Return-Path: linux-nfs-owner@vger.kernel.org Received: from bombadil.infradead.org ([198.137.202.9]:41966 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752322AbaFXMGn (ORCPT ); Tue, 24 Jun 2014 08:06:43 -0400 Date: Tue, 24 Jun 2014 05:06:42 -0700 From: Christoph Hellwig To: Jeff Layton Cc: Christoph Hellwig , bfields@fieldses.org, linux-nfs@vger.kernel.org Subject: Re: [PATCH v1 011/104] NFSd: Lock owners are not per open stateid Message-ID: <20140624120642.GA11055@infradead.org> References: <1403189450-18729-1-git-send-email-jlayton@primarydata.com> <1403189450-18729-12-git-send-email-jlayton@primarydata.com> <20140624115058.GA31935@infradead.org> <20140624080426.75799943@tlielax.poochiereds.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20140624080426.75799943@tlielax.poochiereds.net> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Tue, Jun 24, 2014 at 08:04:26AM -0400, Jeff Layton wrote: > Yeah, I was thinking about that last night. I think the parts that could > be broken out are: > > - the change to multiple stateids per lockowner this one definitively should as it is a significant change in behavior. The second batch that seems easily separatable is the improvements of the client lookup and caching, which includes a few patches. > - moving some of the fields into nfs4_stid > - adding refcounting for stateids and stateowners and fixing the > relationships between them These are indeed deeply interwinded with the locking changes. I still think having them earlier would be useful, but if it's too much work we'll have to live with what we get.