On Mon, 19 Feb 2007, Oliver Neukum wrote:

> Am Donnerstag, 15. Februar 2007 22:06 schrieb Alan Stern:
> > There are two main disadvantages to the less-strict scheme.  One is that 
> > the timing of autosuspends isn't as precise: If the timer goes off every 
> > 10 seconds, then a suspend could occur anywhere between 10 and 20 seconds 
> > after the device becomes idle.  IMO this doesn't matter very much.
> 
> I've looked at this again. It seems to me that if usb_autoresume_device()
> doesn't kill the timer, usb_probe_interface() will race with autosuspend.
> Specifically, you might autosuspend an interface without calling its driver's
> suspend() method, if the timer expires on another CPU.

No, that can't happen.

usb_autoresume_device() always increments udev->pm_usage_cnt.  
autosuspend_check() returns early if that count is positive, preventing 
any autosuspend no matter what state the interface is in.

In fact, usb_autoresume_device() _should_ always stop the timer.  The 
proposal is that usb_autopm_get_interface() might leave the timer running.

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
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to