From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753051Ab3FZWto (ORCPT ); Wed, 26 Jun 2013 18:49:44 -0400 Received: from mail-pb0-f52.google.com ([209.85.160.52]:35778 "EHLO mail-pb0-f52.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752023Ab3FZWtn (ORCPT ); Wed, 26 Jun 2013 18:49:43 -0400 Date: Wed, 26 Jun 2013 15:49:51 -0700 From: Olof Johansson To: Linus Walleij Cc: "linux-arm-kernel@lists.infradead.org" , "linux-kernel@vger.kernel.org" Subject: Re: [PATCH] ARM: ux500: check platform in ux500_l2x0_init Message-ID: <20130626224951.GA911@quad.lixom.net> References: <1372210231-22900-1-git-send-email-olof@lixom.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, On Wed, Jun 26, 2013 at 09:46:08PM +0200, Linus Walleij wrote: > On Wed, Jun 26, 2013 at 3:30 AM, Olof Johansson wrote: > > > Add compatible checks for ux500. This should probably be moved to generic > > DT-based probing whenever DT is populated instead, but until then at > > least don't panic all non-ux500 multiplatform machines. > > > > Signed-off-by: Olof Johansson > > > + /* Ick. I wish there was a single compatible value to compare for, > > + * but there isn't. > > + */ > > + if (!of_machine_is_compatible("st-ericsson,u8500") && > > + !of_machine_is_compatible("st-ericsson,u8540") && > > + !of_machine_is_compatible("st-ericsson,u9500") && > > + !of_machine_is_compatible("st-ericsson,u8540")) > > + return -ENODEV; > > This is checking u8540 twice and does not account for snowball. Snowball has a compatible of: "calaosystems,snowball-a9500", "st-ericsson,u9500" but yeah, I typoed on the 9540. > With new platforms coming we cannot just line up all possible > compat strings like this, we need to rely on the cpu_is* macro > to be reporting something sane, and I think it does. Agreed. > What about just doing this: > > From da838c71fb32d63d7e28c14697702ff48088a09b Mon Sep 17 00:00:00 2001 > From: Linus Walleij > Date: Wed, 26 Jun 2013 21:44:57 +0200 > Subject: [PATCH] ARM: ux500: bail out on alien cpus > > This makes the l2x0 initialization fail gracefully on non-ux500 > systems. > > Signed-off-by: Linus Walleij > --- > arch/arm/mach-ux500/cache-l2x0.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/arch/arm/mach-ux500/cache-l2x0.c b/arch/arm/mach-ux500/cache-l2x0.c > index f58615b..82ccf1d 100644 > --- a/arch/arm/mach-ux500/cache-l2x0.c > +++ b/arch/arm/mach-ux500/cache-l2x0.c > @@ -42,7 +42,8 @@ static int __init ux500_l2x0_init(void) > if (cpu_is_u8500_family() || cpu_is_ux540_family()) > l2x0_base = __io_address(U8500_L2CC_BASE); > else > - ux500_unknown_soc(); > + /* Non-Ux500 platform */ > + return -ENODEV; > > /* Unlock before init */ > ux500_l2x0_unlock(); SGTM. Applied to next/late. -Olof