From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeremy Fitzhardinge Subject: Re: [PATCH 1/10] lguest: Don't rely on last-linked fallthru when no paravirt handler Date: Fri, 09 Feb 2007 12:49:06 -0800 Message-ID: <45CCDE42.2050603@goop.org> References: <1171012296.2718.26.camel@localhost.localdomain> <1171012458.2718.30.camel@localhost.localdomain> <200702091031.20249.ak@muc.de> <1171021966.2718.82.camel@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <1171021966.2718.82.camel@localhost.localdomain> Sender: linux-kernel-owner@vger.kernel.org To: Rusty Russell Cc: Andi Kleen , virtualization@lists.osdl.org, lkml - Kernel Mailing List , Andrew Morton , Jeremy Fitzhardinge List-Id: virtualization@lists.linuxfoundation.org Rusty Russell wrote: > If we make it thought early_fault, this will do just that. > > Given this is a "never happens" situation, however... if you're actually > under Xen or lguest, you won't make it that far (lguest, at least, will > kill you on the cr2 load in early_fault, but it doesn't matter because > we won't get anywhere with early_printk anyway). > > Actually, if we did BUG() here at least lguest would print something... > I wonder what Xen would do... Xen would print a complete register dump and backtrace. There's not a lot else you can do here; you could try early_printk, but if we're in a strange virtual environment, there may be no device on which the output could appear. J