http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48699
--- Comment #13 from janus at gcc dot gnu.org 2011-05-29 21:51:55 UTC --- (In reply to comment #12) > The patch in comment #11 fixes the runtime for the tests in comments #0 and > #9. > However the other tests give a backtrace on x86_64-apple-darwin10. and > valgrind > reports errors of the kind Apparently that already happens without the patch, with a clean trunk (note that comment #4 is select_type_23.f03 in the test suite, but only "do-compile"). Moreover, I think that this test case is in fact invalid, cf. PR 48887.
