Experimental Physics and
Industrial Control System

"Zelazny, Michael Stanley" <[email protected]> · Thu, 4 Oct 2012 15:53:24 -0700

I have an AO record whose ALST & MLST are stuck at nan:

$ caget KLYS:LI23:81:PJTN KLYS:LI23:81:PJTN.PVAL KLYS:LI23:81:PJTN.LALM KLYS:LI23:81:PJTN.ALST KLYS:LI23:81:PJTN.MLST KLYS:LI23:81:PJTN.ADEL KLYS:LI23:81:PJTN.MDEL KLYS:LI25:81:PJTN KLYS:LI25:81:PJTN.PVAL KLYS:LI25:81:PJTN.LALM KLYS:LI25:81:PJTN.ALST KLYS:LI25:81:PJTN.MLST  KLYS:LI25:81:PJTN.ADEL  KLYS:LI25:81:PJTN.MDEL
 KLYS:LI23:81:PJTN              0.164428
 KLYS:LI23:81:PJTN.PVAL         0.164428
 KLYS:LI23:81:PJTN.LALM         0.164428
 KLYS:LI23:81:PJTN.ALST         nan
 KLYS:LI23:81:PJTN.MLST         nan
 KLYS:LI23:81:PJTN.ADEL         0
 KLYS:LI23:81:PJTN.MDEL         0
 KLYS:LI25:81:PJTN              0.0761155
 KLYS:LI25:81:PJTN.PVAL         0.0761155
 KLYS:LI25:81:PJTN.LALM         0.0761155
 KLYS:LI25:81:PJTN.ALST         0.0761155
 KLYS:LI25:81:PJTN.MLST         0.0761155
 KLYS:LI25:81:PJTN.ADEL         0
 KLYS:LI25:81:PJTN.MDEL         0

Once ADEL is stuck at nan, the AO record support has no way to clear this.  From aoRecord.c:

        /* check for archive change */
        delta = prec->alst - prec->val;
        if(delta<0.0) delta = -delta;
        if (delta > prec->adel) {
                /* post events on value field for archive change */
                monitor_mask |= DBE_LOG;
                /* update last archive value monitored */
                prec->alst = prec->val;
        }

According to Matjaz Kobal, COSYLAB, "Every arithmetic expression that includes a NAN value will have a NAN value as a result. So "delta" will have a value of NAN. Also every comparison with a NAN value will return false. So none of the two if statements will execute their bodies and the monitor_mask will not change. Since the ALST field value is not changed anywhere else, there is no way to get inside the if statement."

I have not verified the above statement, but I believe it to be true.

Q1: Have you ever seen this?  If you have, how did you overcome it?  

My testing tells me ALST doesn't get set to nan when VAL goes to nan.  Example:

lcls-zelazny bash:~>caget SIOC:SYS0:ML02:AO201 SIOC:SYS0:ML02:AO201.ALST SIOC:SYS0:ML02:AO201.MLST
SIOC:SYS0:ML02:AO201           0 
SIOC:SYS0:ML02:AO201.ALST      0 
SIOC:SYS0:ML02:AO201.MLST      0 
lcls-zelazny bash:~>caput SIOC:SYS0:ML02:AO201 1
Old : SIOC:SYS0:ML02:AO201           0 
New : SIOC:SYS0:ML02:AO201           1 
lcls-zelazny bash:~>caget SIOC:SYS0:ML02:AO201 SIOC:SYS0:ML02:AO201.ALST SIOC:SYS0:ML02:AO201.MLST
SIOC:SYS0:ML02:AO201           1 
SIOC:SYS0:ML02:AO201.ALST      1 
SIOC:SYS0:ML02:AO201.MLST      1 
lcls-zelazny bash:~>caput SIOC:SYS0:ML02:AO201 nan
Old : SIOC:SYS0:ML02:AO201           1 
New : SIOC:SYS0:ML02:AO201           nan 
lcls-zelazny bash:~>caget SIOC:SYS0:ML02:AO201 SIOC:SYS0:ML02:AO201.ALST SIOC:SYS0:ML02:AO201.MLST
SIOC:SYS0:ML02:AO201           nan 
SIOC:SYS0:ML02:AO201.ALST      1 
SIOC:SYS0:ML02:AO201.MLST      1 
lcls-zelazny bash:~>caput SIOC:SYS0:ML02:AO201 nan
Old : SIOC:SYS0:ML02:AO201           nan 
New : SIOC:SYS0:ML02:AO201           nan 
lcls-zelazny bash:~>caget SIOC:SYS0:ML02:AO201 SIOC:SYS0:ML02:AO201.ALST SIOC:SYS0:ML02:AO201.MLST
SIOC:SYS0:ML02:AO201           nan 
SIOC:SYS0:ML02:AO201.ALST      1 
SIOC:SYS0:ML02:AO201.MLST      1 
lcls-zelazny bash:~>caput SIOC:SYS0:ML02:AO201 2  
Old : SIOC:SYS0:ML02:AO201           nan 
New : SIOC:SYS0:ML02:AO201           2 
lcls-zelazny bash:~>caget SIOC:SYS0:ML02:AO201 SIOC:SYS0:ML02:AO201.ALST SIOC:SYS0:ML02:AO201.MLST
SIOC:SYS0:ML02:AO201           2 
SIOC:SYS0:ML02:AO201.ALST      2 
SIOC:SYS0:ML02:AO201.MLST      2 
lcls-zelazny bash:~>caget SIOC:SYS0:ML02:AO201.ADEL SIOC:SYS0:ML02:AO201.MDEL                 
SIOC:SYS0:ML02:AO201.ADEL      0 
SIOC:SYS0:ML02:AO201.MDEL      0 
lcls-zelazny bash:~>

Q2: Any idea HOW my ALST & MLST got nan?

I'm running a soft IOC on RHEL5 with EPICS 3.14.12.

I'm tempted to change the code to:

        if ((epicsNAN == delta) || (delta > prec->adel)) {
                /* post events on value field for archive change */
                monitor_mask |= DBE_LOG;
                /* update last archive value monitored */
                prec->alst = prec->val;
        }

Q3: Is there any reason I shouldn't?

Thanks in advance for your help!

Mike

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System