From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from endruntechnologies.com ([130.94.228.14]) by canuck.infradead.org with esmtp (Exim 4.33 #1 (Red Hat Linux)) id 1BjC2n-0002HS-J5 for linux-mtd@lists.infradead.org; Sat, 10 Jul 2004 03:13:10 -0400 Received: from unknown (HELO endruntechnologies.com) ([67.169.85.122]) (envelope-sender ) by 130.94.228.14 (qmail-ldap-1.03) with SMTP for ; 10 Jul 2004 07:13:07 -0000 Message-ID: <40EF9703.1010907@endruntechnologies.com> Date: Sat, 10 Jul 2004 00:13:07 -0700 From: Bruce Penrod MIME-Version: 1.0 To: linux-mtd@lists.infradead.org Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Subject: ST Micro M29W641D clone of AM29LV641D needs extra reset command List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Ths ST Micro M29W641D nor flash chip is a clone of the AMD AM29LV641D. As usual, it's not quite a perfect clone and I've solved a problem in using it with the cfi_cmdset_0002() probe that is included in the 2.4.26 kernel mtd driver. The difference in the two chips is that after a cfi query command is performed while in the auto-select state, the ST Micro chip needs two reset commands to return to the read state. The first reset command only causes the ST Micro chip to go back to the auto-select state. The AMD part jumps all the way back to the read state with only one reset comand. So the very simple fix which seems to work for me is to add a second reset command after the existing one at the bottom of cfi_cmdset_0002(). This is the only place in the mtd code where I found cfi query being performed while in the auto-select state. I haven't tried to determine if this is a common problem across the entire ST Micro AMD clone flash product line, but the second reset command can do no harm and might solve problems for many. Bruce