[alsa-devel] Problems with Intel HDA driver
On my new laptop there were no sound with snd-hda-intel driver, so I tried to change model parameter. It seems that with model=hp-m4 my soundcard mostly works (except that main speakers still play when I plug headphones).
I've attached my alsa-info.txt script.
At Mon, 14 Sep 2009 14:33:34 +0300, Andrius Štikonas wrote:
On my new laptop there were no sound with snd-hda-intel driver, so I tried to change model parameter. It seems that with model=hp-m4 my soundcard mostly works (except that main speakers still play when I plug headphones).
Try rather model=hp-dv5. The speaker mute issue could be fixed via enable_msi=1 option.
Takashi
Takashi Iwai wrote On 14-09-2009 15:25:
At Mon, 14 Sep 2009 14:33:34 +0300, Andrius Štikonas wrote:
On my new laptop there were no sound with snd-hda-intel driver, so I tried to change model parameter. It seems that with model=hp-m4 my soundcard mostly works (except that main speakers still play when I plug headphones).
Try rather model=hp-dv5. The speaker mute issue could be fixed via enable_msi=1 option.
grr another dv6. Yes hp-dv5 works on some dv6's but apparently not on everyone of them. Also I managed to get sound without passing dv5 but enabling EAPD through hda-analyzer.
I'll try enable_msi=1 to see what happens.
At Mon, 14 Sep 2009 15:36:54 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 15:25:
At Mon, 14 Sep 2009 14:33:34 +0300, Andrius Štikonas wrote:
On my new laptop there were no sound with snd-hda-intel driver, so I tried to change model parameter. It seems that with model=hp-m4 my soundcard mostly works (except that main speakers still play when I plug headphones).
Try rather model=hp-dv5. The speaker mute issue could be fixed via enable_msi=1 option.
grr another dv6. Yes hp-dv5 works on some dv6's but apparently not on everyone of them. Also I managed to get sound without passing dv5 but enabling EAPD through hda-analyzer.
Yeah, I think we can add EAPD handling as default for IDT92HD71xx codecs. I've never seen machines that can't live with GPIO0 bit up.
Takashi
Takashi Iwai wrote On 14-09-2009 15:36:
Yeah, I think we can add EAPD handling as default for IDT92HD71xx codecs. I've never seen machines that can't live with GPIO0 bit up.
enable_msi doesn't seem to affect the auto-muting problem. I played with every single node in hda-analyzer but can't get auto-mute work. I can manually mute the HP output and the internal speakers by checking/unchecking OUT checkboxes on relevant NIDs. There's a 8xn matrix for GPIO settings, I played with them as well to set unsol but nope.
I want to remember that dell-m4-1 makes automute work but breaks the mixer channels:
http://www.spinics.net/lists/alsa-devel/msg26714.html
Err, I know that I didn't answer the last question in the thread as I'm a little bit lost in the NID pool and didn't have time to investigate further.
At Mon, 14 Sep 2009 16:15:02 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 15:36:
Yeah, I think we can add EAPD handling as default for IDT92HD71xx codecs. I've never seen machines that can't live with GPIO0 bit up.
enable_msi doesn't seem to affect the auto-muting problem. I played with every single node in hda-analyzer but can't get auto-mute work. I can manually mute the HP output and the internal speakers by checking/unchecking OUT checkboxes on relevant NIDs. There's a 8xn matrix for GPIO settings, I played with them as well to set unsol but nope.
I want to remember that dell-m4-1 makes automute work but breaks the mixer channels:
http://www.spinics.net/lists/alsa-devel/msg26714.html
Err, I know that I didn't answer the last question in the thread as I'm a little bit lost in the NID pool and didn't have time to investigate further.
Could you check whether this still valid?
Basically the model option doesn't change the auto-mute behavior. It just changes which pin to be detected and which pin to mute. So, if model=dell-m4-1 makes the auto-muting working, it means that the BIOS setup is broken.
Takashi
At Mon, 14 Sep 2009 15:15:53 +0200, I wrote:
At Mon, 14 Sep 2009 16:15:02 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 15:36:
Yeah, I think we can add EAPD handling as default for IDT92HD71xx codecs. I've never seen machines that can't live with GPIO0 bit up.
enable_msi doesn't seem to affect the auto-muting problem. I played with every single node in hda-analyzer but can't get auto-mute work. I can manually mute the HP output and the internal speakers by checking/unchecking OUT checkboxes on relevant NIDs. There's a 8xn matrix for GPIO settings, I played with them as well to set unsol but nope.
I want to remember that dell-m4-1 makes automute work but breaks the mixer channels:
http://www.spinics.net/lists/alsa-devel/msg26714.html
Err, I know that I didn't answer the last question in the thread as I'm a little bit lost in the NID pool and didn't have time to investigate further.
Could you check whether this still valid?
Basically the model option doesn't change the auto-mute behavior. It just changes which pin to be detected and which pin to mute. So, if model=dell-m4-1 makes the auto-muting working, it means that the BIOS setup is broken.
I looked through alsa-info output now, and found the culprit. NID 0x0f is set as a line-out, and currently the driver doesn't don't do auto-muting for line-out unless a headphone is present. Changing it to a headphone should fix. Or you can give hp_detect=1 hint via sysfs and reconfigure.
Takashi
Takashi Iwai wrote On 14-09-2009 16:23:
I looked through alsa-info output now, and found the culprit. NID 0x0f is set as a line-out, and currently the driver doesn't don't do auto-muting for line-out unless a headphone is present. Changing it to a headphone should fix. Or you can give hp_detect=1 hint via sysfs and reconfigure.
Yes! hp_detect=1 worked altogether with passing hp-dv5 model. I'm ready to test any patch you suggest.
At Mon, 14 Sep 2009 16:41:10 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 16:23:
I looked through alsa-info output now, and found the culprit. NID 0x0f is set as a line-out, and currently the driver doesn't don't do auto-muting for line-out unless a headphone is present. Changing it to a headphone should fix. Or you can give hp_detect=1 hint via sysfs and reconfigure.
Yes! hp_detect=1 worked altogether with passing hp-dv5 model. I'm ready to test any patch you suggest.
Good to hear. The final patch is already merged to sound git tree, and a new snapshot tarball is being created and updated now. FWIW, the essential change is below.
thanks,
Takashi
---
Takashi Iwai wrote On 14-09-2009 16:46:
At Mon, 14 Sep 2009 16:41:10 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 16:23:
I looked through alsa-info output now, and found the culprit. NID 0x0f is set as a line-out, and currently the driver doesn't don't do auto-muting for line-out unless a headphone is present. Changing it to a headphone should fix. Or you can give hp_detect=1 hint via sysfs and reconfigure.
Yes! hp_detect=1 worked altogether with passing hp-dv5 model. I'm ready to test any patch you suggest.
Good to hear. The final patch is already merged to sound git tree, and a new snapshot tarball is being created and updated now. FWIW, the essential change is below.
Ok tested today's snapshot, it works like a charm, thanks again.
At Tue, 15 Sep 2009 12:04:38 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 16:46:
At Mon, 14 Sep 2009 16:41:10 +0300, Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 16:23:
I looked through alsa-info output now, and found the culprit. NID 0x0f is set as a line-out, and currently the driver doesn't don't do auto-muting for line-out unless a headphone is present. Changing it to a headphone should fix. Or you can give hp_detect=1 hint via sysfs and reconfigure.
Yes! hp_detect=1 worked altogether with passing hp-dv5 model. I'm ready to test any patch you suggest.
Good to hear. The final patch is already merged to sound git tree, and a new snapshot tarball is being created and updated now. FWIW, the essential change is below.
Ok tested today's snapshot, it works like a charm, thanks again.
Great. Thank you guys for testing!
Takashi
Hi Takashi, I just updated my driver and libs to 1.0.21 to see whether this is fixed. Unfortunately, when I plug my headphones, the laptop speakers are still playing.
http://www.alsa-project.org/db/?f=c0082231c70ef69de999506f3d59dec0c0cdb633
On Yaum al-Ithnain 24 Ramadan 1430 4:23:05 pm Takashi Iwai wrote:
At Mon, 14 Sep 2009 15:15:53 +0200,
I wrote:
At Mon, 14 Sep 2009 16:15:02 +0300,
Ozan Çağlayan wrote:
Takashi Iwai wrote On 14-09-2009 15:36:
Yeah, I think we can add EAPD handling as default for IDT92HD71xx codecs. I've never seen machines that can't live with GPIO0 bit up.
enable_msi doesn't seem to affect the auto-muting problem. I played with every single node in hda-analyzer but can't get auto-mute work. I can manually mute the HP output and the internal speakers by checking/unchecking OUT checkboxes on relevant NIDs. There's a 8xn matrix for GPIO settings, I played with them as well to set unsol but nope.
I want to remember that dell-m4-1 makes automute work but breaks the mixer channels:
http://www.spinics.net/lists/alsa-devel/msg26714.html
Err, I know that I didn't answer the last question in the thread as I'm a little bit lost in the NID pool and didn't have time to investigate further.
Could you check whether this still valid?
Basically the model option doesn't change the auto-mute behavior. It just changes which pin to be detected and which pin to mute. So, if model=dell-m4-1 makes the auto-muting working, it means that the BIOS setup is broken.
I looked through alsa-info output now, and found the culprit. NID 0x0f is set as a line-out, and currently the driver doesn't don't do auto-muting for line-out unless a headphone is present. Changing it to a headphone should fix. Or you can give hp_detect=1 hint via sysfs and reconfigure.
Takashi
participants (4)
-
Andrius Štikonas -
Munzir Taha (منذر طه) -
Ozan Çağlayan -
Takashi Iwai