From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mta-rtd-2.nl.easynet.net ([194.165.92.5]) by pentafluge.infradead.org with esmtp (Exim 4.30 #5 (Red Hat Linux)) id 1AetMA-0002sU-TW for linux-mtd@lists.infradead.org; Fri, 09 Jan 2004 09:55:06 +0000 Received: from unspecified.host ([195.86.30.10])i099qxCt026859 for ; Fri, 9 Jan 2004 10:53:00 +0100 (CET) Message-ID: <002e01c3d696$8a421190$1800a8c0@inetrjasper> From: "Jasper Jackson" To: "MTD Mailinglist" Date: Fri, 9 Jan 2004 10:54:14 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Subject: EDB9312 Flash chips Reply-To: Jasper Jackson List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hi All, We've been working around for a while with the EDB9312 developmentboard from Cirrus Logic, Now the time has come that we want a filesystem on our flashchips, the problem is, that the flash chips don't seem to respond to a CFI probe. There is a memory map written by David for the edb7312 (the little brother). It seems that the only thing that has to be done is ajust the WINDOW_START and the WINDOW_SIZE, but that doesn't work. CFI Flash device in physical memory map, seems to work when set to: Physical start: 0x60000000 Physical length: 0x2000000 But reading from those adresses results in crep, Almost everything I read is 'R' or 'U' and some dots, no binary data from ie redboot which resides at adress 0x60000000 is found. When lowlevel writing to those adresses (from a homebrew kernel module), the data written can be read back that goes OK, but after a reboot it's gone, not exactly what you expect from flash memory. According to the Cirrus Documentation those are the right adresses, it is also said that there is no difference between the physical and the virual adress (not that I believe that, but ok). Before I go ahead with JFFS I think that it's important that my flashchips are detected. Ok, ready for that part, My question however, is, are there people familiar with MTD and the EDB9312, i've been working for a week now on it, I think i'm doing something terribly stupid wrong, I should be ashamed for that, I know, but I don't see it at the moment. Some extra information: Kernel: 2.4.21-rmk1 Config: MTD Support ON Direct char device access ON Caching block device access ON ROM/Flashchip drivers -> Detect flash by CFI ROM/Flashchip drivers -> Intel/sharp support Thanks in advance, Jasper Jackson