@addy75, Whatever software you have installed, do some research on this. It's not standard behavior. Perhaps what you've been seeing have something to do with this. [ 30.106761] softdog: Software Watchdog Timer: 0.08 initialized. soft_noboot=0 soft_margin=60 sec soft_panic=0 (nowayout=0)by bodhi - Debian
Hi Jan, Those pogoplugs were installed with Arch Linux ARM: /lib/modules/2.6.31.6_SMP_820 - MTD definition is done with u-boot envs. You can define them anyway you want and the kernel will pick it up as specified. The boot mtd is what need to be correctly specified, since that's where u-boot resides. The rest is up to us. Stock mtd2 contains rootfs because that's where the manuby bodhi - Off-Topic
Ah, I see. Well done :) And to start netconsole: setenv preboot 'run preboot_nc'by bodhi - uBoot
You are using old kernel? dt_sata_boot=ext2load ide 0:1 $uimage_addr /boot/uImage_2.6.31.14_OX820_1.2_shv.Pro; bootm $uimage_addrby bodhi - uBoot
Jan1973, I remember now. Your MTD layout is orginal or Arch (WarheadsSE version). We have new u-boot and kernel, which are most users in this forum use to install on Pogo V3 Pro/Classic: http://forum.doozan.com/read.php?2,16044 http://forum.doozan.com/read.php?3,16017,page=1 The MTD content is still basically the same, except that we've redefined it: from dev: size erasesizeby bodhi - Off-Topic
devinger, Remove the $ > setenv set_bootargs_usb 'setenv bootargs > console=$console root=$usb_root > rootdelay=$rootdelay rootfstype=$rootfstype > $mtdparts usb-storage.quirks=0x0bc2:0x3321:u' >by bodhi - uBoot
Cool! However, one more thing to keep in mind is: If you have multiple drives plugged in, then the boot drive is not guarantee to be 0: it could be 1:, because the way the Pogo E02 starts the drives, whichever come up first will be 0:, then 1:, … so on. A very good way (but not foolproof) is to plug that boot drive into the USB port right next to the Ethernet port. All things equal (speedby bodhi - Debian
BTW, I'm creating a new scan script that will be a little bit more resilient for multiple disks types. I was planning include it in the default envs for the next u-boot release.by bodhi - Debian
@Almaz, First, do you have 1 partition where your rootfs resides and no separate /boot partition? and do you have only one disk or multiple disks? What ever partition that contains /boot/uImage is the device being detected by usb_scan. If /boot is the 1st partition in the 1st disk then usb_device is 0:1 (if there is only one device). So if you change the /boot to sda2 then usb_device wilby bodhi - Debian
According to the above, you are running with 128M RAM, so it does look OK: KiB Mem: 121756 total, 119448 used, 2308 free, 16208 buffersby bodhi - uBoot
begrenzungsfaktor, As ingmar said, I did not include it in the kernel build for 3.16. So either you compile it or wait until 3.17 release, perhaps a week from now, time permitted.by bodhi - Debian
jan1973 Wrote: ------------------------------------------------------- > Hi, > > I have a bricked pogoplug pro. As far as I can > tell only the mtd2 rootfs has been corrupted, so > I'm looking for a nanddump with the orignal > pogoplug software on it. > > Thanks for your help! > > Jan It's been so long ago, I'm not sure I can find my mby bodhi - Off-Topic
addy75, See above: http://forum.doozan.com/read.php?2,14351,18279#msg-18279by bodhi - Debian
Almaz Wrote: ------------------------------------------------------- > I guess, it's impossible to boot from raid. > Before I start playing with uboot setting, is it > possible to boot from another partition such as > /dev/sda2? Yes, you can boot from any partition, providing the u-boot env is set correctly to boot that specific partition. The device in ext2load commandby bodhi - Debian
@Dale, > If I set any machid regardless of arcNumber it > will hang at uncompressing Kernel, I will see if > any machid's work, maybe this kernel doesn't > support or is not compiled to anything other than > OpenRD? Probably. If you can find the link to the source code to build VAMP, post here and I'll take a look. As right now, it is running fine, so thereby bodhi - uBoot
Dale, > Finallly, check TOP, very little CPU/Mem > usage, everyone said the V4 was no good for this, > I think it's running just fine :) Very nice :) > [ 0.000000] Machine: Marvell OpenRD Ultimate > Board This is the reason. It was not identified as Pogo V4. The kernel is running it as OpenRD plug. My u-boot build has a wrong machid inside the code, but becauby bodhi - uBoot
Robert, > Can you tell why "every second " boot says > > host 192.168.178.26 is alive > and sometimes it say not alive? Because you turned on setconsole (if_netconsole is the ping before netconsole start running) preboot=run if_netconsole start_netconsole Since you have serial console, there is no need for this, so you can temporarily turn it off (in case youby bodhi - uBoot
denexo, Regarding: http://forum.doozan.com/read.php?2,16044,17821#msg-17821 Is there any news about this problem, i.e. patch in 3.17?by bodhi - Debian
Robert, The problem is not the bad block. It is way out in higher NAND, not in mtd0 where u-boot is. This is the problem with mounting rootfs. Interrupt serial console at the count down and printenv Please post this u-boot envs listing here.by bodhi - uBoot
Dale, Please post these: dmesg cat /proc/cpuinfo ls /sys/class/leds/ -l I want to see if the kernel identifies the box as Pogo V4. If not then the LEDs can not be controlled correctly. However, there might be a work around if we can take a look at (where is the Git) the source tree used to build the Fedora kernel.by bodhi - uBoot
Instead of using $usb_customparams in set_bootargs_usb, put that quirk string in the env directly. See if it will take.by bodhi - uBoot
Almaz, In theory it should be possible. However, I'm not sure because I've never looked into it, since the E02 and other plugs are so low power (5W) and no noise, there is really no need to do WOL.by bodhi - Debian
Awesome :) Yes, please post your script, or instruction here so others looking for installing VAMP or just Fedora can find it.by bodhi - uBoot
WarheadsSE Wrote: ------------------------------------------------------- > I think something might make sense here: > u-boot disables watchdog when it starts, then, we > re-enable it just before starting the kernel. > > Why? In case the kernel hangs. > > Obviously, for we'd do this via a memory write, > instead of in c code for u-boot. Why? Because thenby bodhi - Debian
@Dale, The new u-boot is a drop-in replacement for davygravy's u-boot, which is what you have installed on Pogo E02. All envs are reusable as they are. Of course, each machine needs a different u-boot binary, but the code is common. As for kernel, I suspect there might be a different issue lurking, but you have not got to that point yet! Currently you are still at the rootfs mounting stagby bodhi - uBoot
iptester, Excellent! to set a good default envs set, you can either type them in one by one, or use my default envs script uboot.2014.07-tld-1.environment.scrip. Dowload the attached script binary in this post and follow the instruction to execute it: http://forum.doozan.com/read.php?3,18105,18124#msg-18124 And download the latest rootfs if you have not already done so, put it on eitherby bodhi - uBoot
addy75 Wrote: ------------------------------------------------------- > bodhi, > > ok. I though the dog was "killed" by the new > uboot.... > > Anyway - Thank you for the good work !!! New u-boot did! But it seems the older kernel version turned it back on. Install the latest kernel and try again.by bodhi - Debian
@WarheadsSE, I got the NSA325 DTS running OK on 3.17 now.by bodhi - Debian
addy75, You're not running the latest kernel (3.16.0-kirkwood-tld-3).by bodhi - Debian
iptester, Congrats! you have unbricked it. Wait …. > 39 % [......^C > # Did it complete loading? did you enter the Control-C here ? I hope you did :) Try again and let it boot until the u-boot banner appears and the countdown starts. Then interrupt the count down. At u-boot prompt, get the listing of the evironments: printenvby bodhi - uBoot