[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <TYCPR01MB69489CB7170CFB42F0CBC70383B49@TYCPR01MB6948.jpnprd01.prod.outlook.com>
Date: Fri, 24 Jun 2022 10:14:58 +0000
From: NOMURA JUNICHI(野村 淳一)
<junichi.nomura@....com>
To: Borislav Petkov <bp@...en8.de>
CC: "brijesh.singh@....com" <brijesh.singh@....com>,
LKML <linux-kernel@...r.kernel.org>,
"michael.roth@....com" <michael.roth@....com>,
"x86@...nel.org" <x86@...nel.org>
Subject: RE: [Regression v5.19-rc1] crash kexec fails to boot the 2nd kernel
(Re: [PATCH v12 38/46] x86/sev: Add SEV-SNP feature detection/setup)
From: Borislav Petkov <bp@...en8.de>
> What system is this? Full dmesg and .config pls.
>
> How exactly are you launching the kexec kernel, etc, etc?
dmesg, kernel config and a script to load kexec is attached.
I could reproduce the problem as followings:
# ./kexec-p-noinitrd.sh
# cat /sys/kernel/kexec_crash_loaded
1
# echo c > /proc/sysrq-trigger
If the problem reproduces, the system stops or resets just after
showing a call trace of sysrq-panic on console.
On kernels without problem, it should show boot messages from
the 2nd kernel after the sysrq-panic call trace.
When I use "--noefi" option of kexec, the problem didn't occur.
So I think you need EFI-based system to reproduce the problem.
--
Jun'ichi Nomura, NEC Corporation / NEC Solution Innovators, Ltd.
Download attachment "repro.tgz" of type "application/x-compressed" (81948 bytes)
Download attachment "smime.p7s" of type "application/pkcs7-signature" (5766 bytes)
Powered by blists - more mailing lists