On Mon, 19 Feb 2007, Oliver Neukum wrote:
> Hi Alan,
>
> if probe() fails, you fail to restore the old pm count. If the driver
> supports autosuspend, but doesn't bind, the interface, though unbound,
> will not suspend again.
>
> Regards
> Oliver
>
> Signed-off-by: Oliver Neukum <[EMAIL PROTECTED]>
> ------
>
> --- a/drivers/usb/core/driver.c 2007-02-19 14:27:11.000000000 +0100
> +++ b/drivers/usb/core/driver.c 2007-02-19 14:32:13.000000000 +0100
> @@ -189,6 +189,7 @@
> struct usb_interface *intf;
> struct usb_device *udev;
> const struct usb_device_id *id;
> + int old_usage_count;
> int error = -ENODEV;
>
> dev_dbg(dev, "%s\n", __FUNCTION__);
> @@ -219,12 +220,14 @@
> /* The interface should always appear to be in use
> * unless the driver suports autosuspend.
> */
> + old_usage_count = intf->pm_usage_cnt;
> intf->pm_usage_cnt = !(driver->supports_autosuspend);
>
> error = driver->probe(intf, id);
> if (error) {
> mark_quiesced(intf);
> intf->needs_remote_wakeup = 0;
> + intf->pm_usage_cnt = old_usage_count;
> intf->condition = USB_INTERFACE_UNBOUND;
> } else
> intf->condition = USB_INTERFACE_BOUND;
This change isn't needed. The fact that mark_quiesced() is called for the
interface means that the pm_usage_cnt won't be used for anything -- until
another driver is probed.
After all, if no driver is bound to the interface then the pm_usage_cnt
value is meaningless. By definition, the interface cannot be in use.
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
