Optimize flip kernel by eliminating H2D data transfer, test=develop #46046
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Courtesy-Xs
force-pushed
the
flip_optimize
branch
from
8f4a035
to
808b3b3
Compare
Xreki
reviewed
Sep 15, 2022
| std::vector<int> flip_dims = axis; | ||
|
|
||
| template <typename T, typename Context, size_t N> | ||
| void launch_flip_cuda_kernel(const Context& dev_ctx, |
There was a problem hiding this comment.
launch_flip_cuda_kernel -> LaunchFlipCudaKernel，函数命名是大驼峰式。
|
|
||
| int block_size = 512; | ||
| dim3 dim_block(block_size); | ||
| dim3 dim_grid((N + block_size - 1) / block_size); | ||
| dim3 dim_grid((numel + block_size - 1) / block_size); |
There was a problem hiding this comment.
线程配置，可以调用phi::backends::gpu::GetGpuLaunchConfig1D函数
|
|
||
| namespace phi { | ||
|
|
||
| template <typename T> | ||
| template <typename T, size_t Rank> | ||
| __global__ void flip_cuda_kernel(const int N, |
| auto x_dims = x.dims(); | ||
| const int total_dims = x_dims.size(); | ||
| const int N = x.numel(); | ||
| const int numel = x.numel(); |
| for (size_t idx = 0; idx < N; ++idx) { | ||
| stride_a[idx] = x_stride[idx]; | ||
| shape_a[idx] = x_dims[idx]; | ||
| flip_dims_a[idx] = idx < flip_dims_size ? flip_dims_v[idx] : 0; |
There was a problem hiding this comment.
flip_dims_v也没有必要吧，直接写入flip_dims_a就行？
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR types
Performance optimization
PR changes
OPs
Describe
Test environment
pytorch 1.12.1+102
paddle 2.3+102
cuda 11.2
integration times 1000
dype float32
shape [100,1785]
axis 1
Test Result
