linux-kernel - Re: [PATCH] x86,mm: print likely CPU at segfault time

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 for Android: free password hash cracker in your pocket

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

Message-ID: <c8e023cb-6f50-36f5-65d4-c5e25b264029@intel.com>
Date:   Mon, 19 Jul 2021 12:20:24 -0700
From:   Dave Hansen <dave.hansen@...el.com>
To:     Rik van Riel <riel@...riel.com>, linux-kernel@...r.kernel.org
Cc:     Dave Hansen <dave.hansen@...ux.intel.com>,
        Andy Lutomirski <luto@...nel.org>, kernel-team@...com,
        Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
        x86@...nel.org
Subject: Re: [PATCH] x86,mm: print likely CPU at segfault time

On 7/19/21 12:00 PM, Rik van Riel wrote:
> In a large enough fleet of computers, it is common to have a few bad
> CPUs. Those can often be identified by seeing that some commonly run
> kernel code (that runs fine everywhere else) keeps crashing on the
> same CPU core on a particular bad system.

I've encountered a few of these kinds of things over the years.  This is
*definitely* useful.  What you've proposed here is surely the simplest
thing we could print and probably also offers the best bang for our buck.

The only other thing I thought of is that it might be nice to print out
the core id instead of the CPU id.  If there are hardware issues with a
CPU, they're likely to affect both threads.  Seeing to different "CPUs"
in an SMT environment might tempt some folks to think it's not a
core-level hardware issue.

If it's as trivial as:

	printk(KERN_CONT " on cpu/core %d/%d",
		raw_smp_processor_id(),
		topology_core_id(raw_smp_processor_id()));

it would be handy.  But, it's also not hard to look at 10 segfaults, see
that they happened only on 2 CPUs and realize that hyperthreading is
enabled.

Either way, this patch moves things in the right direction, so:

Acked-by: Dave Hansen <dave.hansen@...ux.intel.com>