Experimental Physics and
Industrial Control System

Bill Stewart <[email protected]> · Wed, 16 Apr 97 9:40:04 EDT

> Return-Path: <[email protected]>
> Received: (from wls@localhost) by xcelsior.larc.nasa.gov
>  (8.7.6/server3.4) id JAA21635
>  for [email protected]; Wed, 16 Apr 1997 09:34:23 -0400 (EDT)
> Message-Id: <[email protected]>
> Date: Wed, 16 Apr 1997 09:34:23 -0400 (EDT)
> From: Bill Stewart <[email protected]>
> 
> My name is Bill Stewart.  I am working at the NASA Langley Research
> Center with EPICS.  I was interested in using the dynamic assignment
> and dynamic monitoring capabilities of SNL and have run into a few 
> problems.  Basically, what I attempted to do was to implement code
> similar to the example on page 22 of the "State Notation Language
> and Seq. Users Guide" with an array of floats.  It turns out that
> this example will not compile with my version of snc.  It fails on the
> following assignment:
> 
>    long setpoint[5];
>    assign setpoint to {  };
> 
> I was able to get it to compile by changing the curly brackets to " " 
> as in the following example:
> 
>     char str[30];
>     float pos_actual[42];
>     assign pos_actual to " ";
> 
> Then,  I perform the following to do the dynamic assignment
> and monitoring.
> 
>      /* do dynamic assignments on the array type parameters */
>     for (i=0; i<42;i++){
>        /* Actual jack positions from LVDT's */
>        sprintf(str,"flexcs:pos:j_pos_%d\n",i);
>        pvAssign(pos_actual[i],str);
>        pvMonitor(pos_actual[i]);
>        .
>        .
>        .
>      }
> 
> When this segment gets loaded and run, I get an error.  Oddly enough,
> it never completes the loop.  It dies on the 25th iteration. What I get
> in debug mode is as follows:
> 
> 
> CA.Client.Diagnostic..............................................
>     Message: "The array or data structure specified will not fit in CA message buffer"
>     Severity: "Error" Context: "ca_add_array_event"
>     Source File: ../seq_if.c Line Number: 279
> task: 0X373aa4 taskwd
> task 1b2d18 fcs suspended
> 
> 
> 
> Also, since the documentation states that the functions pvAssign and 
> pvMonitor are defined to return an integer, what are the expected return
> values for successful and unsuccesful completions?  The documentation that
> I have does not state what the return values are supposed to be.  What gets
> returned (with a few mods. to the above code) is a value of "1" from
> both calls prior to the bus error. I don't believe I have a memory problem.  The version of
> EPICS is R3.12 and the version of snc is 1.9.0 for 3.12.1.  Any takers?
> Thanks for your help....
>  
> 
> William L. Stewart  
> [email protected]
> 
> 
> 
> 
> 

Well,  just as I sent the tech talk message about the snl problem that
I was having with dynamic assignment and monitoring, I figured out my 
own problem.  As I stated in my previous message the example will not
compile.  The problem was with the following assignment:

   long setpoint[5];
   assign setpoint to {  };

This example works fine when you place " " within the curly brackets,
such as :

   long setpoint[5];
   assign setpoint to {"  "};

This seems to satisfy the compiler for the empty array.

William L. Stewart  
[email protected]

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System