[BUG] Suspend/resume test failed on CML and TGL chrome devices

[keqiaozhang]

Describe the bug
Recently, our CI detected the suspend/resume issue on 2 Chrome devices, One plaform is TGLU_VOLT_SDW and another one is CML_HEL_RT5682. There's no much logs available during the test, system failed to resume from suspending in the first round of testing.
From the history of our CI testing, we determined this is a kernel regression.
Bisect result shows that this issue is caused by an acpi change from upstream rebase.
The bad commit is:
e38f9ff

To Reproduce
$ rtcwake -m mem -s 5

Reproduction Rate
100%

Expected behavior

Environment

1. kernel: tip sof-dev/d2937505b7b5
2. SOF: tip main/d792b0602dcf
3. Platform: TGL Chromebook Volteer and CML Helios

The DSDT table is attached:
dsdt-dsl.txt

[marc-hb]

@plbossart can we revert e38f9ff while we investigate this issue? This will allow us to catch other, unrelated suspend/resume issues, something we can't do right now on this platform.

[plbossart]

Yes, we can revert, but I was waiting to see an upstream commit for the revert. I'll add this manually and we can revert-reapply later.

[keqiaozhang]

#3525 is merged, closing this issue.