Thursday, 2013-11-21

--- Day changed Thu Nov 21 2013
-!- b2h [~user@180.245.180.142] has quit [Quit: Leaving]02:00
WntSuper! Got Sailfish SDK image running on my N903:34
WntBut I somehow managed to break my harmattan root partition03:35
Wntand running "fsck.ext4 -y /dev/mmcblk0p2" in Ubiboot says "fsck.ext4: Bad magic number in super-block while trying to open /dev/mmcblk0p2"03:36
WntI suspect that the moslo partition resizing messed it up. And that propably was caused by me having some old nitroid related partitions03:40
WntIs there something I can do to rescue my harmattan root filesystem or should I just reflash everything? This is my current partition table: http://upload.egarden.fi/N9_partition_table.txt03:43
-!- user__ [~user@120.168.1.53] has joined #ubiboot08:17
-!- user__ [~user@120.168.1.53] has quit [Client Quit]08:19
@juicemeWnt, good morning 08:47
@juicemeDo you have backups from your Harmattan partition?08:47
@juicemeIf so, then just reflash and restore from backups. If not, how badly is it messed up, can you access it at all?08:48
Wntjuiceme: cannot access it fsck doesn't understand that there would be any filesystem on the partition.13:19
WntI suspect that the partition table was messed up by the moslo partition resizing thing13:20
WntIs there some default start and end blocks that the harmattan root filesystem partition would normally have on all devices?13:21
WntI could try to change the partitions to match those13:23
Wntbut I do have backup of my MyDocs filesystem and just made a backup with the N9's builtin backup software so I could reflash without losing any data13:24
WntI would only need to install all software again13:25
@juicemewnt, just a minute I'll check from my device13:44
@juicemethe partitions on my device are currently like this;13:45
@juiceme~ # /sbin/sfdisk -l13:45
@juicemeDisk /dev/mmcblk0: 1957120 cylinders, 4 heads, 16 sectors/track13:45
@juicemeUnits = cylinders of 32768 bytes, blocks of 1024 bytes, counting from 013:45
@juiceme   Device Boot Start     End   #cyls    #blocks   Id  System13:45
@juiceme/dev/mmcblk0p1         16  1632511  1632496   52239872    c  W95 FAT32 (LBA)13:45
@juicemeend: (c,h,s) expected (1023,3,16) found (1023,63,32)13:45
@juiceme/dev/mmcblk0p2     1760512  1891583  131072    4194304   83  Linux13:45
@juiceme/dev/mmcblk0p3     1891584  1957119   65536    2097152   83  Linux13:45
@juiceme/dev/mmcblk0p4     1632512  1760511  128000    4096000   83  Linux13:45
@juicemestart: (c,h,s) expected (1023,3,16) found (1023,63,32)13:45
@juicemeend: (c,h,s) expected (1023,3,16) found (1023,63,32)13:45
@juicemeYou can try to manually fix your device to resemble that. (it's a 64G device btw, just a minute I'll fetcg the 16G version...)13:47
WntI have a 64G device also13:50
Wntthat's at least something I can try before reflashing13:50
WntI guess there migh also be some tools that could scan the whole block device for filesystem information13:51
@juicemehere's the info13:51
@juicemeRM696-21-3:~# sfdisk -l13:51
@juicemeDisk /dev/mmcblk0: 485120 cylinders, 4 heads, 16 sectors/track13:51
@juicemeUnits = cylinders of 32768 bytes, blocks of 1024 bytes, counting from 013:51
@juicemeDevice Boot Start     End   #cyls    #blocks   Id  System13:51
@juiceme/dev/mmcblk0p1         16  288511  288496    9231872    c  W95 FAT32 (LBA)13:51
@juiceme/dev/mmcblk0p2     288512  419583  131072    4194304   83  Linux13:51
@juiceme/dev/mmcblk0p3     419584  485119   65536    2097152   83  Linux13:51
@juiceme/dev/mmcblk0p4          0       -       0          0    0  Empty13:51
@juicemeah13:51
@juicemewas too quick to paste it :)13:52
@juicemeI am only aware of tools for checking a particular filesystem on a device but there certainly is no reason why whole device could not be checked.13:53
WntUsed testdisk on my laptop while ubiboot exported the whole block device via usb and testdisk was able to determine the correct partition table15:27
WntNow might be a good time to make a backup of the rootfs :)15:36
-!- Basstard` [~sturgeon@c-5eeaaa55-74736162.cust.telenor.se] has joined #ubiboot21:15

Generated by irclog2html.py 2.12.1 by Marius Gedminas - find it at mg.pov.lt!