Repository navigation
CUDA_IPC: Don't show CUDA error on exit, from static destroy - v1.23 - #11852
Conversation
|
🤖 Starting review — findings will be posted here when done. |
|
Test coverage gap: this is a static-cleanup / driver-shutdown ordering behavior that is hard to unit test, and there is no new test exercising the |
|
🤖 CI Triage Agent — TL;DR: The build failed because Full analysisSummary: Root cause: Also visible earlier in the same log, unrelated but worth noting: Implicated commit: unknown (not caused by [REDACTED:Hex High Entropy String]). The relevant mitigations in master are c77bb48 "UCM/BISTRO: Increase grace time to reduce patch race" (#11112, shasson5) and #11088 "UCM/BISTRO: Wait for syscall completion before patching"; the v1.23 branch under test may predate them. File: src/ucm/bistro/bistro.c:70-103 ( Suggested fix:
Related: #11112 (bistro grace time), #11088 (wait for syscall completion before patching), #11852 (this PR), #11846 (master counterpart)
|
What?
Backport of #11846
Address ai-dynamo/nixl#2171
Why?
We should ignore this diagnostic on cleanup, because it's not a real issue, as all this mem handles will be destroyed anyway, as the cuda driver is shutting down
How?
Add log_level param