From mboxrd@z Thu Jan 1 00:00:00 1970 From: "J. Bruce Fields" Subject: Re: [PATCH 2/5] rpc: Use separate spinlock for cred locking in auth_gss.c Date: Sat, 14 Jun 2008 13:45:28 -0400 Message-ID: <20080614174528.GF27041@fieldses.org> References: <1213397442-15611-1-git-send-email-bfields@citi.umich.edu> <1213397442-15611-2-git-send-email-bfields@citi.umich.edu> <1213397442-15611-3-git-send-email-bfields@citi.umich.edu> <1213459135.7149.7.camel@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: aglo@citi.umich.edu, kwc@citi.umich.edu, linux-nfs@vger.kernel.org To: Trond Myklebust Return-path: Received: from mail.fieldses.org ([66.93.2.214]:35450 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752649AbYFNRpa (ORCPT ); Sat, 14 Jun 2008 13:45:30 -0400 In-Reply-To: <1213459135.7149.7.camel@localhost> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Sat, Jun 14, 2008 at 11:58:55AM -0400, Trond Myklebust wrote: > On Fri, 2008-06-13 at 18:50 -0400, J. Bruce Fields wrote: > > The i_lock uses in auth_gss.c are to protect the gc_upcall and gc_ctx > > fields of the gss cred; so we can use a separate per-auth spinlock for > > those instead of the i_lock. This simplifies the locking in the case > > where we have two different pipes (one the current pipe, one the pipe > > for a new more flexible upcall), either of which might be used for gss > > upcalls. > > NACK. I deliberately ripped out the old struct gss_auth spinlock in > order to allow multiple gss_auth per inode (I believe Kevin was asking > for this). OK, makes sense. So what will be the scope of a cred lookup--an rpc client? I wish I had a better picture of what the data structures here will eventually look like. --b.