p60419572
/
InfiniTensor
forked from jiuyuan/InfiniTensor
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Fix CMake USE_CUDA (#36) 

* Fix: build lib without cuda

* Chore: rename GBMM and G2BMM files

* Fix: seperate CUDA tests from operator tests

* Fix: CMake CMP0104

* Chore: fix typo

* Chore: remove unused headers

Co-authored-by: Liyan Zheng <liyan-zheng@outlook.com>
This commit is contained in:
zhengly123 2022-09-21 12:28:00 +08:00 committed by GitHub
parent 8f67a5cc76
commit 2f8f706f1c
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
24 changed files with 400 additions and 385 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
CMakeLists.txt
View File

@ -73,8 +73,15 @@ if(BUILD_TEST)
include_directories(3rd-party/googletest/googletest/include)
endif()
file(GLOB_RECURSE SRC src/*.cc src/*.cu)
# Source files
file(GLOB_RECURSE SRC src/core/*.cc src/kernels/cpu/*.cc src/nnet/*.cc src/operators/*.cc src/utils/*.cc)
if(USE_CUDA)
file(GLOB_RECURSE SRC_CUDA src/cuda/*.cc src/cuda/*.cu src/kernels/cuda/*.cc src/kernels/cuda/*.cu)
list (APPEND SRC ${SRC_CUDA})
endif()
# Libraries
add_library(InfiniTensor SHARED ${SRC})
if(USE_PROTOBUF)
target_link_libraries(InfiniTensor tensor_proto)
@ -93,6 +100,8 @@ if(USE_CUDA)
set(CMAKE_CUDA_HOST_COMPILER
${CMAKE_CXX_COMPILER}
CACHE STRING "Set cuda host compiler path")
# CMP0104 requires CUDA_ARCHITECTURES
set_target_properties(InfiniTensor PROPERTIES CUDA_ARCHITECTURES "70;80")
enable_language(CUDA)
# TODO: find_package seems unnecessary for CMake >= 3.8
find_package(CUDA REQUIRED)
@ -123,6 +132,9 @@ if(BUILD_TEST)
if(BUILD_TEST_CORE)
build_test(test/core/*.cc)
build_test(test/operators/*.cc)
if (USE_CUDA)
build_test(test/kernels/cuda/*.cc)
endif()
endif()
if(BUILD_TEST_PET)
build_test(test/pet/*.cc)

1
include/cuda/cuda_common.h
View File

@ -2,6 +2,7 @@
#include "core/common.h"
#include <cublas_v2.h>
#include <cuda.h>
#include <cuda_profiler_api.h>
#include <cudnn.h>
#include <curand.h>

7
include/custom_ops.cuh → include/cuda/gbmm_g2bmm.cuh
View File

@ -1,5 +1,4 @@
#ifndef CUSTOM_OPS_CUH
#define CUSTOM_OPS_CUH
#pragma once
#include <cassert>
@ -5797,6 +5796,4 @@ inline void sgbmml(float *__restrict__ q, float *__restrict__ k,
}
}
} // namespace infini
#endif // CUSTOM_OPS_CUH
} // namespace infini

5
include/custom_ops.h → include/cuda/gbmm_g2bmm.h
View File

@ -1,5 +1,4 @@
#ifndef CUSTOM_OPS_H
#define CUSTOM_OPS_H
#pragma once
namespace infini {
@ -10,5 +9,3 @@ void _sgbmml(float *__restrict__ q, float *__restrict__ k,
float *__restrict__ y, int bs, int n, int m, int w, int d);
} // namespace infini
#endif // CUSTOM_OPS_H

4
src/core/runtime.cc
View File

@ -4,10 +4,6 @@
#include "core/perf_engine.h"
#include <chrono>
#include <cstring>
#include <cuda.h>
#include <cuda_profiler_api.h>
#include <cudnn.h>
#include <curand.h>
namespace infini {

2
src/kernels/cuda/G2BMM.cc
View File

@ -1,7 +1,7 @@
#include "operators/G2BMM.h"
#include "cuda/cuda_kernel_wihtout_config.h"
#include "cuda/cuda_runtime.h"
#include "custom_ops.h"
#include "cuda/gbmm_g2bmm.h"
#include <chrono>
#include <functional>
#include <tuple>

2
src/kernels/cuda/GBMM.cc
View File

@ -1,7 +1,7 @@
#include "operators/GBMM.h"
#include "cuda/cuda_kernel_wihtout_config.h"
#include "cuda/cuda_runtime.h"
#include "custom_ops.h"
#include "cuda/gbmm_g2bmm.h"
#include <chrono>
#include <functional>
#include <tuple>

4
src/custom_ops.cu → src/kernels/cuda/gbmm_g2bmm.cu
View File

@ -1,5 +1,5 @@
#include "custom_ops.cuh"
#include "custom_ops.h"
#include "cuda/gbmm_g2bmm.cuh"
#include "cuda/gbmm_g2bmm.h"
namespace infini {

4
src/nnet/derivator.cc
View File

@ -20,10 +20,12 @@
#include "nnet/Visitor/HashVisitor.h"
#include "nnet/Visitor/MergeMemboundMutator.h"
#include "nnet/Visitor/Serializer.h"
#include "nnet/test.h"
namespace nnet {
// avoid dependence of "nnet/test.h"
bool checkExprsEquvivalence(VecExpr exprs);
class SaveStateGuard {
Derivator &derivator;

1
src/operators/G2BMM.cc
View File

@ -1,5 +1,4 @@
#include "operators/G2BMM.h"
#include "custom_ops.h"
namespace infini {

1
src/operators/GBMM.cc
View File

@ -1,5 +1,4 @@
#include "operators/GBMM.h"
#include "custom_ops.h"
namespace infini {

2
test/operators/test_G2BMM.cc → test/kernels/cuda/test_cuda_G2BMM.cc
View File

@ -10,7 +10,7 @@
namespace infini {
using ExpectOutput = vector<float>;
TEST(G2BMM, ShapeInference) {
TEST(CUDA_G2BMM, ShapeInference) {
const int bs = 1, seqlen = 10000, w = 1000, featlen = 512, heads = 8, d = 4;
const int hidden = featlen, hiddenPerHead = hidden / heads;
auto cpuRuntime = CpuRuntimeObj::getInstance();

2
test/operators/test_GBMM.cc → test/kernels/cuda/test_cuda_GBMM.cc
View File

@ -9,7 +9,7 @@
namespace infini {
using ExpectOutput = vector<float>;
TEST(GBMM, ShapeInference) {
TEST(CUDA_GBMM, ShapeInference) {
const int bs = 1, seqlen = 10000, w = 1000, featlen = 512, heads = 8, d = 4;
const int hidden = featlen, hiddenPerHead = hidden / heads;
auto cpuRuntime = CpuRuntimeObj::getInstance();

79
test/kernels/cuda/test_cuda_conv.cc Normal file
View File

@ -0,0 +1,79 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/conv.h"
#include "test.h"
namespace infini {
void testConvCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
vector<float> ansVec) {
// Construct Runtime and graph for CPU and CUDA
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({1, 3, 4, 4}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({2, 3, 3, 3}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(generator);
w0Cpu->setData(generator);
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv =
gCuda->addOp<ConvObj>(i0Cuda, w0Cuda, nullptr, 1, 1, 2, 1, 1, 2);
// allocate CUDA memory
gCuda->dataMalloc();
// Execute on CUDA
cuda->run(gCuda);
// copy output from CUDA to CPU
auto o0Cpu = gCpu->cloneTensor(conv->getOutput());
// check results on CPU
EXPECT_TRUE(o0Cpu->equalData(ansVec));
// print a tensor/operator/graph by print()
gCuda->print();
}
TEST(cuDNN_Conv, run) {
testConvCudnn(OneGenerator(),
vector<float>{12, 12, 18, 18, 12, 12, 18, 18});
testConvCudnn(
IncrementalGenerator(),
vector<float>{4794, 4386, 8199, 7506, 11274, 10542, 20835, 19656});
}
TEST(cuDNN_Conv, tune) {
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({1, 3, 224, 224}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({2, 3, 3, 3}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(IncrementalGenerator());
w0Cpu->setData(IncrementalGenerator());
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv =
gCuda->addOp<ConvObj>(i0Cuda, w0Cuda, nullptr, 1, 1, 1, 1, 1, 1);
// allocate CUDA memory
gCuda->dataMalloc();
// Execute on CUDA
bool tune = true;
cuda->run(gCuda, tune);
}
} // namespace infini

89
test/kernels/cuda/test_cuda_conv_transposed_2d.cc Normal file
View File

@ -0,0 +1,89 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/perf_engine.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/conv.h"
#include "test.h"
namespace infini {
void testConvTransposedCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
vector<float> ansVec) {
const auto &[N, C, H, W, F, R, S] = tuple{1, 1, 2, 2, 1, 4, 4};
const int stride = 1, padding = 0, dilation = 1;
// Construct Runtime and graph for CPU and CUDA
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({N, F, H, H}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({F, C, R, S}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(generator);
w0Cpu->setData(generator);
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv = gCuda->addOp<ConvTransposed2dObj>(i0Cuda, w0Cuda, nullptr,
padding, padding, stride,
stride, dilation, dilation);
gCuda->dataMalloc();
// Execute on CUDA
cuda->run(gCuda);
// copy output from CUDA to CPU
auto o0Cpu = gCpu->cloneTensor(conv->getOutput());
// check results on CPU
EXPECT_TRUE(o0Cpu->equalData(ansVec));
}
TEST(cuDNN_ConvTransposed, run) {
testConvTransposedCudnn(IncrementalGenerator(),
vector<float>{0., 0., 1., 2., 3., 0., 6.,
12., 18., 16., 8., 30., 36., 42.,
32., 16., 54., 60., 66., 48., 24.,
62., 67., 72., 45.});
}
TEST(cuDNN_ConvTransposed, tune) {
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({1, 448, 2, 2}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({448, 256, 4, 4}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(IncrementalGenerator());
w0Cpu->setData(IncrementalGenerator());
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv = gCuda->addOp<ConvTransposed2dObj>(i0Cuda, w0Cuda, nullptr);
// allocate CUDA memory
gCuda->dataMalloc();
// Execute on CUDA
bool tune = true;
cuda->run(gCuda, tune);
// print a tensor/operator/graph by print()
gCuda->print();
// check record
auto kernelAttrs =
KernelAttrs{Device::CUDA, conv->getOpType(), DataType::Float32};
auto perfKey = PerfEngine::Key{kernelAttrs, conv->getOpPerfKey()};
std::optional<PerfRecord> perfData =
PerfEngine::getInstance().getPerfData(perfKey);
ASSERT_TRUE(perfData.has_value());
}
} // namespace infini

68
test/kernels/cuda/test_cuda_element_wise.cc Normal file
View File

@ -0,0 +1,68 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/element_wise.h"
#include "test.h"
namespace infini {
using ExpectOutput = vector<float>;
template <class T>
void testElementWiseCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape, const ExpectOutput &ansVec) {
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto cudaRuntime = make_ref<CudaRuntimeObj>();
// Build input data on CPU
Tensor acpu = make_ref<TensorObj>(shape, DataType::Float32, cpuRuntime);
acpu->dataMalloc();
acpu->setData(generator);
Tensor bcpu = make_ref<TensorObj>(shape, DataType::Float32, cpuRuntime);
bcpu->dataMalloc();
bcpu->setData(generator);
// Build CUDA graph
Graph g = make_ref<GraphObj>(cudaRuntime);
auto a = g->cloneTensor(acpu);
auto b = g->cloneTensor(bcpu);
auto op = g->addOp<T>(a, b, nullptr);
// allocate CUDA memory
g->dataMalloc();
// Execute on CUDA
cudaRuntime->run(g);
// clone CUDA output to CPU
auto c = op->getOutput();
auto ccpu = c->clone(cpuRuntime);
// cudaPrintTensor(c);
// check results on CPU
EXPECT_TRUE(ccpu->equalData(ansVec));
}
TEST(cuDNN_ElementWise, run) {
testElementWiseCudnn<AddObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22});
testElementWiseCudnn<SubObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0});
testElementWiseCudnn<MulObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121});
testElementWiseCudnn<DivObj>(
OneGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1});
testElementWiseCudnn<PowObj>(IncrementalGenerator(), Shape{1, 2, 2, 1},
ExpectOutput{1, 1, 4, 27});
}
} // namespace infini

76
test/kernels/cuda/test_cuda_matmul.cc Normal file
View File

@ -0,0 +1,76 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/matmul.h"
#include "test.h"
namespace infini {
using ExpectOutput = vector<float>;
void testMatmulCuda(
const std::function<void(void *, size_t, DataType)> &generatorA,
const std::function<void(void *, size_t, DataType)> &generatorB,
bool transA, bool transB, const Shape &shapeA, const Shape &shapeB,
const ExpectOutput &ansVec) {
auto cpuRuntime = CpuRuntimeObj::getInstance();
Graph gCpu = make_ref<GraphObj>(cpuRuntime);
auto ACpu = gCpu->addTensor(shapeA, DataType::Float32);
auto BCpu = gCpu->addTensor(shapeB, DataType::Float32);
gCpu->dataMalloc();
ACpu->setData(generatorA);
BCpu->setData(generatorB);
auto cudaRuntime = make_ref<CudaRuntimeObj>();
auto gCuda = make_ref<GraphObj>(cudaRuntime);
auto ACuda = gCuda->cloneTensor(ACpu);
auto BCuda = gCuda->cloneTensor(BCpu);
auto matmul =
gCuda->addOp<MatmulObj>(ACuda, BCuda, nullptr, transA, transB);
// allocate CUDA memory
gCuda->dataMalloc();
cudaRuntime->run(gCuda);
auto CCpu = gCpu->cloneTensor(matmul->getOutput());
// CCpu->printData();
// check results on CPU
EXPECT_TRUE(CCpu->equalData(ansVec));
// print a tensor/operator/graph by print()
// gCuda->print();
}
TEST(cuBLAS_Matmul, run) {
testMatmulCuda(IncrementalGenerator(), OneGenerator(), false, false,
Shape{1, 3, 5}, Shape{1, 5, 2},
ExpectOutput{10, 10, 35, 35, 60, 60});
testMatmulCuda(IncrementalGenerator(), IncrementalGenerator(), true, false,
Shape{2, 3, 4}, Shape{2, 3, 2},
ExpectOutput{40, 52, 46, 61, 52, 70, 58, 79, 400, 448, 424,
475, 448, 502, 472, 529});
}
TEST(cuBLAS_Matmul, tune) {
auto cpuRuntime = CpuRuntimeObj::getInstance();
Graph gCpu = make_ref<GraphObj>(cpuRuntime);
auto ACpu = gCpu->addTensor(Shape{1, 3, 5}, DataType::Float32);
auto BCpu = gCpu->addTensor(Shape{1, 5, 2}, DataType::Float32);
gCpu->dataMalloc();
ACpu->setData(IncrementalGenerator());
BCpu->setData(IncrementalGenerator());
auto cudaRuntime = make_ref<CudaRuntimeObj>();
auto gCuda = make_ref<GraphObj>(cudaRuntime);
auto ACuda = gCuda->cloneTensor(ACpu);
auto BCuda = gCuda->cloneTensor(BCpu);
auto matmul = gCuda->addOp<MatmulObj>(ACuda, BCuda, nullptr);
// allocate CUDA memory
gCuda->dataMalloc();
cudaRuntime->run(gCuda, true);
}
}; // namespace infini

61
test/kernels/cuda/test_cuda_pooling.cc Normal file
View File

@ -0,0 +1,61 @@
#include "core/graph.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/pooling.h"
#include "test.h"
namespace infini {
using KDPS = vector<int>;
using ExpectOutput = vector<float>;
template <class T>
void testPoolCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape, const KDPS &kdps, const ExpectOutput &ansVec) {
EXPECT_TRUE(kdps.size() == 8);
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto cudaRuntime = make_ref<CudaRuntimeObj>();
// Build input data on CPU
Tensor i0cpu = make_ref<TensorObj>(shape, DataType::Float32, cpuRuntime);
i0cpu->dataMalloc();
i0cpu->setData(generator);
// Build CUDA graph
Graph g = make_ref<GraphObj>(cudaRuntime);
auto i0 = g->cloneTensor(i0cpu);
auto pool = g->addOp<T>(i0, nullptr, kdps[0], kdps[1], kdps[2], kdps[3],
kdps[4], kdps[5], kdps[6], kdps[7]);
// allocate CUDA memory
g->dataMalloc();
// Execute on CUDA
cudaRuntime->run(g);
// clone CUDA output to CPU
auto o0 = pool->getOutput();
auto cpuo0 = o0->clone(cpuRuntime);
// check results on CPU
EXPECT_TRUE(cpuo0->equalData(ansVec));
}
TEST(cuDNN_MaxPool, run) {
testPoolCudnn<MaxPoolObj>(IncrementalGenerator(), Shape{1, 2, 5, 5},
KDPS{3, 3, 1, 1, 1, 1, 2, 2},
ExpectOutput{6, 8, 9, 16, 18, 19, 21, 23, 24, 31,
33, 34, 41, 43, 44, 46, 48, 49});
}
TEST(cuDNN_AvgPool, run) {
testPoolCudnn<AvgPoolObj>(
IncrementalGenerator(), Shape{1, 2, 5, 5}, KDPS{3, 3, 1, 1, 1, 1, 2, 2},
ExpectOutput{1.333333, 3.0000, 2.666667, 7.0000, 12.0000, 9.0000,
8.0000, 13.0000, 9.333333, 12.44444, 19.666667, 13.777778,
23.666667, 37.0000, 25.666667, 19.111111, 29.666667,
20.444444});
}
} // namespace infini

2
test/operators/test_unary.cc → test/kernels/cuda/test_cuda_unary.cc
View File

@ -39,7 +39,7 @@ void testUnary(const std::function<void(void *, size_t, DataType)> &generator,
EXPECT_TRUE(outputCpu->equalData(outputGpu2Cpu));
}
TEST(Unary, CuDNN) {
TEST(cuDNN_Unary, run) {
testUnary<ReluObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});
testUnary<SoftmaxObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});
testUnary<AbsObj>(IncrementalGenerator(), Shape{1, 2, 2, 3});

69
test/operators/test_conv.cc
View File

@ -1,8 +1,6 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/conv.h"
#include "test.h"
@ -68,71 +66,4 @@ TEST(Conv, NaiveCPU) {
EXPECT_TRUE(conv->getOutput()->equalData(ans));
}
void testConvCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
vector<float> ansVec) {
// Construct Runtime and graph for CPU and CUDA
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({1, 3, 4, 4}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({2, 3, 3, 3}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(generator);
w0Cpu->setData(generator);
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv =
gCuda->addOp<ConvObj>(i0Cuda, w0Cuda, nullptr, 1, 1, 2, 1, 1, 2);
// allocate CUDA memory
gCuda->dataMalloc();
// Execute on CUDA
cuda->run(gCuda);
// copy output from CUDA to CPU
auto o0Cpu = gCpu->cloneTensor(conv->getOutput());
// check results on CPU
EXPECT_TRUE(o0Cpu->equalData(ansVec));
// print a tensor/operator/graph by print()
gCuda->print();
}
TEST(Conv, cuDNN) {
testConvCudnn(OneGenerator(),
vector<float>{12, 12, 18, 18, 12, 12, 18, 18});
testConvCudnn(
IncrementalGenerator(),
vector<float>{4794, 4386, 8199, 7506, 11274, 10542, 20835, 19656});
}
TEST(Conv, tune) {
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({1, 3, 224, 224}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({2, 3, 3, 3}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(IncrementalGenerator());
w0Cpu->setData(IncrementalGenerator());
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv =
gCuda->addOp<ConvObj>(i0Cuda, w0Cuda, nullptr, 1, 1, 1, 1, 1, 1);
// allocate CUDA memory
gCuda->dataMalloc();
// Execute on CUDA
bool tune = true;
cuda->run(gCuda, tune);
}
} // namespace infini

78
test/operators/test_conv_transposed_2d.cc
View File

@ -2,8 +2,6 @@
#include "core/kernel.h"
#include "core/perf_engine.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/conv.h"
#include "test.h"
@ -36,80 +34,4 @@ TEST(ConvTransposed, ShapeInference) {
}
}
void testConvTransposedCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
vector<float> ansVec) {
const auto &[N, C, H, W, F, R, S] = tuple{1, 1, 2, 2, 1, 4, 4};
const int stride = 1, padding = 0, dilation = 1;
// Construct Runtime and graph for CPU and CUDA
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({N, F, H, H}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({F, C, R, S}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(generator);
w0Cpu->setData(generator);
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv = gCuda->addOp<ConvTransposed2dObj>(i0Cuda, w0Cuda, nullptr,
padding, padding, stride,
stride, dilation, dilation);
gCuda->dataMalloc();
// Execute on CUDA
cuda->run(gCuda);
// copy output from CUDA to CPU
auto o0Cpu = gCpu->cloneTensor(conv->getOutput());
// check results on CPU
EXPECT_TRUE(o0Cpu->equalData(ansVec));
}
TEST(ConvTransposed, cuDNN) {
testConvTransposedCudnn(IncrementalGenerator(),
vector<float>{0., 0., 1., 2., 3., 0., 6.,
12., 18., 16., 8., 30., 36., 42.,
32., 16., 54., 60., 66., 48., 24.,
62., 67., 72., 45.});
}
TEST(ConvTransposed, tune) {
Runtime cpu = CpuRuntimeObj::getInstance(); // CPUruntime is singleton
Graph gCpu = make_ref<GraphObj>(cpu);
Runtime cuda = make_ref<CudaRuntimeObj>();
Graph gCuda = make_ref<GraphObj>(cuda);
// Set input data on CPU in a CPU Graph
Tensor i0Cpu = gCpu->addTensor({1, 448, 2, 2}, DataType::Float32);
Tensor w0Cpu = gCpu->addTensor({448, 256, 4, 4}, DataType::Float32);
// Malloc data for all tensors in a graph. Do we need implicit allocation?
gCpu->dataMalloc();
i0Cpu->setData(IncrementalGenerator());
w0Cpu->setData(IncrementalGenerator());
// Copy input tensors from CPU to CUDA
Tensor i0Cuda = gCuda->cloneTensor(i0Cpu);
Tensor w0Cuda = gCuda->cloneTensor(w0Cpu);
// Build CUDA graph
auto conv = gCuda->addOp<ConvTransposed2dObj>(i0Cuda, w0Cuda, nullptr);
// allocate CUDA memory
gCuda->dataMalloc();
// Execute on CUDA
bool tune = true;
cuda->run(gCuda, tune);
// print a tensor/operator/graph by print()
gCuda->print();
// check record
auto kernelAttrs =
KernelAttrs{Device::CUDA, conv->getOpType(), DataType::Float32};
auto perfKey = PerfEngine::Key{kernelAttrs, conv->getOpPerfKey()};
std::optional<PerfRecord> perfData =
PerfEngine::getInstance().getPerfData(perfKey);
ASSERT_TRUE(perfData.has_value());
}
} // namespace infini

100
test/operators/test_element_wise.cc
View File

@ -1,8 +1,6 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/element_wise.h"
#include "test.h"
@ -20,103 +18,5 @@ TEST(ElementWise, ShapeInference) {
EXPECT_EQ(op->getOutput()->getDims(), (Shape{2, 3, 3, 4}));
}
}
/*
template <typename T>
void test_element_wise(
const std::function<void(void *, size_t, DataType)> &generator,
const vector<uint32_t> &ans) {
Runtime runtime = CpuRuntimeObj::getInstance();
Graph g = make_ref<GraphObj>(runtime);
Tensor i0 = g->addTensor({1, 3, 2, 2}, DataType::UInt32);
Tensor i1 = g->addTensor({2, 3, 1, 2}, DataType::UInt32);
auto op = g->addOp<T>(i0, i1, nullptr);
g->dataMalloc();
i0->setData(generator);
i1->setData(generator);
runtime->run(g, true, true);
// check answer
EXPECT_TRUE(op->getOutput()->equalData(ans));
}
TEST(ElementWise, NaiveCPU) {
test_element_wise<AddObj>(IncrementalGenerator(),
vector<uint32_t>{0, 2, 2, 4, 6, 8, 8, 10,
12, 14, 14, 16, 6, 8, 8, 10,
12, 14, 14, 16, 18, 20, 20, 22});
test_element_wise<SubObj>(
IncrementalGenerator(),
vector<uint32_t>{0, 0, 2, 2,
2, 2, 4, 4,
4, 4, 6, 6,
4294967290, 4294967290, 4294967292, 4294967292,
4294967292, 4294967292, 4294967294, 4294967294,
4294967294, 4294967294, 0, 0});
test_element_wise<MulObj>(
IncrementalGenerator(),
vector<uint32_t>{0, 1, 0, 3, 8, 15, 12, 21, 32, 45, 40, 55,
0, 7, 12, 21, 32, 45, 48, 63, 80, 99, 100, 121});
test_element_wise<DivObj>(OneGenerator(),
vector<uint32_t>{
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
});
}
*/
template <class T>
void testElementWiseCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape, const ExpectOutput &ansVec) {
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto cudaRuntime = make_ref<CudaRuntimeObj>();
// Build input data on CPU
Tensor acpu = make_ref<TensorObj>(shape, DataType::Float32, cpuRuntime);
acpu->dataMalloc();
acpu->setData(generator);
Tensor bcpu = make_ref<TensorObj>(shape, DataType::Float32, cpuRuntime);
bcpu->dataMalloc();
bcpu->setData(generator);
// Build CUDA graph
Graph g = make_ref<GraphObj>(cudaRuntime);
auto a = g->cloneTensor(acpu);
auto b = g->cloneTensor(bcpu);
auto op = g->addOp<T>(a, b, nullptr);
// allocate CUDA memory
g->dataMalloc();
// Execute on CUDA
cudaRuntime->run(g);
// clone CUDA output to CPU
auto c = op->getOutput();
auto ccpu = c->clone(cpuRuntime);
// cudaPrintTensor(c);
// check results on CPU
EXPECT_TRUE(ccpu->equalData(ansVec));
}
TEST(ElementWise, CuDNN) {
testElementWiseCudnn<AddObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22});
testElementWiseCudnn<SubObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0});
testElementWiseCudnn<MulObj>(
IncrementalGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121});
testElementWiseCudnn<DivObj>(
OneGenerator(), Shape{1, 2, 2, 3},
ExpectOutput{1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1});
testElementWiseCudnn<PowObj>(IncrementalGenerator(), Shape{1, 2, 2, 1},
ExpectOutput{1, 1, 4, 27});
}
} // namespace infini

63
test/operators/test_matmul.cc
View File

@ -2,8 +2,6 @@
#include "core/graph.h"
#include "core/kernel.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/matmul.h"
#include "test.h"
@ -30,66 +28,5 @@ TEST(Matmul, ShapeInference) {
EXPECT_EQ(C->getDims(), (Shape{3, 4, 2}));
}
}
void testMatmulCuda(
const std::function<void(void *, size_t, DataType)> &generatorA,
const std::function<void(void *, size_t, DataType)> &generatorB,
bool transA, bool transB, const Shape &shapeA, const Shape &shapeB,
const ExpectOutput &ansVec) {
auto cpuRuntime = CpuRuntimeObj::getInstance();
Graph gCpu = make_ref<GraphObj>(cpuRuntime);
auto ACpu = gCpu->addTensor(shapeA, DataType::Float32);
auto BCpu = gCpu->addTensor(shapeB, DataType::Float32);
gCpu->dataMalloc();
ACpu->setData(generatorA);
BCpu->setData(generatorB);
auto cudaRuntime = make_ref<CudaRuntimeObj>();
auto gCuda = make_ref<GraphObj>(cudaRuntime);
auto ACuda = gCuda->cloneTensor(ACpu);
auto BCuda = gCuda->cloneTensor(BCpu);
auto matmul =
gCuda->addOp<MatmulObj>(ACuda, BCuda, nullptr, transA, transB);
// allocate CUDA memory
gCuda->dataMalloc();
cudaRuntime->run(gCuda);
auto CCpu = gCpu->cloneTensor(matmul->getOutput());
// CCpu->printData();
// check results on CPU
EXPECT_TRUE(CCpu->equalData(ansVec));
// print a tensor/operator/graph by print()
// gCuda->print();
}
TEST(Matmul, cuBlas) {
testMatmulCuda(IncrementalGenerator(), OneGenerator(), false, false,
Shape{1, 3, 5}, Shape{1, 5, 2},
ExpectOutput{10, 10, 35, 35, 60, 60});
testMatmulCuda(IncrementalGenerator(), IncrementalGenerator(), true, false,
Shape{2, 3, 4}, Shape{2, 3, 2},
ExpectOutput{40, 52, 46, 61, 52, 70, 58, 79, 400, 448, 424,
475, 448, 502, 472, 529});
}
TEST(Matmul, tune) {
auto cpuRuntime = CpuRuntimeObj::getInstance();
Graph gCpu = make_ref<GraphObj>(cpuRuntime);
auto ACpu = gCpu->addTensor(Shape{1, 3, 5}, DataType::Float32);
auto BCpu = gCpu->addTensor(Shape{1, 5, 2}, DataType::Float32);
gCpu->dataMalloc();
ACpu->setData(IncrementalGenerator());
BCpu->setData(IncrementalGenerator());
auto cudaRuntime = make_ref<CudaRuntimeObj>();
auto gCuda = make_ref<GraphObj>(cudaRuntime);
auto ACuda = gCuda->cloneTensor(ACpu);
auto BCuda = gCuda->cloneTensor(BCpu);
auto matmul = gCuda->addOp<MatmulObj>(ACuda, BCuda, nullptr);
// allocate CUDA memory
gCuda->dataMalloc();
cudaRuntime->run(gCuda, true);
}
}; // namespace infini

51
test/operators/test_pooling.cc
View File

@ -1,7 +1,5 @@
#include "core/graph.h"
#include "core/runtime.h"
#include "cuda/cuda_runtime.h"
#include "cuda/cuda_utility.h"
#include "operators/pooling.h"
#include "test.h"
@ -70,53 +68,4 @@ TEST(AvgPool, NaiveCPU) {
EXPECT_LT(perfTime, 5);
}
template <class T>
void testPoolCudnn(
const std::function<void(void *, size_t, DataType)> &generator,
const Shape &shape, const KDPS &kdps, const ExpectOutput &ansVec) {
EXPECT_TRUE(kdps.size() == 8);
Runtime cpuRuntime = CpuRuntimeObj::getInstance();
auto cudaRuntime = make_ref<CudaRuntimeObj>();
// Build input data on CPU
Tensor i0cpu = make_ref<TensorObj>(shape, DataType::Float32, cpuRuntime);
i0cpu->dataMalloc();
i0cpu->setData(generator);
// Build CUDA graph
Graph g = make_ref<GraphObj>(cudaRuntime);
auto i0 = g->cloneTensor(i0cpu);
auto pool = g->addOp<T>(i0, nullptr, kdps[0], kdps[1], kdps[2], kdps[3],
kdps[4], kdps[5], kdps[6], kdps[7]);
// allocate CUDA memory
g->dataMalloc();
// Execute on CUDA
cudaRuntime->run(g);
// clone CUDA output to CPU
auto o0 = pool->getOutput();
auto cpuo0 = o0->clone(cpuRuntime);
// check results on CPU
EXPECT_TRUE(cpuo0->equalData(ansVec));
}
TEST(MaxPool, CuDNN) {
testPoolCudnn<MaxPoolObj>(IncrementalGenerator(), Shape{1, 2, 5, 5},
KDPS{3, 3, 1, 1, 1, 1, 2, 2},
ExpectOutput{6, 8, 9, 16, 18, 19, 21, 23, 24, 31,
33, 34, 41, 43, 44, 46, 48, 49});
}
TEST(AvgPool, CuDNN) {
testPoolCudnn<AvgPoolObj>(
IncrementalGenerator(), Shape{1, 2, 5, 5}, KDPS{3, 3, 1, 1, 1, 1, 2, 2},
ExpectOutput{1.333333, 3.0000, 2.666667, 7.0000, 12.0000, 9.0000,
8.0000, 13.0000, 9.333333, 12.44444, 19.666667, 13.777778,
23.666667, 37.0000, 25.666667, 19.111111, 29.666667,
20.444444});
}
} // namespace infini