Return-Path: linux-nfs-owner@vger.kernel.org Received: from verein.lst.de ([213.95.11.211]:44818 "EHLO newverein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752223AbaEHG10 (ORCPT ); Thu, 8 May 2014 02:27:26 -0400 Date: Thu, 8 May 2014 08:27:23 +0200 From: Christoph Hellwig To: "J. Bruce Fields" Cc: linux-nfs@vger.kernel.org Subject: Re: [PATCH] nfsd: clean up fh_auth usage Message-ID: <20140508062723.GA9437@lst.de> References: <1399463384-4324-1-git-send-email-hch@lst.de> <20140507160118.GD4240@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20140507160118.GD4240@fieldses.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Wed, May 07, 2014 at 12:01:18PM -0400, J. Bruce Fields wrote: > On Wed, May 07, 2014 at 01:49:44PM +0200, Christoph Hellwig wrote: > > Use fh_fsid when reffering to the fsid part of the filehandle. The > > variable length auth field envisioned in nfsfh wasn't ever implemented. > > Also clean up some lose ends around this and document the file handle > > format better. > > Thanks. > > > Btw, why do we even export nfsfh.h to userspace? The file handle very > > much is kernel private, and nothing in nfs-utils include the header > > either. > > Something like wireshark might use that information? But anyway that > doesn't mean it makes much sense to export. I'm fine with moving that > header. Let's just put this version in for now, I'll do some research if the uapi headers are used anywhere at all, if not I'll send a series to kill them all.