[EXP][CMDBUF] Fix dependency handling for large buffer fill ops in CUDA graph support #1256
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
…DA graph support The CUDA backend does not handle buffer fill larger than 32 bits. The implemented workaround is to add several node that perform 32 bits buffer fill ops. In the previous implementation, all these nodes took all previous node(s) as predecessors. In this version, the first node takes all previous node(s) as predecessors, then the subsequent node takes the newly added node as a predecessor which ensures that the whole fill ops is completed when the last node in this sequence is completed.
EwanC
approved these changes
Jan 17, 2024
|
Linked DPC++ PR: intel/llvm#12405 |
Bensuo
approved these changes
Jan 17, 2024
|
PR #1319 includes this bugfix and other improvements for the buffer fill op in CUDA (Command-Buffer). |
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Outdated suggestions cannot be applied.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 CUDA backend does not handle buffer fill larger than 32 bits.
The implemented workaround is to add several node that perform 32 bits buffer fill ops.
In the previous implementation, all these nodes took all previous node(s) as predecessors.
In this version, the first node takes all previous node(s) as predecessors, then the subsequent node takes the newly added node as a predecessor which ensures that the whole fill ops is completed when the last node in this sequence is completed.