In a message dated 10-22-98, James Mckenzie said to David Noon about "Need
to copy"

Hi James,

[snip]
JM>finally the code at the beginning of calc_offered_load:

JM>void calc_offered_load (EXPT_DESC_PTR expt)
JM>{
JM>  /* Copy of experiment description structure used for offered load
JM>   * calculation */
JM>    EXPT_DESC   ol_ex;
JM>    int     i;
JM>    double  def_off_load, off_load;
JM>    double  avg_ol;
JM>    int     saved_debug_level;
JM>    char    *copyptr;

JM>if (verbose) fprintf (outfile, "\n*** Calculating offered
load.***\n\n");
JM>
JM>    #ifdef DEBUG_SUPPORT
JM>        saved_debug_level = debug_level;
JM>        if (!(debug_level & DEBUG_OFFLD)) debug_level = DEBUG_OFF;
JM>    #endif /* DEBUG_SUPPORT */

JM>/**/ //bcopy ((char *) expt, (char *) &ol_ex, sizeof (EXPT_DESC));
JM>/*original unix code */

Ugly, but will work for POSIX-compliant compilers/libraries.

JM>//    copyptr = (char *) memcpy (&ol_ex,  expt, sizeof (EXPT_DESC));
JM>/* code modified for use under DOS/OS2 by 
JM>original author. Results in a SYS1808 error */

The above is sort of correct. Your compiler might be returning NULL when it
should be returning a valid address as a void *. I don't bother checking
memcpy() calls, as that is why we have protection hardware for memory.

JM>//    copyptr = (char *) memmove (&ol_ex, expt, sizeof (expt));
JM>/*This "works" except INT variables end up with incorrect values*/

The above has the wrong 3rd argument. It should be sizeof(EXPT_DESC).

JM>//    if (!copyptr) {
JM>//  fprintf (stderr, "Experiment file copy failed \n");
JM>//  exit (1);
JM>//    }

JM>old_ex = ex;
JM>/*This works with all values as they are in ex*/

This is a reference to the global storage. It will work, but makes no use of
your expt parameter so it is far less flexible.

You might also try:

 old_ex  = *expt;

Regards

Dave

___
 * MR/2 2.25 #353 * HAL 9000: "Help me, Dave.  I can't run under Windows."

--- Maximus/2 3.01