FW: Fatal seq_monitor: pvVarMonitorOff

["Janez Golob" <janez.golob@cosylab.com>]

Hi Everyone,

 

When running a sequncer program on the vxWorks IOC I’ve noticed the following errors being reported:

 

…

sevr=fatal seq_monitor: pvVarMonitorOff(var setpoint_reached[5], pv MAGDIPL5: IS_SETPOINT_REACHED) failure:

sevr=fatal seq_monitor: pvVarMonitorOff(var setpoint_reached[6], pv MAGDIPL6: IS_SETPOINT_REACHED) failure:

…

 

Firstly I haven’t paid too much attention but later I’ve figured out that the channels do not recover anymore. As far as I can see the errors are reported for all the channels in the array and not only for one or a few of them (this well might be a coincidence).

 

While trying to isolate a problem I found out that the problem disappear if I call sleep function before entering the next state:

 

        state initialize

        {             

               when (pvConnectCount() == pvAssignCount())

               {

            pvGet(mag_count, SYNC);

            for (i = 0; i < mag_count; i++)

            {

               …

               

                epicsSnprintf(recordName, sizeof(recordName), "MAGDIPL%d:IS_SETPOINT_REACHED", i);

                pvAssign(setpoint_reached[i], recordName);

                pvMonitor(setpoint_reached[i]);

 

                …

            }

                   /*epicsThreadSleep(3.0);*/

               } state wait_start

              

               when (delay(10)) { } state initialize

        }

 

        state wait_start

        {             

               when (pvConnectCount() < pvAssignCount()) {} state initialize

 

               when (efTestAndClear(pws_start_flag)) {} state magnets_to_zero_amps

        }

       

I updated the sequncer version to the latest one 2.1.14 and I can reproduce this almost 100% reliably on the vxWorks IOC - in version 2.1.10 it happed less frequently. With sequencer version 2.1.10 I was able to reproduce this also on win-x86 platform but with the newer version not any more.

 

Has anybody else noticed the same problem?

 

Regards,

Janez

 

EPICS and other versions:

   EPICS 3.14.12

   vxWorks 6.8

   sequncer 2.1.14