hda-intel module hanging if ACPI enabled

hda-intel module hanging if ACPI enabled

[Michael Downey]

I have an Intel Desktop D945GTP motherboard and I am running gentoo with 
kernel 2.6.13-r2.  If I enable ACPI then the system will hang whenever 
the hda-intel kernel module is loaded.  I did some more testing and I 
found that if I enable "Local APIC support for uniprocessors" and 
"IO-APIC support on uniprocessors" in the kernel then the hda-intel 
module will load when ACPI is enabled.  The codec that is used on the 
motherboard is a Sigmatel STAC9221 A1.  Is APIC a requirement for this 
setup?  Also I've tried 1.0.10rc1 of the alsa-driver and got the same 
results.

Also I was wondering how to go about retasking the audio jacks?  The 
default mixer only provides PCM, FRONT, CAPTURE, INPUT SELECT, and MIX.  
I would like to know how to set up 2 of the jacks to be output jacks 
which as far as I know is possible with hda-intel.

Thanks,


Michael Downey
Zymeta Corporation



-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl

Re: hda-intel module hanging if ACPI enabled

[Matt Porter]

On Thu, Sep 29, 2005 at 11:36:29AM -0600, Michael Downey wrote:
> I have an Intel Desktop D945GTP motherboard and I am running gentoo with 
> kernel 2.6.13-r2.  If I enable ACPI then the system will hang whenever 
> the hda-intel kernel module is loaded.  I did some more testing and I 
> found that if I enable "Local APIC support for uniprocessors" and 
> "IO-APIC support on uniprocessors" in the kernel then the hda-intel 
> module will load when ACPI is enabled.  The codec that is used on the 
> motherboard is a Sigmatel STAC9221 A1.  Is APIC a requirement for this 
> setup?  Also I've tried 1.0.10rc1 of the alsa-driver and got the same 
> results.
 
I hadn't noticed this myself since I have local apic support on in
my kernel, but have heard the same thing from other users. Haven't
looked into it further yet.

> Also I was wondering how to go about retasking the audio jacks?  The 
> default mixer only provides PCM, FRONT, CAPTURE, INPUT SELECT, and MIX.  
> I would like to know how to set up 2 of the jacks to be output jacks 
> which as far as I know is possible with hda-intel.

I am working on some updates to the SigmaTel HDA patch to support these
motherboards. The basic issue is that the BIOS doesn't set up the
proper configuration default registers on these motherboards so we
need to detect the motherboard and load the proper configuration
defaults. I also added in some switches to retask those jacks.

I have an issue with my boards in that they don't properly reflect
the Subsystem ID that should be used for identifying the motherboard
and jack configuration.  Can you do a
'grep Subsystem /proc/asound/card0/codec#2' and post the results?
I'd like to verify that your board shows the proper ID.

-Matt


-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl

Re: hda-intel module hanging if ACPI enabled

[Michael Downey]

Matt Porter wrote:

>On Thu, Sep 29, 2005 at 11:36:29AM -0600, Michael Downey wrote:
>  
>
>>I have an Intel Desktop D945GTP motherboard and I am running gentoo with 
>>kernel 2.6.13-r2.  If I enable ACPI then the system will hang whenever 
>>the hda-intel kernel module is loaded.  I did some more testing and I 
>>found that if I enable "Local APIC support for uniprocessors" and 
>>"IO-APIC support on uniprocessors" in the kernel then the hda-intel 
>>module will load when ACPI is enabled.  The codec that is used on the 
>>motherboard is a Sigmatel STAC9221 A1.  Is APIC a requirement for this 
>>setup?  Also I've tried 1.0.10rc1 of the alsa-driver and got the same 
>>results.
>>    
>>
> 
>I hadn't noticed this myself since I have local apic support on in
>my kernel, but have heard the same thing from other users. Haven't
>looked into it further yet.
>
>  
>
>>Also I was wondering how to go about retasking the audio jacks?  The 
>>default mixer only provides PCM, FRONT, CAPTURE, INPUT SELECT, and MIX.  
>>I would like to know how to set up 2 of the jacks to be output jacks 
>>which as far as I know is possible with hda-intel.
>>    
>>
>
>I am working on some updates to the SigmaTel HDA patch to support these
>motherboards. The basic issue is that the BIOS doesn't set up the
>proper configuration default registers on these motherboards so we
>need to detect the motherboard and load the proper configuration
>defaults. I also added in some switches to retask those jacks.
>
>I have an issue with my boards in that they don't properly reflect
>the Subsystem ID that should be used for identifying the motherboard
>and jack configuration.  Can you do a
>'grep Subsystem /proc/asound/card0/codec#2' and post the results?
>I'd like to verify that your board shows the proper ID.
>
>-Matt
>  
>
Thanks for the quick reply. Here's my output:
Subsystem Id: 0x0

The board has a BIOS revision string of: 
NT9410J.86A.1348.2005.0609.1737. I see that there is a newer BIOS 2104 
released on Sept. 2 so updating the BIOS might make a difference. I will 
probably try the update in a day or two so I'll post my results then.

Michael Downey
Zymeta Corporation



-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl

Re: hda-intel module hanging if ACPI enabled

[Matt Porter]

On Thu, Sep 29, 2005 at 12:13:49PM -0600, Michael Downey wrote:
> Matt Porter wrote:
> >I am working on some updates to the SigmaTel HDA patch to support these
> >motherboards. The basic issue is that the BIOS doesn't set up the
> >proper configuration default registers on these motherboards so we
> >need to detect the motherboard and load the proper configuration
> >defaults. I also added in some switches to retask those jacks.
> >
> >I have an issue with my boards in that they don't properly reflect
> >the Subsystem ID that should be used for identifying the motherboard
> >and jack configuration.  Can you do a
> >'grep Subsystem /proc/asound/card0/codec#2' and post the results?
> >I'd like to verify that your board shows the proper ID.
> >
> Thanks for the quick reply. Here's my output:
> Subsystem Id: 0x0

Ugh, this is supposed to have a unique identifying value. Did
you get a very early board from Intel?

> The board has a BIOS revision string of: 
> NT9410J.86A.1348.2005.0609.1737. I see that there is a newer BIOS 2104 
> released on Sept. 2 so updating the BIOS might make a difference. I will 
> probably try the update in a day or two so I'll post my results then.

You can check. However, I've looked at the BIOS updates that appear
to have no changes related to the HDA config default regs. My
current patches load up the appropriate config default values but
I currently have an issue with reliably identifying the appropriate
board for each set of values.

You'll also need the patch update to get the retasking controls.
 
I'm working with SigmaTel on this so let me check on some things
regarding board ID and then I'll hopefully have a patch to post
that will work for you.  We may have to resort to manually
specifying the model type for these boards...I'm hoping to
avoid that though.

-Matt


-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl

Re: hda-intel module hanging if ACPI enabled

[Michael Downey]

Matt Porter wrote:

>Ugh, this is supposed to have a unique identifying value. Did
>you get a very early board from Intel?
>
>  
>
>>The board has a BIOS revision string of: 
>>NT9410J.86A.1348.2005.0609.1737. I see that there is a newer BIOS 2104 
>>released on Sept. 2 so updating the BIOS might make a difference. I will 
>>probably try the update in a day or two so I'll post my results then.
>>    
>>
>
>You can check. However, I've looked at the BIOS updates that appear
>to have no changes related to the HDA config default regs. My
>current patches load up the appropriate config default values but
>I currently have an issue with reliably identifying the appropriate
>board for each set of values.
>
>You'll also need the patch update to get the retasking controls.
> 
>I'm working with SigmaTel on this so let me check on some things
>regarding board ID and then I'll hopefully have a patch to post
>that will work for you.  We may have to resort to manually
>specifying the model type for these boards...I'm hoping to
>avoid that though.
>
>-Matt
>  
>
I've updated the board to the latest BIOS revision and I'm still getting 
a Subsystem ID of 0. So updating the BIOS didn't help. As far as I know 
this is a retail board. It looks like it was manufactured sometime after 
June this year so it is possible it was a fairly early board.

Also is the patch update for the driver available now? I've looked 
around and didn't notice any changes to any code in CVS so if you have a 
patch I'd be glad to try it out. I'm running 1.0.9b of ALSA but I don't 
mind updating to the latest and greatest.


Michael



-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl