AMD Ryzen 5000 Temperature Monitoring Support Sent In For Linux 5.12

[jsax]

Wait, does this mean that the temperature I'm seeing on kernel 5.10 is not correct?

I wondered the same and thought maybe the gentoo kernel I used was already patched.
I have just checked and it seems the change is only about the ccd temperature.
After applying the patch to my 5.10 kernel I now also have Tccd1 on my 5800x, where I only had Tctl and Tdie before:

k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +38.4°C
Tdie: +38.4°C
Tccd1: +36.8°C

[cytomax55]

Amd just posted a lot of Linux engineer jobs so hopefully this doesn't happen in the future

[Danny3]

Amd just posted a lot of Linux engineer jobs so hopefully this doesn't happen in the future

If that's the problem.
The article says it's a two line patch so any developer could've done that easily, especially that they don't have to reverse engineer anything like the developers outside AMD.
Or they could've published the documentation, the IDs or whatever is required.
This attitude makes it clear that they intentionally want to have bad Linux support, at least in the beginning.
For what purpose, I don't know, it seems to be counter-intuitive for a hardware company to want that, especially since they pursue open source on the GPU side.
Maybe it's some financial incentive from Microsoft to make Windows look better.

At least the good part for me is that until these CPUs are fully supported I can buy them second-hand which is cheaper.

[cytomax55]

I could be wrong and I don't have the time to research it but if I remember correctly they were a lot of different positions and some of them were more community manager type positions and some of them are more Linux engineer positions so I would hope there would be better coordination... While it is possible this is being done maliciously we could also argue there's been a poor amount of coordination only time will tell once these new positions are filled

[extremesquared]

For what it's worth, debian testing (using 5.10) is monitoring something that seems like ryzen 5950 temps by default. Not sure what it's actually polling. Plug-and-play though. I'm not really familiar with the data pathway for hw monitoring -- kernel / lmsensors / external lib, and it's unclear how accurate the numbers are, but they are clearly within 10% of reasonable numbers and reflect cpu-only load effects. We're not flying totally blind, at least, which is something that has been a problem in the past.
Maybe this change is for per-die monitoring?

[creative]

I have had a temp readout from the get go.

Voltages and clocks are more what I am after for zenpower and zenmonitor.

[creative]

Wait, does this mean that the temperature I'm seeing on kernel 5.10 is not correct?

[sanjuro]uname -r
5.10.16-custom
[~]
[sanjuro]sensors
k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +35.8°C
Tdie: +35.8°C

amdgpu-pci-2b00
Adapter: PCI adapter
vddgfx: 750.00 mV
fan1: 875 RPM (min = 0 RPM, max = 2900 RPM)
edge: +33.0°C (crit = +94.0°C, hyst = -273.1°C)
power1: 8.14 W (cap = 180.00 W)

What chip are you running?

zenpower-pci-00c3
Adapter: PCI adapter
Tdie: +28.6°C (high = +95.0°C)
Tctl: +28.6°C

[creative]

Sanjuro My mirror neurons fired off when I saw your post.

Went on ahead and removed the zenpower dkms module, since much of its functionality was lost.

$ uname -a
Linux name-pc 5.10.15-1-MANJARO #1 SMP PREEMPT Wed Feb 10 10:42:47 UTC 2021 x86_64 GNU/Linux

$ sensors
k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +24.5°C
Tdie: +24.5°C

[brauliobo]

weird news, I have temperature reporting in my 5600X with kernel 5.10 (and I think with 5.9 as well)

[creative]

brauliobo Of course you would, its just k10temp-pci reporting.

As I have mentioned before, while it may be of less value, secondary or even tertiary. The zenpower module reports a lot more on supported CPU's.