From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ulf Hansson Subject: Re: [PATCH] mmc: core: Add option to prevent eMMC sleep command Date: Wed, 14 Dec 2011 17:48:17 +0100 Message-ID: <4EE8D351.50508@stericsson.com> References: <1323876036-10206-1-git-send-email-ulf.hansson@stericsson.com> <000601ccba74$afdb0180$0f910480$@org> <4EE8C1FB.5080205@stericsson.com> <000c01ccba76$787f7cf0$697e76d0$@org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from eu1sys200aog106.obsmtp.com ([207.126.144.121]:53723 "EHLO eu1sys200aog106.obsmtp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755146Ab1LNQtJ (ORCPT ); Wed, 14 Dec 2011 11:49:09 -0500 In-Reply-To: <000c01ccba76$787f7cf0$697e76d0$@org> Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: Subhash Jadavani Cc: "linux-mmc@vger.kernel.org" , 'Chris Ball' , Per FORLIN , Johan RUDHOLM , 'Lee Jones' Subhash Jadavani wrote: > >> -----Original Message----- >> From: Ulf Hansson [mailto:ulf.hansson@stericsson.com] >> Sent: Wednesday, December 14, 2011 9:04 PM >> To: Subhash Jadavani >> Cc: linux-mmc@vger.kernel.org; 'Chris Ball'; Per FORLIN; Johan RUDHOLM; >> 'Lee Jones' >> Subject: Re: [PATCH] mmc: core: Add option to prevent eMMC sleep >> command >> >> Subhash Jadavani wrote: >>> Why is this patch needed? Do you see any issue with eMMC sleep on >> particular >>> cards? >>> >>> Regards, >>> Subhash >> Suppose the eMMC is put in sleep state when doing suspend, at that >> point >> the kernel crashes. The kernel crash is supposed to be dumped to the >> same eMMC by some other software than the kernel itself. But this >> software do not have support for waking up the eMMC from sleep and is >> thus not able to store the dump. > > Ok. This means that SW still has some support to read/write eMMC. As part of > suspend, we also deselect the card. So does that SW is capable of > re-selecting the card and the do read/write transfer? I know this discussion > might not be required. But still it's good to know. Sorry, but I can't go into too much details, since this software is proprietary. :-) > > As far as patch is concerned, it looks good. You might want to mention the > motivation for this patch in commit text. I fix in a V2 patch. > >> Br >> Ulf Hansson > >