http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48786
--- Comment #6 from Tobias Burnus <burnus at gcc dot gnu.org> 2011-05-11 13:45:52 UTC --- A) Regarding the ambiguity issue (cf. comment 0 and comment 3): As written, I believe the interface is indeed ambiguous - if that's the case, there is no regression in gfortran and comment 0 of this bug is INVALID. I have asked at J3: http://j3-fortran.org/pipermail/j3/2011-May/004366.html Hence: Regression but diagnostic is probably correct -> INVALID. Asked J3 for confirmation. B) Regarding the wrong-code bug (example 2, attachment 24141): See patch at http://gcc.gnu.org/ml/fortran/2011-04/msg00318.html The patch fixes the issue but as the review comment suggests, one should factorize the code into a different function to make it easier to maintain and to read, cf. http://gcc.gnu.org/ml/fortran/2011-05/msg00001.html (Note: You might need valgrind or malloc perturb to run into the bug.) Hence: OOP wrong-code issue but no regression; patch exists, but needs to be cleaned up. C) Regarding the segfault (test case: attachment 24110 with the patch of comment 3), see comment 5 Hence: No regression but an OOP ice-on-valid-code (segfault) bug, which still needs to be investigated and fixed.
