Date: Tue, 26 Apr 2016 11:10:18 -0700
From: Davidlohr Bueso <dave@stgolabs.net>
To: "Austin S. Hemmelgarn" <ahferroin7@gmail.com>
Cc: Julius Werner <jwerner@chromium.org>, linux-efi@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-block@vger.kernel.org,
        Gwendal Grignou <gwendal@chromium.org>,
        Doug Anderson <dianders@chromium.org>, Karel Zak <kzak@redhat.com>
Subject: Re: [PATCH] block: partitions: efi: Always check for alternative GPT
 at end of drive
Message-ID: <20160426181018.GA16601@linux-uzut.site>
References: <1461632806-5946-1-git-send-email-jwerner@chromium.org>
 <571F7D6D.8020209@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Disposition: inline
In-Reply-To: <571F7D6D.8020209@gmail.com>
User-Agent: Mutt/1.5.24 (2015-08-30)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 665
Lines: 15

On Tue, 26 Apr 2016, Austin S. Hemmelgarn wrote:

>At the absolute minimum, we should be logging (at least at a warning 
>level) that we had to fall back the the backup GPT.  If somebody is 
>dealing with a disk that had a torn write to the primary GPT, that's 
>one thing, but this could also be caused by any number of other 
>problems (hardware issues, malicious intent, etc), and we need to log 
>that we detected corrupted data.

We already complain about corrupted primary gpt (at a warn level), and
there's also plenty of verbosity about differences between primary and
backup (see compare_gpts()), or are you referring to something else?

Thanks,
Davidlohr