Page MenuHomeFreeBSD

nvme: Give up when we've failed
ClosedPublic

Authored by imp on Sep 15 2023, 4:14 PM.
Tags
None
Referenced Files
F102125459: D41877.diff
Thu, Nov 7, 10:35 PM
F102113509: D41877.diff
Thu, Nov 7, 6:59 PM
Unknown Object (File)
Thu, Oct 10, 11:18 AM
Unknown Object (File)
Oct 2 2024, 8:22 PM
Unknown Object (File)
Oct 1 2024, 7:22 AM
Unknown Object (File)
Oct 1 2024, 6:18 AM
Unknown Object (File)
Sep 23 2024, 8:34 AM
Unknown Object (File)
Sep 21 2024, 10:44 PM
Subscribers

Details

Summary

Normally, we poll the device every so often to see if commands have
timed out. However, we'll go into the recovery state as part of failing
the drive. To account for all possibilties, if we're failed when we get
into the polling function, just stop polling: Party is over.

Sponsored by: Netflix

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Skipped
Unit
Tests Skipped
Build Status
Buildable 53608
Build 50499: arc lint + arc unit

Event Timeline

imp requested review of this revision.Sep 15 2023, 4:14 PM
This revision is now accepted and ready to land.Sep 15 2023, 4:56 PM

It turns out there's a better way to do this: If we create a new state for failed qpairs, then we can do some other checks.
https://reviews.freebsd.org/D41866 has that idea fully explored. Once that's ready to go, I plan on abandoning this.