077940621d
This PR does reduces the verbosity of parser errors for OpenACC block constructs that do not parse correctly because they are missing their trailing end block directive by: - Removing the redundant error messages created by parsing 3 different styles of directive tokens. - Providing a general mechanism of configuring the max number of contexts printed for every syntax error. - Not printing less specific contexts that are at the same location. Prior to the changes: ``` $ flang -fc1 -fopenacc -fsyntax-only flang/test/Parser/acc-data-statement.f90 2>&1 | tee acc-data-statement.prior.log | wc -l 262 ``` [acc-data-statement.prior.log](https://github.com/user-attachments/files/19298165/acc-data-statement.prior.log) ``` $ flang -fc1 -fopenacc -fsyntax-only flang/test/Parser/acc-data-statement.f90 2>&1 | tee acc-data-statement.prior.log | wc -l 73 ``` [acc-data-statement.post.log](https://github.com/user-attachments/files/19298181/acc-data-statement.post.log)
96 lines
2.0 KiB
Fortran
96 lines
2.0 KiB
Fortran
! RUN: %flang_fc1 -fsyntax-only -fopenacc %s 2>&1
|
|
C Test file for OpenACC directives in fixed-form Fortran
|
|
PROGRAM ACCTEST
|
|
IMPLICIT NONE
|
|
INTEGER :: N, I, J
|
|
PARAMETER (N=100)
|
|
REAL :: A(N), B(N), C(N), D(N)
|
|
REAL :: SUM
|
|
|
|
C Initialize arrays
|
|
DO I = 1, N
|
|
A(I) = I * 1.0
|
|
B(I) = I * 2.0
|
|
C(I) = 0.0
|
|
D(I) = 1.0
|
|
END DO
|
|
|
|
C Basic data construct using C$ACC
|
|
C$ACC DATA COPYIN(A,B) COPYOUT(C)
|
|
DO I = 1, N
|
|
C(I) = A(I) + B(I)
|
|
END DO
|
|
C$ACC END DATA
|
|
|
|
* Parallel construct with loop using *$ACC
|
|
*$ACC PARALLEL PRESENT(A,B,C)
|
|
*$ACC LOOP
|
|
DO I = 1, N
|
|
C(I) = C(I) * 2.0
|
|
END DO
|
|
*$ACC END PARALLEL
|
|
|
|
C Nested loops with collapse - C$ACC style
|
|
C$ACC PARALLEL LOOP COLLAPSE(2)
|
|
DO I = 1, N
|
|
DO J = 1, N
|
|
A(J) = A(J) + B(J)
|
|
END DO
|
|
END DO
|
|
C$ACC END PARALLEL LOOP
|
|
|
|
* Combined parallel loop with reduction - *$ACC style
|
|
SUM = 0.0
|
|
*$ACC PARALLEL LOOP REDUCTION(+:SUM)
|
|
DO I = 1, N
|
|
SUM = SUM + C(I)
|
|
END DO
|
|
*$ACC END PARALLEL LOOP
|
|
|
|
C Kernels construct - C$ACC with continuation
|
|
C$ACC KERNELS
|
|
C$ACC+ COPYOUT(A)
|
|
DO I = 1, N
|
|
A(I) = A(I) * 2.0
|
|
END DO
|
|
C$ACC END KERNELS
|
|
|
|
* Data construct with update - *$ACC with continuation
|
|
*$ACC DATA COPY(B)
|
|
*$ACC+ PRESENT(D)
|
|
B(1) = 999.0
|
|
*$ACC UPDATE HOST(B(1:1))
|
|
PRINT *, 'B(1) = ', B(1)
|
|
*$ACC END DATA
|
|
|
|
C Mixed style directives in nested constructs
|
|
C$ACC DATA COPY(A,B,C)
|
|
*$ACC PARALLEL LOOP
|
|
DO I = 1, N
|
|
A(I) = B(I) + C(I)
|
|
END DO
|
|
*$ACC END PARALLEL LOOP
|
|
C$ACC END DATA
|
|
|
|
* Subroutine call within data region - *$ACC style
|
|
*$ACC DATA COPY(A,B,C)
|
|
CALL SUB1(A, B, C, N)
|
|
*$ACC END DATA
|
|
|
|
PRINT *, 'Sum = ', SUM
|
|
END PROGRAM
|
|
|
|
C Subroutine with mixed ACC directive styles
|
|
SUBROUTINE SUB1(X, Y, Z, M)
|
|
INTEGER M, I
|
|
REAL X(M), Y(M), Z(M)
|
|
|
|
*$ACC PARALLEL PRESENT(X,Y)
|
|
C$ACC LOOP PRIVATE(I)
|
|
DO I = 1, M
|
|
Z(I) = X(I) + Y(I)
|
|
END DO
|
|
C$ACC END LOOP
|
|
*$ACC END PARALLEL
|
|
RETURN
|
|
END SUBROUTINE |