linux-kernel - Re: WARNING: CPU: 1 PID: 0 at arch/x86/kernel/cpu/cpuid-deps.c:123 do_clear_cpu_cap+0xdc/0x130 on Intel(R) Atom(TM) CPU N450 system

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAE9TcM+AeG8F0PubxAcmSWd=unBiLwZK1xAvCGjXk0Y-+YBaJg@mail.gmail.com>
Date: Fri, 17 Oct 2025 23:02:38 +0200
From: Laurențiu Păncescu <lpancescu@...il.com>
To: Dave Hansen <dave.hansen@...el.com>
Cc: Salvatore Bonaccorso <carnil@...ian.org>, Thomas Gleixner <tglx@...utronix.de>, 
	Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>, 
	Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org, 
	"H. Peter Anvin" <hpa@...or.com>, Sohil Mehta <sohil.mehta@...el.com>, 
	"Chang S. Bae" <chang.seok.bae@...el.com>, Andi Kleen <ak@...ux.intel.com>, 
	Eric Biggers <ebiggers@...gle.com>, linux-kernel@...r.kernel.org, 1117002@...s.debian.org
Subject: Re: WARNING: CPU: 1 PID: 0 at arch/x86/kernel/cpu/cpuid-deps.c:123
 do_clear_cpu_cap+0xdc/0x130 on Intel(R) Atom(TM) CPU N450 system

Hello everyone,

many thanks for the quick replies!

The netbook is from 2009 and has the latest BIOS update offered by Asus,
version 1202 from 2010. I used the BIOS defaults (it doesn't even have
fail-safe and optimized defaults, just "defaults"), and there are basically
no settings you can change: enable/disable onboard audio, LAN and LAN boot,
set the HDD in IDE or AHCI mode, and a sole CPU option "Max CPUID Limit"
("disabled for WIndows XP", but it makes no difference for the warning if I
enable or disable it).

I have been running Debian Stable on the netbook ever since I bought it,
and it worked without any errors or warnings in all versions before Debian
13 with kernel 6.12.43 (I get the same in Debian Testing with kernel
6.16.9). I attached the dmesg output produced by Debian 12 Bookworm, kernel
6.1.0, which works without warnings, and cpuinfo. There's a single CPU, an
Intel Atom N450, seen as two CPUs due to hyperthreading - but I can't
disable hyperthreading in BIOS, either.

"rdmsr -a 0x000001a0" produces the following output:
364972489
364972488

Kind regards,
Laurentiu



On Fri, Oct 17, 2025 at 4:26 PM Dave Hansen <dave.hansen@...el.com> wrote:

> On 10/17/25 07:05, Salvatore Bonaccorso wrote:
> > [    0.008282] RAX: 0000000000000000 RBX: ffffcd5fc00bfd18 RCX:
> 0000000000000000
> > [    0.008282] RDX: 0000000000000001 RSI: 0000000000000070 RDI:
> ffffcd5fc00bfd78
> > [    0.008282] RBP: 0000000000000070 R08: 0000000000000000 R09:
> 706f20676e697274
> > [    0.008282] R10: 7473207473616620 R11: 64656c6261736944 R12:
> ffffffffb8f20ba0
> > [    0.008282] R13: ffff8ad332718090 R14: 000000000000061c R15:
> 0000000000000000
> > [    0.008282] FS:  0000000000000000(0000) GS:ffff8ad378ec7000(0000)
> knlGS:0000000000000000
> > [    0.008282] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > [    0.008282] CR2: 0000000000000000 CR3: 000000005842c000 CR4:
> 00000000000006f0
> > [    0.008282] Call Trace:
> > [    0.008282]  <TASK>
> > [    0.008282]  do_clear_cpu_cap+0x106/0x130
> > [    0.008282]  early_init_intel.cold+0x4d/0x11a
> > [    0.008282]  init_intel+0x2a/0x2c0
> > [    0.008282]  identify_cpu+0x18c/0x750
> > [    0.008282]  identify_secondary_cpu+0x50/0xa0
> > [    0.008282]  start_secondary+0x7c/0x160
> > [    0.008282]  common_startup_64+0x13e/0x141
> > [    0.008282]  </TASK>
> > [    0.008282] ---[ end trace 0000000000000000 ]---
> > [    0.354753] smp: Brought up 1 node, 2 CPUs
> > [    0.354797] smpboot: Total of 2 processors activated (6666.06
> BogoMIPS)
> >
> > Any ideas?
>
> That's a fun one.
>
> This looks to be this code:
>
>                         pr_info("Disabled fast string operations\n");
>                         setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);
>
> because we can see the pr_info() and I'm assuming that the 'feature' is
> X86_FEATURE_REP_GOOD (0x70) which we can see in RSI/RBP.
>
> But X86_FEATURE_REP_GOOD isn't even a real CPU feature, it's one of the
> synthetic ones. The only way I can see this happening is if
> MSR_IA32_MISC_ENABLE_FAST_STRING is mismatched between the boot CPU and
> a secondary.
>
> Could you boot the system on a known good kernel and run this, please?
>
>         rdmsr -a 0x000001a0
>
> That'll dump out MSR_IA32_MISC_ENABLE on all the CPUs.
>
> When was the last kernel that worked for you? Also, do you have old
> microcode? Can you dump the beginning of /proc/cpuinfo, please?
>
> processor       : 21
> vendor_id       : GenuineIntel
> cpu family      : 6
> model           : 170
> model name      : Intel(R) Core(TM) Ultra 7 155H
> stepping        : 4
> microcode       : 0x24
>
> I can see your microcode version in dmesg, but not the cpu
> model/family/stepping.
>
>
>

Content of type "text/html" skipped

View attachment "cpuinfo.txt" of type "text/plain" (1487 bytes)

View attachment "dmesg_bookworm.txt" of type "text/plain" (58892 bytes)