Comment 18 for bug 1525554

Today I added some debug output to 4.4-rc8. That's what I got so far:

When the codec probing fails, it takes exactly a bit more than 5 seconds from sending the verb
[ 394.482417] snd_hda_intel 0000:00:1f.3: snd_hdac_bus_send_cmd: Try to send val=0xf0000
to getting:
[ 399.524257] snd_hda_intel 0000:00:1f.3: spurious response 0x14f150f4:0x0, last cmd=0x0f0000

When it works, after reloading a few time, the probing looks well:
[ 798.042375] snd_hda_intel 0000:00:1f.3: bound 0000:00:02.0 (ops i915_audio_component_bind_ops [i915])
[ 798.055895] snd_hda_intel 0000:00:1f.3: codec_mask = 0x5
[ 798.057519] snd_hda_intel 0000:00:1f.3: snd_hdac_bus_send_cmd: Try to send val=0xf0000
[ 798.057525] snd_hda_intel 0000:00:1f.3: azx_rirb_get_response starting, last cmd=0x000f0000
[ 798.057578] snd_hda_intel 0000:00:1f.3: codec #0 probed OK
[ 798.057580] snd_hda_intel 0000:00:1f.3: snd_hdac_bus_send_cmd: Try to send val=0x200f0000
[ 798.057583] snd_hda_intel 0000:00:1f.3: azx_rirb_get_response starting, last cmd=0x200f0000
[ 798.057615] snd_hda_intel 0000:00:1f.3: codec #2 probed OK

I'm just starting to understand the code. If somebody with a better background has some advice what could be useful to try, post it here or ping "skoe" at freenode/#ubuntu-kernel.

btw: I'm quite sure that the script alsa_setup.sh has nothing to do with the issue, but the force-reload is sufficient.