snd_hda_intel on linux 2.6.30

snd_hda_intel on linux 2.6.30

[Tony Hoyle]

Hi,

I've been trying to get sound working on a Poulsbo board using 2.6.30. 
Previously, with the available patches sound worked OK.

2.6.30 has most of the patches already in, and the the Intel HDA is 
indeed recognized, but no sound appears.  Boot logs show:

[   30.948427] HDA Intel 0000:00:1b.0: enabling device (0000 -> 0002)
[   30.958108] HDA Intel 0000:00:1b.0: PCI INT A -> GSI 16 (level, low) 
-> IRQ 16
[   30.968030] HDA Intel 0000:00:1b.0: setting latency timer to 64
[   31.192916] input: HDA Intel MID Line In at Ext Rear Jack as 
/class/input/input3
[   31.204980] input: HDA Intel MID Mic at Ext Front Jack as 
/class/input/input4
[   31.218872] input: HDA Intel MID Line Out at Ext Rear Jack as 
/class/input/input5
[   31.234356] input: HDA Intel MID HP Out at Ext Front Jack as 
/class/input/input6

followed later by

[   50.875374] hda-intel: Invalid position buffer, using LPIB read 
method instead.
[   50.875416] hda-intel: IRQ timing workaround is activated for card 
#0. Suggest a bigger bdl_pos_adj.


I tried backporting (forward porting?) the missing fixups from the older 
patch, but apart from adding the message:

[   30.968039] Do fixup for Poulsbo <6>D0 or newer stepping

I still didn't work (of course the possibility I stuffed up my patching 
looms large there, and I'd rather work with virgin source if it's possible).

What's the best way to go about debugging this?  Is there an official 
(or unofficial) patch for Poulsbo on 2.6.30 knocking around somewhere?

Tony

Re: snd_hda_intel on linux 2.6.30

[Takashi Iwai]

At Mon, 22 Jun 2009 20:49:09 +0100,
Tony Hoyle wrote:
> 
> Hi,
> 
> I've been trying to get sound working on a Poulsbo board using 2.6.30. 
> Previously, with the available patches sound worked OK.

What kind of patches are they?

> 2.6.30 has most of the patches already in, and the the Intel HDA is 
> indeed recognized, but no sound appears.  Boot logs show:
> 
> [   30.948427] HDA Intel 0000:00:1b.0: enabling device (0000 -> 0002)
> [   30.958108] HDA Intel 0000:00:1b.0: PCI INT A -> GSI 16 (level, low) 
> -> IRQ 16
> [   30.968030] HDA Intel 0000:00:1b.0: setting latency timer to 64
> [   31.192916] input: HDA Intel MID Line In at Ext Rear Jack as 
> /class/input/input3
> [   31.204980] input: HDA Intel MID Mic at Ext Front Jack as 
> /class/input/input4
> [   31.218872] input: HDA Intel MID Line Out at Ext Rear Jack as 
> /class/input/input5
> [   31.234356] input: HDA Intel MID HP Out at Ext Front Jack as 
> /class/input/input6

These are all OK.  The lines with "input" are for jack input layer.

> followed later by
> 
> [   50.875374] hda-intel: Invalid position buffer, using LPIB read 
> method instead.
> [   50.875416] hda-intel: IRQ timing workaround is activated for card 
> #0. Suggest a bigger bdl_pos_adj.

These are usually harmless.

> I tried backporting (forward porting?) the missing fixups from the older 
> patch, but apart from adding the message:
> 
> [   30.968039] Do fixup for Poulsbo <6>D0 or newer stepping
> 
> I still didn't work (of course the possibility I stuffed up my patching 
> looms large there, and I'd rather work with virgin source if it's possible).
> 
> What's the best way to go about debugging this?  Is there an official 
> (or unofficial) patch for Poulsbo on 2.6.30 knocking around somewhere?

Run alsa-info.sh with --no-upload option on working (old) and non-working
(2.6.30) kernels, and compare the generated files.


Takashi