From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg KH Subject: Re: [stable] [PATCH] percpu_counter: Fix __percpu_counter_sum() Date: Thu, 11 Dec 2008 08:32:26 -0800 Message-ID: <20081211163226.GA15057@kroah.com> References: <4936D287.6090206@cosmosbay.com> <4936EB04.8000609@cosmosbay.com> <20081207204222.d811c00b.akpm@linux-foundation.org> <1228758923.7096.10.camel@mingming-laptop> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Andrew Morton , Theodore Tso , Peter Zijlstra , stable@kernel.org, linux kernel , Eric Dumazet , linux-ext4@vger.kernel.org, "David S. Miller" To: Mingming Cao Return-path: Content-Disposition: inline In-Reply-To: <1228758923.7096.10.camel@mingming-laptop> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org On Mon, Dec 08, 2008 at 09:55:23AM -0800, Mingming Cao wrote: > =E5=9C=A8 2008-12-07=E6=97=A5=E7=9A=84 20:42 -0800=EF=BC=8CAndrew Mor= ton=E5=86=99=E9=81=93=EF=BC=9A > > (cc stable) > >=20 > > On Sun, 7 Dec 2008 10:28:21 -0500 Theodore Tso wrot= e: > >=20 > > > On Sat, Dec 06, 2008 at 08:22:33PM -0800, Andrew Morton wrote: > > > >=20 > > > > I suggest that what we do is to revert both those changes. We = can > > > > worry about the possibly-unneeded spin_lock later, in a separat= e patch. > > > >=20 > > > > It should have been a separate patch anyway. It's conceptually > > > > unrelated and is not a bugfix, but it was mixed in with a bugfi= x. > > > >=20 > > > > Mingming, this needs urgent consideration, please. Note that I= had to > > > > make additional changes to ext4 due to the subsequent introduct= ion of > > > > the dirty_blocks counter. > > >=20 > > > I've looked the two patches which you've queued in the -mm branch= , and > > > they look correct to me. > > >=20 > > > The bugs fixed by these patches can potentially lead to filesyste= m > > > corruption, since we ultimately use these fields to set the super= block > > > values. This in my mind makes them stable candidates at the very > > > least, and if we weren't so late in the 2.6.28 cycle, I'd be stro= ngly > > > tempted to push them to Linus as a bugfix before the merge window= =2E > > >=20 > > > Andrew, any strong objections for me to grab them for the ext4 tr= ee? > > > Or would you rather carry them? I would prefer that they get pus= hed > > > to Linus as soon as the merge window opens, which is one reason w= hy > > > I'd prefer carry them, but we can do this either way. > > >=20 > >=20 > > I'm planning on sending them off to Linus for 2.6.28 this week, > > assuming nobody can think of a plausible reason to not do that. > >=20 > > Now I didn't look _very_ closely at the chronology, but I think tha= t > > revert-percpu-counter-clean-up-percpu_counter_sum_and_set.patch rev= erts > > a post-2.6.27 change, and is not needed in stable. > >=20 > > revert-percpu_counter-new-function-percpu_counter_sum_and_set.patch > > however reverts a pre-2.6.27 change, and should be merged into 2.6.= 27.=20 > > This patch reverts the addition and use of > > percpu_counter_sum_and_set(), which is racy and can corrupt the > > counters. > >=20 > > However > > revert-percpu_counter-new-function-percpu_counter_sum_and_set.patch > > won't apply to 2.6.27 because the dirty_blocks stuff was added and > > generates rejects. > >=20 > > So if all the above is correct, I'd propose that if and when > > revert-percpu_counter-new-function-percpu_counter_sum_and_set.patch > > hits mainline, we should ask the -stable guys to directly revert=20 > >=20 >=20 > Agreed. >=20 > I checked 2.6.27.8, above are correct, the > revert-percpu-counter-clean-up-percpu_counter_sum_and_set.patch is no= t > needed for 2.6.27.x stable tree. Thanks for letting me know, I'll not include it in the 2.6.27-stable tree then. greg k-h