[Bug fortran/45586] [4.6/4.7 Regression] ICE non-trivial conversion at assignment

[Joost.VandeVondele at pci dot uzh.ch]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45586

--- Comment #52 from Joost VandeVondele <Joost.VandeVondele at pci dot uzh.ch> 2011-04-12 05:50:45 UTC ---
(In reply to comment #51)
> (In reply to comment #50)
> > The original problem in comment #0 fails (i.e. the build of CP2K) with trunk
> Could you try to reduce the test case?

I will give it a try later this week, but this kind of LTO ICEs are difficult
to reduce (starting from ~1MLOC), since it can basically not be done
automatically (and the obvious testcase doesn't seem to fail anymore).