openh264/test/encoder/EncUT_Sample.cpp
Sindre Aamås 48a520915a [Encoder/x86] Add AVX2 SATD routines
WelsSampleSatd16x16_avx2 (~2.31x speedup over SSE4.1 on Haswell).
WelsSampleSatd16x8_avx2  (~2.19x speedup over SSE4.1 on Haswell).
WelsSampleSatd8x16_avx2  (~1.68x speedup over SSE4.1 on Haswell).
WelsSampleSatd8x8_avx2   (~1.53x speedup over SSE4.1 on Haswell).
2016-03-08 11:31:17 +01:00

721 lines
27 KiB
C++

#include <gtest/gtest.h>
#include <math.h>
#include "cpu_core.h"
#include "cpu.h"
#include "sample.h"
#include "sad_common.h"
#include "get_intra_predictor.h"
using namespace WelsEnc;
#define GENERATE_Intra16x16_UT(func, ref, ASM, CPUFLAGS) \
TEST (IntraSadSatdFuncTest, func) { \
const int32_t iLineSizeDec = 32; \
const int32_t iLineSizeEnc = 32; \
int32_t tmpa, tmpb; \
int32_t iBestMode_c, iBestMode_a, iLambda = 50; \
if (ASM) {\
int32_t iCpuCores = 0; \
uint32_t m_uiCpuFeatureFlag = WelsCPUFeatureDetect (&iCpuCores); \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
} \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDec, iLineSizeDec << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pEnc, iLineSizeEnc << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDst, 512, 16); \
for (int i = 0; i < (iLineSizeDec << 5); i++) \
pDec[i] = rand() % 256; \
for (int i = 0; i < (iLineSizeEnc << 5); i++) \
pEnc[i] = rand() % 256; \
for (int i = 0; i < 512; i++) \
pDst[i] = rand() % 256; \
tmpa = ref (pDec + 128, iLineSizeDec, pEnc, iLineSizeEnc, &iBestMode_c, iLambda, pDst); \
tmpb = func (pDec + 128, iLineSizeDec, pEnc, iLineSizeEnc, &iBestMode_a, iLambda, pDst); \
ASSERT_EQ (tmpa, tmpb); \
ASSERT_EQ (iBestMode_c, iBestMode_a); \
}
#define GENERATE_Intra4x4_UT(func, ASM, CPUFLAGS) \
TEST (IntraSadSatdFuncTest, func) { \
const int32_t iLineSizeDec = 32; \
const int32_t iLineSizeEnc = 32; \
int32_t tmpa, tmpb; \
int32_t iBestMode_c, iBestMode_a, iLambda = 50; \
int32_t lambda[2] = {iLambda << 2, iLambda}; \
int32_t iPredMode = rand() % 3; \
if (ASM) {\
int32_t iCpuCores = 0; \
uint32_t m_uiCpuFeatureFlag = WelsCPUFeatureDetect (&iCpuCores); \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
} \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDec, iLineSizeDec << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pEnc, iLineSizeEnc << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDst, 512, 16); \
for (int i = 0; i < (iLineSizeDec << 5); i++) \
pDec[i] = rand() % 256; \
for (int i = 0; i < (iLineSizeEnc << 5); i++) \
pEnc[i] = rand() % 256; \
for (int i = 0; i < 512; i++) \
pDst[i] = rand() % 256; \
tmpa = WelsSampleSatdIntra4x4Combined3_c (pDec + 128, iLineSizeDec, pEnc, iLineSizeEnc, pDst, &iBestMode_c, \
lambda[iPredMode == 2], lambda[iPredMode == 1], lambda[iPredMode == 0]); \
tmpb = func (pDec + 128, iLineSizeDec, pEnc, iLineSizeEnc, pDst, &iBestMode_a, \
lambda[iPredMode == 2], lambda[iPredMode == 1], lambda[iPredMode == 0]); \
ASSERT_EQ (tmpa, tmpb); \
ASSERT_EQ (iBestMode_c, iBestMode_a); \
}
#define GENERATE_Intra8x8_UT(func, ref, ASM, CPUFLAGS) \
TEST (IntraSadSatdFuncTest, func) { \
const int32_t iLineSizeDec = 32; \
const int32_t iLineSizeEnc = 32; \
int32_t tmpa, tmpb; \
int32_t iBestMode_c, iBestMode_a, iLambda = 50; \
if (ASM) {\
int32_t iCpuCores = 0; \
uint32_t m_uiCpuFeatureFlag = WelsCPUFeatureDetect (&iCpuCores); \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
} \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDecCb, iLineSizeDec << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pEncCb, iLineSizeEnc << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDecCr, iLineSizeDec << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pEncCr, iLineSizeEnc << 5, 16); \
ENFORCE_STACK_ALIGN_1D (uint8_t, pDstChma, 512, 16); \
for (int i = 0; i < (iLineSizeDec << 5); i++) { \
pDecCb[i] = rand() % 256; \
pDecCr[i] = rand() % 256; \
} \
for (int i = 0; i < (iLineSizeEnc << 5); i++) { \
pEncCb[i] = rand() % 256; \
pEncCr[i] = rand() % 256; \
} \
for (int i = 0; i < 512; i++) \
pDstChma[i] = rand() % 256; \
tmpa = ref (pDecCb + 128, iLineSizeDec, pEncCb, iLineSizeEnc, &iBestMode_c, iLambda, \
pDstChma, pDecCr + 128, pEncCr); \
tmpb = func (pDecCb + 128, iLineSizeDec, pEncCb, iLineSizeEnc, &iBestMode_a, iLambda, \
pDstChma, pDecCr + 128, pEncCr); \
ASSERT_EQ (tmpa, tmpb); \
ASSERT_EQ (iBestMode_c, iBestMode_a); \
}
#ifdef X86_ASM
GENERATE_Intra16x16_UT (WelsIntra16x16Combined3Sad_ssse3, WelsSampleSadIntra16x16Combined3_c, 1, WELS_CPU_SSSE3)
GENERATE_Intra16x16_UT (WelsIntra16x16Combined3Satd_sse41, WelsSampleSatdIntra16x16Combined3_c, 1, WELS_CPU_SSE41)
GENERATE_Intra8x8_UT (WelsIntraChroma8x8Combined3Satd_sse41, WelsSampleSatdIntra8x8Combined3_c, 1, WELS_CPU_SSE41)
GENERATE_Intra4x4_UT (WelsSampleSatdThree4x4_sse2, 1, WELS_CPU_SSE2)
#endif
#ifdef HAVE_NEON
GENERATE_Intra16x16_UT (WelsIntra16x16Combined3Sad_neon, WelsSampleSadIntra16x16Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra16x16_UT (WelsIntra16x16Combined3Satd_neon, WelsSampleSatdIntra16x16Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra8x8_UT (WelsIntra8x8Combined3Satd_neon, WelsSampleSatdIntra8x8Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra8x8_UT (WelsIntra8x8Combined3Sad_neon, WelsSampleSadIntra8x8Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra4x4_UT (WelsIntra4x4Combined3Satd_neon, 1, WELS_CPU_NEON)
#endif
#ifdef HAVE_NEON_AARCH64
GENERATE_Intra16x16_UT (WelsIntra16x16Combined3Sad_AArch64_neon, WelsSampleSadIntra16x16Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra16x16_UT (WelsIntra16x16Combined3Satd_AArch64_neon, WelsSampleSatdIntra16x16Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra8x8_UT (WelsIntra8x8Combined3Satd_AArch64_neon, WelsSampleSatdIntra8x8Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra8x8_UT (WelsIntra8x8Combined3Sad_AArch64_neon, WelsSampleSadIntra8x8Combined3_c, 1, WELS_CPU_NEON)
GENERATE_Intra4x4_UT (WelsIntra4x4Combined3Satd_AArch64_neon, 1, WELS_CPU_NEON)
#endif
#define ASSERT_MEMORY_FAIL2X(A, B) \
if (NULL == B) { \
pMemAlign->WelsFree(A, "Sad_SrcA");\
ASSERT_TRUE(0); \
}
#define ASSERT_MEMORY_FAIL3X(A, B, C) \
if (NULL == C) { \
pMemAlign->WelsFree(A, "Sad_SrcA"); \
pMemAlign->WelsFree(B, "Sad_SrcB"); \
ASSERT_TRUE(0); \
}
#define PIXEL_STRIDE 32
class SadSatdCFuncTest : public testing::Test {
public:
virtual void SetUp() {
pMemAlign = new CMemoryAlign (0);
m_iStrideA = rand() % 256 + PIXEL_STRIDE;
m_iStrideB = rand() % 256 + PIXEL_STRIDE;
m_pPixSrcA = (uint8_t*)pMemAlign->WelsMalloc (m_iStrideA << 5, "Sad_m_pPixSrcA");
ASSERT_TRUE (NULL != m_pPixSrcA);
m_pPixSrcB = (uint8_t*)pMemAlign->WelsMalloc (m_iStrideB << 5, "Sad_m_pPixSrcB");
ASSERT_MEMORY_FAIL2X (m_pPixSrcA, m_pPixSrcB)
m_pSad = (int32_t*)pMemAlign->WelsMalloc (4 * sizeof (int32_t), "m_pSad");
ASSERT_MEMORY_FAIL3X (m_pPixSrcA, m_pPixSrcB, m_pSad)
}
virtual void TearDown() {
pMemAlign->WelsFree (m_pPixSrcA, "Sad_m_pPixSrcA");
pMemAlign->WelsFree (m_pPixSrcB, "Sad_m_pPixSrcB");
pMemAlign->WelsFree (m_pSad, "m_pSad");
delete pMemAlign;
}
public:
uint8_t* m_pPixSrcA;
uint8_t* m_pPixSrcB;
int32_t m_iStrideA;
int32_t m_iStrideB;
int32_t* m_pSad;
CMemoryAlign* pMemAlign;
};
TEST_F (SadSatdCFuncTest, WelsSampleSad4x4_c) {
for (int i = 0; i < (m_iStrideA << 2); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 2); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 4; i++) {
for (int j = 0; j < 4; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad4x4_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSad8x4_c) {
for (int i = 0; i < (m_iStrideA << 2); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 2); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 4; i++) {
for (int j = 0; j < 8; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad8x4_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSad4x8_c) {
for (int i = 0; i < (m_iStrideA << 3); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 3); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 4; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad4x8_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSad8x8_c) {
for (int i = 0; i < (m_iStrideA << 3); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 3); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 8; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad8x8_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSad16x8_c) {
for (int i = 0; i < (m_iStrideA << 3); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 3); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 16; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad16x8_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSad8x16_c) {
for (int i = 0; i < (m_iStrideA << 4); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 4); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 16; i++) {
for (int j = 0; j < 8; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad8x16_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSad16x16_c) {
for (int i = 0; i < (m_iStrideA << 4); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 4); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t iSumSad = 0;
for (int i = 0; i < 16; i++) {
for (int j = 0; j < 16; j++)
iSumSad += abs (pPixA[j] - pPixB[j]);
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
EXPECT_EQ (WelsSampleSad16x16_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSatd4x4_c) {
for (int i = 0; i < (m_iStrideA << 2); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 2); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB;
int32_t W[16], T[16], Y[16], k = 0;
for (int i = 0; i < 4; i++) {
for (int j = 0; j < 4; j++)
W[k++] = pPixA[j] - pPixB[j];
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
T[0] = W[0] + W[4] + W[8] + W[12];
T[1] = W[1] + W[5] + W[9] + W[13];
T[2] = W[2] + W[6] + W[10] + W[14];
T[3] = W[3] + W[7] + W[11] + W[15];
T[4] = W[0] + W[4] - W[8] - W[12];
T[5] = W[1] + W[5] - W[9] - W[13];
T[6] = W[2] + W[6] - W[10] - W[14];
T[7] = W[3] + W[7] - W[11] - W[15];
T[8] = W[0] - W[4] - W[8] + W[12];
T[9] = W[1] - W[5] - W[9] + W[13];
T[10] = W[2] - W[6] - W[10] + W[14];
T[11] = W[3] - W[7] - W[11] + W[15];
T[12] = W[0] - W[4] + W[8] - W[12];
T[13] = W[1] - W[5] + W[9] - W[13];
T[14] = W[2] - W[6] + W[10] - W[14];
T[15] = W[3] - W[7] + W[11] - W[15];
Y[0] = T[0] + T[1] + T[2] + T[3];
Y[1] = T[0] + T[1] - T[2] - T[3];
Y[2] = T[0] - T[1] - T[2] + T[3];
Y[3] = T[0] - T[1] + T[2] - T[3];
Y[4] = T[4] + T[5] + T[6] + T[7];
Y[5] = T[4] + T[5] - T[6] - T[7];
Y[6] = T[4] - T[5] - T[6] + T[7];
Y[7] = T[4] - T[5] + T[6] - T[7];
Y[8] = T[8] + T[9] + T[10] + T[11];
Y[9] = T[8] + T[9] - T[10] - T[11];
Y[10] = T[8] - T[9] - T[10] + T[11];
Y[11] = T[8] - T[9] + T[10] - T[11];
Y[12] = T[12] + T[13] + T[14] + T[15];
Y[13] = T[12] + T[13] - T[14] - T[15];
Y[14] = T[12] - T[13] - T[14] + T[15];
Y[15] = T[12] - T[13] + T[14] - T[15];
int32_t iSumSatd = 0;
for (int i = 0; i < 16; i++)
iSumSatd += abs (Y[i]);
EXPECT_EQ (WelsSampleSatd4x4_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), (iSumSatd + 1) >> 1);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour16x16_c) {
for (int i = 0; i < (m_iStrideA << 5); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 5); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 16; i++) {
for (int j = 0; j < 16; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour16x16_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour16x8_c) {
for (int i = 0; i < (m_iStrideA << 5); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 5); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 16; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour16x8_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour8x16_c) {
for (int i = 0; i < (m_iStrideA << 5); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 5); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 16; i++) {
for (int j = 0; j < 8; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour8x16_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour8x8_c) {
for (int i = 0; i < (m_iStrideA << 4); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 4); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 8; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour8x8_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour4x4_c) {
for (int i = 0; i < (m_iStrideA << 3); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 3); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 4; i++) {
for (int j = 0; j < 4; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour4x4_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour8x4_c) {
for (int i = 0; i < (m_iStrideA << 3); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 3); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 4; i++) {
for (int j = 0; j < 8; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour8x4_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
TEST_F (SadSatdCFuncTest, WelsSampleSadFour4x8_c) {
for (int i = 0; i < (m_iStrideA << 4); i++)
m_pPixSrcA[i] = rand() % 256;
for (int i = 0; i < (m_iStrideB << 4); i++)
m_pPixSrcB[i] = rand() % 256;
uint8_t* pPixA = m_pPixSrcA;
uint8_t* pPixB = m_pPixSrcB + m_iStrideB;
int32_t iSumSad = 0;
for (int i = 0; i < 8; i++) {
for (int j = 0; j < 4; j++) {
iSumSad += abs (pPixA[j] - pPixB[j - 1]);
iSumSad += abs (pPixA[j] - pPixB[j + 1]);
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]);
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]);
}
pPixA += m_iStrideA;
pPixB += m_iStrideB;
}
WelsSampleSadFour4x8_c (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad);
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad);
}
class SadSatdAssemblyFuncTest : public testing::Test {
public:
virtual void SetUp() {
int32_t iCpuCores = 0;
m_uiCpuFeatureFlag = WelsCPUFeatureDetect (&iCpuCores);
pMemAlign = new CMemoryAlign (16);
m_iStrideA = m_iStrideB = PIXEL_STRIDE;
m_pPixSrcA = (uint8_t*)pMemAlign->WelsMalloc (m_iStrideA << 5, "Sad_m_pPixSrcA");
ASSERT_TRUE (NULL != m_pPixSrcA);
m_pPixSrcB = (uint8_t*)pMemAlign->WelsMalloc (m_iStrideB << 5, "Sad_m_pPixSrcB");
ASSERT_MEMORY_FAIL2X (m_pPixSrcA, m_pPixSrcB)
m_pSad = (int32_t*)pMemAlign->WelsMalloc (4 * sizeof (int32_t), "m_pSad");
ASSERT_MEMORY_FAIL3X (m_pPixSrcA, m_pPixSrcB, m_pSad)
}
virtual void TearDown() {
pMemAlign->WelsFree (m_pPixSrcA, "Sad_m_pPixSrcA");
pMemAlign->WelsFree (m_pPixSrcB, "Sad_m_pPixSrcB");
pMemAlign->WelsFree (m_pSad, "m_pSad");
delete pMemAlign;
}
public:
uint32_t m_uiCpuFeatureFlag;
uint8_t* m_pPixSrcA;
uint8_t* m_pPixSrcB;
int32_t m_iStrideA;
int32_t m_iStrideB;
int32_t* m_pSad;
CMemoryAlign* pMemAlign;
};
#define GENERATE_Sad4x4_UT(func, ref, CPUFLAGS) \
TEST_F (SadSatdAssemblyFuncTest, func) { \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
for (int i = 0; i < (m_iStrideA << 2); i++) \
m_pPixSrcA[i] = rand() % 256; \
for (int i = 0; i < (m_iStrideB << 2); i++) \
m_pPixSrcB[i] = rand() % 256; \
EXPECT_EQ (ref (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), func (m_pPixSrcA, \
m_iStrideA, m_pPixSrcB, m_iStrideB)); \
}
#define GENERATE_Sad8x8_UT(func, ref, CPUFLAGS) \
TEST_F (SadSatdAssemblyFuncTest, func) { \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
for (int i = 0; i < (m_iStrideA << 3); i++) \
m_pPixSrcA[i] = rand() % 256; \
for (int i = 0; i < (m_iStrideB << 3); i++) \
m_pPixSrcB[i] = rand() % 256; \
EXPECT_EQ (ref (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), func (m_pPixSrcA, \
m_iStrideA, m_pPixSrcB, m_iStrideB)); \
}
#define GENERATE_Sad8x16_UT(func, ref, CPUFLAGS) \
TEST_F (SadSatdAssemblyFuncTest, func) { \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
for (int i = 0; i < (m_iStrideA << 4); i++) \
m_pPixSrcA[i] = rand() % 256; \
for (int i = 0; i < (m_iStrideB << 4); i++) \
m_pPixSrcB[i] = rand() % 256; \
EXPECT_EQ (ref (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), func (m_pPixSrcA, \
m_iStrideA, m_pPixSrcB, m_iStrideB)); \
}
#define GENERATE_Sad16x8_UT(func, ref, CPUFLAGS) \
TEST_F (SadSatdAssemblyFuncTest, func) { \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
for (int i = 0; i < (m_iStrideA << 3); i++) \
m_pPixSrcA[i] = rand() % 256; \
for (int i = 0; i < (m_iStrideB << 3); i++) \
m_pPixSrcB[i] = rand() % 256; \
EXPECT_EQ (ref (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), func (m_pPixSrcA, \
m_iStrideA, m_pPixSrcB, m_iStrideB)); \
}
#define GENERATE_Sad16x16_UT(func, ref, CPUFLAGS) \
TEST_F (SadSatdAssemblyFuncTest, func) { \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
for (int i = 0; i < (m_iStrideA << 4); i++) \
m_pPixSrcA[i] = rand() % 256; \
for (int i = 0; i < (m_iStrideB << 4); i++) \
m_pPixSrcB[i] = rand() % 256; \
EXPECT_EQ (ref (m_pPixSrcA, m_iStrideA, m_pPixSrcB, m_iStrideB), func (m_pPixSrcA, \
m_iStrideA, m_pPixSrcB, m_iStrideB)); \
}
#ifdef X86_ASM
GENERATE_Sad4x4_UT (WelsSampleSad4x4_mmx, WelsSampleSad4x4_c, WELS_CPU_MMXEXT)
GENERATE_Sad8x8_UT (WelsSampleSad8x8_sse21, WelsSampleSad8x8_c, WELS_CPU_SSE2)
GENERATE_Sad8x16_UT (WelsSampleSad8x16_sse2, WelsSampleSad8x16_c, WELS_CPU_SSE2)
GENERATE_Sad16x8_UT (WelsSampleSad16x8_sse2, WelsSampleSad16x8_c, WELS_CPU_SSE2)
GENERATE_Sad16x16_UT (WelsSampleSad16x16_sse2, WelsSampleSad16x16_c, WELS_CPU_SSE2)
GENERATE_Sad4x4_UT (WelsSampleSatd4x4_sse2, WelsSampleSatd4x4_c, WELS_CPU_SSE2)
GENERATE_Sad8x8_UT (WelsSampleSatd8x8_sse2, WelsSampleSatd8x8_c, WELS_CPU_SSE2)
GENERATE_Sad8x16_UT (WelsSampleSatd8x16_sse2, WelsSampleSatd8x16_c, WELS_CPU_SSE2)
GENERATE_Sad16x8_UT (WelsSampleSatd16x8_sse2, WelsSampleSatd16x8_c, WELS_CPU_SSE2)
GENERATE_Sad16x16_UT (WelsSampleSatd16x16_sse2, WelsSampleSatd16x16_c, WELS_CPU_SSE2)
GENERATE_Sad4x4_UT (WelsSampleSatd4x4_sse41, WelsSampleSatd4x4_c, WELS_CPU_SSE41)
GENERATE_Sad8x8_UT (WelsSampleSatd8x8_sse41, WelsSampleSatd8x8_c, WELS_CPU_SSE41)
GENERATE_Sad8x16_UT (WelsSampleSatd8x16_sse41, WelsSampleSatd8x16_c, WELS_CPU_SSE41)
GENERATE_Sad16x8_UT (WelsSampleSatd16x8_sse41, WelsSampleSatd16x8_c, WELS_CPU_SSE41)
GENERATE_Sad16x16_UT (WelsSampleSatd16x16_sse41, WelsSampleSatd16x16_c, WELS_CPU_SSE41)
GENERATE_Sad8x8_UT (WelsSampleSatd8x8_avx2, WelsSampleSatd8x8_c, WELS_CPU_AVX2)
GENERATE_Sad8x16_UT (WelsSampleSatd8x16_avx2, WelsSampleSatd8x16_c, WELS_CPU_AVX2)
GENERATE_Sad16x8_UT (WelsSampleSatd16x8_avx2, WelsSampleSatd16x8_c, WELS_CPU_AVX2)
GENERATE_Sad16x16_UT (WelsSampleSatd16x16_avx2, WelsSampleSatd16x16_c, WELS_CPU_AVX2)
#endif
#ifdef HAVE_NEON
GENERATE_Sad4x4_UT (WelsSampleSad4x4_neon, WelsSampleSad4x4_c, WELS_CPU_NEON)
GENERATE_Sad8x8_UT (WelsSampleSad8x8_neon, WelsSampleSad8x8_c, WELS_CPU_NEON)
GENERATE_Sad8x16_UT (WelsSampleSad8x16_neon, WelsSampleSad8x16_c, WELS_CPU_NEON)
GENERATE_Sad16x8_UT (WelsSampleSad16x8_neon, WelsSampleSad16x8_c, WELS_CPU_NEON)
GENERATE_Sad16x16_UT (WelsSampleSad16x16_neon, WelsSampleSad16x16_c, WELS_CPU_NEON)
GENERATE_Sad4x4_UT (WelsSampleSatd4x4_neon, WelsSampleSatd4x4_c, WELS_CPU_NEON)
GENERATE_Sad8x8_UT (WelsSampleSatd8x8_neon, WelsSampleSatd8x8_c, WELS_CPU_NEON)
GENERATE_Sad8x16_UT (WelsSampleSatd8x16_neon, WelsSampleSatd8x16_c, WELS_CPU_NEON)
GENERATE_Sad16x8_UT (WelsSampleSatd16x8_neon, WelsSampleSatd16x8_c, WELS_CPU_NEON)
GENERATE_Sad16x16_UT (WelsSampleSatd16x16_neon, WelsSampleSatd16x16_c, WELS_CPU_NEON)
#endif
#ifdef HAVE_NEON_AARCH64
GENERATE_Sad4x4_UT (WelsSampleSad4x4_AArch64_neon, WelsSampleSad4x4_c, WELS_CPU_NEON)
GENERATE_Sad8x8_UT (WelsSampleSad8x8_AArch64_neon, WelsSampleSad8x8_c, WELS_CPU_NEON)
GENERATE_Sad8x16_UT (WelsSampleSad8x16_AArch64_neon, WelsSampleSad8x16_c, WELS_CPU_NEON)
GENERATE_Sad16x8_UT (WelsSampleSad16x8_AArch64_neon, WelsSampleSad16x8_c, WELS_CPU_NEON)
GENERATE_Sad16x16_UT (WelsSampleSad16x16_AArch64_neon, WelsSampleSad16x16_c, WELS_CPU_NEON)
GENERATE_Sad4x4_UT (WelsSampleSatd4x4_AArch64_neon, WelsSampleSatd4x4_c, WELS_CPU_NEON)
GENERATE_Sad8x8_UT (WelsSampleSatd8x8_AArch64_neon, WelsSampleSatd8x8_c, WELS_CPU_NEON)
GENERATE_Sad8x16_UT (WelsSampleSatd8x16_AArch64_neon, WelsSampleSatd8x16_c, WELS_CPU_NEON)
GENERATE_Sad16x8_UT (WelsSampleSatd16x8_AArch64_neon, WelsSampleSatd16x8_c, WELS_CPU_NEON)
GENERATE_Sad16x16_UT (WelsSampleSatd16x16_AArch64_neon, WelsSampleSatd16x16_c, WELS_CPU_NEON)
#endif
#define GENERATE_SadFour_UT(func, CPUFLAGS, width, height) \
TEST_F (SadSatdAssemblyFuncTest, func) { \
if (0 == (m_uiCpuFeatureFlag & CPUFLAGS)) \
return; \
for (int i = 0; i < (m_iStrideA << 5); i++) \
m_pPixSrcA[i] = rand() % 256; \
for (int i = 0; i < (m_iStrideB << 5); i++) \
m_pPixSrcB[i] = rand() % 256; \
uint8_t* pPixA = m_pPixSrcA; \
uint8_t* pPixB = m_pPixSrcB + m_iStrideB; \
int32_t iSumSad = 0; \
for (int i = 0; i < height; i++) { \
for (int j = 0; j < width; j++) { \
iSumSad += abs (pPixA[j] - pPixB[j - 1]); \
iSumSad += abs (pPixA[j] - pPixB[j + 1]); \
iSumSad += abs (pPixA[j] - pPixB[j - m_iStrideB]); \
iSumSad += abs (pPixA[j] - pPixB[j + m_iStrideB]); \
} \
pPixA += m_iStrideA; \
pPixB += m_iStrideB; \
} \
func (m_pPixSrcA, m_iStrideA, m_pPixSrcB + m_iStrideB, m_iStrideB, m_pSad); \
EXPECT_EQ (m_pSad[0] + m_pSad[1] + m_pSad[2] + m_pSad[3], iSumSad); \
}
#ifdef X86_ASM
GENERATE_SadFour_UT (WelsSampleSadFour4x4_sse2, WELS_CPU_SSE2, 4, 4)
GENERATE_SadFour_UT (WelsSampleSadFour8x8_sse2, WELS_CPU_SSE2, 8, 8)
GENERATE_SadFour_UT (WelsSampleSadFour8x16_sse2, WELS_CPU_SSE2, 8, 16)
GENERATE_SadFour_UT (WelsSampleSadFour16x8_sse2, WELS_CPU_SSE2, 16, 8)
GENERATE_SadFour_UT (WelsSampleSadFour16x16_sse2, WELS_CPU_SSE2, 16, 16)
#endif
#ifdef HAVE_NEON
GENERATE_SadFour_UT (WelsSampleSadFour4x4_neon, WELS_CPU_NEON, 4, 4)
GENERATE_SadFour_UT (WelsSampleSadFour8x8_neon, WELS_CPU_NEON, 8, 8)
GENERATE_SadFour_UT (WelsSampleSadFour8x16_neon, WELS_CPU_NEON, 8, 16)
GENERATE_SadFour_UT (WelsSampleSadFour16x8_neon, WELS_CPU_NEON, 16, 8)
GENERATE_SadFour_UT (WelsSampleSadFour16x16_neon, WELS_CPU_NEON, 16, 16)
#endif
#ifdef HAVE_NEON_AARCH64
GENERATE_SadFour_UT (WelsSampleSadFour4x4_AArch64_neon, WELS_CPU_NEON, 4, 4)
GENERATE_SadFour_UT (WelsSampleSadFour8x8_AArch64_neon, WELS_CPU_NEON, 8, 8)
GENERATE_SadFour_UT (WelsSampleSadFour8x16_AArch64_neon, WELS_CPU_NEON, 8, 16)
GENERATE_SadFour_UT (WelsSampleSadFour16x8_AArch64_neon, WELS_CPU_NEON, 16, 8)
GENERATE_SadFour_UT (WelsSampleSadFour16x16_AArch64_neon, WELS_CPU_NEON, 16, 16)
#endif