From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jonghwan Choi Subject: RE: [PATCH v2] cpufreq: exynos: Fix the compile error Date: Fri, 16 May 2014 18:57:38 +0900 Message-ID: <000001cf70ed$451bbec0$cf533c40$@samsung.com> References: <000001cf70dc$9ccfe4e0$d66faea0$@samsung.com> <5375D150.1020109@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: In-reply-to: <5375D150.1020109@gmail.com> Content-language: ko Sender: linux-samsung-soc-owner@vger.kernel.org To: 'Tomasz Figa' , linux-samsung-soc@vger.kernel.org, linux-arm-kernel@lists.infradead.org, devicetree@vger.kernel.org, cpufreq@vger.kernel.org Cc: 'Kukjin Kim' , "'Rafael J. Wysocki'" , 'Jonghwan Choi' , 'Viresh Kumar' List-Id: devicetree@vger.kernel.org On 16.05.2014 5:50 PM, Tomasz Figa wrote: > We have been discussing Exynos cpufreq framework rework, including DT support since quite a long time. > so if you want to know details, please look at respective discussion threads. I don't have time to > find them for you right now, but if you look for threads of cpufreq patches by Thomas Abraham and ones > about Exynos multiplatform support, then you should have quite complete set of information. > I will check that. > In general, I can see three ways to "fix" this for 3.15: > > 1) revert moving soc_is_* macros to mach-exynos/common.h, > > 2) move soc_is_* macros to a header in mach-exynos/include/mach/, > > 3) use of_machine_is_compatible() or of_match_node() on root node in this driver to match directly > with SoC compatible strings. > > Probably option 3) would be preferred, as it would remove some of the dependencies of this driver on > arch code and including plat/cpu.h (or mach/cpu.h as would be introduced by 2)) would be no longer > necessary. Thanks~ Best Regards.