From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751824Ab3LQGZO (ORCPT ); Tue, 17 Dec 2013 01:25:14 -0500 Received: from mx1.redhat.com ([209.132.183.28]:63956 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750930Ab3LQGZM (ORCPT ); Tue, 17 Dec 2013 01:25:12 -0500 Date: Tue, 17 Dec 2013 14:24:17 +0800 From: Dave Young To: Matt Fleming Cc: linux-kernel@vger.kernel.org, linux-efi@vger.kernel.org, x86@kernel.org, mjg59@srcf.ucam.org, hpa@zytor.com, James.Bottomley@HansenPartnership.com, vgoyal@redhat.com, ebiederm@xmission.com, horms@verge.net.au, kexec@lists.infradead.org, bp@alien8.de, greg@kroah.com, toshi.kani@hp.com, akpm@linux-foundation.org, mingo@kernel.org, msalter@redhat.com, leif.lindholm@linaro.org Subject: Re: [PATCH v6 14/14] x86: kdebugfs do not use __va for getting setup_data virt addr Message-ID: <20131217062417.GE6751@dhcp-16-126.nay.redhat.com> References: <1387186235-15972-1-git-send-email-dyoung@redhat.com> <1387186235-15972-15-git-send-email-dyoung@redhat.com> <20131216163501.GB3873@console-pimps.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20131216163501.GB3873@console-pimps.org> 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 On 12/16/13 at 04:35pm, Matt Fleming wrote: > On Mon, 16 Dec, at 05:30:35PM, Dave Young wrote: > > kdump kernel will use memmap=exactmap kernel cmdline, but __va does not > > work in case memmap=exactmap, so let's always use ioremap_cache. > > > > Signed-off-by: Dave Young > > --- > > arch/x86/kernel/kdebugfs.c | 35 +++++++++++------------------------ > > 1 file changed, 11 insertions(+), 24 deletions(-) > > Dave, I've no idea why this change is necessary from the commit log. Is > it required for kexec to function on EFI? Why does __va() not work in > the memmap=exactmap case? > During previous kdump tests I saw panics while reading the setup_data in debugfs. I thought it is caused by some unmapped addresses. At that time I also reproduced it by booting the non-kexec kernel with memmap=exact. Since you are asking about this I'm testing it again but I seems can not reproduce this problem any more, it's weird. I should dug more about it and save the panic messages. So let's drop this patch for now, I will keep an eye on this and address it later if I can find the problem again. -- Thanks Dave