********> bugfix.34
Correction Authors: Julian Tirado-Rives & David A. Pearlman
Correction Date: 3/6/92  

(Note: Julian Tirado-Rives at Yale discovered and tracked down this
       bug. All I did was flesh out the bug report form and determine
       the circumstances under which the bugs would be encountered. Thanks
       Julian! -- dap).

Programs: GIBBS

Severity: serious

Problem: 

       An attempt to run a free energy simulation where the following
       conditions are met will likely fail:

         1) A "belly" has been requested;

               and at least one of the following:

        2a) Contributions from constraints to the free energy
            have been requested (NCORC = 1).
   
        2b) Constraints have been added to the system by setting INTR>0
            and adding the constraints as line(s) 19, with ITOR=2.

        Condition (1) and one or both of conditions (2) must be
        satisified. Otherwise, the simulation will perform correctly.


Affects: 
         The program will likely fail in routine TORCON, with a report
   that "Routine TORCON (internal constraints) failed to converge" or
   that "Routine TORCON (internal constraints) failed. Required change
   is too large". 

Cause: 
        
         For the case where a belly has been defined, the masses of atoms
    which are part of the belly are not being set. The masses of the
    atoms which are not part of the belly are correctly being set to 0.

Fix:  Make the following changes

***To routine TORCON in src/gibbs/torcon.f

*** OLD torcon.f
--- NEW torcon.f
***************
*** 406,412
  C
     15   DO 17 I=1,ITYPE
              IF (BELLY) THEN
!                 IF (IGRP(IM(I)).LE.0) XIMASS(I) = 0.0
              ELSE
                  XIMASS(I) = RIMASS(IM(I))
              END IF

--- 406,416 -----
  C
     15   DO 17 I=1,ITYPE
              IF (BELLY) THEN
!                 IF (IGRP(IM(I)).LE.0) THEN
!                    XIMASS(I) = 0.0
!                 ELSE
!                    XIMASS(I) = RIMASS(IM(I))
!                 END IF
              ELSE
                  XIMASS(I) = RIMASS(IM(I))
              END IF

Temporary workarounds: None. If you wish to run a simulation where all
     the conditions reported in the Description section are met, you must
     install this fix.

Routines affected:

     GIBBS   Routine TORCON in file .../amber4/src/gibbs/torcon.f