On Tue, 22 Jul 2014, James P Michels III wrote: > This patch adds usb quirks to improve support for devices > with non standard bInterval values. Quirks are added to support devices with > bInterval values expressed as microframes or frames. The quirks cause the > parse endpoint function to modify the reported bInterval to the standards > conforming value. Sorry, but this makes no sense. The quirk is a property of the device as a whole, whereas whether to interpret interval values in frames or microframes depends on the speed of the USB connection. What would happen if a device had the USB_QUIRK_INTERVAL_AS_MICROFRAMES flag set but it was connected at full speed? The quirk should instead be something like USB_QUIRK_NON_EXPONENTIAL_BINTERVAL, and maybe there should be separate flags for interrupt vs. isochronous endpoints. Then the fixup code should vary its action depending on the connection speed. Even that might not work in all cases. It's conceivable that a device could get the bInterval wrong for endpoints in one config and right for endpoints in a different config. Anyway, is this even needed? What's wrong with the existing code? What values does this Razer Blade device have in its descriptors? Alan Stern -- 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
