From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chen Baozi Subject: Question about mmc4 initialization on OMAP5432 Date: Wed, 09 Oct 2013 00:29:06 +0800 Message-ID: <525432D2.9010907@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from mail-pa0-f48.google.com ([209.85.220.48]:49931 "EHLO mail-pa0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755149Ab3JHQ3L (ORCPT ); Tue, 8 Oct 2013 12:29:11 -0400 Received: by mail-pa0-f48.google.com with SMTP id bj1so9083159pad.21 for ; Tue, 08 Oct 2013 09:29:10 -0700 (PDT) Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: linux-omap@vger.kernel.org Hi all, I have got a problem when booting dom0 kernel for Xen on OMAP5432 uEVM. With a recent patch, Xen hypervisor is avoid to map "disabled" device memory for dom0. In omap5-uevm.dts, we have got the &mmc4 node marked as "disabled". So the Xen won't populate <0x480d1000 0x400> memory region to the dom0. If dom0 is trying to access any memory address in this area, the hypervisor would be panic with a tranlation fault. However, it turns out that even the mmc4 is disabled in omap5-uevm.dts, Linux kernel is trying to access the 0x480d1010 when booting, which then leads a hypervisor panic as mentioned. The only point I know right now is that the 0x480d1010 is accessed by the omap_hwmod_read(). However, what makes me confused is how the dom0 kernel called omap_hwmod_read() to access 0x480d1010 while the mmc4 is disabled. I guess it might relate to hard coded omap_hwmod structure but have no idea what happen exactly when dom0 access the address due to the lack of calling stack information when booting Xen hypervisor. Thanks a lot. Chen Baozi