From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751082AbXACTY1 (ORCPT ); Wed, 3 Jan 2007 14:24:27 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751089AbXACTY1 (ORCPT ); Wed, 3 Jan 2007 14:24:27 -0500 Received: from gate.crashing.org ([63.228.1.57]:53180 "EHLO gate.crashing.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751090AbXACTY0 (ORCPT ); Wed, 3 Jan 2007 14:24:26 -0500 In-Reply-To: <411976.37299.qm@web50112.mail.yahoo.com> References: <411976.37299.qm@web50112.mail.yahoo.com> Mime-Version: 1.0 (Apple Message framework v623) Content-Type: text/plain; charset=US-ASCII; format=flowed Message-Id: Content-Transfer-Encoding: 7bit Cc: linux-kernel@vger.kernel.org From: Segher Boessenkool Subject: Re: Subject: [PATCH 2/2] EDAC: K8 Memory scrubbing patch Date: Wed, 3 Jan 2007 20:24:59 +0100 To: Doug Thompson X-Mailer: Apple Mail (2.623) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > One would also expect that cache scrubbing requires hardware > support. You could just write-back invalidate the whole cache periodically if there is no explicit hardware support for cache scrubbing. Less efficient, sure :-) > Signed-off-by: doug thompson Shouldn't you spell your name correctly (with capitalisation) in the sign off? It being a formal thing and all. Some spelling and coding style nits: > +/* Valid scrub rates for the K8 hardware memory scrubber. We map > + maps the scrubbing bandwith to a valid bit pattern. The 'set' "map maps" > + Currently, we only do scrubbing of sdram - the caches are assumed > + to be excercised always by running code and if the scrubber is done "excercised" > + search for the bandwith that is eq or gt than the "bandwith" And please just write "greater or equal". > + for (i=0; scrubrates[i].bandwidth != SDRATE_EOD; i++) { i = 0; > + /* find the bandwith matching the memory scrubber configuration "bandwith" again > + for (i=0; scrubrates[i].bandwidth != SDRATE_EOD; i++) { i = 0; > + /* the bit pattern is invalid - we might fix it > + by applying the slowest scrub rate as this is > + closest to the valid value, but we do not! Why not? > + if (scrubrates[i].bandwidth == SDRATE_EOD) { > + edac_printk(KERN_WARNING, EDAC_MC, > + "Invalid sdram scrub control value: %d \n", Space before the newline. > /* FIXME - stolen from msr.c - the calls in msr.c could be exported */ So fix it :-) Segher