Empty coretemp entries in thermal sensors widget

netnerdy

I have started using the thermal sensors widget. It was working fine, but now widget has gone haywire and showing up invalid entries. Invalid entries are increasing slowly: at first there were no invalid entries and slowly it's taken over the dashboard.

Anyone knows what's going on?

stephenw10

That looks like a gui glitch but it could also be broken ACPI tables on that device.

What do you see those as if you edit the widget and enable 'full sensor name'?

If you reboot do they appear exactly the same?

Steve

netnerdy

Pasted the screenshot when I enabled full sensor name.

I also restarted pfsense afterwards, but the widget still has the invalid entries

Gertjan

Strange.

The 'dev' list is parsed for the literal word 'temperature', not 'temp'

https://github.com/pfsense/pfsense/blob/29b42d654071ce58a2e194319cfc6b7447fe2bca/src/usr/local/www/widgets/widgets/thermal_sensors.widget.php#L34

Can you run

/sbin/sysctl -q dev.cpu | grep temperature | sort

on the command (console or SSH) line ?

netnerdy

I get:

[2.4.5-RELEASE][admin@pfSense]/root: /sbin/sysctl -q dev.cpu | grep temperature | sort
dev.cpu.0.temperature: 49.0C
dev.cpu.1.temperature: 49.0C
dev.cpu.2.temperature: 50.0C
dev.cpu.3.temperature: 50.0C

stephenw10

Try using just: sysctl -aq | grep temperature which is what your device would be checking.

https://github.com/pfsense/pfsense/blob/RELENG_2_4_5/src/usr/local/www/widgets/widgets/thermal_sensors.widget.php#L35

netnerdy

This is what I get when I run that command:

coretemp3: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp3: critical temperature detected, suggest system shutdown
coretemp1: critical temperature detected, suggest system shutdown
coretemp3: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp3: critical temperature detected, suggest system shutdown
coretemp1: critical temperature detected, suggest system shutdown
coretemp1: critical temperature detected, suggest system shutdown
coretemp3: critical temperature detected, suggest system shutdown
coretemp3: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
coretemp0: critical temperature detected, suggest system shutdown
hw.acpi.thermal.tz1.temperature: 29.9C
hw.acpi.thermal.tz0.temperature: 27.9C
dev.cpu.3.temperature: 57.0C
dev.cpu.2.temperature: 57.0C
dev.cpu.1.temperature: 64.0C
dev.cpu.0.temperature: 64.0C

stephenw10

Aha! So it's matching the word 'temperature' in the warnings there.

I would expect that to change after a reboot though, you say it appears identically?

Has it actually overheated? That looks like it must be passively cooled.

Steve

stephenw10

You could try changing the grep to match more accurately. So, for example, set that line in thermal_sensors.widget.php to:

		$_gb = exec("/sbin/sysctl -aq | grep temperature:", $dfout);

Steve

stephenw10

This looks like an easy fix, even I can do it! Opened a bug to track:

https://redmine.pfsense.org/issues/10963

Gertjan

@stephenw10 said in Empty coretemp entries in thermal sensors widget:

Try using just: sysctl -aq | grep temperature which is what your device would be checking.

Your right !!
I was linking the master (future 2.5.0 ...) code.
The 2.4.5 code is what most of us are using right now.
I was already using the new '2.5.0' code myself for identical reasons.
Just one line to change and you'll be fine.

netnerdy

Thanks everyone for the help.

Yes, it is a passively cooled machine. Not sure why reboot didn't clear the logs. Maybe it reached critical temp during bootup?

stephenw10

I would think at least the ordering would be somewhat random if it happened at boot. Probably not something that's cleared at boot then.
It would be interesting to see where in the sysctl output that is shown. It's probably possible to clear it manually if we know what is generating it.

Steve

netnerdy

I changed the php file under /usr/local/www/widgets/widgets/thermal_sensors.widget.php and it worked. Thanks!