From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: 4430SDP boot failure Date: Thu, 6 Jan 2011 10:25:56 -0800 Message-ID: <20110106182555.GW7771@atomide.com> References: <20110106170805.GE1198@n2100.arm.linux.org.uk> <20110106180030.GA8249@n2100.arm.linux.org.uk> <20110106181754.GA8975@n2100.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mho-02-ewr.mailhop.org ([204.13.248.72]:46934 "EHLO mho-02-ewr.mailhop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751686Ab1AFS0H (ORCPT ); Thu, 6 Jan 2011 13:26:07 -0500 Content-Disposition: inline In-Reply-To: <20110106181754.GA8975@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-omap@vger.kernel.org * Russell King - ARM Linux [110106 10:17]: > On Thu, Jan 06, 2011 at 06:00:30PM +0000, Russell King - ARM Linux wrote: > > Therefore, I believe the debugging address calculation stuff in > > arch/arm/mach-omap2/include/mach/debug-macros.S to be rather broken, > > causing an abort when it tries to access the serial port. > > I'm sorry, I think this crap has to go. Trying to make it fit every > platform in one kernel is just too complicated. It _needs_ to be > something so damned simple that it always works without fail, every > single time someone wants to use it. > > The current code really hasn't a hope of doing that. > > Yes, that means you only configure it when you're not building a multi- > platform kernel, but that's worth giving up if in return we have something > that is 100% reliable. > > As it is, I need to invent new debugging code to debug this mess. Sorry, > I'm actually not going to bother - as the LL debug stuff was my debugging > code to debug exactly these kinds of boot problems. Someone else can have > the pain of sorting out this breakage, and tell me when they expect OMAP > to work again. Something that could help with the DEBUG_LL code is to pass optional debug UART info masked into the machine_id using the high bits.. Regards, Tony