Am Mittwoch, 10. Februar 2010 20:59:05 schrieb Alan Cox: > > Are you suggesting that the kernel should include its own polling code > > to detect media changes on devices that don't report them? > > I think that might be best. It would give a coherent API and it only > needs to be polling while a user task is requesting notification. I think the kernel knows less than user space about the need to poll a device. It is far easier to build, for example, a whitelist of dongles that cannot physically change their medium in user space. Why would you want to put the execution of polling into the kernel while the user space code can perform the exact equivalent and is easier to adapt, for example to a system of exponential backoffs in the polling frequency or lowering the frequency when the screensaver becomes active or the lid is closed or whatever user space may come up with? 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
