updates based on PR feedback

There are two major changes (and a few minor ones which are not listed here...see PR discussion for details)

1. Eigen::half implementations for HIP and CUDA have been merged.
This means that
- `CUDA/Half.h` and `HIP/hcc/Half.h` got merged to a new file `GPU/Half.h`
- `CUDA/PacketMathHalf.h` and `HIP/hcc/PacketMathHalf.h` got merged to a new file `GPU/PacketMathHalf.h`
- `CUDA/TypeCasting.h` and `HIP/hcc/TypeCasting.h` got merged to a new file `GPU/TypeCasting.h`

After this change the `HIP/hcc` directory only contains one file `math_constants.h`. That will go away too once that file becomes a part of the HIP install.

2. new macros EIGEN_GPUCC, EIGEN_GPU_COMPILE_PHASE and EIGEN_HAS_GPU_FP16 have been added and the code has been updated to use them where appropriate.
- `EIGEN_GPUCC` is the same as `(EIGEN_CUDACC || EIGEN_HIPCC)`
- `EIGEN_GPU_DEVICE_COMPILE` is the same as `(EIGEN_CUDA_ARCH || EIGEN_HIP_DEVICE_COMPILE)`
- `EIGEN_HAS_GPU_FP16` is the same as `(EIGEN_HAS_CUDA_FP16 or EIGEN_HAS_HIP_FP16)`

unsupported/Eigen/src/SpecialFunctions/SpecialFunctionsImpl.h

@@ -190,7 +190,7 @@ template <>
 struct lgamma_impl<float> {
   EIGEN_DEVICE_FUNC
   static EIGEN_STRONG_INLINE float run(float x) {
-#if !defined(EIGEN_CUDA_ARCH) && (defined(_BSD_SOURCE) || defined(_SVID_SOURCE)) && !defined(__APPLE__) && !defined(EIGEN_HIP_DEVICE_COMPILE)
+#if !defined(EIGEN_GPU_COMPILE_PHASE) && (defined(_BSD_SOURCE) || defined(_SVID_SOURCE)) && !defined(__APPLE__) 
     int dummy;
     return ::lgammaf_r(x, &dummy);
 #else
@@ -203,7 +203,7 @@ template <>
 struct lgamma_impl<double> {
   EIGEN_DEVICE_FUNC
   static EIGEN_STRONG_INLINE double run(double x) {
-#if !defined(EIGEN_CUDA_ARCH) && (defined(_BSD_SOURCE) || defined(_SVID_SOURCE)) && !defined(__APPLE__) && !defined(EIGEN_HIP_DEVICE_COMPILE)
+#if !defined(EIGEN_GPU_COMPILE_PHASE) && (defined(_BSD_SOURCE) || defined(_SVID_SOURCE)) && !defined(__APPLE__) 
     int dummy;
     return ::lgamma_r(x, &dummy);
 #else

unsupported/Eigen/src/SpecialFunctions/arch/CUDA/CudaSpecialFunctions.h

@@ -17,7 +17,7 @@ namespace internal {
 // Make sure this is only available when targeting a GPU: we don't want to
 // introduce conflicts between these packet_traits definitions and the ones
 // we'll use on the host side (SSE, AVX, ...)
-#if defined(EIGEN_CUDACC) && defined(EIGEN_USE_GPU)
+#if defined(EIGEN_GPUCC) && defined(EIGEN_USE_GPU)
 
 template<> EIGEN_DEVICE_FUNC EIGEN_STRONG_INLINE
 float4 plgamma<float4>(const float4& a)