PR/60825 Make float64x1_t in arm_neon.h a proper vector type

gcc/ChangeLog:
	PR target/60825
	* config/aarch64/aarch64.c (aarch64_simd_mangle_map): Add entry for
	V1DFmode.
	* config/aarch64/aarch64-builtins.c (aarch64_simd_builtin_type_mode):
	add V1DFmode
	(BUILTIN_VD1): New.
	(BUILTIN_VD_RE): Remove.
	(aarch64_init_simd_builtins): Add V1DF to modes/modenames.
	(aarch64_fold_builtin): Update reinterpret patterns, df becomes v1df.
	* config/aarch64/aarch64-simd-builtins.def (create): Make a v1df
	variant but not df.
	(vreinterpretv1df*, vreinterpret*v1df): New.
	(vreinterpretdf*, vreinterpret*df): Remove.
	* config/aarch64/aarch64-simd.md (aarch64_create, aarch64_reinterpret*):
	Generate V1DFmode pattern not DFmode.
	* config/aarch64/iterators.md (VD_RE): Include V1DF, remove DF.
	(VD1): New.
	* config/aarch64/arm_neon.h (float64x1_t): typedef with gcc extensions.
	(vcreate_f64): Remove cast, use v1df builtin.
	(vcombine_f64): Remove cast, get elements with gcc vector extensions.
	(vget_low_f64, vabs_f64, vceq_f64, vceqz_f64, vcge_f64, vgfez_f64,
	vcgt_f64, vcgtz_f64, vcle_f64, vclez_f64, vclt_f64, vcltz_f64,
	vdup_n_f64, vdupq_lane_f64, vld1_f64, vld2_f64, vld3_f64, vld4_f64,
	vmov_n_f64, vst1_f64): Use gcc vector extensions.
	(vget_lane_f64, vdupd_lane_f64, vmulq_lane_f64, ): Use gcc extensions,
	add range check using __builtin_aarch64_im_lane_boundsi.
	(vfma_lane_f64, vfmad_lane_f64, vfma_laneq_f64, vfmaq_lane_f64,
	vfms_lane_f64, vfmsd_lane_f64, vfms_laneq_f64, vfmsq_lane_f64): Fix
	type signature, use gcc vector extensions.
	(vreinterpret_p8_f64, vreinterpret_p16_f64, vreinterpret_f32_f64,
	vreinterpret_f64_f32, vreinterpret_f64_p8, vreinterpret_f64_p16,
	vreinterpret_f64_s8, vreinterpret_f64_s16, vreinterpret_f64_s32,
	vreinterpret_f64_s64, vreinterpret_f64_u8, vreinterpret_f64_u16,
	vreinterpret_f64_u32, vreinterpret_f64_u64, vreinterpret_s8_f64,
	vreinterpret_s16_f64, vreinterpret_s32_f64, vreinterpret_s64_f64,
	vreinterpret_u8_f64, vreinterpret_u16_f64, vreinterpret_u32_f64,
	vreinterpret_u64_f64): Use v1df builtin not df.

gcc/testsuite/ChangeLog:
	* g++.dg/abi/mangle-neon-aarch64.C: Also test mangling of float64x1_t.
	* gcc.target/aarch64/aapcs/test_64x1_1.c: New test.
	* gcc.target/aarch64/aapcs/func-ret-64x1_1.c: New test.
	* gcc.target/aarch64/simd/ext_f64_1.c (main): Compare vector elements.
	* gcc.target/aarch64/vadd_f64.c: Rewrite with macro to use vector types.
	* gcc.target/aarch64/vsub_f64.c: Likewise.
	* gcc.target/aarch64/vdiv_f.c (INDEX*, RUN_TEST): Remove indexing scheme
	as now the same for all variants.
	* gcc.target/aarch64/vrnd_f64_1.c (compare_f64): Return float64_t not
	float64x1_t.


git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@211892 138bc75d-0d04-0410-961f-82ee72b054a4

6 files changed, 146 insertions, 130 deletions

diff --git a/gcc/config/aarch64/aarch64-builtins.c b/gcc/config/aarch64/aarch64-builtins.c
index a94ef52f71a..eebb7d3978d 100644
--- a/gcc/config/aarch64/aarch64-builtins.c
+++ b/gcc/config/aarch64/aarch64-builtins.c
@@ -53,6 +53,7 @@ enum aarch64_simd_builtin_type_mode
   T_V4HI,
   T_V2SI,
   T_V2SF,
+  T_V1DF,
   T_DI,
   T_DF,
   T_V16QI,
@@ -76,6 +77,7 @@ enum aarch64_simd_builtin_type_mode
 #define v4hi_UP  T_V4HI
 #define v2si_UP  T_V2SI
 #define v2sf_UP  T_V2SF
+#define v1df_UP  T_V1DF
 #define di_UP    T_DI
 #define df_UP    T_DF
 #define v16qi_UP T_V16QI
@@ -346,6 +348,8 @@ aarch64_types_storestruct_lane_qualifiers[SIMD_MAX_BUILTIN_ARGS]
   VAR2 (T, N, MAP, v8qi, v16qi)
 #define BUILTIN_VD(T, N, MAP) \
   VAR4 (T, N, MAP, v8qi, v4hi, v2si, v2sf)
+#define BUILTIN_VD1(T, N, MAP) \
+  VAR5 (T, N, MAP, v8qi, v4hi, v2si, v2sf, v1df)
 #define BUILTIN_VDC(T, N, MAP) \
   VAR6 (T, N, MAP, v8qi, v4hi, v2si, v2sf, di, df)
 #define BUILTIN_VDIC(T, N, MAP) \
@@ -380,8 +384,6 @@ aarch64_types_storestruct_lane_qualifiers[SIMD_MAX_BUILTIN_ARGS]
   VAR3 (T, N, MAP, v8qi, v4hi, v2si)
 #define BUILTIN_VD_HSI(T, N, MAP) \
   VAR2 (T, N, MAP, v4hi, v2si)
-#define BUILTIN_VD_RE(T, N, MAP) \
-  VAR6 (T, N, MAP, v8qi, v4hi, v2si, v2sf, di, df)
 #define BUILTIN_VQ(T, N, MAP) \
   VAR6 (T, N, MAP, v16qi, v8hi, v4si, v2di, v4sf, v2df)
 #define BUILTIN_VQN(T, N, MAP) \
@@ -729,13 +731,13 @@ aarch64_init_simd_builtins (void)
       aarch64_simd_builtin_datum *d = &aarch64_simd_builtin_data[i];
       const char *const modenames[] =
 	{
-	  "v8qi", "v4hi", "v2si", "v2sf", "di", "df",
+	  "v8qi", "v4hi", "v2si", "v2sf", "v1df", "di", "df",
 	  "v16qi", "v8hi", "v4si", "v4sf", "v2di", "v2df",
 	  "ti", "ei", "oi", "xi", "si", "sf", "hi", "qi"
 	};
       const enum machine_mode modes[] =
 	{
-	  V8QImode, V4HImode, V2SImode, V2SFmode, DImode, DFmode,
+	  V8QImode, V4HImode, V2SImode, V2SFmode, V1DFmode, DImode, DFmode,
 	  V16QImode, V8HImode, V4SImode, V4SFmode, V2DImode,
 	  V2DFmode, TImode, EImode, OImode, XImode, SImode,
 	  SFmode, HImode, QImode
@@ -1342,24 +1344,23 @@ aarch64_fold_builtin (tree fndecl, int n_args ATTRIBUTE_UNUSED, tree *args,
 	  return fold_build2 (NE_EXPR, type, and_node, vec_zero_node);
 	  break;
 	}
-      VAR1 (REINTERP_SS, reinterpretdi, 0, df)
-      VAR1 (REINTERP_SS, reinterpretv8qi, 0, df)
-      VAR1 (REINTERP_SS, reinterpretv4hi, 0, df)
-      VAR1 (REINTERP_SS, reinterpretv2si, 0, df)
-      VAR1 (REINTERP_SS, reinterpretv2sf, 0, df)
-      BUILTIN_VD (REINTERP_SS, reinterpretdf, 0)
-      BUILTIN_VD (REINTERP_SU, reinterpretdf, 0)
-      VAR1 (REINTERP_US, reinterpretdi, 0, df)
-      VAR1 (REINTERP_US, reinterpretv8qi, 0, df)
-      VAR1 (REINTERP_US, reinterpretv4hi, 0, df)
-      VAR1 (REINTERP_US, reinterpretv2si, 0, df)
-      VAR1 (REINTERP_US, reinterpretv2sf, 0, df)
-      BUILTIN_VD (REINTERP_SP, reinterpretdf, 0)
-      VAR1 (REINTERP_PS, reinterpretdi, 0, df)
-      VAR1 (REINTERP_PS, reinterpretv8qi, 0, df)
-      VAR1 (REINTERP_PS, reinterpretv4hi, 0, df)
-      VAR1 (REINTERP_PS, reinterpretv2si, 0, df)
-      VAR1 (REINTERP_PS, reinterpretv2sf, 0, df)
+      VAR1 (REINTERP_SS, reinterpretdi, 0, v1df)
+      VAR1 (REINTERP_SS, reinterpretv8qi, 0, v1df)
+      VAR1 (REINTERP_SS, reinterpretv4hi, 0, v1df)
+      VAR1 (REINTERP_SS, reinterpretv2si, 0, v1df)
+      VAR1 (REINTERP_SS, reinterpretv2sf, 0, v1df)
+      BUILTIN_VD (REINTERP_SS, reinterpretv1df, 0)
+      BUILTIN_VD (REINTERP_SU, reinterpretv1df, 0)
+      VAR1 (REINTERP_US, reinterpretdi, 0, v1df)
+      VAR1 (REINTERP_US, reinterpretv8qi, 0, v1df)
+      VAR1 (REINTERP_US, reinterpretv4hi, 0, v1df)
+      VAR1 (REINTERP_US, reinterpretv2si, 0, v1df)
+      VAR1 (REINTERP_US, reinterpretv2sf, 0, v1df)
+      BUILTIN_VD (REINTERP_SP, reinterpretv1df, 0)
+      VAR1 (REINTERP_PS, reinterpretdi, 0, v1df)
+      VAR1 (REINTERP_PS, reinterpretv8qi, 0, v1df)
+      VAR1 (REINTERP_PS, reinterpretv4hi, 0, v1df)
+      VAR1 (REINTERP_PS, reinterpretv2sf, 0, v1df)
 	return fold_build1 (VIEW_CONVERT_EXPR, type, args[0]);
       VAR1 (UNOP, floatv2si, 2, v2sf)
       VAR1 (UNOP, floatv4si, 2, v4sf)
@@ -1539,6 +1540,7 @@ aarch64_atomic_assign_expand_fenv (tree *hold, tree *clear, tree *update)
 #undef BUILTIN_VALL
 #undef BUILTIN_VB
 #undef BUILTIN_VD
+#undef BUILTIN_VD1
 #undef BUILTIN_VDC
 #undef BUILTIN_VDIC
 #undef BUILTIN_VDN
@@ -1554,7 +1556,6 @@ aarch64_atomic_assign_expand_fenv (tree *hold, tree *clear, tree *update)
 #undef BUILTIN_VDW
 #undef BUILTIN_VD_BHSI
 #undef BUILTIN_VD_HSI
-#undef BUILTIN_VD_RE
 #undef BUILTIN_VQ
 #undef BUILTIN_VQN
 #undef BUILTIN_VQW
diff --git a/gcc/config/aarch64/aarch64-simd-builtins.def b/gcc/config/aarch64/aarch64-simd-builtins.def
index faa0858e3be..1b931bede94 100644
--- a/gcc/config/aarch64/aarch64-simd-builtins.def
+++ b/gcc/config/aarch64/aarch64-simd-builtins.def
@@ -39,7 +39,7 @@
    1-9 - CODE_FOR_<name><mode><1-9>
    10 - CODE_FOR_<name><mode>.  */
 
-  BUILTIN_VD_RE (CREATE, create, 0)
+  BUILTIN_VD1 (CREATE, create, 0)
   BUILTIN_VDC (COMBINE, combine, 0)
   BUILTIN_VB (BINOP, pmul, 0)
   BUILTIN_VDQF (UNOP, sqrt, 2)
@@ -51,28 +51,28 @@
   VAR1 (GETLANE, get_lane, 0, di)
   BUILTIN_VALL (GETLANE, be_checked_get_lane, 0)
 
-  VAR1 (REINTERP_SS, reinterpretdi, 0, df)
-  VAR1 (REINTERP_SS, reinterpretv8qi, 0, df)
-  VAR1 (REINTERP_SS, reinterpretv4hi, 0, df)
-  VAR1 (REINTERP_SS, reinterpretv2si, 0, df)
-  VAR1 (REINTERP_SS, reinterpretv2sf, 0, df)
-  BUILTIN_VD (REINTERP_SS, reinterpretdf, 0)
+  VAR1 (REINTERP_SS, reinterpretdi, 0, v1df)
+  VAR1 (REINTERP_SS, reinterpretv8qi, 0, v1df)
+  VAR1 (REINTERP_SS, reinterpretv4hi, 0, v1df)
+  VAR1 (REINTERP_SS, reinterpretv2si, 0, v1df)
+  VAR1 (REINTERP_SS, reinterpretv2sf, 0, v1df)
+  BUILTIN_VD (REINTERP_SS, reinterpretv1df, 0)
 
-  BUILTIN_VD (REINTERP_SU, reinterpretdf, 0)
+  BUILTIN_VD (REINTERP_SU, reinterpretv1df, 0)
 
-  VAR1 (REINTERP_US, reinterpretdi, 0, df)
-  VAR1 (REINTERP_US, reinterpretv8qi, 0, df)
-  VAR1 (REINTERP_US, reinterpretv4hi, 0, df)
-  VAR1 (REINTERP_US, reinterpretv2si, 0, df)
-  VAR1 (REINTERP_US, reinterpretv2sf, 0, df)
+  VAR1 (REINTERP_US, reinterpretdi, 0, v1df)
+  VAR1 (REINTERP_US, reinterpretv8qi, 0, v1df)
+  VAR1 (REINTERP_US, reinterpretv4hi, 0, v1df)
+  VAR1 (REINTERP_US, reinterpretv2si, 0, v1df)
+  VAR1 (REINTERP_US, reinterpretv2sf, 0, v1df)
 
-  BUILTIN_VD (REINTERP_SP, reinterpretdf, 0)
+  BUILTIN_VD (REINTERP_SP, reinterpretv1df, 0)
 
-  VAR1 (REINTERP_PS, reinterpretdi, 0, df)
-  VAR1 (REINTERP_PS, reinterpretv8qi, 0, df)
-  VAR1 (REINTERP_PS, reinterpretv4hi, 0, df)
-  VAR1 (REINTERP_PS, reinterpretv2si, 0, df)
-  VAR1 (REINTERP_PS, reinterpretv2sf, 0, df)
+  VAR1 (REINTERP_PS, reinterpretdi, 0, v1df)
+  VAR1 (REINTERP_PS, reinterpretv8qi, 0, v1df)
+  VAR1 (REINTERP_PS, reinterpretv4hi, 0, v1df)
+  VAR1 (REINTERP_PS, reinterpretv2si, 0, v1df)
+  VAR1 (REINTERP_PS, reinterpretv2sf, 0, v1df)
 
   BUILTIN_VDQ_I (BINOP, dup_lane, 0)
   /* Implemented by aarch64_<sur>q<r>shl<mode>.  */
diff --git a/gcc/config/aarch64/aarch64-simd.md b/gcc/config/aarch64/aarch64-simd.md
index 7fd70941269..1c32f0c4efa 100644
--- a/gcc/config/aarch64/aarch64-simd.md
+++ b/gcc/config/aarch64/aarch64-simd.md
@@ -2134,7 +2134,7 @@
 ;; Patterns for AArch64 SIMD Intrinsics.
 
 (define_expand "aarch64_create<mode>"
-  [(match_operand:VD_RE 0 "register_operand" "")
+  [(match_operand:VD1 0 "register_operand" "")
    (match_operand:DI 1 "general_operand" "")]
   "TARGET_SIMD"
 {
@@ -2224,7 +2224,7 @@
 
 (define_expand "aarch64_reinterpretv8qi<mode>"
   [(match_operand:V8QI 0 "register_operand" "")
-   (match_operand:VDC 1 "register_operand" "")]
+   (match_operand:VD_RE 1 "register_operand" "")]
   "TARGET_SIMD"
 {
   aarch64_simd_reinterpret (operands[0], operands[1]);
@@ -2233,7 +2233,7 @@
 
 (define_expand "aarch64_reinterpretv4hi<mode>"
   [(match_operand:V4HI 0 "register_operand" "")
-   (match_operand:VDC 1 "register_operand" "")]
+   (match_operand:VD_RE 1 "register_operand" "")]
   "TARGET_SIMD"
 {
   aarch64_simd_reinterpret (operands[0], operands[1]);
@@ -2242,7 +2242,7 @@
 
 (define_expand "aarch64_reinterpretv2si<mode>"
   [(match_operand:V2SI 0 "register_operand" "")
-   (match_operand:VDC 1 "register_operand" "")]
+   (match_operand:VD_RE 1 "register_operand" "")]
   "TARGET_SIMD"
 {
   aarch64_simd_reinterpret (operands[0], operands[1]);
@@ -2251,7 +2251,7 @@
 
 (define_expand "aarch64_reinterpretv2sf<mode>"
   [(match_operand:V2SF 0 "register_operand" "")
-   (match_operand:VDC 1 "register_operand" "")]
+   (match_operand:VD_RE 1 "register_operand" "")]
   "TARGET_SIMD"
 {
   aarch64_simd_reinterpret (operands[0], operands[1]);
@@ -2267,8 +2267,8 @@
   DONE;
 })
 
-(define_expand "aarch64_reinterpretdf<mode>"
-  [(match_operand:DF 0 "register_operand" "")
+(define_expand "aarch64_reinterpretv1df<mode>"
+  [(match_operand:V1DF 0 "register_operand" "")
    (match_operand:VD_RE 1 "register_operand" "")]
   "TARGET_SIMD"
 {
diff --git a/gcc/config/aarch64/aarch64.c b/gcc/config/aarch64/aarch64.c
index f0aafbd5b55..1d631f8f6bc 100644
--- a/gcc/config/aarch64/aarch64.c
+++ b/gcc/config/aarch64/aarch64.c
@@ -7358,6 +7358,7 @@ static aarch64_simd_mangle_map_entry aarch64_simd_mangle_map[] = {
   { V2SImode,  "__builtin_aarch64_simd_si",     "11__Int32x2_t" },
   { V2SImode,  "__builtin_aarch64_simd_usi",    "12__Uint32x2_t" },
   { V2SFmode,  "__builtin_aarch64_simd_sf",     "13__Float32x2_t" },
+  { V1DFmode,  "__builtin_aarch64_simd_df",	"13__Float64x1_t" },
   { V8QImode,  "__builtin_aarch64_simd_poly8",  "11__Poly8x8_t" },
   { V4HImode,  "__builtin_aarch64_simd_poly16", "12__Poly16x4_t" },
   /* 128-bit containerized types.  */
diff --git a/gcc/config/aarch64/arm_neon.h b/gcc/config/aarch64/arm_neon.h
index 3ed8a9851b5..701e8c14834 100644
--- a/gcc/config/aarch64/arm_neon.h
+++ b/gcc/config/aarch64/arm_neon.h
@@ -42,7 +42,8 @@ typedef int64_t int64x1_t;
 typedef int32_t int32x1_t;
 typedef int16_t int16x1_t;
 typedef int8_t int8x1_t;
-typedef double float64x1_t;
+typedef __builtin_aarch64_simd_df float64x1_t
+  __attribute__ ((__vector_size__ (8)));
 typedef __builtin_aarch64_simd_sf float32x2_t
   __attribute__ ((__vector_size__ (8)));
 typedef __builtin_aarch64_simd_poly8 poly8x8_t
@@ -461,7 +462,11 @@ typedef struct poly16x8x4_t
 
 #define __aarch64_vget_lane_f32(__a, __b) \
   __aarch64_vget_lane_any (v2sf, , , __a, __b)
-#define __aarch64_vget_lane_f64(__a, __b) (__a)
+#define __aarch64_vget_lane_f64(__a, __b) __extension__	\
+  ({							\
+    __builtin_aarch64_im_lane_boundsi (__b, 1);		\
+    __a[0];						\
+  })
 
 #define __aarch64_vget_lane_p8(__a, __b) \
   __aarch64_vget_lane_any (v8qi, (poly8_t), (int8x8_t), __a, __b)
@@ -518,7 +523,8 @@ typedef struct poly16x8x4_t
 
 #define __aarch64_vdup_lane_f32(__a, __b) \
    __aarch64_vdup_lane_any (f32, , , __a, __b)
-#define __aarch64_vdup_lane_f64(__a, __b) (__a)
+#define __aarch64_vdup_lane_f64(__a, __b) \
+   __aarch64_vdup_lane_any (f64, , , __a, __b)
 #define __aarch64_vdup_lane_p8(__a, __b) \
    __aarch64_vdup_lane_any (p8, , , __a, __b)
 #define __aarch64_vdup_lane_p16(__a, __b) \
@@ -567,7 +573,8 @@ typedef struct poly16x8x4_t
 /* __aarch64_vdupq_lane internal macros.  */
 #define __aarch64_vdupq_lane_f32(__a, __b) \
    __aarch64_vdup_lane_any (f32, q, , __a, __b)
-#define __aarch64_vdupq_lane_f64(__a, __b) (vdupq_n_f64 (__a))
+#define __aarch64_vdupq_lane_f64(__a, __b) \
+   __aarch64_vdup_lane_any (f64, q, , __a, __b)
 #define __aarch64_vdupq_lane_p8(__a, __b) \
    __aarch64_vdup_lane_any (p8, q, , __a, __b)
 #define __aarch64_vdupq_lane_p16(__a, __b) \
@@ -2475,7 +2482,7 @@ vcreate_u64 (uint64_t __a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vcreate_f64 (uint64_t __a)
 {
-  return (float64x1_t) __builtin_aarch64_createdf (__a);
+  return __builtin_aarch64_createv1df (__a);
 }
 
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
@@ -2643,7 +2650,7 @@ vgetq_lane_u64 (uint64x2_t __a, const int __b)
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
 vreinterpret_p8_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv8qidf_ps (__a);
+  return __builtin_aarch64_reinterpretv8qiv1df_ps (__a);
 }
 
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
@@ -2775,7 +2782,7 @@ vreinterpretq_p8_p16 (poly16x8_t __a)
 __extension__ static __inline poly16x4_t __attribute__ ((__always_inline__))
 vreinterpret_p16_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv4hidf_ps (__a);
+  return __builtin_aarch64_reinterpretv4hiv1df_ps (__a);
 }
 
 __extension__ static __inline poly16x4_t __attribute__ ((__always_inline__))
@@ -2907,7 +2914,7 @@ vreinterpretq_p16_p8 (poly8x16_t __a)
 __extension__ static __inline float32x2_t __attribute__ ((__always_inline__))
 vreinterpret_f32_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv2sfdf (__a);
+  return __builtin_aarch64_reinterpretv2sfv1df (__a);
 }
 
 __extension__ static __inline float32x2_t __attribute__ ((__always_inline__))
@@ -3039,67 +3046,67 @@ vreinterpretq_f32_p16 (poly16x8_t __a)
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_f32 (float32x2_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv2sf (__a);
+  return __builtin_aarch64_reinterpretv1dfv2sf (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_p8 (poly8x8_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv8qi_sp (__a);
+  return __builtin_aarch64_reinterpretv1dfv8qi_sp (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_p16 (poly16x4_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv4hi_sp (__a);
+  return __builtin_aarch64_reinterpretv1dfv4hi_sp (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_s8 (int8x8_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv8qi (__a);
+  return __builtin_aarch64_reinterpretv1dfv8qi (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_s16 (int16x4_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv4hi (__a);
+  return __builtin_aarch64_reinterpretv1dfv4hi (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_s32 (int32x2_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv2si (__a);
+  return __builtin_aarch64_reinterpretv1dfv2si (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_s64 (int64x1_t __a)
 {
-  return __builtin_aarch64_createdf ((uint64_t) vget_lane_s64 (__a, 0));
+  return __builtin_aarch64_createv1df ((uint64_t) vget_lane_s64 (__a, 0));
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_u8 (uint8x8_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv8qi_su (__a);
+  return __builtin_aarch64_reinterpretv1dfv8qi_su (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_u16 (uint16x4_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv4hi_su (__a);
+  return __builtin_aarch64_reinterpretv1dfv4hi_su (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_u32 (uint32x2_t __a)
 {
-  return __builtin_aarch64_reinterpretdfv2si_su (__a);
+  return __builtin_aarch64_reinterpretv1dfv2si_su (__a);
 }
 
 __extension__ static __inline float64x1_t __attribute__((__always_inline__))
 vreinterpret_f64_u64 (uint64x1_t __a)
 {
-  return __builtin_aarch64_createdf (vget_lane_u64 (__a, 0));
+  return __builtin_aarch64_createv1df (vget_lane_u64 (__a, 0));
 }
 
 __extension__ static __inline float64x2_t __attribute__((__always_inline__))
@@ -3171,7 +3178,7 @@ vreinterpretq_f64_u64 (uint64x2_t __a)
 __extension__ static __inline int64x1_t __attribute__ ((__always_inline__))
 vreinterpret_s64_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretdidf (__a);
+  return __builtin_aarch64_reinterpretdiv1df (__a);
 }
 
 __extension__ static __inline int64x1_t __attribute__ ((__always_inline__))
@@ -3303,7 +3310,7 @@ vreinterpretq_s64_p16 (poly16x8_t __a)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vreinterpret_u64_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretdidf_us (__a);
+  return __builtin_aarch64_reinterpretdiv1df_us (__a);
 }
 
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
@@ -3435,7 +3442,7 @@ vreinterpretq_u64_p16 (poly16x8_t __a)
 __extension__ static __inline int8x8_t __attribute__ ((__always_inline__))
 vreinterpret_s8_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv8qidf (__a);
+  return __builtin_aarch64_reinterpretv8qiv1df (__a);
 }
 
 __extension__ static __inline int8x8_t __attribute__ ((__always_inline__))
@@ -3567,7 +3574,7 @@ vreinterpretq_s8_p16 (poly16x8_t __a)
 __extension__ static __inline int16x4_t __attribute__ ((__always_inline__))
 vreinterpret_s16_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv4hidf (__a);
+  return __builtin_aarch64_reinterpretv4hiv1df (__a);
 }
 
 __extension__ static __inline int16x4_t __attribute__ ((__always_inline__))
@@ -3699,7 +3706,7 @@ vreinterpretq_s16_p16 (poly16x8_t __a)
 __extension__ static __inline int32x2_t __attribute__ ((__always_inline__))
 vreinterpret_s32_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv2sidf (__a);
+  return __builtin_aarch64_reinterpretv2siv1df (__a);
 }
 
 __extension__ static __inline int32x2_t __attribute__ ((__always_inline__))
@@ -3831,7 +3838,7 @@ vreinterpretq_s32_p16 (poly16x8_t __a)
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
 vreinterpret_u8_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv8qidf_us (__a);
+  return __builtin_aarch64_reinterpretv8qiv1df_us (__a);
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -3963,7 +3970,7 @@ vreinterpretq_u8_p16 (poly16x8_t __a)
 __extension__ static __inline uint16x4_t __attribute__ ((__always_inline__))
 vreinterpret_u16_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv4hidf_us (__a);
+  return __builtin_aarch64_reinterpretv4hiv1df_us (__a);
 }
 
 __extension__ static __inline uint16x4_t __attribute__ ((__always_inline__))
@@ -4095,7 +4102,7 @@ vreinterpretq_u16_p16 (poly16x8_t __a)
 __extension__ static __inline uint32x2_t __attribute__ ((__always_inline__))
 vreinterpret_u32_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_reinterpretv2sidf_us (__a);
+  return __builtin_aarch64_reinterpretv2siv1df_us (__a);
 }
 
 __extension__ static __inline uint32x2_t __attribute__ ((__always_inline__))
@@ -4238,7 +4245,7 @@ vget_low_f32 (float32x4_t __a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vget_low_f64 (float64x2_t __a)
 {
-  return vgetq_lane_f64 (__a, 0);
+  return (float64x1_t) {vgetq_lane_f64 (__a, 0)};
 }
 
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
@@ -4364,7 +4371,7 @@ vcombine_u64 (uint64x1_t __a, uint64x1_t __b)
 __extension__ static __inline float64x2_t __attribute__ ((__always_inline__))
 vcombine_f64 (float64x1_t __a, float64x1_t __b)
 {
-  return (float64x2_t) __builtin_aarch64_combinedf (__a, __b);
+  return __builtin_aarch64_combinedf (__a[0], __b[0]);
 }
 
 __extension__ static __inline poly8x16_t __attribute__ ((__always_inline__))
@@ -13432,7 +13439,7 @@ vabs_f32 (float32x2_t __a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vabs_f64 (float64x1_t __a)
 {
-  return __builtin_fabs (__a);
+  return (float64x1_t) {__builtin_fabs (__a[0])};
 }
 
 __extension__ static __inline int8x8_t __attribute__ ((__always_inline__))
@@ -13916,7 +13923,7 @@ vceq_f32 (float32x2_t __a, float32x2_t __b)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vceq_f64 (float64x1_t __a, float64x1_t __b)
 {
-  return __a == __b ? -1ll : 0ll;
+  return (uint64x1_t) (__a == __b);
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14086,7 +14093,7 @@ vceqz_f32 (float32x2_t __a)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vceqz_f64 (float64x1_t __a)
 {
-  return __a == 0.0 ? -1ll : 0ll;
+  return (uint64x1_t) (__a == (float64x1_t) {0.0});
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14276,7 +14283,7 @@ vcge_f32 (float32x2_t __a, float32x2_t __b)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vcge_f64 (float64x1_t __a, float64x1_t __b)
 {
-  return __a >= __b ? -1ll : 0ll;
+  return (uint64x1_t) (__a >= __b);
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14432,7 +14439,7 @@ vcgez_f32 (float32x2_t __a)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vcgez_f64 (float64x1_t __a)
 {
-  return __a >= 0.0 ? -1ll : 0ll;
+  return (uint64x1_t) (__a[0] >= (float64x1_t) {0.0});
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14536,7 +14543,7 @@ vcgt_f32 (float32x2_t __a, float32x2_t __b)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vcgt_f64 (float64x1_t __a, float64x1_t __b)
 {
-  return __a > __b ? -1ll : 0ll;
+  return (uint64x1_t) (__a > __b);
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14692,7 +14699,7 @@ vcgtz_f32 (float32x2_t __a)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vcgtz_f64 (float64x1_t __a)
 {
-  return __a > 0.0 ? -1ll : 0ll;
+  return (uint64x1_t) (__a > (float64x1_t) {0.0});
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14796,7 +14803,7 @@ vcle_f32 (float32x2_t __a, float32x2_t __b)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vcle_f64 (float64x1_t __a, float64x1_t __b)
 {
-  return __a <= __b ? -1ll : 0ll;
+  return (uint64x1_t) (__a <= __b);
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -14952,7 +14959,7 @@ vclez_f32 (float32x2_t __a)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vclez_f64 (float64x1_t __a)
 {
-  return __a <= 0.0 ? -1ll : 0ll;
+  return (uint64x1_t) (__a <= (float64x1_t) {0.0});
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -15056,7 +15063,7 @@ vclt_f32 (float32x2_t __a, float32x2_t __b)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vclt_f64 (float64x1_t __a, float64x1_t __b)
 {
-  return __a < __b ? -1ll : 0ll;
+  return (uint64x1_t) (__a < __b);
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -15212,7 +15219,7 @@ vcltz_f32 (float32x2_t __a)
 __extension__ static __inline uint64x1_t __attribute__ ((__always_inline__))
 vcltz_f64 (float64x1_t __a)
 {
-  return __a < 0.0 ? -1ll : 0ll;
+  return (uint64x1_t) (__a < (float64x1_t) {0.0});
 }
 
 __extension__ static __inline uint8x8_t __attribute__ ((__always_inline__))
@@ -15821,7 +15828,7 @@ vdup_n_f32 (float32_t __a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vdup_n_f64 (float64_t __a)
 {
-  return __a;
+  return (float64x1_t) {__a};
 }
 
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
@@ -16314,9 +16321,10 @@ vdups_lane_u32 (uint32x2_t __a, const int __b)
 
 /* vdupd_lane  */
 __extension__ static __inline float64_t __attribute__ ((__always_inline__))
-vdupd_lane_f64 (float64x1_t __a, const int __attribute__ ((unused)) __b)
+vdupd_lane_f64 (float64x1_t __a, const int __b)
 {
-  return __a;
+  __builtin_aarch64_im_lane_boundsi (__b, 1);
+  return __a[0];
 }
 
 __extension__ static __inline int64_t __attribute__ ((__always_inline__))
@@ -16704,18 +16712,18 @@ vfma_lane_f32 (float32x2_t __a, float32x2_t __b,
 				    __a);
 }
 
-__extension__ static __inline float64_t __attribute__ ((__always_inline__))
-vfma_lane_f64 (float64_t __a, float64_t __b,
-	       float64_t __c, const int __lane)
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vfma_lane_f64 (float64x1_t __a, float64x1_t __b,
+	       float64x1_t __c, const int __lane)
 {
-  return __builtin_fma (__b, __c, __a);
+  return (float64x1_t) {__builtin_fma (__b[0], __c[0], __a[0])};
 }
 
 __extension__ static __inline float64_t __attribute__ ((__always_inline__))
 vfmad_lane_f64 (float64_t __a, float64_t __b,
-	        float64_t __c, const int __lane)
+	        float64x1_t __c, const int __lane)
 {
-  return __builtin_fma (__b, __c, __a);
+  return __builtin_fma (__b, __c[0], __a);
 }
 
 __extension__ static __inline float32_t __attribute__ ((__always_inline__))
@@ -16736,11 +16744,12 @@ vfma_laneq_f32 (float32x2_t __a, float32x2_t __b,
 				    __a);
 }
 
-__extension__ static __inline float64_t __attribute__ ((__always_inline__))
-vfma_laneq_f64 (float64_t __a, float64_t __b,
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vfma_laneq_f64 (float64x1_t __a, float64x1_t __b,
 	        float64x2_t __c, const int __lane)
 {
-  return __builtin_fma (__b, __aarch64_vgetq_lane_f64 (__c, __lane), __a);
+  float64_t __c0 = __aarch64_vgetq_lane_f64 (__c, __lane);
+  return (float64x1_t) {__builtin_fma (__b[0], __c0, __a[0])};
 }
 
 __extension__ static __inline float64_t __attribute__ ((__always_inline__))
@@ -16770,9 +16779,9 @@ vfmaq_lane_f32 (float32x4_t __a, float32x4_t __b,
 
 __extension__ static __inline float64x2_t __attribute__ ((__always_inline__))
 vfmaq_lane_f64 (float64x2_t __a, float64x2_t __b,
-	        float64_t __c, const int __lane)
+	        float64x1_t __c, const int __lane)
 {
-  return __builtin_aarch64_fmav2df (__b, vdupq_n_f64 (__c), __a);
+  return __builtin_aarch64_fmav2df (__b, vdupq_n_f64 (__c[0]), __a);
 }
 
 /* vfmaq_laneq  */
@@ -16806,18 +16815,18 @@ vfms_lane_f32 (float32x2_t __a, float32x2_t __b,
 				    __a);
 }
 
-__extension__ static __inline float64_t __attribute__ ((__always_inline__))
-vfms_lane_f64 (float64_t __a, float64_t __b,
-	       float64_t __c, const int __lane)
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vfms_lane_f64 (float64x1_t __a, float64x1_t __b,
+	       float64x1_t __c, const int __lane)
 {
-  return __builtin_fma (-__b, __c, __a);
+  return (float64x1_t) {__builtin_fma (-__b[0], __c[0], __a[0])};
 }
 
 __extension__ static __inline float64_t __attribute__ ((__always_inline__))
 vfmsd_lane_f64 (float64_t __a, float64_t __b,
-	        float64_t __c, const int __lane)
+	        float64x1_t __c, const int __lane)
 {
-  return __builtin_fma (-__b, __c, __a);
+  return __builtin_fma (-__b, __c[0], __a);
 }
 
 __extension__ static __inline float32_t __attribute__ ((__always_inline__))
@@ -16838,11 +16847,12 @@ vfms_laneq_f32 (float32x2_t __a, float32x2_t __b,
 				    __a);
 }
 
-__extension__ static __inline float64_t __attribute__ ((__always_inline__))
-vfms_laneq_f64 (float64_t __a, float64_t __b,
+__extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
+vfms_laneq_f64 (float64x1_t __a, float64x1_t __b,
 	        float64x2_t __c, const int __lane)
 {
-  return __builtin_fma (-__b, __aarch64_vgetq_lane_f64 (__c, __lane), __a);
+  float64_t __c0 = __aarch64_vgetq_lane_f64 (__c, __lane);
+  return (float64x1_t) {__builtin_fma (-__b[0], __c0, __a[0])};
 }
 
 __extension__ static __inline float64_t __attribute__ ((__always_inline__))
@@ -16872,9 +16882,9 @@ vfmsq_lane_f32 (float32x4_t __a, float32x4_t __b,
 
 __extension__ static __inline float64x2_t __attribute__ ((__always_inline__))
 vfmsq_lane_f64 (float64x2_t __a, float64x2_t __b,
-	        float64_t __c, const int __lane)
+	        float64x1_t __c, const int __lane)
 {
-  return __builtin_aarch64_fmav2df (-__b, vdupq_n_f64 (__c), __a);
+  return __builtin_aarch64_fmav2df (-__b, vdupq_n_f64 (__c[0]), __a);
 }
 
 /* vfmsq_laneq  */
@@ -16908,7 +16918,7 @@ vld1_f32 (const float32_t *a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vld1_f64 (const float64_t *a)
 {
-  return *a;
+  return (float64x1_t) {*a};
 }
 
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
@@ -17086,8 +17096,8 @@ vld2_f64 (const float64_t * __a)
   float64x1x2_t ret;
   __builtin_aarch64_simd_oi __o;
   __o = __builtin_aarch64_ld2df ((const __builtin_aarch64_simd_df *) __a);
-  ret.val[0] = (float64x1_t) __builtin_aarch64_get_dregoidf (__o, 0);
-  ret.val[1] = (float64x1_t) __builtin_aarch64_get_dregoidf (__o, 1);
+  ret.val[0] = (float64x1_t) {__builtin_aarch64_get_dregoidf (__o, 0)};
+  ret.val[1] = (float64x1_t) {__builtin_aarch64_get_dregoidf (__o, 1)};
   return ret;
 }
 
@@ -17352,9 +17362,9 @@ vld3_f64 (const float64_t * __a)
   float64x1x3_t ret;
   __builtin_aarch64_simd_ci __o;
   __o = __builtin_aarch64_ld3df ((const __builtin_aarch64_simd_df *) __a);
-  ret.val[0] = (float64x1_t) __builtin_aarch64_get_dregcidf (__o, 0);
-  ret.val[1] = (float64x1_t) __builtin_aarch64_get_dregcidf (__o, 1);
-  ret.val[2] = (float64x1_t) __builtin_aarch64_get_dregcidf (__o, 2);
+  ret.val[0] = (float64x1_t) {__builtin_aarch64_get_dregcidf (__o, 0)};
+  ret.val[1] = (float64x1_t) {__builtin_aarch64_get_dregcidf (__o, 1)};
+  ret.val[2] = (float64x1_t) {__builtin_aarch64_get_dregcidf (__o, 2)};
   return ret;
 }
 
@@ -17642,10 +17652,10 @@ vld4_f64 (const float64_t * __a)
   float64x1x4_t ret;
   __builtin_aarch64_simd_xi __o;
   __o = __builtin_aarch64_ld4df ((const __builtin_aarch64_simd_df *) __a);
-  ret.val[0] = (float64x1_t) __builtin_aarch64_get_dregxidf (__o, 0);
-  ret.val[1] = (float64x1_t) __builtin_aarch64_get_dregxidf (__o, 1);
-  ret.val[2] = (float64x1_t) __builtin_aarch64_get_dregxidf (__o, 2);
-  ret.val[3] = (float64x1_t) __builtin_aarch64_get_dregxidf (__o, 3);
+  ret.val[0] = (float64x1_t) {__builtin_aarch64_get_dregxidf (__o, 0)};
+  ret.val[1] = (float64x1_t) {__builtin_aarch64_get_dregxidf (__o, 1)};
+  ret.val[2] = (float64x1_t) {__builtin_aarch64_get_dregxidf (__o, 2)};
+  ret.val[3] = (float64x1_t) {__builtin_aarch64_get_dregxidf (__o, 3)};
   return ret;
 }
 
@@ -18760,7 +18770,7 @@ vmov_n_f32 (float32_t __a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vmov_n_f64 (float64_t __a)
 {
-  return __a;
+  return (float64x1_t) {__a};
 }
 
 __extension__ static __inline poly8x8_t __attribute__ ((__always_inline__))
@@ -18982,7 +18992,8 @@ vmulq_lane_f32 (float32x4_t __a, float32x2_t __b, const int __lane)
 __extension__ static __inline float64x2_t __attribute__ ((__always_inline__))
 vmulq_lane_f64 (float64x2_t __a, float64x1_t __b, const int __lane)
 {
-  return __a * __b;
+  __builtin_aarch64_im_lane_boundsi (__lane, 1);
+  return __a * __b[0];
 }
 
 __extension__ static __inline int16x8_t __attribute__ ((__always_inline__))
@@ -21028,7 +21039,7 @@ vrndn_f32 (float32x2_t __a)
 __extension__ static __inline float64x1_t __attribute__ ((__always_inline__))
 vrndn_f64 (float64x1_t __a)
 {
-  return __builtin_aarch64_frintndf (__a);
+  return (float64x1_t) {__builtin_aarch64_frintndf (__a[0])};
 }
 
 __extension__ static __inline float32x4_t __attribute__ ((__always_inline__))
@@ -22336,7 +22347,7 @@ vst1_f32 (float32_t *a, float32x2_t b)
 __extension__ static __inline void __attribute__ ((__always_inline__))
 vst1_f64 (float64_t *a, float64x1_t b)
 {
-  *a = b;
+  *a = b[0];
 }
 
 __extension__ static __inline void __attribute__ ((__always_inline__))
diff --git a/gcc/config/aarch64/iterators.md b/gcc/config/aarch64/iterators.md
index 5c304bff239..3203c3da7e2 100644
--- a/gcc/config/aarch64/iterators.md
+++ b/gcc/config/aarch64/iterators.md
@@ -138,8 +138,11 @@
 ;; Double vector modes for combines.
 (define_mode_iterator VDIC [V8QI V4HI V2SI])
 
-;; Double vector modes.
-(define_mode_iterator VD_RE [V8QI V4HI V2SI DI DF V2SF])
+;; Double vector modes, inc. V1DF and the DI "vector" mode, for VREINTERPRET.
+(define_mode_iterator VD_RE [V8QI V4HI V2SI DI V1DF V2SF])
+
+;; Double vector modes inc V1DF
+(define_mode_iterator VD1 [V8QI V4HI V2SI V2SF V1DF])
 
 ;; Vector modes except double int.
 (define_mode_iterator VDQIF [V8QI V16QI V4HI V8HI V2SI V4SI V2SF V4SF V2DF])