Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 22 Dec 2023 03:33:37 +0000
From:      bugzilla-noreply@freebsd.org
To:        fs@FreeBSD.org
Subject:   [Bug 275870] [unionfs] [zfs] kernel panic on umount
Message-ID:  <bug-275870-3630-lPps7V8CP0@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-275870-3630@https.bugs.freebsd.org/bugzilla/>
References:  <bug-275870-3630@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D275870

--- Comment #2 from Jason A. Harmening <jah@FreeBSD.org> ---
It looks like this was caused by the vfs_unregister_upper() calls I added in
14.  I'm a bit surprised that zfs rollback obliterates the v_mount field of=
 the
upper/lower vnodes, but then I'm not familiar with that code.  In any case,
unionfs should probably cache the ref'ed upper/lower mount objects returned=
 by
vfs_register_upper_from_vp (as nullfs already does), because there are seve=
ral
cases outside of unmount() in which unionfs directly accesses
ump->um_[lower|upper]vp->v_mount which aren't likely to be safe in the pres=
ence
of a concurrent (recursive) forced unmount.

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-275870-3630-lPps7V8CP0>