[llvm-commits] Fix Bug 13574: Add more double float shrinking optimizations

[Eli Friedman]

On Wed, Aug 15, 2012 at 1:37 PM, Weiming Zhao <weimingz at codeaurora.org> wrote:
> Hi Steve,
>
>
>
> Yes, it transforms (float)func((double) val) to funcf(v), which saves 2
> conversions and possibly cheaper math  function.

That isn't the same transformation that your patch does.
Specifically, the transformation in your patch doesn't check how
result is used (which is fine for floor, but not cos).

-Eli