ThomasKaiser
commented
2 years ago the 'missed opportunity'
Thank you for the additional testing (already processing the data). I just didn't respond to you already since asking again for another test (single channel, single rank) seemed a bit too annoying to me. :)
Thank you a lot for the thermal input. I integrated it as follows on x86 (colorized diff):
- try to get the
x86_pkg_tempsysfsnode - if this fails use your code to search for
$HWMON/temp1_input - if this also fails report at least any sysfs node if existent or otherwise use
0
Would be great if you can give it a try on an AMD box: http://ix.io/3KYC (sbc-bench -m is already sufficient for a test)
On two Intel thingies it looks like this:
root@gaia:~# sbc-bench.sh -m
Kernel: x86_64, Userland: amd64
CPU sysfs topology (clusters, cpufreq members, clockspeeds)
cpufreq min max
CPU cluster policy speed speed core type
0 0 0 800 2800 Celeron N5100 @ 1.10GHz
1 0 1 800 2800 Celeron N5100 @ 1.10GHz
2 0 2 800 2800 Celeron N5100 @ 1.10GHz
3 0 3 800 2800 Celeron N5100 @ 1.10GHz
Thermal source: /sys/devices/virtual/thermal/thermal_zone4/temp
Time CPU load %cpu %sys %usr %nice %io %irq Temp
13:10:26: 884MHz 0.24 13% 3% 9% 0% 0% 0% 42.0°C
13:10:31: 595MHz 0.22 3% 0% 2% 0% 0% 0% 40.0°C
^C
root@gaia:~# cat /sys/devices/virtual/thermal/thermal_zone4/type
x86_pkg_temp
root@gaia:~# cat /sys/class/hwmon/hwmon2/temp1_input
40000
tk@coloc-1:~$ sbc-bench.sh -m
Kernel: x86_64, Userland: amd64
CPU sysfs topology (clusters, cpufreq members, clockspeeds)
cpufreq min max
CPU cluster policy speed speed core type
0 0 0 800 3000 Xeon Silver 4110 @ 2.10GHz
1 0 1 800 3000 Xeon Silver 4110 @ 2.10GHz
2 0 2 800 3000 Xeon Silver 4110 @ 2.10GHz
3 0 3 800 3000 Xeon Silver 4110 @ 2.10GHz
4 0 4 800 3000 Xeon Silver 4110 @ 2.10GHz
5 0 5 800 3000 Xeon Silver 4110 @ 2.10GHz
6 0 6 800 3000 Xeon Silver 4110 @ 2.10GHz
7 0 7 800 3000 Xeon Silver 4110 @ 2.10GHz
8 0 8 800 3000 Xeon Silver 4110 @ 2.10GHz
9 0 9 800 3000 Xeon Silver 4110 @ 2.10GHz
10 0 10 800 3000 Xeon Silver 4110 @ 2.10GHz
11 0 11 800 3000 Xeon Silver 4110 @ 2.10GHz
12 0 12 800 3000 Xeon Silver 4110 @ 2.10GHz
13 0 13 800 3000 Xeon Silver 4110 @ 2.10GHz
14 0 14 800 3000 Xeon Silver 4110 @ 2.10GHz
15 0 15 800 3000 Xeon Silver 4110 @ 2.10GHz
Thermal source: /sys/devices/virtual/thermal/thermal_zone0/temp
Time CPU load %cpu %sys %usr %nice %io %irq Temp
13:18:31: 2659MHz 0.82 4% 1% 2% 0% 0% 0% 35.0°C
13:18:36: 2635MHz 0.75 5% 1% 2% 0% 0% 0% 36.0°C
^C
tk@coloc-1:~$ cat /sys/devices/virtual/thermal/thermal_zone0/type
x86_pkg_temp
tk@coloc-1:~$ cat /sys/class/hwmon/hwmon0/temp1_input
36000
dan-and
linuxium
After realising the 'missed opportunity' (see my latest comment on 'OS and Memory Impact on Mini PC Gaming Performance') I found that the CPU temperature during system health was not reflected correctly in the latest version of ‘sbc-bench’.
As a temporary work-around I ran the following script:
if $(lsmod | grep -qsw "^k10temp"); then # AMD HWMON=$(grep --include=name -rls /sys/devices/pci -e "^k10temp$" | sed 's?.hwmon/??' | sed 's?/name??') elif $(lsmod | grep -qsw "^coretemp"); then # Intel HWMON=$(ls /sys/devices/platform/coretemp.0/hwmon) else HWMON="" fi if [ -f "/sys/class/hwmon/${HWMON}/temp1_input" ]; then echo "sed -i 's?echo 0 >?ln -fs /sys/class/hwmon/${HWMON}/temp1_input ?' sbc-bench.sh" else echo "Cannot determine 'hwmon'" fi
I'm not suggesting this as 'the fix' but rather using it to highlight where the issue is and one possible way of addressing it.