Add support for Armv8.2-a __fp16

Armv8.2-a provides a native half-precision floating point (__fp16 aka.
float16_t). This patch introduces

* __fp16 as underlying type of Eigen::half if this type is available
* the packet types Packet4hf and Packet8hf representing float16x4_t and
  float16x8_t respectively
* packet-math for the above packets with corresponding scalar type Eigen::half

The packet-math functionality has been implemented by Ashutosh Sharma
<ashutosh.sharma@amperecomputing.com>.

This closes #1940.

Eigen/src/Core/util/Meta.h

@@ -684,15 +684,6 @@ template<> EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC
 bool not_equal_strict(const double& x,const double& y) { return std::not_equal_to<double>()(x,y); }
 #endif
 
-/** \internal extract the bits of the float \a x */
-EIGEN_STRONG_INLINE EIGEN_DEVICE_FUNC unsigned int as_uint(float x)
-{
-  unsigned int ret;
-  EIGEN_USING_STD(memcpy)
-  memcpy(&ret, &x, sizeof(float));
-  return ret;
-}
-
 } // end namespace numext
 
 } // end namespace Eigen