In my mind that the major differences of these drivers are 1. PHY setting and control 2. OTG related stuff 3. VBUS detection and clock enabling 4. Power optimization 5. Charger related stuff There may be some other differences. I think if you want make all the platform share the udc driver, you have to abstract the interfaces more clearly. If you have written the shared driver, I can help you to test at the marvell pxa platform for the udc driver. -----Original Message----- From: Li Yang-R58472 [mailto:R58472@xxxxxxxxxxxxx] Sent: Wednesday, February 16, 2011 2:27 PM To: Matthieu CASTET; Greg KH Cc: linux-usb@xxxxxxxxxxxxxxx; Alan Stern; David Brownell; Nguyen Dinh-R00091; Guennadi Liakhovetski; Chao Xie; Pavankumar Kondeti; Xiaochen Shen; Philippe Skowronski Subject: RE: duplicate udc driver for managing tdi core ? >-----Original Message----- >From: Matthieu CASTET [mailto:matthieu.castet@xxxxxxxxxx] >Sent: Wednesday, February 16, 2011 1:56 AM >To: Greg KH >Cc: linux-usb@xxxxxxxxxxxxxxx; Alan Stern; David Brownell; Li Yang-R58472; >Nguyen Dinh-R00091; Guennadi Liakhovetski; Chao Xie; Pavankumar Kondeti; >Xiaochen Shen; Philippe Skowronski >Subject: Re: duplicate udc driver for managing tdi core ? > >Greg KH a écrit : >> On Sat, Feb 12, 2011 at 03:05:21PM +0100, matthieu castet wrote: >>> Hi, >>> >>> grepping for tdi (arc/chipidea/mips otg ehci like controller) >>> registers in driver/usb/gadget show interesting things : it seems we >>> have 4 drivers for handling the same controller. >>> - fsl_udc_core [1] >>> - ci13xxx_udc [2] >>> - mv_udc [3] >>> - langwell [4] >>> >>> fsl is the historic driver (freesale and tegra will be using it). >>> ci13xxx was added by mips and msm chip use it mv_udc/langwell is >>> newer. >>> >>> I didn't look at the driver in details, there may be tiny difference >>> between all this controllers. >>> >>> Shouldn't we try to merge them and rename to a clear name in order >>> people don't rewrite again the same driver ? >> >> Sure, if you think it is possible, please do so. >> >But without help of the current maintainers it will be difficult. I have >no way to test on their platform. > >Also how to select which driver should be the base of the merge : >- fsl_udc_core is the historic driver and seems to have tested pretty well >(and was used on big endian arch) The driver is also used on Freescale ARM based i.mx SoCs which are little endian. - Leo -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
