possible intel8x0 problem with nforce chip-set

possible intel8x0 problem with nforce chip-set

[Richard Lemon]

Hi Guys,

I have been using the alsa api for a while working on some multimedia
stuff.

I believe I have found a bug with the intel8x0 driver when running the
nforce chip-set.

The problem appears to be an uninitialized variable type problem. The
first time you call snd_pcm_mmap_begin() the channels are returned in
their pairs in a random order. That is the rear channels may end up in
the front or Center/LFE slots, etc, the channel order does not change
from then on.

I compiled up the pcm.c test code and tried that out. This shows the
problem, only worse. If you change the generate_sine function to the
following to pick out one pair of channels only.

if ((chn == 2) || (chn == 3))
	*samples[chn] = ires;
else
	*samples[chn] = 0;

then call pcm with the following...
./pcm -r 48000 -c 6 -m direct_noninterleaved

I would expect silence from all but the rear pair. What we get is a
sequencing of the audio through the 6 channels. It makes no difference
which of the modes or rates I use.

Am I just misusing the api or is this a bug?

Here are the hardware specs.

ShuttleX pc
North bridge-Nvidia nForce2, South bridge-nVidia MCP-T
Realtek ALC650 6 Ch, Support Digital SPDIF Output

for more details on hardware.
http://www.us.shuttle.com/specs_access.asp?pro_id=284

running Mandrake Linux 9.0.

I have tested several of the alsa releases and am currently using 0.9.6.

Richard Lemon.




-------------------------------------------------------
This SF.net email is sponsored by OSDN developer relations
Here's your chance to show off your extensive product knowledge
We want to know what you know. Tell us and you have a chance to win $100
http://www.zoomerang.com/survey.zgi?HRPT1X3RYQNC5V4MLNSV3E54

Re: possible intel8x0 problem with nforce chip-set

[Takashi Iwai]

At 21 Oct 2003 13:18:39 +0930,
Richard Lemon wrote:
> 
> 
> Hi Guys,
> 
> I have been using the alsa api for a while working on some multimedia
> stuff.
> 
> I believe I have found a bug with the intel8x0 driver when running the
> nforce chip-set.
> 
> The problem appears to be an uninitialized variable type problem. The
> first time you call snd_pcm_mmap_begin() the channels are returned in
> their pairs in a random order. That is the rear channels may end up in
> the front or Center/LFE slots, etc, the channel order does not change
> from then on.
> 
> I compiled up the pcm.c test code and tried that out. This shows the
> problem, only worse. If you change the generate_sine function to the
> following to pick out one pair of channels only.
> 
> if ((chn == 2) || (chn == 3))
> 	*samples[chn] = ires;
> else
> 	*samples[chn] = 0;
> 
> then call pcm with the following...
> ./pcm -r 48000 -c 6 -m direct_noninterleaved
> 
> I would expect silence from all but the rear pair. What we get is a
> sequencing of the audio through the 6 channels. It makes no difference
> which of the modes or rates I use.
> 
> Am I just misusing the api or is this a bug?

it might be a hardware duplication / downmixing.
check whether all "Duplicate Front", "Surround Down Mix"
and "Center/LFE Down Mix" switches are off on alsamixer.


Takashi


-------------------------------------------------------
This SF.net email is sponsored by OSDN developer relations
Here's your chance to show off your extensive product knowledge
We want to know what you know. Tell us and you have a chance to win $100
http://www.zoomerang.com/survey.zgi?HRPT1X3RYQNC5V4MLNSV3E54