http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57697
--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> --- (In reply to Dominique d'Humieres from comment #2) > Should not this go to PR57696? No. The examples are similar, but they expose rather different bugs. PR57696 has already a test case and the problem there is that no defined assignment is done for an allocatable component. -> Wrong code as the defined assignment is not invoked. This PR (and attachment 30353) is about an unallocated allocatable LHS, where for a component defined assignment is used. -> Wrong code as it segfaults at run time.
