Experimental Physics and
Industrial Control System

Mark Rivers via Tech-talk <tech-talk at aps.anl.gov> · Mon, 1 Mar 2021 13:43:16 +0000

Hi Torsten,

I think that 2 of the strengths of asyn are:

- It is almost completely independent of EPICS at the driver layer.  The only thing it uses it libCom.

- It has well-defined single-purpose interfaces (asynInt32, asynFloat64, etc.) that write or read single values and return status information.

If we were to add the metadata it would impact both of the above strengths.

It seems to me that using "helper" records is not so bad to keep the drivers independent of knowledge of the details of EPICS records and to keep the interfaces clean.

One of the weaknesses of asyn is that it is written in C, which makes it difficult to extend the interfaces in a backwards compatible manner.  If it had been written in C++ this would be much easier.

Mark

________________________________
From: Tech-talk <tech-talk-bounces at aps.anl.gov> on behalf of Torsten Bögershausen via Tech-talk <tech-talk at aps.anl.gov>
Sent: Monday, March 1, 2021 7:27 AM
To: tech-talk at aps.anl.gov; torsten.bogershausen at esss.se
Subject: Accessing Record fields from asyn driver

Folks,

for some IO we have all logic in a PLC-ish system.
If I take an ai record as an example, the the PLC
does most of the work what we normally do in the record.

Especially:
The PLC "knows" values .EGU, HIGH, LOW, HIHI, LOLO
(and probably others) because all the work is done inside the PLC
program already.

We want of cause an EPICS integration for those kind of devices
(it could be as simple as a temperature sensor), it would be nice if we
can use the values from the PLC for e.g. .EGU and the other fields.
As is, without having to define the in the record database, with the
risk of getting thme out of sync.
And yes, we can query the PLC for all this kind of information.

The asyn framework is really excellent in doing all the polling,
but there is no way to "push" meta data into single record fields.
(Unless I create a lot of helper-records, like a stringin to hold
the .EGU field, and forward the value to the .EGU field of my ai
record).

Are there more people having the same issue ?

Enhancing asyn with this functionality could be one option,
as we have code here to influence the alarm state/severity.

Any thoughts about this ?

Thanks
/Torsten

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System