Does anybody know why there are no calls to gfc_error from iresolve.c? In fixing pr29431, I have found it convenient to insert a check to see if the source and the destination are the same size, if there is no padding. It works fine but it is striking that there are no other error calls. Paul