[clang] [Clang][X86] Replace F16C `vcvtph2ps/256` intrinsics with `(convert|shuffle)vector` builtins (PR #152911)

[via cfe-commits]

================
@@ -40,12 +40,17 @@ typedef signed char __v16qs __attribute__((__vector_size__(16)));
 
 #ifdef __SSE2__
 /* Both _Float16 and __bf16 require SSE2 being enabled. */
+typedef _Float16 __v4hf __attribute__((__vector_size__(8)));
 typedef _Float16 __v8hf __attribute__((__vector_size__(16), __aligned__(16)));
 typedef _Float16 __m128h __attribute__((__vector_size__(16), __aligned__(16)));
 typedef _Float16 __m128h_u __attribute__((__vector_size__(16), __aligned__(1)));
 
 typedef __bf16 __v8bf __attribute__((__vector_size__(16), __aligned__(16)));
 typedef __bf16 __m128bh __attribute__((__vector_size__(16), __aligned__(16)));
+#else
+/* Use __fp16 when _Float16 is not supported. */
+typedef __fp16 __v4hf __attribute__((__vector_size__(8)));
+typedef __fp16 __v8hf __attribute__((__vector_size__(16), __aligned__(16)));
----------------
moorabbit wrote:

That sounds better. Thx!

https://github.com/llvm/llvm-project/pull/152911