Asoc: Intel: SST (CHT) regression in asoc/for-5.11

[Pierre-Louis Bossart]

On 11/29/20 6:24 AM, Hans de Goede wrote:
> Hi All,
> 
> To test the code to dynamically switch between SST/SOF support on BYT/CHT
> from the kernel commandline I merged:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git/log/?h=for-5.11
> 
> Into my personal tree (mostly Linus' master + some pending patches from
> myself).
> 
> After this I was getting the following errors in dmesg when using sound on
> a Medion E2228T laptop with a CHT SoC + NAU8824 codec:
> 
> [   53.805205] intel_sst_acpi 808622A8:00: Wait timed-out condition:0x0, msg_id:0x1 fw_state 0
> [   53.805479] intel_sst_acpi 808622A8:00: fw returned err -16
> [   53.806281] sst-mfld-platform sst-mfld-platform: ASoC: PRE_PMD: pcm0_in event failed: -16
> [   54.829548] intel_sst_acpi 808622A8:00: Wait timed-out condition:0x0, msg_id:0x1 fw_state 0
> [   54.829596] intel_sst_acpi 808622A8:00: fw returned err -16
> [   54.829668] sst-mfld-platform sst-mfld-platform: ASoC: POST_PMD: media0_out event failed: -
> [   55.853230] intel_sst_acpi 808622A8:00: Wait timed-out condition:0x0, msg_id:0x1 fw_state 0
> [   55.853244] intel_sst_acpi 808622A8:00: fw returned err -16
> [   55.853269] sst-mfld-platform sst-mfld-platform: ASoC: POST_PMD: codec_out0 mix 0 event fai
> [   56.876435] intel_sst_acpi 808622A8:00: Wait timed-out condition:0x0, msg_id:0x1 fw_state 0
> [   56.876481] intel_sst_acpi 808622A8:00: fw returned err -16
> [   56.876563] sst-mfld-platform sst-mfld-platform: ASoC: POST_PMD: media0_out mix 0 event fai
> [   61.847455] intel_sst_acpi 808622A8:00: FW sent error response 0x40015
> [   61.847564] intel_sst_acpi 808622A8:00: fw returned err -1
> [   61.847659] sst-mfld-platform sst-mfld-platform: ASoC: error at snd_soc_dai_startup on ssp2
> [   61.847722]  SSP2-Codec: ASoC: BE open failed -1
> [   61.847754]  Audio Port: ASoC: failed to start some BEs -1
> [   61.847786] intel_sst_acpi 808622A8:00: FW sent error response 0x40006
> [   64.301284] intel_sst_acpi 808622A8:00: FW sent error response 0x90001
> [   64.301545] intel_sst_acpi 808622A8:00: not suspending FW!!, Err: -2
> 
> Dropping the asoc/for-5.11 merge and just cherry-picking Pierre-Louis' changes
> for the dynamic switching makes these go away. So this seems to be caused by
> other changes in asoc/for-5.11.
> 
> So any clues where to start looking for this, or should I just bisect this?

Thanks for reporting this Hans.

The only thing that comes to my mind is Morimoto-san's series which 
modified snd_soc_dai_startup, but that was back in September and should 
be in 5.10-rcX

Will give it a try on my side as well.