Error while Running Linuxcnc

What CPU is used?

RPI 4 will work if you force turbo mode and use isolcpus 1,2,3
(but 2G will be an issue if you have fancy backplots)

CPU is as follows:
==================================================
raycnc1@debian:~/Desktop$ lscpu
Architecture:            x86_64
  CPU op-mode(s):        32-bit, 64-bit
  Address sizes:         39 bits physical, 48 bits virtual
  Byte Order:            Little Endian
CPU(s):                  2
  On-line CPU(s) list:   0,1
Vendor ID:               GenuineIntel
  Model name:            Intel(R) Core(TM) i3-6100T CPU @ 3.20GHz
    CPU family:          6
    Model:               94
    Thread(s) per core:  1
    Core(s) per socket:  2
    Socket(s):           1
    Stepping:            3
    CPU(s) scaling MHz:  100%
    CPU max MHz:         3200.0000
    CPU min MHz:         800.0000
    BogoMIPS:            6399.96
    Flags:               fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mc
                         a cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss
                         ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art
                          arch_perfmon pebs bts rep_good nopl xtopology nonstop_
                         tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cp
                         l vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1
                          sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsav
                         e avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault
                          invpcid_single pti ssbd ibrs ibpb stibp tpr_shadow vnm
                         i flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1
                          avx2 smep bmi2 erms invpcid mpx rdseed adx smap clflus
                         hopt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm ara
                         t pln pts hwp hwp_notify hwp_act_window hwp_epp md_clea
                         r flush_l1d arch_capabilities
Virtualization features:
  Virtualization:        VT-x
Caches (sum of all):     
  L1d:                   64 KiB (2 instances)
  L1i:                   64 KiB (2 instances)
  L2:                    512 KiB (2 instances)
  L3:                    3 MiB (1 instance)
NUMA:                    
  NUMA node(s):          1
  NUMA node0 CPU(s):     0,1
Vulnerabilities:         
  Gather data sampling:  Vulnerable: No microcode
  Itlb multihit:         KVM: Mitigation: VMX disabled
  L1tf:                  Mitigation; PTE Inversion; VMX conditional cache flushe
                         s, SMT disabled
  Mds:                   Mitigation; Clear CPU buffers; SMT disabled
  Meltdown:              Mitigation; PTI
  Mmio stale data:       Mitigation; Clear CPU buffers; SMT disabled
  Retbleed:              Mitigation; IBRS
  Spec rstack overflow:  Not affected
  Spec store bypass:     Mitigation; Speculative Store Bypass disabled via prctl
  Spectre v1:            Mitigation; usercopy/swapgs barriers and __user pointer
                          sanitization
  Spectre v2:            Mitigation; IBRS, IBPB conditional, RSB filling, PBRSB-
                         eIBRS Not affected
  Srbds:                 Mitigation; Microcode
  Tsx async abort:       Not affected
raycnc1@debian:~/Desktop$
=======================================================
I did some  test as follows:
=======================================================
raycnc1@debian:~/Desktop$ lshw -class network
bash: lshw: command not found

raycnc1@debian:~/Desktop$ lspci | egrep -i 'network|ethernet'
04:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 16)


raycnc1@debian:~/Desktop$ ping -i .2 -c 4 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.
64 bytes from 10.10.10.10: icmp_seq=1 ttl=64 time=0.118 ms
64 bytes from 10.10.10.10: icmp_seq=2 ttl=64 time=0.065 ms
64 bytes from 10.10.10.10: icmp_seq=3 ttl=64 time=0.063 ms
64 bytes from 10.10.10.10: icmp_seq=4 ttl=64 time=0.063 ms

--- 10.10.10.10 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 609ms
rtt min/avg/max/mdev = 0.063/0.077/0.118/0.023 ms
raycnc1@debian:~/Desktop$ sudo chrt 99 ping -i .001 -q 10.10.10.10
[sudo] password for raycnc1:
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.

--- 10.10.10.10 ping statistics ---
246632 packets transmitted, 246631 received, 0.000405462% packet loss, time 248878ms
rtt min/avg/max/mdev = 0.046/0.049/0.102/0.003 ms
^C
raycnc1@debian:~/Desktop$

raycnc1@debian:~/Desktop$ ping 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.
64 bytes from 10.10.10.10: icmp_seq=1 ttl=64 time=0.078 ms
64 bytes from 10.10.10.10: icmp_seq=2 ttl=64 time=0.078 ms
64 bytes from 10.10.10.10: icmp_seq=3 ttl=64 time=0.063 ms
64 bytes from 10.10.10.10: icmp_seq=4 ttl=64 time=0.070 ms
64 bytes from 10.10.10.10: icmp_seq=5 ttl=64 time=0.063 ms
64 bytes from 10.10.10.10: icmp_seq=6 ttl=64 time=0.073 ms
64 bytes from 10.10.10.10: icmp_seq=7 ttl=64 time=0.065 ms
64 bytes from 10.10.10.10: icmp_seq=8 ttl=64 time=0.063 ms
64 bytes from 10.10.10.10: icmp_seq=9 ttl=64 time=0.063 ms
64 bytes from 10.10.10.10: icmp_seq=10 ttl=64 time=0.063 ms
^C
--- 10.10.10.10 ping statistics ---
338 packets transmitted, 338 received, 0% packet loss, time 345086ms
rtt min/avg/max/mdev = 0.062/0.066/0.152/0.007 ms
raycnc1@debian:~/Desktop$

============================================================


Please confirm whether this CPU can be used or not.

I need a reliable performance and that is the ultimate goal.

One confusion I could not clear..My CPU is Intel i3 6100T.

www.intel.com/content/www/us/en/products.../specifications.html

As per this link, it has 2 core. The "lscpu" result is just in above post.

But in latency histogram, core=4. (attached)

Why??

Do you have Hyper-threading enabled? Hyper-threading creates two virtual cores for each real core, so that a 2 core CPU is seen as 4. Hyperthreading gives poor latency results with real time, turn it off or disable it in your bios.

Yes, hyper threading was on. I disabled it. Also turbo boost was on, now disabled. But there is very negligible improvement in latency.

However, should I set the IRQ affinity to CPU=1? It is a realtek network adapter inbuilt in motherboard.

Turbo boost by itself can cause huge latency spikes

You need to install the R8168_dkms driver for your NIC. Also set the CPU affinity per a script on this forum. There are also some settings to deal with energy efficient Ethernet. I cover it all in my help doc
docs.google.com/document/d/1jeV_4VKzVmOI...diY/edit?usp=sharing

I tries ip-a in terminal and result below:
=========================================================
raycnc1@debian:~/Desktop$ ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host noprefixroute
       valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP group default qlen 1000
    link/ether 18:c0:4d:78:76:3b brd ff:ff:ff:ff:ff:ff
    inet 10.10.10.11/24 brd 10.10.10.255 scope global noprefixroute enp4s0
       valid_lft forever preferred_lft forever
3: wlx20e616011575: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether 20:e6:16:01:15:75 brd ff:ff:ff:ff:ff:ff
    inet 192.168.38.167/24 brd 192.168.38.255 scope global dynamic noprefixroute wlx20e616011575
       valid_lft 2357sec preferred_lft 2357sec
    inet6 2409:4060:29a:55eb:f28:5760:799e:7353/64 scope global dynamic noprefixroute
       valid_lft 6913sec preferred_lft 6913sec
    inet6 fe80::c85e:7991:888d:a126/64 scope link noprefixroute
       valid_lft forever preferred_lft forever
raycnc1@debian:~/Desktop$ ethtool enp4s0
bash: ethtool: command not found
raycnc1@debian:~/Desktop$
==========================================================

But ethool command does not work.

Note: My computer have wireless keyboard+mouse and one wifi router in usb port.




 

You might try "locate ethtool" ethtool may be in /sbin on Debian.
_But_ if you have Realtek Ethernet hardware, Ethtool is not needed.
What _is_ required for Realtek hardware  is to install the 8168 DKMS
driver as rodw mentioned.

The locate ethtool result as follows:
==============================================================
raycnc1@debian:~/Desktop$ locate ethtool
/usr/include/linux/ethtool.h
/usr/include/linux/ethtool_netlink.h
/usr/src/linux-headers-6.1.0-18-common/include/linux/ethtool.h
/usr/src/linux-headers-6.1.0-18-common/include/linux/ethtool_netlink.h
/usr/src/linux-headers-6.1.0-18-common/include/uapi/linux/ethtool.h
/usr/src/linux-headers-6.1.0-18-common/include/uapi/linux/ethtool_netlink.h
/usr/src/linux-headers-6.1.0-18-common-rt/include/linux/ethtool.h
/usr/src/linux-headers-6.1.0-18-common-rt/include/linux/ethtool_netlink.h
/usr/src/linux-headers-6.1.0-18-common-rt/include/uapi/linux/ethtool.h
/usr/src/linux-headers-6.1.0-18-common-rt/include/uapi/linux/ethtool_netlink.h
raycnc1@debian:~/Desktop$

=================================================================

Now as per my last post above today (ip a result) , I assumed that my ethernet card is "enp4s0". I am not sure whether I am right, if wrong , please reply.
And I set the IRQ affinity as per this post of course:
forum.linuxcnc.org/38-general-linuxcnc-q...ead-and-irq-affinity

The "journalct1 -b -u irqbalance.service" resulted some command error.

But "watch -n1 -d cat /proc/interrupts" resulted output which is attached in photo (as irq number changes continuously, not possible to copy/paste).

In that result CPU0 is handling all the Interrupts while running LCNC(182008 in exact in photo) for "enp4s0". Is that correct? I was expecting exactly opposite, interrupts to be in CPU1. I have done isolcpu=1. histogram attached.