From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Vrabel Subject: Re: [PATCH 5/5] SDIO: put active devices into 1-bit mode during suspend Date: Wed, 10 Feb 2010 16:03:24 +0000 Message-ID: <4B72D8CC.9030605@csr.com> References: <1265747366-6676-1-git-send-email-nico@fluxnic.net> <1265747366-6676-6-git-send-email-nico@fluxnic.net> <4B7298EF.4070604@csr.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from cluster-g.mailcontrol.com ([208.87.233.190]:37651 "EHLO cluster-g.mailcontrol.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755279Ab0BJQDh (ORCPT ); Wed, 10 Feb 2010 11:03:37 -0500 Received: from rly30g.srv.mailcontrol.com (localhost.localdomain [127.0.0.1]) by rly30g.srv.mailcontrol.com (MailControl) with ESMTP id o1AG3al6028244 for ; Wed, 10 Feb 2010 16:03:37 GMT Received: from localhost.localdomain (localhost.localdomain [127.0.0.1]) by rly30g.srv.mailcontrol.com (MailControl) id o1AG3TVW027953 for ; Wed, 10 Feb 2010 16:03:29 GMT In-Reply-To: Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: Chris Ball Cc: Nicolas Pitre , Andrew Morton , linux-mmc@vger.kernel.org Chris Ball wrote: > Hi David, > > > Why? Is this for a specific card that doesn't support > > asynchronous card interrupts while in 4-bit mode? If so, I'd > > suggest a card quirk for this and only switch modes for these > > cards. > > (As a check that we're on the same page -- sdio_disable_wide() is only > called if MMC_PM_KEEP_POWER is true, which means the system is going > to be suspended but the card will remain externally powered without > its clock.) > > Under the circumstances above, it's a requirement of the SDIO spec > that moving to 1-bit mode must happen before stopping the clock if > the interrupt mechanism is to be used to wake the host, and the > chipset we've been using (Via VX855) fails to generate a wakeup if > the switch isn't made. Do you have experience of another chipset > that doesn't need the switch to 1-bit mode for a clockless wakeup? I couldn't find anything in the SDIO specification (you know the section number?). Part A2 has a requirement that the controller be placed into 1-bit mode but this doesn't mean that the card has to be. Maybe it's not really a big deal, a single command only costs a few 10s of microseconds anyway. David -- David Vrabel, Senior Software Engineer, Drivers CSR, Churchill House, Cambridge Business Park, Tel: +44 (0)1223 692562 Cowley Road, Cambridge, CB4 0WZ http://www.csr.com/ Member of the CSR plc group of companies. CSR plc registered in England and Wales, registered number 4187346, registered office Churchill House, Cambridge Business Park, Cowley Road, Cambridge, CB4 0WZ, United Kingdom