------- Comment #2 from pault at gcc dot gnu dot org 2006-08-30 04:46 ------- (In reply to comment #1) > Confirmed: changing the substring reference to the unambiguous > write (*,*) i, charr(i:i) > gets rid of the problem. Apparently, vene though charr is scalar, the > reference is being interpreted as an array reference. > > Paul > Cooh! Not only was I full of bulls**t last night but had gone sydlexic too!
charr(i:i) is, of course, no more unambiguous than charr(i), except, apparently, in the bowels of gfortran! In fact, using -fdump-parse-tree, I find that charr(i) is being taken as a character function reference. It must be easy enough to put paid to that... Paul -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28890
