You are not logged in.
Just a PSA: there's a regression in ACPI CPU driver introduced in 5.15.35, 5.16.11, 5.17 kernels and not fixed in these branches to this day (1 May 2022).
The system would hang shortly after the boot, or the boot may not complete at all. On my VIA C7 platform, it most of the time evinces after network activity.
The issue was introduced by ACPI: processor idle: Allow playing dead in C3 state commit and fixed by not yet merged into current branches ACPI: processor: idle: Avoid falling back to C3 type C-states commit.
Offline
Yes, current Arch32's linux package does not work on my VIA C7. Lickily, linux-lts is still on 5.15.34, which is not affected.
If you're going to build newer kernels in the closest future, make sure to apply the patch.
Offline
I don't get this despite currently running a 5.17 kernel, but then I'm running an intel chip not a via one. I guess it's significantly easier to hit on via chips.
The git log links are rather confusing. The first was to help people with AMD cpus, but breaks noticeably those with VIA cpus? Looking up the ACPI power states on wikipedia says than nothing beyond C0-C1 needs to be there, so getting all the way to C3 sounds not ideal for my system either, I guess.
Architecture: pentium4, Testing repos: Yes, Hardware: EeePC 901+2GB RAM+OS half on the SD card.
Offline
Fixed in 5.15.38, 5.17.6 mainline.
Offline
Nice work, all!
Architecture: pentium4, Testing repos: Yes, Hardware: EeePC 901+2GB RAM+OS half on the SD card.
Offline