Experimental Physics and
Industrial Control System

"Noboru Yamamoto" <[email protected]> · 3 Jun 99 12:15:08 +0900

Hellor Ron,

>Hello All,
>
>I think I've seen the answer to this, I may have even asked the question 
before.
>
>I have a 10 element waveform (type float) from which I wish to extract 
each 
>element into an individual AI.
>
>I could code 10 individual subArray records, each mapped to another 
element, but 
>a subroutine record seems easier. If my 10 element waveform VAL is linked 
to A 
>of the subroutine record, is the code simply:
>
>psub->b=psub->a[0]
>psub->c=psub->a[1], etc.?
>
>That seems much too easy, and "a" is not an array, right?
>
>/Ron Chestnut
>
I'm sure that one of GURUs in EPICS collaboration will answer this quesion,
but I will report what I studied anyway.

The psub->inpa field keeps a link to your 10 element waveform VAL field.
You can copy the value of the waveform record into your buffer , 
(you must allocate buffer in your subroutine),
using dbGetLink() function call, I think.
Then you can copy values into the other fields, 

	psub->b=pbuf[0];
	psub->c=pbuf[1]; 
	etc. 

Before I wrote this mail, I looked at the code of the subroutine record 
(subRecord.c).

The process() function calls the fetch_value() function 
to update values in the field A to L(?)
before your subroutine is called . 

The fetch_value() function in the subRecord.c get values from link using:

  	for(i=0, plink=&psub->inpa, pvalue=&psub->a;
 		i<ARG_MAX; 
		i++, plink++, pvalue++) {
		status=dbGetLink(plink,DBR_DOUBLE, pvalue,0,0);
...

I'm not sure if this call does not break subroutine record strucuture,
when plink , whic is psub->inpa in this case, is a link to array field 
longer
than ARG_MAX.

Noboru Yamamoto
KEKB control group
KEK, JAPAN

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System