Am Donnerstag, 5. November 2009 00:27:56 schrieb Alan Stern: > On Wed, 4 Nov 2009, Oliver Neukum wrote: > > This has a tendency for an ugly cycle. > > > > Defective SD card -> error handler resets -> reader disconnected > > -> udev switches mode -> SD reader scanned -> defective SD card -> ... > > > > I think we need a quirk for such devices. I think it is cleaner to have > > it add usbcore's level as we cannot be sure this is always limited > > to the storage driver. Comments? > > Bear in mind that we already have the potential for similar ugly > cycles. Pretty much any device that disconnects itself when an error > or reset occurs and then reconnects will cause one. For example, there > have been reports of this sort of thing happening during enumeration of > the built-in Bluetooth interface in some laptops. People complain > about their system logs getting flooded with error and disconnect > messages. > > This is nothing new, and it's not directly related to mode-switching. > The way to fix it is to have the hub driver look out for devices that > reconnect too much and too often. Yes, similar problems exist. But is better to not trigger such problems rather than handle them. Prevention here has additional benefits as kicking off drivers for no good cause is wrong. Therefore a reset that is known to morph a device is no use in error handling and may do harm to do other drivers. Regards Oliver -- 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
