GuruPlug Server Plus REBOOT!!!

[anon98]

I ordered 4 GuruPlug Server Plus boxes because they had dual NIC interfaces.  However on all my devices the GuruPlug reboots as soon as I plug in the second network cable.  This happens at any stage of booting with any network cable.  I have tried 3 different kernels and corresponding root images.  I also updated u-boot on one of them.  Nothing seems to work!

These boxes were suppose to replace several Mikrotik RouterBoards but the stability issues are unacceptable.

Does anyone have any insight on a possible resolution?

UPDATE: PROBABLY FIXED

After trying seemingly everything I tried something stupid... I plugged it into a 10/100 switch.  Dual gigabit does not work, but at least one of ports being 10/100 seems to be stable so far.  Just happy it's not rebooting!

[mjain]

Please try this images for GuruPlug,
http://www.plugcomputer.org/index.php/us/resources/downloads?func=select&id=14

Please post log if you face any further issues.

Thanks
mj

[Blüto]

Please try this images for GuruPlug,
http://www.plugcomputer.org/index.php/us/resources/downloads?func=select&id=14

Why? they don't work. Have you tried them yourself? On an FC 11 box with Minicom installed, as per the readme?

# cd /home/sheevaplug/openocd.binaries/bin
# ./openocd -f target/board/sheevaplug.cfg -c init -c sheevaplug_reflash_uboot -c exit


Open On-Chip Debugger 1.0 (2009-03-08-08:14) svn:exported

BUGS? Read http://svn.berlios.de/svnroot/repos/openocd/trunk/BUGS

$URL: http://svn.berlios.de/svnroot/repos/openocd/trunk/src/openocd.c $
3000 kHz
dcc downloads are enabled
Error: unable to open ftdi device: device not found
Runtime error, file "command.c", line 456:
    You have new mail in /var/spool/mail/root

[root@localhost bin]#

[anon98]

I made the mistake of not ordering the JTAG module because I thought it was going to be like the SheevaPlug and be integrated.  I do have a serial connection from a custom cable I made which allows me to interact with the U-Boot prompt, but unable to run OpenOCD.

From the guruplug downloads page I flashed U-Boot 2010.03-01161-gd91b0a9 (posted May 10th) and kernel 2.6.33.2-00003-g4b82fc0 (May 10th) and the corresponding rootfs.ubi.img (May 11th) which all boots up and works fine if I never plug in the second network cable.  (Also I disabled loading the bluetooth and wireless drivers which has helped with the stability.)

Actually I take that back, a majority of the time if I power on the device with the network cable attached it fails to boot properly.  Only after it boots can I plug in the first network cable.

Just to be clear... there are no logs, perhaps since I don't have the JTAG.  However the LEDs on the NICs blink quickly for a second or two in sync and then the device hard resets and continues to reset every few seconds until I unplug one of the cables.  This happens regardless of the state, e.g. U-Boot environment, booted into Linux, freshly plugged in, one cable before the other, etc.

[anon98]

Small update, I just tried sheeva.with-linux.com 2.6.33.4 (May 12th) with no luck.  Boots fine, albeit a little slow, but still reboots with 2 network cables.

[mjain]

Why? they don't work. Have you tried them yourself? On an FC 11 box with Minicom installed, as per the readme?

# cd /home/sheevaplug/openocd.binaries/bin
# ./openocd -f target/board/sheevaplug.cfg -c init -c sheevaplug_reflash_uboot -c exit

Yes. I have flashed U-Boot using GuruPlug-Installer and also kernel, rootfs(ubi) image as per this page,
http://www.openplug.org/plugwiki/index.php/Reflashing_images_on_the_GuruPlug
I used  GuruPlug-Installer from above page and after flashing U-boot, just wrote environment variables.
Make sure you have connected JTAG board and appropriate cables.

Here is my log,
guruplug-installer]# ./runme.sh /tftpboot/uboot.guruplug.bin
 ****     Preparing environment variables file ...
  ****   Burning uboot and environment variables ... This will take few
  minutes ...
  Open On-Chip Debugger 0.2.0-in-development (2009-05-17-10:32) svn:1800M


  BUGS? Read http://svn.berlios.de/svnroot/repos/openocd/trunk/BUGS


  $URL: http://svn.berlios.de/svnroot/repos/openocd/trunk/src/openocd.c $
  2000 kHz
  dcc downloads are enabled
  Info : JTAG tap: feroceon.cpu tap/device found: 0x20a023d3 (Manufacturer:
  0x1e9, Part: 0x0a02, Version: 0x2)
  Info : JTAG Tap/device matched
  Error: unknown EmbeddedICE version (comms ctrl: 0x00000018)
  Warn : no telnet port specified, using default port 4444
  Warn : no gdb port specified, using default port 3333
  Warn : no tcl port specified, using default port 6666
  target state: halted
  target halted in ARM state due to debug-request, current mode: Supervisor
  cpsr: 0x000000d3 pc: 0xffff0000
  MMU: disabled, D-Cache: disabled, I-Cache: disabled
  0 0 1 0: 00052078
  NAND flash device 'NAND 512MiB 3,3V 8-bit' found
  successfully erased blocks 0 to 2 on NAND flash device 'NAND 512MiB 3,3V
  8-bit'
  wrote file openocd/uboot.bin to NAND flash 0 up to offset 0x0002c000 in
  27.543037s
  target state: halted
  target halted in ARM state due to debug-request, current mode: Supervisor
  cpsr: 0x000000d3 pc: 0xffff0000
  MMU: disabled, D-Cache: disabled, I-Cache: disabled
  0 0 1 0: 00052078
  NAND flash device 'NAND 512MiB 3,3V 8-bit' found
  successfully erased blocks 2 to 4 on NAND flash device 'NAND 512MiB 3,3V
  8-bit'
  wrote file openocd/uboot-env.bin to NAND flash 0 up to offset 0x00060000 in
  19.960396s
   ****   U-boot should be up and running now. Open your console ...

[TheDOC]

I've got the same issue. If I plug in the LAN cable to the second port before booting the plug, it stops booting or reboots after the Bluetooth: vendor=... message.

I'm using the latest u-boot, kernel and rootfs from http://plugcomputer.org/index.php/us/resources/downloads?func=select&id=14

If i plug in the cable after the guruplug is up and running, everything works fine (at least for the minutes, that I've testet it, after reading this topic).

[Blüto]

Why? they don't work. Have you tried them yourself? On an FC 11 box with Minicom installed, as per the readme?

Yes. I have flashed U-Boot using GuruPlug-Installer and also kernel, rootfs(ubi) image as per this page,
http://www.openplug.org/plugwiki/index.php/Reflashing_images_on_the_GuruPlug

Okay, so in other words, no, you didn't. The guruplug.readme.txt requires you to have a Fedora Core 11 box on which to unpack the Guruplug.installer tarball into the /home directory. The tarball contains, among other things, the Ftdi drivers to set up ttyUSB0 (JTAG) and ttyUSB1 (console) necessary to communicate with the GuruBrick.

The Guruplug.installer on  the wiki is an entirely different item, one that (presumably) doesn't require FC 11 so that it can be run on any host, and that begs the question What Ftdi drivers are you using and where did you obtain them, because the link to Ftdi_sio.ko published in GuruPlug Server Quick-Start Guide is dead, and you couldn't have used the JTAG interface to flash the plug without them?

But I couldn't get that installer to run, either. Not on the FC 11 box and not on an Ubuntu box. The main difference between the two is that ttyUSB0 and ttyUSB1 show up on the Fedora box, but only ttyUSB0 shows up on Ubuntu (using the Ftdi_sio.ko module provided with that kernel).

I was able to flash the recommended uboot.guruplug.bin using both bubt (from an earlier U-Boot upgrade attempt) and USB stick, e.g.

usb start
fatload usb 0:1 0x0800000 uboot.guruplug.bin
nand erase 0x0 0xa0000
nand write 0x0800000 0x0 0xa0000

I followed the rest of the instructions as listed at Reflashing images on the GuruPlug, and here is my output:

U-Boot 2010.03-01161-gd91b0a9 (Apr 22 2010 - 03:24:41)
Marvell-GuruPlug

SoC:   Kirkwood 88F6281_A0
DRAM:  512 MB
NAND:  512 MiB
*** Warning - bad CRC or NAND, using default environment

In:    serial
Out:   serial
Err:   serial
Net:   egiga0, egiga1
88E1121 Initialized on egiga0
88E1121 Initialized on egiga1
Hit any key to stop autoboot:  0
(Re)start USB...
USB:   Register 10011 NbrPorts 1
USB EHCI 1.00
scanning bus for devices... 4 USB Device(s) found
       scanning bus for storage devices... Device NOT ready
   Request Sense returned 02 3A 00
2 Storage Device(s) found

NAND read: device 0 offset 0x100000, size 0x400000
NAND read from offset 100000 failed -74
 4194304 bytes read: ERROR[/color]
## Booting kernel from Legacy Image at 06400000 ...
   Image Name:   Linux-2.6.33.2-00003-g4b82fc0
   Image Type:   ARM Linux Kernel Image (uncompressed)
   Data Size:    2688524 Bytes =  2.6 MB
   Load Address: 00008000
   Entry Point:  00008000
   Verifying Checksum ... Bad Data CRC
ERROR: can't get kernel image!
Marvell>>

Any idea what's happening here? Anyone? Bueller?

Update -- cbxbiker61's 2.6.33.4 kernel boots the GuruBrick, so again I have to ask: Have you tried these files yourself?

[TheDOC]

Anyone knows the best way to contact globalscale? Since this is a major issue on the plugs, they should give a statement on that, shouldn't they?

[MoDaX]

eth0 - 100Mbit
eth1 - 1000Mbit

It still crashes/resets/reboots. I narrowed this down to /etc/init.d/bluetooth which is started from /root/init_setup.sh. If you comment that line out in /root/init_setup.sh, reboots should stop. I believe this reset is due to kernel bug somewhere in bluetooth/eth drivers or their interaction. I will continue digging deeper into /etc/init.d/bluetooth but I'm already very confident that I will track this down to some modprobe or the like.

The general impression of /root/init_setup.sh is that it's of very poor quality and full of hacks rather than proper fixes.

[viorel]

do you succeed to boot on G-Switch?
my uboot version is : U-Boot 2010.03-01161-gd91b0a9 (Apr 22 2010 - 03:24:41)

on 100Mb boot well, but no 1000M D-Link switch.

[swaery]

I am having the exact same issue as the GuruPlug simply reboots whenever I plug in the second network table. I am not sure whether the problem is with the kernels since I have tried more than three already. Also please tell whether this has to do anything with the JTAG module?

[pietsnot]

I am having the exact same issue as the GuruPlug simply reboots whenever I plug in the second network table. I am not sure whether the problem is with the kernels since I have tried more than three already. Also please tell whether this has to do anything with the JTAG module?

Notice:  u-boot and uImage/kernel they do have dependencies,

i have 2xgps+  and don't have that problem
using:
eth1 = wan 100mbit
eth0 = lan 100mbit
jtag (is always enabled/connected)

at first flash this great u-boot:
http://plugapps.com/os/guruplug/

do set you uboot-params correctly

retry to boot

if that still doesn't work

read:
http://oinkzwurgl.org/guruplug_kernel

flash:
http://oinkzwurgl.org/dl.php?file=guruplug-linux-2.6.33.3flipflip.tar.gz
guruplug-linux-2.6.33.3flipflip/uImage

or
http://sheeva.with-linux.com/sheeva/2.6.35.10/sheeva-2.6.35.10-uImage