From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nishanth Menon Subject: Re: [PATCH v3 1/2] cpufreq / OPP: Allow boost frequency to be looked up from device tree Date: Wed, 14 May 2014 09:31:49 -0500 Message-ID: <53737E55.9070704@ti.com> References: <1400029380-5372-1-git-send-email-thomas.ab@samsung.com> <1400029380-5372-2-git-send-email-thomas.ab@samsung.com> <20140514080930.781e0557@amdc2363> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-pm-owner@vger.kernel.org To: Viresh Kumar , Lukasz Majewski Cc: Thomas Abraham , "Rafael J. Wysocki" , Rob Herring , "devicetree@vger.kernel.org" , linux-samsung-soc , "linux-pm@vger.kernel.org" , Tomasz Figa , Kukjin Kim , Thomas P Abraham , "linux-arm-kernel@lists.infradead.org" List-Id: devicetree@vger.kernel.org On 05/14/2014 01:24 AM, Viresh Kumar wrote: > On 14 May 2014 11:39, Lukasz Majewski wrote: >> I agree with Nishanth here, that point 1 (as described by Viresh at >> [*]) is a more scalable approach. > > The only reason why I wanted all that to be done at OPP level was to > ensure if somebody else also needs it apart from cpufreq, they don't have > to duplicate code and find it.. As it is present at a central place.. > > But if no other code is going to look for that, it may just be fine as is.. > If we eventually have a need beyond cpufreq (say devfreq) with similar instances, then it makes sense to move it out to a generic place. Either way, code implementation/duplication is a OS problem - and should be looked at independent of the description in dts. If we feel the description is valid hardware description (which, personally, I do), then lets go to the next discussion point of where to put it - generic or cpufreq specific (here, I have no preference), and finally decide the implementation as necessary as a result of the description. -- Regards, Nishanth Menon