Freeze on power state change, acpi callback error

System Details

  • OS: Artix Linux

  • ** Device: Lenovo Ideapad 3 Chromebook
    ** HWID: LICK
    ** Platform: Intel GeminiLake
    ** Fw Type: Stock ChromeOS w/RW_LEGACY
    ** Fw Ver: Google_Lick.11297.296.0 (09/09/2021)
    ** Fw WP: Enabled


  • Internal storage type: MicroSD

Summary of the Issue

When the power cord is plugged in or unplugged, the system freezes. When this is done while still booting or from getty, the tail end of a stack trace is sometimes printed. It’s very cryptic, but suggests that an acpi callback was triggered and never completed.

Other times, the system has a rare chance of remaining half operational. Some parts frozen and others not. A likely related alternate message when viewing a text mode display is that one cpu core has frozen.

This may correlate with a MrChromeBox update I did around May. I’ve been coping with the issue for a while. There were no similar issues when I first rooted my chromebook. I’ve reinstalled my OS and the problem persists. I’ve tested it in FreeBSD, and the problem is not present. As well my own low level OS project where it is also not present (this os is doing 0.001% what a real OS is doing, but the point is the CPU doesn’t lock when power changes).

I have played with quite a few kernel boot flags with no success.

I am curious if anyone else has seen a similar bug. A friend of mine reported a similar bug in Fedora, not on a Chromebook, so maybe the issue is purely with Linux. I just have a hard time believing if it was related to a Linux update that hit multiple distros that it wouldn’t be patched by now, and the correlation with a MrChromeBox update I ran seems to suspicious. Ultimately I think my friend is experiencing a parallel but not related bug. But it is still interesting. Surely if this existed across all of Linux some people here would have experienced it regardless of our Chromebook use.

easy test is to boot a live USB from another distro (eg, ubuntu) and attempt to reproduce there.

RW_LEGACY firmware is just a bootloader, nothing else. You’re running stock Google firmware, so if there is a firmware bug/incompatibility with mainline Linux, it’s something you;ll have to work around on the OS side.

1 Like

Thanks. That is very helpful to know. Is there a chance the full rom setup could have better compatibility?

I might keep this thread open and dump more info if there is an interest in the community with hacking out Google firmware + Ideapad 3 + Artix stability.

it depends on whether it’s a firmware or an OS issue, which is why I recommended testing another distro via Live USB