On Tue, May 15, 2012 at 09:22:07AM +0800, Peter Chen wrote: > On Mon, May 14, 2012 at 03:51:52PM +0300, Felipe Balbi wrote: > > Hi, > > > > On Mon, May 14, 2012 at 08:41:54PM +0800, Peter Chen wrote: > > > On Mon, May 14, 2012 at 03:18:52PM +0300, Felipe Balbi wrote: > > > > On Mon, May 14, 2012 at 01:28:11PM +0200, Marek Vasut wrote: > > > > > > > > Maybe... but that'll, again, mean loads of per-board data to associate > > > > both devices. If everybody was using DT, we could just use phandles :-p > > > > > > > > Then again, when we want to have a kernel-wide generic PHY layer (not > > > > USB PHY layer, but generic), we will have waaaaay too much data to > > > > associate PHYs with their specific Links. > > > > > > Just like you said just now, there are much things to do for PHY Layer (or > > > kernel-side generic PHY), so we need step by step. Does you agree we move > > > USB PHY related things to drivers/usb/phy first? > > > > Sure, that makes sense. But what will drivers/usb/otg become ? Wouldn't > > it be better to just rename that directory instead ? That's holding PHY > > drivers anyway. > drivers/usb/otg will be only for otg driver, like current xxx_otg.c, and > command otg file in future. > > As we have decided to separate PHY from OTG, it is better to put two things > at two different folders. I'm not sure I follow. What do you call OTG driver ? If it's all that timer initialization that should be done on drivers/usb/core/ in a generic manner, so everybody can use. Remember that OTG related to Host and Device functionality and all those OTG timers should be managed by the core framework. -- balbi
Attachment:
signature.asc
Description: Digital signature
