Bug 218837

Summary: On Dell XPS 16 9640 (2024) with Intel(R) Core(TM) Ultra 9 185H, Kernel is detected as Crash Kernel
Product: Platform Specific/Hardware Reporter: Carsten Tolkmit (ctolkmit)
Component: x86-64Assignee: Thomas Gleixner (tglx)
Status: RESOLVED CODE_FIX    
Severity: normal CC: bhtbzilla, ctolkmit, tglx
Priority: P3    
Hardware: Intel   
OS: Linux   
Kernel Version: 6.9 Subsystem:
Regression: No Bisected commit-id:
Attachments: dmesg output containing the problem
Full lshw.txt
Patch to address the problem of broken ACPI/MADT

Description Carsten Tolkmit 2024-05-14 13:28:16 UTC
Created attachment 306293 [details]
dmesg output containing the problem

When I start my Dell XPS 16 9640 (2024) with a 6.9.0 mainline kernel, I am missing a CPU core:

 LC_ALL=C lscpu --all --extended
CPU NODE SOCKET CORE ONLINE    MAXMHZ   MINMHZ       MHZ
  0    0      0    0    yes 4800.0000 400.0000 2000.0000
  1    0      0    1    yes 5100.0000 400.0000 3800.0000
  2    0      0    2    yes 5100.0000 400.0000 1837.8459
  3    0      0    2    yes 5100.0000 400.0000  400.0000
  4    0      0    0    yes 4800.0000 400.0000 1989.2250
  5    0      0    3    yes 4800.0000 400.0000 1998.5601
  6    0      0    3    yes 4800.0000 400.0000 1903.5920
  7    0      0    4    yes 4800.0000 400.0000 1992.6300
  8    0      0    4    yes 4800.0000 400.0000 1995.6570
  9    0      0    5    yes 4800.0000 400.0000 2000.0000
 10    0      0    5    yes 4800.0000 400.0000  400.0000
 11    0      0    6    yes 3800.0000 400.0000 1851.2080
 12    0      0    7    yes 3800.0000 400.0000 1160.6420
 13    0      0    8    yes 3800.0000 400.0000 1776.4320
 14    0      0    9    yes 3800.0000 400.0000 1557.6689
 15    0      0   10    yes 3800.0000 400.0000 2000.6169
 16    0      0   11    yes 3800.0000 400.0000 1427.8760
 17    0      0   12    yes 3800.0000 400.0000 1960.1360
 18    0      0   13    yes 3800.0000 400.0000 1628.4310
 19    0      0   14    yes 2500.0000 400.0000  400.0000
 20    0      0   15    yes 2500.0000 400.0000  400.0000

It seems that the system thinks my kernel is a kdump kernel, see dmesg:

[    0.089955] CPU topo: Boot CPU APIC ID not the first enumerated APIC ID: 20 > 10
[    0.089956] CPU topo: Crash kernel detected. Disabling real BSP to prevent machine INIT

I doubt that this is really the case, since I am working quite fine on it right now. Just missing one of my P-Core things.

I have the newest official firmware, from lshw:

incendara
    description: Notebook
    product: XPS 16 9640 (0C62)
    vendor: Dell Inc.
    serial: D0M3C24
    width: 64 bits
    capabilities: smbios-3.6.0 dmi-3.6.0 smp vsyscall32
    configuration: boot=normal chassis=notebook family=XPS sku=0C62 uuid=4c4c4544-0030-4d10-8033-c4c04f433234
  *-core
       description: Motherboard
       product: 09CK4V
       vendor: Dell Inc.
       physical id: 0
       version: A00
       serial: /D0M3C24/CNWSC0044D113M/
     *-firmware
          description: BIOS
          vendor: Dell Inc.
          physical id: 0
          version: 1.3.0
          date: 03/12/2024
          size: 1MiB
          capacity: 64MiB
          capabilities: pci pnp upgrade shadowing cdboot bootselect edd int5printscreen int9keyboard int14serial int17printer acpi usb smartbattery biosbootspecification netboot uefi
     *-cpu
          description: CPU
          product: Intel(R) Core(TM) Ultra 9 185H
          vendor: Intel Corp.
          physical id: 400
          bus info: cpu@0
          version: 6.170.4
          slot: U3E1
          size: 1947MHz
          capacity: 4800MHz
          width: 64 bits
          clock: 100MHz
          capabilities: lm fpu fpu_exception wp vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp x86-64 constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rdseed adx smap clflushopt clwb intel_pt sha_ni xsaveopt xsavec xgetbv1 xsaves split_lock_detect user_shstk avx_vnni dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp hwp_pkg_req hfi vnmi umip pku ospke waitpkg gfni vaes vpclmulqdq tme rdpid bus_lock_detect movdiri movdir64b fsrm md_clear serialize pconfig arch_lbr ibt flush_l1d arch_capabilities intel_ppin cpufreq
          configuration: cores=16 enabledcores=16 microcode=28 threads=22
Comment 1 Carsten Tolkmit 2024-05-14 13:29:06 UTC
Created attachment 306294 [details]
Full lshw.txt

I added my full lshw.txt for further info
Comment 2 Thomas Gleixner 2024-05-15 13:44:53 UTC
On Tue, May 14 2024 at 13:28, bugzilla-daemon@kernel.org wrote:
> It seems that the system thinks my kernel is a kdump kernel, see dmesg:
>
> [    0.089955] CPU topo: Boot CPU APIC ID not the first enumerated APIC ID:
> 20
>> 10
> [    0.089956] CPU topo: Crash kernel detected. Disabling real BSP to prevent
> machine INIT
>
> I doubt that this is really the case, since I am working quite fine on it
> right
> now. Just missing one of my P-Core things.

You can work on it obviously...

Unfortunately the ACPI table does not conform to the specification:

 "To ensure that the boot processor is supported post initialization,
  two guidelines should be followed. The first is that OSPM should
  initialize processors in the order that they appear in the MADT. The
  second is that platform firmware should list the boot processor as the
  first processor entry in the MADT."

 "Failure of OSPM implementations and platform firmware to abide by
  these guidelines can result in both unpredictable and non optimal
  platform operation."

The OS follows this guideline. The firmware not.

Let me think about it some more.

Thanks,

        tglx
Comment 3 Carsten Tolkmit 2024-05-15 14:58:12 UTC
I am happy to provide any additional information or dumps, if required, just let me know... (Still running the same "crash kernel" from yesterday, by the way, so the system itself is pretty stable).
Comment 4 Thomas Gleixner 2024-05-15 16:19:49 UTC
Created attachment 306297 [details]
Patch to address the problem of broken ACPI/MADT

Can you please test the attached patch?
Comment 5 Carsten Tolkmit 2024-05-16 06:34:24 UTC
I just double checked with Windows and I found out that Windows 11 does see all 22 cores.

I will try you patch later, once I figured out how to compile the kernel nowadays. My last kernel compile is at least 10 years ago, I just installed the mainline kernel from the ubuntu mainline ppa. Will take some time to come back with answers.
Comment 6 Carsten Tolkmit 2024-05-16 17:40:21 UTC
Hi Thomas,

finally found the time for the kernel compile (busy day), but I can confirm that your patch solves my problem, e.g. "nproc" shows the expected number "22", and "lstopo" no longer misses half a P-Core with this kernel!

Thanks so far!
Comment 7 T.Bracht 2024-05-28 19:51:15 UTC
I can't contribute much to the solution but I can confirm the error for MeteorLake i7Ultra 155H on Asrock NUCS-BOX 155H.