[etherlab-users] Waiting for network receive?

Shahbaz Youssefi shabbyx at gmail.com
Sun Jul 1 12:33:21 CEST 2012

I don't specifically know about user-space, but "Jun Yuan" wrote a patch
for me, adding a function called `ecrt_domain_received` that tells you
whether the domain has been received or not (in kernel-space).

I have tested this function with domains spanning into multiple packets and
calling it while some of the packets have been received and others not, and
it is safe. I just checked the latest development branch and unfortunately
this has not been added to the repository (I wonder why).

If the developers of the master are interested, I could present the test we
performed so that they could be sure the patch works and hopefully include
it in the next release.

On Sun, Jul 1, 2012 at 12:00 PM, <etherlab-users-request at etherlab.org>wrote:

> I was looking at the code and they map the datagram directly to user
> space and has a skb ring size of 2 frames. data cannot be available for
> a re-write until it is acked by the interrupt completion of the ethernet
> device.
> You need to query the skb->users counter and check when it is
> decremented back -- meaning when transmission complete takes place.
> I believe the curretn design might cause problem in heavy traffic or in
> preempt rt when the transmit complete threaded interrupt has a "too" low
> priority.
> On Thu, 2012-06-28 at 19:38 +0200, Johnathan Van Why wrote:
> > The software system we are setting up needs to be able to send an
> > EtherCAT frame, then wait until the network data is available for
> > processing to resume (rather than wait until the next cycle time). Is
> > there any way to do that with the EtherCAT Master? This is a userspace
> > application.
> >
> > Thanks for any help,
> > Johnathan Van Why
> > _______________________________________________
> > etherlab-users mailing list
> > etherlab-users at etherlab.org
> > http://lists.etherlab.org/mailman/listinfo/etherlab-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20120701/284b317e/attachment-0004.htm>

More information about the Etherlab-users mailing list