linux-kernel - Re: WARNING in lmce_supported() during reboot.

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: <e1c50570-1b5b-40d2-bab3-05e9ead51a57@intel.com>
Date: Fri, 25 Oct 2024 16:58:47 -0700
From: Dave Hansen <dave.hansen@...el.com>
To: Kuniyuki Iwashima <kuniyu@...zon.com>, x86@...nel.org,
 linux-edac@...r.kernel.org, linux-kernel@...r.kernel.org
Cc: Tony Luck <tony.luck@...el.com>, Borislav Petkov <bp@...en8.de>,
 Thomas Gleixner <tglx@...utronix.de>, Ingo Molnar <mingo@...hat.com>,
 Dave Hansen <dave.hansen@...ux.intel.com>, "H. Peter Anvin" <hpa@...or.com>,
 Benjamin Herrenschmidt <benh@...zon.com>
Subject: Re: WARNING in lmce_supported() during reboot.

On 10/25/24 16:13, Kuniyuki Iwashima wrote:
> We have seen the splat below few times when just rebooting hosts.
> 
> It rarely happens and seems a timing related, so we don't have a
> reproducer.
> 
> Our kernel source in the splat is here,
> https://github.com/amazonlinux/linux/tree/kernel-6.1.61-85.141.amzn2023

Hi Folks,

We really do need it to be reproduced on mainline.  At the very least,
it would be greatly appreciated if you could summarize what your fork is
doing and why you don't think it is responsible.

But I don't see how this could be timing related.  That MSR gets locked
early from what I can tell, long before the system would be rebooting.

Your best bet is going to be getting a handle on what
MSR_IA32_FEAT_CTL's value was after the CPU was brought up and when this
reboot was attempted.  If those values differ, when it got changed.

I'd _suspect_ some kind of BIOS sleep/wakeup wonkiness where something
forgot to re-lock the MSR.