On Mon, 2009-12-07 at 17:34 +0100, Oliver Neukum wrote:
> Am Montag, 7. Dezember 2009 17:26:41 schrieb Francesco Lavra:
> > Hi,
> > this patch adds support for reset_resume in cdc-acm. This function simply
> > calls tty_hangup() if there was any user of the tty at suspend time, and
> > then calls the normal resume function. Tested with a Samsung SGH-U900
> > device. Without reset_resume, every time the system is hibernated during a
> > PPP connection, the device moves from ttyACMx to ttyACM(x+1), an it must
> > be unplugged and replugged in order for its original node ttyACMx to
> > re-appear. Instead, with reset_resume the device retains its node name
> > across hibernation cycles.
>
> Hi,
>
> it would be better if you made calling hangup conditional on the device
> having been opened. Could you do that?
>
> Regards
> Oliver
The check on tty being non NULL serves that purpose (in addition to
making the code safe), because when close() is called on the last open
device, tty_port_tty_set() is called with NULL as its second argument,
thus NULLifying tty.
BTW, I took that piece of code from acm_disconnect().
An equally working alternative would be the patch below, which is
slightly less efficient, though.
Signed-off-by: Francesco Lavra <francescolavra@xxxxxxxxxxxx>
--- a/drivers/usb/class/cdc-acm.c 2009-12-07 16:22:20.000000000 +0100
+++ b/drivers/usb/class/cdc-acm.c 2009-12-07 18:16:55.000000000 +0100
@@ -1460,6 +1460,21 @@ err_out:
return rv;
}
+static int acm_reset_resume(struct usb_interface *intf)
+{
+ struct acm *acm = usb_get_intfdata(intf);
+ struct tty_struct *tty;
+
+ if (acm->port.count) {
+ tty = tty_port_tty_get(&acm->port);
+ if (tty) {
+ tty_hangup(tty);
+ tty_kref_put(tty);
+ }
+ }
+ return acm_resume(intf);
+}
+
#endif /* CONFIG_PM */
/*
* USB driver structure.
@@ -1546,6 +1561,7 @@ static struct usb_driver acm_driver = {
#ifdef CONFIG_PM
.suspend = acm_suspend,
.resume = acm_resume,
+ .reset_resume = acm_reset_resume,
#endif
.id_table = acm_ids,
#ifdef CONFIG_PM
--
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
