#include "Vec128.h"

void F32fromI32(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.4s}, [x1]                         \n\
	scvtf	v1.4s, v0.4s	// float32 <- int32   \n\
	st1	{v1.4s}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}

void I32fromF32(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.4s}, [x1]                         \n\
	fcvtns	v1.4s, v0.4s	// int32 <- float32   \n\
	st1	{v1.4s}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}

void F64fromI64(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.2d}, [x1]                         \n\
	scvtf	v1.2d, v0.2d	// float64 <- int64   \n\
	st1	{v1.2d}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}

void I64fromF64(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.2d}, [x1]                         \n\
	fcvtns	v1.2d, v0.2d	// int32 <- float32   \n\
	st1	{v1.2d}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}


void F32fromU32(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.4s}, [x1]                         \n\
	ucvtf	v1.4s, v0.4s	// float32 <- int32   \n\
	st1	{v1.4s}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}

void U32fromF32(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.4s}, [x1]                         \n\
	fcvtnu	v1.4s, v0.4s	// uint32 <- float32  \n\
	st1	{v1.4s}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}

void F64fromU64(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.2d}, [x1]                         \n\
	ucvtf	v1.2d, v0.2d	// float64 <- int64   \n\
	st1	{v1.2d}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}

void U64fromF64(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.2d}, [x1]                         \n\
	fcvtnu	v1.2d, v0.2d	// uint64 <- float64  \n\
	st1	{v1.2d}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "x0", "x1"
		  );

}


void F32fromF64(Vec128 x[2], const Vec128& a, const Vec128& b) {
  __asm volatile ("\n\
	ld1	{v0.2d}, [x1]                         \n\
	ld1	{v2.2d}, [x2]                         \n\
	fcvtn	v1.2s, v0.2d	// lower-order F32    \n\
	fcvtn2	v1.4s, v2.2d	// higher-order F32   \n\
	st1	{v1.4s}, [x0]   // [x0] = v1          \n\
	"
	:
	:
	: "v0", "v1", "v2", "x0", "x1"
		  );

}

void F64fromF32(Vec128 x[2], const Vec128& a) {
  __asm volatile ("\n\
	ld1	{v0.4s}, [x1]                         \n\
	fcvtl	v1.2d, v0.2s	// lower-order F32    \n\
	fcvtl2	v2.2d, v0.4s	// higher-order F32   \n\
	st1	{v1.2d, v2.2d}, [x0]   // [x0] = v1   \n\
	"
	:
	:
	: "v0", "v1", "v2", "x0", "x1"
		  );

}