Seagate GoFlex Net - A lot of bad block in "DMESG"

Hello All,
Before I start to beg for help, let me introduce myself :
I am 42 years old and a linux newcomer.

I followed this forum years ago when I boought the Goflex Net (GFN) but at that time I decided to use a custom linux called "Squeezeplug". It ran on USB thumbdrive (2 died, exhausted by the R/W cycle). it was a beat tricky to ssh and start the logitechmediaserver. then I moved to a RPI3 that has a best fit for use. The GFNwent to the drawer.

Now I would like to use the GFN to set a drive to host a shared folder ( samba ), to sync the file from the different device around the house (4PC + 2 tablet). Optionally run some module like "transmission". I would like that the GFN boots and runs without the need of the USB thumbdrive because they wear to o fast and are not reliable to run 24/7.
Thati swhy I wanted to refresh the GFN with latest UBOOT, ROOTFS and Rescue Sytem.


I tried to follow the tutorial : 2017.07 U-Boot Kirkwood - GoFlexNet, but at step 4 :
"Be sure there is no bad block in the first 1M of your NAND (check dmesg). This is very important, if there is bad block in the first 1M (8 blocks), don't flash u-boot, because you will almost certainly brick your box. Please post your question here (there is a work around for it). "

DMESG give this log output :( see attachment for detail)
Scanning device for bad blocks
Bad eraseblock 1 at 0x000000020000
Bad eraseblock 3 at 0x000000060000
Bad eraseblock 4 at 0x000000080000
Bad eraseblock 5 at 0x0000000a0000
Bad eraseblock 7 at 0x0000000e0000
Bad eraseblock 8 at 0x000000100000
Bad eraseblock 10 at 0x000000140000
Bad eraseblock 11 at 0x000000160000
Bad eraseblock 12 at 0x000000180000
Bad eraseblock 13 at 0x0000001a0000
Bad eraseblock 15 at 0x0000001e0000
Bad eraseblock 16 at 0x000000200000
Bad eraseblock 17 at 0x000000220000
Bad eraseblock 18 at 0x000000240000
Bad eraseblock 19 at 0x000000260000
Bad eraseblock 20 at 0x000000280000
Bad eraseblock 22 at 0x0000002c0000

thats stopped me .
I tried to implement the rescue system v.2 from Jeff : https://forum.doozan.com/read.php?4,3896
Unfortunately, I got this message :
--2019-10-15 16:06:56-- (try: 2) http://projects.doozan.com/uboot/valid-uboot.md5
Connecting to projects.doozan.com|45.77.104.33|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 1133 (1.1K)
Saving to: `/tmp/valid-uboot.md5'

100%[===========================================================================>] 1,133 --.-K/s in 0s

2019-10-15 16:06:56 (34.2 MB/s) - `/tmp/valid-uboot.md5' saved [1133/1133]

##
## uBoot is not up-to-date
## Please install the newest uBoot and then re-run this installer.
##

What do I do ? proceed with uboot anyway ? I fear to brick the GFN
I have logs from mount , fw_printenvs and uboot version available
Thanks for your support.

Lancer24,

> What do I do ? proceed with uboot anyway ? I fear
> to brick the GFN

You could brick it without being extra careful.There are a few bad blocks in u-boot area in NAND.

Do you have serial console? or willing to connect serial console? It would be safer to have serial console in case you need to recover from a bad flash.

Let me come back and show you how to flash the new u-boot.

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)

Lancer24,

OK so here is the situation. You have only 3 good blocks in the 1st MB of NAND: blocks 0, 2, 6.

So you cannot install the new u-boot with its envs image in mtd0 (the fist 1MB).

However, if you cannot use this box with stock u-boot, then still there is a way to install new u-boot, but it will be a lot more complicated.

If you want to do that, install serial console, and I will guide you how to flash new u-boot to NAND, and use the uEnv.txt (from the rootfs) to load the envs at each boot. See below for what you need to prepare.

=============

Either keep stock u-boot on this box, or install new u-boot.

I would strongly recommend that you install serial console, and then try UART booting.

An example for booting with UART running kwboot:
https://forum.doozan.com/read.php?3,51739,51919#msg-51919

Once you can run kwboot to load and run the new u-boot image, you'll have a unbrickable box. However, (very rare) some old GoFlex Net box can not be run with kwboot..

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)

Hello Bodhi,
Thanks for your support.

I ordered a NOKIA DKU-5 with its driver. I have now to find a bunch of connector.

A few question about the bad block :
- where are they located : in the internal memory of the GFN or in the USB thumbdrive ?
- is it physical damage or soft damage ?

Lancer24

Lancer,

> A few question about the bad block :
> - where are they located : in the internal memory
> of the GFN or in the USB thumbdrive ?

Neither.

It is in the GFN NAND chip. NAND is a non-volatile memory (flash memory). It is where stock u-boot and stock kernel and OS resides permanently, thus you can power down and they will remain intact.

> - is it physical damage or soft damage ?

It is physical damage. When a NAND block (128K) becomes bad, it is flaged as unusable in the NAND chip internal bad block table.

Sometime people would try to do NAND scrub, hoping that the bad block is alive again. But they won't last long that way.

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)

Take a look here for the correct header type (external link):

https://varkey.in/seagate-goflex-net-serial-connection/


If the serial port are holes, you can also plug in the serial module converter 3 wires, temporarily, without soldering.

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)



Edited 1 time(s). Last edit at 10/16/2019 03:56AM by bodhi.

OK,
Thanks for the information.
The DKU should be at home by the end of the week and the weather foreceast anounc rainy week-end. So, I will have plenty of time to "play" with it and serial console.

Do you have any clue of what happened to get so much bad block ?
Too many hard reboot ?
Bad procedure during squeezeplug hacking procedure ?

Lancer24

> Do you have any clue of what happened to get so
> much bad block ?
> Too many hard reboot ?
> Bad procedure during squeezeplug hacking procedure

Could be a bad NAND chip from factory. Or could be it was flashed too many times and worn out.

But I think most likely it is from factory, i.e. a bad batch of production. In this case, more bad blocks will show up later if you try to flash it repeatedly. So UART booting is really good to have (you can boot the box without NAND, in the worst case).

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)

hi,

Lancer24 Wrote:
-------------------------------------------------------
> if there is bad block in the first 1M (8 blocks),
> don't flash u-boot, because you will almost
> certainly brick your box. Please post your
> question here (there is a work around for it). "

i wonder if this workaround really is described nowhere, or if i just cant find it.
at the moment its just academic for me, but i want to know how to in case of...

jay

jay,

> > if there is bad block in the first 1M (8
> blocks),
> > don't flash u-boot, because you will almost
> > certainly brick your box. Please post your
> > question here (there is a work around for it).
> "
>
> i wonder if this workaround really is described
> nowhere, or if i just cant find it.
> at the moment its just academic for me, but i want
> to know how to in case of...
>

The basics was discussed here quite a few years ago (could have been 7 or 8 years). But I forgot to save the link!

The workaround is varied, depending on where the bad NAND block is, and which flavor of Kirkwood box. So that why I said "post the problem", because I am sure can solve any case of bad NAND block (or at least have a good solution for each case). I just did not want to write a long post that would cover all the scenarios.

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)