[ HPF Home | Versions | Compilers | Projects | Publications | Applications | Benchmarks | Events | Contact ] |
DO i = 2, 99 A(I) = B(I-1) + B(I) + B(I+1) END DO
This is one of the simplest examples of an INDEPENDENT loop. (For simplicity, all examples in this section assume there is no storage or sequence association between any variables used in the code.) Every iteration assigns to a different location in the A array, thus satisfying the first condition above. Since no elements of A are used on the right-hand side, no location that is assigned in the loop is also read, thus satisfying the second condition. Note, however, that many elements of B are used repeatedly; this is allowed by the definition of INDEPENDENT. This loop is INDEPENDENT regardless of the values of the variables involved.
!HPF$ INDEPENDENT FORALL ( I=2:N ) A(I) = B(I-1) + B(I) + B(I+1)
This example is equivalent in all respects to the first example.
!HPF$ INDEPENDENT DO I=1, 100 A(P(I)) = B(I) END DO
This INDEPENDENT directive asserts that the array P does not have any repeated entries (else they would cause interference when A was assigned). The DO loop is therefore equivalent to the Fortran statement
A(P(1:100)) = B(1:100)
©2000-2006 Rice University | [ Contact Us | HiPerSoft | Computer Science ] |