Unexpected duplicate activation with ODLY > 0

[Ivo Hanak via Tech-talk <tech-talk at aps.anl.gov>]

Dear madam/sir,

I run into behavior (of EPICS) I am not able to explain. On EPICS 7.0.6.1 (Windows 64-bit build, MSVS, Windows 11), I load the following DB:

record(longin, "$(P):counter") {

    field(PINI, "1")

    field(INP, "$(P):inc PP")

    field(SCAN, ".1 second")

}

record(calc, "$(P):inc") {

    field(INPA, "$(P):counter")

    field(CALC, "A+1")

}

record(fanout, "$(P):trigger") {

    field(SELM, "All")

    field(LNK1, "$(P):delay")

}

record(calcout, "$(P):delay") {

    field(ODLY, "1")

    field(OOPT, "Every Time")

    field(OUT, "$(P):target PP")

}

record(longin, "$(P):target") {

    field(INP, "$(P):counter")

}

And, I run "camonitor test:target" to observe the output in a console. Then:

* When I use IOCSH and run "dbtr test:trigger" or "dbpf test:trigger 0", I notice that "test:target" is updated per call after the delay of 1 second (the ODLY of "$(P):delay"). This is expected.

* When I run "dbtr test:trigger" multiple times within 1 second, I get one update of "test:target". This is expected.

* When I run "dbpf test:trigger 0" multiple times within 1 second, "test:target" is updated twice with 1 second between updates. The same happens when I run "caput test:trigger 0" multiple times within 1 second.

Based on documentation (https://epics.anl.gov/base/R7-0/6-docs/fanoutRecord.html) fanout should ignore the field VAL and it's purpose is to trigger. Therefore, I am lost right now. Is there a chance someone else may try the DB whether this can be reproduced outside my system, please? I might be just my machine.

Thank you.

Sincerely,

Ivo Hanak