Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932100AbWHKLIi (ORCPT ); Fri, 11 Aug 2006 07:08:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932107AbWHKLIh (ORCPT ); Fri, 11 Aug 2006 07:08:37 -0400 Received: from nf-out-0910.google.com ([64.233.182.186]:33305 "EHLO nf-out-0910.google.com") by vger.kernel.org with ESMTP id S932100AbWHKLIh (ORCPT ); Fri, 11 Aug 2006 07:08:37 -0400 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=EpAydcoF02G3AHwtP8Bn3Yl6SzHT1G7a7ts6s9OFC2Uryv+o5gANDRD2Mffg73/n9C5NVJyjp6XWl+3br87nz7CaA3jsA6XX/Hz6JhsNfN+iGm0eNzrlqOqUWCzRubHnWUasdbVsiNPcUnF9v7JgNIy7nko0/2VW0/nPuI+jECQ= Message-ID: Date: Fri, 11 Aug 2006 11:08:32 +0000 From: "gmu 2k6" To: "Simen Thoresen" Subject: Re: Only 3.2G ram out of 4G seen in an i386 box Cc: "Joel Jaeggli" , linux-kernel@vger.kernel.org In-Reply-To: <44DBA06E.40104@dolphinics.no> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <20060808101504.GJ2152@stingr.net> <20060809074815.bec7f32c.joelja@uoregon.edu> <44DBA06E.40104@dolphinics.no> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3103 Lines: 67 On 8/10/06, Simen Thoresen wrote: > gmu 2k6 wrote: > ... > > so what does it mean that one of Xeons here shows me the full 4GiB as > > total physical memory via `free`? > > > > btw, the box I'm getting will have the 975X chip and include 4GiB RAM > > and if I understood the problem correctly even with 3GiB there will be > > some memory lost to mapping-issues besides the HI/LO mem > > kernel-reserving issue. > > this is what I get on ia32 P4 with 3GiB > > total used free shared buffers cached > > Mem: 3116108 2608196 507912 0 246652 2039708 > > and this is what I get on ia32 Xeon with 4GiB > > total used free shared buffers cached > > Mem: 4087660 268828 3818832 0 57168 103568 > > Hi Gmu, > > This is a bit divergent from the discussion, but will hopefully provide some > background. > > The hardware my employer manufactures asks the BIOS to reserve several > largeish ranges (16M and 16M to 512M), and it's been used in various HPC > projects where machines typically have a lot of memory installed. > > From my experience, most pre 750x-P4-Xeon chipsets did not support any kind > of memory remapping, so often the top 0.5-1.5G of ram would be 'lost'. Great > fun. > Since then, this functionality has started appearing, more or less randomly. > I believe it should be present on all later 75xx chipsets (probably the Core > 2 Xeon chipsets as well), but implementation depends on the BIOS (ie the > motherboard vendors), and their grasp of what people do with their boards is > often quite limited. > > On the P4 boards (and again the Core 2 boards as well), this is probably > even less ordered as they are 'desktop' boards rather than the more costly > 'server' boards. > > I'm pretty sure the Athlon64 and Opteron CPUs should support this, but they > too depend on BIOS functionality to rearrange the memory tables. > > Just out of interest - what chipset does your Xeon system use? lspci of the P4 32bit desktop with 3GiB 00:00.0 Host bridge: Intel Corporation 82875P/E7210 Memory Controller Hub (rev 02) 00:01.0 PCI bridge: Intel Corporation 82875P Processor to AGP Controller (rev 02) 00:1d.0 USB Controller: Intel Corporation 82801EB/ER (ICH5/ICH5R) USB UHCI Controller #1 (rev 02) lspci of the Xeon 32bit HP Proliant Server with 4GiB: 00:00.0 Host bridge: Intel Corporation E7520 Memory Controller Hub (rev 0c) 00:02.0 PCI bridge: Intel Corporation E7525/E7520/E7320 PCI Express Port A (rev 0c) 00:06.0 PCI bridge: Intel Corporation E7520 PCI Express Port C (rev 0c) 00:1d.0 USB Controller: Intel Corporation 82801EB/ER (ICH5/ICH5R) USB UHCI Controller #1 (rev 02) lspci of the new box with 975X and Core 2 Duo not available yet for obvious reasons (I don't have the the box yet). - 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/