From mboxrd@z Thu Jan 1 00:00:00 1970 From: Linus Torvalds Subject: Re: 2.6.19-rc5: known regressions Date: Wed, 8 Nov 2006 07:47:07 -0800 (PST) Message-ID: References: <20061108085235.GT4729@stusta.de> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-pm-bounces@lists.osdl.org Errors-To: linux-pm-bounces@lists.osdl.org To: "Eric W. Biederman" Cc: Andrew Morton , Len Brown , gregkh@suse.de, Thierry Vignaud , discuss@x86-64.org, Ernst Herzberg , Aaron Durbin , Olivier Mondoloni , Alex Romosan , ak@suse.de, linux-ide@vger.kernel.org, oprofile-list@lists.sourceforge.net, Jens Axboe , Thomas Gleixner , Arjan van de Ven , Bryan O'Sullivan , linux-pm@osdl.org, Tim Chen , Jeff Chua , Paolo Ornati , phil.el@wanadoo.fr, Matthew Wilcox , Linux Kernel Mailing List , Adrian Bunk , linux-acpi@vger.kernel.org, mingo@redhat.comK List-Id: linux-ide@vger.kernel.org On Wed, 8 Nov 2006, Eric W. Biederman wrote: > = > I haven't seen anyone reproduce this but Tim Chen, and Tim wasn't > able to root cause the problem so I believe we are going to have > this regression :( Note that you really shouldn't look too closely at lmbench scheduling = fluctuations. They can fluctuate a _lot_, especially under SMP, and it can = depend on things like cache layout that has nothing to do with the = scheduler (ie just code movement can make the lmbench numbers change). So there are "regressions" and there are "shit happens". It can sometimes = be hard to tell the two apart, of course ;) Linus