVR5, kuiVR6, kuiVR7 }; const uint8_t kuiList[10] = { kuiVR8, kuiVR0, kuiVR1, kuiVR2, kuiVR3, kuiVR9, kuiVR4, kuiVR5, kuiVR6, kuiVR7 };
ST32A4 (pPred , LD32 (kuiList + 1)); ST32A4 (pPred , LD32 (kuiList + 1));
ST32A4 (pPred + kiStride , LD32 (kuiList + 6)); ST32A4 (pPred + kiStride , LD32 (kuiList + 6));
ST32A4 (pPred + kiStride2, LD32 (kuiList)); ST32A4 (pPred + kiStride2, LD32 (kuiList));
ST32A4 (pPred + kiStride3, LD32 (kuiList + 5)); ST32A4 (pPred + kiStride3, LD32 (kuiList + 5));
@@ -336,7 +336,7 @@ void WelsI4x4LumaPredHU_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiHU5 = (1 + kuiL23 + (kuiL3 << 1)) >> 2; const uint8_t kuiHU5 = (1 + kuiL23 + (kuiL3 << 1)) >> 2;
const uint8_t kuiList[10] = { kuiHU0, kuiHU1, kuiHU2, kuiHU3, kuiHU4, kuiHU5, kuiL3, kuiL3, kuiL3, kuiL3 }; const uint8_t kuiList[10] = { kuiHU0, kuiHU1, kuiHU2, kuiHU3, kuiHU4, kuiHU5, kuiL3, kuiL3, kuiL3, kuiL3 };
ST32A4 (pPred , LD32 (kuiList)); ST32A4 (pPred , LD32 (kuiList));
ST32A4 (pPred + kiStride , LD32 (kuiList + 2)); ST32A4 (pPred + kiStride , LD32 (kuiList + 2));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 4)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 4));
ST32A4 (pPred + kiStride3, LD32 (kuiList + 6)); ST32A4 (pPred + kiStride3, LD32 (kuiList + 6));
@@ -374,7 +374,7 @@ void WelsI4x4LumaPredHD_c (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiHD9 = (kuiL12 + kuiL23) >> 2; const uint8_t kuiHD9 = (kuiL12 + kuiL23) >> 2;
const uint8_t kuiList[10] = { kuiHD8, kuiHD9, kuiHD6, kuiHD7, kuiHD4, kuiHD5, kuiHD0, kuiHD1, kuiHD2, kuiHD3 }; const uint8_t kuiList[10] = { kuiHD8, kuiHD9, kuiHD6, kuiHD7, kuiHD4, kuiHD5, kuiHD0, kuiHD1, kuiHD2, kuiHD3 };
ST32A4 (pPred , LD32 (kuiList + 6)); ST32A4 (pPred , LD32 (kuiList + 6));
ST32A4 (pPred + kiStride , LD32 (kuiList + 4)); ST32A4 (pPred + kiStride , LD32 (kuiList + 4));
ST32A4 (pPred + kiStride2, LD32 (kuiList + 2)); ST32A4 (pPred + kiStride2, LD32 (kuiList + 2));
ST32A4 (pPred + kiStride3, LD32 (kuiList)); ST32A4 (pPred + kiStride3, LD32 (kuiList));

12
codec/encoder/core/src/decode_mb_aux.cpp
View File

@@ -189,8 +189,8 @@ void WelsIDctT4Rec_c (uint8_t* pRec, int32_t iStride, uint8_t* pPred, int32_t iP
const int32_t kiVerDelR = (iTemp[4 + i] >> 1) - iTemp[12 + i]; const int32_t kiVerDelR = (iTemp[4 + i] >> 1) - iTemp[12 + i];
const int32_t kiVerSumR = iTemp[4 + i] + (iTemp[12 + i] >> 1); const int32_t kiVerSumR = iTemp[4 + i] + (iTemp[12 + i] >> 1);
pRec[i ] = WelsClip1 (pPred[i ] + ((kiVerSumL + kiVerSumR + 32) >> 6)); pRec[i ] = WelsClip1 (pPred[i ] + ((kiVerSumL + kiVerSumR + 32) >> 6));
pRec[iStride + i ] = WelsClip1 (pPred[iPredStride + i ] + ((kiVerDelL + kiVerDelR + 32) >> 6)); pRec[iStride + i ] = WelsClip1 (pPred[iPredStride + i ] + ((kiVerDelL + kiVerDelR + 32) >> 6));
pRec[iDstStridex2 + i] = WelsClip1 (pPred[iPredStridex2 + i] + ((kiVerDelL - kiVerDelR + 32) >> 6)); pRec[iDstStridex2 + i] = WelsClip1 (pPred[iPredStridex2 + i] + ((kiVerDelL - kiVerDelR + 32) >> 6));
pRec[iDstStridex3 + i] = WelsClip1 (pPred[iPredStridex3 + i] + ((kiVerSumL - kiVerSumR + 32) >> 6)); pRec[iDstStridex3 + i] = WelsClip1 (pPred[iPredStridex3 + i] + ((kiVerSumL - kiVerSumR + 32) >> 6));
} }
@@ -199,10 +199,10 @@ void WelsIDctT4Rec_c (uint8_t* pRec, int32_t iStride, uint8_t* pPred, int32_t iP
void WelsIDctFourT4Rec_c (uint8_t* pRec, int32_t iStride, uint8_t* pPred, int32_t iPredStride, int16_t* pDct) { void WelsIDctFourT4Rec_c (uint8_t* pRec, int32_t iStride, uint8_t* pPred, int32_t iPredStride, int16_t* pDct) {
int32_t iDstStridex4 = iStride << 2; int32_t iDstStridex4 = iStride << 2;
int32_t iPredStridex4 = iPredStride << 2; int32_t iPredStridex4 = iPredStride << 2;
WelsIDctT4Rec_c (pRec, iStride, pPred, iPredStride, pDct); WelsIDctT4Rec_c (pRec, iStride, pPred, iPredStride, pDct);
WelsIDctT4Rec_c (&pRec[4], iStride, &pPred[4], iPredStride, pDct + 16); WelsIDctT4Rec_c (&pRec[4], iStride, &pPred[4], iPredStride, pDct + 16);
WelsIDctT4Rec_c (&pRec[iDstStridex4 ], iStride, &pPred[iPredStridex4 ], iPredStride, pDct + 32); WelsIDctT4Rec_c (&pRec[iDstStridex4 ], iStride, &pPred[iPredStridex4 ], iPredStride, pDct + 32);
WelsIDctT4Rec_c (&pRec[iDstStridex4 + 4], iStride, &pPred[iPredStridex4 + 4], iPredStride, pDct + 48); WelsIDctT4Rec_c (&pRec[iDstStridex4 + 4], iStride, &pPred[iPredStridex4 + 4], iPredStride, pDct + 48);
} }

14
codec/encoder/core/src/md.cpp
View File

@@ -446,15 +446,11 @@ int32_t AnalysisVaaInfoIntra_c (uint8_t* pDataY, const int32_t kiLineSize) {
for (; j < 16; j += 4) { for (; j < 16; j += 4) {
num = 0; num = 0;
for (i = 0; i < 16; i += 4, num ++) { for (i = 0; i < 16; i += 4, num ++) {
pBlock[num] = pEncData[i ] + pEncData[i + 1 ] + pEncData[i + 2 ] + pEncData[i + pBlock[num] = pEncData[i ] + pEncData[i + 1 ] + pEncData[i + 2 ] + pEncData[i + 3 ];
3 ]; pBlock[num] += pEncData[i + kiLineSize ] + pEncData[i + kiLineSize + 1] + pEncData[i + kiLineSize + 2] + pEncData[i + kiLineSize + 3];
pBlock[num] += pEncData[i + kiLineSize ] + pEncData[i + kiLineSize + 1 ] + pEncData[i + kiLineSize + 2 ] + pEncData[i + pBlock[num] += pEncData[i + kiLineSize2] + pEncData[i + kiLineSize2 + 1] + pEncData[i + kiLineSize2 + 2] + pEncData[i + kiLineSize2 + 3];
kiLineSize + 3 ]; pBlock[num] += pEncData[i + kiLineSize3] + pEncData[i + kiLineSize3 + 1] + pEncData[i + kiLineSize3 + 2] + pEncData[i + kiLineSize3 + 3];
pBlock[num] += pEncData[i + kiLineSize2] + pEncData[i + kiLineSize2 + 1] + pEncData[i + kiLineSize2 + 2] + pEncData[i + pBlock[num] >>= 4;
kiLineSize2 + 3];
pBlock[num] += pEncData[i + kiLineSize3] + pEncData[i + kiLineSize3 + 1] + pEncData[i + kiLineSize3 + 2] + pEncData[i +
kiLineSize3 + 3];
pBlock[num] >>= 4;
} }
pBlock += 4; pBlock += 4;
pEncData += kiLineSize4; pEncData += kiLineSize4;

26
codec/encoder/core/src/svc_encode_mb.cpp
View File

@@ -45,10 +45,10 @@
namespace WelsEnc { namespace WelsEnc {
void WelsDctMb (int16_t* pRes, uint8_t* pEncMb, int32_t iEncStride, uint8_t* pBestPred, PDctFunc pfDctFourT4) { void WelsDctMb (int16_t* pRes, uint8_t* pEncMb, int32_t iEncStride, uint8_t* pBestPred, PDctFunc pfDctFourT4) {
pfDctFourT4 (pRes, pEncMb, iEncStride, pBestPred, 16); pfDctFourT4 (pRes, pEncMb, iEncStride, pBestPred, 16);
pfDctFourT4 (pRes + 64, pEncMb + 8, iEncStride, pBestPred + 8, 16); pfDctFourT4 (pRes + 64, pEncMb + 8, iEncStride, pBestPred + 8, 16);
pfDctFourT4 (pRes + 128, pEncMb + 8 * iEncStride, iEncStride, pBestPred + 128, 16); pfDctFourT4 (pRes + 128, pEncMb + 8 * iEncStride, iEncStride, pBestPred + 128, 16);
pfDctFourT4 (pRes + 192, pEncMb + 8 * iEncStride + 8, iEncStride, pBestPred + 136, 16); pfDctFourT4 (pRes + 192, pEncMb + 8 * iEncStride + 8, iEncStride, pBestPred + 136, 16);
} }
void WelsEncRecI16x16Y (sWelsEncCtx* pEncCtx, SMB* pCurMb, SMbCache* pMbCache) { void WelsEncRecI16x16Y (sWelsEncCtx* pEncCtx, SMB* pCurMb, SMbCache* pMbCache) {
@@ -77,7 +77,7 @@ void WelsEncRecI16x16Y (sWelsEncCtx* pEncCtx, SMB* pCurMb, SMbCache* pMbCache) {
for (i = 0; i < 4; i++) { for (i = 0; i < 4; i++) {
pFuncList->pfQuantizationFour4x4 (pRes, pFF, pMF); pFuncList->pfQuantizationFour4x4 (pRes, pFF, pMF);
pFuncList->pfScan4x4Ac (pBlock, pRes); pFuncList->pfScan4x4Ac (pBlock, pRes);
pFuncList->pfScan4x4Ac (pBlock + 16, pRes + 16); pFuncList->pfScan4x4Ac (pBlock + 16, pRes + 16);
pFuncList->pfScan4x4Ac (pBlock + 32, pRes + 32); pFuncList->pfScan4x4Ac (pBlock + 32, pRes + 32);
pFuncList->pfScan4x4Ac (pBlock + 48, pRes + 48); pFuncList->pfScan4x4Ac (pBlock + 48, pRes + 48);
@@ -126,12 +126,12 @@ void WelsEncRecI16x16Y (sWelsEncCtx* pEncCtx, SMB* pCurMb, SMbCache* pMbCache) {
pRes[224] = aDctT4Dc[14]; pRes[224] = aDctT4Dc[14];
pRes[240] = aDctT4Dc[15]; pRes[240] = aDctT4Dc[15];
pFuncList->pfIDctFourT4 (pPred, kiRecStride, pBestPred, 16, pRes); pFuncList->pfIDctFourT4 (pPred, kiRecStride, pBestPred, 16, pRes);
pFuncList->pfIDctFourT4 (pPred + 8, kiRecStride, pBestPred + 8, 16, pRes + 64); pFuncList->pfIDctFourT4 (pPred + 8, kiRecStride, pBestPred + 8, 16, pRes + 64);
pFuncList->pfIDctFourT4 (pPred + kiRecStride * 8, kiRecStride, pBestPred + 128, 16, pRes + 128); pFuncList->pfIDctFourT4 (pPred + kiRecStride * 8, kiRecStride, pBestPred + 128, 16, pRes + 128);
pFuncList->pfIDctFourT4 (pPred + kiRecStride * 8 + 8, kiRecStride, pBestPred + 136, 16, pRes + 192); pFuncList->pfIDctFourT4 (pPred + kiRecStride * 8 + 8, kiRecStride, pBestPred + 136, 16, pRes + 192);
} else if (uiCountI16x16Dc > 0) { } else if (uiCountI16x16Dc > 0) {
pFuncList->pfIDctI16x16Dc (pPred, kiRecStride, pBestPred, 16, aDctT4Dc); pFuncList->pfIDctI16x16Dc (pPred, kiRecStride, pBestPred, 16, aDctT4Dc);
} else { } else {
pFuncList->pfCopy16x16Aligned (pPred, kiRecStride, pBestPred, 16); pFuncList->pfCopy16x16Aligned (pPred, kiRecStride, pBestPred, 16);
} }
@@ -319,10 +319,10 @@ void WelsRecPskip (SDqLayer* pCurLayer, SWelsFuncPtrList* pFuncList, SMB* pCu
int32_t* iRecStride = pCurLayer->iCsStride; int32_t* iRecStride = pCurLayer->iCsStride;
uint8_t** pCsMb = &pMbCache->SPicData.pCsMb[0]; uint8_t** pCsMb = &pMbCache->SPicData.pCsMb[0];
pFuncList->pfCopy16x16Aligned (pCsMb[0], *iRecStride++, pMbCache->pSkipMb, 16); pFuncList->pfCopy16x16Aligned (pCsMb[0], *iRecStride++, pMbCache->pSkipMb, 16);
pFuncList->pfCopy8x8Aligned (pCsMb[1], *iRecStride++, pMbCache->pSkipMb + 256, 8); pFuncList->pfCopy8x8Aligned (pCsMb[1], *iRecStride++, pMbCache->pSkipMb + 256, 8);
pFuncList->pfCopy8x8Aligned (pCsMb[2], *iRecStride, pMbCache->pSkipMb + 320, 8); pFuncList->pfCopy8x8Aligned (pCsMb[2], *iRecStride, pMbCache->pSkipMb + 320, 8);
pFuncList->pfSetMemZeroSize8 (pCurMb->pNonZeroCount, 24); pFuncList->pfSetMemZeroSize8 (pCurMb->pNonZeroCount, 24);
} }
bool WelsTryPYskip (sWelsEncCtx* pEncCtx, SMB* pCurMb, SMbCache* pMbCache) { bool WelsTryPYskip (sWelsEncCtx* pEncCtx, SMB* pCurMb, SMbCache* pMbCache) {

4
codec/encoder/core/src/svc_encode_slice.cpp
View File

@@ -448,8 +448,8 @@ void WelsPMbChromaEncode (sWelsEncCtx* pEncCtx, SSlice* pSlice, SMB* pCurMb) {
int16_t* pCurRS = pMbCache->pCoeffLevel + 256; int16_t* pCurRS = pMbCache->pCoeffLevel + 256;
uint8_t* pBestPred = pMbCache->pMemPredChroma; uint8_t* pBestPred = pMbCache->pMemPredChroma;
pFunc->pfDctFourT4 (pCurRS, pMbCache->SPicData.pEncMb[1], kiEncStride, pBestPred, 8); pFunc->pfDctFourT4 (pCurRS, pMbCache->SPicData.pEncMb[1], kiEncStride, pBestPred, 8);
pFunc->pfDctFourT4 (pCurRS + 64, pMbCache->SPicData.pEncMb[2], kiEncStride, pBestPred + 64, 8); pFunc->pfDctFourT4 (pCurRS + 64, pMbCache->SPicData.pEncMb[2], kiEncStride, pBestPred + 64, 8);
WelsEncRecUV (pFunc, pCurMb, pMbCache, pCurRS, 1); WelsEncRecUV (pFunc, pCurMb, pMbCache, pCurRS, 1);
WelsEncRecUV (pFunc, pCurMb, pMbCache, pCurRS + 64, 2); WelsEncRecUV (pFunc, pCurMb, pMbCache, pCurRS + 64, 2);

6
codec/processing/src/denoise/denoise_filter.cpp
View File

@@ -117,9 +117,9 @@ void Gauss3x3Filter (uint8_t* pSrc, int32_t iStride) {
uint8_t* pCurLine2 = pCurLine1 + iStride; uint8_t* pCurLine2 = pCurLine1 + iStride;
uint8_t* pCurLine3 = pCurLine2 + iStride; uint8_t* pCurLine3 = pCurLine2 + iStride;
nSum = pCurLine1[0] + (pCurLine1[1] << 1) + pCurLine1[2] + nSum = pCurLine1[0] + (pCurLine1[1] << 1) + pCurLine1[2] +
(pCurLine2[0] << 1) + (pCurLine2[1] << 2) + (pCurLine2[2] << 1) + (pCurLine2[0] << 1) + (pCurLine2[1] << 2) + (pCurLine2[2] << 1) +
pCurLine3[0] + (pCurLine3[1] << 1) + pCurLine3[2]; pCurLine3[0] + (pCurLine3[1] << 1) + pCurLine3[2];
*pSrc = nSum >> 4; *pSrc = nSum >> 4;
} }

16
test/decoder/DecUT_IntraPrediction.cpp
View File

@@ -114,7 +114,7 @@ void WelsI4x4LumaPredDDL_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiDDL6 = (2 + kuiT6 + kuiT7 + (kuiT7 << 1)) >> 2; // kDDL6 const uint8_t kuiDDL6 = (2 + kuiT6 + kuiT7 + (kuiT7 << 1)) >> 2; // kDDL6
const uint8_t kuiList[8] = { kuiDDL0, kuiDDL1, kuiDDL2, kuiDDL3, kuiDDL4, kuiDDL5, kuiDDL6, 0 }; const uint8_t kuiList[8] = { kuiDDL0, kuiDDL1, kuiDDL2, kuiDDL3, kuiDDL4, kuiDDL5, kuiDDL6, 0 };
ST32 (pPred , LD32 (kuiList)); ST32 (pPred , LD32 (kuiList));
ST32 (pPred + kiStride , LD32 (kuiList + 1)); ST32 (pPred + kiStride , LD32 (kuiList + 1));
ST32 (pPred + kiStride2, LD32 (kuiList + 2)); ST32 (pPred + kiStride2, LD32 (kuiList + 2));
ST32 (pPred + kiStride3, LD32 (kuiList + 3)); ST32 (pPred + kiStride3, LD32 (kuiList + 3));
@@ -140,7 +140,7 @@ void WelsI4x4LumaPredDDLTop_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiDLT3 = kuiT33 >> 1; // kDLT3 const uint8_t kuiDLT3 = kuiT33 >> 1; // kDLT3
const uint8_t kuiList[8] = { kuiDLT0, kuiDLT1, kuiDLT2, kuiDLT3, kuiDLT3, kuiDLT3, kuiDLT3 , kuiDLT3 }; const uint8_t kuiList[8] = { kuiDLT0, kuiDLT1, kuiDLT2, kuiDLT3, kuiDLT3, kuiDLT3, kuiDLT3 , kuiDLT3 };
ST32 (pPred, LD32 (kuiList)); ST32 (pPred, LD32 (kuiList));
ST32 (pPred + kiStride, LD32 (kuiList + 1)); ST32 (pPred + kiStride, LD32 (kuiList + 1));
ST32 (pPred + kiStride2, LD32 (kuiList + 2)); ST32 (pPred + kiStride2, LD32 (kuiList + 2));
ST32 (pPred + kiStride3, LD32 (kuiList + 3)); ST32 (pPred + kiStride3, LD32 (kuiList + 3));
@@ -180,7 +180,7 @@ void WelsI4x4LumaPredDDR_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiDDR6 = (kuiL12 + kuiL23) >> 2; // kuiDDR6 const uint8_t kuiDDR6 = (kuiL12 + kuiL23) >> 2; // kuiDDR6
const uint8_t kuiList[8] = { kuiDDR6, kuiDDR5, kuiDDR4, kuiDDR0, kuiDDR1, kuiDDR2, kuiDDR3, 0 }; const uint8_t kuiList[8] = { kuiDDR6, kuiDDR5, kuiDDR4, kuiDDR0, kuiDDR1, kuiDDR2, kuiDDR3, 0 };
ST32 (pPred , LD32 (kuiList + 3)); ST32 (pPred , LD32 (kuiList + 3));
ST32 (pPred + kiStride , LD32 (kuiList + 2)); ST32 (pPred + kiStride , LD32 (kuiList + 2));
ST32 (pPred + kiStride2, LD32 (kuiList + 1)); ST32 (pPred + kiStride2, LD32 (kuiList + 1));
ST32 (pPred + kiStride3, LD32 (kuiList)); ST32 (pPred + kiStride3, LD32 (kuiList));
@@ -218,7 +218,7 @@ void WelsI4x4LumaPredVL_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiVL9 = (kuiT45 + kuiT56) >> 2; // kuiVL9 const uint8_t kuiVL9 = (kuiT45 + kuiT56) >> 2; // kuiVL9
const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL8, kuiVL9 }; const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL8, kuiVL9 };
ST32 (pPred, LD32 (kuiList)); ST32 (pPred, LD32 (kuiList));
ST32 (pPred + kiStride, LD32 (kuiList + 5)); ST32 (pPred + kiStride, LD32 (kuiList + 5));
ST32 (pPred + kiStride2, LD32 (kuiList + 1)); ST32 (pPred + kiStride2, LD32 (kuiList + 1));
ST32 (pPred + kiStride3, LD32 (kuiList + 6)); ST32 (pPred + kiStride3, LD32 (kuiList + 6));
@@ -248,7 +248,7 @@ void WelsI4x4LumaPredVLTop_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiVL7 = kuiVL3; const uint8_t kuiVL7 = kuiVL3;
const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL7 }; const uint8_t kuiList[10] = { kuiVL0, kuiVL1, kuiVL2, kuiVL3, kuiVL3, kuiVL4, kuiVL5, kuiVL6, kuiVL7, kuiVL7 };
ST32 (pPred , LD32 (kuiList)); ST32 (pPred , LD32 (kuiList));
ST32 (pPred + kiStride , LD32 (kuiList + 5)); ST32 (pPred + kiStride , LD32 (kuiList + 5));
ST32 (pPred + kiStride2, LD32 (kuiList + 1)); ST32 (pPred + kiStride2, LD32 (kuiList + 1));
ST32 (pPred + kiStride3, LD32 (kuiList + 6)); ST32 (pPred + kiStride3, LD32 (kuiList + 6));
@@ -280,7 +280,7 @@ void WelsI4x4LumaPredVR_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiVR9 = (2 + kuiL0 + (kuiL1 << 1) + kuiL2) >> 2; // kuiVR9 const uint8_t kuiVR9 = (2 + kuiL0 + (kuiL1 << 1) + kuiL2) >> 2; // kuiVR9
const uint8_t kuiList[10] = { kuiVR8, kuiVR0, kuiVR1, kuiVR2, kuiVR3, kuiVR9, kuiVR4, kuiVR5, kuiVR6, kuiVR7 }; const uint8_t kuiList[10] = { kuiVR8, kuiVR0, kuiVR1, kuiVR2, kuiVR3, kuiVR9, kuiVR4, kuiVR5, kuiVR6, kuiVR7 };
ST32 (pPred , LD32 (kuiList + 1)); ST32 (pPred , LD32 (kuiList + 1));
ST32 (pPred + kiStride , LD32 (kuiList + 6)); ST32 (pPred + kiStride , LD32 (kuiList + 6));
ST32 (pPred + kiStride2, LD32 (kuiList)); ST32 (pPred + kiStride2, LD32 (kuiList));
ST32 (pPred + kiStride3, LD32 (kuiList + 5)); ST32 (pPred + kiStride3, LD32 (kuiList + 5));
@@ -306,7 +306,7 @@ void WelsI4x4LumaPredHU_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiHU5 = (1 + kuiL23 + (kuiL3 << 1)) >> 2; const uint8_t kuiHU5 = (1 + kuiL23 + (kuiL3 << 1)) >> 2;
const uint8_t kuiList[10] = { kuiHU0, kuiHU1, kuiHU2, kuiHU3, kuiHU4, kuiHU5, kuiL3, kuiL3, kuiL3, kuiL3 }; const uint8_t kuiList[10] = { kuiHU0, kuiHU1, kuiHU2, kuiHU3, kuiHU4, kuiHU5, kuiL3, kuiL3, kuiL3, kuiL3 };
ST32 (pPred , LD32 (kuiList)); ST32 (pPred , LD32 (kuiList));
ST32 (pPred + kiStride , LD32 (kuiList + 2)); ST32 (pPred + kiStride , LD32 (kuiList + 2));
ST32 (pPred + kiStride2, LD32 (kuiList + 4)); ST32 (pPred + kiStride2, LD32 (kuiList + 4));
ST32 (pPred + kiStride3, LD32 (kuiList + 6)); ST32 (pPred + kiStride3, LD32 (kuiList + 6));
@@ -344,7 +344,7 @@ void WelsI4x4LumaPredHD_ref (uint8_t* pPred, const int32_t kiStride) {
const uint8_t kuiHD9 = (kuiL12 + kuiL23) >> 2; const uint8_t kuiHD9 = (kuiL12 + kuiL23) >> 2;
const uint8_t kuiList[10] = { kuiHD8, kuiHD9, kuiHD6, kuiHD7, kuiHD4, kuiHD5, kuiHD0, kuiHD1, kuiHD2, kuiHD3 }; const uint8_t kuiList[10] = { kuiHD8, kuiHD9, kuiHD6, kuiHD7, kuiHD4, kuiHD5, kuiHD0, kuiHD1, kuiHD2, kuiHD3 };
ST32 (pPred , LD32 (kuiList + 6)); ST32 (pPred , LD32 (kuiList + 6));
ST32 (pPred + kiStride , LD32 (kuiList + 4)); ST32 (pPred + kiStride , LD32 (kuiList + 4));
ST32 (pPred + kiStride2, LD32 (kuiList + 2)); ST32 (pPred + kiStride2, LD32 (kuiList + 2));
ST32 (pPred + kiStride3, LD32 (kuiList)); ST32 (pPred + kiStride3, LD32 (kuiList));