Thomas Bitsky Jr | 27 Jun 03:40 2013

Cross Compile for ARM

 

I’m trying to cross-compile the EtherCAT master for an ARM board. I successfully built a Linux RT Prempt kernel for the device.  On my development machine, the source tree where I built the kernel is at: ~/srcroot/linux/linux-3.2.46

 

I moved into the EtherLAB directory and issued:

 

./configure --host=arm-linux --enable-generic --with-linux-dir=~/srcroot/linux/linux-3.2.46

 

The result is:

 

No configured Linux kernel sources in ~/srcroot/linux/linux-3.2.46/

 

But I just built a kernel there. Is there something I have to do to the source tree to get it ready to build the ethercat master?

 

Thanks in advance,

Tome

 

<div>
<div class="WordSection1">
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">I&rsquo;m trying to cross-compile the EtherCAT master for an ARM board. I successfully built a Linux RT Prempt kernel for the device. &nbsp;On my development machine, the source tree where I built the kernel is at: ~/srcroot/linux/linux-3.2.46<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">I moved into the EtherLAB directory and issued:<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">./configure --host=arm-linux --enable-generic --with-linux-dir=~/srcroot/linux/linux-3.2.46<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">The result is:<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">No configured Linux kernel sources in ~/srcroot/linux/linux-3.2.46/<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">But I just built a kernel there. Is there something I have to do to the source tree to get it ready to build the ethercat master?<p></p></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
<p class="MsoNormal">Thanks in advance,<p></p></p>
<p class="MsoNormal">Tome<span><p></p></span></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
</div>
</div>
Graeme Foot | 20 Jun 00:42 2013

Re: How to install and use Etherlab for a Xenomaitarget

Hi,

 

We are using a Beckhoff CX2020 computer.  We are running it as a non-gui controller with:

 

Linux 2.6.32.11

RTAI 3.8.1

EtherLabs 1.5.2

 

The CX2020 has two e1000e NICs for the office network.  The EtherCAT NIC is part of the CX2100-0004 power module but this is specialised hardware and we needed to write our own driver for it (submitted to the etherlab forum a while ago).

 

We are building our Linux environment using Buildroot so that it is very small.

 

 

Linux 2.6.32 does not support the graphics card for this PC but we are running it headless so not a problem for us.  In the future we are intending to integrate our gui onto the same PC but we are waiting on RTAI and a few other things to come together first.

 

 

Regards,

 

Graeme Foot

 

From: Steffen Dalgard [mailto:steffen.sint-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org]
Sent: Thursday, 20 June 2013 00:42
To: Graeme Foot
Subject: Re: [etherlab-users] How to install and use Etherlab for a Xenomaitarget

 

Hi,

Thank you for your feedback,

We have tried to find a release supporting newer HW that has support for Ehterlab, Ubuntu and Xenomai.

It seems to be hard ot find, maybe a dead end.

  We found working Xenomai using Ubuntu 12.04 with a patched 3.5 kernel, but it is not supported by Etherlab

 

We do need a kernel with RT support but not necessarily Xenomai...

 

You wrote that you were using Etherlab master via RTDM, but using RTAI.

Can you share which releases / kernels that you are using?

 

Best regards

Steffen Dalgard

SINTEF ICT

Norway

 

On Thu, May 16, 2013 at 1:11 AM, Graeme Foot <GraemeF-/dx94a6pNkxWk0Htik3J/w@public.gmane.org> wrote:

 

Hi,

 

I have a user space application connecting to the Etherlab master via RTDM, but using RTAI.  In theory xenomai should be pretty similar.

 

If you are using the latest 1.5.2 Etherlab master you won't need a patch.  The patch was to give partial RTDM support before it was fully supported.

 

 

To get RTDM working you will need to specify the correct flags when compiling the master:

  --with-xenomai-dir=<xenomai directory>

  --enable-rtdm

 

All going well you should get a library: /usr/lib/libethercat_rtdm.so

 

When compiling your application link to the library using: -lethercat_rtdm

 

 

With RTAI you can check your application is using RTDM by calling 'cat /proc/rtai/scheduler'.  The syscalls value should remain zero.

 

 

Regards,

Graeme Foot.

 

From: etherlab-users-bounces-TgCWn71uKPpg9hUCZPvPmw@public.gmane.org [mailto:etherlab-users-bounces-TgCWn71uKPpg9hUCZPvPmw@public.gmane.org] On Behalf Of Steffen Dalgard
Sent: Wednesday, 15 May 2013 21:50
To: etherlab-users-TgCWn71uKPpg9hUCZPvPmw@public.gmane.org
Subject: [etherlab-users] How to install and use Etherlab for a Xenomaitarget

 

Hi,

We are trying to install Etherlab on a Xenomai target together with ROS and OROCOS.

The plan is to have a user space rt-application calling the Igh master.

 

Does it exist a procedure for how to get this working?

 

I have tried to read through the mail archive and the examples, but have only found piecewise information.

   I have found mails mentioning patching of driver connecting Etherlab master to user space

   It seems to be special lib libetherkat_rtdm.so

 

Hope someone can help :-)

 

Best regards

Steffen Dalgard

SINTEF ICT

Norway

 

 

<div>

<div class="Section1">

<p class="MsoNormal"><span>Hi,<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>We are using a Beckhoff CX2020
computer.&nbsp; We are running it as a non-gui controller with:<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>Linux 2.6.32.11<p></p></span></p>

<p class="MsoNormal"><span>RTAI 3.8.1<p></p></span></p>

<p class="MsoNormal"><span>EtherLabs 1.5.2<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>The CX2020 has two e1000e NICs for the
office network.&nbsp; The EtherCAT NIC is part of the CX2100-0004 power module
but this is specialised hardware and we needed to write our own driver for it
(submitted to the etherlab forum a while ago).<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>We are building our Linux environment
using Buildroot so that it is very small.<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>Linux 2.6.32 does not support the graphics
card for this PC but we are running it headless so not a problem for us.&nbsp; In
the future we are intending to integrate our gui onto the same PC but we are
waiting on RTAI and a few other things to come together first.<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>Regards,<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span>Graeme Foot<p></p></span></p>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<div>

<div class="MsoNormal" align="center"><span lang="EN-US">

</span></div>

<p class="MsoNormal"><span lang="EN-US">From:</span><span lang="EN-US">
Steffen Dalgard [mailto:steffen.sint@...] <br><span>Sent:</span> Thursday, 20 June 2013 00:42<br><span>To:</span> Graeme Foot<br><span>Subject:</span> Re: [etherlab-users] How
to install and use Etherlab for a Xenomaitarget</span><span lang="EN-US"><p></p></span></p>

</div>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<div>

<div>

<p class="MsoNormal"><span>Hi,<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Thank you for your feedback,<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>We have tried to find a&nbsp;release supporting newer HW that has
support for Ehterlab, Ubuntu and Xenomai.<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>It seems to be hard ot find, maybe a dead end.<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp; We found working Xenomai using Ubuntu 12.04 with a patched 3.5
kernel, but it is not supported by Etherlab<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>We do need a kernel with RT support but not necessarily Xenomai...<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>You wrote that you were using Etherlab master via RTDM, but using RTAI.<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Can you share which&nbsp;releases / kernels that you are using?<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Best regards<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Steffen Dalgard<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>SINTEF ICT<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Norway</span><p></p></p>

</div>

</div>

<div>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<div>

<p class="MsoNormal"><span>On Thu, May 16, 2013 at 1:11 AM, Graeme Foot &lt;<a href="mailto:GraemeF@..." target="_blank">GraemeF@...</a>&gt;
wrote:<p></p></span></p>

<div>

<p class="MsoNormal"><span><a href="https://www.boxbe.com/overview" target="_blank"><span></span></a>Graeme Foot (<a href="mailto:GraemeF@..." target="_blank">GraemeF@...</a>)
is not on <a href="https://www.boxbe.com/approved-list?tc_serial=14152444570&amp;tc_rand=266417122&amp;utm_source=stf&amp;utm_medium=email&amp;utm_campaign=ANNO_MWTP&amp;utm_content=001&amp;token=Hjk5w06MCHsim%2BkE%2BXEMZbTtwqA%2BTXXMJ7spABspVmqlfWvM7lNh80hlUB%2F00a8M&amp;key=eJnd8Q%2BeCThkw7PVpCigMAaxHAX%2BsHy%2FiQ0MxxEI0Ik%3D" target="_blank"><span>your Guest List</span></a> | <a href="https://www.boxbe.com/anno?tc_serial=14152444570&amp;tc_rand=266417122&amp;utm_source=stf&amp;utm_medium=email&amp;utm_campaign=ANNO_MWTP&amp;utm_content=001&amp;token=Hjk5w06MCHsim%2BkE%2BXEMZbTtwqA%2BTXXMJ7spABspVmqlfWvM7lNh80hlUB%2F00a8M&amp;key=eJnd8Q%2BeCThkw7PVpCigMAaxHAX%2BsHy%2FiQ0MxxEI0Ik%3D" target="_blank"><span>Approve sender</span></a> | <a href="https://www.boxbe.com/anno?tc_serial=14152444570&amp;tc_rand=266417122&amp;utm_source=stf&amp;utm_medium=email&amp;utm_campaign=ANNO_MWTP&amp;utm_content=001&amp;dom&amp;token=Hjk5w06MCHsim%2BkE%2BXEMZbTtwqA%2BTXXMJ7spABspVmqlfWvM7lNh80hlUB%2F00a8M&amp;key=eJnd8Q%2BeCThkw7PVpCigMAaxHAX%2BsHy%2FiQ0MxxEI0Ik%3D" target="_blank"><span>Approve domain</span></a> <p></p></span></p>

</div>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

<div vlink="purple" link="blue">

<div>

<p class="MsoNormal"><span>Hi,</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>I have a user space application connecting to the Etherlab master
via RTDM, but using RTAI.&nbsp; In theory xenomai should be pretty similar.</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>If you are using the latest 1.5.2 Etherlab master you won't need a
patch.&nbsp; The patch was to give partial RTDM support before it was fully
supported.</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>To get RTDM working you will need to specify the correct flags when
compiling the master:</span><p></p></p>

<p class="MsoNormal"><span>&nbsp; --with-xenomai-dir=&lt;xenomai directory&gt;</span><p></p></p>

<p class="MsoNormal"><span>&nbsp; --enable-rtdm</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>All going well you should get a library:
/usr/lib/libethercat_rtdm.so</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>When compiling your application link to the library using:
-lethercat_rtdm</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>With RTAI you can check your application is using RTDM by calling
'cat /proc/rtai/scheduler'.&nbsp; The syscalls value should remain zero.</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<p class="MsoNormal"><span>Regards,</span><p></p></p>

<p class="MsoNormal"><span>Graeme Foot.</span><p></p></p>

<p class="MsoNormal"><span>&nbsp;</span><p></p></p>

<div>

<div class="MsoNormal" align="center"><span lang="EN-US">

</span></div>

<p class="MsoNormal"><span lang="EN-US">From:</span><span lang="EN-US"> <a href="mailto:etherlab-users-bounces@..." target="_blank">etherlab-users-bounces@...</a>
[mailto:<a href="mailto:etherlab-users-bounces@..." target="_blank">etherlab-users-bounces@...</a>]
<span>On Behalf Of </span>Steffen Dalgard<br><span>Sent:</span> Wednesday, 15 May 2013 21:50<br><span>To:</span> <a href="mailto:etherlab-users@..." target="_blank">etherlab-users@...</a><br><span>Subject:</span> [etherlab-users] How to
install and use Etherlab for a Xenomaitarget</span><p></p></p>

</div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

<div>

<div>

<p class="MsoNormal"><span>Hi,<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>We are trying to
install Etherlab on a Xenomai target together with ROS and OROCOS.<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>The plan is to
have a user space rt-application calling the Igh master.<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Does it exist a
procedure for how to get this working?<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>I have tried to
read through the mail archive and the examples, but have only found piecewise information.<p></p></span></p>

</div>

<div>

<div>

<p class="MsoNormal"><span>&nbsp;&nbsp; I
have found mails mentioning patching of driver connecting Etherlab master to
user space<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;&nbsp; It
seems to be special lib </span><span>libetherkat_rtdm.so</span><p></p></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Hope someone can
help :-)<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>&nbsp;<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Best regards<p></p></span></p>

</div>

</div>

<div>

<p class="MsoNormal"><span>Steffen Dalgard<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>SINTEF ICT<p></p></span></p>

</div>

<div>

<p class="MsoNormal"><span>Norway</span><p></p></p>

</div>

</div>

</div>

</div>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

</div>

<p class="MsoNormal"><span><p>&nbsp;</p></span></p>

</div>

</div>

</div>
Hassan FIdai | 19 Jun 19:22 2013
Picon

Problem in running etherlab

Hi, 
I am trying to run etherLAB. I have installed RTAI, ethercat master 1.5.2 and etherLAB 1.3-rc3. I am using ubuntu 12.04. But now when I try to run it i get the following error

Starting EtherLab AppCoreinsmod: error inserting '/usr/realtime/modules/rtai_up.ko': -1 Invalid module format

Can someone please help me out with this?

Regards

--
M. Hassan Fidai
<div><div dir="ltr">
<span>Hi,&nbsp;</span><div>I am trying to run etherLAB. I have installed RTAI, ethercat master 1.5.2 and etherLAB 1.3-rc3. I am using ubuntu 12.04. But now when I try to run it i get the following error</div>
<div><br></div>
<div>
<div>Starting EtherLab AppCoreinsmod: error inserting '/usr/realtime/modules/rtai_up.ko': -1 Invalid module format</div>
<div><br></div>
<div>Can someone please help me out with this?</div>
<div><br></div>
<div>Regards</div>
</div>
<div><br></div>-- <br>M. Hassan Fidai
</div></div>
Gianluca Medini | 19 Jun 16:35 2013
Picon

LRW datagram on TI-ICE development board

Hi all,
I'm playing with IGH on a Texas instruments ICE development kit.
The process image has input (5 bytes) and outputs (7 bytes) , so the 
master configures FMMU in order to have a single datagram LRW.
The data size of the LRW is 12 bytes, although should optimize to 7; 
I've analyzed the wireshark capture of  twincat on the same hardware and 
it make this optimization (an image of 7 bytes).
With this not optimized configuration of FMMU the boards doesn't work (I 
think is problem on the ESC controller, I've checked this), however I 
was wondering if making this optimization could be a desiderable behaviour.
There is some compilation switch / parameter on IGH  that  can modify 
this behaviour ?

Regards
Gianluca Medini

-- 
Dott. Gianluca Medini
Eurosoft S.r.l.
Via Ettore Majorana, 40
Zona Industriale Fontanelle
Località CAPANNE
Montopoli Valdarno PI
ufficio (+39) 0571 261919
mobile : (+39) 335 7001588

Qualora questo messaggio fosse da Voi ricevuto per errore, vogliate cortesemente darcene notizia a mezzo
fax (oppure e-mail) e distruggere il messaggio stesso. Qualsiasi utilizzo e/o conservazione dei dati
ricevuti per errore, costituisce violazione delle disposizioni del D.Lgs 196/03 sulla tutela dei Dati Personali.

Attachment (gianluca.vcf): text/x-vcard, 276 bytes
Hi all,
I'm playing with IGH on a Texas instruments ICE development kit.
The process image has input (5 bytes) and outputs (7 bytes) , so the 
master configures FMMU in order to have a single datagram LRW.
The data size of the LRW is 12 bytes, although should optimize to 7; 
I've analyzed the wireshark capture of  twincat on the same hardware and 
it make this optimization (an image of 7 bytes).
With this not optimized configuration of FMMU the boards doesn't work (I 
think is problem on the ESC controller, I've checked this), however I 
was wondering if making this optimization could be a desiderable behaviour.
There is some compilation switch / parameter on IGH  that  can modify 
this behaviour ?

Regards
Gianluca Medini

--

-- 
Dott. Gianluca Medini
Eurosoft S.r.l.
Via Ettore Majorana, 40
Zona Industriale Fontanelle
Località CAPANNE
Montopoli Valdarno PI
ufficio (+39) 0571 261919
mobile : (+39) 335 7001588

Qualora questo messaggio fosse da Voi ricevuto per errore, vogliate cortesemente darcene notizia a mezzo
fax (oppure e-mail) e distruggere il messaggio stesso. Qualsiasi utilizzo e/o conservazione dei dati
ricevuti per errore, costituisce violazione delle disposizioni del D.Lgs 196/03 sulla tutela dei Dati Personali.

Steffen Dalgard | 19 Jun 15:39 2013
Picon

Ethernet driver support for kernel 3.5

Hi,

We have been able to get Ubuntu 12.04 with Xenomai patch to run at our target.

The procedure unvolves using a 3.5 kernel (not the original 3.2 kernel).

  Details can be found at http://www.stanley.gatech.edu/skelly/uncategorized/installing-xenomai-on-ubuntu-12-04/

 

The 3.5 kernel is now the default kernel for Ubuntu 12.04.2 so I guess it will be there for a while

 

When running the ./configure tool it fails

./configure --with-linux-dir=/lib/modules/3.5.7-xenomai-2.6.2.1/source --disable-8139too --with-xenomai-dir=$xenomai_root --enable-rtdm --enable-e1000e

error: kernel 3.5 not available for e1000e driver!

if i take out the e1000e option it configures but on the make install it fails...

Anyone played with Etherlab and kernel 3.5?

Anyone having Xenomai running at kernel 3.2?

Is it worth trying compiling the 3.4 drivers with the 3.5 kernel? 

Best regards
Steffen Dalgard
SINTEF ICT
Norway

<div><div dir="ltr">

<p>Hi,</p>

<p>We have been able to get Ubuntu 12.04 with Xenomai patch to
run at our target.</p>

<p>The procedure unvolves using a 3.5 kernel (not the original
3.2 kernel).</p>

<p>&nbsp; Details can be found at <a href="http://www.stanley.gatech.edu/skelly/uncategorized/installing-xenomai-on-ubuntu-12-04/">http://www.stanley.gatech.edu/skelly/uncategorized/installing-xenomai-on-ubuntu-12-04/</a></p>

<p>&nbsp;</p>

<p>The 3.5 kernel is now&nbsp;the default kernel for Ubuntu
12.04.2 so I guess it will be there for&nbsp;a while</p>

<p>&nbsp;</p>

<p>When running the ./configure tool it fails</p>

<p><span>./configure
--with-linux-dir=/lib/modules/3.5.7-xenomai-2.6.2.1/source --disable-8139too
--with-xenomai-dir=$xenomai_root --enable-rtdm --enable-e1000e</span></p>

<p><span>error: kernel 3.5 not available for e1000e
driver!</span></p>

<p><span>if i take out the e1000e option it configures
but on the make install it fails...</span></p>

<p><span>Anyone played with Etherlab and
kernel 3.5?</span></p>

<p><span>Anyone having Xenomai running at
kernel 3.2?</span></p>

<p><span>Is it worth trying compiling the
3.4 drivers&nbsp;with the 3.5 kernel?&nbsp;</span></p>

<p><span>Best regards<br>
Steffen Dalgard<br>
SINTEF ICT<br>
Norway</span></p>

</div></div>
Uriel Roque | 18 Jun 09:11 2013
Picon

loading ENI file in Etherlab Master

Hi,

I generated an ENI (EtherCAT Network Information) file using Twincat and now want to load it into my etherlab master.

Is there available any implementation of the loading of ENI files for Etherlab master?

The ENI file includes not only the order of the slaves and their configuration but also some init commands.

Has anyone already experience with this?

Thanks,

Uriel

 

--

M.Sc. Uriel Roque

Software Engineering

 

ITK Engineering AG

Luitpoldstraße 59

76863 Herxheim

 

Tel.: +49 7276 9885-600

Fax: +49 7276 9885-619

 

mailto:uriel.roque-CYcon8eqb3htfah0icYw3oQuADTiUCJX@public.gmane.org

_____________________________________________________________

ITK Engineering AG Luitpoldstraße 59 76863 Herxheim,

Tel.: +49 7276 9885-600 Fax: +49 7276 9885-619

mailto:info-CYcon8eqb3htfah0icYw3oQuADTiUCJX@public.gmane.org  http://www.itk-engineering.de

 

Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Josef Würth

Vorstand/Executive Board: Michael Englert (Vorsitzender), Dr. Helmuth Stahl

Sitz der Gesellschaft/Registered Office: 76773 Kuhardt/Pfalz

Registergericht/Registered Court: Amtsgericht Landau, HRB 30139

USt.-ID-Nr./VAT-ID-No. DE 813165046

 

<div>
<div class="WordSection1">
<p class="MsoNormal"><span lang="EN-US">Hi,<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">I generated an ENI (<span class="st">EtherCAT Network Information</span>) file using Twincat and now want to load it into my etherlab master.<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Is there available any implementation of the loading of ENI files for Etherlab master?<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">The ENI file includes not only the order of the slaves and their configuration but also some init commands.
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Has anyone already experience with this?<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Thanks,<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Uriel<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US">--
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">M.Sc. Uriel Roque<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Software Engineering<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US">ITK Engineering AG
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Luitpoldstra&szlig;e 59
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">76863 Herxheim<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US">Tel.: +49 7276 9885-600
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Fax: +49 7276 9885-619
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span><a href="mailto:uriel.roque@..."><span lang="EN-US">mailto:uriel.roque@...</span></a></span><span>
<span lang="EN-US"><p></p></span></span></p>
<p class="MsoNormal"><span lang="EN-US">_____________________________________________________________
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">ITK Engineering AG Luitpoldstra&szlig;e 59 76863 Herxheim,
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">Tel.: +49 7276 9885-600 Fax: +49 7276 9885-619
<p></p></span></p>
<p class="MsoNormal"><span><a href="mailto:info@..."><span lang="EN-US">mailto:info@...</span></a></span><span lang="EN-US">&nbsp;
</span><span><a href="http://www.itk-engineering.de/"><span lang="EN-US">http://www.itk-engineering.de</span></a></span><span lang="EN-US"><p></p></span></p>
<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>
<p class="MsoNormal"><span lang="EN-US">Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Josef W&uuml;rth
<p></p></span></p>
<p class="MsoNormal"><span>Vorstand/Executive Board: Michael Englert (Vorsitzender), Dr. Helmuth Stahl
<p></p></span></p>
<p class="MsoNormal"><span>Sitz der Gesellschaft/Registered Office: 76773 Kuhardt/Pfalz
<p></p></span></p>
<p class="MsoNormal"><span>Registergericht/Registered Court: Amtsgericht Landau, HRB 30139
<p></p></span></p>
<p class="MsoNormal"><span lang="EN-US">USt.-ID-Nr./VAT-ID-No.
</span><span>DE 813165046</span><span><p></p></span></p>
<p class="MsoNormal"><p>&nbsp;</p></p>
</div>
</div>
Ankit Goila | 14 Jun 19:49 2013
Picon

Slaves don't go into the OP state

Hello,

I am using EtherCAT master 1.5.2 kernel module on RT-Preempt, and am using EPOS3 slaves provided by Maxon. When I give the sdo commands through the terminal, I am able to get the motors to run, but the slaves remain in the Pre-op state. I am confused - if the motors are running by sending in the correct sdos, shouldn't the slave be in the OP state? This might precisely be the reason why I am unable to run the motors through Simulink model.
I can't seem to find the error in all this. I'll be grateful if anyone could help me out here.

Thanks,

Ankit

--
Ankit Goila
U-M ID: 83509056
Graduate Student (2012-2014)
M.S.E Mechanical Engineering
University of Michigan, Ann Arbor
Contact: 734-747-0829

<div><div dir="ltr">
<div>
<div>
<div>
<div>Hello,<br><br>
</div>I am using EtherCAT master 1.5.2 kernel module on RT-Preempt, and am using EPOS3 slaves provided by Maxon. When I give the sdo commands through the terminal, I am able to get the motors to run, but the slaves remain in the Pre-op state. I am confused - if the motors are running by sending in the correct sdos, shouldn't the slave be in the OP state? This might precisely be the reason why I am unable to run the motors through Simulink model.<br>
</div>I can't seem to find the error in all this. I'll be grateful if anyone could help me out here. <br><br>
</div>Thanks,<br><br>
</div>Ankit<br clear="all"><div><div><div><div><div>
<br>-- <br><div dir="ltr">
<div>

<span>Ankit Goila</span>
</div>
<span>U-M ID: 83509056</span><br><div><span>Graduate Student (2012-2014)</span></div>
<div><span>M.S.E Mechanical Engineering</span></div>

<div><span>University of Michigan, Ann Arbor</span></div>
<div><span>Contact: 734-747-0829</span></div>

<div><br></div>
</div>
</div></div></div></div></div>
</div></div>
pier | 13 Jun 12:41 2013
Picon

Please Help- Configuration of EL4008 (Analog Output), (Sanya kalani)


Dear Kalani

With these things you need :

1- Enable DEBUG mode  with  ... ethercat debug 1

2-Watch messages in /var/log/syslog .... or messages depending on 
distribuition....

3- Do step by step and watch messages...

Best Regards

Pier

--

-- 

Perlsoft GmbH
Sennereigasse 27
3900 Brig
Switzerland

Tel. +41977022933

Cor Meijneke - 3ME | 13 Jun 09:23 2013
Picon
Picon

Issues with starting ethercat master (with e1000e driver)

Hi,

I ran into some trouble with e1000e driver (I think) while installing the ethercat master. I have setup with
a Intel NUC that yields a 82579V Gigabit network connection with Ubuntu 12.04 running on it (which
indicates its using a e1000e driver) en a simple Beckhoff EK1100 with EL3104 and EL4034 terminals. I've
configured the software with: 
./configure --prefix=/vol/opt/etherlab --disable-8139too --enable-e1000e --with-e1000e-kernel=3.4
and did the rest of the installation according to the manual.
Next I adapted the config file and initiated the service, made an udev rule and restarted the service. 

When I start the ethercat service it says "done" and I can request the slaves giving the output:
0  1:0  PREOP  +  EK1100 EtherCAT-Koppler (2A E-Bus)
1  3:0  PREOP  +  EL3104 4K. Ana. Eingang  +/-10V Diff.
2  5:0  PREOP  +  EL4034 4K. Ana. Ausgang +/-10V, 12bit
But the "ethercad config"  command doesn't give any output and when I run a simple Simulink model it, nothing
happens and the Simulink models cannot connect to target if I try manually.
Below you can see the syslog output for starting the service, running a simple compiled simulink model and
stopping the service respectively.

----------------------------------------------------------------------------------------
----------------------------------Start service-----------------------------------------
----------------------------------------------------------------------------------------
[ 6202.019925] EtherCAT: Master driver devel af21f0bdc7c9
[ 6202.020117] EtherCAT: 1 master waiting for devices.
[ 6202.188857] ec_e1000e: EtherCAT-capable Intel(R) PRO/1000 Network Driver - 1.5.1-k-EtherCAT
[ 6202.188862] ec_e1000e: Copyright(c) 1999 - 2011 Intel Corporation.
[ 6202.188903] ec_e1000e 0000:00:19.0: setting latency timer to 64
[ 6202.189062] ec_e1000e 0000:00:19.0: irq 40 for MSI/MSI-X
[ 6202.484470] EtherCAT: Accepting device EC:A8:6B:F4:AC:61 for master 0.
[ 6202.633032] ec_e1000e 0000:00:19.0: irq 40 for MSI/MSI-X
[ 6202.686015] ec_e1000e 0000:00:19.0: (unregistered net_device): MSI interrupt test failed, using
legacy interrupt.
[ 6202.686080] EtherCAT 0: Starting EtherCAT-IDLE thread.
[ 6202.686279] ec_e1000e 0000:00:19.0: (unregistered net_device): (PCI Express:2.5GT/s:Width x1) ec:a8:6b:f4:ac:61
[ 6202.686295] ec_e1000e 0000:00:19.0: (unregistered net_device): Intel(R) PRO/1000 Network Connection
[ 6202.686375] ec_e1000e 0000:00:19.0: (unregistered net_device): MAC: 10, PHY: 11, PBA No: FFFFFF-0FF
[ 6202.686832] ec_e1000e 0000:00:19.0: (unregistered net_device): Reset adapter
[ 6204.690679] e1000e: ec0 NIC Link is Up 100 Mbps Full Duplex, Flow Control: None
[ 6204.690696] ec_e1000e 0000:00:19.0: (unregistered net_device): 10/100 speed: disabling TSO
[ 6204.690702] EtherCAT 0: Link state changed to UP.
[ 6204.695825] EtherCAT 0: 3 slave(s) responding.
[ 6204.695836] EtherCAT 0: Slave states: PREOP.
[ 6204.695943] EtherCAT 0: Scanning bus.
[ 6205.096368] EtherCAT ERROR 0-2: SDO upload 0x1C13:01 aborted.
[ 6205.096373] EtherCAT ERROR 0-2: SDO abort message 0x06090011: "Subindex does not exist".
[ 6205.096375] EtherCAT ERROR 0-2: Failed to read index of assigned PDO 1 from SM3.
[ 6205.096376] EtherCAT 0: Bus scanning completed in 400 ms.

----------------------------------------------------------------------------------------
----------------------------------Run Model-----------------------------------------
----------------------------------------------------------------------------------------
[ 6253.519443] EtherCAT: Requesting master 0...
[ 6253.519448] EtherCAT: Successfully requested master 0.
[ 6253.519491] EtherCAT 0: Domain0: Logical address 0x00000000, 8 byte, expected working counter 1.
[ 6253.519493] EtherCAT 0:   Datagram domain0-0: Logical offset 0x00000000, 8 byte, type LRD.
[ 6253.519565] EtherCAT 0: Master thread exited.
[ 6253.519568] EtherCAT 0: Starting EtherCAT-OP thread.
[ 6253.929194] EtherCAT WARNING 0-1: Slave does not support changing the PDO mapping!
[ 6253.929198] EtherCAT WARNING 0-1: Currently mapped PDO entries: 0x6030:11/16. Entries to map: 0x6040:11/16
[ 6259.153438] EtherCAT WARNING 0-1: Slave did not sync after 5000 ms.
[ 6259.203384] EtherCAT ERROR 0-1: Failed to set SAFEOP state, slave refused state change (PREOP + ERROR).
[ 6259.223362] EtherCAT ERROR 0-1: AL status message 0x0035: "DC Invalid Sync Cycle Time".
[ 6259.253328] EtherCAT 0-1: Acknowledged state PREOP.
[ 6259.283759] EtherCAT 0: Slave states: PREOP, OP.
[ 6259.511476] EtherCAT WARNING 0: 240 datagrams UNMATCHED!
[ 6259.682945] EtherCAT WARNING: Datagram ffff880409434d80 (domain0-0) was SKIPPED 120 times.

----------------------------------------------------------------------------------------
----------------------------------Stop service-----------------------------------------
----------------------------------------------------------------------------------------
[ 7198.032419] EtherCAT 0: Releasing master...
[ 7198.032511] EtherCAT 0: Master thread exited.
[ 7198.032527] EtherCAT 0: Starting EtherCAT-IDLE thread.
[ 7198.032606] EtherCAT 0: Released.
[ 7198.047031] EtherCAT 0: Slave states: PREOP.
[ 7232.575556] EtherCAT 0: Master thread exited.
[ 7232.593050] EtherCAT 0: Link state changed to DOWN.
[ 7232.712710] EtherCAT 0: Releasing main device EC:A8:6B:F4:AC:61.
[ 7232.717499] EtherCAT: Master module cleaned up.
[ 7233.724151] e1000e: Intel(R) PRO/1000 Network Driver - 2.0.0-k
[ 7233.724161] e1000e: Copyright(c) 1999 - 2012 Intel Corporation.
[ 7233.724233] e1000e 0000:00:19.0: setting latency timer to 64
[ 7233.724411] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 7233.724494] e1000e 0000:00:19.0: irq 40 for MSI/MSI-X
[ 7233.980802] e1000e 0000:00:19.0: eth0: (PCI Express:2.5GT/s:Width x1) ec:a8:6b:f4:ac:61
[ 7233.980815] e1000e 0000:00:19.0: eth0: Intel(R) PRO/1000 Network Connection
[ 7233.980890] e1000e 0000:00:19.0: eth0: MAC: 10, PHY: 11, PBA No: FFFFFF-0FF
[ 7234.150630] e1000e 0000:00:19.0: irq 40 for MSI/MSI-X
[ 7234.251416] e1000e 0000:00:19.0: irq 40 for MSI/MSI-X
[ 7234.254712] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[ 7234.256198] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[ 7235.960283] e1000e: eth0 NIC Link is Up 100 Mbps Full Duplex, Flow Control: None
[ 7235.960298] e1000e 0000:00:19.0: eth0: 10/100 speed: disabling TSO
[ 7235.963387] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready

Hope someone can give me some advice how to tacle this issue.

Best regards,

Cor

---------------------------
Ing. C. Meijneke 
Research Engineer 

Delft Biorobotics Laboratory 
Bio Mechanical Engineering 
Delft University of Technology 

Graeme Foot | 13 Jun 05:21 2013

Fix for ecrt_master_select_reference_clock

Hi,

 

I just noticed that ecrt_master_select_reference_clock hasn't been working.

 

On startup the master auto selects the first DC slave in ecrt_master_select_reference_clock.

 

Problem 1:

Calling ecrt_master_select_reference_clock selects the slave config to use, but does not activate it immediately.  ecrt_master_select_reference_clock is only called if the bus is rescanned, at which point problem 2 happens.

 

Problem 2:

ecrt_master_select_reference_clock is called before slaves are attached to the slave configs so trying to set up dc_ref_config fails and then no slave is set up as the dc_ref_clock.

 

 

I have attached a patch that corrects the above problems.  The patch also contains a few other things I change so I've added a description of my changes for this problem below.

 

 

master/master.c

 

ec_master_find_dc_ref_clock

- If using dc_ref_config fails then all slaves are scanned for the first slave to support DC

 

ecrt_master_select_reference_clock

- Calls ec_master_find_dc_ref_clock

- I have removed the dc capabilities check as these are in ec_master_find_dc_ref_clock anyway

 

master/fsm_master.c

 

ec_fsm_master_state_scan_slave

- I have moved ec_master_calc_dc after ec_master_attach_slave_configs

 

 

Regards,

Graeme Foot.

<div>

<div class="Section1">

<p class="MsoNormal"><span lang="EN-US">Hi,<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">I just noticed that ecrt_master_select_reference_clock hasn't been
working.<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">On startup the master auto selects the first DC slave in ecrt_master_select_reference_clock.<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">Problem 1:<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">Calling ecrt_master_select_reference_clock selects the slave config to
use, but does not activate it immediately.&nbsp; ecrt_master_select_reference_clock
is only called if the bus is rescanned, at which point problem 2 happens.<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">Problem 2:<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">ecrt_master_select_reference_clock is called before slaves are attached
to the slave configs so trying to set up dc_ref_config fails and then no slave is
set up as the dc_ref_clock.<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">I have attached a patch that corrects the above problems.&nbsp; The
patch also contains a few other things I change so I've added a description of
my changes for this problem below.<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">master/master.c<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">ec_master_find_dc_ref_clock<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">- If using dc_ref_config fails then all slaves are scanned for the
first slave to support DC<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">ecrt_master_select_reference_clock<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">- Calls ec_master_find_dc_ref_clock<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">- I have removed the dc capabilities check as these are in ec_master_find_dc_ref_clock
anyway<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">master/fsm_master.c<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">ec_fsm_master_state_scan_slave<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">- I have moved ec_master_calc_dc after ec_master_attach_slave_configs<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US"><p>&nbsp;</p></span></p>

<p class="MsoNormal"><span lang="EN-US">Regards,<p></p></span></p>

<p class="MsoNormal"><span lang="EN-US">Graeme Foot.<p></p></span></p>

</div>

</div>
Ankit Goila | 12 Jun 18:54 2013
Picon

EPOS Library for EtherLab

Hello Andreas,

Thank you for the help. I have already tried the steps that you had mentioned in the mail. I'm still not able to run my slaves through the Simulink model. There are two issues that are coming up:

(1) On compiling the model the regular way, I get an error saying:

gmake: *** No rule to make target `rt_defines.h', needed by `test_capi.o'.  Stop.

When I set the 'Configuration Parameters' -> Code Generation -> Interface to 'Classic call interface', this error goes away.

(2) The example.mdl file supplied with EtherLab suggests that I use an 'Infrastructure' block along with my regular model, something that doesn't exist for the EPOS.

I am working on it, trying to figure out where I am going wrong. Any help/information in this regard will be really appreciated.

Thanks and Regards,

Ankit

--
Ankit Goila
U-M ID: 83509056
Graduate Student (2012-2014)
M.S.E Mechanical Engineering
University of Michigan, Ann Arbor
Contact: 734-747-0829

<div><div dir="ltr">
<div>
<div>
<div>Hello Andreas,<br><br>
</div>Thank you for the help. I have already tried the steps that you had mentioned in the mail. I'm still not able to run my slaves through the Simulink model. There are two issues that are coming up: <br><br>
</div>(1) On compiling the model the regular way, I get an error saying: <br><br>gmake: *** No rule to make target `rt_defines.h', needed by `test_capi.o'.&nbsp; Stop. <br><br>When I set the 'Configuration Parameters' -&gt; Code Generation -&gt; Interface to 'Classic call interface', this error goes away. <br><br>
</div>
<div>(2) The example.mdl file supplied with EtherLab suggests that I use an 'Infrastructure' block along with my regular model, something that doesn't exist for the EPOS. <br><br>
</div>
<div>I am working on it, trying to figure out where I am going wrong. Any help/information in this regard will be really appreciated.<br><br>
</div>
<div>Thanks and Regards,<br><br>
</div>
<div>Ankit<br>
</div>
<div><div><div><div><div>
<br>-- <br><div dir="ltr">
<div><span>Ankit Goila</span></div>

<span>U-M ID: 83509056</span><br><div><span>Graduate Student (2012-2014)</span></div>

<div><span>M.S.E Mechanical Engineering</span></div>
<div><span>University of Michigan, Ann Arbor</span></div>

<div><span>Contact: 734-747-0829</span></div>
<div><br></div>
</div>
</div></div></div></div></div>
</div></div>

Gmane