Remove warning about f64->f32 cast being potential UB

As discussed in #15536, the LLVM documentation incorrect described
overflowing f64->f32 casts as being undefined behavior. LLVM never
treated them as such, and the documentation has been adjusted in
https://reviews.llvm.org/rL329065. As such, this warning can now
be removed.

Closes #49622.

Author: nikic

src/librustc_typeck/check/demand.rs

@@ -502,10 +502,6 @@ impl<'a, 'gcx, 'tcx> FnCtxt<'a, 'gcx, 'tcx> {
                                             &format!("{}, producing the closest possible value",
                                                      msg),
                                             cast_suggestion);
-                        err.warn("casting here will cause undefined behavior if the value is \
-                                  finite but larger or smaller than the largest or smallest \
-                                  finite value representable by `f32` (this is a bug and will be \
-                                  fixed)");
                     }
                     true
                 }