From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mail-out.m-online.net ([212.18.0.10])
	by canuck.infradead.org with esmtp (Exim 4.72 #1 (Red Hat Linux))
	id 1PNj8e-0002gO-Av
	for linux-mtd@lists.infradead.org; Wed, 01 Dec 2010 09:38:13 +0000
Date: Wed, 1 Dec 2010 10:38:07 +0100
From: Anatolij Gustschin <agust@denx.de>
To: Anatolij Gustschin <agust@denx.de>, linux-mtd@lists.infradead.org
Subject: Re: UBIFS partition on NOR flash not mountable after power cut test
Message-ID: <20101201103807.1e432cfb@wker>
In-Reply-To: <20101130163541.44c48a94@wker>
References: <20101129195014.19224240@wker> <4CF4D4CA.8040906@aimvalley.nl>
	<20101130163541.44c48a94@wker>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Cc: Detlev Zundel <dzu@denx.de>, Norbert van Bolhuis <nvbolhuis@aimvalley.nl>
List-Id: Linux MTD discussion mailing list <linux-mtd.lists.infradead.org>
List-Unsubscribe: <http://lists.infradead.org/mailman/options/linux-mtd>,
	<mailto:linux-mtd-request@lists.infradead.org?subject=unsubscribe>
List-Archive: <http://lists.infradead.org/pipermail/linux-mtd/>
List-Post: <mailto:linux-mtd@lists.infradead.org>
List-Help: <mailto:linux-mtd-request@lists.infradead.org?subject=help>
List-Subscribe: <http://lists.infradead.org/mailman/listinfo/linux-mtd>,
	<mailto:linux-mtd-request@lists.infradead.org?subject=subscribe>

On Tue, 30 Nov 2010 16:35:41 +0100
Anatolij Gustschin <agust@denx.de> wrote:
...
> On Tue, 30 Nov 2010 11:41:14 +0100
> Norbert van Bolhuis <nvbolhuis@aimvalley.nl> wrote:
> 
> > Did you read the mail discussions
> > "UBIFS Corrupt during power failure"
> > dated april 2009 from Eric Holmberg.
> > 
> > He did some great work in making UBIFS on NOR flash
> > more reliable for power-cuts.
> 
> I have partially read the thread "UBIFS Corrupt during power 
> failure", especially the discussion from March, May, June and July
> 2009 but not the whole discussion from April.
> 
> > One of the problems was UBI/UBIFS assumes a (CFI) write
> > buffer size of 8 bytes. Maybe this is still the case.
> > I think for your NOR flash it is 64 bytes though.
> 
> Thanks for the hint!
> 
> Yes, I remember Eric saying that he used 8 byte write buffers but
> still had some failures. Therefore I didn't try to reduce the write
> buffer size up to now. But these problems seem to be fixed meanwhile.
> 
> The write buffer size used in the CFI driver is indeed 64 bytes. Now
> I'm running the tests with write buffer size reduced to 8 bytes and
> it looks much better now. After 295 test cycles I haven't seen the
> issue so far.

Further testing with a CFI 8 byte write buffer shows that there are
other issues. Now we end up with the corrupted VID header and cannot
attach the MTD device. I'm going to send the debug log to the list.

Thanks,
Anatolij

--
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-0 Fax: +49-8142-66989-80  Email: office@denx.de