[KUBE-54] gimamdsmi: gate watcher device polling on watch subscribers - #87
Open
bhatturu wants to merge 1 commit into
Open
[KUBE-54] gimamdsmi: gate watcher device polling on watch subscribers#87bhatturu wants to merge 1 commit into
bhatturu wants to merge 1 commit into
Conversation
The GIM watcher's timer callback opened /dev/gim-smi0 and issued full-access SMI reads every tick unconditionally, even with no GPUWatch subscribers -- a permanent 1Hz gim poll with no consumer, and the exposure surface for the KUBE-54 hang (a wedged gim ioctl in this path livelocks the caller in-kernel). Port the subscriber guard the baremetal amdsmi path already has. Validated by strace on banff MI300X (GIM) and localhost MI210 (baremetal): patched build opens the device only at startup discovery, zero periodic opens after; stock polls gim-smi0 ~1Hz indefinitely.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The GIM watcher's timer callback opened /dev/gim-smi0 and issued full-access SMI reads every tick unconditionally, even with no GPUWatch subscribers -- a permanent 1Hz gim poll with no consumer, and the exposure surface for the KUBE-54 hang (a wedged gim ioctl in this path livelocks the caller in-kernel). Port the subscriber guard the baremetal amdsmi path already has.
Validated by strace on MI300X (GIM): patched build opens the device only at startup discovery, zero periodic opens after; stock polls gim-smi0 ~1Hz indefinitely.
Motivation
Technical Details
Test Plan
Test Result
Submission Checklist