[gcc(refs/users/marxin/heads/vect_cond_expr-rework-v3)] [ARM][GCC][4/5x]: MVE load intrinsics with zero(_z) suffix.

Martin Liska marxin@gcc.gnu.org
Tue Mar 24 13:57:34 GMT 2020


https://gcc.gnu.org/g:429d607bc468828ea1e40852bbf8a1e9bbc34e7a

commit 429d607bc468828ea1e40852bbf8a1e9bbc34e7a
Author: Srinath Parvathaneni <srinath.parvathaneni@arm.com>
Date:   Wed Mar 18 18:35:17 2020 +0000

    [ARM][GCC][4/5x]: MVE load intrinsics with zero(_z) suffix.
    
    This patch supports the following MVE ACLE load intrinsics with zero(_z) suffix.
    * ``_z`` (zero) which indicates false-predicated lanes are filled with zeroes, these are only used for load instructions.
    
    vldrbq_gather_offset_z_s16, vldrbq_gather_offset_z_u8, vldrbq_gather_offset_z_s32, vldrbq_gather_offset_z_u16, vldrbq_gather_offset_z_u32, vldrbq_gather_offset_z_s8, vldrbq_z_s16, vldrbq_z_u8, vldrbq_z_s8, vldrbq_z_s32, vldrbq_z_u16, vldrbq_z_u32, vldrwq_gather_base_z_u32, vldrwq_gather_base_z_s32.
    
    Please refer to M-profile Vector Extension (MVE) intrinsics [1]  for more details.
    [1]  https://developer.arm.com/architectures/instruction-sets/simd-isas/helium/mve-intrinsics
    
    2020-03-18  Andre Vieira  <andre.simoesdiasvieira@arm.com>
                Mihail Ionescu  <mihail.ionescu@arm.com>
                Srinath Parvathaneni  <srinath.parvathaneni@arm.com>
    
            * config/arm/arm-builtins.c (LDRGBS_Z_QUALIFIERS): Define builtin
            qualifier.
            (LDRGBU_Z_QUALIFIERS): Likewise.
            (LDRGS_Z_QUALIFIERS): Likewise.
            (LDRGU_Z_QUALIFIERS): Likewise.
            (LDRS_Z_QUALIFIERS): Likewise.
            (LDRU_Z_QUALIFIERS): Likewise.
            * config/arm/arm_mve.h (vldrbq_gather_offset_z_s16): Define macro.
            (vldrbq_gather_offset_z_u8): Likewise.
            (vldrbq_gather_offset_z_s32): Likewise.
            (vldrbq_gather_offset_z_u16): Likewise.
            (vldrbq_gather_offset_z_u32): Likewise.
            (vldrbq_gather_offset_z_s8): Likewise.
            (vldrbq_z_s16): Likewise.
            (vldrbq_z_u8): Likewise.
            (vldrbq_z_s8): Likewise.
            (vldrbq_z_s32): Likewise.
            (vldrbq_z_u16): Likewise.
            (vldrbq_z_u32): Likewise.
            (vldrwq_gather_base_z_u32): Likewise.
            (vldrwq_gather_base_z_s32): Likewise.
            (__arm_vldrbq_gather_offset_z_s8): Define intrinsic.
            (__arm_vldrbq_gather_offset_z_s32): Likewise.
            (__arm_vldrbq_gather_offset_z_s16): Likewise.
            (__arm_vldrbq_gather_offset_z_u8): Likewise.
            (__arm_vldrbq_gather_offset_z_u32): Likewise.
            (__arm_vldrbq_gather_offset_z_u16): Likewise.
            (__arm_vldrbq_z_s8): Likewise.
            (__arm_vldrbq_z_s32): Likewise.
            (__arm_vldrbq_z_s16): Likewise.
            (__arm_vldrbq_z_u8): Likewise.
            (__arm_vldrbq_z_u32): Likewise.
            (__arm_vldrbq_z_u16): Likewise.
            (__arm_vldrwq_gather_base_z_s32): Likewise.
            (__arm_vldrwq_gather_base_z_u32): Likewise.
            (vldrbq_gather_offset_z): Define polymorphic variant.
            * config/arm/arm_mve_builtins.def (LDRGBS_Z_QUALIFIERS): Use builtin
            qualifier.
            (LDRGBU_Z_QUALIFIERS): Likewise.
            (LDRGS_Z_QUALIFIERS): Likewise.
            (LDRGU_Z_QUALIFIERS): Likewise.
            (LDRS_Z_QUALIFIERS): Likewise.
            (LDRU_Z_QUALIFIERS): Likewise.
            * config/arm/mve.md (mve_vldrbq_gather_offset_z_<supf><mode>): Define
            RTL pattern.
            (mve_vldrbq_z_<supf><mode>): Likewise.
            (mve_vldrwq_gather_base_z_<supf>v4si): Likewise.
    
    gcc/testsuite/ChangeLog: Likewise.
    
    2020-03-18  Andre Vieira  <andre.simoesdiasvieira@arm.com>
                Mihail Ionescu  <mihail.ionescu@arm.com>
                Srinath Parvathaneni  <srinath.parvathaneni@arm.com>
    
            * gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s16.c: New test.
            * gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s32.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s8.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u16.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u32.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u8.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_z_s16.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_z_s32.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_z_s8.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_z_u16.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_z_u32.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrbq_z_u8.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_s32.c: Likewise.
            * gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_u32.c: Likewise.

Diff:
---
 gcc/ChangeLog                                      |  52 +++++++++
 gcc/config/arm/arm-builtins.c                      |  34 ++++++
 gcc/config/arm/arm_mve.h                           | 118 ++++++++++++++++++++-
 gcc/config/arm/arm_mve_builtins.def                |   6 ++
 gcc/config/arm/mve.md                              |  66 ++++++++++++
 gcc/testsuite/ChangeLog                            |  19 ++++
 .../mve/intrinsics/vldrbq_gather_offset_z_s16.c    |  22 ++++
 .../mve/intrinsics/vldrbq_gather_offset_z_s32.c    |  22 ++++
 .../arm/mve/intrinsics/vldrbq_gather_offset_z_s8.c |  22 ++++
 .../mve/intrinsics/vldrbq_gather_offset_z_u16.c    |  22 ++++
 .../mve/intrinsics/vldrbq_gather_offset_z_u32.c    |  22 ++++
 .../arm/mve/intrinsics/vldrbq_gather_offset_z_u8.c |  22 ++++
 .../gcc.target/arm/mve/intrinsics/vldrbq_z_s16.c   |  14 +++
 .../gcc.target/arm/mve/intrinsics/vldrbq_z_s32.c   |  14 +++
 .../gcc.target/arm/mve/intrinsics/vldrbq_z_s8.c    |  14 +++
 .../gcc.target/arm/mve/intrinsics/vldrbq_z_u16.c   |  14 +++
 .../gcc.target/arm/mve/intrinsics/vldrbq_z_u32.c   |  14 +++
 .../gcc.target/arm/mve/intrinsics/vldrbq_z_u8.c    |  14 +++
 .../arm/mve/intrinsics/vldrwq_gather_base_z_s32.c  |  14 +++
 .../arm/mve/intrinsics/vldrwq_gather_base_z_u32.c  |  14 +++
 20 files changed, 537 insertions(+), 2 deletions(-)

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 33c02ebaa47..4d4f9b1ec63 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,55 @@
+2020-03-18  Andre Vieira  <andre.simoesdiasvieira@arm.com>
+            Mihail Ionescu  <mihail.ionescu@arm.com>
+            Srinath Parvathaneni  <srinath.parvathaneni@arm.com>
+
+	* config/arm/arm-builtins.c (LDRGBS_Z_QUALIFIERS): Define builtin
+	qualifier.
+	(LDRGBU_Z_QUALIFIERS): Likewise.
+	(LDRGS_Z_QUALIFIERS): Likewise.
+	(LDRGU_Z_QUALIFIERS): Likewise.
+	(LDRS_Z_QUALIFIERS): Likewise.
+	(LDRU_Z_QUALIFIERS): Likewise.
+	* config/arm/arm_mve.h (vldrbq_gather_offset_z_s16): Define macro.
+	(vldrbq_gather_offset_z_u8): Likewise.
+	(vldrbq_gather_offset_z_s32): Likewise.
+	(vldrbq_gather_offset_z_u16): Likewise.
+	(vldrbq_gather_offset_z_u32): Likewise.
+	(vldrbq_gather_offset_z_s8): Likewise.
+	(vldrbq_z_s16): Likewise.
+	(vldrbq_z_u8): Likewise.
+	(vldrbq_z_s8): Likewise.
+	(vldrbq_z_s32): Likewise.
+	(vldrbq_z_u16): Likewise.
+	(vldrbq_z_u32): Likewise.
+	(vldrwq_gather_base_z_u32): Likewise.
+	(vldrwq_gather_base_z_s32): Likewise.
+	(__arm_vldrbq_gather_offset_z_s8): Define intrinsic.
+	(__arm_vldrbq_gather_offset_z_s32): Likewise.
+	(__arm_vldrbq_gather_offset_z_s16): Likewise.
+	(__arm_vldrbq_gather_offset_z_u8): Likewise.
+	(__arm_vldrbq_gather_offset_z_u32): Likewise.
+	(__arm_vldrbq_gather_offset_z_u16): Likewise.
+	(__arm_vldrbq_z_s8): Likewise.
+	(__arm_vldrbq_z_s32): Likewise.
+	(__arm_vldrbq_z_s16): Likewise.
+	(__arm_vldrbq_z_u8): Likewise.
+	(__arm_vldrbq_z_u32): Likewise.
+	(__arm_vldrbq_z_u16): Likewise.
+	(__arm_vldrwq_gather_base_z_s32): Likewise.
+	(__arm_vldrwq_gather_base_z_u32): Likewise.
+	(vldrbq_gather_offset_z): Define polymorphic variant.
+	* config/arm/arm_mve_builtins.def (LDRGBS_Z_QUALIFIERS): Use builtin
+	qualifier.
+	(LDRGBU_Z_QUALIFIERS): Likewise.
+	(LDRGS_Z_QUALIFIERS): Likewise.
+	(LDRGU_Z_QUALIFIERS): Likewise.
+	(LDRS_Z_QUALIFIERS): Likewise.
+	(LDRU_Z_QUALIFIERS): Likewise.
+	* config/arm/mve.md (mve_vldrbq_gather_offset_z_<supf><mode>): Define
+	RTL pattern.
+	(mve_vldrbq_z_<supf><mode>): Likewise.
+	(mve_vldrwq_gather_base_z_<supf>v4si): Likewise.
+
 2020-03-18  Andre Vieira  <andre.simoesdiasvieira@arm.com>
             Mihail Ionescu  <mihail.ionescu@arm.com>
             Srinath Parvathaneni  <srinath.parvathaneni@arm.com>
diff --git a/gcc/config/arm/arm-builtins.c b/gcc/config/arm/arm-builtins.c
index c87fa311851..c3deb9efc88 100644
--- a/gcc/config/arm/arm-builtins.c
+++ b/gcc/config/arm/arm-builtins.c
@@ -677,6 +677,40 @@ arm_ldrgbu_qualifiers[SIMD_MAX_BUILTIN_ARGS]
   = { qualifier_unsigned, qualifier_unsigned, qualifier_immediate};
 #define LDRGBU_QUALIFIERS (arm_ldrgbu_qualifiers)
 
+static enum arm_type_qualifiers
+arm_ldrgbs_z_qualifiers[SIMD_MAX_BUILTIN_ARGS]
+  = { qualifier_none, qualifier_unsigned, qualifier_immediate,
+      qualifier_unsigned};
+#define LDRGBS_Z_QUALIFIERS (arm_ldrgbs_z_qualifiers)
+
+static enum arm_type_qualifiers
+arm_ldrgbu_z_qualifiers[SIMD_MAX_BUILTIN_ARGS]
+  = { qualifier_unsigned, qualifier_unsigned, qualifier_immediate,
+      qualifier_unsigned};
+#define LDRGBU_Z_QUALIFIERS (arm_ldrgbu_z_qualifiers)
+
+static enum arm_type_qualifiers
+arm_ldrgs_z_qualifiers[SIMD_MAX_BUILTIN_ARGS]
+  = { qualifier_none, qualifier_pointer, qualifier_unsigned,
+      qualifier_unsigned};
+#define LDRGS_Z_QUALIFIERS (arm_ldrgs_z_qualifiers)
+
+static enum arm_type_qualifiers
+arm_ldrgu_z_qualifiers[SIMD_MAX_BUILTIN_ARGS]
+  = { qualifier_unsigned, qualifier_pointer, qualifier_unsigned,
+      qualifier_unsigned};
+#define LDRGU_Z_QUALIFIERS (arm_ldrgu_z_qualifiers)
+
+static enum arm_type_qualifiers
+arm_ldrs_z_qualifiers[SIMD_MAX_BUILTIN_ARGS]
+  = { qualifier_none, qualifier_pointer, qualifier_unsigned};
+#define LDRS_Z_QUALIFIERS (arm_ldrs_z_qualifiers)
+
+static enum arm_type_qualifiers
+arm_ldru_z_qualifiers[SIMD_MAX_BUILTIN_ARGS]
+  = { qualifier_unsigned, qualifier_pointer, qualifier_unsigned};
+#define LDRU_Z_QUALIFIERS (arm_ldru_z_qualifiers)
+
 /* End of Qualifier for MVE builtins.  */
 
    /* void ([T element type] *, T, immediate).  */
diff --git a/gcc/config/arm/arm_mve.h b/gcc/config/arm/arm_mve.h
index deed81cfb31..4570a0b16c3 100644
--- a/gcc/config/arm/arm_mve.h
+++ b/gcc/config/arm/arm_mve.h
@@ -1744,6 +1744,20 @@ typedef struct { uint8x16_t val[4]; } uint8x16x4_t;
 #define vstrbq_scatter_offset_p_u16( __base, __offset, __value, __p) __arm_vstrbq_scatter_offset_p_u16( __base, __offset, __value, __p)
 #define vstrwq_scatter_base_p_s32(__addr,  __offset, __value, __p) __arm_vstrwq_scatter_base_p_s32(__addr,  __offset, __value, __p)
 #define vstrwq_scatter_base_p_u32(__addr,  __offset, __value, __p) __arm_vstrwq_scatter_base_p_u32(__addr,  __offset, __value, __p)
+#define vldrbq_gather_offset_z_s16(__base, __offset, __p) __arm_vldrbq_gather_offset_z_s16(__base, __offset, __p)
+#define vldrbq_gather_offset_z_u8(__base, __offset, __p) __arm_vldrbq_gather_offset_z_u8(__base, __offset, __p)
+#define vldrbq_gather_offset_z_s32(__base, __offset, __p) __arm_vldrbq_gather_offset_z_s32(__base, __offset, __p)
+#define vldrbq_gather_offset_z_u16(__base, __offset, __p) __arm_vldrbq_gather_offset_z_u16(__base, __offset, __p)
+#define vldrbq_gather_offset_z_u32(__base, __offset, __p) __arm_vldrbq_gather_offset_z_u32(__base, __offset, __p)
+#define vldrbq_gather_offset_z_s8(__base, __offset, __p) __arm_vldrbq_gather_offset_z_s8(__base, __offset, __p)
+#define vldrbq_z_s16(__base, __p) __arm_vldrbq_z_s16(__base, __p)
+#define vldrbq_z_u8(__base, __p) __arm_vldrbq_z_u8(__base, __p)
+#define vldrbq_z_s8(__base, __p) __arm_vldrbq_z_s8(__base, __p)
+#define vldrbq_z_s32(__base, __p) __arm_vldrbq_z_s32(__base, __p)
+#define vldrbq_z_u16(__base, __p) __arm_vldrbq_z_u16(__base, __p)
+#define vldrbq_z_u32(__base, __p) __arm_vldrbq_z_u32(__base, __p)
+#define vldrwq_gather_base_z_u32(__addr,  __offset, __p) __arm_vldrwq_gather_base_z_u32(__addr,  __offset, __p)
+#define vldrwq_gather_base_z_s32(__addr,  __offset, __p) __arm_vldrwq_gather_base_z_s32(__addr,  __offset, __p)
 #endif
 
 __extension__ extern __inline void
@@ -11330,6 +11344,105 @@ __arm_vstrwq_scatter_base_p_u32 (uint32x4_t __addr, const int __offset, uint32x4
 {
   __builtin_mve_vstrwq_scatter_base_p_uv4si (__addr, __offset, __value, __p);
 }
+
+__extension__ extern __inline int8x16_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_gather_offset_z_s8 (int8_t const * __base, uint8x16_t __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_gather_offset_z_sv16qi ((__builtin_neon_qi *) __base, __offset, __p);
+}
+
+__extension__ extern __inline int32x4_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_gather_offset_z_s32 (int8_t const * __base, uint32x4_t __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_gather_offset_z_sv4si ((__builtin_neon_qi *) __base, __offset, __p);
+}
+
+__extension__ extern __inline int16x8_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_gather_offset_z_s16 (int8_t const * __base, uint16x8_t __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_gather_offset_z_sv8hi ((__builtin_neon_qi *) __base, __offset, __p);
+}
+
+__extension__ extern __inline uint8x16_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_gather_offset_z_u8 (uint8_t const * __base, uint8x16_t __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_gather_offset_z_uv16qi ((__builtin_neon_qi *) __base, __offset, __p);
+}
+
+__extension__ extern __inline uint32x4_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_gather_offset_z_u32 (uint8_t const * __base, uint32x4_t __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_gather_offset_z_uv4si ((__builtin_neon_qi *) __base, __offset, __p);
+}
+
+__extension__ extern __inline uint16x8_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_gather_offset_z_u16 (uint8_t const * __base, uint16x8_t __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_gather_offset_z_uv8hi ((__builtin_neon_qi *) __base, __offset, __p);
+}
+
+__extension__ extern __inline int8x16_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_z_s8 (int8_t const * __base, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_z_sv16qi ((__builtin_neon_qi *) __base, __p);
+}
+
+__extension__ extern __inline int32x4_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_z_s32 (int8_t const * __base, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_z_sv4si ((__builtin_neon_qi *) __base, __p);
+}
+
+__extension__ extern __inline int16x8_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_z_s16 (int8_t const * __base, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_z_sv8hi ((__builtin_neon_qi *) __base, __p);
+}
+
+__extension__ extern __inline uint8x16_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_z_u8 (uint8_t const * __base, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_z_uv16qi ((__builtin_neon_qi *) __base, __p);
+}
+
+__extension__ extern __inline uint32x4_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_z_u32 (uint8_t const * __base, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_z_uv4si ((__builtin_neon_qi *) __base, __p);
+}
+
+__extension__ extern __inline uint16x8_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrbq_z_u16 (uint8_t const * __base, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrbq_z_uv8hi ((__builtin_neon_qi *) __base, __p);
+}
+
+__extension__ extern __inline int32x4_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrwq_gather_base_z_s32 (uint32x4_t __addr, const int __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrwq_gather_base_z_sv4si (__addr, __offset, __p);
+}
+
+__extension__ extern __inline uint32x4_t
+__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
+__arm_vldrwq_gather_base_z_u32 (uint32x4_t __addr, const int __offset, mve_pred16_t __p)
+{
+  return __builtin_mve_vldrwq_gather_base_z_uv4si (__addr, __offset, __p);
+}
+
 #if (__ARM_FEATURE_MVE & 2) /* MVE Floating point.  */
 
 __extension__ extern __inline void
@@ -13471,6 +13584,7 @@ __arm_vsubq_m_n_f16 (float16x8_t __inactive, float16x8_t __a, float16_t __b, mve
 {
   return __builtin_mve_vsubq_m_n_fv8hf (__inactive, __a, __b, __p);
 }
+
 #endif
 
 enum {
@@ -18034,6 +18148,8 @@ extern void *__ARM_undef;
   int (*)[__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_base_p_s32 (p0, p1, __ARM_mve_coerce(__p2, int32x4_t), p3), \
   int (*)[__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_base_p_u32 (p0, p1, __ARM_mve_coerce(__p2, uint32x4_t), p3));})
 
+#endif /* MVE Integer.  */
+
 #define vldrbq_gather_offset_z(p0,p1,p2) __arm_vldrbq_gather_offset_z(p0,p1,p2)
 #define __arm_vldrbq_gather_offset_z(p0,p1,p2) ({ __typeof(p0) __p0 = (p0); \
   __typeof(p1) __p1 = (p1); \
@@ -18045,8 +18161,6 @@ extern void *__ARM_undef;
   int (*)[__ARM_mve_type_uint8_t_const_ptr][__ARM_mve_type_uint16x8_t]: __arm_vldrbq_gather_offset_z_u16 (__ARM_mve_coerce(__p0, uint8_t const *), __ARM_mve_coerce(__p1, uint16x8_t), p2), \
   int (*)[__ARM_mve_type_uint8_t_const_ptr][__ARM_mve_type_uint32x4_t]: __arm_vldrbq_gather_offset_z_u32 (__ARM_mve_coerce(__p0, uint8_t const *), __ARM_mve_coerce(__p1, uint32x4_t), p2));})
 
-#endif /* MVE Integer.  */
-
 #define vqrdmlahq_m(p0,p1,p2,p3) __arm_vqrdmlahq_m(p0,p1,p2,p3)
 #define __arm_vqrdmlahq_m(p0,p1,p2,p3) ({ __typeof(p0) __p0 = (p0); \
   __typeof(p1) __p1 = (p1); \
diff --git a/gcc/config/arm/arm_mve_builtins.def b/gcc/config/arm/arm_mve_builtins.def
index c6e065d52d0..0f466e4cec2 100644
--- a/gcc/config/arm/arm_mve_builtins.def
+++ b/gcc/config/arm/arm_mve_builtins.def
@@ -703,3 +703,9 @@ VAR3 (STRSS_P, vstrbq_scatter_offset_p_s, v16qi, v8hi, v4si)
 VAR3 (STRSU_P, vstrbq_scatter_offset_p_u, v16qi, v8hi, v4si)
 VAR1 (STRSBS_P, vstrwq_scatter_base_p_s, v4si)
 VAR1 (STRSBU_P, vstrwq_scatter_base_p_u, v4si)
+VAR1 (LDRGBS_Z, vldrwq_gather_base_z_s, v4si)
+VAR1 (LDRGBU_Z, vldrwq_gather_base_z_u, v4si)
+VAR3 (LDRGS_Z, vldrbq_gather_offset_z_s, v16qi, v8hi, v4si)
+VAR3 (LDRGU_Z, vldrbq_gather_offset_z_u, v16qi, v8hi, v4si)
+VAR3 (LDRS_Z, vldrbq_z_s, v16qi, v8hi, v4si)
+VAR3 (LDRU_Z, vldrbq_z_u, v16qi, v8hi, v4si)
diff --git a/gcc/config/arm/mve.md b/gcc/config/arm/mve.md
index 0e8b04f531c..03a90ab3212 100644
--- a/gcc/config/arm/mve.md
+++ b/gcc/config/arm/mve.md
@@ -8142,3 +8142,69 @@
    return "";
 }
   [(set_attr "length" "8")])
+
+;;
+;; [vldrbq_gather_offset_z_s vldrbq_gather_offset_z_u]
+;;
+(define_insn "mve_vldrbq_gather_offset_z_<supf><mode>"
+  [(set (match_operand:MVE_2 0 "s_register_operand" "=&w")
+	(unspec:MVE_2 [(match_operand:<MVE_B_ELEM> 1 "memory_operand" "Us")
+		       (match_operand:MVE_2 2 "s_register_operand" "w")
+		       (match_operand:HI 3 "vpr_register_operand" "Up")]
+	 VLDRBGOQ))
+  ]
+  "TARGET_HAVE_MVE"
+{
+   rtx ops[4];
+   ops[0] = operands[0];
+   ops[1] = operands[1];
+   ops[2] = operands[2];
+   ops[3] = operands[3];
+   if (!strcmp ("<supf>","s") && <V_sz_elem> == 8)
+     output_asm_insn ("vpst\n\tvldrbt.u8\t%q0, [%m1, %q2]",ops);
+   else
+     output_asm_insn ("vpst\n\tvldrbt.<supf><V_sz_elem>\t%q0, [%m1, %q2]",ops);
+   return "";
+}
+  [(set_attr "length" "8")])
+
+;;
+;; [vldrbq_z_s vldrbq_z_u]
+;;
+(define_insn "mve_vldrbq_z_<supf><mode>"
+  [(set (match_operand:MVE_2 0 "s_register_operand" "=w")
+	(unspec:MVE_2 [(match_operand:<MVE_B_ELEM> 1 "memory_operand" "Us")
+		       (match_operand:HI 2 "vpr_register_operand" "Up")]
+	 VLDRBQ))
+  ]
+  "TARGET_HAVE_MVE"
+{
+   rtx ops[2];
+   int regno = REGNO (operands[0]);
+   ops[0] = gen_rtx_REG (TImode, regno);
+   ops[1]  = operands[1];
+   output_asm_insn ("vpst\n\tvldrbt.<supf><V_sz_elem>\t%q0, %E1",ops);
+   return "";
+}
+  [(set_attr "length" "8")])
+
+;;
+;; [vldrwq_gather_base_z_s vldrwq_gather_base_z_u]
+;;
+(define_insn "mve_vldrwq_gather_base_z_<supf>v4si"
+  [(set (match_operand:V4SI 0 "s_register_operand" "=&w")
+	(unspec:V4SI [(match_operand:V4SI 1 "s_register_operand" "w")
+		      (match_operand:SI 2 "immediate_operand" "i")
+		      (match_operand:HI 3 "vpr_register_operand" "Up")]
+	 VLDRWGBQ))
+  ]
+  "TARGET_HAVE_MVE"
+{
+   rtx ops[3];
+   ops[0] = operands[0];
+   ops[1] = operands[1];
+   ops[2] = operands[2];
+   output_asm_insn ("vpst\n\tvldrwt.u32\t%q0, [%q1, %2]",ops);
+   return "";
+}
+  [(set_attr "length" "8")])
diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index 1998f3eb62a..451f0eee276 100644
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,22 @@
+2020-03-18  Andre Vieira  <andre.simoesdiasvieira@arm.com>
+            Mihail Ionescu  <mihail.ionescu@arm.com>
+            Srinath Parvathaneni  <srinath.parvathaneni@arm.com>
+
+	* gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s16.c: New test.
+	* gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s32.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s8.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u16.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u32.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u8.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_z_s16.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_z_s32.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_z_s8.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_z_u16.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_z_u32.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrbq_z_u8.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_s32.c: Likewise.
+	* gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_u32.c: Likewise.
+
 2020-03-18  Andre Vieira  <andre.simoesdiasvieira@arm.com>
             Mihail Ionescu  <mihail.ionescu@arm.com>
             Srinath Parvathaneni  <srinath.parvathaneni@arm.com>
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s16.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s16.c
new file mode 100644
index 00000000000..a51689f7740
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s16.c
@@ -0,0 +1,22 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int16x8_t
+foo (int8_t const * base, uint16x8_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z_s16 (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s16"  }  } */
+
+int16x8_t
+foo1 (int8_t const * base, uint16x8_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s16"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s32.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s32.c
new file mode 100644
index 00000000000..4b3b8ba3249
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s32.c
@@ -0,0 +1,22 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int32x4_t
+foo (int8_t const * base, uint32x4_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z_s32 (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s32"  }  } */
+
+int32x4_t
+foo1 (int8_t const * base, uint32x4_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s32"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s8.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s8.c
new file mode 100644
index 00000000000..14e0f301d72
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_s8.c
@@ -0,0 +1,22 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int8x16_t
+foo (int8_t const * base, uint8x16_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z_s8 (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u8"  }  } */
+
+int8x16_t
+foo1 (int8_t const * base, uint8x16_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u8"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u16.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u16.c
new file mode 100644
index 00000000000..7b075001c81
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u16.c
@@ -0,0 +1,22 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint16x8_t
+foo (uint8_t const * base, uint16x8_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z_u16 (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u16"  }  } */
+
+uint16x8_t
+foo1 (uint8_t const * base, uint16x8_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u16"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u32.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u32.c
new file mode 100644
index 00000000000..c81ef6d365a
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u32.c
@@ -0,0 +1,22 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint32x4_t
+foo (uint8_t const * base, uint32x4_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z_u32 (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u32"  }  } */
+
+uint32x4_t
+foo1 (uint8_t const * base, uint32x4_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u32"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u8.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u8.c
new file mode 100644
index 00000000000..3c91278ddc9
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_gather_offset_z_u8.c
@@ -0,0 +1,22 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint8x16_t
+foo (uint8_t const * base, uint8x16_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z_u8 (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u8"  }  } */
+
+uint8x16_t
+foo1 (uint8_t const * base, uint8x16_t offset, mve_pred16_t p)
+{
+  return vldrbq_gather_offset_z (base, offset, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u8"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s16.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s16.c
new file mode 100644
index 00000000000..c709db6c3ea
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s16.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int16x8_t
+foo (int8_t const * base, mve_pred16_t p)
+{
+  return vldrbq_z_s16 (base, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s16"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s32.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s32.c
new file mode 100644
index 00000000000..aa47e239c4b
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s32.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int32x4_t
+foo (int8_t const * base, mve_pred16_t p)
+{
+  return vldrbq_z_s32 (base, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s32"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s8.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s8.c
new file mode 100644
index 00000000000..85d7ef43667
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_s8.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int8x16_t
+foo (int8_t const * base, mve_pred16_t p)
+{
+  return vldrbq_z_s8 (base, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.s8"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u16.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u16.c
new file mode 100644
index 00000000000..054305797eb
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u16.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint16x8_t
+foo (uint8_t const * base, mve_pred16_t p)
+{
+  return vldrbq_z_u16 (base, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u16"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u32.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u32.c
new file mode 100644
index 00000000000..73b9ea19a6e
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u32.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint32x4_t
+foo (uint8_t const * base, mve_pred16_t p)
+{
+  return vldrbq_z_u32 (base, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u32"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u8.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u8.c
new file mode 100644
index 00000000000..3925c9f17a2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrbq_z_u8.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint8x16_t
+foo (uint8_t const * base, mve_pred16_t p)
+{
+  return vldrbq_z_u8 (base, p);
+}
+
+/* { dg-final { scan-assembler "vldrbt.u8"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_s32.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_s32.c
new file mode 100644
index 00000000000..2e61760ede7
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_s32.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+int32x4_t
+foo (uint32x4_t addr, mve_pred16_t p)
+{
+  return vldrwq_gather_base_z_s32 (addr, 4, p);
+}
+
+/* { dg-final { scan-assembler "vldrwt.u32"  }  } */
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_u32.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_u32.c
new file mode 100644
index 00000000000..a71c562722d
--- /dev/null
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/vldrwq_gather_base_z_u32.c
@@ -0,0 +1,14 @@
+/* { dg-do compile  } */
+/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
+/* { dg-additional-options "-O2" } */
+
+#include "arm_mve.h"
+
+uint32x4_t
+foo (uint32x4_t addr, mve_pred16_t p)
+{
+  return vldrwq_gather_base_z_u32 (addr, 4, p);
+}
+
+/* { dg-final { scan-assembler "vldrwt.u32"  }  } */


More information about the Gcc-cvs mailing list