********> bugfix.52
Correction Author: David A. Pearlman
Correction Date: 11/05/92

Programs: GIBBS

Severity: Slight

Problem: 

       If a simulation is run where the free energy contributions of
       constraints are requested (NCORC=1) and slow growth is being used,
       a format overflow can result when writing the restart file. 

       Nothing at all will be wrong with the simulation, and no problem
       at all will be encountered unless one attempts to restart using
       the file with the overflowed field.

Affects: 

       A field near the end of the restart file will appear as "*****".
       One will not be able to carry out a restart (IREST=1) with this
       file unless the overflowed field is removed. (This is easily
       accomplished; see "workarounds" below).

Cause: 

        A variable which is not being used is being incremented after every
        step of slow growth, and becomes too big to fit in the i5 field.

Fix:  Make the following change to routine CCPRT

---To routine CCPRT in file connrg.f -----------------------------------------

*** OLD connrg.f
--- NEW connrg.f
***************
*** 2963,2968 ****
--- 2963,2969 ----
        IOUT = IOUT1
    601 IF (ISAV.EQ.1) IOUT=IOUT3
  C
+       IF (IPERON.EQ.0) IPRNOW(2) = 0
        IF (FORM) THEN
            WRITE(IOUT,1020) 'CONSTRAINTS'
            WRITE(IOUT,1014) INDMAX,IPERON,IDWIDE,IBNDLM,NBRST,NARST,
------------------------------------------------------------------------------

Temporary workarounds: 

     You can edit the restart file and replace the overflowed field
              "*****" 
     with 
              "    0"             . 
     This variable is not being used with slow growth, so the resulting
     restart file will be correct after the change.

Routines affected:

     GIBBS   Routine CCNRG in file ...amber4/src/gibbs/connrg.f