beta2

[mk01]

(6th Sep, 2013 04:04 AM)namtih Wrote:  My cmdline.txt:

Code:

sdhci-bcm2708.sync_after_dma=0 dwc_otg.lpm_enable=0 console=tty1 root=/dev/sda2 rootfsopts=subvol=root/@,noatime,compress=lzo,thread_pool=1,space_cache,autodefrag rootfstype=btrfs rootwait smsc95xx.turbo_mode=N elevator=noop nohdparm noswap mod_scsi.scan=async

the two errors during upgrade are ok, both because the needed devices/services will be available only after restart with proper kernel.

just the cmdline.txt looks wrong. like it would not be upgraded. i'm just wondering how your /etc/fstab looked like at the time of running upgrade, specially "/boot" record - and whether /boot could be mounted at that time from device /dev/mmcblk0p1.

nevertheless the cmdline.txt should be looking like this:

Code:

sdhci-bcm2708.sync_after_dma=0 dwc_otg.lpm_enable=0 console=tty1 root=/dev/sda2  rootfsopts=subvol=root/@,noatime,autodefrag,thread_pool=3,compress=lzo rootfstype=btrfs rootwait=10 smsc95xx.turbo_mode=Y elevator=deadline logo.nologo quiet noswap loglevel=0 mod_scsi.scan=sync partswap startevent=xbian splash nohdparm

specially the "startevent=xbian" is important. otherwise it can do "nothing" as you describe it.

---

(6th Sep, 2013 05:06 AM)namtih Wrote:  Before the update I've made an backup image of the USB drive with win32diskimager. I've now restored the image to the USB drive, but it's also hanging now while booting....
http://www.imagebanana.com/view/z5pvit3f/20130905_210327.jpg

just to provide more info - beta1 needs to be booted without "startevent=" parameter, beta2 with it.

the "btrfs generations" just mean, you mounted the filesystem with newer kernel, then reverted to old(er) one. btrfs driver (from older kernel currently booting) found records it can't understand.

---

if you created the /dev/sda copy with dd with copying whole /dev/mmcblk0, then the BOOT first partition got copied including the same LABEL and UUID which could cause that during update your 1st partition from /dev/sda got mounted instead of the right one which caused kernel & initramfs & cmdline.txt update on wrong partition - definitely not the one which is used by RPI firmware to boot the system up (/dev/mmcblk0p1).

can confirm this? or can you post output of running "sudo blkid" ?