Jump to content

Recommended Posts

Posted

Hi again,

 

I've a strange problem again.

 

I've several Lime2 and tested them with USB-Sticks, works fine.

 

Now I got a Fan-Tec case (USB 3.0 DB-ALU3-6G) containing a 4TB-SATA from a customer.

 

As soon as I plug the disk into one of the USB-Plugs of a lime2, I got a kernel panic:

 

Message from syslogd@localhost at Oct 19 17:49:20 ... kernel:[  460.550693] Internal error: Oops: 17 [#1] PREEMPT SMP ARM
Message from syslogd@localhost at Oct 19 17:49:20 ... kernel:[  461.304238] Process kworker/u:1 (pid: 55, stack limit = 0xee9702f0)
Message from syslogd@localhost at Oct 19 17:49:20 ... kernel:[  461.314387] Stack: (0xee971d78 to 0xee972000)

....

 

Some facts:

* USB-Sticks are no problem at all

* No difference if formatted or not (2* FAT or 1 NTFS)

* Kernel 3.105, 3.108, 3.109 crashes --> original images

* Disabled any found *OTG and *USB-Gadgets Kernel parameter --> no success --> build with Igors script :)

* Kernel 4.2.3-sunxi seems to work (but can't read, probably because its not a MBR), at least it dosn't crash
* Lime2 with nand

 

Unfortunatelly, 4.2.3 is not an option for me because NAND is not supported and I can't install a sd on every board.

 

Found only this: https://github.com/raspberrypi/linux/issues/268, but couldn't follow completely.

 

Any thoughs or hints? Or how can I get the nand to work on 4.2.3?

 

Thanks for any help...

Posted

The original image from olimex (Linux a20-olimex 3.4.90+ #11 SMP PREEMPT Wed Aug 20 08:20:32 EEST 2014 armv7l GNU/Linux) crashes too...

Tested also different boards.

 

Do normal USB-Hard disks work generally? (May it's the chipset, maybe 4TB, maybe all USB-Disks?)

Posted

The only issue regarding USB (that i remember now) is related to OTG port. Are you using it?

Posted

Hello Igor,

 

no, I'm using only the two normal ports at the front side. Disabled also nearly anything in the kernel except EHCI, OHCI Allwinnner E/OHCI and USB Mass Storage...

 

I will check tomorrow, if another harddisk will work in this fantec case.

Posted

Check a script.bin and play with those settings too. Kernel might be just fine.

Posted

Checked:

 

- Crash: script: disabled OTG-Port usbc0: crash

- Don't work: script: set host_init_state=0: no usb at all

- Works: used a 120GB HDD inside the fantec case

 

So I think there is a 2TB barrier somewhere inside the kernel. But CONFIG_LBDAF=y seems to be set according to config-file.

Posted

Have no idea why is this happening :(

 

Better wait for NAND support on mainline kernel. I heard it's already done but not implemented yet.

 

If problem happen there it won't be a problem to fix it since there are a lot of people working on it. On the other hand no one is taking care on legacy kernel any more ... except minor things.

 

I am using 3T but on SATA and it works fine :) With Lime2 and legacy kernel.

Posted

Ok, 3.4.110 too :(

 

Message from syslogd@localhost at Jan  1 21:38:52 ...
 kernel:[  556.108806] Internal error: Oops: 17 [#1] PREEMPT SMP ARM

Message from syslogd@localhost at Jan  1 21:38:52 ...
 kernel:[  556.839929] Process kworker/u:0 (pid: 5, stack limit = 0xef05a2f0)

Message from syslogd@localhost at Jan  1 21:38:52 ...
 kernel:[  556.849870] Stack: (0xef05bd78 to 0xef05c000)
 

Guest
This topic is now closed to further replies.
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines