From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Shilimkar, Santosh" Subject: Re: [PATCH 3/3] cpufreq: Add a generic cpufreq-cpu0 driver Date: Sat, 21 Jul 2012 10:34:29 +0530 Message-ID: References: <1342713281-31114-1-git-send-email-shawn.guo@linaro.org> <1342713281-31114-4-git-send-email-shawn.guo@linaro.org> <20120720123317.GA32263@S2101-09.ap.freescale.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Return-path: In-Reply-To: Sender: cpufreq-owner@vger.kernel.org To: "Turquette, Mike" Cc: Shawn Guo , Kevin Hilman , Nishanth Menon , linux-pm@vger.kernel.org, devicetree-discuss@lists.ozlabs.org, cpufreq@vger.kernel.org, "Rafael J. Wysocki" , Richard Zhao , Russell King - ARM Linux , linux-arm-kernel@lists.infradead.org List-Id: devicetree@vger.kernel.org On Fri, Jul 20, 2012 at 9:20 PM, Turquette, Mike wrote: > On Fri, Jul 20, 2012 at 5:33 AM, Shawn Guo wrote: >> On Fri, Jul 20, 2012 at 12:22:14PM +0530, Shilimkar, Santosh wrote: >>> This *_CPU0 doesn't seems to be appropriate since this infrastructure will be >>> used on SMP systems where CPUs shares clock/voltage rails. May be you can >>> get rid of that unless there is a strong need. >>> >> All the resource handlers, clk, regulator, opp, DT node, are retrieved >> from CPU0 device even for SMP. I hope the naming of the driver could >> tell: >> >> - The driver supports UP >> - The driver supports SMP where CPUs shares clock/voltage rails by >> managing CPU0 (resource) >> - The driver does not support SMP where individual CPU can scale >> clock/voltage independently. >> > > How about cpufreq-single-thread.c and CONFIG_CPUFREQ_SINGLE_THREAD? > That makes sense for both UP and SMP. > Indeed. This sounds more appropriate and also reflects what actually happens with a UP or shared clock SMP case. Regards Santosh