[EN] CPUs APIC ID mismatch seems to relate to a firmware bug

[humon]

Hi folks,
I set up the following OS: siduction 2023.1.1 giants - xfce on an Acer Aspire E5-523G notebook type laptop over the past month.

I just updated the old 2016 firmware in an attempt to address the following message:
[Firmware Bug]: CPU   0: APIC ID mismatch. CPUID 0x0010
[Firmware Bug]: CPU   1: APIC ID mismatch. CPUID 0x0011
The most up-to-date firmware available from Acer was from 2017.

BIOS is Insyde ver 1.18 now.  I boot UEFI and the partition Giants resides in is GPT formated ext4 (although grub rescue> shows it as ext2 for some reason).

The following is what I see when the machine boots up on the Giants side (it is dual boot with Win 11):
[    0.114575] [Firmware Bug]: CPU   0: APIC ID mismatch. CPUID 0x0010
[    0.001342] [Firmware Bug]: CPU   1: APIC ID mismatch. CPUID 0x0011
[    0.093907] piix4_smbus 0000:00:14.0SMBus region 0xb20 already in use!
[    0.110410] i2c_hid_acpi i2c_ELAN0501:00: failed to fetch HID descriptor: -121
[    0.110442] i2c_hid_acpi i2c_ELAN0501:00: failed to fetch the HID descriptor
/dev/sda5: clean, 307404/3276800 files. 5224927/13107200 blocks

Welcome to siduction Giants XFCE!

And then the usual rolling display as packages are loading.

My questions relate to the mismatch situation.  I have had several glitchy starts on the laptop, but I haven't been able to pinpoint any one source (although tumblerd often seems to be involved in the system hanging).

I only saw one post on the subject of CPU APIC ID mismatches and didn't gain much info from it.

Ideas?

[dibl]

No direct knowledge, but this google hit suggests to me that it's about the touchpad:

https://forum.manjaro.org/t/touchpad-not-detected-i2c-hid-acpi-i2c-msft0001-failed-to-fetch-the-hid-descriptor/119311

There are some troubleshooting suggestions down that page.

[humon]

Thanks for the great answer.  It turns out that information did solve one part of the problem.

The touchpad was misconfigured.   I changed the BIOS entry related to the touchpad and, poof, those two ELAN0501 error lines are now gone.

However, it appears the ACPI problem persists.  It looks like the problem could be any combination of OS incorrectly interacting with the hardware, ACPI not well configured at the OEM level,  other things not playing well with ACPI, UEFI having some strange role, or, who knows.

For now I will leave the Win 11 side of things alone, temporarily remove siduction and load some other OS.  I notice in the ACPI wikipedia page the Haiku and FreeBSD are supposed to have coding related to updating the ACPI subsystem.  I've played a bit with Haiku but not with FreeBSD.

The wikipedia page says that "linux-based operating systems can provide handling of ACPI events via acpid", but I notice the word "can" only indicates possibility.

Testing other OSes will take a few days.

[ro_sid]

Be careful with the meaning (and consequences) of the boot messages:
APIC = Advanced Programmable Interrupt Controller
ACPI = Advanced Configuration and Power Interface

Both have common issues, but are not the same.
[Especially one can often ignore ACPI issues (e.g. malformed tables).]

[dibl]

It is also true that kernal messages about the hardware are sometimes only for information, and not always indicating a problem. i had a laptop for years that had "possibly missing firmware" messages. Possibly. But it wasn't.