From: Mark Fasheh Date: Thu Nov 30 21:52:02 2006 Subject: [Ocfs2-devel] [patch 3/3] ocfs2-timeout-protocol.patch In-Reply-To: <20061130182225.818592000@suse.de> References: <20061130181128.860671000@suse.de> <20061130182225.818592000@suse.de> Message-ID: <20061201055158.GB19617@ca-server1.us.oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com Hi Andrew, This patch is much better. The new comments and the removal of prototype cleanups made it a lot easier to read through too. I have just one comment below :) On Thu, Nov 30, 2006 at 07:11:31PM +0100, abeekhof@suse.de wrote: > Index: fs/ocfs2/cluster/tcp.c > =================================================================== > --- fs/ocfs2/cluster/tcp.c.orig 2006-11-30 18:53:24.000000000 +0100 > +++ fs/ocfs2/cluster/tcp.c 2006-11-30 19:02:30.000000000 +0100 > @@ -380,6 +380,13 @@ static void o2net_sc_cancel_delayed_work > sc_put(sc); > } > > +atomic_t o2net_connected_peers = ATOMIC_INIT(0); > + > +int o2net_num_connected_peers(void) > +{ > + return o2net_connected_peers.counter; > +} I think maybe you missed the definition of 'atomic_read', which needs to be used instead of bare access to the counter value. I see a couple places where you access the counter below - those need to be updated as well. Thanks, --Mark -- Mark Fasheh Senior Software Developer, Oracle mark.fasheh@oracle.com