All PLX based chips - Important Upgrade Note [updated]

[WarheadsSE]

updater is updated w/ kernel that automounts devtmpfs to /dev. I have also uploaded an up-to-date as of 30 minutes ago rootfs, that I have *not yet* moved to "latest" staus.

I am going to bed!

[pogoman]

$this->bbcode_second_pass_quote('WarheadsSE', 'u')pdater is updated w/ kernel that automounts devtmpfs to /dev. I have also uploaded an up-to-date as of 30 minutes ago rootfs, that I have *not yet* moved to "latest" staus.

I am going to bed!

[Geoff]

$this->bbcode_second_pass_quote('pogoman', 'A')fter upgrade, make sure check you /dev/ and see if random/tty*/mtd*/sdb* are there.
if not, mknod of them, otherwise you will not be able to boot up.

[Onnline]

I updated my kernel via the script and took the latest rootfs (06-2012). When trying to run transmissiond the configs were not created, will looking in the logs I stumbled upon the fact the /dev/null is created as 600 instead 666. The post on https://bbs.archlinux.org/viewtopic.php?id=142816 hints this to be something with udev, only rules I see in udev are automount and automount SD. How do I set /dev/null to 666 at boot?

[azraleus]

Hey, I found a workaround. I added those lines to my /etc/rc.local and I can run my transmissiond without problems.

$this->bbcode_second_pass_code('', 'chmod 777 /dev/null
rm /dev/random
rm /dev/urandom
mknod -m 666 /dev/random c 1 8
mknod -m 666 /dev/urandom c 1 9
')

[slycat]

Bit of a bind:
System reboot failed, used a spare flashdrive with latest rootfs to get a chroot environment for my drive and ran the update script (w/ uImage.pci since PPPro) with no errors returned. Umounted chroot folders after exiting, reboot into original drive, but no luck. So I tried to boot again from the spare flashdrive, but that didn't work either. Decided to reinstall the latest rootfs on spare flashdrive, again with mac_addr added, and still nothing.

Front led is blinking but then goes solid green. There are no other drives connected when booting ensuring the flashdrive is the first to boot from. However I am getting no connectivity (does not show up on router list) and, therefore, cannot access the plug at all at this point.

I am positive the drives are intact and free of hardware-related errors and that the spare flashdrive is with the latest rootfs for oxnas along with a properly formatted mac_addr. Any ideas what might've gone wrong?

[WarheadsSE]

Try http://archlinuxarm.org/os/oxnas/ArchLi ... nas.tar.gz I just haven't made it an official latest yet.

[slycat]

That was able to get my spare flashdrive to boot, thanks.
However, after formatting my harddrive to a back up I made, chroot, and running the update-oxnas (w/ PCI), it still won't boot off the harddrive. The usual constant blinking and no connection to router (I have it scripted to go orange after initialization and then green once rc.local is complete). Just to note: when chrooting the HDD is on usb, but when trying to operate it is by eSata.

Could there be another package that is in conflict, like a list of baddies I've mistakenly upgraded to? Would it be necessary to upgrade/downgrade (pacman -Syu) before leaving the chroot?

[Geoff]

Just a thought for those who re-run update-oxnas: try to avoid erasing and re-flashing the same kernel over again to NAND. That might wear out the NAND, and there's no need to do that.

$this->bbcode_second_pass_quote('slycat', 'J')ust to note: when chrooting the HDD is on usb, but when trying to operate it is by eSata.

[WarheadsSE]

I've never had to Zero them, but yes Geoff is right on the dd to /dev/sdX1.

The re-flashing the update, true, but a couple of times at most shouldn't be breaking things.