2014-05-17 06:08:05 +08:00
|
|
|
// This file is part of Eigen, a lightweight C++ template library
|
|
|
|
// for linear algebra.
|
|
|
|
//
|
|
|
|
// Copyright (C) 2014 Benoit Steiner <benoit.steiner.goog@gmail.com>
|
|
|
|
//
|
|
|
|
// This Source Code Form is subject to the terms of the Mozilla
|
|
|
|
// Public License v. 2.0. If a copy of the MPL was not distributed
|
|
|
|
// with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
|
|
|
|
|
|
|
|
#define EIGEN_USE_THREADS
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
|
2014-05-17 06:08:05 +08:00
|
|
|
#include "main.h"
|
2015-01-15 07:46:04 +08:00
|
|
|
#include <iostream>
|
2014-05-17 06:08:05 +08:00
|
|
|
#include <Eigen/CXX11/Tensor>
|
|
|
|
|
|
|
|
using Eigen::Tensor;
|
2015-02-11 04:06:19 +08:00
|
|
|
|
2014-05-17 06:08:05 +08:00
|
|
|
|
2014-10-14 08:02:09 +08:00
|
|
|
static void test_multithread_elementwise()
|
2014-05-17 06:08:05 +08:00
|
|
|
{
|
2014-10-14 08:02:09 +08:00
|
|
|
Tensor<float, 3> in1(2,3,7);
|
|
|
|
Tensor<float, 3> in2(2,3,7);
|
|
|
|
Tensor<float, 3> out(2,3,7);
|
2014-05-17 06:08:05 +08:00
|
|
|
|
|
|
|
in1.setRandom();
|
|
|
|
in2.setRandom();
|
|
|
|
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(internal::random<int>(3, 11));
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&tp, internal::random<int>(3, 11));
|
2014-06-06 01:49:34 +08:00
|
|
|
out.device(thread_pool_device) = in1 + in2 * 3.14f;
|
2014-05-17 06:08:05 +08:00
|
|
|
|
|
|
|
for (int i = 0; i < 2; ++i) {
|
|
|
|
for (int j = 0; j < 3; ++j) {
|
|
|
|
for (int k = 0; k < 7; ++k) {
|
2014-10-11 06:20:37 +08:00
|
|
|
VERIFY_IS_APPROX(out(i,j,k), in1(i,j,k) + in2(i,j,k) * 3.14f);
|
2014-05-17 06:08:05 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
|
|
|
|
static void test_multithread_compound_assignment()
|
|
|
|
{
|
|
|
|
Tensor<float, 3> in1(2,3,7);
|
|
|
|
Tensor<float, 3> in2(2,3,7);
|
|
|
|
Tensor<float, 3> out(2,3,7);
|
|
|
|
|
|
|
|
in1.setRandom();
|
|
|
|
in2.setRandom();
|
|
|
|
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(internal::random<int>(3, 11));
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&tp, internal::random<int>(3, 11));
|
2014-10-14 08:02:09 +08:00
|
|
|
out.device(thread_pool_device) = in1;
|
|
|
|
out.device(thread_pool_device) += in2 * 3.14f;
|
|
|
|
|
|
|
|
for (int i = 0; i < 2; ++i) {
|
|
|
|
for (int j = 0; j < 3; ++j) {
|
|
|
|
for (int k = 0; k < 7; ++k) {
|
|
|
|
VERIFY_IS_APPROX(out(i,j,k), in1(i,j,k) + in2(i,j,k) * 3.14f);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
template<int DataLayout>
|
2014-10-14 08:02:09 +08:00
|
|
|
static void test_multithread_contraction()
|
|
|
|
{
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 4, DataLayout> t_left(30, 50, 37, 31);
|
|
|
|
Tensor<float, 5, DataLayout> t_right(37, 31, 70, 2, 10);
|
|
|
|
Tensor<float, 5, DataLayout> t_result(30, 50, 70, 2, 10);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
t_left.setRandom();
|
|
|
|
t_right.setRandom();
|
|
|
|
|
|
|
|
// this contraction should be equivalent to a single matrix multiplication
|
|
|
|
typedef Tensor<float, 1>::DimensionPair DimPair;
|
|
|
|
Eigen::array<DimPair, 2> dims({{DimPair(2, 0), DimPair(3, 1)}});
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
typedef Map<Matrix<float, Dynamic, Dynamic, DataLayout>> MapXf;
|
2014-10-14 08:02:09 +08:00
|
|
|
MapXf m_left(t_left.data(), 1500, 1147);
|
|
|
|
MapXf m_right(t_right.data(), 1147, 1400);
|
2015-01-15 07:46:04 +08:00
|
|
|
Matrix<float, Dynamic, Dynamic, DataLayout> m_result(1500, 1400);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(4);
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&tp, 4);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
// compute results by separate methods
|
|
|
|
t_result.device(thread_pool_device) = t_left.contract(t_right, dims);
|
|
|
|
m_result = m_left * m_right;
|
|
|
|
|
|
|
|
for (ptrdiff_t i = 0; i < t_result.size(); i++) {
|
|
|
|
VERIFY(&t_result.data()[i] != &m_result.data()[i]);
|
|
|
|
if (fabs(t_result.data()[i] - m_result.data()[i]) >= 1e-4) {
|
|
|
|
std::cout << "mismatch detected: " << t_result.data()[i] << " vs " << m_result.data()[i] << std::endl;
|
|
|
|
assert(false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
template<int DataLayout>
|
2014-10-14 08:02:09 +08:00
|
|
|
static void test_contraction_corner_cases()
|
|
|
|
{
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 2, DataLayout> t_left(32, 500);
|
|
|
|
Tensor<float, 2, DataLayout> t_right(32, 28*28);
|
|
|
|
Tensor<float, 2, DataLayout> t_result(500, 28*28);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
t_left = (t_left.constant(-0.5f) + t_left.random()) * 2.0f;
|
|
|
|
t_right = (t_right.constant(-0.6f) + t_right.random()) * 2.0f;
|
|
|
|
t_result = t_result.constant(NAN);
|
|
|
|
|
|
|
|
// this contraction should be equivalent to a single matrix multiplication
|
|
|
|
typedef Tensor<float, 1>::DimensionPair DimPair;
|
|
|
|
Eigen::array<DimPair, 1> dims{{DimPair(0, 0)}};
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
typedef Map<Matrix<float, Dynamic, Dynamic, DataLayout>> MapXf;
|
2014-10-14 08:02:09 +08:00
|
|
|
MapXf m_left(t_left.data(), 32, 500);
|
|
|
|
MapXf m_right(t_right.data(), 32, 28*28);
|
2015-01-15 07:46:04 +08:00
|
|
|
Matrix<float, Dynamic, Dynamic, DataLayout> m_result(500, 28*28);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(12);
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&tp, 12);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
// compute results by separate methods
|
|
|
|
t_result.device(thread_pool_device) = t_left.contract(t_right, dims);
|
|
|
|
m_result = m_left.transpose() * m_right;
|
|
|
|
|
|
|
|
for (ptrdiff_t i = 0; i < t_result.size(); i++) {
|
2015-08-16 20:00:02 +08:00
|
|
|
assert(!(numext::isnan)(t_result.data()[i]));
|
2014-10-14 08:02:09 +08:00
|
|
|
if (fabs(t_result.data()[i] - m_result.data()[i]) >= 1e-4) {
|
|
|
|
std::cout << "mismatch detected at index " << i << " : " << t_result.data()[i] << " vs " << m_result.data()[i] << std::endl;
|
|
|
|
assert(false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
t_left.resize(32, 1);
|
|
|
|
t_left = (t_left.constant(-0.5f) + t_left.random()) * 2.0f;
|
|
|
|
t_result.resize (1, 28*28);
|
|
|
|
t_result = t_result.constant(NAN);
|
|
|
|
t_result.device(thread_pool_device) = t_left.contract(t_right, dims);
|
|
|
|
new(&m_left) MapXf(t_left.data(), 32, 1);
|
|
|
|
m_result = m_left.transpose() * m_right;
|
|
|
|
for (ptrdiff_t i = 0; i < t_result.size(); i++) {
|
2015-08-16 20:00:02 +08:00
|
|
|
assert(!(numext::isnan)(t_result.data()[i]));
|
2014-10-14 08:02:09 +08:00
|
|
|
if (fabs(t_result.data()[i] - m_result.data()[i]) >= 1e-4) {
|
|
|
|
std::cout << "mismatch detected: " << t_result.data()[i] << " vs " << m_result.data()[i] << std::endl;
|
|
|
|
assert(false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
t_left.resize(32, 500);
|
|
|
|
t_right.resize(32, 4);
|
|
|
|
t_left = (t_left.constant(-0.5f) + t_left.random()) * 2.0f;
|
|
|
|
t_right = (t_right.constant(-0.6f) + t_right.random()) * 2.0f;
|
|
|
|
t_result.resize (500, 4);
|
|
|
|
t_result = t_result.constant(NAN);
|
|
|
|
t_result.device(thread_pool_device) = t_left.contract(t_right, dims);
|
|
|
|
new(&m_left) MapXf(t_left.data(), 32, 500);
|
|
|
|
new(&m_right) MapXf(t_right.data(), 32, 4);
|
|
|
|
m_result = m_left.transpose() * m_right;
|
|
|
|
for (ptrdiff_t i = 0; i < t_result.size(); i++) {
|
2015-08-16 20:00:02 +08:00
|
|
|
assert(!(numext::isnan)(t_result.data()[i]));
|
2014-10-14 08:02:09 +08:00
|
|
|
if (fabs(t_result.data()[i] - m_result.data()[i]) >= 1e-4) {
|
|
|
|
std::cout << "mismatch detected: " << t_result.data()[i] << " vs " << m_result.data()[i] << std::endl;
|
|
|
|
assert(false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
t_left.resize(32, 1);
|
|
|
|
t_right.resize(32, 4);
|
|
|
|
t_left = (t_left.constant(-0.5f) + t_left.random()) * 2.0f;
|
|
|
|
t_right = (t_right.constant(-0.6f) + t_right.random()) * 2.0f;
|
|
|
|
t_result.resize (1, 4);
|
|
|
|
t_result = t_result.constant(NAN);
|
|
|
|
t_result.device(thread_pool_device) = t_left.contract(t_right, dims);
|
|
|
|
new(&m_left) MapXf(t_left.data(), 32, 1);
|
|
|
|
new(&m_right) MapXf(t_right.data(), 32, 4);
|
|
|
|
m_result = m_left.transpose() * m_right;
|
|
|
|
for (ptrdiff_t i = 0; i < t_result.size(); i++) {
|
2015-08-16 20:00:02 +08:00
|
|
|
assert(!(numext::isnan)(t_result.data()[i]));
|
2014-10-14 08:02:09 +08:00
|
|
|
if (fabs(t_result.data()[i] - m_result.data()[i]) >= 1e-4) {
|
|
|
|
std::cout << "mismatch detected: " << t_result.data()[i] << " vs " << m_result.data()[i] << std::endl;
|
|
|
|
assert(false);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
template<int DataLayout>
|
2014-10-14 08:02:09 +08:00
|
|
|
static void test_multithread_contraction_agrees_with_singlethread() {
|
|
|
|
int contract_size = internal::random<int>(1, 5000);
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 3, DataLayout> left(internal::random<int>(1, 80),
|
|
|
|
contract_size,
|
|
|
|
internal::random<int>(1, 100));
|
2014-10-14 08:02:09 +08:00
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 4, DataLayout> right(internal::random<int>(1, 25),
|
|
|
|
internal::random<int>(1, 37),
|
|
|
|
contract_size,
|
|
|
|
internal::random<int>(1, 51));
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
left.setRandom();
|
|
|
|
right.setRandom();
|
|
|
|
|
|
|
|
// add constants to shift values away from 0 for more precision
|
|
|
|
left += left.constant(1.5f);
|
|
|
|
right += right.constant(1.5f);
|
|
|
|
|
|
|
|
typedef Tensor<float, 1>::DimensionPair DimPair;
|
|
|
|
Eigen::array<DimPair, 1> dims({{DimPair(1, 2)}});
|
|
|
|
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(internal::random<int>(2, 11));
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&tp, internal::random<int>(2, 11));
|
2014-10-14 08:02:09 +08:00
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 5, DataLayout> st_result;
|
2014-10-14 08:02:09 +08:00
|
|
|
st_result = left.contract(right, dims);
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 5, DataLayout> tp_result(st_result.dimensions());
|
2014-10-14 08:02:09 +08:00
|
|
|
tp_result.device(thread_pool_device) = left.contract(right, dims);
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
VERIFY(dimensions_match(st_result.dimensions(), tp_result.dimensions()));
|
2014-10-14 08:02:09 +08:00
|
|
|
for (ptrdiff_t i = 0; i < st_result.size(); i++) {
|
|
|
|
// if both of the values are very small, then do nothing (because the test will fail
|
|
|
|
// due to numerical precision issues when values are small)
|
|
|
|
if (fabs(st_result.data()[i] - tp_result.data()[i]) >= 1e-4) {
|
|
|
|
VERIFY_IS_APPROX(st_result.data()[i], tp_result.data()[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2015-07-01 04:08:12 +08:00
|
|
|
template<int DataLayout>
|
|
|
|
static void test_multithreaded_reductions() {
|
|
|
|
const int num_threads = internal::random<int>(3, 11);
|
|
|
|
ThreadPool thread_pool(num_threads);
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&thread_pool, num_threads);
|
|
|
|
|
|
|
|
const int num_rows = internal::random<int>(13, 732);
|
|
|
|
const int num_cols = internal::random<int>(13, 732);
|
|
|
|
Tensor<float, 2, DataLayout> t1(num_rows, num_cols);
|
|
|
|
t1.setRandom();
|
|
|
|
|
|
|
|
Tensor<float, 1, DataLayout> full_redux(1);
|
|
|
|
full_redux = t1.sum();
|
|
|
|
|
|
|
|
Tensor<float, 1, DataLayout> full_redux_tp(1);
|
|
|
|
full_redux_tp.device(thread_pool_device) = t1.sum();
|
|
|
|
|
|
|
|
// Check that the single threaded and the multi threaded reductions return
|
|
|
|
// the same result.
|
|
|
|
VERIFY_IS_APPROX(full_redux(0), full_redux_tp(0));
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2014-10-14 08:02:09 +08:00
|
|
|
static void test_memcpy() {
|
|
|
|
|
|
|
|
for (int i = 0; i < 5; ++i) {
|
|
|
|
const int num_threads = internal::random<int>(3, 11);
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(num_threads);
|
|
|
|
Eigen::ThreadPoolDevice thread_pool_device(&tp, num_threads);
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
const int size = internal::random<int>(13, 7632);
|
|
|
|
Tensor<float, 1> t1(size);
|
|
|
|
t1.setRandom();
|
|
|
|
std::vector<float> result(size);
|
|
|
|
thread_pool_device.memcpy(&result[0], t1.data(), size*sizeof(float));
|
2015-06-18 01:14:13 +08:00
|
|
|
for (int j = 0; j < size; j++) {
|
|
|
|
VERIFY_IS_EQUAL(t1(j), result[j]);
|
2014-10-14 08:02:09 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
static void test_multithread_random()
|
|
|
|
{
|
2015-05-21 04:52:07 +08:00
|
|
|
Eigen::ThreadPool tp(2);
|
|
|
|
Eigen::ThreadPoolDevice device(&tp, 2);
|
2015-01-15 07:46:04 +08:00
|
|
|
Tensor<float, 1> t(1 << 20);
|
|
|
|
t.device(device) = t.random<Eigen::internal::NormalRandomGenerator<float>>();
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2014-10-14 08:02:09 +08:00
|
|
|
void test_cxx11_tensor_thread_pool()
|
|
|
|
{
|
|
|
|
CALL_SUBTEST(test_multithread_elementwise());
|
|
|
|
CALL_SUBTEST(test_multithread_compound_assignment());
|
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
CALL_SUBTEST(test_multithread_contraction<ColMajor>());
|
|
|
|
CALL_SUBTEST(test_multithread_contraction<RowMajor>());
|
2014-10-14 08:02:09 +08:00
|
|
|
|
2015-01-15 07:46:04 +08:00
|
|
|
CALL_SUBTEST(test_multithread_contraction_agrees_with_singlethread<ColMajor>());
|
|
|
|
CALL_SUBTEST(test_multithread_contraction_agrees_with_singlethread<RowMajor>());
|
2014-10-14 08:02:09 +08:00
|
|
|
|
|
|
|
// Exercise various cases that have been problematic in the past.
|
2015-01-15 07:46:04 +08:00
|
|
|
CALL_SUBTEST(test_contraction_corner_cases<ColMajor>());
|
|
|
|
CALL_SUBTEST(test_contraction_corner_cases<RowMajor>());
|
2014-10-14 08:02:09 +08:00
|
|
|
|
2015-07-01 04:08:12 +08:00
|
|
|
CALL_SUBTEST(test_multithreaded_reductions<ColMajor>());
|
|
|
|
CALL_SUBTEST(test_multithreaded_reductions<RowMajor>());
|
|
|
|
|
2014-10-14 08:02:09 +08:00
|
|
|
CALL_SUBTEST(test_memcpy());
|
2015-01-15 07:46:04 +08:00
|
|
|
|
|
|
|
CALL_SUBTEST(test_multithread_random());
|
2014-10-14 08:02:09 +08:00
|
|
|
}
|