On Thu, 12 Oct 2017, Daniel Drake wrote: > On Wed, Oct 11, 2017 at 11:04 PM, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote: > > Also, you should check whether this is for a runtime suspend or a > > system suspend. You don't want to go around disconnecting a device > > whenever it gets runtime suspended! > > Good point, I had not considered runtime suspend. It's not quite so > simple though. > > I make your suggested change (testing PMSG_IS_AUTO in this codepath), > then enable autosuspend with: > echo auto > /sys/bus/usb/devices/1-9/power/control > > then the device gets suspended (no interface drivers are loaded) > without the port disconnect happening. > > Now if I go into S3 suspend, the original problem returns: the system > wakes up immediately. Nasty! > So that is an imperfection with this approach. Any suggestions for how > to proceed? How about moving the test into usb_suspend() in driver.c? If the quirk flag is set at that point, call hub_port_disable(). You'll have to add a glue routine to hub.c for this purpose. 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
