From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH 0/4] Hacks to allow booting ARM SMP kernel on UP ARMv7 Date: Mon, 6 Sep 2010 11:03:47 -0700 Message-ID: <20100906180346.GC20849@atomide.com> References: <20100817104414.19061.38999.stgit@baageli.muru.com> <20100906104413.GD20903@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]:49703 "EHLO mho-01-ewr.mailhop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754134Ab0IFSEB (ORCPT ); Mon, 6 Sep 2010 14:04:01 -0400 Content-Disposition: inline In-Reply-To: <20100906104413.GD20903@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.infradead.org, linux-omap@vger.kernel.org, bryan.wu@canonical.com * Russell King - ARM Linux [100906 03:36]: > Here's my latest patch (which is combined from two patches.) > > Tony, could you follow up with patches for anything which is still > required - I think there's two things you've addressed which this > currently misses: > > 1. not initializing twd_base (I'm not convinced this is safe - rather > making smp_prepare_cpus() return early is probably a better idea.) > 2. __flush_icache_all() > > Note that (2) seems to be complicated by the instruction only being > available on ARMv7 and later. Yeah will check probably on Tuesday, heading to a BBQ as it's a holiday here. Also my laptop cooling fan broke a few days ago, got it patched up temporarily though: An external fan duct taped to the side of the machine until the replacement arrives :) > Also note that this should only be used for comparing SMP vs UP versions > of the same architecture - in other words, not ARMv6 vs ARMv6K. Right, ARMv6 vs ARMv6K can be dealt with in another patch once this is done. > Lastly, what's happening about ARMv6 and ARMv7 processor setup functions? > Are we going to split them into those which need SMP bits twiddled and > those which don't? If we are going to split them, someone needs to follow > up on this email with patches to do it (preferably to be slotted in > before these changes.) No need to split them any longer AFAIK. Or do you have some case in mind that's not being handled now, maybe I don't follow you? Regards, Tony