osc: Fix rdma component when not using ob1 #13319
Merged
Conversation
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
When the ob1 PML was not eligible for selection (such as when the user sets --mca pml cm), the BML and BTL frameworks are not initialized and the rdma osc component will later fail as there are no BTLs available. This patch resolves the issue by having the rdma osc component initialize the BML interface. Making this change required two additional, related changes. First, since the BTLs use the modex, the rdma initialization must be moved before the modex point, so that putting data in the modex works as expected. Second, BTLs can require loading the entire world during init (such as TCP when there are multiple threads and multiple NICs or usnic), so we extend the world loading checks to include OSC. Since the other Portals4 components say that they do require world loading, we also assume the Portals4 osc component also requires world loading. Signed-off-by: Brian Barrett <[email protected]> (cherry picked from commit 4215325)
hppritcha
approved these changes
Jul 1, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
When the ob1 PML was not eligible for selection (such as when the user sets --mca pml cm), the BML and BTL frameworks are not initialized and the rdma osc component will later fail as there are no BTLs available. This patch resolves the issue by having the rdma osc component initialize the BML interface.
Making this change required two additional, related changes. First, since the BTLs use the modex, the rdma initialization must be moved before the modex point, so that putting data in the modex works as expected. Second, BTLs can require loading the entire world during init (such as TCP when there are multiple threads and multiple NICs or usnic), so we extend the world loading checks to include OSC.
Since the other Portals4 components say that they do require world loading, we also assume the Portals4 osc component also requires world loading.
(cherry picked from commit 4215325)