linux-6.7rc5 breaks sof-rt5682

[135e2]

Describe the bug
Somewhere between 6.7rc4 and 6.7rc5 breaks the chromebook audio. Upon inspecting the kernel log, I've identified the only differences (omitting identical lines):

// 6.7rc4, working
[   61.572587] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx error for 0x60010000 (msg/reply size: 108/20): -22
[   61.572593] sof-audio-pci-intel-tgl 0000:00:1f.3: HW params ipc failed for stream 1
[   61.572594] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_hw_params on 0000:00:1f.3: -22
[   61.573247] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx error for 0x60010000 (msg/reply size: 108/20): -22
[   61.573250] sof-audio-pci-intel-tgl 0000:00:1f.3: HW params ipc failed for stream 1
[   61.573251] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_hw_params on 0000:00:1f.3: -22
[   61.573888] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx error for 0x60010000 (msg/reply size: 108/20): -22
[   61.573892] sof-audio-pci-intel-tgl 0000:00:1f.3: HW params ipc failed for stream 1
[   61.573893] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_hw_params on 0000:00:1f.3: -22
[   61.574570] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx error for 0x60010000 (msg/reply size: 108/20): -22
[   61.574572] sof-audio-pci-intel-tgl 0000:00:1f.3: HW params ipc failed for stream 1
[   61.574573] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_hw_params on 0000:00:1f.3: -22

// 6.7rc6, broken
[   48.930740] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx error for 0x30020000 (msg/reply size: 12/0): -13
[   48.930762] sof-audio-pci-intel-tgl 0000:00:1f.3: failed to free widget DMIC0.IN
[   57.235697] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx error for 0x30030000 (msg/reply size: 16/0): -22
[   57.235701] sof-audio-pci-intel-tgl 0000:00:1f.3: sof_ipc3_route_setup: route DMIC0.IN -> BUF4.0 failed
[   57.235703] sof-audio-pci-intel-tgl 0000:00:1f.3: sof_ipc3_set_up_all_pipelines: route set up failed
[   57.235704] sof-audio-pci-intel-tgl 0000:00:1f.3: Failed to restore pipeline after resume -22
[   57.235706] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.235926] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.235966] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236006] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236041] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236074] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236107] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236141] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236173] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236205] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22
[   57.236239] sof-audio-pci-intel-tgl 0000:00:1f.3: ASoC: error at snd_soc_pcm_component_pm_runtime_get on 0000:00:1f.3: -22

Looking through recent commits in kernel, I found no changes made to intel sof drivers. Might it be a kernel regression, or should I report to the upstream kernel bugzilla? Any guidance or suggestions would be much appreciated.

---

Edit: Now I suspect it's a regression from the sof project. Tried upgrading sof-firmware to v2023.12 (they released it 5 hrs ago xD), rebooted, but got no luck.

OS
Arch Linux

Kernel
linux-mainline built from AUR

Boardname
Voxel

[WeirdTreeThing]

Unfortunate how intel keeps breaking things in the kernel. The only thing that can really be done is a git bisect to find the bad commit.

[ninelore]

Can also confirm on linux-lts 6.6.13-1

[ninelore]

@135e2 Workaround is downgrade to 6.6.12

[135e2]

UPDATE:
2024/02/01
Hours of git bisecting led me to the first bad commit causing the regression:
torvalds/linux@31ed8da
@WeirdTreeThing any ideas or should I report to the linux kernel maintainer mailing list?

---

2024/02/06
@WeirdTreeThing @ninelore
Bug reported to sof: thesofproject/linux#4807

---

2024/02/26
Fix (ASoC: SOF: ipc3-topology: Fix pipeline tear down logic) arrives in 6.7.6 & 6.8-rc5. Closing the issue.

[WeirdTreeThing]

Thanks for finding that! Yeah maybe report it to an alsa mailing list or just the regular linux mailing list.

[ninelore]

Fixed in 6.6.18 and 6.7.6

Please close

[BinBashBanana]

swag

[WeirdTreeThing]

swag indeed