[APPLIED] [PATCH natty,oneiric] UBUNTU: SAUCE: Unregister input device only if it is registered

Jesse Sung jesse.sung at canonical.com
Tue Sep 20 02:03:58 UTC 2011


On 09/19/2011 09:41 PM, Leann Ogasawara wrote:
> On Mon, 2011-09-19 at 09:25 +0800, Jesse Sung wrote:
>> https://bugs.launchpad.net/bugs/839238
>>
>> SRU Justification:
>>
>> Using an ALPS touchpad in the quirk list makes unloading psmouse.ko hang
>> the system.
>> Also, due to the protocol problem of ALPS touchpad, the disconnect
>> function would be called when system is running, and cause a random hang.
>>
>> Fix:
>>
>>   From e14874429daac619a2bea5030494ffd68230bfa1 Mon Sep 17 00:00:00 2001
>> From: Wen-chien Jesse Sung<jesse.sung at canonical.com>
>> Date: Fri, 16 Sep 2011 15:20:22 +0800
>> Subject: [PATCH] UBUNTU: SAUCE: Unregister input device only if it is
>>    registered
>>
>> BugLink: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/839238
>>
>> dev2 is not registered in alps_model_quirk_enabled mode, do not
>> unregister while disconnecting.
>>
>> Signed-off-by: Wen-chien Jesse Sung<jesse.sung at canonical.com>
>
> Signed-off-by: Leann Ogasawara<leann.ogasawara at canonical.com>
>
> Applied to Oneiric master-next.  Jesse, I'm also curious if you have
> sent this upstream as well?
>
> Thanks,
> Leann
>
>> ---
>>    drivers/input/mouse/alps.c |    3 ++-
>>    1 files changed, 2 insertions(+), 1 deletions(-)
>>
>> diff --git a/drivers/input/mouse/alps.c b/drivers/input/mouse/alps.c
>> index da20b62..9adff60 100644
>> --- a/drivers/input/mouse/alps.c
>> +++ b/drivers/input/mouse/alps.c
>> @@ -745,7 +745,8 @@ static void alps_disconnect(struct psmouse *psmouse)
>>
>>    	psmouse_reset(psmouse);
>>    	del_timer_sync(&priv->timer);
>> -	input_unregister_device(priv->dev2);
>> +	if (!alps_model_quirk_enabled)
>> +		input_unregister_device(priv->dev2);
>>    	kfree(priv);
>>    }
>>
>> --
>> 1.7.5.4
>>
>>

Hi Leann and rtg,

Just as Seth pointed out, alps_model_quirk_enabled is only exists in 
ubuntu kernel (natty and oneiric), so I didn't send this to upstream.

Thanks,
Jesse Sung




More information about the kernel-team mailing list