[07:36] PR snapcraft#2389 opened: repo: run `apt update` before `apt install` [09:22] hi, anyone that could help with a rpi3 issue with ubuntu core? [09:48] is there a tool to modify the uboot.env file? [11:51] ackk: I'm here but I don't know enough about rpi3 to help [11:51] ackk: the config system can toggle some options in the boot process [11:51] ackk: but not sure what you really need [12:14] zyga, hi, so, I have rpi3 with ubuntu core, at some point it was rebooted and didn't come back. turns out the pi2 kernel snap was updated but the var in the uboot env wasn't, so it was trying to boot from a nonexistent kernel [12:14] zyga, I fixed that (I wonder if it's a ubuntu bug), but now the initramfs can't mount the root partition for some reason. it's ext4, but it fails after trying ext3 and ext2, saying there's some incompatibility [12:14] I can mount it fine on my pc [12:43] PR snapd#6089 closed: tests: install dependencies during prepare [13:32] ackk, from the runninf system you can use fw_printenv and fw_setenv to modify uboot.env [13:32] *running [13:33] and indeed you can manage uboot env from the uboot shell from the serial console like on all embedded boards we support [13:34] ogra yeah I was looking for something to change it from my pc, found about uboot-go from mvo :) [13:34] i have never seen it fail to set the vars though ... that would definitely be a bug if you didnt tinker with the card (not properly unmounting it or some such when you used it in a PC between boots) [13:34] ogra, so it turns out the snaps for core and kerner were wrong in the env [13:34] not sure why they didn't get updated [13:34] i have pi's here that run the same image since 2 years without ever failing [13:35] ogra, yeah that's weird [13:36] the ext2/3/4 warnings are nroaml btw ... the ext4 driver loops over all possible extX versions it can support and pronts that nonsense [13:36] *prints [13:36] *normal [13:37] if you can capture a full serial boot log and dump it to a pastebin i can take a look [13:41] ogra, is it normal that the core snap is marked as disabled? [13:42] ogra, (same for the pi2 kernel) https://paste.ubuntu.com/p/F2mrphCvfr/ [13:46] no, something is seriously broken [13:46] what image was that originally ? [13:47] core kernel and gadget should never be marked as disabled on UC [13:48] ogra, the official pi3 image [13:48] for ubuntu core [13:48] hmm, strange [13:48] ogra, so I tried to enable core, it rebooted but didn't come back [13:49] i still wonder how you got into that situation in the first place [13:49] ogra, should the pi2-kernel be enaled on a pi3? [13:49] yes [13:49] ogra, it's my nextcloud box, it was just there running [13:49] ogra, I had a power loss, maybe something broke there [13:49] the name is a bit confusing, it should be pi-kernel [13:50] a power loss shouldnt cause this ... the system should recover [13:50] nothing else happened [13:50] snap_core=core_4916.snap [13:50] snap_kernel=pi2-kernel_73.snap [13:50] snap_mode= [13:50] snap_try_core=core_5331.snap [13:50] this looks better [13:50] not really ... [13:51] why? [13:51] sndp should have unset snap_try_core after it verified the boot [13:51] well it hasn't yet [13:51] *snapd [13:51] I just reenabled the snap [13:52] lemme try to reboot [13:52] echnically you should just have rebooted until the system fixed itself ... without touching anything [13:52] if you touched any vars they might be out of sync with what snapd expects [13:53] ogra, I did touch snap_core and snap_kernel as they were set to values that didn't exist and the pi wasn't booting [13:53] i assume there is something wrong with the watchdog ... the system should have done this on its own (rebooting until it fxed itself) [13:54] could it be that your SD is/was full ? [13:54] ogra, no [13:54] we dont really have protection against that yet [13:54] ogra, I rebooted, but it's still on the old core [13:54] and the try is still there [13:54] thats the only way i can imagine that the vars point to something non-existing [13:54] try "snap refresh core" [13:55] ack@nextcloud:~$ df -h | grep mmc [13:55] /dev/mmcblk0p2 3.6G 964M 2.4G 29% /writable [13:55] /dev/mmcblk0p1 127M 38M 89M 30% /boot/uboot [13:55] ogra, ^ fwiw [13:55] ogra, ok it's downloading a newer core now [13:55] lets see how that goes [13:57] rebooting now [13:57] good [13:58] well, it's supposed to, it doesn't seem it is [13:58] whats on the console ? [13:58] ok finally [13:58] nothing [13:58] but it's rebooting now [13:58] finally ? you mean it booted ? [13:59] it's booting [13:59] k [14:09] ogra, ok, everything seems green now [14:09] how does the kernel look in snap list ? [14:09] enable again ? [14:10] *enabled [14:10] pi2-kernel 4.4.0-1098.106 73 stable canonical✓ kernel [14:10] yep, everything enabled [14:10] ok .. thats fine but outddated ... [14:10] snap refresh pi2-kernel [14:10] then you should be back up to date [14:11] (current stable is rev 74) [14:13] cool, thanks === fayg0 is now known as gh0st^ === gh0st^ is now known as fayg0 === phoenix_firebrd is now known as murthy [22:20] Bug #1600136 changed: App indicator does not show icon for Qt apps or with custom icons

(Ubuntu):Fix Released by 3v1n0>