NEON f32 <-> f16 and f32 <-> u16

src/opts/SkNx_neon.h

Index: src/opts/SkNx_neon.h
diff --git a/src/opts/SkNx_neon.h b/src/opts/SkNx_neon.h
index 641e9d296b54014e124ab306a24df328bc1544f9..be37baf3cb20c8c0582797c1011b8e4a126f5c2d 100644
--- a/src/opts/SkNx_neon.h
+++ b/src/opts/SkNx_neon.h
@@ -365,6 +365,14 @@ public:
 #undef SHIFT16
 #undef SHIFT8
 
+template<> inline Sk4h SkNx_cast<uint16_t, float>(const Sk4f& src) {
+    return vqmovn_u32(vcvtq_u32_f32(src.fVec));
+}
+
+template<> inline Sk4f SkNx_cast<float, uint16_t>(const Sk4h& src) {
+    return vcvtq_f32_u32(vmovl_u16(src.fVec));
+}
+
 template<> inline Sk4b SkNx_cast<uint8_t, float>(const Sk4f& src) {
     uint32x4_t _32 = vcvtq_u32_f32(src.fVec);
     uint16x4_t _16 = vqmovn_u32(_32);