[etherlab-users] Fwd: unofficial patch: check domain received

Shahbaz Youssefi shabbyx at gmail.com
Wed Nov 20 18:31:15 CET 2013


Hi

I'm forwarding the patch as I received it (with the name of the author, but
I'd wait for his consent before disclosing his other info although just in
case).

In fact, the functions therein don't touch the functionality of the master.
In other words if you don't use those functions, it doesn't absolutely
incur any penalty whatsoever (except perhaps an extra `case` in `ioctl`
which I think we can all agree is negligible). The functions themselves are
very simple and simply just loop over the domains and tell whether they are
received.

Here's how it's tested (quoting from the past):


*--- quote ---*











*Alright, I did the test with the following configuration:2 slaves, each
producing 1.2k data, split equally in 2 datagrams. A bridge placed in the
network that introduces a delay between the 2 datagrams. The delays I
tested with were 100us, 500us and 1ms. That is, the bridge makes sure there
is at least 1ms delay between passing frames. I logged the passing times
and this is done correctly. Now, with ecrt_domain_received, I could see
that it returns true after a correct time. That is, when I increase the gap
in the bridge, the receive time in the network is increased. With another
program I checked the data received from the slaves and they are correct
(both datagrams arrive). So I guess everything works just fine! :)Thank you
for this patch, hope it makes its way to the official release.ShahbazP.S.
You had forgot the EXPORT_SYMBOL in your patch (just in case you wanted to
give it to the developers)*
*--- /quote ---*

Furthermore, a peculiarity that I discovered is (quoted from another email):





*--- quote ---Something you might want to know:If ecrt_domain_received is
called after ecrt_domain_process, the kernel crashes. I did so to only read
data of received domains (after I had processed them) only to discover this
behavior.*

*No big deal though, as I can just read whatever is available and I don't
really gain much by ignoring timed-out domains.*

*--- /quote ---*

There was a long discussion on this and it was also reminded at another
time. You can read the whole discussion here:

http://lists.etherlab.org/pipermail/etherlab-users/2011/date.html
(search for "Knowing when the packet"...)

and at a later time:

http://lists.etherlab.org/pipermail/etherlab-users/2012/001755.html
http://lists.etherlab.org/pipermail/etherlab-users/2012/001756.html

---------- Forwarded message ----------
From: Jun Yuan

patch file for stable 1.5 ;)

-- 
Jun Yuan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20131120/6a218189/attachment-0002.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: check_received_stable1.5_2266.patch
Type: text/x-patch
Size: 3988 bytes
Desc: not available
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20131120/6a218189/attachment-0003.bin>


More information about the Etherlab-users mailing list