From: Mark Brown Subject: Re: [PATCH 4/4] S3C64XX DSP: Added the CPU driver for PCM controllers Date: Tue, 3 Nov 2009 09:52:48 +0000 Message-ID: <20091103095247.GB1520@rakim.wolfsonmicro.main> References: <1257127791-5595-1-git-send-email-jassisinghbrar@gmail.com> <1257127791-5595-2-git-send-email-jassisinghbrar@gmail.com> <1257127791-5595-3-git-send-email-jassisinghbrar@gmail.com> <1257127791-5595-4-git-send-email-jassisinghbrar@gmail.com> <20091102111753.GB29351@sirena.org.uk> <1b68c6790911021842s47de7fbdtfec7831b837a98dd@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from opensource.wolfsonmicro.com ([80.75.67.52]:42910 "EHLO opensource2.wolfsonmicro.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1756001AbZKCJwo (ORCPT ); Tue, 3 Nov 2009 04:52:44 -0500 Content-Disposition: inline In-Reply-To: <1b68c6790911021842s47de7fbdtfec7831b837a98dd@mail.gmail.com> Sender: linux-samsung-soc-owner@vger.kernel.org List-Id: linux-samsung-soc@vger.kernel.org To: jassi brar Cc: linux-samsung-soc@vger.kernel.org, Jassi Brar On Tue, Nov 03, 2009 at 11:42:21AM +0900, jassi brar wrote: > On Mon, Nov 2, 2009 at 8:17 PM, Mark Brown > >> + =A0 =A0 =A0 =A0 =A0 =A0 local_irq_save(irqs); > > Why local_irq_save()? =A0On existing uniprocessor systems they shou= ld be > Don't we need to protect critical mode change in PCM control register= ? > Since the s3c64xx_dsp_snd_tx/rxctrl is small and non-blocking, I thin= k > protection by simply enabling/disabling IRQs should be fine. Yes, but you're only disabling IRQs on the local CPU. This means that = a future multi-core CPU variant won't get the benefit of the lock. Since the locking primitives are implemented so that the SMP-specifics compil= e out there should be no performance overhead from being prepared for that.