Fix guards for using scratch space with SYCL #6003
Merged
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
|
Retest this please. |
dalg24
reviewed
Mar 22, 2023
dalg24
approved these changes
Mar 27, 2023
|
I plan on ignoring the Clang+CUDA build failure |
nliber
approved these changes
Mar 28, 2023
Closed
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Outdated suggestions cannot be applied.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.
This fixes
team_scratch_1_queuesforSYCL+Cudaby properly guarding scratch space forTeamPolicy. The changes here mirror for the most part what we do for Cuda and HIP.Previously, we only had a
m_team_scratch_mutexmutex that would be locked in the constructor of aTeamPolicyParallelFororParallelReduceand unlocked in the respective destructor. As the test failure demonstrates this is not sufficient since subsequent calls could resize and invalidate the level 1 team scratch allocation. InCudaandHIP, this problem is solved by having a pool of scratch allocations that we cycle through until an unused one is found. Similar to what we were previously doing, a scratch allocation would be acquired in the constructor and marked as unused in the destructor. To make appropriate synchronization more obvious, I opted for storing asycl::eventfor each item in the pool and waiting for that event before acquiring the respective team allocation.