1From: Mingming Cao <mcao@us.ibm.com> 2Subject: Re: [RFC][PATCH 06/10] cifs: define inode-level cache object and 3 register them 4Date: Fri, 25 Jun 2010 17:52:24 -0700 5Lines: 92 6Message-ID: <OFB55E8EC7.E8DD23D5-ON8725774E.0004921E-8825774E.0004CC31@us.ibm.com> 7References: <20100625125306.7f9b1966@tlielax.poochiereds.net> <4C24A606.5040001@suse.de> 8 <1277220214-3597-1-git-send-email-sjayaraman@suse.de> <9822.1277312573@redhat.com> 9 <22697.1277470549@redhat.com> <18628.1277502398@redhat.com> <20100625182651.36800d06@tlielax.poochiereds.net> 10 <AANLkTilOTrHLvLv4XWYZO6xCnYZgYT7gO2M-oKZ6VvqM@mail.gmail.com> 11Mime-Version: 1.0 12Content-Type: text/plain; charset=ISO-8859-1 13Content-Transfer-Encoding: quoted-printable 14Cc: linux-cifs@vger.kernel.org, Jeff Layton <jlayton@redhat.com>, 15 samba-technical@lists.samba.org, linux-kernel@vger.kernel.org, 16 David Howells <dhowells@redhat.com>, linux-fsdevel@vger.kernel.org, 17 "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com> 18To: Steve French <smfrench@gmail.com> 19X-From: samba-technical-bounces@lists.samba.org Sat Jun 26 13:36:56 2010 20Return-path: <samba-technical-bounces@lists.samba.org> 21Envelope-to: gnsi-samba-technical@m.gmane.org 22Received: from fn.samba.org ([216.83.154.106] helo=lists.samba.org) 23 by lo.gmane.org with esmtp (Exim 4.69) 24 (envelope-from <samba-technical-bounces@lists.samba.org>) 25 id 1OSTgu-00025d-6P 26 for gnsi-samba-technical@m.gmane.org; Sat, 26 Jun 2010 13:36:56 +0200 27Received: from fn.samba.org (localhost [127.0.0.1]) 28 by lists.samba.org (Postfix) with ESMTP id 1ED11AD2C4; 29 Sat, 26 Jun 2010 05:36:45 -0600 (MDT) 30X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on fn.samba.org 31X-Spam-Level: 32X-Spam-Status: No, score=-6.6 required=3.8 tests=BAYES_00,HTML_MESSAGE, 33 RCVD_IN_DNSWL_MED,SPF_PASS autolearn=ham version=3.2.5 34X-Original-To: samba-technical@lists.samba.org 35Delivered-To: samba-technical@lists.samba.org 36Received: from e34.co.us.ibm.com (e34.co.us.ibm.com [32.97.110.152]) 37 by lists.samba.org (Postfix) with ESMTP id 30F90AD282 38 for <samba-technical@lists.samba.org>; 39 Fri, 25 Jun 2010 18:52:24 -0600 (MDT) 40Received: from d03relay01.boulder.ibm.com (d03relay01.boulder.ibm.com 41 [9.17.195.226]) 42 by e34.co.us.ibm.com (8.14.4/8.13.1) with ESMTP id o5Q0iN1h017083 43 for <samba-technical@lists.samba.org>; Fri, 25 Jun 2010 18:44:23 -0600 44Received: from d03av01.boulder.ibm.com (d03av01.boulder.ibm.com [9.17.195.167]) 45 by d03relay01.boulder.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id 46 o5Q0qQTN175324 47 for <samba-technical@lists.samba.org>; Fri, 25 Jun 2010 18:52:26 -0600 48Received: from d03av01.boulder.ibm.com (loopback [127.0.0.1]) 49 by d03av01.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVout) with ESMTP 50 id o5Q0qPCF006767 51 for <samba-technical@lists.samba.org>; Fri, 25 Jun 2010 18:52:26 -0600 52Received: from d03nm128.boulder.ibm.com (d03nm128.boulder.ibm.com 53 [9.17.195.32]) 54 by d03av01.boulder.ibm.com (8.14.4/8.13.1/NCO v10.0 AVin) with ESMTP id 55 o5Q0qPrh006760; Fri, 25 Jun 2010 18:52:25 -0600 56In-Reply-To: <AANLkTilOTrHLvLv4XWYZO6xCnYZgYT7gO2M-oKZ6VvqM@mail.gmail.com> 57X-KeepSent: B55E8EC7:E8DD23D5-8725774E:0004921E; 58 type=4; name=$KeepSent 59X-Mailer: Lotus Notes Build V852_M2_03302010 March 30, 2010 60X-MIMETrack: Serialize by Router on D03NM128/03/M/IBM(Release 8.0.1|February 61 07, 2008) at 06/25/2010 18:52:25 62X-Mailman-Approved-At: Sat, 26 Jun 2010 05:36:42 -0600 63X-Content-Filtered-By: Mailman/MimeDel 2.1.12 64X-BeenThere: samba-technical@lists.samba.org 65X-Mailman-Version: 2.1.12 66Precedence: list 67List-Id: "Discussions on Samba internals. For general questions please 68 subscribe to the list samba@samba.org" 69 <samba-technical.lists.samba.org> 70List-Unsubscribe: <https://lists.samba.org/mailman/options/samba-technical>, 71 <mailto:samba-technical-request@lists.samba.org?subject=unsubscribe> 72List-Archive: <http://lists.samba.org/pipermail/samba-technical> 73List-Post: <mailto:samba-technical@lists.samba.org> 74List-Help: <mailto:samba-technical-request@lists.samba.org?subject=help> 75List-Subscribe: <https://lists.samba.org/mailman/listinfo/samba-technical>, 76 <mailto:samba-technical-request@lists.samba.org?subject=subscribe> 77Sender: samba-technical-bounces@lists.samba.org 78Errors-To: samba-technical-bounces@lists.samba.org 79Archived-At: <http://permalink.gmane.org/gmane.linux.kernel/1003208> 80 81 82 83Steve French <smfrench@gmail.com> wrote on 06/25/2010 04:05:30 PM: 84 85> Steve French <smfrench@gmail.com> 86> 06/25/2010 04:05 PM 87> 88> To 89> 90> Jeff Layton <jlayton@samba.org>, "Aneesh Kumar K.V" 91> <aneesh.kumar@linux.vnet.ibm.com>, Mingming Cao/Beaverton/IBM@IBMUS 92> 93> cc 94> 95> David Howells <dhowells@redhat.com>, Suresh Jayaraman 96> <sjayaraman@suse.de>, linux-cifs@vger.kernel.org, linux- 97> fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, samba- 98> technical@lists.samba.org, Jeff Layton <jlayton@redhat.com> 99> 100> Subject 101> 102> Re: [RFC][PATCH 06/10] cifs: define inode-level cache object and 103> register them 104> 105> On Fri, Jun 25, 2010 at 5:26 PM, Jeff Layton <jlayton@samba.org> wrot= 106e: 107> > 108> > On Fri, 25 Jun 2010 22:46:38 +0100 109> > David Howells <dhowells@redhat.com> wrote: 110> > 111> > > Jeff Layton <jlayton@samba.org> wrote: 112> > > 113> > > > Looks like it mostly uses the ctime. IMO, the mtime would be a 114better 115> > > > choice since it changes less frequently, but I don't guess that= 116 it 117> > > > matters very much. 118> > > 119> > > I'd've thought mtime changes more frequently since that's 120> altered when data is 121> > > written. =A0ctime is changed when attributes are changed. 122> > > 123> > 124> > IIUC, updating mtime for a write is also an attribute change, and t= 125hat 126> > affects ctime. According to the stat(2) manpage: 127> > 128> > =A0 =A0 =A0 The field st_ctime is changed by writing or by setting 129> =A0inode =A0informa- 130> > =A0 =A0 =A0 tion (i.e., owner, group, link count, mode, etc.). 131> > 132> > > Note that Ext4 appears to have a file creation time field in its 133inode 134> > > (struct ext4_inode::i_crtime[_extra]). =A0Can Samba be made to us= 135e 136that? 137> > > 138> > 139> > Is it exposed to userspace in any (standard) way? It would be handy= 140 to 141> > have that. While we're wishing...it might also be nice to have a 142> > standard way to get at the i_generation from userspace too. 143> > 144> 145> Yes - I have talked with MingMing and Aneesh about those (NFS may 146> someday be able to use those too).=A0 An obstacle in the past had bee= 147n 148> that samba server stores its own fake creation time in an ndr encoded= 149 150> xattr which complicates things. 151> 152> MingMing/Annesh - 153> Xattr or other way to get at birth time? 154> 155> 156 157Not yet, 158 The ext4 file creation time only accesable from the kernel at the mome= 159nt. 160There were discussion 161to make this information avaliable via xattr before, but was rejected, 162since most people 163agree that making this info avalibele via stat() is more standard. Howe= 164ver 165modifying stat() would imply 166big interface change. thus no action has been taken yet. 167 168> -- 169> Thanks, 170> 171> Steve= 172 173 174 175