Experimental Physics and
Industrial Control System

Michael Davidsaver via Tech-talk <tech-talk at aps.anl.gov> · Thu, 26 Nov 2020 11:27:32 -0800

On 11/24/20 1:49 PM, Hu, Yong wrote:
> Michael,
> 
> I tested those records in 3.14.12, 3.15.6, and 7.0.4.1, the same result. Although I set the reprocessing field RPRO to 1, the asynchronous calcout record is not reprocessed after the asynchronous delay is completed.

Firstly, RPRO is an internal flag, not a configuration option.
It should never be set in a .db file.

In this case, I don't think setting RPRO is causing the issue you see.

I think the fundamental problem is that the internal state tracking which
calcoutRecord uses to make "Transition To ..." work is incompatible with
the way asynchronous processing queued.

Specifically, transitions are detected by comparing VAL and PVAL
(previous value).  However, PVAL doesn't get updated if a request
to process comes while the calcout record is busy (PACT=1).

In this case, the RPRO field is set to indicate the the request to
process was skipped, so that this record will be re-processed after
the inprogress asynchronous operation completes.

However, in your case the edge is missed if PMStatus-I has changed
twice during this time.

It looks like the goal of setting ODLY is to act as a holdoff timer?
I think you'll need to do the timer with a separate record.

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System