Experimental Physics and
Industrial Control System

Tim Mooney <[email protected]> · Mon, 26 Jul 1999 15:10:04 -0500 (CDT)

Re...

> Suppose, for the sake of this
> example, you take a genSub record with an integer output
> 
>      +--------+
>      | genSub |
>      |        +-OUTA-
>      |        |
>      +--------+
>      FTVA: LONG
>      SNAM: myFunction
> 
> and your user defined function writes to its VALA field like this
> 
>      long myFunction( struct genSubRecord *pgensub )
>      {
>          *(long *)pgensub->vala = value;
>          return 0;
>      }
> 
> ...
>  I ... saw some very
> strange behaviour (a different record generating an "illegal choice"
> message when I tried to write a valid value to one of its inputs) until I
> explicitly protected pgensub->vala with a mutex semaphore.
> 
> Any advice gratefully received.

Wait a minute... *pgensub->vala is not a shared resource unless two gensub records
have the same value for pgensub->vala, and that would indicate an error in the
initialization of vala (unless you're trying something tricky).

Anyway, for whatever it's worth, I've seen illegal-choice messages in only two
situations: 1) mismatch between the .dbd file and the code; and 2) somebody
stepping on somebody else's memory.

Tim Mooney

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System