From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 8 Aug 2002 15:15:09 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 8 Aug 2002 15:15:09 -0400 Received: from holomorphy.com ([66.224.33.161]:45722 "EHLO holomorphy") by vger.kernel.org with ESMTP id ; Thu, 8 Aug 2002 15:15:08 -0400 Date: Thu, 8 Aug 2002 12:18:51 -0700 From: William Lee Irwin III To: "Martin J. Bligh" Cc: linux-kernel@vger.kernel.org Subject: Re: further IO-APIC oddities Message-ID: <20020808191851.GE15685@holomorphy.com> Mail-Followup-To: William Lee Irwin III , "Martin J. Bligh" , linux-kernel@vger.kernel.org References: <20020808162856.GD6256@holomorphy.com> <70720000.1028829388@flay> <20020808180743.GD15685@holomorphy.com> <73720000.1028830446@flay> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Description: brief message Content-Disposition: inline In-Reply-To: <73720000.1028830446@flay> User-Agent: Mutt/1.3.25i Organization: The Domain of Holomorphy Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org At some point in the past, I wrote: >> It's different from 2.5.29, I can follow up with that. 2.5.29 saw all 0's, >> so whatever it was that was scribbling over the MPC table and making the >> ID's all 0, it's scribbling on something else now (probably mem_map). On Thu, Aug 08, 2002 at 11:14:06AM -0700, Martin J. Bligh wrote: > I thought that was only if you reduced NR_CPUS? And you shouldn't > need to read the IOAPIC tables to do that - the basic array was getting > overwritten. Or am I confusing at least two different bugs? Sorry, I'm being obtuse. What I had in mind here was: (1) there's been a longstanding bug where the MPC table entries with the IO-APIC ID's gets zeroed out (2) this bug mysteriously went away, even though we never tracked it down (3) bugs that smell of mem_map getting stomped on are cropping up ... and I suspect these three things are related, e.g. the MPC stuff moved or link order changed or some such Heisenbug-ish nonsense and now it's hitting something else. The NR_CPUS bit isn't even bootable (the panic() happens before console_init() IIRC) so it's not quite that one. =) Cheers, Bill