On Fri, Aug 10, 2018 at 01:13:57PM +0300, Andy Shevchenko wrote: > +Cc Darren > > On Fri, 2018-08-10 at 08:16 +0200, Florian Eckert wrote: > > Hello Andy, > > > > first thanks for your review. > > > > > And why it's under arch/x86? Is it uncore part? > > > I suppose it should go under drivers/platform/x86. > > > > I am using a "PC Engines" alix6 board. This is the precursor of the > > new > > APU boards for "PC Engines". > > The button device declaration for the alix borads is under > > "arch/platform/x86/geode/alix.c". > > So I thought the new one should go there as well. > > > > But I will change this. Thanks for the hint. > > Thomas already has discussion with me and others about some platform > drivers under arch/x86. As far as I remember his point is that PDx86 > subsystem for platform stuff suits better. We even move couple of > drivers out from arch/x86 (pmc_atom one of them for sure). > > Geode is quite old SoC, right? Ideally it maybe even go to > drivers/platform/amd. > > Darren, what's your opinion about location of the new platform code for > AMD platforms? Agree. We have a todo to move non architectural platform drivers to drivers/platform/*. The new ones should go to drivers/platform/*, and as we can, we should be migrating the old ones. -- Darren Hart VMware Open Source Technology Center