• Home
  • Help
  • Search
  • Login
  • Register
  Show Posts
Pages: [1]
1  Linux Stuff / Kernel / Re: 2.6.30-rc2 uImage and kmods ipkg feed on: April 19, 2009, 11:31:21 AM
I had not flashed the image.   

It now gets past the point it was stuck at, but now I have a I get a Kernel panic:
Quote
rtc-mv rtc-mv: setting system clock to 2009-01-26 18:14:20 UTC (1232993660)
List of all partitions:
1f00            1024 mtdblock0 (driver?)
1f01            4096 mtdblock1 (driver?)
1f02          519168 mtdblock2 (driver?)
No filesystem could mount root, tried:  ext3 ext2 cramfs
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(31,2)
[<c0029640>] (unwind_backtrace+0x0/0xdc) from [<c0034a38>] (panic+0x58/0x124)
[<c0034a38>] (panic+0x58/0x124) from [<c0008e88>] (mount_block_root+0x1c4/0x204)
[<c0008e88>] (mount_block_root+0x1c4/0x204) from [<c00090d8>] (prepare_namespace+0x11c/0x174)
[<c00090d8>] (prepare_namespace+0x11c/0x174) from [<c0008884>] (kernel_init+0xb8/0xe4)
[<c0008884>] (kernel_init+0xb8/0xe4) from [<c0037838>] (do_exit+0x0/0x6cc)
[<c0037838>] (do_exit+0x0/0x6cc) from [<00000000>] (0x0)
2  Linux Stuff / Kernel / Re: 2.6.30-rc2 uImage and kmods ipkg feed on: April 19, 2009, 09:54:12 AM
yep, both are set
3  Linux Stuff / Kernel / Re: 2.6.30-rc2 uImage and kmods ipkg feed on: April 19, 2009, 12:19:33 AM
I already had the system booting from the Ubuntu distro from nand, and attempted to reflash the uImage by setting up the optware feed, and ipkg installing the kernel.

However, when I did this, the kernel-image_2.6.30-rc2-1_arm.ipk no longer existed, but was replaced with the newer kernel-image_2.6.30-rc2-2_arm.ipk


I'm guessing that this was the cause for the problem that developed:

After rebooting, and setting the two new parameters in uboot, my device no longer boots.
I get:
Quote
Starting kernel ...
Uncompressing Linux ................................................................
.............................................................................. done, boot
ing the kernel.

At this point, the device freezes and give no more output to the serial terminal.



Any suggestion?
4  General Category / Application ideas and development Q/A / Re: Hamachi VPN will not install on: April 15, 2009, 07:51:12 AM
I tried the Nokia770 version of Hamachi.  It installed the tuncfg correctly, and hamachi-init ran correctly.

However, attempting to run "hamachi start" failed.  It looks like its a problem with the tun device.

Quote
root@debian:~# hamachi start
26 18:08:16.911 [   0] [ 1364] tap: connect() failed 2 (No such file or directory)
root@debian:~# modprobe tun
FATAL: Could not load /lib/modules/2.6.22.18/modules.dep: No such file or directory
root@debian:~# ls -la /dev/net/tun
crw------- 1 root root 10, 200 Jan 26 18:04 /dev/net/tun
root@debian:~# cat /dev/net/tun
cat: /dev/net/tun: No such device

So, /dev/net/tun exists (i think), but does not seem to want to load.

I'm not sure if this matters or not, but not only does the module that modprobe fail to run not exist, but the /lib/modules directory is also completely empty.

Update:
I found that the module.dep file is created by running depmod.  It fails at first, but a mkdir /lib/modules/[version number here] allows it to complete. 

However, still have the problem of:
Quote
root@debian:~# modprobe tun
FATAL: Module tun not found.



Would this mean that the current kernel does not have the tun module?


5  General Category / Application ideas and development Q/A / Hamachi VPN will not install on: April 14, 2009, 10:46:19 PM
I have attempted to install Hamachi on my plug, but have not had any success with it.

I'm running the Ubuntu distribution from the Marvell site.


The instructions for installing it on a linux machine are quite simple, untar and make install, but for some reason this is failing. 
The Makefile uses strip (from gcc and/or binutils packages), but strip fails, giving the message:
Quote
strip: Unable to recognise the format of the input file '/sbin/tuncfg'

I have reinstalled the binutils package to make sure strip was properly installed, re-untared the package, attempted the make install again, manually copying and chmoding the fils (though, without striping), and even attempted porting the /sbin/tuncfg file from my Ubuntu laptop to see if that would work, but in every case, I will get the previous error, or get a message of:
Quote
-bash: /sbin/tuncfg: cannot execute binary file
when I attempt to run the executable myself.

As a bit of a newbie to Linux, I would not be surprised if I am missing something completely obvious.
Any ideas?
6  Linux Stuff / Kernel / Re: "Bad Magic Number" on: April 14, 2009, 12:04:53 PM
Yeah, I found that while looking at the thread on the uImage size.  The new uImage runs succesfully.
7  Linux Stuff / Kernel / Re: "Bad Magic Number" on: April 14, 2009, 09:12:12 AM
Thank you.
8  Linux Stuff / Kernel / "Bad Magic Number" on: April 13, 2009, 10:04:00 PM
I followed the instructions provided in the "SheevaPlug - Writing Jaunty Filesystem on the NAND flash" document, and have copied over the new uImage and the ubuntu.jffs2 file.

However, when I rebooted the device, I received the following:

Quote
NAND read: device 0 offset 0x100000, size 0x400000

Reading data from 0x4ff800 -- 100% complete.
 4194304 bytes read: OK
## Booting image at 00800000 ...
Bad Magic Number
Marvell>>

To be clear, that third to last line does correctly read "00800000", not "0x800000".  It stuck out to me, so I checked the bootargs and bootcmd, but they are correct.

While I nandwrote the uImage file, I did get a message allong the lines of:
Quote
image partially written
Success

Would this possibly be part of the problem?  or does anyone have any idea what could have gone wrong?
Pages: [1]