From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mike Turquette Subject: Re: [RFC] cpufreq: Add bindings for CPU clock sharing topology Date: Wed, 23 Jul 2014 17:33:14 -0700 Message-ID: <20140724003314.6419.51564@quantum> References: <7e097b71342c9f5f63b07ff2e135eb7beb626aab.1405661369.git.viresh.kumar@linaro.org> <53CA8D95.8010108@ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8BIT Return-path: In-Reply-To: Sender: linux-pm-owner@vger.kernel.org To: Viresh Kumar , Santosh Shilimkar Cc: Olof Johansson , Rob Herring , "Rafael J. Wysocki" , Grant Likely , "linaro-kernel@lists.linaro.org" , Nishanth Menon , Sudeep Holla , Stephen Boyd , "linux-arm-kernel@lists.infradead.org" , "linux-pm@vger.kernel.org" , "devicetree@vger.kernel.org" , Lorenzo Pieralisi , Arvind Chauhan , Arnd Bergmann List-Id: devicetree@vger.kernel.org Quoting Viresh Kumar (2014-07-20 05:07:32) > On 19 July 2014 20:54, Santosh Shilimkar wrote: > > Sorry for jumping late > > No, you aren't late. Its just 2 days old thread :) > > > but one of the point I was raising as part of your > > other series was to extend the CPU topology bindings to cover the voltage > > domain information which is probably what is really needed to let the > > CPUfreq extract the information. Not sure if it was already discussed. > > Not it wasn't. > > > After all the CPU clocks, cluster, clock-gating, power domains are pretty much > > related. So instead of having new binding for CPUFreq, I was wondering whether > > we can extend the CPU topology binding information to include missing information. > > Scheduler work anyway needs that information. > > > > Ref: Documentation/devicetree/bindings/arm/topology.txt > > > > Does that make sense ? > > Yeah it does, but I am not sure what exactly the bindings should look then. > So, the most basic step could be moving the new bindings to topology.txt > and name clock-master to dvfs-master. > > What else? If we're going to model the hardware then the binding should not use the CPU phandles in "clock-master" or "dvfs-master". The correct thing to model for a given CPU is which clock consumes. It's not accurate to say that one CPU is the "master", at least not in this context. A previous approach tried to compare struct clk pointers, which is a bad idea since those are just cookies and should not be deref'd by drivers. However a similar approach would be to compare the phandle, right? Regards, Mike > > If its going to be much controversial then we *can* go for just dvfs bindings > for now and then update them later. > > Doesn't make sense? :) From mboxrd@z Thu Jan 1 00:00:00 1970 From: mike.turquette@linaro.org (Mike Turquette) Date: Wed, 23 Jul 2014 17:33:14 -0700 Subject: [RFC] cpufreq: Add bindings for CPU clock sharing topology In-Reply-To: References: <7e097b71342c9f5f63b07ff2e135eb7beb626aab.1405661369.git.viresh.kumar@linaro.org> <53CA8D95.8010108@ti.com> Message-ID: <20140724003314.6419.51564@quantum> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Quoting Viresh Kumar (2014-07-20 05:07:32) > On 19 July 2014 20:54, Santosh Shilimkar wrote: > > Sorry for jumping late > > No, you aren't late. Its just 2 days old thread :) > > > but one of the point I was raising as part of your > > other series was to extend the CPU topology bindings to cover the voltage > > domain information which is probably what is really needed to let the > > CPUfreq extract the information. Not sure if it was already discussed. > > Not it wasn't. > > > After all the CPU clocks, cluster, clock-gating, power domains are pretty much > > related. So instead of having new binding for CPUFreq, I was wondering whether > > we can extend the CPU topology binding information to include missing information. > > Scheduler work anyway needs that information. > > > > Ref: Documentation/devicetree/bindings/arm/topology.txt > > > > Does that make sense ? > > Yeah it does, but I am not sure what exactly the bindings should look then. > So, the most basic step could be moving the new bindings to topology.txt > and name clock-master to dvfs-master. > > What else? If we're going to model the hardware then the binding should not use the CPU phandles in "clock-master" or "dvfs-master". The correct thing to model for a given CPU is which clock consumes. It's not accurate to say that one CPU is the "master", at least not in this context. A previous approach tried to compare struct clk pointers, which is a bad idea since those are just cookies and should not be deref'd by drivers. However a similar approach would be to compare the phandle, right? Regards, Mike > > If its going to be much controversial then we *can* go for just dvfs bindings > for now and then update them later. > > Doesn't make sense? :)