From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH 1/5] ARM: OMAP2/3: Remove OMAP2_32KSYNCT_BASE Date: Thu, 28 May 2009 08:54:31 -0700 Message-ID: <20090528155431.GC32453@atomide.com> References: <20090514231820.19331.65419.stgit@localhost> <20090514232306.19331.92192.stgit@localhost> <20090528152517.GB20736@n2100.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mho-01-ewr.mailhop.org ([204.13.248.71]:60663 "EHLO mho-01-ewr.mailhop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754554AbZE1Pyl (ORCPT ); Thu, 28 May 2009 11:54:41 -0400 Content-Disposition: inline In-Reply-To: <20090528152517.GB20736@n2100.arm.linux.org.uk> Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: Russell King - ARM Linux Cc: linux-arm-kernel@lists.arm.linux.org.uk, linux-omap@vger.kernel.org * Russell King - ARM Linux [090528 08:25]: > On Thu, May 14, 2009 at 04:23:06PM -0700, Tony Lindgren wrote: > > Use processor specific defines instead. > > > > As an extra bonus, this patch fixes the problem of CONFIG_DEBUG_SPINLOCK > > calling sched_clock before we have things initialized: > > Hmm. I don't think sched_clock peculiarities should be handed by the > clock source implementation. > > This is really a generic kernel bug: lots of ARM architecture sched_clock > implementations can not be called before things like IO memory has been > setup - and that's something that mainline people have been made aware > of repeatedly, and they've fixed the problems when they've come up. Yeah agreed. > Moreover, the issue of using clocksources for sched_clock is being > discussed on the main ARM lists, with the possibility of the generic > kernel-wide sched_clock implementation being based upon clock sources. > > Whether that changes anything with this patch, I'm not entirely sure. The clocksource and sched_clock thing is a separate issue, this patch is still needed to remove the conflicting defines. Tony