[Omp] The parallel

[Shengyan Hong]

Every OMP member,
       subroutine evolve(u0, u1, t, indexmap, d1, d2, d3)

c---------------------------------------------------------------------
c---------------------------------------------------------------------

c---------------------------------------------------------------------
c evolve u0 -> u1 (t time steps) in fourier space
c---------------------------------------------------------------------

       implicit none
       include 'global.h'
       integer d1, d2, d3
       double complex u0(d1+1,d2,d3)
       double complex u1(d1+1,d2,d3)
       integer  indexmap(d1+1,d2,d3)
       integer t
       integer i, j, k
       integer TID,OMP_GET_THREAD_NUM
!$omp parallel default(shared) private(i,j,k)
C      TID = OMP_GET_THREAD_NUM()
C      PRINT *, 'thread = ', TID
C      print *, "March 12"
        CALL MAGIC_BRK_SIM_START()
!$omp do
       do k = 1, d3
         do j = 1, d2
             do i = 1, d1
                 u1(i,j,k) = u0(i,j,k)*ex(t*indexmap(i,j,k))
             end do
          end do
        end do
!$omp end do nowait
       CALL MAGIC_BRK_SIM_MIDDLE()
!$omp BARRIER
       CALL MAGIC_BRK_SIM_STOP()
!$omp end parallel
       return
       end

       For the above code, it generates 8 threads each time when it is 
called by the parent subroutine. And the OS allocates the 8 threads to 8 
processors. When the subroutine evolve is over, the 8 threads is over too.
Next time, the subroutine evolve is called again. And the code will 
generate 8 new threads. I want that the 8 threads can be assigned to the 8 processors in the same mapping 
relationship as it happens last time?
      First, is the new 8 threads the similar to the 8 ones the subroutine evolve generates last time? Second, 
can I use some methods to make the new threads binding to the 
processors in the same mapping form the OS does last time?
      Thank you.
                                       Shengyan Hong