From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964828AbVJDPnm (ORCPT ); Tue, 4 Oct 2005 11:43:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S964827AbVJDPnm (ORCPT ); Tue, 4 Oct 2005 11:43:42 -0400 Received: from ebiederm.dsl.xmission.com ([166.70.28.69]:20444 "EHLO ebiederm.dsl.xmission.com") by vger.kernel.org with ESMTP id S964826AbVJDPnl (ORCPT ); Tue, 4 Oct 2005 11:43:41 -0400 To: Andi Kleen Cc: Andrew Morton , linux-kernel@vger.kernel.org, fastboot@osdl.org Subject: Re: [PATCH 2/2] x86_64: move apic init in init_IRQs References: <200510041724.36535.ak@suse.de> From: ebiederm@xmission.com (Eric W. Biederman) Date: Tue, 04 Oct 2005 09:42:12 -0600 In-Reply-To: <200510041724.36535.ak@suse.de> (Andi Kleen's message of "Tue, 4 Oct 2005 17:24:36 +0200") Message-ID: User-Agent: Gnus/5.1007 (Gnus v5.10.7) Emacs/21.4 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Andi Kleen writes: > On Tuesday 04 October 2005 17:13, Eric W. Biederman wrote: >> All kinds of ugliness exists because we don't initialize >> the apics during init_IRQs. >> - We calibrate jiffies in non apic mode even when we are using apics. >> - We have to have special code to initialize the apics when non-smp. >> - The legacy i8259 must exist and be setup correctly, even >> we won't use it past initialization. > > Actually some setups use its timer even in ACPI mode. Do you take this into > account? The patch is basically code motion and duplication removal with just a few tweaks thrown in to account for the change in location in the kernel boot process when all of occurs. So everything that was working should continue to work. > In theory it looks reasonable, but I guess it won't work without the bogus > watchdog change. That would need to be resolved first. Sure. That should be straight forward. The only part that breaks is the watchdog initialization/testing. As I recall there was no obvious place to put it. The call has to happen after everything is running so it is a bit tricky to find a good place to put it. Eric