From: "Chuck Lever" Subject: Re: [PATCH 001 of 4] Copy intr and soft flags to portmap-bind client Date: Wed, 15 Nov 2006 14:38:45 -0500 Message-ID: <76bd70e30611151138p1b41c85as5b954ed886051ff2@mail.gmail.com> References: <20061024122646.4426.patches@notabene> <1061024024851.4734@suse.de> <76bd70e30610241605x13ae0bamdd306aee2f67d4e1@mail.gmail.com> <1163610689.5691.174.camel@lade.trondhjem.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Cc: NeilBrown , Olaf Kirch , Chuck Lever , nfs@lists.sourceforge.net Return-path: Received: from sc8-sf-mx1-b.sourceforge.net ([10.3.1.91] helo=mail.sourceforge.net) by sc8-sf-list2-new.sourceforge.net with esmtp (Exim 4.43) id 1GkQav-00041C-EN for nfs@lists.sourceforge.net; Wed, 15 Nov 2006 11:38:50 -0800 Received: from ug-out-1314.google.com ([66.249.92.174]) by mail.sourceforge.net with esmtp (Exim 4.44) id 1GkQat-00065t-Il for nfs@lists.sourceforge.net; Wed, 15 Nov 2006 11:38:49 -0800 Received: by ug-out-1314.google.com with SMTP id z38so220651ugc for ; Wed, 15 Nov 2006 11:38:46 -0800 (PST) To: "Trond Myklebust" In-Reply-To: <1163610689.5691.174.camel@lade.trondhjem.org> List-Id: "Discussion of NFS under Linux development, interoperability, and testing." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: nfs-bounces@lists.sourceforge.net Errors-To: nfs-bounces@lists.sourceforge.net On 11/15/06, Trond Myklebust wrote: > On Tue, 2006-10-24 at 16:05 -0700, Chuck Lever wrote: > > Looking at this has forced me to actually think about it ;-) > > > > Since portmapper is always a child task, I think it is safe and > > reasonable to use a "hard,intr" RPC client all the time. The parent > > task will timeout automatically if it is a soft mount, right? The RPC > > client should terminate the child if the parent dies. > > ...or to set it to always be soft. I suppose that would be OK, since call_bind_status will retry timed-out rpcbind requests for hard mounts. > Hmm... Looking at that code, though there is something that bothers me. > Why are we calling pmap_wake_portmap_waiters every time the call to > xprt_test_and_set_binding(xprt)!=0? That looks like a nasty bug. Well the idea is to make sure the task isn't left on the binding queue if pmap_getport can't actually queue an rpcbind request. I don't think it should be waking up all waiters, though -- that part is probably wrong. -- "We who cut mere stones must always be envisioning cathedrals" -- Quarry worker's creed ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs