mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-06-04 23:59:56 +00:00
6b89c7ad04
1.Let mlas use session thread pool 2.Remove onnxruntime_USE_MLAS cmake option 3. Remove the win32 thread pool code inside mlas mlas will: 1.use ort thread pool if it get passed in 2.use openmp if the threadpool parameter is nullptr 3.run single threaded if the threadpool parameter is nullptr and openmp is disabled.
203 lines
6.8 KiB
C++
203 lines
6.8 KiB
C++
/**
|
|
* Copyright (c) 2016-present, Facebook, Inc.
|
|
*
|
|
* Licensed under the Apache License, Version 2.0 (the "License");
|
|
* you may not use this file except in compliance with the License.
|
|
* You may obtain a copy of the License at
|
|
*
|
|
* http://www.apache.org/licenses/LICENSE-2.0
|
|
*
|
|
* Unless required by applicable law or agreed to in writing, software
|
|
* distributed under the License is distributed on an "AS IS" BASIS,
|
|
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
* See the License for the specific language governing permissions and
|
|
* limitations under the License.
|
|
*/
|
|
// Modifications Copyright (c) Microsoft.
|
|
|
|
#include "core/util/math.h"
|
|
#include <gtest/gtest.h>
|
|
#include "core/platform/threadpool.h"
|
|
#include "core/util/math_cpuonly.h"
|
|
namespace onnxruntime {
|
|
|
|
#define VECTOR_HEAD(x) x.size() > 0 ? &x[0] : NULL
|
|
|
|
//parameter is thread pool size
|
|
class MathGemmTest : public testing::TestWithParam<int> {
|
|
protected:
|
|
static concurrency::ThreadPool* CreateThreadPool(int size) {
|
|
if (size == 1)
|
|
return nullptr;
|
|
return new concurrency::ThreadPool("test", size);
|
|
}
|
|
std::unique_ptr<concurrency::ThreadPool> tp{CreateThreadPool(GetParam())};
|
|
};
|
|
|
|
TEST_P(MathGemmTest, GemmNoTransNoTrans) {
|
|
auto& provider = CPUMathUtil::Instance();
|
|
std::vector<float> X(50); // 5 * 10
|
|
std::vector<float> W(60); // 10 * 6
|
|
std::vector<float> Y(30); // 5 * 6
|
|
math::Set<float, CPUMathUtil>(X.size(), 1, VECTOR_HEAD(X), &provider);
|
|
math::Set<float, CPUMathUtil>(W.size(), 1, VECTOR_HEAD(W), &provider);
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < X.size(); ++i) {
|
|
EXPECT_EQ(X[i], 1);
|
|
}
|
|
for (size_t i = 0; i < W.size(); ++i) {
|
|
EXPECT_EQ(W[i], 1);
|
|
}
|
|
|
|
const float kOne = 1.0;
|
|
const float kPointFive = 0.5;
|
|
const float kZero = 0.0;
|
|
math::Gemm<float>(CblasNoTrans, CblasNoTrans, 5, 6, 10, kOne,
|
|
VECTOR_HEAD(X), VECTOR_HEAD(W), kZero, VECTOR_HEAD(Y),
|
|
tp.get());
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 10) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemm<float>(CblasNoTrans, CblasNoTrans, 5, 6, 10, kOne,
|
|
VECTOR_HEAD(X), VECTOR_HEAD(W), kPointFive,
|
|
VECTOR_HEAD(Y), tp.get());
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 15) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemm<float>(CblasNoTrans, CblasNoTrans, 5, 6, 10,
|
|
kPointFive,
|
|
VECTOR_HEAD(X), VECTOR_HEAD(W), kOne, VECTOR_HEAD(Y),
|
|
tp.get());
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 20) << i;
|
|
}
|
|
}
|
|
|
|
TEST_P(MathGemmTest, GemmNoTransTrans) {
|
|
auto& provider = CPUMathUtil::Instance();
|
|
|
|
std::vector<float> X(50); // 5 * 10
|
|
std::vector<float> W(60); // 10 * 6
|
|
std::vector<float> Y(30); // 5 * 6
|
|
math::Set<float, CPUMathUtil>(X.size(), 1, VECTOR_HEAD(X), &provider);
|
|
math::Set<float, CPUMathUtil>(W.size(), 1, VECTOR_HEAD(W), &provider);
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < X.size(); ++i) {
|
|
EXPECT_EQ(X[i], 1);
|
|
}
|
|
for (size_t i = 0; i < W.size(); ++i) {
|
|
EXPECT_EQ(W[i], 1);
|
|
}
|
|
|
|
const float kOne = 1.0;
|
|
const float kPointFive = 0.5;
|
|
const float kZero = 0.0;
|
|
math::Gemm<float>(CblasNoTrans, CblasTrans, 5, 6, 10, kOne,
|
|
VECTOR_HEAD(X), VECTOR_HEAD(W), kZero, VECTOR_HEAD(Y),
|
|
tp.get());
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 10) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemm<float>(CblasNoTrans, CblasTrans, 5, 6, 10, kOne,
|
|
VECTOR_HEAD(X), VECTOR_HEAD(W), kPointFive,
|
|
VECTOR_HEAD(Y), tp.get());
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 15) << i;
|
|
}
|
|
math::Gemm<float>(CblasNoTrans, CblasTrans, 5, 6, 10, kPointFive,
|
|
VECTOR_HEAD(X), VECTOR_HEAD(W), kOne, VECTOR_HEAD(Y),
|
|
tp.get());
|
|
EXPECT_EQ(Y.size(), 30);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 20) << i;
|
|
}
|
|
}
|
|
|
|
INSTANTIATE_TEST_CASE_P(MathGemmTests, MathGemmTest,
|
|
testing::Values(1, 4));
|
|
|
|
TEST(MathTest, GemvNoTrans) {
|
|
auto& provider = CPUMathUtil::Instance();
|
|
std::vector<float> A(50); // 5 * 10
|
|
std::vector<float> X(10);
|
|
std::vector<float> Y(5);
|
|
math::Set<float, CPUMathUtil>(A.size(), 1, VECTOR_HEAD(A), &provider);
|
|
math::Set<float, CPUMathUtil>(X.size(), 1, VECTOR_HEAD(X), &provider);
|
|
EXPECT_EQ(Y.size(), 5);
|
|
for (size_t i = 0; i < A.size(); ++i) {
|
|
EXPECT_EQ(A[i], 1);
|
|
}
|
|
for (size_t i = 0; i < X.size(); ++i) {
|
|
EXPECT_EQ(X[i], 1);
|
|
}
|
|
|
|
const float kOne = 1.0;
|
|
const float kPointFive = 0.5;
|
|
const float kZero = 0.0;
|
|
math::Gemv<float, CPUMathUtil>(CblasNoTrans, 5, 10, kOne, VECTOR_HEAD(A), VECTOR_HEAD(X),
|
|
kZero, VECTOR_HEAD(Y), &provider);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 10) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemv<float, CPUMathUtil>(CblasNoTrans, 5, 10, kOne, VECTOR_HEAD(A), VECTOR_HEAD(X),
|
|
kPointFive, VECTOR_HEAD(Y), &provider);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 15) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemv<float, CPUMathUtil>(CblasNoTrans, 5, 10, kPointFive, VECTOR_HEAD(A),
|
|
VECTOR_HEAD(X), kOne, VECTOR_HEAD(Y),
|
|
&provider);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 20) << i;
|
|
}
|
|
}
|
|
|
|
TEST(MathTest, GemvTrans) {
|
|
auto& provider = CPUMathUtil::Instance();
|
|
std::vector<float> A(60); // 6 * 10
|
|
std::vector<float> X(6);
|
|
std::vector<float> Y(10);
|
|
math::Set<float, CPUMathUtil>(A.size(), 1, VECTOR_HEAD(A), &provider);
|
|
math::Set<float, CPUMathUtil>(X.size(), 1, VECTOR_HEAD(X), &provider);
|
|
EXPECT_EQ(Y.size(), 10);
|
|
for (size_t i = 0; i < A.size(); ++i) {
|
|
EXPECT_EQ(A[i], 1);
|
|
}
|
|
for (size_t i = 0; i < X.size(); ++i) {
|
|
EXPECT_EQ(X[i], 1);
|
|
}
|
|
|
|
const float kOne = 1.0;
|
|
const float kPointFive = 0.5;
|
|
const float kZero = 0.0;
|
|
math::Gemv<float, CPUMathUtil>(CblasTrans, 6, 10, kOne, VECTOR_HEAD(A), VECTOR_HEAD(X),
|
|
kZero, VECTOR_HEAD(Y), &provider);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 6) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemv<float, CPUMathUtil>(CblasTrans, 6, 10, kOne, VECTOR_HEAD(A), VECTOR_HEAD(X),
|
|
kPointFive, VECTOR_HEAD(Y), &provider);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 9) << i;
|
|
}
|
|
// Test Accumulate
|
|
math::Gemv<float, CPUMathUtil>(CblasTrans, 6, 10, kPointFive, VECTOR_HEAD(A),
|
|
VECTOR_HEAD(X), kOne, VECTOR_HEAD(Y),
|
|
&provider);
|
|
for (size_t i = 0; i < Y.size(); ++i) {
|
|
EXPECT_EQ(Y[i], 12) << i;
|
|
}
|
|
}
|
|
|
|
} // namespace onnxruntime
|