http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59700
--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> --- > Created attachment 31760 [details] > list_io.diff > > The attached patch fixes the code in comment #1. I'm currently > doing the regtest. The patch fixes the problem (at least integer and real) without regression. I don't know how to test that it does not reintroduce memory leak (though I doubt it will). Would it help if I write a test case and do the packaging?
