Qubes not shutting down

[Loren Rogers]

Hi all,

I've successfully installed Qubes on my Thinkpad X201 tablet, but it has
issues shutting down. When I explicitly tell it to reboot or shutdown,
it goes through the entire shutdown sequence, but hangs on an empty
black screen. Occasionally, I see an unchanging white underscore (_)
character displayed in the top left when it hangs.

I tried leaving it in this state for about an hour, and no change--I've
always had to force-reset. I assume this is not normal?

Also, I find that the system randomly begins the shutdown sequence on
its own. (And hangs on the black screen at the end.)

Thanks,
Loren

[Grzesiek Chodzicki]

The same issue occurs on my system only if I shut the system down while a VM with a PCI device without FLR support is running

[Loren Rogers]

Interesting - thanks for the info. Are you saying it doesn't shut down,
or that it shuts down automatically?

[Loren Rogers]

On 11/16/2016 02:33 PM, Grzesiek Chodzicki wrote:

Also, I just confirmed that it shuts down cleanly with all VMs off and
no USB devices plugged in.

[Andrew David Wong]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

It sounds like you've encountered one (or both) of these issues:

https://github.com/QubesOS/qubes-issues/issues/1581
https://github.com/QubesOS/qubes-issues/issues/1826

- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
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=bRaI
-----END PGP SIGNATURE-----

[Loren Rogers]

Another correlation I've noticed is that my machine randomly shuts
itself down without warning when I'm browsing in the Anon-Whonix VM. It
seems that simply having the Whonix browser open causes the problem.
I've not been able to pin down an exact cause, but it seems to happen
after about 5-20min. When this happens, the machine sometimes ends up in
a hung state (black screen) at the end of the shutdown process.

I've also noticed that the fan speeds up right at it starts to shutdown.
(The screen turns to the Qubes logo with the progress bar, then the fan
cranks up.) Sometimes the bar makes it all the way to the end, other
times it seems to simply crash to a hault. As I mentioned elsewhere, the
Thinkpad X201t is known to have overheating issues, but I'm not sure if
this is related. I'm not working the machine particularly hard (just
browsing articles on the web), and the hardware is not particularly hot
to the touch.

[raah...@gmail.com]

ya i just usually have to shutdown all vms except for the sys-vms that autostart.

[Drew White]

On Sunday, 20 November 2016 04:56:03 UTC+11, Loren Rogers wrote:
> Another correlation I've noticed is that my machine randomly shuts
> itself down without warning when I'm browsing in the Anon-Whonix VM. It
> seems that simply having the Whonix browser open causes the problem.
> I've not been able to pin down an exact cause, but it seems to happen
> after about 5-20min. When this happens, the machine sometimes ends up in
> a hung state (black screen) at the end of the shutdown process.
>
> I've also noticed that the fan speeds up right at it starts to shutdown.
> (The screen turns to the Qubes logo with the progress bar, then the fan
> cranks up.) Sometimes the bar makes it all the way to the end, other
> times it seems to simply crash to a hault. As I mentioned elsewhere, the
> Thinkpad X201t is known to have overheating issues, but I'm not sure if
> this is related. I'm not working the machine particularly hard (just
> browsing articles on the web), and the hardware is not particularly hot
> to the touch.

When it gets to the qubes logo screen, press ESC to see what it's actually doing.

If you wish to always know what it's doing, turn off rhgb and quiet in the boot config.

Then you will see where the issue is.

[Loren Rogers]

Thanks, I'll give that a shot next time it happens. I feel like it'll go
by too quickly for me to see what's happening; does it also write its
activity to a log somewhere?

[Loren Rogers]

On 11/21/2016 12:42 AM, Drew White wrote:

[Loren Rogers]

I can now confirm that it's an over heating issue. When it went into the
automatic shutdown sequence, I pressed escape and managed to take note
of a few of the messages. One of the very first ones was something about
"thermal_zone_0 critical temperature reached: 128C", which I assume is
the cause. (This isn't an exact quote, since I noted it from memory.)

This raises some questions:
- What could be causing this overheating issue in Whonix?
- Is 128C a normal temperature for the safety shutdown to kick in?
- Does Qubes have a warning / alert system for potential overheat? (Like
low battery)

[Drew White]

It is a high temperature, but does it ONLY happen in Whonix?
Or if you push the PC does it happen also?
Have you tried limiting the threads Whonix can use?

Sometimes CPUs have shutdown at 99 degrees.
So 128 degrees is a bit high in my own opinion.

I recommend you check the CPU Fan and heatsinks (if it has them).

[Loren Rogers]

Thanks for the input - I just dusted out the fan, and we'll see if it
helps. It wasn't too bad, but we'll see if there's an improvement.

No, it also randomly goes into auto-shutdown when backing up VMs.
However, that happens about 20% of the time. Whonix seems to do it about
80% of the time, the other 20% I figure I shut it down before it does so
on its own. I figure there may be something in the Whonix VM that's
causing my processor to over work itself. The auto-shutdowns may be
ultimately linked to dust in the fan or something like that, but if
there's something processor intensive in Whonix, it may be worth looking
into.

Also, a heat warning message would be nice. I assume the thresholds are
set via the bios - is there a standard way of monitoring this? (I'm not
particularly well versed in this sort of thing.)

[Drew White]

I recommend you get your HDD checked, and your RAM.

Test both thoroughly.
Could be some bad sectors.
Also run a smartd check.

Some PCs have system diagnostics built in for RAM in the startup sequence.

[Pawel Debski]

Thinkpads have known problem that after long time of usage GPU radiator glued to the chip goes loose and X201 is rather older model. Not sure how to measure GPU temp but if this is the case you can improve it yourself with superglue & silver smear or any repair shop around shall do it for a few coppers.

[Loren Rogers]

Thanks everyone -- this is great info.

I think the dust removal was really helpful. I've noticed that the fan
is working a lot less, and it's not crashing during backups like it did
before. I'll check the HDD and RAM soon, and I'll keep the GPU radiator
issue in mind. (I had no idea this was an issue - thanks for the pointer!)

Any thoughts on a temperature warning message?

[Loren Rogers]

Also, returning to the original issue, I've attached a photo of the
output logs when the machine hangs during shutdown. (I pressed esc while
the Qubes logo was showing, before it locked up.) The last few messages
confirm that I was having an overheating issue, but the device-mapper
messages are new to me. Does anyone have a lead on what this may be?

[Drew White]

I would recommend taking it to a real computer shop and getting it checked out.

There could be several things wrong.. It may need a reseat of the heatsinks, and also a check of the sensors. They can sometimes be old and go bad and read wrong.

If you have a temperature sensor that you can use, check the temperature of the pc case near the CPU and GPU just to see how hot it is on the case. That may give you a good indication of something being very wrong.

But it's cold to the touch, then either it'd doing well in moving the heat, or the sensor is knackered.

Either way, it may be a good idea to get someone to give it a quick checkup.

[raah...@gmail.com]

maybet thats your hdd reported temp? Fedora use smartctl by default I think? could be your hdd is going if its the case.

[raah...@gmail.com]

test the hdd, if not probably cpu.

[raah...@gmail.com]

you can install lm-sensors to dom0 and check temps while using the pc.