[Bugs] [Bug 1421721] volume start command hangs

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1421721



--- Comment #4 from Jeff Darcy <jdarcy at redhat.com> ---
There is one place in the multiplexing code where I added a call to
gf_timer_call_cancel.  It's in glusterd_volume_start_glusterfs, which is part
of the stack above.  Suspiciously, this is done without a lock on conn->lock as
is done in other places (e.g. rpc_clnt_reconnect_cleanup).  This fits with the
theory that it's a race.  In gf_timer_call_cancel we do call list_del instead
of list_del_init, and there seems to be no other protection against being
called twice, so I suspect that when we hit the race (because of the missing
lock) we corrupt the list.  Unfortunately, since the test already passes
consistently for me, I won't be able to test whether the fix has any effect.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=PeLSaDL2Ha&a=cc_unsubscribe