Thread: Introducing ubiboot N9 (multiboot OS loader)
View Single Post
juiceme is offline juiceme
2013-04-04 , 21:42
Community Council | Posts: 4,920 | Thanked: 12,867 times | Joined on May 2012 @ Southerrn Finland
#239
Originally Posted by peterleinchen View Post
What about updating from 01 to 02. I think as easy as grab source/ubifs.tar and overwrite? Or do we need to do some more.
Well, updating from 01 -> 02 is really easy, just grab the ubiboot-02_040413.tar, put the ubiboot.conf and ubiboot-02.menus.cpio in the root of your ubifs partition (since you already have prepared that) and check the content of the ubiboot.conf.

Originally Posted by peterleinchen View Post
Thanks also for the USB/RTC thingy. I hope it works, will see soon.
I had modified the init to only enable the maintenance mode when bootreason was pwr_key and USB attached (so first push power button and then attach USB). But this seems way better.
The fix should work now. The logic is 2-staged, first it is detected that powerup reason was "usb" and then sysfs is mined for the actual event, whether it is "charger" or "data" connection. Device only enters maintanance mode if it was plugged to a computer.

Originally Posted by peterleinchen View Post
Another report:
I had a lot of wd32_to bootreasons. Almost always after shutting down device from any other kernel. And then boot up and let menu time out ( so default kernel l2fix was chosen). Then first boot did not succeed and a second boot (also timeout) did work. For this second boot the bootreason was wd32_to. Any ideas?
Did not go completely through sources, but have seen you have command line in logger (which holds also bootreason, right?)
Just realised that therr is some unclean umounting going on... propably that amounts to the problems coderus has been also having. Just took me 30min to boot back to Harmattan

Originally Posted by peterleinchen View Post
Another report (which you may change maybe):
in new sources: when/if the kexec loading fails we get a log, but this log will not show the error number given back
I'll have a look at it when I need to change init code anyways to fix the current issues.
 

The Following User Says Thank You to juiceme For This Useful Post: