From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754698AbaBRJYe (ORCPT ); Tue, 18 Feb 2014 04:24:34 -0500 Received: from opensource.wolfsonmicro.com ([80.75.67.52]:43545 "EHLO opensource.wolfsonmicro.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754473AbaBRJYb (ORCPT ); Tue, 18 Feb 2014 04:24:31 -0500 Date: Tue, 18 Feb 2014 09:24:29 +0000 From: Charles Keepax To: Mark Brown Cc: alsa-devel@alsa-project.org, lars@metafoo.de, eric.y.miao@gmail.com, dmitry.torokhov@gmail.com, lgirdwood@gmail.com, haojian.zhuang@gmail.com, linux-kernel@vger.kernel.org, peter.ujfalusi@ti.com, cw00.choi@samsung.com, myungjoo.ham@samsung.com, patches@opensource.wolfsonmicro.com, linux@arm.linux.org.uk, jarkko.nikula@bitmer.com Subject: Re: [alsa-devel] [PATCH 04/15] ASoC: wm5100: Update locking around use of DAPM pin API Message-ID: <20140218092429.GB18502@opensource.wolfsonmicro.com> References: <1392655903-24537-1-git-send-email-ckeepax@opensource.wolfsonmicro.com> <1392655903-24537-5-git-send-email-ckeepax@opensource.wolfsonmicro.com> <20140218004729.GK2669@sirena.org.uk> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20140218004729.GK2669@sirena.org.uk> User-Agent: Mutt/1.5.17+20080114 (2008-01-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Feb 18, 2014 at 09:47:29AM +0900, Mark Brown wrote: > On Mon, Feb 17, 2014 at 04:51:32PM +0000, Charles Keepax wrote: > > > + snd_soc_dapm_force_enable_pin_locked(&codec->dapm, "CP2"); > > + snd_soc_dapm_force_enable_pin_locked(&codec->dapm, "SYSCLK"); > > + > > + mutex_unlock(&codec->dapm.card->dapm_mutex); > > + > > snd_soc_dapm_sync(&codec->dapm); > > With all these patches it seems weird that we have to drop the lock to > do the sync which will immediately retake it. It's not broken but it > looks off - it would be better to have a version of _sync() that we can > call within the lock. No problem to add a version of sync that can be called from within the lock, should help out with Dimtry's comments as well. > > Regarding the naming issue that Lars mentioned I think the current > operations are probably fine but calling them _unlocked() meaning they > don't do any locking (as distinct from the existing _locked() which take > locks) might be OK. Yeah that would be good, since my original aim here was to avoid updating every single usage of these functions. Thanks, Charles