From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267625AbUHUScc (ORCPT ); Sat, 21 Aug 2004 14:32:32 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267630AbUHUScb (ORCPT ); Sat, 21 Aug 2004 14:32:31 -0400 Received: from ppp2-adsl-157.the.forthnet.gr ([193.92.233.157]:28455 "EHLO ppp1-100.the.forthnet.gr") by vger.kernel.org with ESMTP id S267625AbUHUScA (ORCPT ); Sat, 21 Aug 2004 14:32:00 -0400 From: V13 To: "Barry K. Nathan" Subject: Re: Possible dcache BUG Date: Sat, 21 Aug 2004 21:31:31 +0300 User-Agent: KMail/1.7 Cc: Gene Heskett , linux-kernel@vger.kernel.org, Marc Ballarin References: <200408201608.51038.gene.heskett@verizon.net> <20040821092556.GA14991@ip68-4-98-123.oc.oc.cox.net> In-Reply-To: <20040821092556.GA14991@ip68-4-98-123.oc.oc.cox.net> MIME-Version: 1.0 Content-Type: multipart/signed; boundary="nextPart12569762.SpULFajyao"; protocol="application/pgp-signature"; micalg=pgp-sha1 Content-Transfer-Encoding: 7bit Message-Id: <200408212131.38019.v13@priest.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org --nextPart12569762.SpULFajyao Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline On Saturday 21 August 2004 12:25, Barry K. Nathan wrote: > > Memtest86 may not know howto enable it if its an > > nforce2 option. Whatever cache shown as switchable in the bios, > > turning it off makes a very sick bird out of the machine, like a > > 33mhz 386sx? > > Yeah, disabling the L2 cache on a modern CPU makes it really slow. But, > it's still a useful troubleshooting option... When I had the problem described in my previous mail I came to the concluss= ion=20 that it was related with cache *BUT* it seemed that the cache was just=20 caching wrong data. Disabling the cache would just reduce the problem. One reason for this is that when the program detected errors in a buffer (i= =2Ee.=20 0x1234 instead of 0x1111) then they would NOT go away if the program was=20 reading from this buffer all the time. This means that the cache always=20 returned the same data. The error was 'gone' every time the program was=20 suspended for a while or when something else used a lot of memory (i.e.=20 another instance of this program). So, I'm not suggesting that his cache is faulty but that there can be a CPU= =20 (or even a M/B) problem that corrupts data when they are transfered from=20 memory to the processor. > -Barry K. Nathan <> --nextPart12569762.SpULFajyao Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) iD8DBQBBJ5UJVEjwdyuhmSoRAig6AJ904SwtzHs4j+kPwwbp5+l4REVfZwCgs0N/ 1Q6uq26BVoQ54jmr8MdjfJM= =Ka+W -----END PGP SIGNATURE----- --nextPart12569762.SpULFajyao--