prov/shm: allow sender-side CMA for RMA writes with FI_REMOTE_CQ_DATA#12300
Open
yinliaws wants to merge 1 commit into
Open
prov/shm: allow sender-side CMA for RMA writes with FI_REMOTE_CQ_DATA#12300yinliaws wants to merge 1 commit into
yinliaws wants to merge 1 commit into
Conversation
For RMA writes > SMR_INJECT_SIZE with FI_REMOTE_CQ_DATA (writedata), allow smr_rma_fast (sender-side CMA) instead of falling through to the receiver-side CMA path. The sender does process_vm_writev directly into the target's registered MR, then posts ofi_op_write_async with the cq_data embedded. The receiver generates the remote CQ entry on seeing SMR_REMOTE_CQ_DATA in the write_async notification. This eliminates the freestack allocation and return-queue round-trip for writedata, recovering 20-25% bandwidth regression at 1-8MB on Graviton and AMD platforms. Signed-off-by: Yin Li <yinliq@amazon.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
For RMA writes > SMR_INJECT_SIZE with FI_REMOTE_CQ_DATA (writedata), allow smr_rma_fast (sender-side CMA) instead of falling through to the receiver-side CMA path. The sender does process_vm_writev directly into the target's registered MR, then posts ofi_op_write_async with the cq_data embedded. The receiver generates the remote CQ entry on seeing SMR_REMOTE_CQ_DATA in the write_async notification.
This eliminates the freestack allocation and return-queue round-trip for writedata, recovering 20-25% bandwidth regression at 1-8MB on Graviton and AMD platforms.