M. Koehrer wrote:
> Hi!
> 
> I have tested two boards, one with revision 0c and one with revision 0d.
> No difference here. 
> I have also implemented a status array, the returns the last 16 status codes
> before the missing signal:
> The result is:
> 2050 0050 4050 0050 4050 0050 2050 0050 4050 0050 4050 0050 2050 0050 4050 
> 0050
> I have the sequence: Receive two messages, send one message, receive two, 
> send one...
> However, at the end I have received only one frame and not two (at least 
> status indicates that).
> Or: Within one interrupt both frames come in.
> 
> When I increase the size of my short, second message from 80 to 200 byte 
> everything works
> perfectly. Then I never get the communication error.
> This is the workaround I use currently.

This really sounds like some timing issue. Extending the messages size
delays it.

Do those two incoming frames normally trigger two separate IRQ events?
Or do they arrive within the same round, causing the IRQ handler to loop?

> 
> This all looks really strange to me. I have no idea what the reason for this 
> could be...
> 

In order to get hold of this issue, maybe reproduce it somewhere else as
well, could you try to abstract your scenario? First question for me
would be if just receiving this kind of packet duo already causes
problems. That would allow to record such a traffic pattern (tcpdump)
and replay it from some arbitrary linux host (tcpreplay).

Otherwise, writing two simple RTnet applications that generate this
packet pattern would be great. If they trigger the issue for you, I
could run them here as well and cross-check it.

Jan

Attachment: signature.asc
Description: OpenPGP digital signature

-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
RTnet-users mailing list
RTnet-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rtnet-users

Reply via email to