Experimental Physics and
Industrial Control System

On Jun 30, 2021, at 11:07 AM, Sobhani, Bayan via Tech-talk <tech-talk at aps.anl.gov> wrote:

After upgrading EPICS, I started getting this message for some PVs:

 

missing record support entry table Record Support Routine (get_enum_str) Record Type ai PV XF:28ID1-CT{RG:B1}T-I.SEVR  dbGetField

 

It seems to be because I am converting an enum to a string in:

 

field(INPA, "$(Sys)$(Dev)T$(N_T)-I.SEVR")

field(FTA, "STRING")

 

This used to work in previous versions of EPICS, and in fact I can get the problem to go away if I read the PV over channel access, like: "$(Sys)$(Dev)T$(N_T)-I.SEVR CP”

 

So I am wondering: What is the recommended way to get the severity string from a PV in recent versions of EPICS?

That should work as you expected using a DB link, but it seems to have been broken by some recent changes to the IOC on the 7.0 branch (it’s fine in 7.0.4, not sure about 7.0.4.1):

woz$ cat ~/db/test.db 

record(ai, "ai") {}

record(stringin, "si") {

  field(INP, "ai.SEVR")

}

woz$ ./base-3.15/bin/darwin-x86/softIoc -d ~/db/test.db

Starting iocInit

############################################################################

## EPICS R3.15.9

## EPICS Base built Jun 30 2021

############################################################################

iocRun: All initialization complete

epics> dbpf si.PROC 1

DBR_UCHAR:          1         0x1                 

epics> dbgf si

DBR_STRING:          "INVALID"          

epics> woz$ 

woz$ ./base-7.0/bin/darwin-x86/softIoc -d ~/db/test.db

Starting iocInit

############################################################################

## EPICS R7.0.5.1-DEV

## Rev. R7.0.5-264-g7ca25d3c888c56844877-dirty

############################################################################

iocRun: All initialization complete

epics> dbpf si.PROC 1

missing record support entry table Record Support Routine (get_enum_str) Record Type ai PV ai.SEVR  dbGetField

DBF_UCHAR:          1 = 0x1

I’m pretty sure I know where the problem came from, in the mean-time your CA work-around is probably the best solution, until we can ship a fix.

Thanks for the bug report, I have filed it as an issue in GitHub.

- Andrew