Experimental Physics and
Industrial Control System

[email protected] (William Lupton) · Thu, 3 Feb 2000 08:51:04 -1000

Dear all,

Fragments from Ralph and Deb's postings...
>>> We might consider one additional subtlety
>> 
>> Just one?
>> 
>>> for those of us who use STAT and/or
>>> SEVR, particularly to display value with a color rule that indicates these
>>> values.  If an output record had readback but was not processed, would it
>>> affect STAT and/or SEVR values.  Right now I think these are usually UDF
>>> and INVALID respectively until the record is processed. 
>> 
>> Good point. As the record is defined after reading back the value,
>> UDF/INVALID is inappropriate. Would you think we need to have another
>> status that announces that the value has been read back and is valid, but
>> the record hasn't been processed yet? Or would that be too subtle a
>> difference to be signalled?

The existing output record device support that I have looked at (_not_
an exhaustive test) looks something like this (from devBoXVme220.c)...

        status = xy220_read(pvmeio->card,pbo->mask,&value);
        if(status == 0) pbo->rbv = pbo->rval = value;
        else status = 2;
	.
	return(status);

...so if the readback succeeds, RVAL is set to the readback value and a
zero status is returned, which (for bo anyway) will do...

        if(status==0) {
                if(pbo->rval==0) pbo->val = 0;
                else pbo->val = 1;
                pbo->udf = FALSE;
        } else if (status==2) status=0;

...so RVAL will be converted to VAL and UDF will be cleared. If the readback
failed, then a status of 2 is returned, no conversion to VAL occurs, and UDF
is left unchanged.

So I think that the UDF/INVALID side of things is already handled.

William

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System