On Fri, 9 Mar 2007, Oliver Neukum wrote: > > I'm still not sure how good an idea it is. But let's say we do it. > > Then the new power/level attribute file would have 4 possible values: > > > > on, auto, suspend, and suspend-without-autoresume > > Why? What's the difference to remote wakeup which we don't encode here?
There is no difference for remote wakeup, but there is a difference for autoresume. Remember, these are two different things. Remote wakeup occurs when the device signals that something has happened; autoresume occurs when the driver receives a new I/O request. When the device has been autosuspended normally, its power/level attribute would be "suspend". Or the user could write that value to the attribute, forcing the device to suspend immediately. But autoresume would still be enabled, so the device would resume whenever new I/O appeared -- as might happen if, for example, some process tried to send a packet out over a USB network interface. The new fourth state, "suspend-without-autoresume", specifically disables autoresume. It would allow you to suspend a network interface and keep it suspended -- even if someone tried to send a packet out that interface -- without doing "ifconfig down" and breaking all the existing sessions. Isn't that what you wanted? Alan Stern ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ linux-usb-devel@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
