Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754013Ab0DVXYN (ORCPT ); Thu, 22 Apr 2010 19:24:13 -0400 Received: from mail.tpi.com ([70.99.223.143]:4668 "EHLO mail.tpi.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753914Ab0DVXYL (ORCPT ); Thu, 22 Apr 2010 19:24:11 -0400 X-Greylist: delayed 1426 seconds by postgrey-1.27 at vger.kernel.org; Thu, 22 Apr 2010 19:24:11 EDT Message-ID: <4BD0D4F0.9090605@canonical.com> Date: Thu, 22 Apr 2010 17:00:00 -0600 From: Tim Gardner Reply-To: tim.gardner@canonical.com User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100415 Thunderbird/3.0.4 MIME-Version: 1.0 To: Amit Kucheria CC: Sascha Hauer , Amit Kucheria , Russell King , =?ISO-8859-1?Q?Uwe_Kleine-K=F6n?= =?ISO-8859-1?Q?ig?= , linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH 1/1] arm: mxc: utilise usecount field in clock operations References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2846 Lines: 100 On 04/22/2010 02:30 PM, Amit Kucheria wrote: > From: Amit Kucheria > > This patch fixes the clock refcounting when reparenting is used. > > Sascha just pointed out a good explanation of refcounting here: > http://www.spinics.net/lists/arm-kernel/msg85879.html > > Signed-off-by: Amit Kucheria > --- > arch/arm/plat-mxc/clock.c | 34 ++++++++++++++++++++++------------ > 1 files changed, 22 insertions(+), 12 deletions(-) > > diff --git a/arch/arm/plat-mxc/clock.c b/arch/arm/plat-mxc/clock.c > index 323ff8c..14781cc 100644 > --- a/arch/arm/plat-mxc/clock.c > +++ b/arch/arm/plat-mxc/clock.c > @@ -50,15 +50,16 @@ static DEFINE_MUTEX(clocks_mutex); > > static void __clk_disable(struct clk *clk) > { > - if (clk == NULL || IS_ERR(clk)) > + WARN_ON(!clk->usecount); > + if (clk == NULL || IS_ERR(clk) || !clk->usecount) > return; > The clk==NULL check seem superfluous if WARN_ON(!clk->usecount) has already dereferenced clk (and possibly crashed). You might need two statements if its likely that clk could be NULL. > - __clk_disable(clk->parent); > - __clk_disable(clk->secondary); > - > - WARN_ON(!clk->usecount); > - if (!(--clk->usecount)&& clk->disable) > - clk->disable(clk); > + if (!(--clk->usecount)) { > + if (clk->disable) > + clk->disable(clk); > + __clk_disable(clk->parent); > + __clk_disable(clk->secondary); > + } > } > > static int __clk_enable(struct clk *clk) > @@ -66,12 +67,13 @@ static int __clk_enable(struct clk *clk) > if (clk == NULL || IS_ERR(clk)) > return -EINVAL; > > - __clk_enable(clk->parent); > - __clk_enable(clk->secondary); > - > - if (clk->usecount++ == 0&& clk->enable) > - clk->enable(clk); > + if (clk->usecount++ == 0) { > + __clk_enable(clk->parent); > + __clk_enable(clk->secondary); > > + if (clk->enable) > + clk->enable(clk); > + } > return 0; > } > > @@ -160,17 +162,25 @@ EXPORT_SYMBOL(clk_set_rate); > int clk_set_parent(struct clk *clk, struct clk *parent) > { > int ret = -EINVAL; > + struct clk *prev_parent = clk->parent; > > if (clk == NULL || IS_ERR(clk) || parent == NULL || > IS_ERR(parent) || clk->set_parent == NULL) > return ret; > > + if (clk->usecount != 0) > + clk_enable(parent); > + > mutex_lock(&clocks_mutex); > ret = clk->set_parent(clk, parent); > if (ret == 0) > clk->parent = parent; > + > mutex_unlock(&clocks_mutex); > > + if (clk->usecount != 0) > + clk_disable(prev_parent); > + > return ret; > } > EXPORT_SYMBOL(clk_set_parent); -- Tim Gardner tim.gardner@canonical.com -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/