------- Comment #2 from dominiq at lps dot ens dot fr 2010-07-01 21:24 ------- > I think it is not really a regression in terms of the code path.
Well, my definition of a regression is "something working as expected at revision n, no longer working as expected at revision n+1". Indeed revision n+1 can expose a latent/older bug, such as comparing the rhs to its temporary, but not to the lhs as it should. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44744
