On Thu, 2 Aug 2012, B, Ravi wrote: > > > --- a/drivers/usb/storage/usb.c > > > +++ b/drivers/usb/storage/usb.c > > > @@ -844,6 +844,8 @@ static void > > quiesce_and_remove_host(struct us_data *us) > > > */ > > > scsi_lock(host); > > > set_bit(US_FLIDX_DISCONNECTING, &us->dflags); > > > + /* stop the current urbs when the device got disconnected */ > > > + usb_stor_stop_transport(us); > > > > This shouldn't be necessary. This code runs after > > scsi_remove_host() returns, so there should not be any URBs > > running at this point. > > > > Have you actually encountered a problem that this patch fixes? > > In specific condition, where the transmit request is in progress and > device is unplugged from host, found that this current tx request is > not dequeued/unlinked during disconnect. Please provide more information. What specific condition? What was the current request? How did this happen? If the device was unplugged while an URB was active, the URB should have completed very quickly with a failure. Why didn't it fail? As far as I can tell, this shouldn't be needed. scsi_remove_host() calls scsi_forget_host(), which calls __scsi_remove_device(), which calls scsi_free_queue(), which calls blk_cleanup_queue(), which calls blk_drain_queue(), which should wait until all pending requests are finished. 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
