Re: [linux-usb-devel] [PATCH 1/2] usb_gigaset: suspend support

[Tilman Schmidt <tilman@imap.cc>]

[-- Attachment #1: Type: text/plain, Size: 1584 bytes --]

Am 30.10.2007 09:52 schrieb Oliver Neukum:
> Am Montag 29 Oktober 2007 schrieb Tilman Schmidt:

>> --- a/drivers/isdn/gigaset/usb-gigaset.c	2007-10-23 01:48:12.000000000 +0200
>> +++ b/drivers/isdn/gigaset/usb-gigaset.c	2007-10-24 00:47:32.000000000 +0200
>> @@ -847,6 +859,63 @@ static void gigaset_disconnect(struct us
>>  	gigaset_unassign(cs);
>>  }
>>  
>> +/* gigaset_suspend
>> + * This function is called before the USB connection is suspended or reset.
>> + */
>> +static int gigaset_suspend(struct usb_interface *intf, pm_message_t message)
>> +{
>> +	struct cardstate *cs;
>> +	struct usb_cardstate *ucs;
>> +	int rc;
>> +
>> +	if ((cs = usb_get_intfdata(intf)) == NULL ||
>> +	    (ucs = cs->hw.usb) == NULL) {
>> +		err("%s: no cardstate", __func__);
>> +		return -EFAULT;
>> +	}
>> +
>> +	//FIXME stop common module activities? ISDN_STAT_STOP? block open()?
>> +
>> +	/* stop submitting bulk URBs */
>> +	tasklet_disable(&cs->write_tasklet);
>> +
>> +	/* kill pending read URB */
>> +	usb_kill_urb(ucs->read_urb);
> 
> no pending write URB?

There shouldn't be. A read URB is always posted in case the device
sends anything, even while the device is idle, but a write URB is
only submitted if there's actual activity. But you're right, it's
safer to add a "usb_kill_urb(ucs->bulk_out_urb);" here just in case.

Thanks,
Tilman

-- 
Tilman Schmidt                          E-Mail: tilman@imap.cc
Bonn, Germany
Diese Nachricht besteht zu 100% aus wiederverwerteten Bits.
Ungeöffnet mindestens haltbar bis: (siehe Rückseite)


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 253 bytes --]