[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Stuart Wood]

I'm having some new problems with my jffs2 root file system when it is
mounted the first time from small page nand flash.
I'm pretty sure its some sort of configuration problem but the web
hasn't helped so far.

NAND Flash,
erase block = 0x4000, page size

The jffs2 image is being built with this command from Buildroot:

# Use fakeroot so mkfs.jffs2 believes the previous fakery
echo "/home/stuart/ArmLinux/buildroot/toolchain_build_arm_nofpu/mtd_orig/util/mkfs.jffs2
-e 0x4000 -p -l -s 0x200 -n -D target/generic/device_table.txt -d
/home/stuart/ArmLinux/buildroot/build_arm_nofpu/root -o
/home/stuart/ArmLinux/buildroot/rootfs.arm_nofpu.jffs2" \
                >> /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
chmod a+x /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
/opt/buildroot/usr/bin/fakeroot --
/opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
rootdir=/home/stuart/ArmLinux/buildroot/build_arm_nofpu/root
table='target/generic/device_table.txt'
mkfs.jffs2: skipping device_table entry '/dev': no parent directory!
mkfs.jffs2: skipping device_table entry '/tmp': no parent directory!
mkfs.jffs2: skipping device_table entry '/etc': no parent directory!
-rw-r--r-- 1 stuart stuart 2736128 2008-06-13 09:52
/home/stuart/ArmLinux/buildroot/rootfs.arm_nofpu.jffs2
/sbin/ldconfig -r /home/stuart/ArmLinux/buildroot/build_arm_nofpu/root
2>/dev/null

I'm using u-boot 1.3.3 to load the image like this:

-- 
Stuart Wood

Lab X Technologies, LLC
176 Anderson Ave.
Suite 302
Rochester, NY 14607
Phone: (585) 271-7790 x207
Fax: (585) 473.4707

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Stuart Wood]

Sorry for the initial incomplete post......

I'm having some new problems with my jffs2 root file system when it is
mounted the first time from small page nand flash.
I'm pretty sure its some sort of configuration problem but the web
hasn't helped so far.

NAND Flash,
  erase block = 0x4000 (16K)
  page size = 0x200 (512)

The jffs2 image is being built with this command from Buildroot:

# Use fakeroot so mkfs.jffs2 believes the previous fakery
echo "/home/stuart/ArmLinux/buildroot/toolchain_build_arm_nofpu/mtd_orig/util/mkfs.jffs2
-e 0x4000 -p -l -s 0x200 -n -D target/generic/device_table.txt -d
/home/stuart/ArmLinux/buildroot/build_arm_nofpu/root -o
/home/stuart/ArmLinux/buildroot/rootfs.arm_nofpu.jffs2" \
               >> /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
chmod a+x /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
/opt/buildroot/usr/bin/fakeroot --
/opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
rootdir=/home/stuart/ArmLinux/buildroot/build_arm_nofpu/root
table='target/generic/device_table.txt'
mkfs.jffs2: skipping device_table entry '/dev': no parent directory!
mkfs.jffs2: skipping device_table entry '/tmp': no parent directory!
mkfs.jffs2: skipping device_table entry '/etc': no parent directory!
-rw-r--r-- 1 stuart stuart 2736128 2008-06-13 09:52
/home/stuart/ArmLinux/buildroot/rootfs.arm_nofpu.jffs2
/sbin/ldconfig -r /home/stuart/ArmLinux/buildroot/build_arm_nofpu/root
2>/dev/null

I'm using u-boot 1.3.3 to load the image like this:

u-boot>
u-boot> loady
## Ready for binary (ymodem) download to 0xE1000000 at 115200 bps...
C## Total Size      = 0x0029c000 = 2736128 Bytes
u-boot> nand erase clean 500000 1b00000

NAND erase: device 0 offset 0x500000, size 0x1b00000
Erasing at 0x1ffc000 -- 100% complete. Cleanmarker written at 0x1ffc000.
OK
u-boot> nand write.jffs2 e1000000 500000

NAND write: device 0 offset 0x500000, size 0x1b00000

Writing data at 0x1fffe00 -- 100% complete.
 28311552 bytes written: OK
u-boot>

And then when I boot the image I get these errors:

jffs_check_nand_cleanmarker returned 0
jffs_check_nand_cleanmarker returned 0
jffs_check_nand_cleanmarker returned 0
<CLIPPED>
jffs_check_nand_cleanmarker returned 0
jffs_check_nand_cleanmarker returned 0
jffs_check_nand_cleanmarker returned 0
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c000:
0x138e instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c004:
0x1358 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c008:
0x33cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c00c:
0x73c8 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c010:
0x31ec instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c014:
0x35cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c018:
0x93cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c01c:
0x31cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c020:
0x13ce instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c024:
0x728c instead
Further such events for this erase block will not be printed
JFFS2: Erase block at 0x0029c000 is not formatted. It will be erased
jffs_check_nand_cleanmarker returned 0
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0000:
0x138e instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0004:
0x1378 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0008:
0x33cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a000c:
0x73c8 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0010:
0x31ac instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0014:
0x35cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0018:
0x93cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a001c:
0x31cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0020:
0x13ce instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x002a0024:
0x728c instead
Further such events for this erase block will not be printed
JFFS2: Erase block at 0x002a0000 is not formatted. It will be erased
<CLIPPED>
JFFS2: Erase block at 0x01af8000 is not formatted. It will be erased
jffs_check_nand_cleanmarker returned 0
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc000:
0x33c8 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc004:
0x37c4 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc008:
0x33c8 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc00c:
0x33cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc010:
0x1284 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc014:
0x338c instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc018:
0x33c8 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc01c:
0x03e1 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc020:
0x21c8 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x01afc024:
0x33cc instead
Further such events for this erase block will not be printed
JFFS2: Erase block at 0x01afc000 is not formatted. It will be erased
VFS: Mounted root (jffs2 filesystem).
Freeing init memory: 84K
Initializing random number generator... done.
Starting network...
ip: RTNETLINK answers: File exists
udhcpc (v1.2.2) started
Jan  1 00:21:07 udhcpc[171]: udhcpc (v1.2.2) started
Sending discover...
Jan  1 00:21:07 udhcpc[171]: Sending discover...
Sending select for 192.168.0.144...
Jan  1 00:21:08 udhcpc[171]: Sending select for 192.168.0.144...
Lease of 192.168.0.144 obtained, lease time 86400
Jan  1 00:21:08 udhcpc[171]: Lease of 192.168.0.144 obtained, lease time 86400
deleting routers
route: SIOC[ADD|DEL]RT: No such process
adding dns 66.133.170.2
adding dns 67.50.135.146
 Creating FPGA Device Nodes
Spartan IIIe: INITIALIZATION OK
  Found image: MIX_EVM_FPGA
Spartan IIIe: Finished
Reseting SRC4193's
Wrote:  0x 200,
Ininitalizing SRC4193
Ininitalizing AK4620B Codec
Auto Select Mode
Wrote:  0x  72,
MIX Module driver loaded
Mix_IO Initialized
modprobe: module cm2_module not found.
modprobe: failed to load module cm2_module
Creating Mix Module Device Nodes


Welcome to the Mix-EVM.

mix-evm login:

It takes approx. 20 minutes to get through the JFFS@ warnings.

The kernel is 2.6.18 and the JFFS2 settings are:

# CONFIG_JFFS_FS is not set
CONFIG_JFFS2_FS=y
CONFIG_JFFS2_FS_DEBUG=0
CONFIG_JFFS2_FS_WRITEBUFFER=y
CONFIG_JFFS2_SUMMARY=y
# CONFIG_JFFS2_FS_XATTR is not set
# CONFIG_JFFS2_COMPRESSION_OPTIONS is not set
CONFIG_JFFS2_ZLIB=y
CONFIG_JFFS2_RTIME=y
# CONFIG_JFFS2_RUBIN is not set

Any thoughts on what might be wrong would be appriciated!


Thanks

--
Stuart Wood

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Grant Erickson]

On 6/13/08 7:45 AM, Stuart Wood wrote:
> I'm having some new problems with my jffs2 root file system when it is
> mounted the first time from small page nand flash.
> I'm pretty sure its some sort of configuration problem but the web
> hasn't helped so far.
> 
> [ ... ]
> 
> The jffs2 image is being built with this command from Buildroot:
> 
> # Use fakeroot so mkfs.jffs2 believes the previous fakery
> echo 
> "/home/stuart/ArmLinux/buildroot/toolchain_build_arm_nofpu/mtd_orig/util/mkfs.
> jffs2
> -e 0x4000 -p -l -s 0x200 -n -D target/generic/device_table.txt -d
> /home/stuart/ArmLinux/buildroot/build_arm_nofpu/root -o
> /home/stuart/ArmLinux/buildroot/rootfs.arm_nofpu.jffs2" \
>>> /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
> chmod a+x /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
> /opt/buildroot/usr/bin/fakeroot --
> /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
> rootdir=/home/stuart/ArmLinux/buildroot/build_arm_nofpu/root
> table='target/generic/device_table.txt'
> mkfs.jffs2: skipping device_table entry '/dev': no parent directory!
> mkfs.jffs2: skipping device_table entry '/tmp': no parent directory!
> mkfs.jffs2: skipping device_table entry '/etc': no parent directory!

Regarding the above three errors, you may be interested in the following
patch and referenced thread:

http://lists.infradead.org/pipermail/linux-mtd/2008-June/022008.html

> u-boot> nand erase clean 500000 1b00000

I cannot speak to your specific board; however, I found that 'clean' was not
necessary and used the following sequence to write a JFFS2 image to Samsung
small-page NAND on my PPC405EXr-based board:

    => tftp 800000 root.img
    => nand erase 800000 1400000
    => nand write.i ${fileaddr} 800000 ${filesize}

Adjust sizes and offsets accordingly for your board.

Regards,

Grant Erickson

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Stuart Wood]

Hi Grant,

Tried what you said including patching mkfs.jffs2 and I'm now getting
these errors

jffs_check_nand_cleanmarker returned 1
jffs2_check_nand_cleanmarker(): Cleanmarker node not detected in block
at 00298000
OOB at 00298000 was cf 0c 3f c3 ff ff cc 0f ff ff ff ff ff ff ff ff
jffs_check_nand_cleanmarker returned 1
jffs2_check_nand_cleanmarker(): Cleanmarker node not detected in block
at 0029c000
OOB at 0029c000 was 3f 00 cf f0 ff ff 0f 33 ff ff ff ff ff ff ff ff
jffs_check_nand_cleanmarker returned 1
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c000:
0x138e instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c004:
0x1358 instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c008:
0x33cc instead
jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x0029c00c:
0x73c8 instead

This seems to imply that the kernel is looking for the clean markers.
If they are not needed
with NAND flash then why would it look for them? I did not see
anything in the kernel configs
for JFFS2 that would tell it to use use them.

As I said before this is kernel 2.6.18, an older version of build root
using mtd_20050122

Any other thoughts?

Stuart
On Fri, Jun 13, 2008 at 11:54 AM, Grant Erickson
<gerickson@nuovations.com> wrote:

> On 6/13/08 7:45 AM, Stuart Wood wrote:
>> I'm having some new problems with my jffs2 root file system when it is
>> mounted the first time from small page nand flash.
>> I'm pretty sure its some sort of configuration problem but the web
>> hasn't helped so far.
>>
>> [ ... ]
>>
>> The jffs2 image is being built with this command from Buildroot:
>>
>> # Use fakeroot so mkfs.jffs2 believes the previous fakery
>> echo
>> "/home/stuart/ArmLinux/buildroot/toolchain_build_arm_nofpu/mtd_orig/util/mkfs.
>> jffs2
>> -e 0x4000 -p -l -s 0x200 -n -D target/generic/device_table.txt -d
>> /home/stuart/ArmLinux/buildroot/build_arm_nofpu/root -o
>> /home/stuart/ArmLinux/buildroot/rootfs.arm_nofpu.jffs2" \
>>>> /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
>> chmod a+x /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
>> /opt/buildroot/usr/bin/fakeroot --
>> /opt/buildroot/_fakeroot.rootfs.arm_nofpu.jffs2
>> rootdir=/home/stuart/ArmLinux/buildroot/build_arm_nofpu/root
>> table='target/generic/device_table.txt'
>> mkfs.jffs2: skipping device_table entry '/dev': no parent directory!
>> mkfs.jffs2: skipping device_table entry '/tmp': no parent directory!
>> mkfs.jffs2: skipping device_table entry '/etc': no parent directory!
>
> Regarding the above three errors, you may be interested in the following
> patch and referenced thread:
>
> http://lists.infradead.org/pipermail/linux-mtd/2008-June/022008.html
>
>> u-boot> nand erase clean 500000 1b00000
>
> I cannot speak to your specific board; however, I found that 'clean' was not
> necessary and used the following sequence to write a JFFS2 image to Samsung
> small-page NAND on my PPC405EXr-based board:
>
>    => tftp 800000 root.img
>    => nand erase 800000 1400000
>    => nand write.i ${fileaddr} 800000 ${filesize}
>
> Adjust sizes and offsets accordingly for your board.
>
> Regards,
>
> Grant Erickson
>
>
>



-- 
Stuart Wood

Lab X Technologies, LLC
176 Anderson Ave.
Suite 302
Rochester, NY 14607
Phone: (585) 271-7790 x207
Fax: (585) 473.4707

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Grant Erickson]

On 6/26/08 10:27 AM, Stuart Wood wrote:
> Tried what you said including patching mkfs.jffs2 and I'm now getting
> these errors
> 
> [ errors omitted ]
>
> This seems to imply that the kernel is looking for the clean markers.
> If they are not needed
> with NAND flash then why would it look for them? I did not see
> anything in the kernel configs
> for JFFS2 that would tell it to use use them.
> 
> As I said before this is kernel 2.6.18, an older version of build root
> using mtd_20050122
> 
> Any other thoughts?

Stuart:

Without knowing more about the specific origin and configuration of your
kernel, version and configuration of u-boot and model, revision and geometry
of NAND, it is difficult to speculate.

I am using:

    AMCC PowerPC 405EXr
    Samsung K9F1208U0B NAND
    U-boot 1.3.3
    Linux 2.6.25-rc3-DENX

Are you certain that u-boot and linux are in agreement about the NAND layout
and format? I had some initial problems in that area until I applied a patch
to u-boot (prior to 1.3.3) that addressed the issue and brought the two into
agreement.

Regards,

Grant

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Stuart Wood]

On Thu, Jun 26, 2008 at 1:55 PM, Grant Erickson
<gerickson@nuovations.com> wrote:

> Without knowing more about the specific origin and configuration of your
> kernel, version and configuration of u-boot and model, revision and geometry
> of NAND, it is difficult to speculate.
>
> I am using:
>
>    AMCC PowerPC 405EXr
>    Samsung K9F1208U0B NAND
>    U-boot 1.3.3
>    Linux 2.6.25-rc3-DENX
>

I'm using the following:
      Cirrus Logic EP9302 (ARM920T)
      ST NAND256W3A2BNG   NAND
      U-boot 1.3.3 modified for the EP93xx
      Linux 2.8.18 + EP93xx patches

> Are you certain that u-boot and linux are in agreement about the NAND layout
> and format? I had some initial problems in that area until I applied a patch
> to u-boot (prior to 1.3.3) that addressed the issue and brought the two into
> agreement.

Can you expand on this? I've set the partition tables to be the same
between the two,
but I'm not sure what else needs to synchronized. MTD version?


-- 
Stuart Wood

Lab X Technologies, LLC
176 Anderson Ave.
Suite 302
Rochester, NY 14607
Phone: (585) 271-7790 x207
Fax: (585) 473.4707

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Kenneth Johansson]

On Thu, 2008-06-26 at 14:16 -0400, Stuart Wood wrote:
> On Thu, Jun 26, 2008 at 1:55 PM, Grant Erickson
> > Are you certain that u-boot and linux are in agreement about the NAND layout
> > and format? I had some initial problems in that area until I applied a patch
> > to u-boot (prior to 1.3.3) that addressed the issue and brought the two into
> > agreement.
> 
> Can you expand on this? I've set the partition tables to be the same
> between the two,
> but I'm not sure what else needs to synchronized. MTD version?

jffs2 stores the cleanmarker in the oob area. u-boot and linux needs to
be configured so it is stored in the same place in the oob area.

do an erase with cleanmarker from u-boot and read out what is stored in
oob then do the same under linux and you will probably have the
cleanmarker att different places.

if you do not see the cleanmarker you have it in the same place as ecc
and that is also wrong.

fixing it is specific to the nand chip and nand controller that is used.

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Andrejs Cainikovs]

Stuart Wood wrote:
> I'm using the following:
>       Cirrus Logic EP9302 (ARM920T)
>       ST NAND256W3A2BNG   NAND
>       U-boot 1.3.3 modified for the EP93xx
>       Linux 2.8.18 + EP93xx patches
>   
Woohah! Could you please send me a copy of your kernel source?

Kind regards,
Andrejs Cainikovs.

[U-Boot-Users] Problems with root JFFS2 on NAND Flash

[Stuart Wood]

Kenneth,

Thanks for the suggestions. I tried testing the clen markers and they
appear to be fine.
So I don't know why the kernel complains so much.....

This is what I did;

u-boot> nand erase clean kernel_2

NAND erase: device 0 offset 0x300000, size 0x200000
Erasing at 0x4fc000 -- 100% complete. Cleanmarker written at 0x4fc000.
written at 0x388000.
OK
u-boot> nand dump 0x300000
Page 00300000 dump:
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
        ff ff ff ff ff ff ff ff  ff ff ff ff ff ff ff ff
OOB:
        ff ff ff ff ff ff ff ff
        85 19 03 20 08 00 00 00
u-boot>

# nanddump -p  -l 512 /dev/mtd5
Block size 16384, page size 512, OOB size 16
Dumping data starting at 0x00000000 and ending at 0x00000200...
0x00000000: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000010: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000020: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000030: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000040: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000050: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000060: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000070: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000080: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000090: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000a0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000b0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000c0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000d0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000e0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000f0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000100: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000110: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000120: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000130: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000140: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000150: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000160: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000170: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000180: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000190: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001a0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001b0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001c0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001d0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001e0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001f0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
  OOB Data: ff ff ff ff ff ff ff ff 85 19 03 20 08 00 00 00

# nanddump -p  -l 512 /dev/mtd5
Block size 16384, page size 512, OOB size 16
Dumping data starting at 0x00000000 and ending at 0x00000200...
0x00000000: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000010: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000020: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000030: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000040: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000050: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000060: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000070: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000080: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000090: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000a0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000b0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000c0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000d0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000e0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000000f0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000100: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000110: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000120: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000130: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000140: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000150: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000160: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000170: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000180: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x00000190: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001a0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001b0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001c0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001d0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001e0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
0x000001f0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
  OOB Data: ff ff ff ff ff ff ff ff 85 19 03 20 08 00 00 00

As you can see the clean marker is in the same place.

I did not understand this part here

On Thu, Jun 26, 2008 at 2:51 PM, Kenneth Johansson <kenneth@southpole.se> wrote:
> if you do not see the cleanmarker you have it in the same place as ecc
> and that is also wrong.
>
> fixing it is specific to the nand chip and nand controller that is used.

There is no nand controller, the interface is bit banged with software ecc.
The ECC data seems to be loaded in the oob bytes 0-3, 6, 7 witch
should not conflict
right?

-- 
Stuart Wood

Lab X Technologies, LLC
176 Anderson Ave.
Suite 302
Rochester, NY 14607
Phone: (585) 271-7790 x207
Fax: (585) 473.4707