Backporting the HDA Jack sense patches to Precise
Brad Figg
brad.figg at canonical.com
Thu Jan 5 19:25:04 UTC 2012
On 01/05/2012 10:28 AM, David Henningsson wrote:
> On 01/05/2012 06:12 PM, Brad Figg wrote:
>> On 01/05/2012 08:48 AM, David Henningsson wrote:
>>> On 01/03/2012 03:46 PM, Leann Ogasawara wrote:
>>>> On Tue, 2012-01-03 at 11:44 +0100, David Henningsson wrote:
>>>>> A follow-up to the earlier thoughts about having the new jack detection
>>>>> interface in Precise. [1] Takashi has now merged the code into his main
>>>>> branch, which means it'll make it into 3.3. I think now would be a good
>>>>> time to start testing/backporting it into the Precise kernel.
>>>>>
>>>>> Is there anything I need to know about how do we go about this? I was
>>>>> thinking to do this:
>>>>>
>>>>> 1) I'll export the relevant patches from Takashi's tree with git
>>>>> format-patch
>>>>
>>>> You may want to cherry-pick instead.
>>>>
>>>>> 2) I'll create a local branch of the Precise git tree and applying them
>>>>> one by one, backporting/changing as necessary
>>>>
>>>> For tracking purposes, if you do cherry-pick, could you use the -x
>>>> option. Could you also be sure also modify the commit message to note
>>>> which git repo it was cherry-picked from, eg:
>>>>
>>>> (cherry-picked from<sha1>
>>>> git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound.git)
>>>>
>>>> In the event that the patch does not cherry-pick cleanly, could you
>>>> modify the commit message to note it's a backport rather than a clean
>>>> cherry-pick, eg:
>>>>
>>>> (backported from<sha1>
>>>> git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound.git)
>>>>
>>>>> 3) I'll build a test kernel
>>>>> 4) I'll test the new feature on at least two machines I have here and
>>>>> confirm that it works.
>>>>> 5) I'll push the branch for review/inclusion into a public tree at
>>>>> kernel.ubuntu.com.
>>>>>
>>>>> Does this seem reasonable?
>>>>
>>>> Sounds fine.
>>>
>>> Ok, I have now done this. I've tested the new kernel on two machines,
>>> on which I've tested analog playback, auto-mute, HDMI playback and a
>>> quick recording test as well. No regressions were found - in fact,
>>> these patches also improve the old jack detection
>>> interface, so PulseAudio works better already. :-)
>>>
>>> For reference, one machine had a Realtek codec and Nvidia for HDMI,
>>> and the other Sigmatel codec and Intel for HDMI.
>>>
>>> The branch you want to review/include is at
>>> http://kernel.ubuntu.com/git?p=diwic/ubuntu-precise.git;a=shortlog;h=refs/heads/jack-detection-backport
>>>
>>>
>>
>> I'm a little confused. Is this a request to pull the changes from
>> this branch into the Precise kernel tree?
>
> Well, I guess so, unless you (as in the "kernel team") would like to do some reviewing/testing yourself first. I'm assuming you would know the workflow better than I do. :-)
>
>> I'd suggest before that happens, that a set of kernels be built
>> and a call for community testing be done. Based on the positive
>> feedback from that testing, then pull this into Precise.
>
> That works for me, assuming you know how to "call for community testing"? Also, I just tried to upload the entire kernel source package to one of my ppa's, but the uploading failed at the final byte, or something like that.
>
If you look at what Colin King has been doing for power management
testing, that's what I consider a call for community testing.
>> Also, 3.2 just went final and the dev kernel is being rebased on
>> that. Your changes and test kernels should be on the rebased
>> Precise kernel.
>
> Hmm, do you suggest I wait until the master branch is updated (five minutes ago it still wasn't)?
>
If you look at the master-next branch and if you followed any of the
discussion on #ubuntu-kernel today you would have noticed that 3.2
final got rebased onto the master-next branch. There were some other
housekeeping changes that got done. Note, had your request come in
before today, I would not have suggested this. However, since we are
so close to having 3.2 on the master branch, I think it's warranted.
Brad
--
Brad Figg brad.figg at canonical.com http://www.canonical.com
More information about the kernel-team
mailing list