[Bug tree-optimization/69320] [6 Regression] wrong code generation at -O2 and higher

[law at redhat dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69320

--- Comment #8 from Jeffrey A. Law <law at redhat dot com> ---
Jakub.  I'll certainly backtest against the referenced BZs

While the fix is trivial, I'm seriously considering just turning the
exploitation of ranges off for gcc-6.  It didn't help as much as I had wanted
in the path splitting bits, so it's not terribly important for this release.