Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 03 Oct 2018 06:02:25 +0000
From:      bugzilla-noreply@freebsd.org
To:        fs@FreeBSD.org
Subject:   [Bug 229614] ZFS lockup in zil_commit_impl
Message-ID:  <bug-229614-3630-gUgeXyqafE@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-229614-3630@https.bugs.freebsd.org/bugzilla/>
References:  <bug-229614-3630@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=3D229614

--- Comment #19 from vi@endrift.com ---
This is the line that the _sx_xlock_hard call that's hanging is on:
https://github.com/freebsd/freebsd/blob/release/11.2.0/sys/cddl/contrib/ope=
nsolaris/uts/common/fs/zfs/zil.c#L2190
(beginning of zil_commit_writer). That means that the issuer lock is being =
held
by something.

Notably, the issuer lock can be held while waiting on a txg to sync in
zil_commit_waiter_timeout (or more precisely, its call to
zil_commit_writer_stall). I'm not entirely sure (since I'm unfamiliar with =
this
code), but I think this may be the source of the deadlock. Upon inspection,
both the quiesce and sync threads appear to be idle, and I can't find any
evidence that there's any other thread that's in that function, so I'm not =
sure
that's what's going on. Plus I have no idea what's going on with the
un-ptraceable process as mentioned in comment 9, which I've observed too. It
appears to be livelocked on my system (100% of one thread at all times), bu=
t I
can't inspect it at all--even kgdb can't touch it.

--=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-229614-3630-gUgeXyqafE>