From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sudip Mukherjee Subject: Re: [PATCH] sound: oss: ad1848: Fix returned errno code in ad1848_init() Date: Wed, 23 Sep 2015 11:56:58 +0530 Message-ID: <20150923062658.GE4010@sudip-pc> References: <1442936257-8988-1-git-send-email-luisbg@osg.samsung.com> <20150922164642.GA14353@sudip-pc> <5601A7D1.3010008@osg.samsung.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <5601A7D1.3010008@osg.samsung.com> Sender: linux-kernel-owner@vger.kernel.org To: Luis de Bethencourt Cc: linux-kernel@vger.kernel.org, perex@perex.cz, tiwai@suse.com, alsa-devel@alsa-project.org List-Id: alsa-devel@alsa-project.org On Tue, Sep 22, 2015 at 08:11:13PM +0100, Luis de Bethencourt wrote: > On 22/09/15 17:46, Sudip Mukherjee wrote: > > On Tue, Sep 22, 2015 at 04:37:37PM +0100, Luis de Bethencourt wrote: > > Hi Sudip, > > True! I missed that. > > This change will mean the block for 'if (hw_config->slots[0] != -1)' will > run and it shouldn't. > > In ad1848.c:1998 sound_install_audiodrv() can return -ENOMEM as well, but this > is turned into -1 also. All errno codes are ignored in sound/oss/pss.c, not > worth it. > > I'm withdrawing my patch. Sorry for this. > > Is there any other way to silence the smatch warning? Yes, ad1848_init() is only called from 3 places and everytime the return is stored in hw_config->slots[0]. So change in all these 3 places and then slots[0] is being checked in sound_unload_audiodev() and also in sound/oss/pss.c, so you need to modify those checks also. But these are all just from a quick grep, if you really want to change then you need to see the driver carefully for other places that i might have missed. regards sudip