linux-kernel - Re: [PATCH] x86/mce: Unify vendors grading logic and provide AMD machine error checks

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]

Date:   Tue, 8 Mar 2022 20:32:51 +0100
From:   Borislav Petkov <bp@...en8.de>
To:     Carlos Bilbao <carlos.bilbao@....com>
Cc:     tglx@...utronix.de, mingo@...hat.com, dave.hansen@...ux.intel.com,
        x86@...nel.org, yazen.ghannam@....com,
        linux-kernel@...r.kernel.org, linux-edac@...r.kernel.org,
        bilbao@...edu
Subject: Re: [PATCH] x86/mce: Unify vendors grading logic and provide AMD
 machine error checks

On Tue, Mar 08, 2022 at 12:41:34PM -0600, Carlos Bilbao wrote:
> AMD's severity grading covers very few machine errors. In the graded cases
> there are no user-readable messages, complicating debugging of critical
> hardware errors. Furthermore, with the current implementation AMD MCEs have
> no support for the severities-coverage file. Adding new severities for AMD
> with the current logic would be too convoluted.
> 
> Fix the above issues including AMD severities to the severity table, in
> combination with Intel MCEs. Unify the severity grading logic of both
> vendors. Label the vendor-specific cases (e.g. cases with different
> registers) where checks cannot be implicit with the available features.
> 
> Signed-off-by: Carlos Bilbao <carlos.bilbao@....com>
> ---
>  arch/x86/include/asm/mce.h         |   7 ++
>  arch/x86/kernel/cpu/mce/severity.c | 188 +++++++++++++++--------------
>  2 files changed, 103 insertions(+), 92 deletions(-)

Sorry, maybe you're too new to this and you probably haven't read the
old discussions we have had about the severity grading turd. In order to
save you some time: adding more to that macro insanity is not going to
happen.

The AMD severity grading functions are *actually* readable vs this
abomination which I hate with passion.

If you want to add more logic, you should add to mce_severity_amd(),
perhaps call other helper functions which grade based on a certain
aspect of the error type, split the logic, use comments, etc, but
*definitely* not this.

Thx.

-- 
Regards/Gruss,
    Boris.

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