From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Dooks Date: Fri, 07 Mar 2014 14:30:11 +0000 Subject: Re: Renesas RCar device-tree USB series Message-Id: <5319D7F3.2060209@codethink.co.uk> List-Id: References: <1394128887-4197-1-git-send-email-ben.dooks@codethink.co.uk> In-Reply-To: <1394128887-4197-1-git-send-email-ben.dooks@codethink.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-sh@vger.kernel.org On 07/03/14 05:36, Magnus Damm wrote: > Hi Ben, > > On Fri, Mar 7, 2014 at 3:01 AM, Ben Dooks wrote: >> This is a new series covering enabling the RCar series of SoCs USB >> with device-tree based booting. It has been tested on the R8A7790 >> Lager board. >> >> Improvements from the previous series include: >> >> - mapping usb to the relevant phy by dt >> - better use of existing pci of functions >> >> Note, there is still an issue with the second gigabyte of memory on >> the Lager, which with current kernels causes the system to abort on >> startup. This series will only work if the top memory area is disabled. > > Thanks for these patches. I think they start looking really good. > > In my mind there are two outstanding issues: > > 1) Is the USB core code change acceptable or not? > > The patch "[PATCH 4/9] usb: phy: check for of_node when getting phy" > looks quite fine to me, but I wonder if the USB maintainers will > accept it as-is or require some rework. I would say that the rest of > this series depends on that USB core change. I'm not sure, but without it we need to hack the PHY driver to turn on the PHY at start time. > 2) Per-port USB PHY driver configuration via DT > > Right now each USB host controller points to the same PHY device. > Thanks for working on describing the topology! As you know, the PHY > driver itself handles several USB ports, and I'd like to use DT to > represent the mapping between which PHY port that maps to what USB > Host controller to allow proper run time configuration. Right now in > this version of the series there is no such mapping. Of course, that > depends on proper USB PHY DT bindings... Hmm, given the shared phy is shared and referenced counted then I don't /think/ there is much more to be done for this. If we add more PHYs then I would assume that each one of them would -- Ben Dooks http://www.codethink.co.uk/ Senior Engineer Codethink - Providing Genius