Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

!26605 Fix gpu mem leak bug, add cuda memmalloc result check 

Merge pull request !26605 from wuwenbing/master
This commit is contained in:
i-robot 2021-11-23 01:33:10 +00:00 committed by Gitee
parent 0d11abc7c2 26d4bf6350
commit a04fdd04c9
1 changed files with 14 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
mindspore/ccsrc/backend/kernel_compiler/gpu/math/eigh_c_gpu_kernel.h
View File

@ -114,10 +114,14 @@ class EighcGpuKernel : public GpuKernel {
cudaMalloc(reinterpret_cast<void **>(&dev_input_shape), kShape2dDims * sizeof(size_t));
size_t *dev_input_axis = nullptr;
cudaMalloc(reinterpret_cast<void **>(&dev_input_axis), kShape2dDims * sizeof(size_t));
cudaMemcpyAsync(dev_input_shape, input_shape, kShape2dDims * sizeof(size_t), cudaMemcpyHostToDevice,
reinterpret_cast<cudaStream_t>(stream_ptr));
cudaMemcpyAsync(dev_input_axis, input_axis, kShape2dDims * sizeof(size_t), cudaMemcpyHostToDevice,
reinterpret_cast<cudaStream_t>(stream_ptr));
CHECK_CUDA_RET_WITH_EXCEPT(kernel_node_,
cudaMemcpyAsync(dev_input_shape, input_shape, kShape2dDims * sizeof(size_t),
cudaMemcpyHostToDevice, reinterpret_cast<cudaStream_t>(stream_ptr)),
"malloc input shape workspace failed");
CHECK_CUDA_RET_WITH_EXCEPT(kernel_node_,
cudaMemcpyAsync(dev_input_axis, input_axis, kShape2dDims * sizeof(size_t),
cudaMemcpyHostToDevice, reinterpret_cast<cudaStream_t>(stream_ptr)),
"malloc input shape workspace failed");
CalTranspose(m_ * m_, output_v_addr, dev_input_shape, dev_input_axis, kShape2dDims, w_v_addr,
reinterpret_cast<cudaStream_t>(stream_ptr));
@ -145,6 +149,12 @@ class EighcGpuKernel : public GpuKernel {
reinterpret_cast<cudaStream_t>(stream_ptr));
CalTranspose(m_ * m_, w_v_addr, dev_input_shape, dev_input_axis, kShape2dDims, output_v_addr,
reinterpret_cast<cudaStream_t>(stream_ptr));
if (dev_input_shape) {
cudaFree(dev_input_shape);
}
if (dev_input_axis) {
cudaFree(dev_input_axis);
}
// convert real scalar to complex
if (d_work) {
cudaFree(d_work);