Re: snd_hda : sometimes sound sometimes not

On 28/05/2011 15:46, Jeremy Chadwick wrote:
> On Sat, May 28, 2011 at 03:30:26PM +0200, David Demelier wrote:
>> On 12/05/2011 08:47, David Demelier wrote:
>>> Hello,
>>>
>>> I don't know if there is a lot of changes in the snd_hda driver in the
>>> -STABLE branch but since I upgraded to it sometimes I have sound and
>>> sometimes not.
>>>
>>> The mixer are exactly the same when these event occurs. This happened
>>> this morning. After booting I do not have any sound. I rebooted and
>>> suddenly I've got sound again...
>>>
>>> I only tweak snd_hda(4) for a pin sense on the front panel (it has no
>>> sound neither)
>>>
>>> So I added in /boot/devices.hints :
>>> hint.hdac.1.cad0.nid27.config="as=1 seq=15"
>>>
>>> And there's the both dmesg ok.txt when sound is here and not.txt when
>>> there isn't as you can see there is no difference related to the hda
>>> driver.
>>>
>>> http://markand.malikania.fr/ok.txt
>>> http://markand.malikania.fr/nok.txt
>>>
>>> I'm guessing something. My laptop has a mute shortcut, if I press it at
>>> the BIOS stage I will not have sound neither thus is it possible that my
>>> chipset is muted from anything?
>>>
>>> Cheers,
>>>
>>
>> Sorry to cross-post again, but I just wanted to tell you that the
>> problem disappeared in -CURRENT so now I just how the unknown bogus
>> code will be MFC before 8.3-RELEASE
>
> Unless someone can chime in with details of the commits which changed,
> assuming "the magic change" will be MFC'd is a bad one.  It's safe to
> say that when 8.3-RELEASE comes out if this problem haunts you again,
> you will be mailing the list about it, and this cycle will continue
> until 9.0-RELEASE comes out.
>
> Does any developer/committer have familiarity with this issue and have
> some ideas as to what may have changed in CURRENT that addresses David's
> issue?  And if so, can that code be MFC'd safely or patches provided to
> David for RELENG_8 that he can try out?
>
> I'm CC'ing mav_at_ here (snd_hda(4) says he's one of the authors), although
> he may not have any knowledge of the code which may need to be MFC'd.
> He may be able to point us to who has a better idea though.
>

No worries Jeremy but thanks for your interest you seems to be the only 
one who believed my problem. The problem has been fixed in -STABLE, I 
don't know where but it works now ....

-- 
David Demelier