[alsa-devel] Realtek alc269 questions
I'm trying to fix the sound on some Dell Optiplex machines, which are running the ALC269 chip. I can "fix" them with some hda-verb commands, but that won't get me auto-mute, so I'll have to make a model (or possibly more than one) for them.
Anyway, I have a pair of questions:
1) In dmesg, there is a "SKU not ready" message at startup. What is SKU, and is it a problem that it is not ready?
2) There are two DACs, and all models so far seem to use NID 0x03 for headphones and NID 0x02 for everything else. But is there a single stream going through the controller and then splitted into two on the codec side, or is just one of the DACs activated at a time, depending on the jack sensing state?
At Mon, 26 Jul 2010 12:01:50 +0200, David Henningsson wrote:
I'm trying to fix the sound on some Dell Optiplex machines, which are running the ALC269 chip. I can "fix" them with some hda-verb commands, but that won't get me auto-mute, so I'll have to make a model (or possibly more than one) for them.
What is actually a problem?
Anyway, I have a pair of questions:
- In dmesg, there is a "SKU not ready" message at startup. What is SKU,
and is it a problem that it is not ready?
Realtek codecs suppose that the device PCI SSID is set up for indicating hardware functionalities. This is Realtek-specific, and most vendors don't follow. The driver spews some messages when the SSID isn't compliant.
- There are two DACs, and all models so far seem to use NID 0x03 for
headphones and NID 0x02 for everything else. But is there a single stream going through the controller and then splitted into two on the codec side, or is just one of the DACs activated at a time, depending on the jack sensing state?
Both DACs use the same single stream at the same time.
Takashi
2010-07-26 12:29, Takashi Iwai skrev:
At Mon, 26 Jul 2010 12:01:50 +0200, David Henningsson wrote:
I'm trying to fix the sound on some Dell Optiplex machines, which are running the ALC269 chip. I can "fix" them with some hda-verb commands, but that won't get me auto-mute, so I'll have to make a model (or possibly more than one) for them.
What is actually a problem?
It varies between the different models, but most problems I think come from that pin NIDs aren't the same as the auto-parser (or any of the models) assume.
This causes playback, recording, auto-mute etc to fail.
In addition, a minor issue is that one of the models are being mislabeled as ALC259.
Anyway, I have a pair of questions:
- In dmesg, there is a "SKU not ready" message at startup. What is SKU,
and is it a problem that it is not ready?
Realtek codecs suppose that the device PCI SSID is set up for indicating hardware functionalities. This is Realtek-specific, and most vendors don't follow. The driver spews some messages when the SSID isn't compliant.
- There are two DACs, and all models so far seem to use NID 0x03 for
headphones and NID 0x02 for everything else. But is there a single stream going through the controller and then splitted into two on the codec side, or is just one of the DACs activated at a time, depending on the jack sensing state?
Both DACs use the same single stream at the same time.
Thanks for your answers!
At Mon, 26 Jul 2010 13:03:41 +0200, David Henningsson wrote:
2010-07-26 12:29, Takashi Iwai skrev:
At Mon, 26 Jul 2010 12:01:50 +0200, David Henningsson wrote:
I'm trying to fix the sound on some Dell Optiplex machines, which are running the ALC269 chip. I can "fix" them with some hda-verb commands, but that won't get me auto-mute, so I'll have to make a model (or possibly more than one) for them.
What is actually a problem?
It varies between the different models, but most problems I think come from that pin NIDs aren't the same as the auto-parser (or any of the models) assume.
Then it's a BIOS bug :) You can override the pin default values in fixup table.
This causes playback, recording, auto-mute etc to fail.
In addition, a minor issue is that one of the models are being mislabeled as ALC259.
ALC259 / ALC271X identification was recently added, so the old drivers may show a different codec name.
thanks,
Takashi
participants (2)
-
David Henningsson -
Takashi Iwai