From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lukasz Majewski Subject: Re: [PATCH 2/6] thermal:exynos4: TMU Common clock framework support for TMU (Thermal Monitoring Unit) Date: Tue, 23 Apr 2013 10:06:47 +0200 Message-ID: <20130423100647.20d80c73@amdc308.digital.local> References: <1366389493-8239-1-git-send-email-l.majewski@samsung.com> <1366389493-8239-3-git-send-email-l.majewski@samsung.com> <20130423081724.03b3b422@amdc308.digital.local> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: Received: from mailout1.samsung.com ([203.254.224.24]:28258 "EHLO mailout1.samsung.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755457Ab3DWIHI (ORCPT ); Tue, 23 Apr 2013 04:07:08 -0400 In-reply-to: Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: Sachin Kamat Cc: Kukjin Kim , Mike Turquette , Zhang Rui , "devicetree-discuss@lists.ozlabs.org" , "linux-samsung-soc@vger.kernel.org" , Linux PM list , Amit Daniel Kachhap , Kyungmin Park , "eduardo.valentin@ti.com" Hi Sachin, > Hi Lukasz, > > On 23 April 2013 11:47, Lukasz Majewski > wrote: > > Hi Sachin, > > > >> On 19 April 2013 22:08, Lukasz Majewski > >> wrote: > >> > This patch modifies exynos_thermal.c file to use > >> > clk_disable_unprepare() and clk_prepare_enable() instead of > >> > clk_{enable|disable}. > >> > > >> > Signed-off-by: Lukasz Majewski > >> > Signed-off-by: Kyungmin Park > >> > --- > >> > >> I have already submitted a similar patch for this: > >> http://permalink.gmane.org/gmane.linux.power-management.general/33310 > >> > >> > > > > Thanks for pointing to the correct patch. > > > > However, I've got a question: > > > > The patch only changes clock names at exynos_tmu_{probe|remove}. > > > > Correct me if I'm wrong, but shouldn't we also change clock_enable > > to clk_prepare_enable at exynos_tmu_read()? (Are we guaranteed, > > that we will NOT sleep there?) > > Since clk_prepare does not do anything, i thought it was sufficient to > have it once in probe and then unprepare in remove. > Do you see a real problem in this implementation. If so I can update > it to use clk_prepare_enable at other places as well. > I just wanted to stick to the common clock new API. It seems to me that exynos_tmu_read() might sleep, but I'm quite novice at TMU :-). If yours patches work, then we shall apply them. I will do my best to test yours three patches ASAP on my setup. -- Best regards, Lukasz Majewski Samsung R&D Poland (SRPOL) | Linux Platform Group