On Fri, Nov 18, 2011 at 8:46 PM, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote: > On Fri, 18 Nov 2011, Amit Nagal wrote: > >> Hi , >> >> If an USB interface is currently being probed ( usb driver probe >> routine is executing ), >> and another driver or userspace program triggers a usb reset on the >> device containing this interface , >> during this time : >> >> 1) how a usb interface driver can handle such usb reset requests >> while currently being probed ? > > It can't. It will get messed up. > >> 2) will linux usb subsystem execute probe routine completely first >> and then then do usb reset the device ? > > That depends. If the reset is triggered directly by another driver > then it will occur during the probe. But if it is triggered by a user > program then it will wait for the probe to finish. > > However... If the other driver is written correctly, it will always > call usb_lock_device_for_reset() before issuing the reset. That > routine will wait for the probe to finish before returning. Therefore > the probe won't get messed up. > > You should read the kerneldoc for usb_reset_and_verify_device() and > usb_lock_device_for_reset(). > >> 3) if a usb interface is already claimed , why is usb reset of this >> interface allowed by another driver or userspace program . > > There is no such thing as a USB reset of an interface, there is only a > USB reset of an entire device. Another driver might own one of the > other interfaces on the same device and might need to reset it. > > Alan Stern > > Thank you for sharing very useful insights . Actually in my embedded target , usb audio modules after probe is finished , gets unbind . when i investigated , the reason is usb reset . i am debugging now to know whether this usb reset has come from userspace or from some other driver . Thanks and Regards Amit Nagal -- 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
