Closed LouisCyfer closed 3 years ago
Hi, I don't know if it's just me or my lack of experience in utilizing bash-scripts.
I really love the idea of this tool and need itto have any idea about my CPU temperature (lm-sensors doesn't work well for my AMD Ryzen 2700).
I figured the option -C prints out a sensor monitor continuously. is there a way to make that one single-print so it can be used like grepping a value from sensors?
Hi
It is made to be continuous but I'm adding your idea into the todo list.
Meanwhile, you can use the JSON output, argument -j
There are a lot of data but piping through a json parser, you could limit the output to the temperature values. Those are mainly per CPU for each core.
Hello,
For your testings, the enhancement is available in the develop
branch.
corefreq-cli -h
CoreFreq. Copyright (C) 2015-2021 CYRIL INGENIERIE
Usage: corefreq-cli [-option
Interface options
-0,1,2 Memory unit in K,M,G Byte
-F Temperature in Fahrenheit
-J # SMBIOS string index number
-Y Show Secret Data
-t <v> Show Top (default) with optional <v>iew:
{ frequency, instructions, core, idle, package,
tasks, interrupts, sensors, voltage, power, slices }
-d Show Dashboard
Command options
-C <#> Monitor Sensors
-V <#> Monitor Voltage
-W <#> Monitor Power
-g <#> Monitor Package
-c <#> Monitor Counters
-i <#> Monitor Instructions
-s Print System Information
-j Print System Information (json-encoded)
-M Print Memory Controller
-R Print System Registers
-m Print Topology
-u Print CPUID
-B Print SMBIOS
-k Print Kernel
-h Print out this message
-v Print the version number
Exit status: 0 SUCCESS Successful execution 3 CMD_SYNTAX Command syntax error 4 SHM_FILE Shared memory file error 5 SHM_MMAP Shared memory mapping error 6 PERM_ERR Execution not permitted 7 MEM_ERR Memory operation error 8 EXEC_ERR General execution error 9 SYS_CALL System call error
Report bugs to labs[at]cyring.fr
* To make a single-print from the sensor monitor:
corefreq-cli -C 1 CPU Freq(MHz) VID Vcore TMP(C) Accumulator Energy(J) Power(W) 000 4.84 103 0.9062 30 000000000000000349 0.005325317 0.005325317 001 4.35 103 0.9062 30 000000000000000201 0.003067017 0.003067017 002 4.06 101 0.9187 30 000000000000000185 0.002822876 0.002822876 003 4.39 101 0.9187 30 000000000000000178 0.002716064 0.002716064 ... 031 3.80 101 0.9187 31 000000000000000000 0.000000000 0.000000000
Package Cores Uncore Memory
Energy(J): 16.452529907 0.051834106 16.002548218 0.000000000 Power(W) : 16.452529907 0.051834106 16.002548218 0.000000000
* Going further, you can also combine various iterations
corefreq-cli -C 3 -V 2 -W 5
> which means 3 Sensors loops, 2 Voltages loops, 5 Power loops
Thanks for any comment.
That is awesome, ty for the insight. This helps alot!
Have you checked sensors
recently? The module k10temp
, after some kernel updates, has added more and more, and now shows a lot of values for at least X570 and any Ryzen 3 CPU. A bit strange if it doesn't go for the 2XXX CPUs as well?
I get all this from my 3950X, ccds and everything:
k10temp-pci-00c3
Adapter: PCI adapter
Vcore: 1.48 V
Vsoc: 1.08 V
Tctl: +35.4°C
Tdie: +35.4°C
Tccd1: +34.0°C
Tccd2: +37.5°C
Icore: 14.00 A
Isoc: 8.50 A
I don't discourage the use of CoreFreq of course. It's awesome 👍 But for simple grepping etc, just using the built-in kernel module is simpler.
NOTE: There's NO reason to run # sensors-detect
. As long as k10temp
loads at boot it just works, as with Coretemp for Intel.
Closing the issue.
As a reminder for anyone reading this: don't run any k10temp
or any other sensors driver in the same time than CoreFreq. This will overwhelm the SMU of Ryzen, EPYC, Threadripper processors which may freeze sooner or later.
Hi, I don't know if it's just me or my lack of experience in utilizing bash-scripts.
I really love the idea of this tool and need itto have any idea about my CPU temperature (lm-sensors doesn't work well for my AMD Ryzen 2700).
I figured the option -C prints out a sensor monitor continuously. is there a way to make that one single-print so it can be used like grepping a value from sensors?