diff mbox series

[v2,3/3,GCC] arm: vld1_types_x4 ACLE intrinsics

Message ID 20231207154106.4808-4-Ezra.Sitorus@arm.com
State Unresolved
Headers
Series arm: vld1_types_xN ACLE intrinsics |

Checks

Context Check Description
snail/gcc-patch-check warning Git am fail log

Commit Message

Ezra Sitorus Dec. 7, 2023, 3:41 p.m. UTC 
  From: Ezra Sitorus <ezra.sitorus@arm.com>

This patch is part of a series of patches implementing the _xN
variants of the vld1 intrinsic for the arm port. This patch adds the
_x4 variants of the vld1 intrinsic.

The previous vld1_x4 has been updated to vld1q_x4 to take into
account that it works with 4-word-length types. vld1_x4 is now
only for 2-word-length types.

ACLE documents:
https://developer.arm.com/documentation/ihi0053/latest/

ISA documents:
https://developer.arm.com/documentation/ddi0487/latest/

gcc/ChangeLog:
        * config/arm/arm_neon.h
        (vld1_u8_x4, vld1_u16_x4, vld1_u32_x4, vld1_u64_x4): New
        (vld1_s8_x4, vld1_s16_x4, vld1_s32_x4, vld1_s64_x4): New.
        (vld1_f16_x4, vld1_f32_x4): New.
        (vld1_p8_x4, vld1_p16_x4, vld1_p64_x4): New.
        (vld1_bf16_x4): New.
        (vld1q_types_x4): Updated to use vld1q_x4
        from arm_neon_builtins.def
        * config/arm/arm_neon_builtins.def
        (vld1_x4): Updated entries.
        (vld1q_x4): New entries, but comes from the old vld1_x2
        * config/arm/neon.md (neon_vld1q_x4<mode>):
        Updated from neon_vld1_x4<mode>.

gcc/testsuite/ChangeLog:
        * gcc.target/arm/simd/vld1_base_xN_1.c: Add new tests.
        * gcc.target/arm/simd/vld1_bf16_xN_1.c: Add new tests.
        * gcc.target/arm/simd/vld1_fp16_xN_1.c: Add new tests.
        * gcc.target/arm/simd/vld1_p64_xN_1.c: Add new tests.
---
 gcc/config/arm/arm_neon.h                     | 156 ++++++++++++++++--
 gcc/config/arm/arm_neon_builtins.def          |   3 +-
 gcc/config/arm/neon.md                        |  10 ++
 .../gcc.target/arm/simd/vld1_base_xN_1.c      |  63 ++++++-
 .../gcc.target/arm/simd/vld1_bf16_xN_1.c      |   7 +-
 .../gcc.target/arm/simd/vld1_fp16_xN_1.c      |   7 +-
 .../gcc.target/arm/simd/vld1_p64_xN_1.c       |   7 +-
 7 files changed, 231 insertions(+), 22 deletions(-)
diff mbox series

Patch

diff --git a/gcc/config/arm/arm_neon.h b/gcc/config/arm/arm_neon.h
index dbc37cafe28..8bcf1d6325e 100644
--- a/gcc/config/arm/arm_neon.h
+++ b/gcc/config/arm/arm_neon.h
@@ -10325,6 +10325,15 @@  vld1_p64_x3 (const poly64_t * __a)
   return __rv.__i;
 }
 
+__extension__ extern __inline poly64x1x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_p64_x4 (const poly64_t * __a)
+{
+  union { poly64x1x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4di ((const __builtin_neon_di *) __a);
+  return __rv.__i;
+}
+
 #pragma GCC pop_options
 __extension__ extern __inline int8x8_t
 __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
@@ -10426,6 +10435,42 @@  vld1_s64_x3 (const int64_t * __a)
   return __rv.__i;
 }
 
+__extension__ extern __inline int8x8x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_s8_x4 (const int8_t * __a)
+{
+  union { int8x8x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v8qi ((const __builtin_neon_qi *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline int16x4x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_s16_x4 (const int16_t * __a)
+{
+  union { int16x4x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v4hi ((const __builtin_neon_hi *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline int32x2x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_s32_x4 (const int32_t * __a)
+{
+  union { int32x2x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v2si ((const __builtin_neon_si *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline int64x1x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_s64_x4 (const int64_t * __a)
+{
+  union { int64x1x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4di ((const __builtin_neon_di *) __a);
+  return __rv.__i;
+}
+
 #if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
 __extension__ extern __inline float16x4_t
 __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
@@ -10482,6 +10527,26 @@  vld1_f32_x3 (const float32_t * __a)
   return __rv.__i;
 }
 
+#if defined (__ARM_FP16_FORMAT_IEEE) || defined (__ARM_FP16_FORMAT_ALTERNATIVE)
+__extension__ extern __inline float16x4x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_f16_x4 (const float16_t * __a)
+{
+  union { float16x4x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v4hf (__a);
+  return __rv.__i;
+}
+#endif
+
+__extension__ extern __inline float32x2x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_f32_x4 (const float32_t * __a)
+{
+  union { float32x2x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v2sf ((const __builtin_neon_sf *) __a);
+  return __rv.__i;
+}
+
 __extension__ extern __inline uint8x8_t
 __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1_u8 (const uint8_t * __a)
@@ -10582,6 +10647,42 @@  vld1_u64_x3 (const uint64_t * __a)
   return __rv.__i;
 }
 
+__extension__ extern __inline uint8x8x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_u8_x4 (const uint8_t * __a)
+{
+  union { uint8x8x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v8qi ((const __builtin_neon_qi *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline uint16x4x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_u16_x4 (const uint16_t * __a)
+{
+  union { uint16x4x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v4hi ((const __builtin_neon_hi *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline uint32x2x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_u32_x4 (const uint32_t * __a)
+{
+  union { uint32x2x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v2si ((const __builtin_neon_si *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline uint64x1x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_u64_x4 (const uint64_t * __a)
+{
+  union { uint64x1x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4di ((const __builtin_neon_di *) __a);
+  return __rv.__i;
+}
+
 __extension__ extern __inline poly8x8_t
 __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1_p8 (const poly8_t * __a)
@@ -10632,6 +10733,24 @@  vld1_p16_x3 (const poly16_t * __a)
   return __rv.__i;
 }
 
+__extension__ extern __inline poly8x8x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_p8_x4 (const poly8_t * __a)
+{
+  union { poly8x8x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v8qi ((const __builtin_neon_qi *) __a);
+  return __rv.__i;
+}
+
+__extension__ extern __inline poly16x4x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_p16_x4 (const poly16_t * __a)
+{
+  union { poly16x4x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v4hi ((const __builtin_neon_hi *) __a);
+  return __rv.__i;
+}
+
 #pragma GCC push_options
 #pragma GCC target ("fpu=crypto-neon-fp-armv8")
 __extension__ extern __inline poly64x2_t
@@ -10664,7 +10783,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_p64_x4 (const poly64_t * __a)
 {
   union { poly64x2x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v2di ((const __builtin_neon_di *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v2di ((const __builtin_neon_di *) __a);
   return __rv.__i;
 }
 
@@ -10774,7 +10893,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_s8_x4 (const uint8_t * __a)
 {
   union { int8x16x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v16qi ((const __builtin_neon_qi *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v16qi ((const __builtin_neon_qi *) __a);
   return __rv.__i;
 }
 
@@ -10783,7 +10902,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_s16_x4 (const uint16_t * __a)
 {
   union { int16x8x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v8hi ((const __builtin_neon_hi *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v8hi ((const __builtin_neon_hi *) __a);
   return __rv.__i;
 }
 
@@ -10792,7 +10911,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_s32_x4 (const int32_t * __a)
 {
   union { int32x4x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v4si ((const __builtin_neon_si *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v4si ((const __builtin_neon_si *) __a);
   return __rv.__i;
 }
 
@@ -10801,7 +10920,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_s64_x4 (const int64_t * __a)
 {
   union { int64x2x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v2di ((const __builtin_neon_di *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v2di ((const __builtin_neon_di *) __a);
   return __rv.__i;
 }
 
@@ -10867,7 +10986,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_f16_x4 (const float16_t * __a)
 {
   union { float16x8x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v8hf (__a);
+  __rv.__o = __builtin_neon_vld1q_x4v8hf (__a);
   return __rv.__i;
 }
 #endif
@@ -10877,7 +10996,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_f32_x4 (const float32_t * __a)
 {
   union { float32x4x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v4sf ((const __builtin_neon_sf *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v4sf ((const __builtin_neon_sf *) __a);
   return __rv.__i;
 }
 
@@ -10986,7 +11105,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_u8_x4 (const uint8_t * __a)
 {
   union { uint8x16x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v16qi ((const __builtin_neon_qi *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v16qi ((const __builtin_neon_qi *) __a);
   return __rv.__i;
 }
 
@@ -10995,7 +11114,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_u16_x4 (const uint16_t * __a)
 {
   union { uint16x8x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v8hi ((const __builtin_neon_hi *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v8hi ((const __builtin_neon_hi *) __a);
   return __rv.__i;
 }
 
@@ -11004,7 +11123,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_u32_x4 (const uint32_t * __a)
 {
   union { uint32x4x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v4si ((const __builtin_neon_si *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v4si ((const __builtin_neon_si *) __a);
   return __rv.__i;
 }
 
@@ -11013,7 +11132,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_u64_x4 (const uint64_t * __a)
 {
   union { uint64x2x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v2di ((const __builtin_neon_di *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v2di ((const __builtin_neon_di *) __a);
   return __rv.__i;
 }
 
@@ -11072,7 +11191,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_p8_x4 (const poly8_t * __a)
 {
   union { poly8x16x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v16qi ((const __builtin_neon_qi *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v16qi ((const __builtin_neon_qi *) __a);
   return __rv.__i;
 }
 
@@ -11081,7 +11200,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_p16_x4 (const poly16_t * __a)
 {
   union { poly16x8x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v8hi ((const __builtin_neon_hi *) __a);
+  __rv.__o = __builtin_neon_vld1q_x4v8hi ((const __builtin_neon_hi *) __a);
   return __rv.__i;
 }
 
@@ -21072,6 +21191,15 @@  vld1_bf16_x3 (const bfloat16_t * __ptr)
   return __rv.__i;
 }
 
+__extension__ extern __inline bfloat16x4x4_t
+__attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
+vld1_bf16_x4 (const bfloat16_t * __ptr)
+{
+  union { bfloat16x4x4_t __i; __builtin_neon_oi __o; } __rv;
+  __rv.__o = __builtin_neon_vld1_x4v4bf ((const __builtin_neon_bf *) __ptr);
+  return __rv.__i;
+}
+
 __extension__ extern __inline bfloat16x8_t
 __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_bf16 (const bfloat16_t * __ptr)
@@ -21102,7 +21230,7 @@  __attribute__  ((__always_inline__, __gnu_inline__, __artificial__))
 vld1q_bf16_x4 (const bfloat16_t * __ptr)
 {
   union { bfloat16x8x4_t __i; __builtin_neon_xi __o; } __rv;
-  __rv.__o = __builtin_neon_vld1_x4v8bf ((const __builtin_neon_bf *) __ptr);
+  __rv.__o = __builtin_neon_vld1q_x4v8bf ((const __builtin_neon_bf *) __ptr);
   return __rv.__i;
 }
 
diff --git a/gcc/config/arm/arm_neon_builtins.def b/gcc/config/arm/arm_neon_builtins.def
index c74f0db645b..20dfcae7de5 100644
--- a/gcc/config/arm/arm_neon_builtins.def
+++ b/gcc/config/arm/arm_neon_builtins.def
@@ -305,7 +305,8 @@  VAR7 (LOAD1, vld1_x2, v8qi, v4hi, v2si, di, v4hf, v2sf, v4bf)
 VAR7 (LOAD1, vld1q_x2, v16qi, v8hi, v4si, v2di, v8hf, v4sf, v8bf)
 VAR7 (LOAD1, vld1_x3, v8qi, v4hi, v2si, di, v4hf, v2sf, v4bf)
 VAR7 (LOAD1, vld1q_x3, v16qi, v8hi, v4si, v2di, v8hf, v4sf, v8bf)
-VAR7 (LOAD1, vld1_x4, v16qi, v8hi, v4si, v2di, v8hf, v4sf, v8bf)
+VAR7 (LOAD1, vld1_x4, v8qi, v4hi, v2si, di, v4hf, v2sf, v4bf)
+VAR7 (LOAD1, vld1q_x4, v16qi, v8hi, v4si, v2di, v8hf, v4sf, v8bf)
 VAR12 (LOAD1LANE, vld1_lane,
 	v8qi, v4hi, v2si, v2sf, di, v16qi, v8hi, v4si, v4sf, v2di, v4bf, v8bf)
 VAR10 (LOAD1, vld1_dup,
diff --git a/gcc/config/arm/neon.md b/gcc/config/arm/neon.md
index e67cbc247d9..30f5bf8e40e 100644
--- a/gcc/config/arm/neon.md
+++ b/gcc/config/arm/neon.md
@@ -5005,6 +5005,16 @@  if (BYTES_BIG_ENDIAN)
 )
 
 (define_insn "neon_vld1_x4<mode>"
+  [(set (match_operand:OI 0 "s_register_operand" "=w")
+        (unspec:OI [(match_operand:OI 1 "neon_struct_operand" "Um")
+                    (unspec:VDQX [(const_int 0)] UNSPEC_VSTRUCTDUMMY)]
+                    UNSPEC_VLD1))]
+  "TARGET_NEON"
+  "vld1.<V_sz_elem>\t%h0, %A1"
+  [(set_attr "type" "neon_load1_4reg<q>")]
+)
+
+(define_insn "neon_vld1q_x4<mode>"
   [(set (match_operand:XI 0 "s_register_operand" "=w")
         (unspec:XI [(match_operand:OI 1 "neon_struct_operand" "Um")
                     (unspec:VQXBF [(const_int 0)] UNSPEC_VSTRUCTDUMMY)]
diff --git a/gcc/testsuite/gcc.target/arm/simd/vld1_base_xN_1.c b/gcc/testsuite/gcc.target/arm/simd/vld1_base_xN_1.c
index 95314bbe0de..a5686ffac01 100644
--- a/gcc/testsuite/gcc.target/arm/simd/vld1_base_xN_1.c
+++ b/gcc/testsuite/gcc.target/arm/simd/vld1_base_xN_1.c
@@ -115,7 +115,62 @@  poly16x4x3_t test_vld1_p16_x3 (poly16_t * a)
     return vld1_p16_x3 (a);
 }
 
-/* { dg-final { scan-assembler-times {vld1.8\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 6 } }  */
-/* { dg-final { scan-assembler-times {vld1.16\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 6 } }  */
-/* { dg-final { scan-assembler-times {vld1.32\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 6 } }  */
-/* { dg-final { scan-assembler-times {vld1.64\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+:64\]\n} 4 } }  */
\ No newline at end of file
+uint8x8x4_t test_vld1_u8_x4 (uint8_t * a)
+{
+    return vld1_u8_x4 (a);
+}
+
+uint16x4x4_t test_vld1_u16_x4 (uint16_t * a)
+{
+    return vld1_u16_x4 (a);
+}
+
+uint32x2x4_t test_vld1_u32_x4 (uint32_t * a)
+{
+    return vld1_u32_x4 (a);
+}
+
+uint64x1x4_t test_vld1_u64_x4 (uint64_t * a)
+{
+    return vld1_u64_x4 (a);
+}
+
+int8x8x4_t test_vld1_s8_x4 (int8_t * a)
+{
+    return vld1_s8_x4 (a);
+}
+
+int16x4x4_t test_vld1_s16_x4 (int16_t * a)
+{
+    return vld1_s16_x4 (a);
+}
+
+int32x2x4_t test_vld1_s32_x4 (int32_t * a)
+{
+    return vld1_s32_x4 (a);
+}
+
+int64x1x4_t test_vld1_s64_x4 (int64_t * a)
+{
+    return vld1_s64_x4 (a);
+}
+
+float32x2x4_t test_vld1_f32_x4 (float32_t * a)
+{
+    return vld1_f32_x4 (a);
+}
+
+poly8x8x4_t test_vld1_p8_x4 (poly8_t * a)
+{
+    return vld1_p8_x4 (a);
+}
+
+poly16x4x4_t test_vld1_p16_x4 (poly16_t * a)
+{
+    return vld1_p16_x4 (a);
+}
+
+/* { dg-final { scan-assembler-times {vld1.8\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 9 } }  */
+/* { dg-final { scan-assembler-times {vld1.16\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 9 } }  */
+/* { dg-final { scan-assembler-times {vld1.32\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 9 } }  */
+/* { dg-final { scan-assembler-times {vld1.64\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+:64\]\n} 6 } }  */
\ No newline at end of file
diff --git a/gcc/testsuite/gcc.target/arm/simd/vld1_bf16_xN_1.c b/gcc/testsuite/gcc.target/arm/simd/vld1_bf16_xN_1.c
index c1935da0a4c..7ed17834ccf 100644
--- a/gcc/testsuite/gcc.target/arm/simd/vld1_bf16_xN_1.c
+++ b/gcc/testsuite/gcc.target/arm/simd/vld1_bf16_xN_1.c
@@ -15,4 +15,9 @@  bfloat16x4x3_t test_vld1_bf16_x3 (bfloat16_t * a)
     return vld1_bf16_x3 (a);
 }
 
-/* { dg-final { scan-assembler-times {vld1.16\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 2 } }  */
\ No newline at end of file
+bfloat16x4x4_t test_vld1_bf16_x4 (bfloat16_t * a)
+{
+    return vld1_bf16_x4 (a);
+}
+
+/* { dg-final { scan-assembler-times {vld1.16\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 3 } }  */
\ No newline at end of file
diff --git a/gcc/testsuite/gcc.target/arm/simd/vld1_fp16_xN_1.c b/gcc/testsuite/gcc.target/arm/simd/vld1_fp16_xN_1.c
index 20363239f5b..82e7211ebbf 100644
--- a/gcc/testsuite/gcc.target/arm/simd/vld1_fp16_xN_1.c
+++ b/gcc/testsuite/gcc.target/arm/simd/vld1_fp16_xN_1.c
@@ -15,4 +15,9 @@  float16x4x3_t test_vld1_f16_x3 (float16_t * a)
     return vld1_f16_x3 (a);
 }
 
-/* { dg-final { scan-assembler-times {vld1.16\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 2 } }  */
+float16x4x4_t test_vld1_f16_x4 (float16_t * a)
+{
+    return vld1_f16_x4 (a);
+}
+
+/* { dg-final { scan-assembler-times {vld1.16\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+\]\n} 3 } }  */
diff --git a/gcc/testsuite/gcc.target/arm/simd/vld1_p64_xN_1.c b/gcc/testsuite/gcc.target/arm/simd/vld1_p64_xN_1.c
index 210de511c71..644371b89ea 100644
--- a/gcc/testsuite/gcc.target/arm/simd/vld1_p64_xN_1.c
+++ b/gcc/testsuite/gcc.target/arm/simd/vld1_p64_xN_1.c
@@ -15,4 +15,9 @@  poly64x1x3_t test_vld1_p64_x3 (poly64_t * a)
     return vld1_p64_x3 (a);
 }
 
-/* { dg-final { scan-assembler-times {vld1.64\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+:64\]\n} 2 } }  */
+poly64x1x4_t test_vld1_p64_x4 (poly64_t * a)
+{
+    return vld1_p64_x4 (a);
+}
+
+/* { dg-final { scan-assembler-times {vld1.64\t\{d[0-9]+-d[0-9]+\}, \[r[0-9]+:64\]\n} 3 } }  */