On Fri, 14 May 2010, Dan Carpenter wrote:
> This is a Smatch thing.
>
> fs/ceph/mds_client.c +2217 send_mds_reconnect(82) error: we previously assumed 'session' could be null.
> 2176 down_read(&mdsc->snap_rwsem);
> 2177
> 2178 if (!session)
> 2179 goto send;
>
> If we hit this goto then we will dereference session and oops.
>
> 2180 dout("session %p state %s\n", session,
> 2181 session_state_name(session->s_state));
Yeah, this is a bit of a mess. There is already a fix (and big cleanup of
this function) in the unstable branch, queued up for the next merge
window. I forgot it's theoretically possible to hit this with the current
code (although only with the clustered mds). If I send a final set fixes
for .34 I'll include something for this too...
Thanks!
sage
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
