Re: [Emc-developers] Ethercat Working Group

[gene heskett]

On 2/18/23 19:28, Rod Webster wrote:

Progress has been slow with  the kernel stuff. I was able to build a non
PREEMPT_RT image for the Odroid N2++ and get it to boot yesterday.
Initially you need to boot from petitboot which sits on flash storage.
Step 2, adding the PREEMPT_RT patches failed after that.  It seems the
armbian kernel 6.1 is a bit behind kernel.org so some of the patches from
November failed.

Kernel 6.2 is released today. My plan is to move to Armbian Edge which
already runs the 6.2 kernel before continuing. I need to get that sorted
bere trying xenomai

Gene, the Armbian build environment is run on an amd64 environment. I have
to say playing with kernels has shown you need a PC with 8-10 cores to get
it done in a reasonable time which is something my 10 year old i7 does not
have!

My rpi4 can build lcnc from git, buster master branch in around an hour 
the last time I fired it off a couple months ago, but probably slower 
today with all the new gfx in the docs. The docs take way more time than 
lcnc.


I haven't actually built a new, running kernel in going on 3 years, but 
I've built what I'm running in a bit over an hour. My build environment 
is on a 240 gig SSD on the usb3 adaptor, so the only traffic for all 
that does not touch the u-sd it boots from unless I install the debs it 
makes for linuxcnc.


This is the kernel I built in Fed 2020:
Linux rpi4.coyote.den 4.19.71-rt24-v7l+ #1 SMP PREEMPT RT Thu Feb 6 
07:09:18 EST 2020 armv7l GNU/Linux


latency-test shows about 12 u-secs for the 1khz servo-thread unless FF 
is running, FF is a pig but you knew that. Even watching youtube the 
lathe rarely stutters. but latency will hit 200 u-secs pretty easy then.


I have a standby and a small ups on the pi, so I determine uptimes, it 
has never crashed,

pi@rpi4:~ $ uptime
 01:38:13 up 124 days, 12:17,  3 users,  load average: 0.46, 0.34, 0.26

I hope the next one I build is as stable.

Anyway, that 240g is about half full and I can clear out whatever it 
take to make a new test mule out of it. Probably 100g of non-precious stuff.



Rod Webster
*1300 896 832*
+61 435 765 611
Vehicle Modifications Network
www.vehiclemods.net.au


On Sun, 19 Feb 2023 at 03:15, gene heskett  wrote:


On 2/18/23 11:31, Small Shop Concepts wrote:

That sounds awesome Rod!  I am interested in your Xenomai 4 test results.

On Fri, Feb 10, 2023 at 3:53 AM Rod Webster  wrote:


I was lucky enough to attend an Ethercat working group seminar today in
Brisbane, Australia where German based Martin Rostan,  Executive

Director

from the EtherCAT Technology Group (ETG) was the main presenter.

It was all very interesting and the detailed low level description of
ethercat packets and alternatives based on TCP makes Ethercat streets in
front of other alternatives because of the sheer efficiency of the
protocol. This let me understand the issues causing the error finishing
read error being experienced with hm2eth connected devices (ie mesa
ethernet cards) in later kernels (5.10 and above).

I was also able to speak with Martin about the licensing issue
preventing us from including the Linuxcnc ethercat driver in our
distribution and have a greater understanding of the problem. Their
license requires users of an ethercat master (in this case the iGh

Etherlab

master we use) to become members of the ETG (which is free). Anybody
selling machines would need to adhere to this. So technically, we can't

put

that condition on to our terms because the existing GPL licence
from Linuxcnc propagates to the ethercat driver. He did go on to say

that

they would never take action against somebody promoting use of ethercat
masters.

So back to the error finishing read problem (that can't be solved by

wide

scale adoption of ethercat), I recently purchased an Odroid N2+ Arm

based

PC which arrived today. I was hoping the different platform may improve
network latency. I have been able to compile an Armbian bookworm image
which booted OK.  A user on the Armbian forum  is willing to help me

with

building a PREEMPT_RT image following this prerequisite first step. but

his

main interest is in Xenomai 4 and he is keen to help me subsequently

build

Xenomai for the N2+.  Xenomai seems to have a significant advantage over
PREEMPT_RT re latency. Given we already support Xenomai 3, I am hoping

it

will be a fairly simple matter to support xenomai 4 and build a test for
network latency to look at differences between xenomai and preempt_rt.

So I am sure progress will be slow, but I am hopeful we will come up

with a

solution to error finishing read!

I hope some of you are interested in this ChatGPT free update!

Rod Webster


I too would be interested in Xenomai 4 as applied to armhf/arm64 stuff.
The kernel I first built as preempt-rt for the pi's has been decent but
is getting long in the tooth and although the installation of it on
armhf stuff is dead simple, its not really well fitting into the apt
update scheme, apt in fact ignores 

Re: [Emc-developers] Ethercat Working Group

[Rod Webster]

Progress has been slow with  the kernel stuff. I was able to build a non
PREEMPT_RT image for the Odroid N2++ and get it to boot yesterday.
Initially you need to boot from petitboot which sits on flash storage.
Step 2, adding the PREEMPT_RT patches failed after that.  It seems the
armbian kernel 6.1 is a bit behind kernel.org so some of the patches from
November failed.

Kernel 6.2 is released today. My plan is to move to Armbian Edge which
already runs the 6.2 kernel before continuing. I need to get that sorted
bere trying xenomai

Gene, the Armbian build environment is run on an amd64 environment. I have
to say playing with kernels has shown you need a PC with 8-10 cores to get
it done in a reasonable time which is something my 10 year old i7 does not
have!

Rod Webster
*1300 896 832*
+61 435 765 611
Vehicle Modifications Network
www.vehiclemods.net.au


On Sun, 19 Feb 2023 at 03:15, gene heskett  wrote:

> On 2/18/23 11:31, Small Shop Concepts wrote:
> > That sounds awesome Rod!  I am interested in your Xenomai 4 test results.
> >
> > On Fri, Feb 10, 2023 at 3:53 AM Rod Webster  wrote:
> >>
> >> I was lucky enough to attend an Ethercat working group seminar today in
> >> Brisbane, Australia where German based Martin Rostan,  Executive
> Director
> >> from the EtherCAT Technology Group (ETG) was the main presenter.
> >>
> >> It was all very interesting and the detailed low level description of
> >> ethercat packets and alternatives based on TCP makes Ethercat streets in
> >> front of other alternatives because of the sheer efficiency of the
> >> protocol. This let me understand the issues causing the error finishing
> >> read error being experienced with hm2eth connected devices (ie mesa
> >> ethernet cards) in later kernels (5.10 and above).
> >>
> >> I was also able to speak with Martin about the licensing issue
> >> preventing us from including the Linuxcnc ethercat driver in our
> >> distribution and have a greater understanding of the problem. Their
> >> license requires users of an ethercat master (in this case the iGh
> Etherlab
> >> master we use) to become members of the ETG (which is free). Anybody
> >> selling machines would need to adhere to this. So technically, we can't
> put
> >> that condition on to our terms because the existing GPL licence
> >> from Linuxcnc propagates to the ethercat driver. He did go on to say
> that
> >> they would never take action against somebody promoting use of ethercat
> >> masters.
> >>
> >> So back to the error finishing read problem (that can't be solved by
> wide
> >> scale adoption of ethercat), I recently purchased an Odroid N2+ Arm
> based
> >> PC which arrived today. I was hoping the different platform may improve
> >> network latency. I have been able to compile an Armbian bookworm image
> >> which booted OK.  A user on the Armbian forum  is willing to help me
> with
> >> building a PREEMPT_RT image following this prerequisite first step. but
> his
> >> main interest is in Xenomai 4 and he is keen to help me subsequently
> build
> >> Xenomai for the N2+.  Xenomai seems to have a significant advantage over
> >> PREEMPT_RT re latency. Given we already support Xenomai 3, I am hoping
> it
> >> will be a fairly simple matter to support xenomai 4 and build a test for
> >> network latency to look at differences between xenomai and preempt_rt.
> >>
> >> So I am sure progress will be slow, but I am hopeful we will come up
> with a
> >> solution to error finishing read!
> >>
> >> I hope some of you are interested in this ChatGPT free update!
> >>
> >> Rod Webster
> >>
> I too would be interested in Xenomai 4 as applied to armhf/arm64 stuff.
> The kernel I first built as preempt-rt for the pi's has been decent but
> is getting long in the tooth and although the installation of it on
> armhf stuff is dead simple, its not really well fitting into the apt
> update scheme, apt in fact ignores it. And while it can & does install
> new kernels, it does not disable mine. So its safe to sudo apt update etc.
>
> I am equipt to build kernels on my rpi4, or linuxcnc.  And I keep
> linuxcnc uptodate from github here. So I can supply another native build
> environment for testing.  Just ask.
>
> >> VMN®
> >>
> >> www.vmn.com.au
> >>
> >> Ph: 1300 896 832
> >>
> >> Mob: +61 435 765 611
> >>
> >> ___
> >> Emc-developers mailing list
> >> Emc-developers@lists.sourceforge.net
> >> https://lists.sourceforge.net/lists/listinfo/emc-developers
> >
> >
> > ___
> > Emc-developers mailing list
> > Emc-developers@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-developers
>
> Cheers, Gene Heskett.
> --
> "There are four boxes to be used in defense of liberty:
>   soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
>   - Louis D. Brandeis
> Genes Web page 

Re: [Emc-developers] Ethercat Working Group

[gene heskett]

On 2/18/23 11:31, Small Shop Concepts wrote:

That sounds awesome Rod!  I am interested in your Xenomai 4 test results.

On Fri, Feb 10, 2023 at 3:53 AM Rod Webster  wrote:


I was lucky enough to attend an Ethercat working group seminar today in
Brisbane, Australia where German based Martin Rostan,  Executive Director
from the EtherCAT Technology Group (ETG) was the main presenter.

It was all very interesting and the detailed low level description of
ethercat packets and alternatives based on TCP makes Ethercat streets in
front of other alternatives because of the sheer efficiency of the
protocol. This let me understand the issues causing the error finishing
read error being experienced with hm2eth connected devices (ie mesa
ethernet cards) in later kernels (5.10 and above).

I was also able to speak with Martin about the licensing issue
preventing us from including the Linuxcnc ethercat driver in our
distribution and have a greater understanding of the problem. Their
license requires users of an ethercat master (in this case the iGh Etherlab
master we use) to become members of the ETG (which is free). Anybody
selling machines would need to adhere to this. So technically, we can't put
that condition on to our terms because the existing GPL licence
from Linuxcnc propagates to the ethercat driver. He did go on to say that
they would never take action against somebody promoting use of ethercat
masters.

So back to the error finishing read problem (that can't be solved by wide
scale adoption of ethercat), I recently purchased an Odroid N2+ Arm based
PC which arrived today. I was hoping the different platform may improve
network latency. I have been able to compile an Armbian bookworm image
which booted OK.  A user on the Armbian forum  is willing to help me with
building a PREEMPT_RT image following this prerequisite first step. but his
main interest is in Xenomai 4 and he is keen to help me subsequently build
Xenomai for the N2+.  Xenomai seems to have a significant advantage over
PREEMPT_RT re latency. Given we already support Xenomai 3, I am hoping it
will be a fairly simple matter to support xenomai 4 and build a test for
network latency to look at differences between xenomai and preempt_rt.

So I am sure progress will be slow, but I am hopeful we will come up with a
solution to error finishing read!

I hope some of you are interested in this ChatGPT free update!

Rod Webster

I too would be interested in Xenomai 4 as applied to armhf/arm64 stuff. 
The kernel I first built as preempt-rt for the pi's has been decent but 
is getting long in the tooth and although the installation of it on 
armhf stuff is dead simple, its not really well fitting into the apt 
update scheme, apt in fact ignores it. And while it can & does install 
new kernels, it does not disable mine. So its safe to sudo apt update etc.


I am equipt to build kernels on my rpi4, or linuxcnc.  And I keep 
linuxcnc uptodate from github here. So I can supply another native build 
environment for testing.  Just ask.



VMN®

www.vmn.com.au

Ph: 1300 896 832

Mob: +61 435 765 611

___
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers



___
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers


Cheers, Gene Heskett.
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 



___
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers

Re: [Emc-developers] Ethercat Working Group

[Small Shop Concepts]

That sounds awesome Rod!  I am interested in your Xenomai 4 test results.

On Fri, Feb 10, 2023 at 3:53 AM Rod Webster  wrote:
>
> I was lucky enough to attend an Ethercat working group seminar today in
> Brisbane, Australia where German based Martin Rostan,  Executive Director
> from the EtherCAT Technology Group (ETG) was the main presenter.
>
> It was all very interesting and the detailed low level description of
> ethercat packets and alternatives based on TCP makes Ethercat streets in
> front of other alternatives because of the sheer efficiency of the
> protocol. This let me understand the issues causing the error finishing
> read error being experienced with hm2eth connected devices (ie mesa
> ethernet cards) in later kernels (5.10 and above).
>
> I was also able to speak with Martin about the licensing issue
> preventing us from including the Linuxcnc ethercat driver in our
> distribution and have a greater understanding of the problem. Their
> license requires users of an ethercat master (in this case the iGh Etherlab
> master we use) to become members of the ETG (which is free). Anybody
> selling machines would need to adhere to this. So technically, we can't put
> that condition on to our terms because the existing GPL licence
> from Linuxcnc propagates to the ethercat driver. He did go on to say that
> they would never take action against somebody promoting use of ethercat
> masters.
>
> So back to the error finishing read problem (that can't be solved by wide
> scale adoption of ethercat), I recently purchased an Odroid N2+ Arm based
> PC which arrived today. I was hoping the different platform may improve
> network latency. I have been able to compile an Armbian bookworm image
> which booted OK.  A user on the Armbian forum  is willing to help me with
> building a PREEMPT_RT image following this prerequisite first step. but his
> main interest is in Xenomai 4 and he is keen to help me subsequently build
> Xenomai for the N2+.  Xenomai seems to have a significant advantage over
> PREEMPT_RT re latency. Given we already support Xenomai 3, I am hoping it
> will be a fairly simple matter to support xenomai 4 and build a test for
> network latency to look at differences between xenomai and preempt_rt.
>
> So I am sure progress will be slow, but I am hopeful we will come up with a
> solution to error finishing read!
>
> I hope some of you are interested in this ChatGPT free update!
>
> Rod Webster
>
> VMN®
>
> www.vmn.com.au
>
> Ph: 1300 896 832
>
> Mob: +61 435 765 611
>
> ___
> Emc-developers mailing list
> Emc-developers@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-developers


___
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers

[Emc-developers] Ethercat Working Group

[Rod Webster]

I was lucky enough to attend an Ethercat working group seminar today in
Brisbane, Australia where German based Martin Rostan,  Executive Director
from the EtherCAT Technology Group (ETG) was the main presenter.

It was all very interesting and the detailed low level description of
ethercat packets and alternatives based on TCP makes Ethercat streets in
front of other alternatives because of the sheer efficiency of the
protocol. This let me understand the issues causing the error finishing
read error being experienced with hm2eth connected devices (ie mesa
ethernet cards) in later kernels (5.10 and above).

I was also able to speak with Martin about the licensing issue
preventing us from including the Linuxcnc ethercat driver in our
distribution and have a greater understanding of the problem. Their
license requires users of an ethercat master (in this case the iGh Etherlab
master we use) to become members of the ETG (which is free). Anybody
selling machines would need to adhere to this. So technically, we can't put
that condition on to our terms because the existing GPL licence
from Linuxcnc propagates to the ethercat driver. He did go on to say that
they would never take action against somebody promoting use of ethercat
masters.

So back to the error finishing read problem (that can't be solved by wide
scale adoption of ethercat), I recently purchased an Odroid N2+ Arm based
PC which arrived today. I was hoping the different platform may improve
network latency. I have been able to compile an Armbian bookworm image
which booted OK.  A user on the Armbian forum  is willing to help me with
building a PREEMPT_RT image following this prerequisite first step. but his
main interest is in Xenomai 4 and he is keen to help me subsequently build
Xenomai for the N2+.  Xenomai seems to have a significant advantage over
PREEMPT_RT re latency. Given we already support Xenomai 3, I am hoping it
will be a fairly simple matter to support xenomai 4 and build a test for
network latency to look at differences between xenomai and preempt_rt.

So I am sure progress will be slow, but I am hopeful we will come up with a
solution to error finishing read!

I hope some of you are interested in this ChatGPT free update!

Rod Webster

VMN®

www.vmn.com.au

Ph: 1300 896 832

Mob: +61 435 765 611

___
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers