[alsa-devel] seg fault with 1.0.17rc2

[Jerry Geis]

Takashi Iwai wrote:
> At Thu, 26 Jun 2008 12:03:24 -0400,
> Jerry Geis wrote:
>   
>>
>> Takashi Iwai wrote:
>>     
>>> At Thu, 26 Jun 2008 10:38:57 -0400,
>>> Jerry Geis wrote:
>>>   
>>>       
>>>> #0  0xb7e892ff in memcpy () from /lib/tls/libc.so.6
>>>> #1  0xb74e40a9 in snd_pcm_area_copy (dst_area=0xb7125b70, dst_offset=0, 
>>>> src_area=0x81dc1c0, src_offset=170, samples=0, 
>>>> format=SND_PCM_FORMAT_S16_LE) at pcm.c:2589
>>>>     
>>>>         
>>> samples = 0 and...
>>>
>>>   
>>>       
>>>> #2  0xb74e438d in snd_pcm_areas_copy (dst_areas=0xb7125b7c, 
>>>> dst_offset=0, src_areas=0x81dc1cc, src_offset=170, channels=1, 
>>>> frames=122, format=SND_PCM_FORMAT_S16_LE) at pcm.c:2736
>>>>     
>>>>         
>>> ... here frames = 122.  Something inconsistent around here.
>>> snd_pcm_areas_copy() must passe samples=frames when channels=1.
>>> Could you check the values via gdb?
>>>
>>>
>>> Takashi
>>>
>>>   
>>>       
>> Takashi,
>>
>> I am not sure what your asking me. The output I provided is gdb what else
>> can I check? Really anxious to get this USB sound device playing 
>> consistantly.
>>     
>
> Check whether frames still 122 in frame#1, for example.
>
>   
>> Is there a better asound.conf to use?
>>     
>
> The strange thing is that the recent config for usb-audio also uses
> dmix/dsnoop.  And you don't get any errors with the system-default
> config?
>
>
> Takashi
>
>   
Takashi,

checking frames still 122 in frame #1 is way over my expertise.

With this asound.conf file It plays but choppy audio.

defaults.ctl.card 0
defaults.pcm.card 0

pcm.card0 {
  type hw
  card 0
}

pcm.dmixer {
  type dmix
  ipc_key 1025
  slave {
    pcm "hw:0,0"
    period_time 0
    period_size 2048
    buffer_size 32768
    rate 48000
  }
  bindings {
    0 0
    1 1
  }
}
pcm.skype {
  type asym

  playback.pcm "dmixer"
  capture.pcm "card0"
}

pcm.!default {
  type plug
  slave.pcm "skype"
}



Jerry