Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S936709AbXFFXYe (ORCPT ); Wed, 6 Jun 2007 19:24:34 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S935647AbXFFXY0 (ORCPT ); Wed, 6 Jun 2007 19:24:26 -0400 Received: from lucidpixels.com ([75.144.35.66]:39725 "EHLO lucidpixels.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1763575AbXFFXYZ (ORCPT ); Wed, 6 Jun 2007 19:24:25 -0400 Date: Wed, 6 Jun 2007 19:24:24 -0400 (EDT) From: Justin Piszcz X-X-Sender: jpiszcz@p34.internal.lan To: Jesse Barnes cc: Andi Kleen , linux-kernel@vger.kernel.org, "Eric W. Biederman" Subject: Re: [PATCH] trim memory not covered by WB MTRRs In-Reply-To: <200706061620.10627.jesse.barnes@intel.com> Message-ID: References: <200706061229.24486.jesse.barnes@intel.com> <200706061528.44070.jesse.barnes@intel.com> <200706061620.10627.jesse.barnes@intel.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1547 Lines: 44 On Wed, 6 Jun 2007, Jesse Barnes wrote: > On Wednesday, June 6, 2007 3:57 pm Justin Piszcz wrote: >> On Wed, 6 Jun 2007, Jesse Barnes wrote: >>> On Wednesday, June 6, 2007 3:26 pm Justin Piszcz wrote: >>>> Nope, I booted with only netconsole= options. I have a lot of HW >>>> in the box and I guess the buffer is too small. Not sure where to >>>> change it in the kernel. Looking.. >>> >>> It's called "kernel log buffer size" and it's in "General setup". >>> >>> Jesse >> >> Did the dmesg output get you what you needed? Why the few KB >> difference? >> >> :) > > Yeah, looked at your e820 and your MTRR settings and I think my patch is > doing the right thing (i.e. trimming just the right amount of memory, > leaving you with as much as possible). > > The mem= approach though looks slightly off, but I haven't looked at > x86_64's mem= handling to see why. From a high level though, adjusting > end_pfn is the right thing to do, since theoretically mem= could choose > to make holes in your low memory and keep your high memory in the > allocation pools (though it's not generally implemented this way). > > Jesse > Ahh, ok! Sounds great, I will keep running the kernel with your patch without mem= and let you know if I see any issues. Chances of getting this into 2.6.22-rc5? Justin. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/