Received: from parabel.levigo.net ([62.206.214.16] helo=parabel.matrix.de) by bombadil.infradead.org with esmtp (Exim 4.68 #1 (Red Hat Linux)) id 1JhPos-0000Bd-TA for linux-mtd@lists.infradead.org; Thu, 03 Apr 2008 13:49:35 +0000 Received: from localhost (finite.levigo.net [10.208.4.86]) by parabel.matrix.de (Postfix) with ESMTP id 0E63D342A6 for ; Thu, 3 Apr 2008 15:49:22 +0200 (CEST) Received: from parabel.matrix.de ([10.208.4.80]) by localhost (finite [10.208.4.86]) (amavisd-new, port 10024) with ESMTP id 30054-10 for ; Thu, 3 Apr 2008 15:49:10 +0200 (CEST) Received: from mailrelay.gdsys.de (unknown [217.6.197.20]) by parabel.matrix.de (Postfix) with ESMTP id F039434500 for ; Thu, 3 Apr 2008 15:49:08 +0200 (CEST) From: eibach@gdsys.de To: linux-mtd@lists.infradead.org Date: 03 Apr 2008 13:49:04 UT Message-ID: MIME-Version: 1.0 Subject: [SOLVED] jffs2: filesystem corruption caused by writing and powercycle Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7Bit List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , > I have a problem with jffs2 filesystem corruption. > My environment is kernel 2.6.23.4, running on PPC. > Flash is Spansion 29GL512 NOR-Flash. > > The problem can be reproduced simply by turning off power while writing. > After 3 or 4 powercycles I get errormessages when trying to copy the file I wrote before turning off power: > > / # cp /data/tst /data/tst2 > Data CRC ee9173b1 != calculated CRC 6af6806d for node at 0221f108 > cp: read error: Input/output error > / # The problem is solved by patch "[JFFS2] Fix data CRC checking on NOR flash." from David Woodhouse, 21 Nov 2007. Cheers Dirk