List tags on Micro850 inconsistent

[Richard Porter]

Has anyone had issues listing tags on a Micro850 (model 2080-L50E-24QWB) where the example code list_tags_micro8x0.c only works a few times then stops?

I ran list_tags_micro8x0 successfully 3 times. Between each run I added a global variable of a different type. At this time it was working so I didn't keep a record of the console output. However, after the third attempt, list_tags_micro8x0 fails every time. I have tried the following:

• Reboot PC
• Restart PLC
• Remove additional global variables
• Download original project to PLC
• Create new project and download

Throughout all these iterations the simple.c example (modified for a micro850) was able to run successfully.

I'm starting to think I was dreaming when I saw it working!

[Kyle]

I do not have a Micro8x0 PLC on which to test.  We've found out that:

• You cannot get the UDT definitions on a Micro800 using the same method as that for a ControlLogix.
• You cannot see BOOL array tags in the tag listing. 

Both of these are _really_ annoying and so far we have not found a work around.  There are several of us in various CIP-related open source projects that keep poking at this.  Nothing we have found documents how to do this.

If you can capture logging output from the library that might help.

Best,

Kyle

[Kyle]

I hit send way too soon.

I looked through the attached log and it is weird.  There are two requests for tag info sent to the Micro850.  It responds to both.

The first one returns 526 bytes and the CIP status is 0x06 which means that this is a partial result.  So another request is sent and gets back 474 bytes and the status is 0x00 which means no error.  So that was the end of the transfer.  Then... nothing... nothing at all.  I just see the internal tickler thread running every 100ms with nothing to do.  

I don't have a Micro850 or any 800 model on which to test.  Are you willing to be a guinea pig?  This feels like a bug in the library.

My tag listing tests still work on my ControlLogix, so whatever it is it did not break everything.

Best,

Kyle

[Richard Porter]

Hi Kyle,

Yes, I am happy to be a guinea pig.

Let me know what you want tested.

Richard

[Kyle]

I will get back to you when I am back in my home lab.  I am leaving on vacation shortly and won't be back until next week.

Best,

Kyle

[Kyle]

I have been poking at this one, and another report with issue #569 looks like it may be the same thing.   I found some flaws with the refactor a couple releases ago, and especially with my tests that were passing things that should not pass.

Please try with version 2.6.10.   

Best,

Kyle

[Richard Porter]

It is working consistently. Great work Kyle.

Regards,

Richard

[Kyle]

Thanks for confirming.  I will be reworking my tests a bit to make sure that this is caught and does not regress.

Best,

Kyle