From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Layton Subject: Re: [ 105/175] CIFS: Respect negotiated MaxMpxCount Date: Sun, 1 Apr 2012 09:55:39 -0400 Message-ID: <20120401095539.31323c95@corrin.poochiereds.net> References: <20120330195801.GA31806@kroah.com> <20120330194846.517484801@linuxfoundation.org> <20120331201825.036fe1d8@corrin.poochiereds.net> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Greg KH , linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, stable-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, torvalds-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org, akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org, alan-qBU/x9rampVanCEyBjwyrvXRex20P6io@public.gmane.org, Pavel Shilovsky , linux-cifs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Steve French Return-path: In-Reply-To: Sender: linux-cifs-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: On Sun, 1 Apr 2012 02:26:58 -0500 Steve French wrote: > On Sat, Mar 31, 2012 at 7:18 PM, Jeff Layton wrot= e: > > On Fri, 30 Mar 2012 12:50:10 -0700 > > Greg KH wrote: > > > >> 3.3-stable review patch. =C2=A0If anyone has any objections, pleas= e let me know. > >> > > > > I'm not sure it's a good idea to put this patch into any stable rel= ease > > just yet. I think it's a correct and necessary fix, but it has some > > potential to cause regressions too (or uncover other preexisting > > problems with this code). Might it be best to wait until we have so= me > > more experience with this before we push it into stable? >=20 > Delaying it a little is probably ok, but weigh that against the > frequency/severity > of the problem that Pavel's fix addresses: ie sequential writes to s= ome > versions of Windows 7 and Windows Vista can fail partway > through large file copy (which also causes those servers > to become unresponsive to subsequent connection requests > as well) so the fix does address a hot problem. >=20 >=20 >=20 I'm not disputing whether this patch is correct. It's clearly a bug that cifs.ko has never respected this value, and it's quite problematic with certain servers. The problem is that we don't have any confirmation that: a) this fixes any of the problems that we think it will b) this doesn't introduce any regressions The request slot allocation code is quite fiddly and fragile, and I think the potential for "b" is somewhat high. I'd feel more comfortable if we waited until we have more experience with this patch before merging it into stable. --=20 Jeff Layton From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752343Ab2DANzK (ORCPT ); Sun, 1 Apr 2012 09:55:10 -0400 Received: from mail-gx0-f174.google.com ([209.85.161.174]:61199 "EHLO mail-gx0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751872Ab2DANzG convert rfc822-to-8bit (ORCPT ); Sun, 1 Apr 2012 09:55:06 -0400 Date: Sun, 1 Apr 2012 09:55:39 -0400 From: Jeff Layton To: Steve French Cc: Greg KH , linux-kernel@vger.kernel.org, stable@vger.kernel.org, torvalds@linux-foundation.org, akpm@linux-foundation.org, alan@lxorguk.ukuu.org.uk, Pavel Shilovsky , linux-cifs@vger.kernel.org Subject: Re: [ 105/175] CIFS: Respect negotiated MaxMpxCount Message-ID: <20120401095539.31323c95@corrin.poochiereds.net> In-Reply-To: References: <20120330195801.GA31806@kroah.com> <20120330194846.517484801@linuxfoundation.org> <20120331201825.036fe1d8@corrin.poochiereds.net> X-Mailer: Claws Mail 3.8.0 (GTK+ 2.24.8; x86_64-redhat-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sun, 1 Apr 2012 02:26:58 -0500 Steve French wrote: > On Sat, Mar 31, 2012 at 7:18 PM, Jeff Layton wrote: > > On Fri, 30 Mar 2012 12:50:10 -0700 > > Greg KH wrote: > > > >> 3.3-stable review patch.  If anyone has any objections, please let me know. > >> > > > > I'm not sure it's a good idea to put this patch into any stable release > > just yet. I think it's a correct and necessary fix, but it has some > > potential to cause regressions too (or uncover other preexisting > > problems with this code). Might it be best to wait until we have some > > more experience with this before we push it into stable? > > Delaying it a little is probably ok, but weigh that against the > frequency/severity > of the problem that Pavel's fix addresses: ie sequential writes to some > versions of Windows 7 and Windows Vista can fail partway > through large file copy (which also causes those servers > to become unresponsive to subsequent connection requests > as well) so the fix does address a hot problem. > > > I'm not disputing whether this patch is correct. It's clearly a bug that cifs.ko has never respected this value, and it's quite problematic with certain servers. The problem is that we don't have any confirmation that: a) this fixes any of the problems that we think it will b) this doesn't introduce any regressions The request slot allocation code is quite fiddly and fragile, and I think the potential for "b" is somewhat high. I'd feel more comfortable if we waited until we have more experience with this patch before merging it into stable. -- Jeff Layton