From mboxrd@z Thu Jan 1 00:00:00 1970 From: Olof Johansson Date: Mon, 07 Oct 2013 16:37:26 +0000 Subject: Re: [PATCH/RFC] clocksource: Consolidate SH and ARM mach-shmobile Kconfig bits Message-Id: <20131007163726.GC15413@quad.lixom.net> List-Id: References: <20130925233709.16540.66903.sendpatchset@w520> <20130927044323.GC16541@verge.net.au> In-Reply-To: <20130927044323.GC16541@verge.net.au> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Simon Horman Cc: Magnus Damm , linux-sh@vger.kernel.org, khilman@linaro.org, arnd@arndb.de, daniel.lezcano@linaro.org, linux-kernel@vger.kernel.org, tglx@linutronix.de On Fri, Sep 27, 2013 at 01:43:25PM +0900, Simon Horman wrote: > On Thu, Sep 26, 2013 at 08:37:09AM +0900, Magnus Damm wrote: > > From: Magnus Damm > > > > Based on request from Olof Johansson, move ARM mach-shmobile > > clocksource entries to drivers/clocksource/Kconfig, and at > > the same time make the SH architecture use these entries. > > > > Kconfig entries for CMT, MTU2, TMU and STI are with this patch > > now kept in one place. The SH SYS_SUPPORTS_nnn bits are dropped > > to simplye let the user select the appropriate ones using Kconfig > > instead of trying to filter available timers based on SoC. > > > > Known issues: > > 1) Not yet wrappend in menu/endmenu, populates the root driver menu > > -> proposal: Create separate patch to wrap all the clocksource entries? > > That seems reasonable to me. > > > 2) Should the entries be sorted? > > -> proposal: Ignore cosmetic bits > > I propose that you sort the entries are adding > but leave the rest of the entries as is. > > > 3) What is correct dependency for multiplatform and single platform? > > -> proposal: Perhaps we should simply default to no? > > I'm not sure that I understand the issue here. > > > 4) Is single patch across 2 architectures and drivers/clocksource OK? > > -> proposal: Keep it like this if possible > > Olof, I would like your opinion on this but from my pov this seems > reasonable. Especially as one of the architectures (sh) is seeing very little > change at this time. (Sorry for the delayed response, the renesas volume email can be large and I don't always read it when it comes in). Yeah, as long as you have appropriate acks and don't expect conflicts. -Olof