From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mailout3.w1.samsung.com ([210.118.77.13]:63311 "EHLO mailout3.w1.samsung.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755712Ab2DTSYk (ORCPT ); Fri, 20 Apr 2012 14:24:40 -0400 MIME-version: 1.0 Content-transfer-encoding: 7BIT Content-type: text/plain; charset=ISO-8859-1 Date: Fri, 20 Apr 2012 20:24:38 +0200 From: Sylwester Nawrocki Subject: Re: s5p-fimc VIDIOC_STREAMOFF bug In-reply-to: <6126533.0OoG4qIlQU@flatron> To: Tomasz Figa Cc: linux-samsung-soc@vger.kernel.org, "linux-media@vger.kernel.org" Message-id: <4F91A9E6.3060402@samsung.com> References: <6126533.0OoG4qIlQU@flatron> Sender: linux-media-owner@vger.kernel.org List-ID: Hi Tomasz, On 04/19/2012 11:45 PM, Tomasz Figa wrote: > Hi, > > I have been working on adapting s5p-fimc driver for s3c6410 and everything > seems to be working just fine after some minor changes (except minor loss > of functionality - only codec path is supported, but for most use cases it > does not matter). > > However I think that I have spotted a bug in capture stop / capture suspend > handling. In fimc_capture_state_cleanup() the ST_CAPT_SUSPENDED status bit > of fimc->state field is being set regardless of suspend parameter, which > confuses the driver that FIMC is suspended and might not accept buffers > into active queue and so the driver will never start the capture process > unless the device gets closed and reopened (because of the condition > checking the count of active buffers). > > In my fork for s3c6410 I have moved the set_bit call into > fimc_capture_suspend(), so the bit gets set only when the device gets > suspended. This seems to solve the problem and I do not see any issues that > this could introduce, so it might be a good solution. > > Let me know if I am wrong in anything I have written. Your conclusions are correct, there was indeed a bug like that. There is already a patch fixing this [1], but it is going to be available just from v3.4. I'm considering sending it to Greg for inclusion in the stable releases, after it gets upstream. Once I did some preliminary work for the s3c-fimc driver, but dropped this due to lack of time. If you ever decide you want to mainline your work, just send the patches to linux-media@vger.kernel.org (and perhaps also to samsung-soc and ARM Linux) for review. > > Best regards, > Tomasz Figa [1] http://patchwork.linuxtv.org/patch/10417 Thanks, -- Sylwester Nawrocki Samsung Poland R&D Center