[AArch64] vld{2,3,4}{,_lane,_dup}, vcombine, vcreate

gcc/:

	* config/aarch64/aarch64.c (aarch64_split_simd_combine): Add V4HFmode.
	* config/aarch64/aarch64-builtins.c (VAR13, VAR14): New.
	(aarch64_scalar_builtin_types, aarch64_init_simd_builtin_scalar_types):
	Add __builtin_aarch64_simd_hf.
	* config/aarch64/arm_neon.h (float16x4x2_t, float16x8x2_t,
	float16x4x3_t, float16x8x3_t, float16x4x4_t, float16x8x4_t,
	vcombine_f16, vst2_lane_f16, vst2q_lane_f16, vst3_lane_f16,
	vst3q_lane_f16, vst4_lane_f16, vst4q_lane_f16, vld2_f16, vld2q_f16,
	vld3_f16, vld3q_f16, vld4_f16, vld4q_f16, vld2_dup_f16, vld2q_dup_f16,
	vld3_dup_f16, vld3q_dup_f16, vld4_dup_f16, vld4q_dup_f16,
	vld2_lane_f16, vld2q_lane_f16, vld3_lane_f16, vld3q_lane_f16,
	vld4_lane_f16, vld4q_lane_f16, vst2_f16, vst2q_f16, vst3_f16,
	vst3q_f16, vst4_f16, vst4q_f16, vcreate_f16): New.

	* config/aarch64/iterators.md (VALLDIF, Vtype, Vetype, Vbtype,
	V_cmp_result, v_cmp_result): Add cases for V4HF and V8HF.
	(VDC, Vdbl): Add V4HF.

gcc/testsuite/:

	* gcc.target/aarch64/vldN_1.c: Add float16x4_t and float16x8_t cases.
	* gcc.target/aarch64/vldN_dup_1.c: Likewise.
	* gcc.target/aarch64/vldN_lane_1.c: Likewise.
	(main): update orig_data to avoid float16 NaN on bigendian.

From-SVN: r227543

4 files changed, 305 insertions, 9 deletions

diff --git a/gcc/config/aarch64/aarch64-builtins.c b/gcc/config/aarch64/aarch64-builtins.c
index 243fa9f..5a04263 100644
--- a/gcc/config/aarch64/aarch64-builtins.c
+++ b/gcc/config/aarch64/aarch64-builtins.c
@@ -297,6 +297,12 @@ aarch64_types_storestruct_lane_qualifiers[SIMD_MAX_BUILTIN_ARGS]
 #define VAR12(T, N, MAP, A, B, C, D, E, F, G, H, I, J, K, L) \
   VAR11 (T, N, MAP, A, B, C, D, E, F, G, H, I, J, K) \
   VAR1 (T, N, MAP, L)
+#define VAR13(T, N, MAP, A, B, C, D, E, F, G, H, I, J, K, L, M) \
+  VAR12 (T, N, MAP, A, B, C, D, E, F, G, H, I, J, K, L) \
+  VAR1 (T, N, MAP, M)
+#define VAR14(T, X, MAP, A, B, C, D, E, F, G, H, I, J, K, L, M, N) \
+  VAR13 (T, X, MAP, A, B, C, D, E, F, G, H, I, J, K, L, M) \
+  VAR1 (T, X, MAP, N)
 
 #include "aarch64-builtin-iterators.h"
 
@@ -374,6 +380,7 @@ const char *aarch64_scalar_builtin_types[] = {
   "__builtin_aarch64_simd_qi",
   "__builtin_aarch64_simd_hi",
   "__builtin_aarch64_simd_si",
+  "__builtin_aarch64_simd_hf",
   "__builtin_aarch64_simd_sf",
   "__builtin_aarch64_simd_di",
   "__builtin_aarch64_simd_df",
@@ -661,6 +668,8 @@ aarch64_init_simd_builtin_scalar_types (void)
 					     "__builtin_aarch64_simd_qi");
   (*lang_hooks.types.register_builtin_type) (intHI_type_node,
 					     "__builtin_aarch64_simd_hi");
+  (*lang_hooks.types.register_builtin_type) (aarch64_fp16_type_node,
+					     "__builtin_aarch64_simd_hf");
   (*lang_hooks.types.register_builtin_type) (intSI_type_node,
 					     "__builtin_aarch64_simd_si");
   (*lang_hooks.types.register_builtin_type) (float_type_node,
diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index 5951045..7967170 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -1335,6 +1335,9 @@ aarch64_split_simd_combine (rtx dst, rtx src1, rtx src2)
 	case V2SImode:
 	  gen = gen_aarch64_simd_combinev2si;
 	  break;
+	case V4HFmode:
+	  gen = gen_aarch64_simd_combinev4hf;
+	  break;
 	case V2SFmode:
 	  gen = gen_aarch64_simd_combinev2sf;
 	  break;
diff --git a/gcc/config/aarch64/arm_neon.h b/gcc/config/aarch64/arm_neon.h
index 5f61b5b..2bb75bb0e 100644
--- a/gcc/config/aarch64/arm_neon.h
+++ b/gcc/config/aarch64/arm_neon.h
@@ -152,6 +152,16 @@ typedef struct uint64x2x2_t
   uint64x2_t val[2];
 } uint64x2x2_t;
 
+typedef struct float16x4x2_t
+{
+  float16x4_t val[2];
+} float16x4x2_t;
+
+typedef struct float16x8x2_t
+{
+  float16x8_t val[2];
+} float16x8x2_t;
+
 typedef struct float32x2x2_t
 {
   float32x2_t val[2];
@@ -272,6 +282,16 @@ typedef struct uint64x2x3_t
   uint64x2_t val[3];
 } uint64x2x3_t;
 
+typedef struct float16x4x3_t
+{
+  float16x4_t val[3];
+} float16x4x3_t;
+
+typedef struct float16x8x3_t
+{
+  float16x8_t val[3];
+} float16x8x3_t;
+
 typedef struct float32x2x3_t
 {
   float32x2_t val[3];
@@ -392,6 +412,16 @@ typedef struct uint64x2x4_t
   uint64x2_t val[4];
 } uint64x2x4_t;
 
+typedef struct float16x4x4_t
+{
+  float16x4_t val[4];
+} float16x4x4_t;
+
+typedef struct float16x8x4_t
+{
+  float16x8_t val[4];
+} float16x8x4_t;
+
 typedef struct float32x2x4_t
 {
   float32x2_t val[4];
@@ -2643,6 +2673,12 @@ vcreate_s64 (uint64_t __a)
   return (int64x1_t) {__a};
 }
 
+__extension__ static __inline float16x4_t __attribute__ ((__always_inline__))
+vcreate_f16 (uint64_t __a)
+{
+  return (float16x4_t) __a;
+}
+
 __extension__ static __inline float32x2_t __attribute__ ((__always_inline__))
 vcreate_f32 (uint64_t __a)
 {
@@ -4779,6 +4815,12 @@ vcombine_s64 (int64x1_t __a, int64x1_t __b)
   return __builtin_aarch64_combinedi (__a[0], __b[0]);
 }
 
+__extension__ static __inline float16x8_t __attribute__ ((__always_inline__))
+vcombine_f16 (float16x4_t __a, float16x4_t __b)
+{
+  return __builtin_aarch64_combinev4hf (__a, __b);
+}
+
 __extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
 vcombine_f32 (float32x2_t __a, float32x2_t __b)
 {
@@ -9907,7 +9949,7 @@ vtstq_p16 (poly16x8_t a, poly16x8_t b)
    +------+----+----+----+----+
    |uint  | Y  | Y  | N  | N  |
    +------+----+----+----+----+
-   |float | -  | -  | N  | N  |
+   |float | -  | Y  | N  | N  |
    +------+----+----+----+----+
    |poly  | Y  | Y  | -  | -  |
    +------+----+----+----+----+
@@ -9921,7 +9963,7 @@ vtstq_p16 (poly16x8_t a, poly16x8_t b)
    +------+----+----+----+----+
    |uint  | Y  | Y  | Y  | Y  |
    +------+----+----+----+----+
-   |float | -  | -  | Y  | Y  |
+   |float | -  | Y  | Y  | Y  |
    +------+----+----+----+----+
    |poly  | Y  | Y  | -  | -  |
    +------+----+----+----+----+
@@ -9935,7 +9977,7 @@ vtstq_p16 (poly16x8_t a, poly16x8_t b)
    +------+----+----+----+----+
    |uint  | Y  | N  | N  | Y  |
    +------+----+----+----+----+
-   |float | -  | -  | N  | Y  |
+   |float | -  | N  | N  | Y  |
    +------+----+----+----+----+
    |poly  | Y  | N  | -  | -  |
    +------+----+----+----+----+
@@ -9951,6 +9993,7 @@ __STRUCTN (int, 8, 2)
 __STRUCTN (int, 16, 2)
 __STRUCTN (uint, 8, 2)
 __STRUCTN (uint, 16, 2)
+__STRUCTN (float, 16, 2)
 __STRUCTN (poly, 8, 2)
 __STRUCTN (poly, 16, 2)
 /* 3-element structs.  */
@@ -9962,6 +10005,7 @@ __STRUCTN (uint, 8, 3)
 __STRUCTN (uint, 16, 3)
 __STRUCTN (uint, 32, 3)
 __STRUCTN (uint, 64, 3)
+__STRUCTN (float, 16, 3)
 __STRUCTN (float, 32, 3)
 __STRUCTN (float, 64, 3)
 __STRUCTN (poly, 8, 3)
@@ -9999,6 +10043,8 @@ vst2_lane_ ## funcsuffix (ptrtype *__ptr,				     \
 				     __ptr, __o, __c);			     \
 }
 
+__ST2_LANE_FUNC (float16x4x2_t, float16x8x2_t, float16_t, v4hf, v8hf, hf, f16,
+		 float16x8_t)
 __ST2_LANE_FUNC (float32x2x2_t, float32x4x2_t, float32_t, v2sf, v4sf, sf, f32,
 		 float32x4_t)
 __ST2_LANE_FUNC (float64x1x2_t, float64x2x2_t, float64_t, df, v2df, df, f64,
@@ -10037,6 +10083,7 @@ vst2q_lane_ ## funcsuffix (ptrtype *__ptr,				    \
 				    __ptr, __temp.__o, __c);		    \
 }
 
+__ST2_LANE_FUNC (float16x8x2_t, float16_t, v8hf, hf, f16)
 __ST2_LANE_FUNC (float32x4x2_t, float32_t, v4sf, sf, f32)
 __ST2_LANE_FUNC (float64x2x2_t, float64_t, v2df, df, f64)
 __ST2_LANE_FUNC (poly8x16x2_t, poly8_t, v16qi, qi, p8)
@@ -10078,6 +10125,8 @@ vst3_lane_ ## funcsuffix (ptrtype *__ptr,				     \
 				     __ptr, __o, __c);			     \
 }
 
+__ST3_LANE_FUNC (float16x4x3_t, float16x8x3_t, float16_t, v4hf, v8hf, hf, f16,
+		 float16x8_t)
 __ST3_LANE_FUNC (float32x2x3_t, float32x4x3_t, float32_t, v2sf, v4sf, sf, f32,
 		 float32x4_t)
 __ST3_LANE_FUNC (float64x1x3_t, float64x2x3_t, float64_t, df, v2df, df, f64,
@@ -10116,6 +10165,7 @@ vst3q_lane_ ## funcsuffix (ptrtype *__ptr,				    \
 				    __ptr, __temp.__o, __c);		    \
 }
 
+__ST3_LANE_FUNC (float16x8x3_t, float16_t, v8hf, hf, f16)
 __ST3_LANE_FUNC (float32x4x3_t, float32_t, v4sf, sf, f32)
 __ST3_LANE_FUNC (float64x2x3_t, float64_t, v2df, df, f64)
 __ST3_LANE_FUNC (poly8x16x3_t, poly8_t, v16qi, qi, p8)
@@ -10162,6 +10212,8 @@ vst4_lane_ ## funcsuffix (ptrtype *__ptr,				     \
 				     __ptr, __o, __c);			     \
 }
 
+__ST4_LANE_FUNC (float16x4x4_t, float16x8x4_t, float16_t, v4hf, v8hf, hf, f16,
+		 float16x8_t)
 __ST4_LANE_FUNC (float32x2x4_t, float32x4x4_t, float32_t, v2sf, v4sf, sf, f32,
 		 float32x4_t)
 __ST4_LANE_FUNC (float64x1x4_t, float64x2x4_t, float64_t, df, v2df, df, f64,
@@ -10200,6 +10252,7 @@ vst4q_lane_ ## funcsuffix (ptrtype *__ptr,				    \
 				    __ptr, __temp.__o, __c);		    \
 }
 
+__ST4_LANE_FUNC (float16x8x4_t, float16_t, v8hf, hf, f16)
 __ST4_LANE_FUNC (float32x4x4_t, float32_t, v4sf, sf, f32)
 __ST4_LANE_FUNC (float64x2x4_t, float64_t, v2df, df, f64)
 __ST4_LANE_FUNC (poly8x16x4_t, poly8_t, v16qi, qi, p8)
@@ -15255,6 +15308,17 @@ vld2_u32 (const uint32_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x4x2_t __attribute__ ((__always_inline__))
+vld2_f16 (const float16_t * __a)
+{
+  float16x4x2_t ret;
+  __builtin_aarch64_simd_oi __o;
+  __o = __builtin_aarch64_ld2v4hf (__a);
+  ret.val[0] = __builtin_aarch64_get_dregoiv4hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_dregoiv4hf (__o, 1);
+  return ret;
+}
+
 __extension__ static __inline float32x2x2_t __attribute__ ((__always_inline__))
 vld2_f32 (const float32_t * __a)
 {
@@ -15376,6 +15440,17 @@ vld2q_u64 (const uint64_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x8x2_t __attribute__ ((__always_inline__))
+vld2q_f16 (const float16_t * __a)
+{
+  float16x8x2_t ret;
+  __builtin_aarch64_simd_oi __o;
+  __o = __builtin_aarch64_ld2v8hf (__a);
+  ret.val[0] = __builtin_aarch64_get_qregoiv8hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_qregoiv8hf (__o, 1);
+  return ret;
+}
+
 __extension__ static __inline float32x4x2_t __attribute__ ((__always_inline__))
 vld2q_f32 (const float32_t * __a)
 {
@@ -15530,6 +15605,18 @@ vld3_u32 (const uint32_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x4x3_t __attribute__ ((__always_inline__))
+vld3_f16 (const float16_t * __a)
+{
+  float16x4x3_t ret;
+  __builtin_aarch64_simd_ci __o;
+  __o = __builtin_aarch64_ld3v4hf (__a);
+  ret.val[0] = __builtin_aarch64_get_dregciv4hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_dregciv4hf (__o, 1);
+  ret.val[2] = __builtin_aarch64_get_dregciv4hf (__o, 2);
+  return ret;
+}
+
 __extension__ static __inline float32x2x3_t __attribute__ ((__always_inline__))
 vld3_f32 (const float32_t * __a)
 {
@@ -15662,6 +15749,18 @@ vld3q_u64 (const uint64_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x8x3_t __attribute__ ((__always_inline__))
+vld3q_f16 (const float16_t * __a)
+{
+  float16x8x3_t ret;
+  __builtin_aarch64_simd_ci __o;
+  __o = __builtin_aarch64_ld3v8hf (__a);
+  ret.val[0] = __builtin_aarch64_get_qregciv8hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_qregciv8hf (__o, 1);
+  ret.val[2] = __builtin_aarch64_get_qregciv8hf (__o, 2);
+  return ret;
+}
+
 __extension__ static __inline float32x4x3_t __attribute__ ((__always_inline__))
 vld3q_f32 (const float32_t * __a)
 {
@@ -15829,6 +15928,19 @@ vld4_u32 (const uint32_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x4x4_t __attribute__ ((__always_inline__))
+vld4_f16 (const float16_t * __a)
+{
+  float16x4x4_t ret;
+  __builtin_aarch64_simd_xi __o;
+  __o = __builtin_aarch64_ld4v4hf (__a);
+  ret.val[0] = __builtin_aarch64_get_dregxiv4hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_dregxiv4hf (__o, 1);
+  ret.val[2] = __builtin_aarch64_get_dregxiv4hf (__o, 2);
+  ret.val[3] = __builtin_aarch64_get_dregxiv4hf (__o, 3);
+  return ret;
+}
+
 __extension__ static __inline float32x2x4_t __attribute__ ((__always_inline__))
 vld4_f32 (const float32_t * __a)
 {
@@ -15972,6 +16084,19 @@ vld4q_u64 (const uint64_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x8x4_t __attribute__ ((__always_inline__))
+vld4q_f16 (const float16_t * __a)
+{
+  float16x8x4_t ret;
+  __builtin_aarch64_simd_xi __o;
+  __o = __builtin_aarch64_ld4v8hf (__a);
+  ret.val[0] = __builtin_aarch64_get_qregxiv8hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_qregxiv8hf (__o, 1);
+  ret.val[2] = __builtin_aarch64_get_qregxiv8hf (__o, 2);
+  ret.val[3] = __builtin_aarch64_get_qregxiv8hf (__o, 3);
+  return ret;
+}
+
 __extension__ static __inline float32x4x4_t __attribute__ ((__always_inline__))
 vld4q_f32 (const float32_t * __a)
 {
@@ -16033,6 +16158,17 @@ vld2_dup_s32 (const int32_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x4x2_t __attribute__ ((__always_inline__))
+vld2_dup_f16 (const float16_t * __a)
+{
+  float16x4x2_t ret;
+  __builtin_aarch64_simd_oi __o;
+  __o = __builtin_aarch64_ld2rv4hf ((const __builtin_aarch64_simd_hf *) __a);
+  ret.val[0] = __builtin_aarch64_get_dregoiv4hf (__o, 0);
+  ret.val[1] = (float16x4_t) __builtin_aarch64_get_dregoiv4hf (__o, 1);
+  return ret;
+}
+
 __extension__ static __inline float32x2x2_t __attribute__ ((__always_inline__))
 vld2_dup_f32 (const float32_t * __a)
 {
@@ -16242,6 +16378,17 @@ vld2q_dup_u64 (const uint64_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x8x2_t __attribute__ ((__always_inline__))
+vld2q_dup_f16 (const float16_t * __a)
+{
+  float16x8x2_t ret;
+  __builtin_aarch64_simd_oi __o;
+  __o = __builtin_aarch64_ld2rv8hf ((const __builtin_aarch64_simd_hf *) __a);
+  ret.val[0] = (float16x8_t) __builtin_aarch64_get_qregoiv8hf (__o, 0);
+  ret.val[1] = __builtin_aarch64_get_qregoiv8hf (__o, 1);
+  return ret;
+}
+
 __extension__ static __inline float32x4x2_t __attribute__ ((__always_inline__))
 vld2q_dup_f32 (const float32_t * __a)
 {
@@ -16396,6 +16543,18 @@ vld3_dup_u32 (const uint32_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x4x3_t __attribute__ ((__always_inline__))
+vld3_dup_f16 (const float16_t * __a)
+{
+  float16x4x3_t ret;
+  __builtin_aarch64_simd_ci __o;
+  __o = __builtin_aarch64_ld3rv4hf ((const __builtin_aarch64_simd_hf *) __a);
+  ret.val[0] = (float16x4_t) __builtin_aarch64_get_dregciv4hf (__o, 0);
+  ret.val[1] = (float16x4_t) __builtin_aarch64_get_dregciv4hf (__o, 1);
+  ret.val[2] = (float16x4_t) __builtin_aarch64_get_dregciv4hf (__o, 2);
+  return ret;
+}
+
 __extension__ static __inline float32x2x3_t __attribute__ ((__always_inline__))
 vld3_dup_f32 (const float32_t * __a)
 {
@@ -16528,6 +16687,18 @@ vld3q_dup_u64 (const uint64_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x8x3_t __attribute__ ((__always_inline__))
+vld3q_dup_f16 (const float16_t * __a)
+{
+  float16x8x3_t ret;
+  __builtin_aarch64_simd_ci __o;
+  __o = __builtin_aarch64_ld3rv8hf ((const __builtin_aarch64_simd_hf *) __a);
+  ret.val[0] = (float16x8_t) __builtin_aarch64_get_qregciv8hf (__o, 0);
+  ret.val[1] = (float16x8_t) __builtin_aarch64_get_qregciv8hf (__o, 1);
+  ret.val[2] = (float16x8_t) __builtin_aarch64_get_qregciv8hf (__o, 2);
+  return ret;
+}
+
 __extension__ static __inline float32x4x3_t __attribute__ ((__always_inline__))
 vld3q_dup_f32 (const float32_t * __a)
 {
@@ -16695,6 +16866,19 @@ vld4_dup_u32 (const uint32_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x4x4_t __attribute__ ((__always_inline__))
+vld4_dup_f16 (const float16_t * __a)
+{
+  float16x4x4_t ret;
+  __builtin_aarch64_simd_xi __o;
+  __o = __builtin_aarch64_ld4rv4hf ((const __builtin_aarch64_simd_hf *) __a);
+  ret.val[0] = (float16x4_t) __builtin_aarch64_get_dregxiv4hf (__o, 0);
+  ret.val[1] = (float16x4_t) __builtin_aarch64_get_dregxiv4hf (__o, 1);
+  ret.val[2] = (float16x4_t) __builtin_aarch64_get_dregxiv4hf (__o, 2);
+  ret.val[3] = (float16x4_t) __builtin_aarch64_get_dregxiv4hf (__o, 3);
+  return ret;
+}
+
 __extension__ static __inline float32x2x4_t __attribute__ ((__always_inline__))
 vld4_dup_f32 (const float32_t * __a)
 {
@@ -16838,6 +17022,19 @@ vld4q_dup_u64 (const uint64_t * __a)
   return ret;
 }
 
+__extension__ static __inline float16x8x4_t __attribute__ ((__always_inline__))
+vld4q_dup_f16 (const float16_t * __a)
+{
+  float16x8x4_t ret;
+  __builtin_aarch64_simd_xi __o;
+  __o = __builtin_aarch64_ld4rv8hf ((const __builtin_aarch64_simd_hf *) __a);
+  ret.val[0] = (float16x8_t) __builtin_aarch64_get_qregxiv8hf (__o, 0);
+  ret.val[1] = (float16x8_t) __builtin_aarch64_get_qregxiv8hf (__o, 1);
+  ret.val[2] = (float16x8_t) __builtin_aarch64_get_qregxiv8hf (__o, 2);
+  ret.val[3] = (float16x8_t) __builtin_aarch64_get_qregxiv8hf (__o, 3);
+  return ret;
+}
+
 __extension__ static __inline float32x4x4_t __attribute__ ((__always_inline__))
 vld4q_dup_f32 (const float32_t * __a)
 {
@@ -16890,6 +17087,8 @@ vld2_lane_##funcsuffix (const ptrtype * __ptr, intype __b, const int __c)  \
   return __b;								   \
 }
 
+__LD2_LANE_FUNC (float16x4x2_t, float16x4_t, float16x8x2_t, float16_t, v4hf,
+		 v8hf, hf, f16, float16x8_t)
 __LD2_LANE_FUNC (float32x2x2_t, float32x2_t, float32x4x2_t, float32_t, v2sf, v4sf,
 		 sf, f32, float32x4_t)
 __LD2_LANE_FUNC (float64x1x2_t, float64x1_t, float64x2x2_t, float64_t, df, v2df,
@@ -16934,6 +17133,7 @@ vld2q_lane_##funcsuffix (const ptrtype * __ptr, intype __b, const int __c) \
   return ret;								   \
 }
 
+__LD2_LANE_FUNC (float16x8x2_t, float16x8_t, float16_t, v8hf, hf, f16)
 __LD2_LANE_FUNC (float32x4x2_t, float32x4_t, float32_t, v4sf, sf, f32)
 __LD2_LANE_FUNC (float64x2x2_t, float64x2_t, float64_t, v2df, df, f64)
 __LD2_LANE_FUNC (poly8x16x2_t, poly8x16_t, poly8_t, v16qi, qi, p8)
@@ -16981,6 +17181,8 @@ vld3_lane_##funcsuffix (const ptrtype * __ptr, intype __b, const int __c)  \
   return __b;								   \
 }
 
+__LD3_LANE_FUNC (float16x4x3_t, float16x4_t, float16x8x3_t, float16_t, v4hf,
+		 v8hf, hf, f16, float16x8_t)
 __LD3_LANE_FUNC (float32x2x3_t, float32x2_t, float32x4x3_t, float32_t, v2sf, v4sf,
 		 sf, f32, float32x4_t)
 __LD3_LANE_FUNC (float64x1x3_t, float64x1_t, float64x2x3_t, float64_t, df, v2df,
@@ -17027,6 +17229,7 @@ vld3q_lane_##funcsuffix (const ptrtype * __ptr, intype __b, const int __c) \
   return ret;								   \
 }
 
+__LD3_LANE_FUNC (float16x8x3_t, float16x8_t, float16_t, v8hf, hf, f16)
 __LD3_LANE_FUNC (float32x4x3_t, float32x4_t, float32_t, v4sf, sf, f32)
 __LD3_LANE_FUNC (float64x2x3_t, float64x2_t, float64_t, v2df, df, f64)
 __LD3_LANE_FUNC (poly8x16x3_t, poly8x16_t, poly8_t, v16qi, qi, p8)
@@ -17082,6 +17285,8 @@ vld4_lane_##funcsuffix (const ptrtype * __ptr, intype __b, const int __c)  \
 
 /* vld4q_lane */
 
+__LD4_LANE_FUNC (float16x4x4_t, float16x4_t, float16x8x4_t, float16_t, v4hf,
+		 v8hf, hf, f16, float16x8_t)
 __LD4_LANE_FUNC (float32x2x4_t, float32x2_t, float32x4x4_t, float32_t, v2sf, v4sf,
 		 sf, f32, float32x4_t)
 __LD4_LANE_FUNC (float64x1x4_t, float64x1_t, float64x2x4_t, float64_t, df, v2df,
@@ -17130,6 +17335,7 @@ vld4q_lane_##funcsuffix (const ptrtype * __ptr, intype __b, const int __c) \
   return ret;								   \
 }
 
+__LD4_LANE_FUNC (float16x8x4_t, float16x8_t, float16_t, v8hf, hf, f16)
 __LD4_LANE_FUNC (float32x4x4_t, float32x4_t, float32_t, v4sf, sf, f32)
 __LD4_LANE_FUNC (float64x2x4_t, float64x2_t, float64_t, v2df, df, f64)
 __LD4_LANE_FUNC (poly8x16x4_t, poly8x16_t, poly8_t, v16qi, qi, p8)
@@ -22491,6 +22697,18 @@ vst2_u32 (uint32_t * __a, uint32x2x2_t val)
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
+vst2_f16 (float16_t * __a, float16x4x2_t val)
+{
+  __builtin_aarch64_simd_oi __o;
+  float16x8x2_t temp;
+  temp.val[0] = vcombine_f16 (val.val[0], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  temp.val[1] = vcombine_f16 (val.val[1], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  __o = __builtin_aarch64_set_qregoiv8hf (__o, temp.val[0], 0);
+  __o = __builtin_aarch64_set_qregoiv8hf (__o, temp.val[1], 1);
+  __builtin_aarch64_st2v4hf (__a, __o);
+}
+
+__extension__ static __inline void __attribute__ ((__always_inline__))
 vst2_f32 (float32_t * __a, float32x2x2_t val)
 {
   __builtin_aarch64_simd_oi __o;
@@ -22593,6 +22811,15 @@ vst2q_u64 (uint64_t * __a, uint64x2x2_t val)
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
+vst2q_f16 (float16_t * __a, float16x8x2_t val)
+{
+  __builtin_aarch64_simd_oi __o;
+  __o = __builtin_aarch64_set_qregoiv8hf (__o, val.val[0], 0);
+  __o = __builtin_aarch64_set_qregoiv8hf (__o, val.val[1], 1);
+  __builtin_aarch64_st2v8hf (__a, __o);
+}
+
+__extension__ static __inline void __attribute__ ((__always_inline__))
 vst2q_f32 (float32_t * __a, float32x4x2_t val)
 {
   __builtin_aarch64_simd_oi __o;
@@ -22765,6 +22992,20 @@ vst3_u32 (uint32_t * __a, uint32x2x3_t val)
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
+vst3_f16 (float16_t * __a, float16x4x3_t val)
+{
+  __builtin_aarch64_simd_ci __o;
+  float16x8x3_t temp;
+  temp.val[0] = vcombine_f16 (val.val[0], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  temp.val[1] = vcombine_f16 (val.val[1], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  temp.val[2] = vcombine_f16 (val.val[2], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  __o = __builtin_aarch64_set_qregciv8hf (__o, (float16x8_t) temp.val[0], 0);
+  __o = __builtin_aarch64_set_qregciv8hf (__o, (float16x8_t) temp.val[1], 1);
+  __o = __builtin_aarch64_set_qregciv8hf (__o, (float16x8_t) temp.val[2], 2);
+  __builtin_aarch64_st3v4hf ((__builtin_aarch64_simd_hf *) __a, __o);
+}
+
+__extension__ static __inline void __attribute__ ((__always_inline__))
 vst3_f32 (float32_t * __a, float32x2x3_t val)
 {
   __builtin_aarch64_simd_ci __o;
@@ -22879,6 +23120,16 @@ vst3q_u64 (uint64_t * __a, uint64x2x3_t val)
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
+vst3q_f16 (float16_t * __a, float16x8x3_t val)
+{
+  __builtin_aarch64_simd_ci __o;
+  __o = __builtin_aarch64_set_qregciv8hf (__o, (float16x8_t) val.val[0], 0);
+  __o = __builtin_aarch64_set_qregciv8hf (__o, (float16x8_t) val.val[1], 1);
+  __o = __builtin_aarch64_set_qregciv8hf (__o, (float16x8_t) val.val[2], 2);
+  __builtin_aarch64_st3v8hf ((__builtin_aarch64_simd_hf *) __a, __o);
+}
+
+__extension__ static __inline void __attribute__ ((__always_inline__))
 vst3q_f32 (float32_t * __a, float32x4x3_t val)
 {
   __builtin_aarch64_simd_ci __o;
@@ -23075,6 +23326,22 @@ vst4_u32 (uint32_t * __a, uint32x2x4_t val)
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
+vst4_f16 (float16_t * __a, float16x4x4_t val)
+{
+  __builtin_aarch64_simd_xi __o;
+  float16x8x4_t temp;
+  temp.val[0] = vcombine_f16 (val.val[0], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  temp.val[1] = vcombine_f16 (val.val[1], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  temp.val[2] = vcombine_f16 (val.val[2], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  temp.val[3] = vcombine_f16 (val.val[3], vcreate_f16 (__AARCH64_UINT64_C (0)));
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) temp.val[0], 0);
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) temp.val[1], 1);
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) temp.val[2], 2);
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) temp.val[3], 3);
+  __builtin_aarch64_st4v4hf ((__builtin_aarch64_simd_hf *) __a, __o);
+}
+
+__extension__ static __inline void __attribute__ ((__always_inline__))
 vst4_f32 (float32_t * __a, float32x2x4_t val)
 {
   __builtin_aarch64_simd_xi __o;
@@ -23201,6 +23468,17 @@ vst4q_u64 (uint64_t * __a, uint64x2x4_t val)
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
+vst4q_f16 (float16_t * __a, float16x8x4_t val)
+{
+  __builtin_aarch64_simd_xi __o;
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) val.val[0], 0);
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) val.val[1], 1);
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) val.val[2], 2);
+  __o = __builtin_aarch64_set_qregxiv8hf (__o, (float16x8_t) val.val[3], 3);
+  __builtin_aarch64_st4v8hf ((__builtin_aarch64_simd_hf *) __a, __o);
+}
+
+__extension__ static __inline void __attribute__ ((__always_inline__))
 vst4q_f32 (float32_t * __a, float32x4x4_t val)
 {
   __builtin_aarch64_simd_xi __o;
diff --git a/gcc/config/aarch64/iterators.md b/gcc/config/aarch64/iterators.md
index 3e4f07d..58cc000 100644
--- a/gcc/config/aarch64/iterators.md
+++ b/gcc/config/aarch64/iterators.md
@@ -111,9 +111,9 @@
 (define_mode_iterator VALLDI_F16 [V8QI V16QI V4HI V8HI V2SI V4SI V2DI
 				  V4HF V8HF V2SF V4SF V2DF DI])
 
-;; All vector modes barring HF modes, plus DI and DF.
+;; All vector modes, plus DI and DF.
 (define_mode_iterator VALLDIF [V8QI V16QI V4HI V8HI V2SI V4SI
-			       V2DI V2SF V4SF V2DF DI DF])
+			       V2DI V4HF V8HF V2SF V4SF V2DF DI DF])
 
 ;; Vector modes for Integer reduction across lanes.
 (define_mode_iterator VDQV [V8QI V16QI V4HI V8HI V4SI V2DI])
@@ -134,7 +134,7 @@
 (define_mode_iterator VQW [V16QI V8HI V4SI])
 
 ;; Double vector modes for combines.
-(define_mode_iterator VDC [V8QI V4HI V2SI V2SF DI DF])
+(define_mode_iterator VDC [V8QI V4HI V4HF V2SI V2SF DI DF])
 
 ;; Vector modes except double int.
 (define_mode_iterator VDQIF [V8QI V16QI V4HI V8HI V2SI V4SI V2SF V4SF V2DF])
@@ -361,7 +361,8 @@
                          (V2SI "2s") (V4SI  "4s")
                          (DI   "1d") (DF    "1d")
                          (V2DI "2d") (V2SF "2s")
-			 (V4SF "4s") (V2DF "2d")])
+			 (V4SF "4s") (V2DF "2d")
+			 (V4HF "4h") (V8HF "8h")])
 
 (define_mode_attr Vrevsuff [(V4HI "16") (V8HI "16") (V2SI "32")
                             (V4SI "32") (V2DI "64")])
@@ -387,7 +388,8 @@
 (define_mode_attr Vetype [(V8QI "b") (V16QI "b")
 			  (V4HI "h") (V8HI  "h")
                           (V2SI "s") (V4SI  "s")
-			  (V2DI "d") (V2SF  "s")
+			  (V2DI "d") (V4HF "h")
+			  (V8HF "h") (V2SF  "s")
 			  (V4SF "s") (V2DF  "d")
 			  (SF   "s") (DF  "d")
 			  (QI "b")   (HI "h")
@@ -397,7 +399,8 @@
 (define_mode_attr Vbtype [(V8QI "8b")  (V16QI "16b")
 			  (V4HI "8b") (V8HI  "16b")
 			  (V2SI "8b") (V4SI  "16b")
-			  (V2DI "16b") (V2SF  "8b")
+			  (V2DI "16b") (V4HF "8b")
+			  (V8HF "16b") (V2SF  "8b")
 			  (V4SF "16b") (V2DF  "16b")
 			  (DI   "8b")  (DF    "8b")
 			  (SI   "8b")])
@@ -448,6 +451,7 @@
 
 ;; Double modes of vector modes (lower case).
 (define_mode_attr Vdbl [(V8QI "v16qi") (V4HI "v8hi")
+			(V4HF "v8hf")
 			(V2SI "v4si")  (V2SF "v4sf")
 			(SI   "v2si")  (DI   "v2di")
 			(DF   "v2df")])
@@ -522,6 +526,7 @@
 				(V4HI "V4HI") (V8HI  "V8HI")
 				(V2SI "V2SI") (V4SI  "V4SI")
 				(DI   "DI")   (V2DI  "V2DI")
+				(V4HF "V4HI") (V8HF  "V8HI")
 				(V2SF "V2SI") (V4SF  "V4SI")
 				(V2DF "V2DI") (DF    "DI")
 				(SF   "SI")])
@@ -531,6 +536,7 @@
 				(V4HI "v4hi") (V8HI  "v8hi")
 				(V2SI "v2si") (V4SI  "v4si")
 				(DI   "di")   (V2DI  "v2di")
+				(V4HF "v4hi") (V8HF  "v8hi")
 				(V2SF "v2si") (V4SF  "v4si")
 				(V2DF "v2di") (DF    "di")
 				(SF   "si")])