From mboxrd@z Thu Jan 1 00:00:00 1970 From: Joel Becker Date: Fri, 20 Jun 2008 15:36:14 -0700 Subject: [Cluster-devel] Re: [Ocfs2-devel] [RFC] configfs: Pin configfs subsystems separately from new config_items. In-Reply-To: <20080620124644.GQ30804@localhost> References: <1213742460-26331-1-git-send-email-joel.becker@oracle.com> <20080618123134.GC30804@localhost> <20080618161215.GA16780@ca-server1.us.oracle.com> <20080618165101.GI30804@localhost> <20080618200713.GE16780@ca-server1.us.oracle.com> <20080619111357.GM30804@localhost> <20080619220739.GC10888@mail.oracle.com> <20080620124644.GQ30804@localhost> Message-ID: <20080620223614.GD21416@mail.oracle.com> List-Id: To: cluster-devel.redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit On Fri, Jun 20, 2008 at 02:46:44PM +0200, Louis Rilling wrote: > But keeping this last config_item_put() prevents the simplification that I'm > defending. And I think that the simplification is worth moving this > config_item_put() before client_drop_item(). I guess I'm not seeing what's simpler. Four lines of try_module_get() aren't very complex, really. Conversely, having some functions that *don't* do config_item_get_item() is weird. Not invalid, just harder to read. Let's put it on the shelf. What I have with this patch is really no different in effective behavior. Joel -- "When ideas fail, words come in very handy." - Goethe Joel Becker Principal Software Developer Oracle E-mail: joel.becker at oracle.com Phone: (650) 506-8127