Fixed critical issue in CPUAccelerator runtime related to Shared Memory allocations and Warp operations. #836
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
m4rs-mt
force-pushed
the
fixed_cpu_allocations
branch
3 times, most recently
from
ee56c8e
to
1d66d80
Compare
MoFtZ
approved these changes
Sep 1, 2022
m4rs-mt
force-pushed
the
fixed_cpu_allocations
branch
from
1d66d80
to
730ffdc
Compare
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Outdated suggestions cannot be applied.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.
The current GPU emulation runtime environment (currently referred to as
CPUAccelerator) has a critical bug which can lead to invalid program output in certain scenarios. This bug is considered critical because programs may crash, but not in all cases, generally speaking. In other words, this problem can lead to non-deterministic incorrect output that may go unnoticed.The problem is related to the performance of
SharedMemoryassignments in divergent branches that "return" to the main execution thread, which in turn performs a number of specific operations targetingWarpandSharedMemoryoperations.This PR addresses this critical problem by refining the way
SharedMemoryallocations work in the context of theCPUAcceleratorruntime.