From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sujit Reddy Thumma Subject: Re: [PATCH] mmc: core: Prevent too long response times for suspend Date: Wed, 19 Oct 2011 22:55:07 +0530 Message-ID: <4E9F07F3.2050103@codeaurora.org> References: <1318514638-32452-1-git-send-email-ulf.hansson@stericsson.com> <4f62290baad48cd8655f3f82f7b162b9.squirrel@www.codeaurora.org> <4E9BFA84.1040709@stericsson.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from wolverine01.qualcomm.com ([199.106.114.254]:11824 "EHLO wolverine01.qualcomm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751469Ab1JSRZO (ORCPT ); Wed, 19 Oct 2011 13:25:14 -0400 In-Reply-To: <4E9BFA84.1040709@stericsson.com> Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: Ulf Hansson Cc: "linux-mmc@vger.kernel.org" , Chris Ball , Per FORLIN , Lee Jones On 10/17/2011 3:21 PM, Ulf Hansson wrote: >> >> Why would there be pending requests while host is suspending? Is the >> kernel framework not handling sync before going to suspend? However, the >> mmc_blk_suspend() would be called before the host driver suspends (as all >> the driver suspend routines are serialized) which means it stops block >> layer to queue more I/O requests well before the host driver start >> suspend. Does this sequence break in your case? > > I have observed this issue for different cases (one case was logging to > eMMC). The idea is simply that we would like to be sure that we do not > wait "forever", no matter if the "upper layers" misbehaved in the > suspend sequence. > Agree. I was just curious to know why would there be a pending requests. If the issue is seen while logging to eMMC then there may be something wrong, probably, we may see issue while running mmc_test and initiating suspend. Your patch looks good to me though. Reviewed-by: Sujit Reddy Thumma >> >> Your concern seems to be valid for SDIO case, but again the function >> driver must be intelligent enough to return -EBUSY as it knows that it >> had >> posted a request to MMC. >> > > For SDIO, should we really assume that function driver has implemented a > suspend function and moreover that it actually always behaves as we expect? I hope so, but yes, in some cases that may be broken. I agree with you. > > Br > Ulf Hansson -- Thanks & Regards, Sujit Reddy Thumma