From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris Ball Subject: Re: SDHCI regression since 2.6.39 Date: Wed, 14 Sep 2011 09:48:21 -0400 Message-ID: References: <4E6EA3E4.5080401@goop.org> <4E6F8C24.3090207@goop.org> <4E6FA44F.1050705@goop.org> <4E6FBCA1.6040307@goop.org> <4E6FC422.9080306@goop.org> <7C3ED924-555D-4C50-9000-36077650328C@marvell.com> <4E6FCE28.5040403@goop.org> <4E6FDA0B.9060007@goop.org> <4E6FDC3E.8030702@goop.org> <003A4041-99F9-4598-AB9A-D08C76CDAECB@marvell.com> <4E6FFADE.5090004@goop.org> Mime-Version: 1.0 Content-Type: text/plain Return-path: Received: from void.printf.net ([89.145.121.20]:44189 "EHLO void.printf.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751076Ab1INNse (ORCPT ); Wed, 14 Sep 2011 09:48:34 -0400 In-Reply-To: <4E6FFADE.5090004@goop.org> (Jeremy Fitzhardinge's message of "Tue, 13 Sep 2011 17:52:46 -0700") Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: Jeremy Fitzhardinge Cc: Philip Rakity , "linux-mmc@vger.kernel.org" , Linux Kernel Mailing List , Manoj Iyer Hi Jeremy, On Tue, Sep 13 2011, Jeremy Fitzhardinge wrote: > Very strange. I reinstalled the original 2.6.38-based F15 kernel > (kernel-2.6.38.6-26.rc1.fc15.x86_64) which I'm sure did work, and it is > failing in the same way. So I guess some BIOS setting changed, or the > hardware itself has died? But it's still there; it does appear on the > PCI bus, and it does respond to the driver to some extent (register > dumps aren't all 0xff, etc). Long shot, but did you try removing power completely between the reboots? I wonder if the PCI config space write patch could be (a) bogus for your device, and (b) preserved across warm reboot by your PCI hardware. So, you could just try removing the battery and PSU, hitting the power button a few times with them unplugged to discharge caps, and then boot into the old kernel again. Thanks, - Chris. -- Chris Ball One Laptop Per Child