[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200618154444.GB18007@redhat.com>
Date: Thu, 18 Jun 2020 11:44:45 -0400
From: Mike Snitzer <snitzer@...hat.com>
To: JeongHyeon Lee <jhs2.lee@...sung.com>
Cc: agk@...hat.com, dm-devel@...hat.com, corbet@....net,
linux-doc@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: New mode DM-Verity error handling
On Thu, Jun 18 2020 at 2:56am -0400,
JeongHyeon Lee <jhs2.lee@...sung.com> wrote:
> Hello, Dear devcice-mapper maintainers.
>
> I'm JeongHyeon Lee, work in Samsung. I'm chage of DM-Verity feature with
> Mr. sunwook eom.
> I have a patch or suggestion about DM-Verity error handling.
>
> Our device (smart phone) need DM-Verity feature. So I hope there is new
> mode DM-Verity error handling.
> This new mode concept is When detect corrupted block, will be go to panic.
>
> Because our team policy is found device DM-Verity error, device will go
> panic.
> And then analyze what kind of device fault (crash UFS, IO error, DRAM
> bit flip etc)
>
> In addition to the smart phone, I would like to have an option that
> users or administrators can use accordingly.
> There are patch contents in the attachment. I would really appreciate it
> if you could check it.
>
> I will look forward to hearing from yours.
> Thank you :)
>
I do not accept that panicing the system because of verity failure is
reasonable.
In fact, even rebooting (via DM_VERITY_MODE_RESTART) looks very wrong.
The device should be put in a failed state and left for admin recovery.
Mike
Powered by blists - more mailing lists