linux-kernel - Re: [tip:x86/alternatives] [x86/alternatives] ee8962082a: WARNING:at_arch/x86/kernel/cpu/cpuid-deps.c:#do_clear_cpu

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: <20240508082438.GBZjs2xsn4diYXU4ud@fat_crate.local>
Date: Wed, 8 May 2024 10:24:38 +0200
From: Borislav Petkov <bp@...en8.de>
To: Oliver Sang <oliver.sang@...el.com>,
	Sean Christopherson <seanjc@...gle.com>
Cc: oe-lkp@...ts.linux.dev, lkp@...el.com, linux-kernel@...r.kernel.org,
	x86@...nel.org, Ingo Molnar <mingo@...nel.org>,
	Srikanth Aithal <sraithal@....com>, fengwei.yin@...el.com,
	yujie.liu@...el.com
Subject: Re: [tip:x86/alternatives] [x86/alternatives] ee8962082a:
 WARNING:at_arch/x86/kernel/cpu/cpuid-deps.c:#do_clear_cpu_cap

On Wed, May 08, 2024 at 04:08:18PM +0800, Oliver Sang wrote:
> we really don't do any manual change recently, however, per your reminding,
> we rerun tests on both ee8962082a and its parent v6.9-rc3 again.
> 
> while we made original report, we always saw below for both commits.
> "x86/cpu: VMX (outside TXT) disabled by BIOS"
> 
> by rerun today, we cannot see it again, on both commits.

Yeah, from the last dump you shouldn't be seeing it.

> then for ee8962082a, the reported
>   WARNING:at_arch/x86/kernel/cpu/cpuid-deps.c:#do_clear_cpu_cap
> disappeared also while no "x86/cpu: VMX (outside TXT) disabled by BIOS"

Yeah, it does feel like something underneath the OS has changed/is
changing for this to happen.

> we are using kexec to boot into kernel, and the kernel images keep same,
> so we are sure this strage phenomenon is not caused by kernel build.

Hmm, kexec won't go through BIOS so I can imagine if the kernel has left
FEAT_CTL in some weird state... but then if it has been locked, nothing
should change it anymore.

> now we doubt it's really a bios issue, we found another similar machine,
> which also show "x86/cpu: VMX (outside TXT) disabled by BIOS", but after
> several rounds of cold reboot, the message disappeared, too.

If you see that again, please run my branch with the debug patch - that
should at least tell us what's weird in FEAT_CTL.

> we will investigate this BIOS further and avoid this kind of misleading
> report in the future.
> 
> sorry again :(

No worries, thanks for testing kernels and thanks Sean too. :-)

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette