Added <0, 0, 0, 0> <1, 1, 1, 1> and <3, 3, 3, 3> to NeonShuffleFloat32x4 (#1989)

Author: jrouwe

Jolt/Core/ARMNeon.h

@@ -37,6 +37,12 @@
 	}
 
 	// Specializations
+	template <>
+	JPH_INLINE float32x4_t NeonShuffleFloat32x4<0, 0, 0, 0>(float32x4_t inV1, float32x4_t inV2)
+	{
+		return vdupq_laneq_f32(inV1, 0);
+	}
+
 	template <>
 	JPH_INLINE float32x4_t NeonShuffleFloat32x4<0, 1, 0, 0>(float32x4_t inV1, float32x4_t inV2)
 	{
@@ -139,6 +145,12 @@
 		return vrev64q_f32(inV1);
 	}
 
+	template <>
+	JPH_INLINE float32x4_t NeonShuffleFloat32x4<1, 1, 1, 1>(float32x4_t inV1, float32x4_t inV2)
+	{
+		return vdupq_laneq_f32(inV1, 1);
+	}
+
 	template <>
 	JPH_INLINE float32x4_t NeonShuffleFloat32x4<1, 1, 2, 2>(float32x4_t inV1, float32x4_t inV2)
 	{
@@ -308,6 +320,12 @@
 		return vcombine_f32(zy, zy);
 	}
 
+	template <>
+	JPH_INLINE float32x4_t NeonShuffleFloat32x4<3, 3, 3, 3>(float32x4_t inV1, float32x4_t inV2)
+	{
+		return vdupq_laneq_f32(inV1, 3);
+	}
+
 	// Shuffle a vector
 	#define JPH_NEON_SHUFFLE_F32x4(vec1, vec2, index1, index2, index3, index4) NeonShuffleFloat32x4<index1, index2, index3, index4>(vec1, vec2)
 	#define JPH_NEON_SHUFFLE_U32x4(vec1, vec2, index1, index2, index3, index4) vreinterpretq_u32_f32((NeonShuffleFloat32x4<index1, index2, index3, index4>(vreinterpretq_f32_u32(vec1), vreinterpretq_f32_u32(vec2))))