From mboxrd@z Thu Jan 1 00:00:00 1970 From: Zhang Rui Subject: Re: [PATCH 1/1] thermal: Exynos: Add missing dependency Date: Thu, 15 Nov 2012 12:14:26 +0800 Message-ID: <1352952866.2143.25.camel@rzhang1-mobl4> References: <1352875704-2178-1-git-send-email-sachin.kamat@linaro.org> <1352938284.2143.5.camel@rzhang1-mobl4> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: Received: from mga09.intel.com ([134.134.136.24]:41647 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1767553Ab2KOEOa (ORCPT ); Wed, 14 Nov 2012 23:14:30 -0500 In-Reply-To: Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: Sachin Kamat Cc: linux-pm@vger.kernel.org, durgadoss.r@intel.com, patches@linaro.org, akpm@linux-foundation.org, Amit Daniel Kachhap On Thu, 2012-11-15 at 08:52 +0530, Sachin Kamat wrote: > Hi Zhang, > > Thanks for your review comments. > > On 15 November 2012 05:41, Zhang Rui wrote: > > Hi, Sachin, > > > > thanks for catching the problem. > > > > On Wed, 2012-11-14 at 12:18 +0530, Sachin Kamat wrote: > >> CPU_FREQ_TABLE depends on CPU_FREQ. Selecting CPU_FREQ_TABLE without checking > >> for dependencies gives the following compilation warnings: > >> warning: (ARCH_TEGRA_2x_SOC && ARCH_TEGRA_3x_SOC && UX500_SOC_DB8500 && > >> CPU_THERMAL && EXYNOS_THERMAL) selects CPU_FREQ_TABLE which has unmet > >> direct dependencies (ARCH_HAS_CPUFREQ && CPU_FREQ) > >> > > Amit, > > > > how is exynos driver supposed to work? > > do you want the exynos driver still be loaded without CPU_THERMAL? > > If yes, EXYNOS_THERMAL should not select CPU_FREQ_TABLE. > > If no, EXYNOS_THERMAL should depends on CPU_THERMAL instead of THERMAL. > > and CPU_THERMAL will select CPU_FREQ_TABLE instead. > > > > IMO, either of the above solution will be more proper to fix this > > warning. > > I will discuss with Amit and send an updated patch to fix this. > I checked the code of exynos driver, it fails to load if no cpufreq_cooling device found, which means it depends on CPU_THERMAL. I'm not sure if this is the behavior you want, but IMO, exynos driver should register its critical trip point to the generic thermal layer instead, when CPU_THERMAL is not set. thanks, rui > > > > > thanks, > > rui > > > >> Cc: Amit Daniel Kachhap > >> Signed-off-by: Sachin Kamat > >> --- > >> Build tested using exynos4_defconfig on linux-next tree of 20121114. > >> --- > >> drivers/thermal/Kconfig | 2 +- > >> 1 files changed, 1 insertions(+), 1 deletions(-) > >> > >> diff --git a/drivers/thermal/Kconfig b/drivers/thermal/Kconfig > >> index 266c15e..197b7db 100644 > >> --- a/drivers/thermal/Kconfig > >> +++ b/drivers/thermal/Kconfig > >> @@ -50,7 +50,7 @@ config RCAR_THERMAL > >> > >> config EXYNOS_THERMAL > >> tristate "Temperature sensor on Samsung EXYNOS" > >> - depends on (ARCH_EXYNOS4 || ARCH_EXYNOS5) && THERMAL > >> + depends on (ARCH_EXYNOS4 || ARCH_EXYNOS5) && THERMAL && CPU_FREQ > >> select CPU_FREQ_TABLE > >> help > >> If you say yes here you get support for TMU (Thermal Managment > > > > > > >