Migrate crossKernel from THC to ATen (CUDA) #60039
Closed
Conversation
This file contains hidden or 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
peterbell10
added
module: porting
💊 CI failures summary and remediationsAs of commit 697c739 (more details on the Dr. CI page and at hud.pytorch.org/pr/60039):
🕵️ 3 new failures recognized by patternsThe following CI failures do not appear to be due to upstream breakages:
|
|
I measure a performance improvement here:
|
anjali411
added
the
triaged
ngimel
reviewed
Jun 22, 2021
| int64_t x1stride, int64_t x2stride) { | ||
| const auto N = iter.numel(); | ||
| auto offset_calculator = make_element_offset_calculator<3>(iter); | ||
| TORCH_INTERNAL_ASSERT(N > 0 && N <= std::numeric_limits<int32_t>::max()); |
There was a problem hiding this comment.
Is this check enough? e.g. if you have a tensors with <INT_MAX element, but it's not memory dense, so maximum offset is >INT_MAX (or UINT_MAX)
There was a problem hiding this comment.
I could assert iter.can_use_32bit_indexing() but that's a more costly check and is guarunteed by cross_impl using with_32bit_indexing anyway.
There was a problem hiding this comment.
Then with_32bit_indexing should guarantee that N is less than INT_MAX? Or it's here to protect people directly calling launch_cross_kernel, not going through cross_impl? In this case probably should be TORCH_INTERNAL_ASSERT_DEBUG_ONLY (unlikely someone will call launch_cross_kernel directly).
|
@ngimel has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
ngimel
approved these changes
Jun 22, 2021
|
@ngimel has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
Ref #24507 (There doesn't seem to be an actual issue for cross)
This also moves the remaining operator functors in
THCTensorMathPointwise.cuhtoSparseCUDATensorMath.cuwhich is the only file using them.