Cannot reboot/shutdown after upgrade to new kernel (4.2.3)

mzsade · 10-23-2015, 12:01 AM

Was so delighted with the upgrade, everything worked perfectly apparently, had no issues with sound or graphics, the system freeze problem from before was no more and iceweasel was never snappier; it was enough to convince me that i could safely remove the old kernel and it's modules and now this, drat!
No journal files in the /var/log folder
Booting with the recovery option also fails. Please, where do i start looking?

Timothy Miller · 10-23-2015, 12:04 AM

What DE/WM are you running? Does running the shutdown command from a console still work?

mzsade · 10-23-2015, 12:20 AM

Standalone OpenBox. Neither <sudo reboot> nor <sudo shutdown -h now> works.

Timothy Miller · 10-23-2015, 12:32 AM

Do you have your system as sysvinit or systemd?

mzsade · 10-23-2015, 01:41 AM

Systemd and systemd-networkd instead of NetworkManager. Would the output of dmesg be of use?

Same thing with "poweroff -i -f". Uninstalling acpid didn't help.

cynwulf · 10-23-2015, 04:33 AM

As you're only running openbox, is it possible that you can purge systemd and install sysvinit-core and see if that causes, halt, shutdown, poweroff, reboot etc to start working again?

In systemd those are just links provided by the systemd-sysv package. So check if that package is still installed and didn't 'disappear'?

More info is also need besides "doesn't work". What output do you see when you run those commands?

Apart from that I can only state the obvious: You're running unstable and breakage is quite normal.

Timothy Miller · 10-23-2015, 07:44 AM

Yeah, sysvinit would be the next step to rule out a breakage in systemd (not like systemd would EVER have issues...).

mzsade · 10-23-2015, 01:00 PM

Systemd is the defacto standard and there's no going back to "the good old days" so please don't offer that as a solution. If you google the title of this thread you'll see that my issue, or my hardware are not unique to this problem after an upgrade. The output of the commands that you asked for is invariably, "shutting down processes, unmounting file systems, preparing to shutdown or powering off" at which point the system freezes or hangs, or a blank screen, so i don't see what specific "more info" you were expecting from me which would have helped you arrive at a definitive solution.

Upgrading the kernel from the repos has never gone well for me, if i can reboot into that miserable partition again i will try to compile the new kernel from source and let you know if that takes care of the issue before i reinstall. Thanks.

Timothy Miller · 10-23-2015, 03:14 PM

Depends on which distro you have. Jessie still has SysvInit installed unless explicitly uninstalled by the user, even though SystemD is the default. I specifically had to use this the other day because SystemD was hanging on boot. Had to boot to SysVInit, fix SystemD, then reboot with SystemD.

Shadow_7 · 10-23-2015, 03:18 PM

The shutdown syntax has changed with systemd it seems.

Where the old way is:

# shutdown -h -P now

The NEW way is:

# shutdown -H -P +0

No idea why they changed that.

angryfirelord · 10-23-2015, 03:35 PM

Debian has its own wayback machine in regards to packages. If it happened after a kernel update, then you can enable a snapshot of a repository and try to install an older kernel.

http://snapshot.debian.org/

The issue tracker will tell you when a package was uploaded, so you can use the timestamp of that for your snapshot repo.

https://tracker.debian.org/pkg/linux

Does journalctl show anything? Are you on testing or unstable?

mzsade · 10-23-2015, 03:57 PM

Unstable, last few lines from the output of

Code:

sudo shutdown -H -P +0
systemd-shutdown[1]: All DM devices detached
systemd-shutdown[1]: Powering off
kvm: exiting hardware virtualization
sd 0:0:0:0: [sda] synchronizing SCSI cache
sd 0:0:0:0: [sda] Stopping disk

This is where the screen freezes so tell me what you make of it. I still think upgrading the kernel from the repos buggers up the system somehow and recompiling the new kernel from source will fix it. It's a wonder that i can still boot into it (albeit after more than one attempt now) and that it's fully functional apart from this glitch. I've been treating my poor hard disk like a rag doll these past few days and i hope it will survive long enough for me to try recompiling the kernel from source and test my uneducated surmise.

273 · 10-23-2015, 04:19 PM

Can you not just install an older kernel and boot that instead until things calm down
I had some weird stalls on shutdown on both my machines after some fairly recent dist-upgrades but they did take a few minutes to shut down at one point so, perhaps, you have to wait a long time (I think it took about 5 minutes or so the first few times)? Sorry, I have no explanation as to why this would be the case.

Shadow_7 · 10-23-2015, 04:31 PM

Perhaps leave off the -P if it's a virtualized thing that you're exiting. Since -P is power off (on both versions of shutdown). And tends to be x86 specific.

mzsade · 10-23-2015, 04:47 PM

Quote:

Originally Posted by 273

Can you not just install an older kernel and boot that instead until things calm down

Wish i hadn't been in such a darned hurry to remove the old kernel and it's modules. Anyway, give me a couple of days, i will compile from source this time and if that fixes the issue we will know for certain it's not the kernel but the method by which it was upgraded to that was at fault.

Edit: Nope, it finally died on me so we'll never know. R.I.P. Sid, we had a pretty good run together.