Dear All, 2016-01-31 Paul Thomas <pa...@gcc.gnu.org>
PR fortran/67564 * trans-expr.c (gfc_conv_procedure_call): For the vtable copy subroutines, add a string length argument, when the actual argument is an unlimited polymorphic class object. 2016-01-31 Paul Thomas <pa...@gcc.gnu.org> PR fortran/67564 * gfortran.dg/allocate_with_source_17.f03: New test. Committed as 'obvious' in revision 233016. Dominique had tested a more elaborate version of the patch, which included a big chunk in trans-array.c. It turns out that this is not needed and the one-liner above is truly obvious. I have posted on the PR another one-liner that fixes the problem on 5-branch and allows the testcase to run correctly. Slightly mysteriously, the unlimited polymorphic argument is not arriving with the unlimited attribute. That said, there is no reason, logically, to test for unlimitedness, since it is the only condition in which a character can be copied to a class object. It regtests OK, so OK for 5-branch too? Cheers Paul