linux-kernel - RE: [PATCH] drivers: hv: log when enabling crash_kexec_post

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: <MWHPR21MB1593EAD0167AA9EC4139042AD7369@MWHPR21MB1593.namprd21.prod.outlook.com>
Date:   Thu, 17 Feb 2022 16:44:13 +0000
From:   "Michael Kelley (LINUX)" <mikelley@...rosoft.com>
To:     Stephen Brennan <stephen.s.brennan@...cle.com>,
        KY Srinivasan <kys@...rosoft.com>,
        Haiyang Zhang <haiyangz@...rosoft.com>,
        Stephen Hemminger <sthemmin@...rosoft.com>,
        Wei Liu <wei.liu@...nel.org>, Dexuan Cui <decui@...rosoft.com>
CC:     "linux-hyperv@...r.kernel.org" <linux-hyperv@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH] drivers: hv: log when enabling crash_kexec_post_notifiers

From: Stephen Brennan <stephen.s.brennan@...cle.com> Sent: Monday, February 14, 2022 5:38 PM
> 
> Recently I went down a rabbit hole looking at a race condition in
> panic() on a Hyper-V guest. I assumed, since it was missing from the
> command line, that crash_kexec_post_notifiers was disabled. Only after
> a rather long reproduction and analysis process did I learn that Hyper-V
> actually enables this setting unconditionally.
> 
> Users and debuggers alike would like to know when these things happen. I
> think it would be good to print a message to the kernel log when this
> happens, so that a grep for "crash_kexec_post_notifiers" shows relevant
> results.

I'm OK with adding this output line.  However, you have probably
seen the two other LKML threads [1] and [2] about reorganizing the
panic notifiers to clearly distinguish between notifiers that always run
vs. those controlled by "crash_kexec_post_notifiers".  If the changes
proposed in those threads are submitted and accepted, it is likely that
the kernel log message in this patch would become unnecessary.
But since we don't know when those proposed changes might come
to fruition, adding the message for now makes sense.

Reviewed-by: Michael Kelley <mikelley@...rosoft.com>

[1] https://lore.kernel.org/lkml/20220108153451.195121-1-gpiccoli@igalia.com/
[2] https://lore.kernel.org/lkml/20220114183046.428796-1-gpiccoli@igalia.com/

> 
> Signed-off-by: Stephen Brennan <stephen.s.brennan@...cle.com>
> ---
>  drivers/hv/hv_common.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> index 181d16bbf49d..c1dd21d0d7ef 100644
> --- a/drivers/hv/hv_common.c
> +++ b/drivers/hv/hv_common.c
> @@ -79,8 +79,10 @@ int __init hv_common_init(void)
>  	 * calling crash enlightment interface before running kdump
>  	 * kernel.
>  	 */
> -	if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE)
> +	if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) {
>  		crash_kexec_post_notifiers = true;
> +		pr_info("Hyper-V: enabling crash_kexec_post_notifiers\n");
> +	}
> 
>  	/*
>  	 * Allocate the per-CPU state for the hypercall input arg.
> --
> 2.30.2