[Solved] Peculiar USB-Stick: read=y write=n and more

Soukyuu · 2014-05-05 17:48:44

Solved by low level formatting with "Phison Low Level Formatter v 2.9.0.0" (Russian site)

------------- Original Message ---------------

It's not really a arch-specific issue. I can't access the contents of the stick on windows at all, it sees the partition but hangs when I try to open it or start chkdsk/diskpart etc.
On arch, I can list the partitions with the following result:

Disk /dev/sdc: 1.9 GiB, 2062548992 bytes, 4028416 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0x6f20736b
This doesn't look like a partition table. Probably you selected the wrong device.

Device    Boot      Start        End    Blocks  Id System
/dev/sdc1 ?     778135908 1919645538 570754815+ 72 Unknown
/dev/sdc2 ?     168689522 2104717761 968014120  65 Novell Netware 386
/dev/sdc3 ?    1869881465 3805909656 968014096  79 Unknown
/dev/sdc4 ?    2885681152 2885736650     27749+  d Unknown

Partition table entries are not in disk order.

That's not really what the partition table is supposed to be looking like, it should be one fat32 partition, which is also what I see in dolphin.
Writing a new partition table with fdisk results in a

/dev/sdc: close device failed: Input/output error

Interestingly, I can run badblocks and no errors are reported. The LED on the stick is blinking all the time.

badblocks -s /dev/sdc
Checking for bad blocks (read-only test): done

I have tried doing a fsck -r on it, and it finds some issues. When I answer "yes" to correct an issue, it suddenly says "leaving filesystem unchanged" and aborts. There seem to be 2 FATs on the device, the above happens whether I select the first or the second FAT.
I have also tried using dd if=/dev/zero of=/dev/sdc. It starts writing but seems to hang after writing 2MB of data. That's also the moment the LED on the stick stops blinking.

In summary, it seems read operations, at least on linux, are working. Any sort of write operation fail though. The stick doesn't have a physical "read only"-switch, so I'm half-expecting to be told it's busted. Is there anything else I can try doing to completely zero it out? I don't care about the data on it as I was able to read out everything from it.

Last edited by Soukyuu (2014-05-06 15:49:27)

Inxsible · 2014-05-05 17:59:02

maybe it is write protected kinda like the SD cards that have a slider to make it write protected.

Soukyuu · 2014-05-05 18:11:01

Nope, it doesn't. I even opened it up so all I have is the USB connector and the chip attached to it. No switches/sliders. The contacts between the usb connector and the chip seem intact as well.

kokoko3k · 2014-05-05 18:27:19

I don't think read operation on linux are working.
Try to make a complete dump of the stick on a file using dd.
meanwhile watch out for related kernel messages...

Soukyuu · 2014-05-05 19:59:20

The data I copied from it (mostly text files) are openable and not corrupted. Reading out with dd made my VM freeze, but on a physical machine, I was able to read it out multiple times. No related kernel messages in journalctl -f
The physical machine can't perform any write operations on it either, though, same behavior as on the VM.

edit: got a few messages after a few tries:

5月 05 21:56:53 dynabook kernel: sd 2:0:0:0: [sdb] Unhandled sense code
 5月 05 21:56:53 dynabook kernel: sd 2:0:0:0: [sdb]  
 5月 05 21:56:53 dynabook kernel: Result: hostbyte=0x00 driverbyte=0x08
 5月 05 21:56:53 dynabook kernel: sd 2:0:0:0: [sdb]  
 5月 05 21:56:53 dynabook kernel: Sense Key : 0x7 [current] 
 5月 05 21:56:53 dynabook kernel: sd 2:0:0:0: [sdb]  
 5月 05 21:56:53 dynabook kernel: ASC=0x27 ASCQ=0x0
 5月 05 21:56:53 dynabook kernel: sd 2:0:0:0: [sdb] CDB: 
 5月 05 21:56:53 dynabook kernel: cdb[0]=0x2a: 2a 00 00 00 00 00 00 00 01 00
 5月 05 21:56:53 dynabook kernel: end_request: critical target error, dev sdb, sector 0
 5月 05 21:56:53 dynabook kernel: Buffer I/O error on device sdb, logical block 0
 5月 05 21:56:53 dynabook kernel: lost page write due to I/O error on sdb
 5月 05 21:56:53 dynabook sudo[915]: pam_unix(sudo:session): session closed for user root
 5月 05 21:57:05 dynabook sudo[937]: haruyo : TTY=pts/0 ; PWD=/home/haruyo ; USER=root ; COMMAND=/usr/bin/fdisk /dev/sdb
 5月 05 21:57:05 dynabook sudo[937]: pam_unix(sudo:session): session opened for user root by haruyo(uid=0)
 5月 05 21:57:09 dynabook logger[944]: ACPI action undefined: PNP0C0A:00
 5月 05 21:57:13 dynabook kernel: sd 2:0:0:0: [sdb] Unhandled sense code
 5月 05 21:57:13 dynabook kernel: sd 2:0:0:0: [sdb]  
 5月 05 21:57:13 dynabook kernel: Result: hostbyte=0x00 driverbyte=0x08
 5月 05 21:57:13 dynabook kernel: sd 2:0:0:0: [sdb]  
 5月 05 21:57:13 dynabook kernel: Sense Key : 0x7 [current] 
 5月 05 21:57:13 dynabook kernel: sd 2:0:0:0: [sdb]  
 5月 05 21:57:13 dynabook kernel: ASC=0x27 ASCQ=0x0
 5月 05 21:57:13 dynabook kernel: sd 2:0:0:0: [sdb] CDB: 
 5月 05 21:57:13 dynabook kernel: cdb[0]=0x2a: 2a 00 00 00 00 00 00 00 01 00
 5月 05 21:57:13 dynabook kernel: end_request: critical target error, dev sdb, sector 0
 5月 05 21:57:13 dynabook kernel: Buffer I/O error on device sdb, logical block 0
 5月 05 21:57:13 dynabook kernel: lost page write due to I/O error on sdb

I guess that's a sign it's dying?

Last edited by Soukyuu (2014-05-05 20:04:05)

emeres · 2014-05-06 12:49:52

Soukyuu wrote:

The stick doesn't have a physical "read only"-switch, so I'm half-expecting to be told it's busted. Is there anything else I can try doing to completely zero it out? I don't care about the data on it as I was able to read out everything from it.

Soukyuu wrote:

Nope, it doesn't. I even opened it up so all I have is the USB connector and the chip attached to it. No switches/sliders. The contacts between the usb connector and the chip seem intact as well.

What IC is on the stick? Write everything down and upload an image if you can.

Soukyuu · 2014-05-06 13:18:02

Here you go:
Side A
Side B

edit: it's actually "eh FL2", the e is very small, had to take a magnifying glass to read it.

Last edited by Soukyuu (2014-05-06 14:56:03)

emeres · 2014-05-06 13:41:58

The most interesting part is under the sticker. That is the flash memory, PS2231 is just a usb 2.0 flash controller.

Soukyuu · 2014-05-06 14:02:15

Under the sticker:
Toshiba
TH58NVG4D4CXLV3
P31977
Japan
0644KAD

edit: i'll try to make a photo, but the text is pretty unreadable.

Last edited by Soukyuu (2014-05-06 14:53:18)

emeres · 2014-05-06 14:49:25

The objective was to find the datasheet for this flash NAND. They do have certain norms and standards, but sometimes even different revision have different means of handling data. You should check its write cycles if you figure out its model (I could not find it), you may have used that up.
You could try writing zeros directly though the controller, its datasheet will be helpful for that. Hand clocking the controller, setting operation mode and address and finally clocking (preferably using a square function generator) either zeros or ones as data input. Use an isolated environment for this, meaning do not attend this using your pc usb port, since you may short-circuit the device and damage your board.
Another possibility is the controllers code got corrupted, there are some "utilities" to recover firmware/format on low level/etc. (usually russian made and windows only), you might want to try that.
Probably not worth the whole work though. If you want a "secure wipe" destroy the flash ic.

PS. Not sure the admin will be happy about uploading images here directly.

Soukyuu · 2014-05-06 14:55:27

Pretty hard to see anything on these: Try A | Try B | Try C
I also found this and was able to correct the number following that guide:

Toshiba
TH58NVG4D4CXLV3
P31977
Japan
0644KAD

@images: as far as I understood, direct displaying of images was only a problem if they're very big, either size or resolution wise. I replaced them with links.

Soukyuu · 2014-05-06 15:45:24

Solved! Thanks for suggesting me to search for a low level formatter. Was able to find one that worked here (namely the "Phison Low Level Formatter"). After following the instructions the USB drive is now formatted, accessible and writeable.

emeres · 2014-05-06 16:06:16

Good, but be careful with storing sensible data on it. You never know.

Arch Linux

#1 2014-05-05 17:48:44

[Solved] Peculiar USB-Stick: read=y write=n and more

#2 2014-05-05 17:59:02

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#3 2014-05-05 18:11:01

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#4 2014-05-05 18:27:19

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#5 2014-05-05 19:59:20

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#6 2014-05-06 12:49:52

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#7 2014-05-06 13:18:02

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#8 2014-05-06 13:41:58

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#9 2014-05-06 14:02:15

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#10 2014-05-06 14:49:25

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#11 2014-05-06 14:55:27

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#12 2014-05-06 15:45:24

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

#13 2014-05-06 16:06:16

Re: [Solved] Peculiar USB-Stick: read=y write=n and more

Board footer