Linux 5.5 To Finally Expose NVMe Drive Temperatures Via HWMON

[phoronix]

Phoronix: Linux 5.5 To Finally Expose NVMe Drive Temperatures Via HWMON

Linux for years has supported monitoring NVMe drive temperatures when installing the nvme user-space utility and run as root, etc. But now finally with Linux 5.5 the kernel is supporting NVMe drive temperature reporting through the hardware monitoring "HWMON" infrastructure alongside other hardware sensors...

http://www.phoronix.com/scan.php?page=news_item&px=Linux-5.5-NVMe-HWMON-Support

[shmerl]

Finally! I had to use nvme-cli to read that.

[muncrief]

I just applied the patches to my Manjaro 5.4.0-rc8 kernel build and it works!

I didn't even have to rescan, after reboot there was simply a new sensor named nvme-pci-0100 reading the correct temperature under the label "Composite." So I just changed the name in Psensor and now I can finally monitor all my disk temperatures again.

Now if AMD could just fix the the fan control/RPM sensor issues with my R9-390 all my hardware issues would be solved for now. I actually had to use a service I found in a Wiki, with a few slight modifications, to manually monitor my GPU temp every second and control its fans, which is kind of crazy. But if I didn't do it my GPU would start melting when I tried to play a game.

[Danny3]

I could never understand the broken logic of Linux where you would need a utility to run as root to read the sensors.
I'm glad that some developers want to bring some sanity into this madness.

[skeevy420]

I could never understand the broken logic of Linux where you would need a utility to run as root to read the sensors.
I'm glad that some developers want to bring some sanity into this madness.

I can, but only to prevent certain hardware from doing unexpected behavior...like a remote user logging on and checking CD/DVD drive speeds before configuring a program to write a backup disk, but a few minutes earlier the sys-admin got tired of waiting on Joe to log in and do is damn job so he used that dive to burn the backup disk and the two commands clash leaving the disk burnt with some errors.

Hey Joe, that was our last disk so you're not working remote today and you need to pick up a 100 pack of DVD-Rs on the way in. I'll reimburse you when you get here.

[bug77]

I could never understand the broken logic of Linux where you would need a utility to run as root to read the sensors.
I'm glad that some developers want to bring some sanity into this madness.

If it's not exposed in proper, hwmon way, then probably you need to query some lower level stuff. Hence the need for root access.

[starshipeleven]

I could never understand the broken logic of Linux where you would need a utility to run as root to read the sensors.
I'm glad that some developers want to bring some sanity into this madness.

Broken logic of the NVMe module you mean?
Because for pretty much everything else you don't need root to read sensors.

user@hostname:~> sensors
nct6779-isa-0290
Adapter: ISA adapter
Vcore:                  +0.59 V  (min =  +0.00 V, max =  +1.74 V)
in1:                    +0.27 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
AVCC:                   +3.42 V  (min =  +2.98 V, max =  +3.63 V)
+3.3V:                  +3.42 V  (min =  +2.98 V, max =  +3.63 V)
in4:                    +1.88 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in5:                    +0.99 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in6:                    +0.28 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
3VSB:                   +3.49 V  (min =  +2.98 V, max =  +3.63 V)
Vbat:                   +3.30 V  (min =  +2.70 V, max =  +3.63 V)
in9:                    +0.00 V  (min =  +0.00 V, max =  +0.00 V)
in10:                   +0.22 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in11:                   +1.04 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in12:                   +1.70 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in13:                   +1.04 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
in14:                   +1.82 V  (min =  +0.00 V, max =  +0.00 V)  ALARM
fan1:                     0 RPM  (min =    0 RPM)
fan2:                  1439 RPM  (min =    0 RPM)
fan3:                     0 RPM  (min =    0 RPM)
fan4:                     0 RPM  (min =    0 RPM)
fan5:                     0 RPM  (min =    0 RPM)
SYSTIN:                 +30.0°C  (high =  +0.0°C, hyst =  +0.0°C)  ALARM  sensor = thermistor
CPUTIN:                 +30.0°C  (high = +80.0°C, hyst = +75.0°C)  sensor = thermistor
AUXTIN0:                +93.5°C    sensor = thermistor
AUXTIN1:               +100.0°C    sensor = thermistor
AUXTIN2:                +24.0°C    sensor = thermistor
AUXTIN3:                -24.0°C    sensor = thermistor
SMBUSMASTER 0:          +29.0°C 
PCH_CHIP_CPU_MAX_TEMP:   +0.0°C 
PCH_CHIP_TEMP:           +0.0°C 
PCH_CPU_TEMP:            +0.0°C 
intrusion0:            OK
intrusion1:            ALARM
beep_enable:           disabled

amdgpu-pci-0600
Adapter: PCI adapter
vddgfx:           N/A 
vddnb:            N/A 
edge:         +29.0°C  (crit = +80.0°C, hyst =  +0.0°C)

k10temp-pci-00c3
Adapter: PCI adapter
Tdie:         +29.1°C  (high = +70.0°C)
Tctl:         +29.1°C

[starshipeleven]

I can, but only to prevent certain hardware from doing unexpected behavior

It was just not hooked up to hwmon interface, there is no reason to wall off sensor readings.