add maximum limit for grid of reduce, elementwise, gather and scatter #40813
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
ZzSean
reviewed
Mar 23, 2022
| @@ -128,6 +128,8 @@ inline GpuLaunchConfig GetGpuLaunchConfig1D( | |||
| // Number of threads per block shall be larger than 64. | |||
| threads = std::max(64, threads); | |||
| int blocks = DivUp(DivUp(numel, vec_size), threads); | |||
| int limit_blocks = context.GetCUDAMaxGridDimSize()[0]; | |||
| if (blocks > limit_blocks) blocks = limit_blocks; | |||
ZzSean
reviewed
Mar 23, 2022
| @@ -1044,7 +1056,7 @@ void ReduceKernel(const KPDevice& dev_ctx, | |||
|
|
|||
| auto x_dim = phi::vectorize<int>(x.dims()); | |||
| auto config = ReduceConfig<Ty>(origin_reduce_dims, x_dim); | |||
| config.Run(); | |||
| config.Run(x.place()); | |||
There was a problem hiding this comment.
可不可以把这个LimitGridDim写在外部，就可以直接用dev_ctx了
There was a problem hiding this comment.
config.Run()里面有对block数量做限制，所以把thread数量限制一起放在config.Run()里面了，这样可被复用
|
能不能给一下出错的算子配置、修改前的线程数和修改后的线程数？ |
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Outdated suggestions cannot be applied.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
Bug fixes
PR changes
OPs
Describe
The grid number of reduce、elementwise and masked_select has not been limited, which may raise a bug like:
So we add a maximum limit for grid of reduce, elementwise, gather and scatter kernel.