Add comments regarding potential slow down

Author: meltq

libc/src/__support/FPUtil/cast.h

@@ -18,6 +18,8 @@
 
 namespace LIBC_NAMESPACE::fputil {
 
+// TODO: Add optimization for known good targets with fast 
+// float to float16 conversion: https://github.com/llvm/llvm-project/issues/133517
 template <typename OutType, typename InType>
 LIBC_INLINE constexpr cpp::enable_if_t<cpp::is_floating_point_v<OutType> &&
                                            cpp::is_floating_point_v<InType>,

libc/src/math/generic/hypotf16.cpp

@@ -18,6 +18,8 @@
 
 namespace LIBC_NAMESPACE_DECL {
 
+// For targets where conversion from float to float16 has to be
+// emulated, fputil::hypot<float16> is faster
 LLVM_LIBC_FUNCTION(float16, hypotf16, (float16 x, float16 y)) {
   using FloatBits = fputil::FPBits<float>;
   using FPBits = fputil::FPBits<float16>;