[BUG] When ASYNC is enabled GDS needs to handle cudaMalloced bounce buffers #5268
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
…d buffers Signed-off-by: Alessandro Bellina <abellina@nvidia.com>
jlowe
approved these changes
Apr 18, 2022
|
build |
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Outdated suggestions cannot be applied.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.
Given the ASYNC allocator is enabled, the UCX bounce buffers are allocated directly, bypassing the ASYNC pool (because memory from the async pool can't be mapped for purposes of GPUDirectRDMA).
This PR fixes GDS copies where it assumed the bounce buffer was a
DeviceMemoryBufferwhen it was really aCudaMemoryBuffer(e.g. straight from cudaMalloc). It also fixes a couple of leaks where.slicewas used, but the sliced buffer was not closed in the stack.