Since the late 90s, most PC's have come with built-in health monitoring chips that sense temperatures, voltages and fan speeds, and all this useful -- and potentially PC-saving -- information is available to users.

In Linux, the best way of monitoring it is via the lm-sensors package that comes ready-installed in most distributions. It does however require a little configuration.

• On a command line, run sensors-detect to kick off a routine to check out your hardware and see what kernel modules should be loaded to make the most effective use of lm-sensors.

• At the end you'll be presented with a summary of findings and given the option of adding recommended drivers to /etc/modules. Type 'yes' if you want these modules loaded automatically every time you boot.

There's no need to reboot to try out lm-sensors. Just 'modprobe' each of the listed modules then do a sensors -s to evaluate the new configuration settings. In my case I got the message

To load everything that is needed, add this to /etc/modules: #----cut here---- # Chip drivers coretemp w83627ehf #----cut here----

... so I did

modprobe coretemp
modprobe w83627ehfupdate-grub
sensors -s

Simply running sensors then detailed all the information for my hardware.



Troubleshooting
I did have a little trouble loading w83627ehf driver as modprobe returned

FATAL: Error inserting w83627ehf (/lib/modules/2.6.32-24-generic/kernel/drivers/hwmon/w83627ehf.ko): Device or resource busy

but a little googling found the answer. In this case, adding the line

GRUB_CMDLINE_LINUX="acpi_enforce_resources=lax"

to /etc/default/grub then running update-grub solved the problem and loaded the driver when I rebooted.



Extensions
Once you have lm-sensors up and running you can get visual feedback from the likes of desktop widgets:



or run a daemon and plot your information over time:



There's also a ton of information on the lm-sensors website.