------- Comment #12 from rguenth at gcc dot gnu dot org 2010-02-17 21:32 ------- (In reply to comment #11) > Mark, with your permission, I'm restoring P3: indeed, strictly speaking the > failure of that specific testcase isn't a regression, because it didn't exist > in gcc4.4.x, but I understand it uncovered a rather subtle problem in the C++ > front-end. Anyway, I agree that we should figure out a plain C++ testcase, > outside the libstdc++ testsuite, which fails only in 4.5, otherwise, I will > understand if you release managers eventually decide to xfail the library > testcase. But Jason is on it already... ;)
The C++ FE issue is a regression (we removed the expr_size langhook) that causes wrong-code. -- rguenth at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P3 |P1 Summary|20_util/bind/ref2.cc FAILs |[4.5 Regression] | |20_util/bind/ref2.cc FAILs http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43075