From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <lagrange@fr.oleane.com>
Received: from smtp04.msg.oleane.net (smtp04.msg.oleane.net [62.161.4.4])
	by ozlabs.org (Postfix) with ESMTP id 784B5DDFE1
	for <linuxppc-embedded@ozlabs.org>;
	Tue, 12 Jun 2007 01:46:02 +1000 (EST)
Received: from GEG2400 ([194.3.133.184]) (authenticated)
	by smtp04.msg.oleane.net (MTA) with ESMTP id l5BFKkP6029675
	for <linuxppc-embedded@ozlabs.org>; Mon, 11 Jun 2007 17:20:47 +0200
From: "Laurent Lagrange" <lagrange@fr.oleane.com>
To: <linuxppc-embedded@ozlabs.org>
Subject: Linux 2.6.x, MTD drivers and NAND flashes
Date: Mon, 11 Jun 2007 17:20:43 +0200
Message-ID: <000001c7ac3c$15e7da80$5201a8c0@GEG2400>
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_NextPart_000_0001_01C7AC4C.D972F470"
In-Reply-To: 
List-Id: Linux on Embedded PowerPC Developers Mail List
	<linuxppc-embedded.ozlabs.org>
List-Unsubscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=unsubscribe>
List-Archive: <http://ozlabs.org/pipermail/linuxppc-embedded>
List-Post: <mailto:linuxppc-embedded@ozlabs.org>
List-Help: <mailto:linuxppc-embedded-request@ozlabs.org?subject=help>
List-Subscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=subscribe>

This is a multi-part message in MIME format.

------=_NextPart_000_0001_01C7AC4C.D972F470
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit

Hello,

I use a Linux 2.6.9 on a 85xx custom board with a NAND flash.

This flash has some bad blocks.
NAND device: Manufacturer ID: 0xec, Chip ID: 0xf1 (Samsung NAND 128MiB 3,3V
8-bit)
Scanning device for bad blocks
Bad eraseblock 8 at 0x00100000
Bad eraseblock 430 at 0x035c0000
...
Creating 1 MTD partitions on "NAND 128MiB 3,3V 8-bit":
0x00000000-0x08000000 : "flash partition 1".

I have selected in the kernel :
1) the MTD character and block supports for NAND,
2) the CRAMFS and JFFS2 supports for NAND.

I can erase the flash but receive IO errors for each bad blocks (that seems
right).
I can mount an empty JFFS2 partition on the NAND and untar some directories
and files in it.

I would like to generate some JFFS2 images and program them in the flash
with a character
command like "cat img > /dev/mtd/0". If I dont encounter a bad block, it
runs right else the
command fails.

This is a little bit annoying. So my question is simple :
Is there a MTD character driver which can detect and ignore the bad blocks
in a NAND flash ?

Thanks for any tips.
Bye
Laurent




------=_NextPart_000_0001_01C7AC4C.D972F470
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<TITLE></TITLE>

<META content=3D"MSHTML 6.00.6000.16441" name=3DGENERATOR></HEAD>
<BODY>
<DIV><FONT face=3DArial size=3D2>Hello,<BR><BR>I use a Linux 2.6.9 on a =
85xx custom=20
board with a NAND flash. </FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>This flash has some bad =
blocks.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>NAND device: Manufacturer ID: 0xec, =
Chip ID: 0xf1=20
(Samsung NAND 128MiB 3,3V 8-bit)<BR>Scanning device for bad =
blocks<BR>Bad=20
eraseblock 8 at 0x00100000<BR>Bad eraseblock 430 at =
0x035c0000</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>...<BR>Creating 1 MTD partitions on =
"NAND 128MiB=20
3,3V 8-bit":<BR>0x00000000-0x08000000 : "flash partition =
1".</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>I have selected in the kernel =
:</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>1) the MTD character and block supports =
for=20
NAND,</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>2) the CRAMFS and JFFS2 supports for=20
NAND.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>I can erase the flash but receive IO =
errors for=20
each bad blocks (that seems right).</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>I can mount an empty JFFS2 partition on =
the NAND=20
and untar some directories and files in it.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>I would like to generate some JFFS2 =
images and=20
program them in the flash </FONT><FONT face=3DArial size=3D2>with a =
character=20
</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>command like "cat img &gt; /dev/mtd/0". =

</FONT><FONT face=3DArial size=3D2>If I dont encounter a bad block, it =
runs=20
right&nbsp;else the </FONT></DIV>
<DIV><FONT face=3DArial size=3D2>command fails. </FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>This is a little bit annoying. So =
m</FONT><FONT=20
face=3DArial size=3D2>y question is simple : </FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Is there a MTD character driver which =
can detect=20
and ignore the bad blocks in a NAND flash&nbsp;?&nbsp;</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT face=3DArial size=3D2>Thanks for any tips.</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Bye</FONT></DIV>
<DIV><FONT face=3DArial size=3D2>Laurent</FONT></DIV>
<DIV><FONT face=3DArial size=3D2></FONT>&nbsp;</DIV>
<DIV><FONT><FONT face=3DArial size=3D2>&nbsp;</FONT></DIV>
<DIV><FONT face=3DArial><BR><FONT =
size=3D2></FONT></FONT></DIV></FONT></BODY></HTML>

------=_NextPart_000_0001_01C7AC4C.D972F470--

From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <bgat@billgatliff.com>
Received: from smtp122.sbc.mail.re3.yahoo.com (smtp122.sbc.mail.re3.yahoo.com
	[66.196.96.95]) by ozlabs.org (Postfix) with SMTP id 21DBBDDF1D
	for <linuxppc-embedded@ozlabs.org>;
	Tue, 12 Jun 2007 02:12:18 +1000 (EST)
Message-ID: <466D7169.1000407@billgatliff.com>
Date: Mon, 11 Jun 2007 10:59:37 -0500
From: Bill Gatliff <bgat@billgatliff.com>
MIME-Version: 1.0
To: Laurent Lagrange <lagrange@fr.oleane.com>
Subject: Re: Linux 2.6.x, MTD drivers and NAND flashes
References: <000001c7ac3c$15e7da80$5201a8c0@GEG2400>
In-Reply-To: <000001c7ac3c$15e7da80$5201a8c0@GEG2400>
Content-Type: multipart/alternative;
	boundary="------------050102040909040701090807"
Cc: linuxppc-embedded@ozlabs.org
List-Id: Linux on Embedded PowerPC Developers Mail List
	<linuxppc-embedded.ozlabs.org>
List-Unsubscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=unsubscribe>
List-Archive: <http://ozlabs.org/pipermail/linuxppc-embedded>
List-Post: <mailto:linuxppc-embedded@ozlabs.org>
List-Help: <mailto:linuxppc-embedded-request@ozlabs.org?subject=help>
List-Subscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=subscribe>

This is a multi-part message in MIME format.
--------------050102040909040701090807
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Laurent Lagrange wrote:
> Hello,
>
> I use a Linux 2.6.9 on a 85xx custom board with a NAND flash.
>  
> This flash has some bad blocks.
> NAND device: Manufacturer ID: 0xec, Chip ID: 0xf1 (Samsung NAND 128MiB 
> 3,3V 8-bit)
> Scanning device for bad blocks
> Bad eraseblock 8 at 0x00100000
> Bad eraseblock 430 at 0x035c0000
> ...
> Creating 1 MTD partitions on "NAND 128MiB 3,3V 8-bit":
> 0x00000000-0x08000000 : "flash partition 1".
>  
> I have selected in the kernel :
> 1) the MTD character and block supports for NAND,
> 2) the CRAMFS and JFFS2 supports for NAND.
>  
> I can erase the flash but receive IO errors for each bad blocks (that 
> seems right).
> I can mount an empty JFFS2 partition on the NAND and untar some 
> directories and files in it.
>  
> I would like to generate some JFFS2 images and program them in the 
> flash with a character
> command like "cat img > /dev/mtd/0". If I dont encounter a bad block, 
> it runs right else the
> command fails.
>  
> This is a little bit annoying. So my question is simple :
> Is there a MTD character driver which can detect and ignore the bad 
> blocks in a NAND flash ?

Wrong question.  :)

With NAND, working directly with the media is indeed a pain because of 
bad blocks.  What's done instead is to leave that to the filesystem.  
Pick a NAND-aware one like YAFFS.


b.g.

-- 
Bill Gatliff
bgat@billgatliff.com


--------------050102040909040701090807
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Laurent Lagrange wrote:
<blockquote cite="mid000001c7ac3c$15e7da80$5201a8c0@GEG2400" type="cite">
  <meta http-equiv="Content-Type" content="text/html; ">
  <title></title>
  <meta content="MSHTML 6.00.6000.16441" name="GENERATOR">
  <div><font face="Arial" size="2">Hello,<br>
  <br>
I use a Linux 2.6.9 on a 85xx custom board with a NAND flash. </font></div>
  <div>&nbsp;</div>
  <div><font face="Arial" size="2">This flash has some bad blocks.</font></div>
  <div><font face="Arial" size="2">NAND device: Manufacturer ID: 0xec,
Chip ID: 0xf1 (Samsung NAND 128MiB 3,3V 8-bit)<br>
Scanning device for bad blocks<br>
Bad eraseblock 8 at 0x00100000<br>
Bad eraseblock 430 at 0x035c0000</font></div>
  <div><font face="Arial" size="2">...<br>
Creating 1 MTD partitions on "NAND 128MiB 3,3V 8-bit":<br>
0x00000000-0x08000000 : "flash partition 1".</font></div>
  <div>&nbsp;</div>
  <div><font face="Arial" size="2">I have selected in the kernel :</font></div>
  <div><font face="Arial" size="2">1) the MTD character and block
supports for NAND,</font></div>
  <div><font face="Arial" size="2">2) the CRAMFS and JFFS2 supports for
NAND.</font></div>
  <div>&nbsp;</div>
  <div><font face="Arial" size="2">I can erase the flash but receive IO
errors for each bad blocks (that seems right).</font></div>
  <div><font face="Arial" size="2">I can mount an empty JFFS2 partition
on the NAND and untar some directories and files in it.</font></div>
  <div>&nbsp;</div>
  <div><font face="Arial" size="2">I would like to generate some JFFS2
images and program them in the flash </font><font face="Arial" size="2">with
a character </font></div>
  <div><font face="Arial" size="2">command like "cat img &gt;
/dev/mtd/0". </font><font face="Arial" size="2">If I dont encounter a
bad block, it runs right&nbsp;else the </font></div>
  <div><font face="Arial" size="2">command fails. </font></div>
  <div>&nbsp;</div>
  <div><font face="Arial" size="2">This is a little bit annoying. So m</font><font
 face="Arial" size="2">y question is simple : </font></div>
  <div><font face="Arial" size="2">Is there a MTD character driver
which can detect and ignore the bad blocks in a NAND flash&nbsp;? <br>
  </font></div>
</blockquote>
<br>
Wrong question.&nbsp; :)<br>
<br>
With NAND, working directly with the media is indeed a pain because of
bad blocks.&nbsp; What's done instead is to leave that to the filesystem.&nbsp;
Pick a NAND-aware one like YAFFS.<br>
<br>
<br>
b.g.<br>
<pre class="moz-signature" cols="72">-- 
Bill Gatliff
<a class="moz-txt-link-abbreviated" href="mailto:bgat@billgatliff.com">bgat@billgatliff.com</a>
</pre>
</body>
</html>

--------------050102040909040701090807--

From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gkenaley@echelon.com>
Received: from [205.229.50.10] (ntfw.echelon.com [205.229.50.10])
	by ozlabs.org (Postfix) with ESMTP id 3053BDDEC6
	for <linuxppc-embedded@ozlabs.org>;
	Tue, 12 Jun 2007 05:59:55 +1000 (EST)
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Subject: RE: Linux 2.6.x, MTD drivers and NAND flashes
Date: Mon, 11 Jun 2007 12:59:53 -0700
Message-ID: <DDBD7B17DB2ECE48BCD94C593F7255B4046EA149@monk.echelon.echcorp.com>
In-reply-to: <466D7169.1000407@billgatliff.com>
References: <000001c7ac3c$15e7da80$5201a8c0@GEG2400>
	<466D7169.1000407@billgatliff.com>
From: "Gary Kenaley" <gkenaley@echelon.com>
To: "Bill Gatliff" <bgat@billgatliff.com>,
	"Laurent Lagrange" <lagrange@fr.oleane.com>
Cc: linuxppc-embedded@ozlabs.org
List-Id: Linux on Embedded PowerPC Developers Mail List
	<linuxppc-embedded.ozlabs.org>
List-Unsubscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=unsubscribe>
List-Archive: <http://ozlabs.org/pipermail/linuxppc-embedded>
List-Post: <mailto:linuxppc-embedded@ozlabs.org>
List-Help: <mailto:linuxppc-embedded-request@ozlabs.org?subject=help>
List-Subscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=subscribe>

Laurent,

Can't you just use the flashcp copy command from the mtd-utils package?
-----------------------------------------------------------------------
flash_eraseall -j /dev/mtd0

mkfs.jffs2 -b -n --eraseblock=3D131072 --pagesize=3D2048 -p -r =
/test_root -v -o test_root.img

flashcp test_root.img /dev/mtd0


The other way is what Bill mentions:
------------------------------------
flash_eraseall -j /dev/mtd0
mount -t jffs2 /dev/mtdblock0 /tmp/mnt
tar -xzf test_root.tgz -C /tmp/mnt
umount /tmp/mnt

Gary

> > ________________________________________
> > From: linuxppc-embedded-bounces+gkenaley=3Dechelon.com@ozlabs.org > =
> > [mailto:linuxppc-embedded-bounces+gkenaley=3Dechelon.com@ozlabs.org] =
On Behalf > > Of Bill Gatliff
> > Sent: Monday, June 11, 2007 9:00 AM
> > To: Laurent Lagrange
> > Cc: linuxppc-embedded@ozlabs.org
> > Subject: Re: Linux 2.6.x, MTD drivers and NAND flashes
> >
> > Laurent Lagrange wrote:=20
> > Hello,
> >
> > I use a Linux 2.6.9 on a 85xx custom board with a NAND flash.=20
> >=A0
> > This flash has some bad blocks.
> > NAND device: Manufacturer ID: 0xec, Chip ID: 0xf1 (Samsung NAND =
128MiB > > 3,3V 8-bit)
> > Scanning device for bad blocks
> > Bad eraseblock 8 at 0x00100000
> > Bad eraseblock 430 at 0x035c0000
> >...
> > Creating 1 MTD partitions on "NAND 128MiB 3,3V 8-bit":
> > 0x00000000-0x08000000 : "flash partition 1".
> >=A0
> > I have selected in the kernel :
> > 1) the MTD character and block supports for NAND,
> > 2) the CRAMFS and JFFS2 supports for NAND.
> >=A0
> > I can erase the flash but receive IO errors for each bad blocks =
(that > > seems right).
> > I can mount an empty JFFS2 partition on the NAND and untar some > > =
directories and files in it.
> >=A0
> > I would like to generate some JFFS2 images and program them in the =
flash > > with a character=20
> > command like "cat img > /dev/mtd/0". If I dont encounter a bad =
block, it > > runs right=A0else the=20
> > command fails.=20
> >=A0
> > This is a little bit annoying. So my question is simple :=20
> > Is there a MTD character driver which can detect and ignore the bad =
> blocks > in a NAND flash=A0?=20
> >
> Wrong question.=A0 :)

> With NAND, working directly with the media is indeed a pain because of =
bad > blocks.=A0 What's done instead is to leave that to the =
filesystem.=A0 Pick a > NAND-aware one like YAFFS.
>
>
> b.g.
>
> --=20
> Bill Gatliff
> bgat@billgatliff.com

From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <lagrange@fr.oleane.com>
Received: from smtp02.msg.oleane.net (smtp02.msg.oleane.net [62.161.4.2])
	by ozlabs.org (Postfix) with ESMTP id 1516CDDE02
	for <linuxppc-embedded@ozlabs.org>;
	Tue, 12 Jun 2007 20:29:07 +1000 (EST)
From: "Laurent Lagrange" <lagrange@fr.oleane.com>
To: "'Gary Kenaley'" <gkenaley@echelon.com>,
	"'Bill Gatliff'" <bgat@billgatliff.com>
Subject: RE: Linux 2.6.x, MTD drivers and NAND flashes
Date: Tue, 12 Jun 2007 12:06:29 +0200
Message-ID: <000f01c7acd9$633bc290$5201a8c0@GEG2400>
MIME-Version: 1.0
Content-Type: multipart/mixed;
	boundary="----=_NextPart_000_0010_01C7ACEA.26C70390"
In-Reply-To: <DDBD7B17DB2ECE48BCD94C593F7255B4046EA149@monk.echelon.echcorp.com>
Cc: linuxppc-embedded@ozlabs.org
List-Id: Linux on Embedded PowerPC Developers Mail List
	<linuxppc-embedded.ozlabs.org>
List-Unsubscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=unsubscribe>
List-Archive: <http://ozlabs.org/pipermail/linuxppc-embedded>
List-Post: <mailto:linuxppc-embedded@ozlabs.org>
List-Help: <mailto:linuxppc-embedded-request@ozlabs.org?subject=help>
List-Subscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=subscribe>

This is a multi-part message in MIME format.

------=_NextPart_000_0010_01C7ACEA.26C70390
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 8bit


Gary and Bill,

My need is to load a basic Linux kernel on boards to program multiple 
NOR and NAND flashes with CRAMFS and JFFS2 (perhaps YAFFS) images.

I usually used mtd character accesses to do the job on NOR flashes.
But flash_eraseall and flashcp seem to be excellent tools.
I'll try them quickly.

Many thanks for your help
Bye
Laurent

> -----Message d'origine-----
> De : Gary Kenaley [mailto:gkenaley@echelon.com]
> Envoyé : lun. 11 juin 2007 21:00
> À : Bill Gatliff; Laurent Lagrange
> Cc : linuxppc-embedded@ozlabs.org
> Objet : RE: Linux 2.6.x, MTD drivers and NAND flashes
> 
> 
> Laurent,
> 
> Can't you just use the flashcp copy command from the 
> mtd-utils package?
> --------------------------------------------------------------
> ---------
> flash_eraseall -j /dev/mtd0
> 
> mkfs.jffs2 -b -n --eraseblock=131072 --pagesize=2048 -p -r 
> /test_root -v -o test_root.img
> 
> flashcp test_root.img /dev/mtd0
> 
> 
> The other way is what Bill mentions:
> ------------------------------------
> flash_eraseall -j /dev/mtd0
> mount -t jffs2 /dev/mtdblock0 /tmp/mnt
> tar -xzf test_root.tgz -C /tmp/mnt
> umount /tmp/mnt
> 
> Gary

------=_NextPart_000_0010_01C7ACEA.26C70390
Content-Type: application/ms-tnef;
	name="winmail.dat"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
	filename="winmail.dat"
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------=_NextPart_000_0010_01C7ACEA.26C70390--

From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <glppe-linuxppc-embedded-2@m.gmane.org>
Received: from ciao.gmane.org (main.gmane.org [80.91.229.2])
	(using TLSv1 with cipher AES256-SHA (256/256 bits))
	(Client did not present a certificate)
	by ozlabs.org (Postfix) with ESMTP id C7824DDDF8
	for <linuxppc-embedded@ozlabs.org>;
	Tue, 12 Jun 2007 23:38:41 +1000 (EST)
Received: from list by ciao.gmane.org with local (Exim 4.43)
	id 1Hy6ZZ-000078-Lu
	for linuxppc-embedded@ozlabs.org; Tue, 12 Jun 2007 15:38:13 +0200
Received: from p5497d0b8.dip.t-dialin.net ([84.151.208.184])
	by main.gmane.org with esmtp (Gmexim 0.1 (Debian))
	id 1AlnuQ-0007hv-00
	for <linuxppc-embedded@ozlabs.org>; Tue, 12 Jun 2007 15:38:13 +0200
Received: from dzu by p5497d0b8.dip.t-dialin.net with local (Gmexim 0.1
	(Debian)) id 1AlnuQ-0007hv-00
	for <linuxppc-embedded@ozlabs.org>; Tue, 12 Jun 2007 15:38:13 +0200
To: linuxppc-embedded@ozlabs.org
From: Detlev Zundel <dzu@denx.de>
Subject: Re: Linux 2.6.x, MTD drivers and NAND flashes
Date: Tue, 12 Jun 2007 15:37:43 +0200
Message-ID: <m2k5u9xpi0.fsf@sowhat.denx.de>
References: <000001c7ac3c$15e7da80$5201a8c0@GEG2400>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: news <news@sea.gmane.org>
List-Id: Linux on Embedded PowerPC Developers Mail List
	<linuxppc-embedded.ozlabs.org>
List-Unsubscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=unsubscribe>
List-Archive: <http://ozlabs.org/pipermail/linuxppc-embedded>
List-Post: <mailto:linuxppc-embedded@ozlabs.org>
List-Help: <mailto:linuxppc-embedded-request@ozlabs.org?subject=help>
List-Subscribe: <https://ozlabs.org/mailman/listinfo/linuxppc-embedded>,
	<mailto:linuxppc-embedded-request@ozlabs.org?subject=subscribe>

Hi Laurent,

> I use a Linux 2.6.9 on a 85xx custom board with a NAND flash.
>  
> This flash has some bad blocks.
> NAND device: Manufacturer ID: 0xec, Chip ID: 0xf1 (Samsung NAND 128MiB 3,3V
> 8-bit)
> Scanning device for bad blocks
> Bad eraseblock 8 at 0x00100000
> Bad eraseblock 430 at 0x035c0000
> ...
> Creating 1 MTD partitions on "NAND 128MiB 3,3V 8-bit":
> 0x00000000-0x08000000 : "flash partition 1".
>  
> I have selected in the kernel :
> 1) the MTD character and block supports for NAND,
> 2) the CRAMFS and JFFS2 supports for NAND.
>  
> I can erase the flash but receive IO errors for each bad blocks (that seems
> right).
> I can mount an empty JFFS2 partition on the NAND and untar some directories and
> files in it.
>  
> I would like to generate some JFFS2 images and program them in the flash with a
> character
> command like "cat img > /dev/mtd/0". If I dont encounter a bad block, it runs
> right else the
> command fails.
>  
> This is a little bit annoying. So my question is simple :
> Is there a MTD character driver which can detect and ignore the bad blocks in a
> NAND flash ? 

Actually there is the UBI[1] subsystem dedicated to this problem which
entered mainline late in April.  If you have that in your kernel, you
can feed your flash to UBI and get an bad-block aware mtd device out
of it.  This should allow your intended usage.  Of course you can have
JFFS2 sit on top of that.

Cheers
  Detlev

[1] http://www.linux-mtd.infradead.org/faq/ubi.html

-- 
I have always observed that the pretensions of all people are in
exact inverse ratio to their merits; this is one of the axioms of
morals.                            -- Joseph Lagrange
--
DENX Software Engineering GmbH,      MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich,  Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-40 Fax: (+49)-8142-66989-80 Email: dzu@denx.de