------- Comment #13 from manu at gcc dot gnu dot org 2010-02-21 17:57 ------- (In reply to comment #12) > Subject: Re: [4.5 Regression] c-c++-common/pr41779.c doesn't > work
Sorry I do not understand completely your answer. Shouldn't we warn at all? If the semantics were implemented as intended, should we warn? Also, the casting to long double seems to be introduced by the float, not by the integer, if I debugged the code correctly. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43128
