From mboxrd@z Thu Jan 1 00:00:00 1970 From: ben-linux@fluff.org (Ben Dooks) Date: Wed, 16 Jun 2010 22:14:58 +0100 Subject: [RFC,PATCH 1/2] Add a common struct clk In-Reply-To: <20100614093459.GD12159@pengutronix.de> References: <19473.61547.684572.647641@ipc1.ka-ro> <201006111718.47426.jeremy.kerr@canonical.com> <19474.172.742782.972629@ipc1.ka-ro> <20100611095839.GC10894@pengutronix.de> <19474.2817.333749.485028@ipc1.ka-ro> <1276319643.1962.181.camel@pasglop> <19477.52889.982995.407051@ipc1.ka-ro> <20100614064028.GA12159@pengutronix.de> <19477.53673.644128.558827@ipc1.ka-ro> <20100614093459.GD12159@pengutronix.de> Message-ID: <20100616211458.GB30005@fluff.org.uk> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Mon, Jun 14, 2010 at 11:34:59AM +0200, Uwe Kleine-K?nig wrote: > Hello, > > > And arch/arm/plat-s3c/clock.c has the following comment: > > |/* We originally used an mutex here, but some contexts (see resume) > > | * are calling functions such as clk_set_parent() with IRQs disabled > > | * causing an BUG to be triggered. > > | */ > > |DEFINE_SPINLOCK(clocks_lock); > I wonder why it's needed to reparent clocks during resume. And where > exactly IRQs are disabled. Hmm, this comment was initially introduced > by v2.6.28-rc7-180-gc3391e3, its commit log talks about cpufreq, not > resume. > > Ben (Dooks): Is this still relevant? Yes, unfortunately the system may not resume with the clock registers in the same state as they where before, and there are things that may need to set the parents before the resume can continue. We may have ended up saving quite a bit of the clock state by force, but it isn't the nicest way, and there's still issues with any clocks that require pll stabilisation. -- Ben (ben at fluff.org, http://www.fluff.org/) 'a smiley only costs 4 bytes' From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755669Ab0FPVPR (ORCPT ); Wed, 16 Jun 2010 17:15:17 -0400 Received: from aeryn.fluff.org.uk ([87.194.8.8]:26661 "EHLO kira.home.fluff.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753307Ab0FPVPP (ORCPT ); Wed, 16 Jun 2010 17:15:15 -0400 Date: Wed, 16 Jun 2010 22:14:58 +0100 From: Ben Dooks To: Uwe Kleine-K?nig Cc: Lothar Wa?mann , Benjamin Herrenschmidt , Jeremy Kerr , linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, Ben Dooks Subject: Re: [RFC,PATCH 1/2] Add a common struct clk Message-ID: <20100616211458.GB30005@fluff.org.uk> References: <19473.61547.684572.647641@ipc1.ka-ro> <201006111718.47426.jeremy.kerr@canonical.com> <19474.172.742782.972629@ipc1.ka-ro> <20100611095839.GC10894@pengutronix.de> <19474.2817.333749.485028@ipc1.ka-ro> <1276319643.1962.181.camel@pasglop> <19477.52889.982995.407051@ipc1.ka-ro> <20100614064028.GA12159@pengutronix.de> <19477.53673.644128.558827@ipc1.ka-ro> <20100614093459.GD12159@pengutronix.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20100614093459.GD12159@pengutronix.de> X-Disclaimer: These are my own opinions, so there! User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Jun 14, 2010 at 11:34:59AM +0200, Uwe Kleine-K?nig wrote: > Hello, > > > And arch/arm/plat-s3c/clock.c has the following comment: > > |/* We originally used an mutex here, but some contexts (see resume) > > | * are calling functions such as clk_set_parent() with IRQs disabled > > | * causing an BUG to be triggered. > > | */ > > |DEFINE_SPINLOCK(clocks_lock); > I wonder why it's needed to reparent clocks during resume. And where > exactly IRQs are disabled. Hmm, this comment was initially introduced > by v2.6.28-rc7-180-gc3391e3, its commit log talks about cpufreq, not > resume. > > Ben (Dooks): Is this still relevant? Yes, unfortunately the system may not resume with the clock registers in the same state as they where before, and there are things that may need to set the parents before the resume can continue. We may have ended up saving quite a bit of the clock state by force, but it isn't the nicest way, and there's still issues with any clocks that require pll stabilisation. -- Ben (ben@fluff.org, http://www.fluff.org/) 'a smiley only costs 4 bytes'