[Pkg-alsa-devel] Bug#408506: alsa-utils: Tries /dev/snd/controlC0, but only /dev/controlC0 exists

[Lionel Elie Mamane]

On Fri, Jan 26, 2007 at 07:31:36PM +0100, Elimar Riesebieter wrote:
> On Fri, 26 Jan 2007 the mental interface of
> Lionel Elie Mamane told:

>> Running alsamixer, it tries to open /dev/snd/controlC0 and falls back
>> to /dev/aloadC0, but only /dev/controlC0 exists on my system. I use
>> udev.

>> I then symlinked /dev/snd to ., and suddenly everything started to
>> work _much_ better: Before that I could use my sound card only through
>> OSS emulation, now all my programs recognise alsa.

>> So maybe the bug is in alsalib for not trying /dev/controlC0 (and the
>> same for the pcm* devices), maybe in the udev rules for creating these
>> devices in /dev/ and not /dev/snd/, ...

> To dive in your suggestion we need some more infos.
> Kerneldrivers or build from alsa-source?

Kernel drivers, self-compiled kernel with make-kpkg from 2.6.17-5
sources. amd64 kernel, i386 "main" userland, but the same happens with
in a amd64 chroot. But this means the problem could be specific to an
amd64 kernel with an i386 udev.

> Soundcard ?

00:06.0 Multimedia audio controller: nVidia Corporation nForce3 Audio (rev a2)

> Loaded modules ?

lionelm at harif:~$ lsmod|grep snd
snd_seq_dummy           4740  0
snd_seq_oss            39232  0
snd_seq_midi           10304  0
snd_rawmidi            30176  1 snd_seq_midi
snd_seq_midi_event      9280  2 snd_seq_oss,snd_seq_midi
snd_seq                63104  6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
snd_seq_device         10448  5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
snd_pcm_oss            48864  0
snd_mixer_oss          20672  1 snd_pcm_oss
snd_intel8x0           38440  1
snd_ac97_codec        117528  1 snd_intel8x0
snd_ac97_bus            3136  1 snd_ac97_codec
snd_pcm               104076  3 snd_pcm_oss,snd_intel8x0,snd_ac97_codec
snd_timer              26952  2 snd_seq,snd_pcm
snd                    68904  12 snd_seq_oss,snd_rawmidi,snd_seq,snd_seq_device,snd_pcm_oss,snd_mixer_oss,snd_intel8x0,snd_ac97_codec,snd_pcm,snd_timer
soundcore              12000  1 snd
snd_page_alloc         11536  2 snd_intel8x0,snd_pcm

> Output of $ cat /proc/asound/version
> Output of $ cat /proc/asound/modules

lionelm at harif:~$ cat /proc/asound/version
Advanced Linux Sound Architecture Driver Version 1.0.11rc4 (Wed Mar 22 10:27:24 2006 UTC).
lionelm at harif:~$ cat /proc/asound/modules
 0 snd_intel8x0
lionelm at harif:~$ cat /proc/asound/cards
 0 [nForce3        ]: NFORCE - NVidia nForce3
                      NVidia nForce3 with ALC650E at 0xff6fb000, irq 185
lionelm at harif:~$ cat /proc/asound/devices
  0: [ 0]   : control
  1:        : sequencer
 16: [ 0- 0]: digital audio playback
 18: [ 0- 2]: digital audio playback
 24: [ 0- 0]: digital audio capture
 25: [ 0- 1]: digital audio capture
 33:        : timer
lionelm at harif:~$ cat /proc/asound/pcm
00-02: Intel ICH - IEC958 : NVidia nForce3 - IEC958 : playback 1
00-01: Intel ICH - MIC ADC : NVidia nForce3 - MIC ADC : capture 1
00-00: Intel ICH : NVidia nForce3 : playback 1 : capture 1

>> Anyway, if this is not just a local problen, it is the kind of
>> things I would want to be solved before etch releases...

> Ok, I' ve set severity to wishlist then.

I don't understand why you think this is a "feature request" or "very
difficult to fix due to major design considerations" (that is,
wishlist severity), rather than a genuine bug, a problem.

>> Extract of strace of alsamixer (without the symlink):

>>  open("/dev/snd/controlC0", O_RDONLY)    = -1 ENOENT (No such file or directory)
>>  open("/dev/aloadC0", O_RDONLY)          = -1 ENOENT (No such file or directory)

> We are always looking in /dev/snd for alsa drivers so I dont know
> how aloadC0 was created.

It was not created, it does not exist, but when dev/snd/controlC0 does
not exist, alsamixer tries to open it.

> Do you use a custom ~/.asoundrc?

No. Nor a /etc/asound.conf, not anything like that.

-- 
Lionel