[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170309081736.7jbbbcjwehjnhjwg@pd.tnic>
Date: Thu, 9 Mar 2017 09:17:36 +0100
From: Borislav Petkov <bp@...en8.de>
To: Fengguang Wu <fengguang.wu@...el.com>
Cc: Ye Xiaolong <xiaolong.ye@...el.com>, X86 ML <x86@...nel.org>,
Andy Lutomirski <luto@...capital.net>,
Peter Zijlstra <peterz@...radead.org>,
LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: Re: [lkp-robot] [x86] ed3ce2a917: BUG:unable_to_handle_kernel
On Thu, Mar 09, 2017 at 10:30:52AM +0800, Fengguang Wu wrote:
> One possible improvement is to provide "lkp qemu" reproduce steps for
> kernel oops -- it would be way more convenient and safe to follow than
> "lkp run", since the later risks hang the physical machine.
Right, that would mean that the bug should be triggerable in qemu. Some
bugs happen on baremetal only though.
> As for the test description, the dmesg carries markers for the user
> space test start/stop points, so the robot can easily tell whether the
> oops happen during the test or before/after the test -- the latter may
> well (but not always) indicate the oops is not relevant to the testcase,
> but to the regular kernel boot/reboot/kexec process.
Right, if that is made part of the report, it would make the reports
better. Something like: "this happens when preparing our test env" or
"this happens while running this test blabla".
Thanks guys and keep up the good work!
--
Regards/Gruss,
Boris.
Good mailing practices for 400: avoid top-posting and trim the reply.
Powered by blists - more mailing lists