Basically my issue reads very very similar to the one from bug id 104952.
I am running Ubuntu 18.04.1 LTS on an Gigabyte Prix GB-BLPD-5005 (Intel Pentium J5005 Gemini Lake). Running mainline kernel:
Linux brix 4.20.5-042005-generic #201901260434 SMP Sat Jan 26 09:36:17 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
I have attached a Yamaha AVR via HDMI and after that a Samsung TV:
Whenever I switch off the AVR and turn it back on after some longer time (> 60 seconds or a couple of minutes - not tested that much yet) - I loos HDMI audio and have the following messages in dmesg:
[52092.976001] snd_hda_intel 0000:00:0e.0: azx_get_response timeout, switching to polling mode: last cmd=0x20bf8100
[52093.983688] snd_hda_intel 0000:00:0e.0: No response from codec, disabling MSI: last cmd=0x20bf8100
[52094.988008] snd_hda_intel 0000:00:0e.0: azx_get_response timeout, switching to single_cmd mode: last cmd=0x20bf8100
[52095.272444] snd_hda_codec_hdmi hdaudioC0D2: Unable to sync register 0x2f0d00. -5
For getting HDMI audio restored I need to shutdown X and unload/reload snd-hda-intel.
Please advice what kind of debug logs are needed for investigation and I will happily provide them.
Can you try to reproduce the issue using latest drm-tip (https://cgit.freedesktop.org/drm-tip) and kernel parameters drm.debug=0x1e log_buf_len=4M, and if the problem persists attach the full dmesg from boot.
I tried drm-tip from here:
hope this is the right place.
Linux brix 5.0.0-994-generic #201901310201 SMP Thu Jan 31 02:03:36 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
Seems the issue is not happening with that version. At least I was not able to reproduce it now. So very nice :)
Question is how this drm state will hit a stable mainline kernel? Will it be automagically in Kernel 5.0? Or does it make sense (or is it even possible) to apply drm-tip to my former used 4.20.5 kernel by self compiling?
I spoke to soon. Had the issue again. I will switch on debugging again and will post a log once I can reproduce again.
Created attachment 143281 [details]
no audio after avr power on
Here you go. Was able to reproduce it. Hope the dmesg shows something useful. Let me know what else you need from me.
Hi, I had the exact same problem (J5005 NUC) and I _think_ I fixed it by accident just the other day by swapping HDMI cables. It has not happened for two days so far, so fingers crossed. It it happens again, I will try to recreate the issue with drm-tip.
Otherwise the procedure to get it up and running was the same as you describe,
- shutdown xorg
- unload module
- load module
- restart xorg
No idea on why the HDMI cable would be an issue though.
My setup is kinda similar to yours:
NUC --> HDMI splitter (4 inputs, 1 output) -+---> HDMI Projector
+---> S/PDIF Speakers
Contacted audio team to respond on this issue. If there is any feedback from the audio team I will post the details here.
Can you try with latest upstream kernel, 5.2 or git?