(CC: fortran@ as it relates to Fortran.)
Hi all,
On 1/7/20 12:16 PM, Tobias Burnus wrote:
in terms of the check, it looks fine to me – but I am not sure about
the spec.
* [OpenACC] Actually, I simply missed the bit (here: OpenACC 3; OpenACC
2.6 is same): “Any array or subarray in a data
Hi Julian, hi Thomas,
in terms of the check, it looks fine to me – but I am not sure about the
spec.
At least the following test case seems to work fine:
integer :: A(10,10), out(12)
A = reshape([(i, i=0,100)], shape(A))
!$omp target map(A(3:6,3:5), out)
!$acc parallel copy(A(3:6,3:5), out)
Hi,
This patch tightens up error checking for array references used in
OpenACC clauses such that they must now be contiguous. I believe this
matches up to the spec (as of 2.6). I've tried to make it so an error
only triggers if the compiler is sure that a given array expression must
be
