Re: [RFC PATCH] USB: enable "power/wakeup" to control remote wakeup in the runtime suspend

[Sarah Sharp <sarah.a.sharp@xxxxxxxxxxxxxxx>]

Ok, let's step back a bit here.

Tianyu, I think you're confused about what exactly USB device suspend
is, and what remote wakeup is.  I suggest you read
Documentation/power-management.txt.

I think Tianyu is concerned that the port power off mechanism will not
be used very often because many USB devices are suspended with remote
wakeup on.  That is a fact of life, and adding new sysfs files will not
help with that.

Take a USB keyboard, for instance.  If you turn on auto-suspend by
echoing "auto" to power/control, the keyboard will suspend after no
keystrokes are transmitted for two seconds.  The usbhid driver suspends
the keyboard with remote wakeup enabled.  Then when the user presses a
key, the keyboard sends remote wakeup signaling, and the usbhid driver
sends a request to get the keystoke.

If you attempt to disable remote wakeup, the keyboard will never send a
wakeup when a key is pressed.  To the user, the keyboard is "broken".

Several USB drivers that need remote wakeup work this way.  USB 3G
modems, for example, need to send a remote wakeup signal when someone
receives a text message.

This is just the way USB devices are designed.  We will just have to
accept that the port power off mechanism won't be used with these USB
devices.

Sarah Sharp
--
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