Bizhub C308/C368 firmware update problem [POTENTIAL SOLUTION]

[jotunn]

In case of machine hanging with two fins while firmware upgrade with USB memory - try to rename two files on your USB memory FW0002 folder:

- initramfs.gz.uboot to initra~1.ubo
and
- mb86s73eb.dtb to mb86s7~1.dtb

[emujo]

Can you go little further as to what this is doing, and why? Emujo

[jotunn]

OK

During firmware upgrade with some USB memories machine stops with two fins displayed on screen (and stays in this stage forever), so I started digging and found that machine can't find proper ramdisk image on USB memory:

[...]
reading FW0002/Image
5335136 bytes read
reading FW0002/initra~1.ubo
** Unable to read file FW0002/initra~1.ubo **
reading FW0002/mb86s7~1.dtb
28708 bytes read
Wrong Ramdisk Image Format
Ramdisk image is corrupt or invalid
u-boot> <- machine stops here

Environment variable for bootup process looks like:
bootcmd=fatload usb 0 80008000 FW0002/Image;fatload usb 0 82000000 FW0002/initra~1.ubo;fatload usb 0 81000000 FW0002/mb86s7~1.dtb;bootm 80008000 82000000 81000000

So someone in Konica Minolta has defined file names in old FAT16 style (SFN - short file names, 8.3 style), although all the files on USB memory has newer LFN names typical for FAT32.
In theory (and in most cases) it should work without problems, but in this case it doesn't. I don't know if this is a problem of buggy FAT implementation in U-Boot used in KM machines,
or a problem with FAT formatting in newer OS'es, but definitely files with long names such as initramfs.gz.uboot and mb86s73eb.dtb can't be used to boot as their SFN doesn't exist for U-Boot.
So making a copy of those files with short filenames in FAT16 style - initra~1.ubo and mb86s7~1.dtb can solve problems.

But all theory need a proof - I have only one machine in this moment, a machine with a broken SATA controller (HDD is unreadable) and I can't upgrade firmware (kernel panic because of non-existent HDD),
but after filename change my machine started to boot farther than described above (<- machine stops here).

So maybe someone can try this solution.

[DadO]

Can u try not to upload firmware and wait few day's for new release?

[jotunn]

You don't get it at all

This problem cannot be solved by other firmware version - it doesn't matter if you have G00-16, G00-54 or G00-84.
I thought that C224(e) series was picky about USB memories, but current machines are much worse in this case (for me only?) - memories that worked for C224 (and I could format them, delete, repartition and so on), for new machines doesn't work at all. Moreover, I had one working memory for C308 which stopped working after quick format under Windows 7 - formatting has changed something definitely, so it was logical for me that maybe it's a problem in my computer, but nothing could help - neither formatting in another computer with Windows 7 (even with special tools), nor formatting under different OS'es (Linux, MacOS X).

So the problem lies (mainly) in the machine - I think someone has just used buggy/not patched version of U-Boot while designing firmware for new machine line.

[tech51]

Just adding to this.
We have found that if you hold the stop button down too long after the black dot appears on the screen, the update will stop at the two fin stage.
Release the button as soon as the dot appears.
I know it doesn't sound likely but we've had two engineers on areas 150 miles apart report this back.
Cheers...

[toshibadale]

For 8 series machines you must create a folder in the root called FW0002 and put firmware in it. For 7 series it is FW0001. Must be FAT32 also. And these machines are picky about USB sticks, and all security functions on stick must be removed.

[gerryproulx]

I have tried multiple usb sticks and different firmwares to no result on a brand new c368 just sticks on 2 bars. I have tried changing the file names and get a firmware error so that doesn't work.

[jotunn]

What kind of error?

[habik]

There is a firmware bug. See the TABT1600559EN