[SRU[[Trusty] Synaptics HID touchpad driver

[AceLan Kao]

Hi Tim and Brad,

I'm asking help from users on the below 2 bugs.
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1305522
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1265885

And could you point out explicitly what kind of regression test we are
expected to do.
I can find some other machines with synaptics i2c touchpad, is there
anything that I could do?

Best regards,
AceLan Kao.

2014-04-23 23:31 GMT+08:00 Brad Figg <brad.figg at canonical.com>:
> On 04/23/2014 07:56 AM, Tim Gardner wrote:
>> On 04/23/2014 12:59 AM, AceLan Kao wrote:
>>> SRU JUSTIFICATION
>>>
>>> BugLink: https://bugs.launchpad.net/bugs/1305522
>>>
>>> IMPACT: The Synaptics touchpad with i2c interface doesn't work well
>>> after kernel starting supporting i2c-hid device. And the solution
>>> currently we provided is to blacklist i2c_hid driver and let the
>>> touchpad fallback to use PS/2 interface.
>>>
>>> FIX: Synaptics already submitted patches to upstream and will be
>>> merged into kernel v3.6. I cherry-picked the essential patch set from
>>>     git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid.git rmi4
>>> The touchpad supports multi-touch up to 4 fingers(the most I can
>>> verified) after applying those patches.
>>>
>>> TEST: I have built a kernel for test and it works on a Dell laptop
>>> with Synaptics i2c touchpad I have on hands.
>>>     http://people.canonical.com/~acelan/bugs/lp1305522/
>>>
>>> git://kernel.ubuntu.com/acelan/ubuntu-trusty.git rmi4
>>>
>>
>> So, the first 5 patches are actually upstream cherry-picks from 3.15-rc1. Though large, they are mostly mechanical. The last 2 patches, which introduce a new driver, can be cherry-picked from linux-next.
>>
>> I'm sure there can't _possibly_ be any bugs in merge window patches, but I'd feel better if there was some information in the LP report that indicates you've at least done a little regression testing.
>>
>> rtg
>
> AceLan,
>
> I'm in total agreement with Tim here. You folks should all be doing more
> regression testing of the patches you send to us.
>
> Brad
> --
> Brad Figg brad.figg at canonical.com http://www.canonical.com