Show all posts by user

Re: Linux Kernel 3.12.0 Kirkwood package and rootfs (Non Flattened Device Tree) - 11 years ago

You're the first to get the NSA325 to boot on Debian 3.12.x that I knew of :) Awesome! The key point is setenv bootcmd_linux 'setenv bootargs $(bootargs_linux); usb reset; ext2load usb 0:1 $(loadaddr) /boot/uImage; ext2load usb 0:1 0x01100000 /boot/uInitrd; bootm $(loadaddr) 0x01100000' So you can adjust this address if uBoot ever have problem loading the kernel and initrd

by bodhi - Debian

Re: Linux Kernel 3.12.0 Kirkwood package and rootfs (Non Flattened Device Tree) - 11 years ago

Rootfs Debian-3.12.0-kirkwood-tld-3 was uploaded. Please see 1st post for download link.

by bodhi - Debian

Re: Linux Kernel 3.12.0 Kirkwood package and rootfs (Non Flattened Device Tree) - 11 years ago

Buttzy, Did you check to see if the NSA has UART booting (bootROM v 1.21): http://forum.doozan.com/read.php?3,7852 The nanddump could just be: nanddump -nof uboot.mtd0.bak /dev/mtd0 Or you could use Davygravy's image, too in this thread http://forum.doozan.com/read.php?2,7806 This command is correct (if the image is ready-for-NAND format, then -p must be used for kwboot to

by bodhi - Debian

Re: Linux Kernel 3.12.0 Kirkwood package and rootfs (Non Flattened Device Tree) - 11 years ago

Buttzy, Do you have serial console connected? it's best to do that before you play with uBoot envs. The stock uBoot does not have netconsole. That loadaddr I believe is where stock uBoot load the kernel uImage (that's how I understood WarheadsSE set it up so users can restore to stock easily). I don't have the NSA325, so I can help you but it must be done with serial console.

by bodhi - Debian

Re: U-Boot loading compressed uImage - 11 years ago

ricke Wrote: ------------------------------------------------------- > What compression does u-boot 2013.11 support for > uImage if any?, i tried lzma2 and it didn't boot. I dont think 2013.11 has been released yet! Which uBoot are you running?

by bodhi - uBoot

Re: Linux Kernel 3.12.0 Kirkwood package and rootfs (Non Flattened Device Tree) - 11 years ago

Buttzy, I will upload a new rootfs tonight. Your problem is still in uBoot. Nothing about kernel yet. Check the arcNumber for NSA325. It should match the machid.

by bodhi - Debian

Re: After Debian install, it boots back to PogoPlug, E02 - 11 years ago

Almaz, The dialog is just another package that usually used whenever you install a package, missing it does not hurt anything, because the apt-get falls back to teletype. I did not install it since I expect users to install it when they see the errors (will do that for the next release to keep it simpler, it looks like a problem but it is not). So: apt-get install dialog should fix it.

by bodhi - Debian

Re: After Debian install, it boots back to PogoPlug, E02 - 11 years ago

Almaz Wrote: ------------------------------------------------------- > Problem solved. UBOOT env were set to EXT3, where > Debian script automatically format the drive to > EXT2. I'm just amazed no one even posted anything > about it. Right after you run Jeff script, uBoot env is set to Ext2, and the drive is set to Ext2. The mismatched usbrootfs type problem is very

by bodhi - Debian

Re: Rescue V2.8.2 : many enhancements + full kernel support for GoFlexNet/Home, Zyxel NSA3x0 and many others - 11 years ago

No this rescue system won't work for Pogo Pro. For Debian on Pogo Pro, see this thread: http://forum.doozan.com/read.php?3,6336

by bodhi - Rescue System

Re: Rescue V3.0 : A work in progress ... wanna help? - 11 years ago

sylvester Wrote: ------------------------------------------------------- > Oh... checking through the script again and I > think I found where I went wrong. I didn't > notice the last couple of output line of the uboot > install script (which is part of > kirkwood.debian-wheezy.sh) to do fw_setenv > arcnumber and the debian installation script just > started r

by bodhi - Rescue System

Re: Rescue V3.0 : A work in progress ... wanna help? - 11 years ago

Yes. arcNumber needs to be set to the GoFlex Home number. The script does not set arcNumber. Because it has no way of knowing if you have a kernel that supports it. Main line kernel does not have the patch for GF Home. Hence, the purpose of my kernel and rootfs builds.

by bodhi - Rescue System

Re: Debian on NSA320 in stalled only on USB - 11 years ago

canotto Wrote: ------------------------------------------------------- > Hi to everyone ! > > > I'm new on this forum and 'ive read many post on > how to install > debian on this little box > > I'd like to try debian installed on usb without > touching the original firmware > northe hard disk content but i'm little confused > on

by bodhi - Debian

Re: After Debian install, it boots back to PogoPlug, E02 - 11 years ago

Mount the stick on another box and examine /boot directory. uImage and uInitrd are missing from that directory then Squeeze installation was not successful.

by bodhi - Debian

Re: After Debian install, it boots back to PogoPlug, E02 - 11 years ago

Don't need to use gparted. Use fdisk to create partition, and e2fsck to format as ext3. Just a precaution to make sure that it is MBR partition. If you can use fdisk to list partition, and see that it is not GPT, then it's fine. Sometime you need to power down and cold boot, instead of reboot to see if Squueze was installed OK.

by bodhi - Debian

Re: After Debian install, it boots back to PogoPlug, E02 - 11 years ago

You should install Squeeze and and then upgrade to Wheezy: http://projects.doozan.com/debian/ then change /etc/apt/sources.list to point to wheezy, and apt-get update apt-get upgrade Or use this ready-made rootfs: http://forum.doozan.com/read.php?2,12096

by bodhi - Debian

Re: 2013.10 U-Boot for FDT and Non-FDT Kernel - GoFlex Net - 11 years ago

I'd like to hear any feedback on these uBoot images! did it work? … problem?

by bodhi - uBoot

Re: Rescue V3.0 : A work in progress ... wanna help? - 11 years ago

sylvester, My kernel build has the patch for the "too few good blocks.." already incorporated for GoFlex Home. You could post your dmesg here so we can take a look. Also post your cat /proc/mtd cat /proc/cpuinfo About the md5 from nanddump, did you get mtdparts definition printed out after installation? is it different now or the same?

by bodhi - Rescue System

Re: Rescue V3.0 : A work in progress ... wanna help? - 11 years ago

Use this latest rootfs: http://forum.doozan.com/read.php?2,12096

by bodhi - Rescue System

Re: Can boot, but cannot reboot - 11 years ago

Roman, I've overlooked that you have rescue system installed! so to really test the reset, the command would have to be: setenv bootcmd 'usb start; usb start; usb start; run force_rescue_bootcmd; run ubifs_bootcmd; run usb_bootcmd; usb stop; reset' In any case, it is just to see if we have more one alternative workaround. Try only if you feel like testing it!

by bodhi - uBoot

Re: 2013.10 U-Boot for FDT and Non-FDT Kernel - GoFlex Net - 11 years ago

I've uploaded the 2013.10 uBoot image for Pogo E02. Please see 1st post for download link.

by bodhi - uBoot

Re: Can boot, but cannot reboot - 11 years ago

RomanG, I have another idea to work around this problem. I have not scripted a nice sequence of commands to test it. But perhaps you can try it manually in serial or net console. Reboot, interrupt your console and change the bootcmd to take out the run pogo_bootcmd part: setenv bootcmd 'usb start; usb start; usb start; run force_rescue_bootcmd; run ubifs_bootcmd; run usb_bootcmd;

by bodhi - uBoot

Re: Pogoplug E02 Bad Block - 11 years ago

musselmanb > Is the dump extractable? It could be. We don't know exactly how much data to dump, without mounting it. But if you would want to recover from only a handful of bad blocks, then I would dump 32MB - 1 block, 32MB - 4 blocks and so on. See if that will be ok to flash it back. Or you could try mounting it, copy the content with cp -aR to a USB thumb, and then repack the

by bodhi - uBoot

Re: Pogoplug E02 Bad Block - 11 years ago

Congrat! I am glad it worked out for you! Hope my usage of the block numbers did not confuse anybody reading this. When I said block 1 to 4 above, I meant 1st block to 4th block. Very possible the reason it was corrupted because of fw_setenv. It would write a different checksum to 0xa0000, and blparam expects different checksum. Usually the checksum got rebuilt during reboot a second t

by bodhi - uBoot

Re: some questions ... before installing - 11 years ago

Kristoff, See the updated 1st post in the orginal working thread for links. Actually people are working on the latest Debian/Arch kernel!

by bodhi - Debian

Re: Pogoplug E02 Bad Block - 11 years ago

musselmanb , I think you're almost out of option here. The Pogo E02 does not have UART booting, so if we flash around the bad block without using block 5 (I had an idea how to do that), and it does not work, you can't recover without JTAG. Here what I was thinking, but please do not attempt this if you have no way of recovering :) The stock uBoot size must have been 3 blocks

by bodhi - uBoot

Re: Install testing from the very beginning? - 11 years ago

tdd, These little boxes will do all of the above, pretty efficiently. I would move all those chores to the Pogo E02(s) without hesitation. Here is a very good thread we have recently about serial console: http://forum.doozan.com/read.php?8,13263 And you probably already knew this: it is pefectly OK to have a USB thumb for rootfs formatted as Ext3, if you move all your logging to /tmp.

by bodhi - Debian

Re: Pogoplug E02 Bad Block - 11 years ago

musselmanb, Your theory is correct :) because your bad block was on the first 4 blocks (I did not notice if you've posted the output of flashing uBoot on bad block). Original uBoot uses 0xa0000 for the envs. Jeff/Davy's uBoot uses 0xc0000. The reason was so that original uBoot envs would stay intact so it could be used in booting Pogo OS. But because of bad block, we wiped out block

by bodhi - uBoot

Re: Install testing from the very beginning? - 11 years ago

tdd, - The very first thing you should do is to setup netconsole, if you don't have serial console. netconsole will let you recover from a bad rootfs to a certain degree (till before the kernel is loaded), and hopefully to observe more of kernel booting. http://forum.doozan.com/read.php?3,14 - /dev/sda1 does not need to be in fstab. You have /dev/root. That's the rootfs where you c

by bodhi - Debian

Re: Pogoplug E02 Bad Block - 11 years ago

In pogo OS, what is your output of cat /etc/fw_env.config does it show 0xa0000 as the address? If you have use fw_setenv in Pogo OS. Then continue using fw_setenv, don't use blparam. But if you have not, then use blparam. Because they have different checksum.

by bodhi - uBoot

Re: Wheezy for Pogoplug B01? - 11 years ago

twinclouds, If you are running Debian Squeeze rootfs on this OXNAS B01, you can upgrade the rootfs to Wheezy. But udev will cause problem, so it will need to be hold back to the current version you have. t's very simple as running a command: http://forum.doozan.com/read.php?3,6336,13058#msg-13058 So: - hold udev - changing apt/source.list to wheezy - apt-get update - apt-get upgr

by bodhi - Debian