Fortran allocatable problem with openmp 3

[unrue]

Dear all,

i have a problem with ifort using Fortran allocatable arrays delared as lastprivate, using openmp 3.0. The simple example under ifort 11.1 doesn't not work, it dies on the first matrix access:

[fortran]program prova

implicit none
 complex ,dimension(:,:),allocatable:: matrix
 integer  :: i,j

 allocate(matrix(100,100))
 matrix=0

!$OMP PARALLEL DO lastprivate(matrix) private(j)
  do i = 1, 100
    do j = 1, 100
     matrix(i,j) = matrix(i,j) + i + j
    end do
  enddo
!$OMP END PARALLEL DO 

 WRITE(*,*) " matrix " , matrix

 deallocate(matrix)

end program
[/fortran]

According to this paper:

http://wikis.sun.com/display/openmp/Fortran+Allocatable+Arrays

if a compiler agrees to openmp 3.0 standard should permits to declare allocatable arrays as private, lastprivate, and so on.

Maybe is a ifort bug?

Thanks a lot.

[Kevin_D_Intel]

This is related to the earlier OpenMP bug discussion.

It is not clear if you are on Linux or Mac OS X, but as with the earlier report, on Linux I verified this test case fails with a segmentation fault when using the 11.1 Update 6 (11.1.072 - Linux) and runs successfully when using 11.1 Update 7 (11.1.073 - Linux).

Please try upgrading to the latest 11.1 Update 7 and let us know if that resolves the issue.

$ ifort -V -openmp u78882.f90
Intel Fortran Intel 64 Compiler Professional for applications running on Intel 64, Version 11.1 Build 20100414 Package ID: l_cprof_p_11.1.072
Copyright (C) 1985-2010 Intel Corporation. All rights reserved.

Intel Fortran 11.1-2739
GNU ld version 2.18.50.0.6-7.fc9 20080403

$ export OMP_NUM_THREADS=2
$ ./a.out
forrtl: severe (174): SIGSEGV, segmentation fault occurred

$ ifort -V -openmp u78882.f90
Intel Fortran Intel 64 Compiler Professional for applications running on Intel 64, Version 11.1 Build 20100806 Package ID: l_cprof_p_11.1.073
Copyright (C) 1985-2010 Intel Corporation. All rights reserved.

Intel Fortran 11.1-2755
GNU ld version 2.18.50.0.6-7.fc9 20080403

$ export OMP_NUM_THREADS=2
$ ./a.out
matrix (0.0000000E+00,0.0000000E+00) (0.0000000E+00,0.0000000E+00)

(198.0000,0.0000000E+00) (218.2004,0.0000000E+00) (200.0000,0.0000000E+00)

[unrue]

Hi Kevin,

thanks for your quick reply. I forgot to specify the platorm. I'm on Linux. I'll try with upgraded version.

Thanks!