------- Comment #13 from kargl at gcc dot gnu dot org 2010-04-24 16:56 -------
(In reply to comment #12)
>
> Just for completeness, this testcase again emphasizes that this is an ICE on
> valid code.
>
> MODULE fft_tools
> IMPLICIT NONE
> INTEGER, PARAMETER :: sp=4, dp=8
> INTEGER, PARAMETER :: lp = dp
> CONTAINS
> SUBROUTINE sparse_alltoall ( rs, rq, rcount, rreq, group )
> COMPLEX(KIND=lp), DIMENSION(:, :), &
> POINTER :: rs
> COMPLEX(KIND=lp), DIMENSION(:, :), &
> POINTER :: rq
> INTEGER, DIMENSION(:) :: rcount,rreq
> INTEGER :: group, pos
> IF ( rcount(pos) /= 0 ) THEN
> rq(1:rcount(pos),pos) = rs(1:rcount(pos),pos)
> END IF
> END SUBROUTINE sparse_alltoall
> END MODULE fft_tools
For even more completeness, the code in comment #12 is still
invalid. Don't worry I'll fix it before I eventually commit
the patch as noted in comment #11.
--
kargl at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
CC|rguenther at suse dot de |
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43793
