Re: [Ac100] New test deb for Oneiric

2011-10-10 Thread Marc Dietrich
Am Montag, 10. Oktober 2011, 12:22:47 schrieb Andrey Danin:
 10 октября 2011, 12:08 от Marc Dietrich marvi...@gmx.de:
  Am Samstag, 8. Oktober 2011, 11:45:20 schrieb Oliver Grawert:
   hi,
   
   Am Samstag, den 08.10.2011, 11:41 +0200 schrieb Oliver Grawert:
works fine for me after setting ...
   
   stops working after resume from suspend, we might need some
   re-initialization here, if that's possible from user space through some
   command we could add the re-init to a gnome-power-manager quirk, though
   i suspect the kernel needs to handle it rather than user space...
  
  maybe the amplifier just needs to be enabled again. Should be easy to add to
  nvec- event.c.
 
 Headphones don't work too (if hp are related to nvec, please point how it is 
 done).
 alc5632 registers state is restored after resume but with no effect.
 snd_soc_tegra_paz00 and snd_soc_alc5632 must be reloaded to fix this issue: 
 sudo
 rmmod snd_soc_tegra_paz00
 sudo rmmod snd_soc_alc5632
 sudo modprobe snd_soc_alc5632
 sudo modprobe snd_soc_tegra_paz00

aha, so it is not related to nvec (would be too easy to blame it for 
everything), but 
more like a codec init issue. At least this isolates the problem a bit.

Thanks

Marc



___
Mailing list: https://launchpad.net/~ac100
Post to : ac100@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ac100
More help   : https://help.launchpad.net/ListHelp


Re: [Ac100] New test deb for Oneiric

2011-10-10 Thread Oliver Grawert
hi,
Am Montag, den 10.10.2011, 13:10 +0200 schrieb Marc Dietrich:

  Headphones don't work too (if hp are related to nvec, please point how it 
  is done).
  alc5632 registers state is restored after resume but with no effect.
  snd_soc_tegra_paz00 and snd_soc_alc5632 must be reloaded to fix this issue: 
  sudo
  rmmod snd_soc_tegra_paz00
  sudo rmmod snd_soc_alc5632
  sudo modprobe snd_soc_alc5632
  sudo modprobe snd_soc_tegra_paz00
 
 aha, so it is not related to nvec (would be too easy to blame it for 
 everything), but 
 more like a codec init issue. At least this isolates the problem a bit.
 
removing/re-adding the modules does nothing here, nor does putting them
into the pm-utils config for modules to un/reload which does also walk
the the modules tree and unloads the dependencies (and re-loads them on
resume in the right order indeed) (... see the SUSPEND_MODULES variable)

i spent 3h on this now and never managed to get to a state where sound
played again after resume.

ciao
oli


___
Mailing list: https://launchpad.net/~ac100
Post to : ac100@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ac100
More help   : https://help.launchpad.net/ListHelp


Re: [Ac100] New test deb for Oneiric

2011-10-10 Thread Andrey Danin
Epic fail :)
cat /sys/kernel/debug/asoc/tegra-paz00/alc5632.0-001e/codec_reg
and read from i2c device give different results. According to dump from i2c 
device registers state after resume is different to state before suspend. 
Restoring regs state didn't fix sound after resume.
Additional investigations are needed.


10 октября 2011, 15:10 от Marc Dietrich marvi...@gmx.de:
 Am Montag, 10. Oktober 2011, 12:22:47 schrieb Andrey Danin:
  10 октября 2011, 12:08 от Marc Dietrich marvi...@gmx.de:
   Am Samstag, 8. Oktober 2011, 11:45:20 schrieb Oliver Grawert:
hi,
   
Am Samstag, den 08.10.2011, 11:41 +0200 schrieb Oliver Grawert:
 works fine for me after setting ...
   
stops working after resume from suspend, we might need some
re-initialization here, if that's possible from user space through some
command we could add the re-init to a gnome-power-manager quirk, though
i suspect the kernel needs to handle it rather than user space...
  
   maybe the amplifier just needs to be enabled again. Should be easy to add 
   to
   nvec- event.c.
 
  Headphones don't work too (if hp are related to nvec, please point how it 
  is done).
  alc5632 registers state is restored after resume but with no effect.
  snd_soc_tegra_paz00 and snd_soc_alc5632 must be reloaded to fix this issue: 
  sudo
  rmmod snd_soc_tegra_paz00
  sudo rmmod snd_soc_alc5632
  sudo modprobe snd_soc_alc5632
  sudo modprobe snd_soc_tegra_paz00
 
 aha, so it is not related to nvec (would be too easy to blame it for 
 everything), but
 more like a codec init issue. At least this isolates the problem a bit.
 
 Thanks
 
 Marc
 
 
___
Mailing list: https://launchpad.net/~ac100
Post to : ac100@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ac100
More help   : https://help.launchpad.net/ListHelp


Re: [Ac100] New test deb for Oneiric

2011-10-08 Thread Oliver Grawert
hi,
Am Samstag, den 08.10.2011, 11:41 +0200 schrieb Oliver Grawert:

 works fine for me after setting ...
stops working after resume from suspend, we might need some
re-initialization here, if that's possible from user space through some
command we could add the re-init to a gnome-power-manager quirk, though
i suspect the kernel needs to handle it rather than user space...

ciao
oli


___
Mailing list: https://launchpad.net/~ac100
Post to : ac100@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ac100
More help   : https://help.launchpad.net/ListHelp


Re: [Ac100] New test deb for Oneiric

2011-10-08 Thread Leon Romanovsky
On Sat, Oct 8, 2011 at 11:41, Oliver Grawert o...@ubuntu.com wrote:

 hi,
 Am Freitag, den 07.10.2011, 23:37 +0300 schrieb Jani Monoses:
  There's an alleged fix for the speakers (works for Marc) but others
  need to try reproducing it.
 ...
 
 http://startx.ro/~jani/linux-image-2.6.38-1001-ac100_2.6.38-1001.1_armel.deb
 ...
  linux-ac100 (2.6.38-1001.1) oneiric; urgency=low
 ...
* Disable PCI, build modules for MD/DM/RAID and Zcache.
* Update to latest git upstream, fixes sound from speakers.
* Based on upstream commit dc61519faf9523819e783b08970b0a1b5a50ec73

 works fine for me after setting the following in alsamixer:
 SpeakerOut Mux [HPOut Mix]

 (i guess we will need some alsaucm or alsactrl files in an SRU to
 initialize that properly)

 The funny thing that me and stuw have no speakers working with latest
kernel.


 ciao
oli





 ___
 Mailing list: https://launchpad.net/~ac100
 Post to : ac100@lists.launchpad.net
 Unsubscribe : https://launchpad.net/~ac100
 More help   : https://help.launchpad.net/ListHelp




-- 
Leon Romanovsky | Independent Linux Consultant
www.leon.nu | l...@leon.nu
___
Mailing list: https://launchpad.net/~ac100
Post to : ac100@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ac100
More help   : https://help.launchpad.net/ListHelp


Re: [Ac100] New test deb for Oneiric

2011-10-08 Thread Oliver Grawert
hi,
Am Samstag, den 08.10.2011, 12:19 +0200 schrieb hiro:
 here's my asound.state. use alsactl to restore it's settings to your setup.
 what about an init file /usr/share/alsa/init/00main ?
you should never use asound.state manually, it is modified and parsed by
alsa all the time and you might confuse it if you manually change the
file. this is a 100% system file that shouldnt be touched by users.

00main would be an option, but the whole ancient init hackery is
considered obsolete upstream, so that would also only be temporary ... 

ASoC sound is supposed to use ucm nowadays, so we need to create the
proper ucm configuration to initialize the sound properly. alongside an
SRU upload for oneiric this indeed needs to go upstream.

ciao
oli



___
Mailing list: https://launchpad.net/~ac100
Post to : ac100@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ac100
More help   : https://help.launchpad.net/ListHelp