Re: igc0 crash
Robert Elz <[email protected]>
| Newsgroups | gmane.os.netbsd.current |
|---|---|
| Message-ID | <[email protected]> |
Date: Wed, 12 Aug 2026 04:43:23 +1000
From: matthew green <[email protected]>
Message-ID: <[email protected]>
| and igc_disable_intr() only writes to device regs. ie, there
| should be no way for sc_core_lock to be not held here.
I know nothing about this at all, but what does sc_core_lock being
held or not have to do with anything?
Paul's panic is from kern_timeout.c/callout_stop() which is apparently
failing the
KASSERT(interlock == NULL || mutex_owned(interlock));
which might mean some other lock needs to be taken before the callout_stop()
(or something like that), but private igc locks can't have anything (direcly)
to do with it.
kre