Commit 03c58dc6 authored by fenglei.tian's avatar fenglei.tian

fix merge bugs

parent a4f01e62
mkdir build
cd build
cmake .. -DNGRAPH_GPU_ENABLE=TRUE -DNGRAPH_CPU_ENABLE=TRUE -DCUDNN_ROOT_DIR=/usr/lib/x86_64-linux-gnu/ -DCUDNN_INCLUDE_DIR=/usr/include -DZLIB_LIBRARY=/usr/lib/x86_64-linux/gpu/libz.so -DZLIB_INCLUDE_DIR=/usr/include/ -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DNGRPH_COMPILER_DIAG_ENABLE=TRUE
make -j24 all
./build/test/unit-test --gtest_filter=GPU.abs
#./build/test/unit-test --gtest_filter=GPU.ab
#./build/test/unit-test --gtest_filter=GPU.abc
#./build/test/unit-test --gtest_filter=GPU.maximum
#./build/test/unit-test --gtest_filter=GPU.minimum
#./build/test/unit-test --gtest_filter=GPU.multiple*
#./build/test/unit-test --gtest_filter=GPU.sqrt
#./build/test/unit-test --gtest_filter=GPU.nagtive
#./build/test/unit-test --gtest_filter=GPU.abs
#./build/test/unit-test --gtest_filter=GPU.dot*
......@@ -21,35 +21,11 @@
#include "ngraph/runtime/gpu/gpu_call_frame.hpp"
#include "ngraph/runtime/gpu/gpu_external_function.hpp"
#include "ngraph/runtime/gpu/gpu_tensor_view.hpp"
#include "ngraph/runtime/gpu/gpu_util.hpp"
using namespace std;
using namespace ngraph;
#define NVRTC_SAFE_CALL(x) \
do \
{ \
nvrtcResult result = x; \
if (result != NVRTC_SUCCESS) \
{ \
std::cerr << "\nerror: " #x " failed with error " << nvrtcGetErrorString(result) \
<< '\n'; \
exit(1); \
} \
} while (0)
#define CUDA_SAFE_CALL(x) \
do \
{ \
CUresult result = x; \
if (result != CUDA_SUCCESS) \
{ \
const char* msg; \
cuGetErrorName(result, &msg); \
std::cerr << "\nerror: " #x " failed with error " << msg << '\n'; \
exit(1); \
} \
} while (0)
runtime::gpu::GPU_CallFrame::GPU_CallFrame(std::shared_ptr<GPU_ExternalFunction> external_function,
EntryPoint compiled_function)
: m_external_function(external_function)
......
......@@ -16,7 +16,9 @@
#pragma once
#include <memory>
#include <string>
#include "ngraph/runtime/gpu/gpu_util.hpp"
namespace ngraph
......
......@@ -17,16 +17,11 @@
#include <algorithm>
#include <cmath>
#include <cublas_v2.h>
#include <cublas_v2.h>
#include <cuda.h>
#include <cuda.h>
#include <cuda_runtime.h>
#include <cuda_runtime.h>
#include <cudnn_v7.h>
#include <cudnn_v7.h>
#include <iostream>
#include <nvrtc.h>
#include <nvrtc.h>
#include <string>
#include <typeindex>
#include <unordered_map>
......
......@@ -256,10 +256,10 @@ void runtime::gpu::GPU_ExternalFunction::compile()
string pch_header_source = writer.get_code();
// writer += R"(
// using namespace ngraph;
// using namespace std;
//)";
writer += R"(
using namespace ngraph;
using namespace std;
)";
if (m_emit_timing)
{
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment