Hi, On Tue, May 24, 2011 at 10:20:37AM -0400, Alan Stern wrote: > On Tue, 24 May 2011, Tanya Brokhman wrote: > > > Hi Felipe > > > > > > Yes :) The driver->speed is updated in usb_composite_probe() if > > > > CONFIG_USB_GADGET_SUPERSPEED is defined. > > > > > > > > So, are we ok with this solution? The module parameter I mean? > > > > Are you going to try the v13 in your branch? Please let me know how > > > it > > > > goes and of course if you have any comments. > > > > > > I think it still gives the possibility for failure. I would rather not > > > take that until all gadget drivers are fixed. We can help you doing > > > that and we only change driver->speed after all gadget drivers have > > > their "sensible defaults" SuperSpeed descriptors. > > > > By "until all gadget drivers are fixed" you mean until all gadget drivers > > provide SS descriptors? This will take for ever... > > I wasn't about to modify all gadget drivers and to add SS descriptors for > > them. I can add default values (as generate_ss_descriptors() did if you > > remember) but I don't think this is the right approach because as you said - > > different gadget drivers might have different SS descriptors and I don't > > feel confident enough to set these values. Nor do I have the ability to test > > each of the gadget drivers the way I would like to after this change. > > The only gadget driver I felt confident adding SS descriptors for is UASP, > > which I tested properly. > > > > Actually if the CONFIG_USB_GADGET_SUPERSPEED is turned off, which is the > > default of it, the speed won't be updated and all these series won't be > > functional so I don't see any possibilities for failure in such > > configuration. Or am I missing something? > > dummy_hcd should work when CONFIG_USB_GADGET_SUPERSPEED is enabled, > even if the usb_gadget_driver structure is initialized with the speed > field set to USB_SPEED_HIGH. This will be true for all the standalone > gadget drivers until they are updated. > > Which leaves a question about the composite gadget framework. Should > it be updated with SS support? Probably not until the various function > drivers have all been updated. We can add support for all the USB3 standard requests, just don't change the driver->speed field until all gadget drivers are fixed. -- balbi
Attachment:
signature.asc
Description: Digital signature
