From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from c-24-91-108-216.hsd1.ma.comcast.net ([24.91.108.216] helo=ahgu.homeunix.com) by canuck.infradead.org with esmtps (Exim 4.52 #1 (Red Hat Linux)) id 1EEr4s-0003EE-F5 for linux-mtd@lists.infradead.org; Mon, 12 Sep 2005 12:22:51 -0400 Received: from maaxiang (fwma1.atitech.com [204.176.13.6]) (authenticated bits=0) by ahgu.homeunix.com (8.12.11/8.12.11) with ESMTP id j8CGF5bu045859 for ; Mon, 12 Sep 2005 12:15:05 -0400 (EDT) (envelope-from ahgu@ahgu.homeunix.com) Message-ID: <075201c5b7b6$87f49090$1f1a12ac@atitech.com> From: "ahgu" To: Date: Mon, 12 Sep 2005 12:24:58 -0400 MIME-Version: 1.0 Content-Type: text/plain; format=flowed; charset="ISO-8859-1"; reply-type=response Content-Transfer-Encoding: 7bit Subject: NAND Write fails Reply-To: ahgu List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , There is problem with the recover function in the latest CVS 2 days ago. Instead of using the breakme in the wbuf.c I put something like this in the nand.c, the lowest write function. after write, in the verification: if(data dont match || block==###) return error. I got the driver to crash in case like this. Has anyone seen this condition? Also, is there a design document for the wbuf.c? I have trouble understanding the jffs2_flash_writev function. thanks Andrew