[PATCH] UBUNTU: SAUCE: acer-wmi: Add wireless quirk for Lenovo 3000 N200

[Colin Ian King]

On 07/10/11 01:58, Seth Forshee wrote:
> On Fri, Oct 07, 2011 at 12:47:41AM +0100, Colin Ian King wrote:
>> On 06/10/11 21:35, Seth Forshee wrote:
>>> This patch has been submitted upstream and acked by the acer-wmi
>>> maintainer, but it has not yet been merged. Without this fix the
>>> affected model has a permanent soft block on wireless whenever acer-wmi
>>> is loaded, and the regression potential of the fix is very small as it
>>> is a DMI quirk for a single model. Therefore, I'm proposing we SRU this
>>> as a sauce patch at this time to get the fix into oneiric sooner.
>>>
>>> == SRU Justification ==
>>>
>>> Impact: acer-wmi is lacking a required quirk for the Lenovo 3000 N200
>>> wlan rfkill. As a result, wireless is permanently blocked on this
>>> machine whenever acer-wmi is loaded.
>>>
>>> Fix: Add quirk to use the correct EC register for updating the wlan
>>> rfkill status.
>>>
>>> Testcase: Loading acer-wmi without the fix results in a permanent wlan
>>> soft block. With the fix the rfkill switch behaves as expected. Verified
>>> against natty on LP #857297.
>>>
>>> BugLink: http://bugs.launchpad.net/bugs/857297
>>>
>> I've got a Lenovo 3000 N200, so I'm very curious why the acer-wmi
>> module is being loaded. By default this platform driver is not
>> loaded.
> acer-wmi actually seems to support a number of Lenovo models. The
> machine the bug was reported on has WMI guid
> 67C3371D-95A3-4C37-BB61-DD47B491DAAB, which is one of those that
> acer-wmi lists in its module aliases. as well as another guid the driver
> uses, 431F16ED-0C2B-444C-B267-27DEB140CF9C. And the driver does seem to
> generally work on the laptop, I'm even reusing an existing quirk in the
> patch.
Got it, my Lenovo 3000N200 is definitely a different version, it has 
totally different GUIDs.
> It does seem that there are a couple of sub-models sold under this name,
> so if the driver doesn't load on yours you might compare your DMI
> information and DSDT to that in the bug. They may not be the same.
>
Thanks for the clarification!