ref: f5817fa612d39b3b46b3164c17ab9d5f3d7bc8c8
dir: /test/vp9_quantize_test.cc/
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <math.h>
#include <stdlib.h>
#include <string.h>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_config.h"
#include "./vpx_dsp_rtcd.h"
#include "./vp9_rtcd.h"
#include "test/acm_random.h"
#include "test/buffer.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "test/util.h"
#include "vp9/common/vp9_entropy.h"
#include "vp9/common/vp9_scan.h"
#include "vpx/vpx_codec.h"
#include "vpx/vpx_integer.h"
#include "vpx_ports/vpx_timer.h"
using libvpx_test::ACMRandom;
using libvpx_test::Buffer;
namespace {
const int number_of_iterations = 100;
typedef void (*QuantizeFunc)(const tran_low_t *coeff, intptr_t count,
int skip_block, const int16_t *zbin,
const int16_t *round, const int16_t *quant,
const int16_t *quant_shift, tran_low_t *qcoeff,
tran_low_t *dqcoeff, const int16_t *dequant,
uint16_t *eob, const int16_t *scan,
const int16_t *iscan);
typedef std::tr1::tuple<QuantizeFunc, QuantizeFunc, vpx_bit_depth_t,
int /*max_size*/>
QuantizeParam;
// Wrapper for FP version which does not use zbin or quant_shift.
typedef void (*QuantizeFPFunc)(const tran_low_t *coeff, intptr_t count,
int skip_block, const int16_t *round,
const int16_t *quant, tran_low_t *qcoeff,
tran_low_t *dqcoeff, const int16_t *dequant,
uint16_t *eob, const int16_t *scan,
const int16_t *iscan);
template <QuantizeFPFunc fn>
void QuantFPWrapper(const tran_low_t *coeff, intptr_t count, int skip_block,
const int16_t *zbin, const int16_t *round,
const int16_t *quant, const int16_t *quant_shift,
tran_low_t *qcoeff, tran_low_t *dqcoeff,
const int16_t *dequant, uint16_t *eob, const int16_t *scan,
const int16_t *iscan) {
(void)zbin;
(void)quant_shift;
fn(coeff, count, skip_block, round, quant, qcoeff, dqcoeff, dequant, eob,
scan, iscan);
}
class VP9QuantizeBase {
public:
VP9QuantizeBase(vpx_bit_depth_t bit_depth, int max_size)
: bit_depth_(bit_depth), max_size_(max_size) {
max_value_ = (1 << bit_depth_) - 1;
zbin_ptr_ =
reinterpret_cast<int16_t *>(vpx_memalign(16, 8 * sizeof(*zbin_ptr_)));
round_ptr_ =
reinterpret_cast<int16_t *>(vpx_memalign(16, 8 * sizeof(*round_ptr_)));
quant_ptr_ =
reinterpret_cast<int16_t *>(vpx_memalign(16, 8 * sizeof(*quant_ptr_)));
quant_shift_ptr_ = reinterpret_cast<int16_t *>(
vpx_memalign(16, 8 * sizeof(*quant_shift_ptr_)));
dequant_ptr_ = reinterpret_cast<int16_t *>(
vpx_memalign(16, 8 * sizeof(*dequant_ptr_)));
}
~VP9QuantizeBase() {
vpx_free(zbin_ptr_);
vpx_free(round_ptr_);
vpx_free(quant_ptr_);
vpx_free(quant_shift_ptr_);
vpx_free(dequant_ptr_);
zbin_ptr_ = NULL;
round_ptr_ = NULL;
quant_ptr_ = NULL;
quant_shift_ptr_ = NULL;
dequant_ptr_ = NULL;
libvpx_test::ClearSystemState();
}
protected:
int16_t *zbin_ptr_;
int16_t *round_ptr_;
int16_t *quant_ptr_;
int16_t *quant_shift_ptr_;
int16_t *dequant_ptr_;
const vpx_bit_depth_t bit_depth_;
int max_value_;
const int max_size_;
};
class VP9QuantizeTest : public VP9QuantizeBase,
public ::testing::TestWithParam<QuantizeParam> {
public:
VP9QuantizeTest()
: VP9QuantizeBase(GET_PARAM(2), GET_PARAM(3)), quantize_op_(GET_PARAM(0)),
ref_quantize_op_(GET_PARAM(1)) {}
protected:
const QuantizeFunc quantize_op_;
const QuantizeFunc ref_quantize_op_;
};
void GenerateHelperArrays(ACMRandom *rnd, int16_t *zbin, int16_t *round,
int16_t *quant, int16_t *quant_shift,
int16_t *dequant) {
for (int j = 0; j < 2; j++) {
// Values determined by deconstructing vp9_init_quantizer().
// zbin may be up to 1143 for 8 and 10 bit Y values, or 1200 for 12 bit Y
// values or U/V values of any bit depth. This is because y_delta is not
// factored into the vp9_ac_quant() call.
zbin[j] = rnd->RandRange(1200);
// round may be up to 685 for Y values or 914 for U/V.
round[j] = rnd->RandRange(914);
// quant ranges from 1 to -32703
quant[j] = static_cast<int>(rnd->RandRange(32704)) - 32703;
// quant_shift goes up to 1 << 16.
quant_shift[j] = rnd->RandRange(16384);
// dequant maxes out at 1828 for all cases.
dequant[j] = rnd->RandRange(1828);
}
for (int j = 2; j < 8; j++) {
zbin[j] = zbin[1];
round[j] = round[1];
quant[j] = quant[1];
quant_shift[j] = quant_shift[1];
dequant[j] = dequant[1];
}
}
TEST_P(VP9QuantizeTest, OperationCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
Buffer<tran_low_t> coeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 16);
ASSERT_TRUE(coeff.Init());
Buffer<tran_low_t> qcoeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(qcoeff.Init());
Buffer<tran_low_t> dqcoeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(dqcoeff.Init());
Buffer<tran_low_t> ref_qcoeff =
Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(ref_qcoeff.Init());
Buffer<tran_low_t> ref_dqcoeff =
Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(ref_dqcoeff.Init());
uint16_t eob, ref_eob;
for (int i = 0; i < number_of_iterations; ++i) {
// Test skip block for the first three iterations to catch all the different
// sizes.
const int skip_block = 0;
TX_SIZE sz;
if (max_size_ == 16) {
sz = static_cast<TX_SIZE>(i % 3); // TX_4X4, TX_8X8 TX_16X16
} else {
sz = TX_32X32;
}
const TX_TYPE tx_type = static_cast<TX_TYPE>((i >> 2) % 3);
const scan_order *scan_order = &vp9_scan_orders[sz][tx_type];
const int count = (4 << sz) * (4 << sz);
coeff.Set(&rnd, -max_value_, max_value_);
GenerateHelperArrays(&rnd, zbin_ptr_, round_ptr_, quant_ptr_,
quant_shift_ptr_, dequant_ptr_);
ref_quantize_op_(coeff.TopLeftPixel(), count, skip_block, zbin_ptr_,
round_ptr_, quant_ptr_, quant_shift_ptr_,
ref_qcoeff.TopLeftPixel(), ref_dqcoeff.TopLeftPixel(),
dequant_ptr_, &ref_eob, scan_order->scan,
scan_order->iscan);
ASM_REGISTER_STATE_CHECK(
quantize_op_(coeff.TopLeftPixel(), count, skip_block, zbin_ptr_,
round_ptr_, quant_ptr_, quant_shift_ptr_,
qcoeff.TopLeftPixel(), dqcoeff.TopLeftPixel(),
dequant_ptr_, &eob, scan_order->scan, scan_order->iscan));
EXPECT_TRUE(qcoeff.CheckValues(ref_qcoeff));
EXPECT_TRUE(dqcoeff.CheckValues(ref_dqcoeff));
EXPECT_EQ(eob, ref_eob);
if (HasFailure()) {
printf("Failure on iteration %d.\n", i);
qcoeff.PrintDifference(ref_qcoeff);
dqcoeff.PrintDifference(ref_dqcoeff);
return;
}
}
}
TEST_P(VP9QuantizeTest, EOBCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
Buffer<tran_low_t> coeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 16);
ASSERT_TRUE(coeff.Init());
Buffer<tran_low_t> qcoeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(qcoeff.Init());
Buffer<tran_low_t> dqcoeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(dqcoeff.Init());
Buffer<tran_low_t> ref_qcoeff =
Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(ref_qcoeff.Init());
Buffer<tran_low_t> ref_dqcoeff =
Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(ref_dqcoeff.Init());
uint16_t eob, ref_eob;
for (int i = 0; i < number_of_iterations; ++i) {
const int skip_block = 0;
TX_SIZE sz;
if (max_size_ == 16) {
sz = static_cast<TX_SIZE>(i % 3); // TX_4X4, TX_8X8 TX_16X16
} else {
sz = TX_32X32;
}
const TX_TYPE tx_type = static_cast<TX_TYPE>((i >> 2) % 3);
const scan_order *scan_order = &vp9_scan_orders[sz][tx_type];
int count = (4 << sz) * (4 << sz);
// Two random entries
coeff.Set(0);
coeff.TopLeftPixel()[rnd(count)] =
static_cast<int>(rnd.RandRange(max_value_ * 2)) - max_value_;
coeff.TopLeftPixel()[rnd(count)] =
static_cast<int>(rnd.RandRange(max_value_ * 2)) - max_value_;
GenerateHelperArrays(&rnd, zbin_ptr_, round_ptr_, quant_ptr_,
quant_shift_ptr_, dequant_ptr_);
ref_quantize_op_(coeff.TopLeftPixel(), count, skip_block, zbin_ptr_,
round_ptr_, quant_ptr_, quant_shift_ptr_,
ref_qcoeff.TopLeftPixel(), ref_dqcoeff.TopLeftPixel(),
dequant_ptr_, &ref_eob, scan_order->scan,
scan_order->iscan);
ASM_REGISTER_STATE_CHECK(
quantize_op_(coeff.TopLeftPixel(), count, skip_block, zbin_ptr_,
round_ptr_, quant_ptr_, quant_shift_ptr_,
qcoeff.TopLeftPixel(), dqcoeff.TopLeftPixel(),
dequant_ptr_, &eob, scan_order->scan, scan_order->iscan));
EXPECT_TRUE(qcoeff.CheckValues(ref_qcoeff));
EXPECT_TRUE(dqcoeff.CheckValues(ref_dqcoeff));
EXPECT_EQ(eob, ref_eob);
if (HasFailure()) {
printf("Failure on iteration %d.\n", i);
qcoeff.PrintDifference(ref_qcoeff);
dqcoeff.PrintDifference(ref_dqcoeff);
return;
}
}
}
TEST_P(VP9QuantizeTest, DISABLED_Speed) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
Buffer<tran_low_t> coeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 16);
ASSERT_TRUE(coeff.Init());
Buffer<tran_low_t> qcoeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(qcoeff.Init());
Buffer<tran_low_t> dqcoeff = Buffer<tran_low_t>(max_size_, max_size_, 0, 32);
ASSERT_TRUE(dqcoeff.Init());
uint16_t eob;
TX_SIZE starting_sz, ending_sz;
if (max_size_ == 16) {
starting_sz = TX_4X4;
ending_sz = TX_16X16;
} else {
starting_sz = TX_32X32;
ending_sz = TX_32X32;
}
for (TX_SIZE sz = starting_sz; sz <= ending_sz; ++sz) {
// zbin > coeff, zbin < coeff.
for (int i = 0; i < 2; ++i) {
const int skip_block = 0;
// TX_TYPE defines the scan order. That is not relevant to the speed test.
// Pick the first one.
const TX_TYPE tx_type = DCT_DCT;
const scan_order *scan_order = &vp9_scan_orders[sz][tx_type];
const int count = (4 << sz) * (4 << sz);
GenerateHelperArrays(&rnd, zbin_ptr_, round_ptr_, quant_ptr_,
quant_shift_ptr_, dequant_ptr_);
if (i == 0) {
// When |coeff values| are less than zbin the results are 0.
int threshold = 100;
if (max_size_ == 32) {
// For 32x32, the threshold is halved. Double it to keep the values
// from clearing it.
threshold = 200;
}
for (int j = 0; j < 8; ++j) zbin_ptr_[j] = threshold;
coeff.Set(&rnd, -99, 99);
} else if (i == 1) {
for (int j = 0; j < 8; ++j) zbin_ptr_[j] = 50;
coeff.Set(&rnd, -500, 500);
}
vpx_usec_timer timer;
vpx_usec_timer_start(&timer);
for (int j = 0; j < 100000000 / count; ++j) {
quantize_op_(coeff.TopLeftPixel(), count, skip_block, zbin_ptr_,
round_ptr_, quant_ptr_, quant_shift_ptr_,
qcoeff.TopLeftPixel(), dqcoeff.TopLeftPixel(),
dequant_ptr_, &eob, scan_order->scan, scan_order->iscan);
}
vpx_usec_timer_mark(&timer);
const int elapsed_time = static_cast<int>(vpx_usec_timer_elapsed(&timer));
if (i == 0) printf("Bypass calculations.\n");
if (i == 1) printf("Full calculations.\n");
printf("Quantize %dx%d time: %5d ms\n", 4 << sz, 4 << sz,
elapsed_time / 1000);
}
printf("\n");
}
}
using std::tr1::make_tuple;
#if HAVE_SSE2
#if CONFIG_VP9_HIGHBITDEPTH
// TODO(johannkoenig): Fix vpx_quantize_b_sse2 in highbitdepth builds.
// make_tuple(&vpx_quantize_b_sse2, &vpx_highbd_quantize_b_c, VPX_BITS_8),
INSTANTIATE_TEST_CASE_P(
SSE2, VP9QuantizeTest,
::testing::Values(
make_tuple(&vpx_highbd_quantize_b_sse2, &vpx_highbd_quantize_b_c,
VPX_BITS_8, 16),
make_tuple(&vpx_highbd_quantize_b_sse2, &vpx_highbd_quantize_b_c,
VPX_BITS_10, 16),
make_tuple(&vpx_highbd_quantize_b_sse2, &vpx_highbd_quantize_b_c,
VPX_BITS_12, 16),
make_tuple(&vpx_highbd_quantize_b_32x32_sse2,
&vpx_highbd_quantize_b_32x32_c, VPX_BITS_8, 32),
make_tuple(&vpx_highbd_quantize_b_32x32_sse2,
&vpx_highbd_quantize_b_32x32_c, VPX_BITS_10, 32),
make_tuple(&vpx_highbd_quantize_b_32x32_sse2,
&vpx_highbd_quantize_b_32x32_c, VPX_BITS_12, 32)));
#else
INSTANTIATE_TEST_CASE_P(SSE2, VP9QuantizeTest,
::testing::Values(make_tuple(&vpx_quantize_b_sse2,
&vpx_quantize_b_c,
VPX_BITS_8, 16)));
#endif // CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
DISABLED_SSE2, VP9QuantizeTest,
::testing::Values(make_tuple(&QuantFPWrapper<vp9_quantize_fp_sse2>,
&QuantFPWrapper<vp9_quantize_fp_c>, VPX_BITS_8,
16)));
#endif // HAVE_SSE2
#if HAVE_SSSE3 && !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(SSSE3, VP9QuantizeTest,
::testing::Values(make_tuple(&vpx_quantize_b_ssse3,
&vpx_quantize_b_c,
VPX_BITS_8, 16)));
#if ARCH_X86_64
// TODO(johannkoenig): SSSE3 optimizations do not yet pass this test.
INSTANTIATE_TEST_CASE_P(
DISABLED_SSSE3, VP9QuantizeTest,
::testing::Values(make_tuple(&vpx_quantize_b_32x32_ssse3,
&vpx_quantize_b_32x32_c, VPX_BITS_8, 32),
make_tuple(&QuantFPWrapper<vp9_quantize_fp_ssse3>,
&QuantFPWrapper<vp9_quantize_fp_c>, VPX_BITS_8,
16),
make_tuple(&QuantFPWrapper<vp9_quantize_fp_32x32_ssse3>,
&QuantFPWrapper<vp9_quantize_fp_32x32_c>,
VPX_BITS_8, 32)));
#endif // ARCH_X86_64
#endif // HAVE_SSSE3 && !CONFIG_VP9_HIGHBITDEPTH
// TODO(johannkoenig): AVX optimizations do not yet pass the 32x32 test or
// highbitdepth configurations.
#if HAVE_AVX && !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
AVX, VP9QuantizeTest,
::testing::Values(make_tuple(&vpx_quantize_b_avx, &vpx_quantize_b_c,
VPX_BITS_8, 16),
// Even though SSSE3 and AVX do not match the reference
// code, we can keep them in sync with each other.
make_tuple(&vpx_quantize_b_32x32_avx,
&vpx_quantize_b_32x32_ssse3, VPX_BITS_8, 32)));
#endif // HAVE_AVX && !CONFIG_VP9_HIGHBITDEPTH
// TODO(webm:1448): dqcoeff is not handled correctly in HBD builds.
#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
NEON, VP9QuantizeTest,
::testing::Values(
make_tuple(&vpx_quantize_b_neon, &vpx_quantize_b_c, VPX_BITS_8, 16),
make_tuple(&vpx_quantize_b_32x32_neon, &vpx_quantize_b_32x32_c,
VPX_BITS_8, 32),
make_tuple(&QuantFPWrapper<vp9_quantize_fp_neon>,
&QuantFPWrapper<vp9_quantize_fp_c>, VPX_BITS_8, 16),
make_tuple(&QuantFPWrapper<vp9_quantize_fp_32x32_neon>,
&QuantFPWrapper<vp9_quantize_fp_32x32_c>, VPX_BITS_8, 32)));
#endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH
// Only useful to compare "Speed" test results.
INSTANTIATE_TEST_CASE_P(
DISABLED_C, VP9QuantizeTest,
::testing::Values(
make_tuple(&vpx_quantize_b_c, &vpx_quantize_b_c, VPX_BITS_8, 16),
make_tuple(&vpx_quantize_b_32x32_c, &vpx_quantize_b_32x32_c, VPX_BITS_8,
32),
make_tuple(&QuantFPWrapper<vp9_quantize_fp_c>,
&QuantFPWrapper<vp9_quantize_fp_c>, VPX_BITS_8, 16),
make_tuple(&QuantFPWrapper<vp9_quantize_fp_32x32_c>,
&QuantFPWrapper<vp9_quantize_fp_32x32_c>, VPX_BITS_8, 32)));
} // namespace