Fireware Update Manager snap misbehaving: won't unmount at shutdown


#1

Dear all

EDIT: other snaps have trouble uloading too: I get a message about snap ‘core’ not unmounting, as well.
Hello!

I have the Fireware Update Manager installed as a snap:

$  snap list
Name   Version    Rev   Tracking  Publisher       Notes
core   16-2.39.3  7270  stable    canonical✓      core
fwupd  1.2.10     880   stable    richard-hughes  classic

I see this at computer shutdown:

Failed unmounting Mount unit for fwupd

Can anyone rid me of this troublesome error message? So far as I can tell, admittedly, the message - or what causes it - has no consequences.

My system:

$  neofetch --off [edited]
OS: Linux Mint 19.1 Tessa x86_64 
Host: 20KHCTO1WW ThinkPad X1 Carbon 6th 
Kernel: 5.2.3-050203-generic 
Packages: 3158 (dpkg), 6 (flatpak), 3 (snap) 
Shell: bash 4.4.20 
Resolution: 2560x1440 @ 60.01Hz 
DE: Cinnamon 4.2.3 
WM: Mutter (Muffin) 
CPU: Intel i7-8550U (8) @ 4.000GHz 
GPU: Intel UHD Graphics 620 
Memory: 2320MiB / 15927MiB 
Disk (/dev/nvme0n1p4): 17G / 40G (44%) 
Disk (/dev/nvme0n1p2): 185G / 405G (48%) 
Disk (udev): 0 / 7.8G (0%)

#2

I think the problem may have been that a snap called ‘core18’ had become - as ‘snap list’ put it - broken. I reinstalled that snap and now it no longer reports as broken.

Possibly that problem was caused by installing a very recent (but release as against release candidate) Linux kernel.

However: that problem with ‘core18’ may not have been the cause of my original problem. The original problem was intermittent (I think) so although I am not seeing it currently it might recur.


#3

while this is indeed ugly, snap packages are readonly squashfs filesystems, so it isn’t fatal, nothing gets written to them so no data can get corrupted when they do not unmount cleanly.


#4

Thank you for the information.


#5

There is a known bug in the 5.2.* series of Linux. This has only very recently been patched, and is unlikely to have made it into your RC build.


#6

Thanks. I’m on kernel 5.2.6 (which is not a release candidate). I’ll way wait for the fix to come through on some future version of kernel 5.2. I suppose I’m asking for (at least limited) trouble by running such a recent kernel.