linux-kernel - Re: [PATCH] x86/tsc: Add debugfs entry to mark TSC as unstable after boot

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: <b43e2353-41ff-f2de-881c-c9a3348552b7@igalia.com>
Date: Fri, 21 Mar 2025 16:26:02 -0300
From: "Guilherme G. Piccoli" <gpiccoli@...lia.com>
To: "H. Peter Anvin" <hpa@...or.com>, bp@...en8.de
Cc: tglx@...utronix.de, x86@...nel.org, linux-kernel@...r.kernel.org,
 mingo@...hat.com, dave.hansen@...ux.intel.com, kernel@...ccoli.net,
 kernel-dev@...lia.com
Subject: Re: [PATCH] x86/tsc: Add debugfs entry to mark TSC as unstable after
 boot

On 17/03/2025 15:42, H. Peter Anvin wrote:
> On March 17, 2025 7:35:45 AM PDT, "Guilherme G. Piccoli" <gpiccoli@...lia.com> wrote:
>> On 26/02/2025 10:27, Guilherme G. Piccoli wrote:
>>> Right now, we can force the TSC to be marked as unstable through
>>> boot parameter. There are debug / test cases though in which would
>>> be preferable to simulate the clocksource watchdog behavior, i.e.,
>>> marking TSC as unstable during the system run. Some paths might
>>> change, for example: the tracing clock is auto switched to global
>>> if TSC is marked as unstable on boot, but it could remain local if
>>> TSC gets marked as unstable after tracing initialization.
>>>
>>> Hence, the proposal here is to have a simple debugfs file that
>>> gets TSC marked as unstable when written.
>>>
>>> Signed-off-by: Guilherme G. Piccoli <gpiccoli@...lia.com>
>> Guilherme
> 
> To be honest I don't think this belongs in debugfs; rather it belongs in sysfs.
> 
> Debugfs should not be necessarily in serious production systems – it is way too large of an attack surface, which is a very good reason why it is its own filesystem – but if this is a real issue on hardware then it may be needed.
> 
>    -hpa

Thanks! I agree, but as per Boris comment, he suggested to drop this
one, right?

In other words, we have 2 options in my understanding:

(a) Drop it;

(b) Re-implement using sysfs entry instead of debugfs;

What do you all think?
Thanks again,

Guilherme