UCP/CORE: Restore cached mem_flags for UCX-owned allocations - #11681
Closed
guy-ealey-morag wants to merge 1 commit into
Closed
UCP/CORE: Restore cached mem_flags for UCX-owned allocations#11681guy-ealey-morag wants to merge 1 commit into
guy-ealey-morag wants to merge 1 commit into
Conversation
Signed-off-by: Guy Ealey Morag <gealeymorag@nvidia.com>
Contributor
Author
|
Possibly already resolved by a different PR |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
When
ucp_memh_create_from_mem()builds a memory handle for a UCX-owned allocation, restore the memory-type flags (notablyUCS_MEM_FLAG_REGISTRABLE) from the cached allocation MD instead of relying solely on pointer detection.Why
memh->mem_flagsdrives which MDs the rendezvous protocol must register the buffer on. For CUDA memory, derivingUCS_MEM_FLAG_REGISTRABLEvia pointer detection can depend on a thread-local CUDA context on the calling thread.NIXL allocates/registers buffers from a thread that does not have the CUDA context bound (a worker/progress thread separate from the one that created the context). On that thread,
REGISTRABLEis silently dropped from the handle.Internal Bug