------- Comment #2 from burnus at gcc dot gnu dot org 2008-09-10 20:34 ------- Close as INVALID.
I hope your problem is solved now. If not, I suggest to write to the mailing list. (By the way, on can find quite often code like integer, parameter :: dp = kind(1.0d0) real(kind=dp) :: sq2 = sqrt(2.0) which has a similar problem. Only using sqrt(2.0d0) or sqrt(2.0_dp) one really uses the full precision. Even if one knows about the problem it easily slips in while programming.) -- burnus at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |INVALID http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37458