"Firmware Bug (...) APIC ID mismatch" in dmesg

[dobrowolski-lukasz]

Component

Dasharo firmware

Device

NovaCustom NV4x 12th Gen

Dasharo version

v1.8.0

Dasharo Tools Suite version

No response

Test case ID

No response

Brief summary

I get "Firmware Bug" prints in dmesg when running Linux on Xen.

How reproducible

No response

How to reproduce

Install Debian 13 with kernel form backports and xen.
xen: 4.20
linux: 7.0.10+deb13
Boot the system and run dmesg.

I get similar errors on QubesOS 4.3.

Expected behavior

No errors/warnings in dmesg.

Actual behavior

Following show in dmesg:

(...)
[    2.517933] NMI watchdog: Perf NMI watchdog permanently disabled
[    2.518126] smp: Bringing up secondary CPUs ...
[    2.518284] installing Xen timer for CPU 2
[    2.518731] installing Xen timer for CPU 4
[    2.519229] installing Xen timer for CPU 6
[    2.519734] installing Xen timer for CPU 8
[    2.520224] installing Xen timer for CPU 9
[    2.520491] installing Xen timer for CPU 10
[    2.520769] installing Xen timer for CPU 11
[    2.524194] installing Xen timer for CPU 12
[    2.524482] installing Xen timer for CPU 13
[    2.524728] installing Xen timer for CPU 14
[    2.527028] installing Xen timer for CPU 15
[    0.008155] [Firmware Bug]: CPU   2: APIC ID mismatch. CPUID: 0x0002 APIC: 0x0008
[    2.528085] cpu 2 spinlock event irq 181
[    0.008155] [Firmware Bug]: CPU   4: APIC ID mismatch. CPUID: 0x0004 APIC: 0x0010
[    2.528103] cpu 4 spinlock event irq 182
[    0.008155] [Firmware Bug]: CPU   6: APIC ID mismatch. CPUID: 0x0006 APIC: 0x0018
[    2.528103] cpu 6 spinlock event irq 183
[    0.008155] [Firmware Bug]: CPU   8: APIC ID mismatch. CPUID: 0x0008 APIC: 0x0020
[    2.528103] cpu 8 spinlock event irq 184
[    0.008155] [Firmware Bug]: CPU   9: APIC ID mismatch. CPUID: 0x0009 APIC: 0x0022
[    2.528104] cpu 9 spinlock event irq 185
[    0.008155] [Firmware Bug]: CPU  10: APIC ID mismatch. CPUID: 0x000a APIC: 0x0024
[    2.528104] cpu 10 spinlock event irq 186
[    0.008155] [Firmware Bug]: CPU  11: APIC ID mismatch. CPUID: 0x000b APIC: 0x0026
[    2.528104] cpu 11 spinlock event irq 187
[    0.008155] [Firmware Bug]: CPU  12: APIC ID mismatch. CPUID: 0x000c APIC: 0x0028
[    2.528104] cpu 12 spinlock event irq 188
[    0.008155] [Firmware Bug]: CPU  13: APIC ID mismatch. CPUID: 0x000d APIC: 0x002a
[    2.528104] cpu 13 spinlock event irq 189
[    0.008155] [Firmware Bug]: CPU  14: APIC ID mismatch. CPUID: 0x000e APIC: 0x002c
[    2.532062] cpu 14 spinlock event irq 190
[    0.008155] [Firmware Bug]: CPU  15: APIC ID mismatch. CPUID: 0x000f APIC: 0x002e
[    2.532072] cpu 15 spinlock event irq 191
[    2.532184] installing Xen timer for CPU 1
[    2.532611] installing Xen timer for CPU 3
[    2.532895] installing Xen timer for CPU 5
[    2.533156] installing Xen timer for CPU 7
[    2.533234] cpu 1 spinlock event irq 212
[    0.008155] [Firmware Bug]: CPU   3: APIC ID mismatch. CPUID: 0x0003 APIC: 0x0009
[    2.533234] cpu 3 spinlock event irq 213
[    0.008155] [Firmware Bug]: CPU   5: APIC ID mismatch. CPUID: 0x0005 APIC: 0x0011
[    2.533234] cpu 5 spinlock event irq 214
[    0.008155] [Firmware Bug]: CPU   7: APIC ID mismatch. CPUID: 0x0007 APIC: 0x0019
[    2.533234] cpu 7 spinlock event irq 215
[    2.533234] smp: Brought up 1 node, 16 CPUs
(...)

Screenshots

No response

Additional context

I think You should add some tests for new dmesg/lx dmesg/... for all new release candidates of Dasharo.

Solutions you've tried

No response

[filipleple]

EDIT: comment applies to NUC BOX

This has come up in:

• STB002.001 encounters unlisted error #1013 (comment)
• Unlisted error in dmesg | STB002.002 | QubesOS | Heads & UEFI #1201 (comment)

It's listed as a known issue since v0.9.1, not fixed so far since it doesn't seem to impact actual functionality. Seems to be an issue with MTL-wide core enumeration code. - seems to be an even broader issue, since it appears on other SoCs as well.

[dobrowolski-lukasz]

Just checked and It's not on known issues list for Dasharo 1.8.0 for nv41. Also nv41 is ADL not MTL. I don't remember those prints being present on Dasharo 1.7.2 for nv41.

[filipleple]

Oh sorry, I'd clicked into this issue from a NUC BOX context, and blindly assumed it's also reporting on the NUC. Thank you for the input, I think this hasn't been spotted on ADL yet. So that invalidates narrowing down to MTL, it must be a regression after introducing/rebasing onto a wider-scoped change.