@@ -610,18 +610,6 @@ arm_quadop_unone_unone_unone_none_pred_qualifiers[SIMD_MAX_BUILTIN_ARGS]
#define QUADOP_UNONE_UNONE_UNONE_NONE_PRED_QUALIFIERS \
(arm_quadop_unone_unone_unone_none_pred_qualifiers)
-static enum arm_type_qualifiers
-arm_strss_qualifiers[SIMD_MAX_BUILTIN_ARGS]
- = { qualifier_void, qualifier_pointer, qualifier_unsigned,
- qualifier_none};
-#define STRSS_QUALIFIERS (arm_strss_qualifiers)
-
-static enum arm_type_qualifiers
-arm_strsu_qualifiers[SIMD_MAX_BUILTIN_ARGS]
- = { qualifier_void, qualifier_pointer, qualifier_unsigned,
- qualifier_unsigned};
-#define STRSU_QUALIFIERS (arm_strsu_qualifiers)
-
static enum arm_type_qualifiers
arm_strsbs_qualifiers[SIMD_MAX_BUILTIN_ARGS]
= { qualifier_void, qualifier_unsigned, qualifier_immediate, qualifier_none};
@@ -633,18 +621,6 @@ arm_strsbu_qualifiers[SIMD_MAX_BUILTIN_ARGS]
qualifier_unsigned};
#define STRSBU_QUALIFIERS (arm_strsbu_qualifiers)
-static enum arm_type_qualifiers
-arm_strsu_p_qualifiers[SIMD_MAX_BUILTIN_ARGS]
- = { qualifier_void, qualifier_pointer, qualifier_unsigned,
- qualifier_unsigned, qualifier_predicate};
-#define STRSU_P_QUALIFIERS (arm_strsu_p_qualifiers)
-
-static enum arm_type_qualifiers
-arm_strss_p_qualifiers[SIMD_MAX_BUILTIN_ARGS]
- = { qualifier_void, qualifier_pointer, qualifier_unsigned,
- qualifier_none, qualifier_predicate};
-#define STRSS_P_QUALIFIERS (arm_strss_p_qualifiers)
-
static enum arm_type_qualifiers
arm_strsbs_p_qualifiers[SIMD_MAX_BUILTIN_ARGS]
= { qualifier_void, qualifier_unsigned, qualifier_immediate,
@@ -245,6 +245,16 @@ class vstrq_scatter_impl : public store_truncating
public:
using store_truncating::store_truncating;
+ CONSTEXPR vstrq_scatter_impl (bool shifted,
+ scalar_mode to_int_mode,
+ opt_scalar_mode to_float_mode)
+ : store_truncating (to_int_mode, to_float_mode),
+ m_shifted (shifted)
+ {}
+
+ /* Shifted offset (true) or plain offset (false). */
+ bool m_shifted;
+
rtx expand (function_expander &e) const override
{
insn_code icode;
@@ -255,15 +265,23 @@ public:
case PRED_none:
icode = (e.vector_mode (0) == memory_mode
/* Non-truncating store case. */
- ? code_for_mve_vstrq_scatter_offset (memory_mode)
+ ? (m_shifted
+ ? code_for_mve_vstrq_scatter_shifted_offset (memory_mode)
+ : code_for_mve_vstrq_scatter_offset (memory_mode))
/* Truncating store case. */
- : code_for_mve_vstrq_truncate_scatter_offset (memory_mode));
+ : (m_shifted
+ ? CODE_FOR_mve_vstrq_truncate_scatter_shifted_offset_v4si
+ : code_for_mve_vstrq_truncate_scatter_offset (memory_mode)));
break;
case PRED_p:
icode = (e.vector_mode (0) == memory_mode
- ? code_for_mve_vstrq_scatter_offset_p (memory_mode)
- : code_for_mve_vstrq_truncate_scatter_offset_p (memory_mode));
+ ? (m_shifted
+ ? code_for_mve_vstrq_scatter_shifted_offset_p (memory_mode)
+ : code_for_mve_vstrq_scatter_offset_p (memory_mode))
+ : (m_shifted
+ ? CODE_FOR_mve_vstrq_truncate_scatter_shifted_offset_p_v4si
+ : code_for_mve_vstrq_truncate_scatter_offset_p (memory_mode)));
break;
default:
@@ -1241,12 +1259,15 @@ FUNCTION_ONLY_N_NO_F (vsliq, VSLIQ)
FUNCTION_ONLY_N_NO_F (vsriq, VSRIQ)
FUNCTION (vst1q, vst1_impl,)
FUNCTION (vstrbq, vstrq_impl, (QImode, opt_scalar_mode ()))
-FUNCTION (vstrbq_scatter, vstrq_scatter_impl, (QImode, opt_scalar_mode ()))
-FUNCTION (vstrdq_scatter, vstrq_scatter_impl, (DImode, opt_scalar_mode ()))
+FUNCTION (vstrbq_scatter, vstrq_scatter_impl, (false, QImode, opt_scalar_mode ()))
+FUNCTION (vstrdq_scatter, vstrq_scatter_impl, (false, DImode, opt_scalar_mode ()))
+FUNCTION (vstrdq_scatter_shifted, vstrq_scatter_impl, (true, DImode, opt_scalar_mode ()))
FUNCTION (vstrhq, vstrq_impl, (HImode, HFmode))
-FUNCTION (vstrhq_scatter, vstrq_scatter_impl, (HImode, HFmode))
+FUNCTION (vstrhq_scatter, vstrq_scatter_impl, (false, HImode, HFmode))
+FUNCTION (vstrhq_scatter_shifted, vstrq_scatter_impl, (true, HImode, HFmode))
FUNCTION (vstrwq, vstrq_impl, (SImode, SFmode))
-FUNCTION (vstrwq_scatter, vstrq_scatter_impl, (SImode, SFmode))
+FUNCTION (vstrwq_scatter, vstrq_scatter_impl, (false, SImode, SFmode))
+FUNCTION (vstrwq_scatter_shifted, vstrq_scatter_impl, (true, SImode, SFmode))
FUNCTION_WITH_RTX_M_N (vsubq, MINUS, VSUBQ)
FUNCTION (vuninitializedq, vuninitializedq_impl,)
@@ -174,9 +174,12 @@ DEF_MVE_FUNCTION (vstrbq, store, all_integer, p_or_none)
DEF_MVE_FUNCTION (vstrbq_scatter, store_scatter_offset, all_integer, p_or_none)
DEF_MVE_FUNCTION (vstrhq, store, integer_16_32, p_or_none)
DEF_MVE_FUNCTION (vstrhq_scatter, store_scatter_offset, integer_16_32, p_or_none)
+DEF_MVE_FUNCTION (vstrhq_scatter_shifted, store_scatter_offset, integer_16_32, p_or_none)
DEF_MVE_FUNCTION (vstrwq, store, integer_32, p_or_none)
DEF_MVE_FUNCTION (vstrwq_scatter, store_scatter_offset, integer_32, p_or_none)
+DEF_MVE_FUNCTION (vstrwq_scatter_shifted, store_scatter_offset, integer_32, p_or_none)
DEF_MVE_FUNCTION (vstrdq_scatter, store_scatter_offset, integer_64, p_or_none)
+DEF_MVE_FUNCTION (vstrdq_scatter_shifted, store_scatter_offset, integer_64, p_or_none)
DEF_MVE_FUNCTION (vsubq, binary_opt_n, all_integer, mx_or_none)
DEF_MVE_FUNCTION (vuninitializedq, inherent, all_integer_with_64, none)
#undef REQUIRES_FLOAT
@@ -247,8 +250,10 @@ DEF_MVE_FUNCTION (vrndxq, unary, all_float, mx_or_none)
DEF_MVE_FUNCTION (vst1q, store, all_float, p_or_none)
DEF_MVE_FUNCTION (vstrhq, store, float_16, p_or_none)
DEF_MVE_FUNCTION (vstrhq_scatter, store_scatter_offset, float_16, p_or_none)
+DEF_MVE_FUNCTION (vstrhq_scatter_shifted, store_scatter_offset, float_16, p_or_none)
DEF_MVE_FUNCTION (vstrwq, store, float_32, p_or_none)
DEF_MVE_FUNCTION (vstrwq_scatter, store_scatter_offset, float_32, p_or_none)
+DEF_MVE_FUNCTION (vstrwq_scatter_shifted, store_scatter_offset, float_32, p_or_none)
DEF_MVE_FUNCTION (vsubq, binary_opt_n, all_float, mx_or_none)
DEF_MVE_FUNCTION (vuninitializedq, inherent, all_float, none)
#undef REQUIRES_FLOAT
@@ -208,10 +208,13 @@ extern const function_base *const vst1q;
extern const function_base *const vstrbq;
extern const function_base *const vstrbq_scatter;
extern const function_base *const vstrdq_scatter;
+extern const function_base *const vstrdq_scatter_shifted;
extern const function_base *const vstrhq;
extern const function_base *const vstrhq_scatter;
+extern const function_base *const vstrhq_scatter_shifted;
extern const function_base *const vstrwq;
extern const function_base *const vstrwq_scatter;
+extern const function_base *const vstrwq_scatter_shifted;
extern const function_base *const vsubq;
extern const function_base *const vuninitializedq;
@@ -58,14 +58,8 @@
#define vldrwq_gather_offset_z(__base, __offset, __p) __arm_vldrwq_gather_offset_z(__base, __offset, __p)
#define vldrwq_gather_shifted_offset(__base, __offset) __arm_vldrwq_gather_shifted_offset(__base, __offset)
#define vldrwq_gather_shifted_offset_z(__base, __offset, __p) __arm_vldrwq_gather_shifted_offset_z(__base, __offset, __p)
-#define vstrhq_scatter_shifted_offset(__base, __offset, __value) __arm_vstrhq_scatter_shifted_offset(__base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_p(__base, __offset, __value, __p) __arm_vstrhq_scatter_shifted_offset_p(__base, __offset, __value, __p)
#define vstrdq_scatter_base_p(__addr, __offset, __value, __p) __arm_vstrdq_scatter_base_p(__addr, __offset, __value, __p)
#define vstrdq_scatter_base(__addr, __offset, __value) __arm_vstrdq_scatter_base(__addr, __offset, __value)
-#define vstrdq_scatter_shifted_offset_p(__base, __offset, __value, __p) __arm_vstrdq_scatter_shifted_offset_p(__base, __offset, __value, __p)
-#define vstrdq_scatter_shifted_offset(__base, __offset, __value) __arm_vstrdq_scatter_shifted_offset(__base, __offset, __value)
-#define vstrwq_scatter_shifted_offset_p(__base, __offset, __value, __p) __arm_vstrwq_scatter_shifted_offset_p(__base, __offset, __value, __p)
-#define vstrwq_scatter_shifted_offset(__base, __offset, __value) __arm_vstrwq_scatter_shifted_offset(__base, __offset, __value)
#define vuninitializedq(__v) __arm_vuninitializedq(__v)
#define vstrdq_scatter_base_wb(__addr, __offset, __value) __arm_vstrdq_scatter_base_wb(__addr, __offset, __value)
#define vstrdq_scatter_base_wb_p(__addr, __offset, __value, __p) __arm_vstrdq_scatter_base_wb_p(__addr, __offset, __value, __p)
@@ -153,32 +147,12 @@
#define vldrwq_gather_shifted_offset_z_f32(__base, __offset, __p) __arm_vldrwq_gather_shifted_offset_z_f32(__base, __offset, __p)
#define vldrwq_gather_shifted_offset_z_s32(__base, __offset, __p) __arm_vldrwq_gather_shifted_offset_z_s32(__base, __offset, __p)
#define vldrwq_gather_shifted_offset_z_u32(__base, __offset, __p) __arm_vldrwq_gather_shifted_offset_z_u32(__base, __offset, __p)
-#define vstrhq_scatter_shifted_offset_s32( __base, __offset, __value) __arm_vstrhq_scatter_shifted_offset_s32( __base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_s16( __base, __offset, __value) __arm_vstrhq_scatter_shifted_offset_s16( __base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_u32( __base, __offset, __value) __arm_vstrhq_scatter_shifted_offset_u32( __base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_u16( __base, __offset, __value) __arm_vstrhq_scatter_shifted_offset_u16( __base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_p_s32( __base, __offset, __value, __p) __arm_vstrhq_scatter_shifted_offset_p_s32( __base, __offset, __value, __p)
-#define vstrhq_scatter_shifted_offset_p_s16( __base, __offset, __value, __p) __arm_vstrhq_scatter_shifted_offset_p_s16( __base, __offset, __value, __p)
-#define vstrhq_scatter_shifted_offset_p_u32( __base, __offset, __value, __p) __arm_vstrhq_scatter_shifted_offset_p_u32( __base, __offset, __value, __p)
-#define vstrhq_scatter_shifted_offset_p_u16( __base, __offset, __value, __p) __arm_vstrhq_scatter_shifted_offset_p_u16( __base, __offset, __value, __p)
#define vstrdq_scatter_base_p_s64(__addr, __offset, __value, __p) __arm_vstrdq_scatter_base_p_s64(__addr, __offset, __value, __p)
#define vstrdq_scatter_base_p_u64(__addr, __offset, __value, __p) __arm_vstrdq_scatter_base_p_u64(__addr, __offset, __value, __p)
#define vstrdq_scatter_base_s64(__addr, __offset, __value) __arm_vstrdq_scatter_base_s64(__addr, __offset, __value)
#define vstrdq_scatter_base_u64(__addr, __offset, __value) __arm_vstrdq_scatter_base_u64(__addr, __offset, __value)
-#define vstrdq_scatter_shifted_offset_p_s64(__base, __offset, __value, __p) __arm_vstrdq_scatter_shifted_offset_p_s64(__base, __offset, __value, __p)
-#define vstrdq_scatter_shifted_offset_p_u64(__base, __offset, __value, __p) __arm_vstrdq_scatter_shifted_offset_p_u64(__base, __offset, __value, __p)
-#define vstrdq_scatter_shifted_offset_s64(__base, __offset, __value) __arm_vstrdq_scatter_shifted_offset_s64(__base, __offset, __value)
-#define vstrdq_scatter_shifted_offset_u64(__base, __offset, __value) __arm_vstrdq_scatter_shifted_offset_u64(__base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_f16(__base, __offset, __value) __arm_vstrhq_scatter_shifted_offset_f16(__base, __offset, __value)
-#define vstrhq_scatter_shifted_offset_p_f16(__base, __offset, __value, __p) __arm_vstrhq_scatter_shifted_offset_p_f16(__base, __offset, __value, __p)
#define vstrwq_scatter_base_f32(__addr, __offset, __value) __arm_vstrwq_scatter_base_f32(__addr, __offset, __value)
#define vstrwq_scatter_base_p_f32(__addr, __offset, __value, __p) __arm_vstrwq_scatter_base_p_f32(__addr, __offset, __value, __p)
-#define vstrwq_scatter_shifted_offset_f32(__base, __offset, __value) __arm_vstrwq_scatter_shifted_offset_f32(__base, __offset, __value)
-#define vstrwq_scatter_shifted_offset_p_f32(__base, __offset, __value, __p) __arm_vstrwq_scatter_shifted_offset_p_f32(__base, __offset, __value, __p)
-#define vstrwq_scatter_shifted_offset_p_s32(__base, __offset, __value, __p) __arm_vstrwq_scatter_shifted_offset_p_s32(__base, __offset, __value, __p)
-#define vstrwq_scatter_shifted_offset_p_u32(__base, __offset, __value, __p) __arm_vstrwq_scatter_shifted_offset_p_u32(__base, __offset, __value, __p)
-#define vstrwq_scatter_shifted_offset_s32(__base, __offset, __value) __arm_vstrwq_scatter_shifted_offset_s32(__base, __offset, __value)
-#define vstrwq_scatter_shifted_offset_u32(__base, __offset, __value) __arm_vstrwq_scatter_shifted_offset_u32(__base, __offset, __value)
#define vuninitializedq_u8(void) __arm_vuninitializedq_u8(void)
#define vuninitializedq_u16(void) __arm_vuninitializedq_u16(void)
#define vuninitializedq_u32(void) __arm_vuninitializedq_u32(void)
@@ -737,62 +711,6 @@ __arm_vldrwq_gather_shifted_offset_z_u32 (uint32_t const * __base, uint32x4_t __
return __builtin_mve_vldrwq_gather_shifted_offset_z_uv4si ((__builtin_neon_si *) __base, __offset, __p);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_s32 (int16_t * __base, uint32x4_t __offset, int32x4_t __value)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_sv4si ((__builtin_neon_hi *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_s16 (int16_t * __base, uint16x8_t __offset, int16x8_t __value)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_sv8hi ((__builtin_neon_hi *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_u32 (uint16_t * __base, uint32x4_t __offset, uint32x4_t __value)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_uv4si ((__builtin_neon_hi *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_u16 (uint16_t * __base, uint16x8_t __offset, uint16x8_t __value)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_uv8hi ((__builtin_neon_hi *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p_s32 (int16_t * __base, uint32x4_t __offset, int32x4_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_p_sv4si ((__builtin_neon_hi *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p_s16 (int16_t * __base, uint16x8_t __offset, int16x8_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_p_sv8hi ((__builtin_neon_hi *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p_u32 (uint16_t * __base, uint32x4_t __offset, uint32x4_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_p_uv4si ((__builtin_neon_hi *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p_u16 (uint16_t * __base, uint16x8_t __offset, uint16x8_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_p_uv8hi ((__builtin_neon_hi *) __base, __offset, __value, __p);
-}
-
__extension__ extern __inline void
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vstrdq_scatter_base_p_s64 (uint64x2_t __addr, const int __offset, int64x2_t __value, mve_pred16_t __p)
@@ -821,62 +739,6 @@ __arm_vstrdq_scatter_base_u64 (uint64x2_t __addr, const int __offset, uint64x2_t
__builtin_mve_vstrdq_scatter_base_uv2di (__addr, __offset, __value);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset_p_s64 (int64_t * __base, uint64x2_t __offset, int64x2_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrdq_scatter_shifted_offset_p_sv2di ((__builtin_neon_di *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset_p_u64 (uint64_t * __base, uint64x2_t __offset, uint64x2_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrdq_scatter_shifted_offset_p_uv2di ((__builtin_neon_di *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset_s64 (int64_t * __base, uint64x2_t __offset, int64x2_t __value)
-{
- __builtin_mve_vstrdq_scatter_shifted_offset_sv2di ((__builtin_neon_di *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset_u64 (uint64_t * __base, uint64x2_t __offset, uint64x2_t __value)
-{
- __builtin_mve_vstrdq_scatter_shifted_offset_uv2di ((__builtin_neon_di *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_p_s32 (int32_t * __base, uint32x4_t __offset, int32x4_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrwq_scatter_shifted_offset_p_sv4si ((__builtin_neon_si *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_p_u32 (uint32_t * __base, uint32x4_t __offset, uint32x4_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrwq_scatter_shifted_offset_p_uv4si ((__builtin_neon_si *) __base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_s32 (int32_t * __base, uint32x4_t __offset, int32x4_t __value)
-{
- __builtin_mve_vstrwq_scatter_shifted_offset_sv4si ((__builtin_neon_si *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_u32 (uint32_t * __base, uint32x4_t __offset, uint32x4_t __value)
-{
- __builtin_mve_vstrwq_scatter_shifted_offset_uv4si ((__builtin_neon_si *) __base, __offset, __value);
-}
-
__extension__ extern __inline int64x2_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vldrdq_gather_base_wb_s64 (uint64x2_t * __addr, const int __offset)
@@ -1513,20 +1375,6 @@ __arm_vldrwq_gather_shifted_offset_z_f32 (float32_t const * __base, uint32x4_t _
return __builtin_mve_vldrwq_gather_shifted_offset_z_fv4sf ((__builtin_neon_si *) __base, __offset, __p);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_f16 (float16_t * __base, uint16x8_t __offset, float16x8_t __value)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_fv8hf ((__builtin_neon_hi *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p_f16 (float16_t * __base, uint16x8_t __offset, float16x8_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrhq_scatter_shifted_offset_p_fv8hf ((__builtin_neon_hi *) __base, __offset, __value, __p);
-}
-
__extension__ extern __inline void
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vstrwq_scatter_base_f32 (uint32x4_t __addr, const int __offset, float32x4_t __value)
@@ -1541,20 +1389,6 @@ __arm_vstrwq_scatter_base_p_f32 (uint32x4_t __addr, const int __offset, float32x
__builtin_mve_vstrwq_scatter_base_p_fv4sf (__addr, __offset, __value, __p);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_f32 (float32_t * __base, uint32x4_t __offset, float32x4_t __value)
-{
- __builtin_mve_vstrwq_scatter_shifted_offset_fv4sf ((__builtin_neon_si *) __base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_p_f32 (float32_t * __base, uint32x4_t __offset, float32x4_t __value, mve_pred16_t __p)
-{
- __builtin_mve_vstrwq_scatter_shifted_offset_p_fv4sf ((__builtin_neon_si *) __base, __offset, __value, __p);
-}
-
__extension__ extern __inline float32x4_t
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vldrwq_gather_base_wb_f32 (uint32x4_t * __addr, const int __offset)
@@ -2057,62 +1891,6 @@ __arm_vldrwq_gather_shifted_offset_z (uint32_t const * __base, uint32x4_t __offs
return __arm_vldrwq_gather_shifted_offset_z_u32 (__base, __offset, __p);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset (int16_t * __base, uint32x4_t __offset, int32x4_t __value)
-{
- __arm_vstrhq_scatter_shifted_offset_s32 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset (int16_t * __base, uint16x8_t __offset, int16x8_t __value)
-{
- __arm_vstrhq_scatter_shifted_offset_s16 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset (uint16_t * __base, uint32x4_t __offset, uint32x4_t __value)
-{
- __arm_vstrhq_scatter_shifted_offset_u32 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset (uint16_t * __base, uint16x8_t __offset, uint16x8_t __value)
-{
- __arm_vstrhq_scatter_shifted_offset_u16 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p (int16_t * __base, uint32x4_t __offset, int32x4_t __value, mve_pred16_t __p)
-{
- __arm_vstrhq_scatter_shifted_offset_p_s32 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p (int16_t * __base, uint16x8_t __offset, int16x8_t __value, mve_pred16_t __p)
-{
- __arm_vstrhq_scatter_shifted_offset_p_s16 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p (uint16_t * __base, uint32x4_t __offset, uint32x4_t __value, mve_pred16_t __p)
-{
- __arm_vstrhq_scatter_shifted_offset_p_u32 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p (uint16_t * __base, uint16x8_t __offset, uint16x8_t __value, mve_pred16_t __p)
-{
- __arm_vstrhq_scatter_shifted_offset_p_u16 (__base, __offset, __value, __p);
-}
-
__extension__ extern __inline void
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vstrdq_scatter_base_p (uint64x2_t __addr, const int __offset, int64x2_t __value, mve_pred16_t __p)
@@ -2141,62 +1919,6 @@ __arm_vstrdq_scatter_base (uint64x2_t __addr, const int __offset, uint64x2_t __v
__arm_vstrdq_scatter_base_u64 (__addr, __offset, __value);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset_p (int64_t * __base, uint64x2_t __offset, int64x2_t __value, mve_pred16_t __p)
-{
- __arm_vstrdq_scatter_shifted_offset_p_s64 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset_p (uint64_t * __base, uint64x2_t __offset, uint64x2_t __value, mve_pred16_t __p)
-{
- __arm_vstrdq_scatter_shifted_offset_p_u64 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset (int64_t * __base, uint64x2_t __offset, int64x2_t __value)
-{
- __arm_vstrdq_scatter_shifted_offset_s64 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrdq_scatter_shifted_offset (uint64_t * __base, uint64x2_t __offset, uint64x2_t __value)
-{
- __arm_vstrdq_scatter_shifted_offset_u64 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_p (int32_t * __base, uint32x4_t __offset, int32x4_t __value, mve_pred16_t __p)
-{
- __arm_vstrwq_scatter_shifted_offset_p_s32 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_p (uint32_t * __base, uint32x4_t __offset, uint32x4_t __value, mve_pred16_t __p)
-{
- __arm_vstrwq_scatter_shifted_offset_p_u32 (__base, __offset, __value, __p);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset (int32_t * __base, uint32x4_t __offset, int32x4_t __value)
-{
- __arm_vstrwq_scatter_shifted_offset_s32 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset (uint32_t * __base, uint32x4_t __offset, uint32x4_t __value)
-{
- __arm_vstrwq_scatter_shifted_offset_u32 (__base, __offset, __value);
-}
-
__extension__ extern __inline void
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vstrdq_scatter_base_wb (uint64x2_t * __addr, const int __offset, int64x2_t __value)
@@ -2563,20 +2285,6 @@ __arm_vldrwq_gather_shifted_offset_z (float32_t const * __base, uint32x4_t __off
return __arm_vldrwq_gather_shifted_offset_z_f32 (__base, __offset, __p);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset (float16_t * __base, uint16x8_t __offset, float16x8_t __value)
-{
- __arm_vstrhq_scatter_shifted_offset_f16 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrhq_scatter_shifted_offset_p (float16_t * __base, uint16x8_t __offset, float16x8_t __value, mve_pred16_t __p)
-{
- __arm_vstrhq_scatter_shifted_offset_p_f16 (__base, __offset, __value, __p);
-}
-
__extension__ extern __inline void
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vstrwq_scatter_base (uint32x4_t __addr, const int __offset, float32x4_t __value)
@@ -2591,20 +2299,6 @@ __arm_vstrwq_scatter_base_p (uint32x4_t __addr, const int __offset, float32x4_t
__arm_vstrwq_scatter_base_p_f32 (__addr, __offset, __value, __p);
}
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset (float32_t * __base, uint32x4_t __offset, float32x4_t __value)
-{
- __arm_vstrwq_scatter_shifted_offset_f32 (__base, __offset, __value);
-}
-
-__extension__ extern __inline void
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vstrwq_scatter_shifted_offset_p (float32_t * __base, uint32x4_t __offset, float32x4_t __value, mve_pred16_t __p)
-{
- __arm_vstrwq_scatter_shifted_offset_p_f32 (__base, __offset, __value, __p);
-}
-
__extension__ extern __inline void
__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
__arm_vstrwq_scatter_base_wb (uint32x4_t * __addr, const int __offset, float32x4_t __value)
@@ -3101,42 +2795,6 @@ extern void *__ARM_undef;
int (*)[__ARM_mve_type_float16_t_ptr][__ARM_mve_type_float16x8x2_t]: __arm_vst2q_f16 (__ARM_mve_coerce_f16_ptr(p0, float16_t *), __ARM_mve_coerce(__p1, float16x8x2_t)), \
int (*)[__ARM_mve_type_float32_t_ptr][__ARM_mve_type_float32x4x2_t]: __arm_vst2q_f32 (__ARM_mve_coerce_f32_ptr(p0, float32_t *), __ARM_mve_coerce(__p1, float32x4x2_t)));})
-#define __arm_vstrhq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t), p3), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t), p3), \
- int (*)[__ARM_mve_type_float16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_float16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_f16 (__ARM_mve_coerce_f16_ptr(p0, float16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, float16x8_t), p3));})
-
-#define __arm_vstrhq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t)), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t)), \
- int (*)[__ARM_mve_type_float16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_float16x8_t]: __arm_vstrhq_scatter_shifted_offset_f16 (__ARM_mve_coerce_f16_ptr(p0, float16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, float16x8_t)));})
-
-#define __arm_vstrhq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t)), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t)), \
- int (*)[__ARM_mve_type_float16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_float16x8_t]: __arm_vstrhq_scatter_shifted_offset_f16 (__ARM_mve_coerce_f16_ptr(p0, float16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, float16x8_t)));})
-
-#define __arm_vstrhq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t), p3), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t), p3), \
- int (*)[__ARM_mve_type_float16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_float16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_f16 (__ARM_mve_coerce_f16_ptr(p0, float16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, float16x8_t), p3));})
-
#define __arm_vstrwq_scatter_base(p0,p1,p2) ({ __typeof(p2) __p2 = (p2); \
_Generic( (int (*)[__ARM_mve_typeid(__p2)])0, \
int (*)[__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_base_s32 (p0, p1, __ARM_mve_coerce(__p2, int32x4_t)), \
@@ -3149,34 +2807,6 @@ extern void *__ARM_undef;
int (*)[__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_base_p_u32(p0, p1, __ARM_mve_coerce(__p2, uint32x4_t), p3), \
int (*)[__ARM_mve_type_float32x4_t]: __arm_vstrwq_scatter_base_p_f32(p0, p1, __ARM_mve_coerce(__p2, float32x4_t), p3));})
-#define __arm_vstrwq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int32_t_ptr][__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s32_ptr(p0, int32_t *), __p1, __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __p1, __ARM_mve_coerce(__p2, uint32x4_t)), \
- int (*)[__ARM_mve_type_float32_t_ptr][__ARM_mve_type_float32x4_t]: __arm_vstrwq_scatter_shifted_offset_f32 (__ARM_mve_coerce_f32_ptr(p0, float32_t *), __p1, __ARM_mve_coerce(__p2, float32x4_t)));})
-
-#define __arm_vstrwq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int32_t_ptr][__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s32_ptr(p0, int32_t *), __p1, __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __p1, __ARM_mve_coerce(__p2, uint32x4_t), p3), \
- int (*)[__ARM_mve_type_float32_t_ptr][__ARM_mve_type_float32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_f32 (__ARM_mve_coerce_f32_ptr(p0, float32_t *), __p1, __ARM_mve_coerce(__p2, float32x4_t), p3));})
-
-#define __arm_vstrwq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int32_t_ptr][__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s32_ptr(p0, int32_t *), __p1, __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __p1, __ARM_mve_coerce(__p2, uint32x4_t), p3), \
- int (*)[__ARM_mve_type_float32_t_ptr][__ARM_mve_type_float32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_f32 (__ARM_mve_coerce_f32_ptr(p0, float32_t *), __p1, __ARM_mve_coerce(__p2, float32x4_t), p3));})
-
-#define __arm_vstrwq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int32_t_ptr][__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s32_ptr(p0, int32_t *), __p1, __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __p1, __ARM_mve_coerce(__p2, uint32x4_t)), \
- int (*)[__ARM_mve_type_float32_t_ptr][__ARM_mve_type_float32x4_t]: __arm_vstrwq_scatter_shifted_offset_f32 (__ARM_mve_coerce_f32_ptr(p0, float32_t *), __p1, __ARM_mve_coerce(__p2, float32x4_t)));})
-
#define __arm_vuninitializedq(p0) ({ __typeof(p0) __p0 = (p0); \
_Generic( (int (*)[__ARM_mve_typeid(__p0)])0, \
int (*)[__ARM_mve_type_int8x16_t]: __arm_vuninitializedq_s8 (), \
@@ -3327,22 +2957,6 @@ extern void *__ARM_undef;
int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4x2_t]: __arm_vst2q_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __ARM_mve_coerce(__p1, uint32x4x2_t)));})
-#define __arm_vstrhq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t), p3), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t), p3));})
-
-#define __arm_vstrhq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t)), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t)));})
-
#define __arm_vstrdq_scatter_base_p(p0,p1,p2,p3) ({ __typeof(p2) __p2 = (p2); \
_Generic( (int (*)[__ARM_mve_typeid(__p2)])0, \
int (*)[__ARM_mve_type_int64x2_t]: __arm_vstrdq_scatter_base_p_s64 (p0, p1, __ARM_mve_coerce(__p2, int64x2_t), p3), \
@@ -3353,34 +2967,6 @@ extern void *__ARM_undef;
int (*)[__ARM_mve_type_int64x2_t]: __arm_vstrdq_scatter_base_s64 (p0, p1, __ARM_mve_coerce(__p2, int64x2_t)), \
int (*)[__ARM_mve_type_uint64x2_t]: __arm_vstrdq_scatter_base_u64 (p0, p1, __ARM_mve_coerce(__p2, uint64x2_t)));})
-#define __arm_vstrhq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t)), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t)), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t)));})
-
-#define __arm_vstrhq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p1)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_int16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_s16 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, int16x8_t), p3), \
- int (*)[__ARM_mve_type_int16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_int32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s16_ptr(p0, int16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint16x8_t][__ARM_mve_type_uint16x8_t]: __arm_vstrhq_scatter_shifted_offset_p_u16 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint16x8_t), __ARM_mve_coerce(__p2, uint16x8_t), p3), \
- int (*)[__ARM_mve_type_uint16_t_ptr][__ARM_mve_type_uint32x4_t][__ARM_mve_type_uint32x4_t]: __arm_vstrhq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u16_ptr(p0, uint16_t *), __ARM_mve_coerce(__p1, uint32x4_t), __ARM_mve_coerce(__p2, uint32x4_t), p3));})
-
-#define __arm_vstrwq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int32_t_ptr][__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_shifted_offset_s32 (__ARM_mve_coerce_s32_ptr(p0, int32_t *), __p1, __ARM_mve_coerce(__p2, int32x4_t)), \
- int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_shifted_offset_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __p1, __ARM_mve_coerce(__p2, uint32x4_t)));})
-
-#define __arm_vstrwq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p1) __p1 = (p1); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int32_t_ptr][__ARM_mve_type_int32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_s32 (__ARM_mve_coerce_s32_ptr(p0, int32_t *), __p1, __ARM_mve_coerce(__p2, int32x4_t), p3), \
- int (*)[__ARM_mve_type_uint32_t_ptr][__ARM_mve_type_uint32x4_t]: __arm_vstrwq_scatter_shifted_offset_p_u32 (__ARM_mve_coerce_u32_ptr(p0, uint32_t *), __p1, __ARM_mve_coerce(__p2, uint32x4_t), p3));})
-
#define __arm_vuninitializedq(p0) ({ __typeof(p0) __p0 = (p0); \
_Generic( (int (*)[__ARM_mve_typeid(__p0)])0, \
int (*)[__ARM_mve_type_int8x16_t]: __arm_vuninitializedq_s8 (), \
@@ -3489,18 +3075,6 @@ extern void *__ARM_undef;
int (*)[__ARM_mve_type_int64x2_t]: __arm_vstrdq_scatter_base_p_s64 (p0, p1, __ARM_mve_coerce(__p2, int64x2_t), p3), \
int (*)[__ARM_mve_type_uint64x2_t]: __arm_vstrdq_scatter_base_p_u64 (p0, p1, __ARM_mve_coerce(__p2, uint64x2_t), p3));})
-#define __arm_vstrdq_scatter_shifted_offset_p(p0,p1,p2,p3) ({ __typeof(p0) __p0 = (p0); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(__p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int64_t_ptr][__ARM_mve_type_int64x2_t]: __arm_vstrdq_scatter_shifted_offset_p_s64 (__ARM_mve_coerce_s64_ptr(__p0, int64_t *), p1, __ARM_mve_coerce(__p2, int64x2_t), p3), \
- int (*)[__ARM_mve_type_uint64_t_ptr][__ARM_mve_type_uint64x2_t]: __arm_vstrdq_scatter_shifted_offset_p_u64 (__ARM_mve_coerce_u64_ptr(__p0, uint64_t *), p1, __ARM_mve_coerce(__p2, uint64x2_t), p3));})
-
-#define __arm_vstrdq_scatter_shifted_offset(p0,p1,p2) ({ __typeof(p0) __p0 = (p0); \
- __typeof(p2) __p2 = (p2); \
- _Generic( (int (*)[__ARM_mve_typeid(__p0)][__ARM_mve_typeid(__p2)])0, \
- int (*)[__ARM_mve_type_int64_t_ptr][__ARM_mve_type_int64x2_t]: __arm_vstrdq_scatter_shifted_offset_s64 (__ARM_mve_coerce_s64_ptr(__p0, int64_t *), p1, __ARM_mve_coerce(__p2, int64x2_t)), \
- int (*)[__ARM_mve_type_uint64_t_ptr][__ARM_mve_type_uint64x2_t]: __arm_vstrdq_scatter_shifted_offset_u64 (__ARM_mve_coerce_u64_ptr(__p0, uint64_t *), p1, __ARM_mve_coerce(__p2, uint64x2_t)));})
-
#endif /* __cplusplus */
#endif /* __ARM_FEATURE_MVE */
#endif /* _GCC_ARM_MVE_H. */
@@ -713,28 +713,12 @@ VAR1 (LDRGU_Z, vldrdq_gather_offset_z_u, v2di)
VAR1 (LDRGU_Z, vldrdq_gather_shifted_offset_z_u, v2di)
VAR1 (LDRGU_Z, vldrwq_gather_offset_z_u, v4si)
VAR1 (LDRGU_Z, vldrwq_gather_shifted_offset_z_u, v4si)
-VAR2 (STRSU_P, vstrhq_scatter_shifted_offset_p_u, v8hi, v4si)
-VAR2 (STRSU, vstrhq_scatter_shifted_offset_u, v8hi, v4si)
-VAR2 (STRSS_P, vstrhq_scatter_shifted_offset_p_s, v8hi, v4si)
-VAR2 (STRSS, vstrhq_scatter_shifted_offset_s, v8hi, v4si)
VAR1 (STRSBS, vstrdq_scatter_base_s, v2di)
VAR1 (STRSBS, vstrwq_scatter_base_f, v4sf)
VAR1 (STRSBS_P, vstrdq_scatter_base_p_s, v2di)
VAR1 (STRSBS_P, vstrwq_scatter_base_p_f, v4sf)
VAR1 (STRSBU, vstrdq_scatter_base_u, v2di)
VAR1 (STRSBU_P, vstrdq_scatter_base_p_u, v2di)
-VAR1 (STRSS, vstrdq_scatter_shifted_offset_s, v2di)
-VAR1 (STRSS, vstrhq_scatter_shifted_offset_f, v8hf)
-VAR1 (STRSS, vstrwq_scatter_shifted_offset_f, v4sf)
-VAR1 (STRSS, vstrwq_scatter_shifted_offset_s, v4si)
-VAR1 (STRSS_P, vstrdq_scatter_shifted_offset_p_s, v2di)
-VAR1 (STRSS_P, vstrhq_scatter_shifted_offset_p_f, v8hf)
-VAR1 (STRSS_P, vstrwq_scatter_shifted_offset_p_f, v4sf)
-VAR1 (STRSS_P, vstrwq_scatter_shifted_offset_p_s, v4si)
-VAR1 (STRSU, vstrdq_scatter_shifted_offset_u, v2di)
-VAR1 (STRSU, vstrwq_scatter_shifted_offset_u, v4si)
-VAR1 (STRSU_P, vstrdq_scatter_shifted_offset_p_u, v2di)
-VAR1 (STRSU_P, vstrwq_scatter_shifted_offset_p_u, v4si)
VAR1 (STRSBWBU, vstrwq_scatter_base_wb_u, v4si)
VAR1 (STRSBWBU, vstrdq_scatter_base_wb_u, v2di)
VAR1 (STRSBWBU_P, vstrwq_scatter_base_wb_p_u, v4si)
@@ -274,6 +274,7 @@ (define_mode_iterator MVE_types [V16QI V8HI V4SI V2DI TI V8HF V4SF V2DF])
(define_mode_iterator MVE_vecs [V16QI V8HI V4SI V2DI V8HF V4SF V2DF])
(define_mode_iterator MVE_VLD_ST [V16QI V8HI V4SI V8HF V4SF])
(define_mode_iterator MVE_VLD_ST_scatter [V16QI V8HI V4SI V8HF V4SF V2DI])
+(define_mode_iterator MVE_VLD_ST_scatter_shifted [V8HI V4SI V8HF V4SF V2DI])
(define_mode_iterator MVE_0 [V8HF V4SF])
(define_mode_iterator MVE_1 [V16QI V8HI V4SI V2DI])
(define_mode_iterator MVE_3 [V16QI V8HI])
@@ -292,6 +293,7 @@ (define_mode_attr MVE_wide_n_type [(V8QI "v8hi") (V4QI "v4si") (V4HI "v4si")])
(define_mode_attr MVE_wide_n_sz_elem [(V8QI "16") (V4QI "32") (V4HI "32")])
(define_mode_attr MVE_wide_n_VPRED [(V8QI "V8BI") (V4QI "V4BI") (V4HI "V4BI")])
(define_mode_attr MVE_scatter_offset [(V16QI "V16QI") (V8HI "V8HI") (V4SI "V4SI") (V8HF "V8HI") (V4SF "V4SI") (V2DI "V2DI")])
+(define_mode_attr MVE_scatter_shift [(V8HI "1") (V4SI "2") (V8HF "1") (V4SF "2") (V2DI "3")])
;;----------------------------------------------------------------------------
;; Code iterators
@@ -2532,11 +2534,8 @@ (define_int_attr supf [(VCVTQ_TO_F_S "s") (VCVTQ_TO_F_U "u") (VREV16Q_S "s")
(VLDRDQGO_S "s") (VLDRDQGO_U "u") (VLDRDQGSO_S "s")
(VLDRDQGSO_U "u") (VLDRWQGO_S "s") (VLDRWQGO_U "u")
(VLDRWQGSO_S "s") (VLDRWQGSO_U "u")
- (VSTRHQSSO_S "s") (VSTRHQSSO_U "u")
(VSTRDQSB_S "s") (VSTRDQSB_U "u")
- (VSTRDQSSO_S "s") (VSTRDQSSO_U "u")
- (VSTRWQSSO_U "u")
- (VSTRWQSSO_S "s") (VSTRWQSBWB_S "s") (VSTRWQSBWB_U "u")
+ (VSTRWQSBWB_S "s") (VSTRWQSBWB_U "u")
(VLDRWQGBWB_S "s") (VLDRWQGBWB_U "u") (VLDRDQGBWB_S "s")
(VLDRDQGBWB_U "u") (VSTRDQSBWB_S "s") (VADCQ_M_S "s")
(VSTRDQSBWB_U "u") (VSBCQ_U "u") (VSBCQ_M_U "u")
@@ -2947,10 +2946,7 @@ (define_int_iterator VLDRDGOQ [VLDRDQGO_S VLDRDQGO_U])
(define_int_iterator VLDRDGSOQ [VLDRDQGSO_S VLDRDQGSO_U])
(define_int_iterator VLDRWGOQ [VLDRWQGO_S VLDRWQGO_U])
(define_int_iterator VLDRWGSOQ [VLDRWQGSO_S VLDRWQGSO_U])
-(define_int_iterator VSTRHSSOQ [VSTRHQSSO_S VSTRHQSSO_U])
(define_int_iterator VSTRDSBQ [VSTRDQSB_S VSTRDQSB_U])
-(define_int_iterator VSTRDSSOQ [VSTRDQSSO_S VSTRDQSSO_U])
-(define_int_iterator VSTRWSSOQ [VSTRWQSSO_S VSTRWQSSO_U])
(define_int_iterator VSTRWSBWBQ [VSTRWQSBWB_S VSTRWQSBWB_U])
(define_int_iterator VLDRWGBWBQ [VLDRWQGBWB_S VLDRWQGBWB_U])
(define_int_iterator VSTRDSBWBQ [VSTRDQSBWB_S VSTRDQSBWB_U])
@@ -4072,68 +4072,80 @@ (define_insn "mve_vldrwq_gather_shifted_offset_z_<supf>v4si"
[(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vldrwq_gather_shifted_offset_<supf>v4si"))
(set_attr "length" "8")])
+;; Vector scatter stores with shifted offset
;;
-;; [vstrhq_scatter_shifted_offset_p_s vstrhq_scatter_shifted_offset_p_u]
+;; [vstrhq_scatter_shifted_offset_s vstrhq_scatter_shifted_offset_u]
+;; [vstrhq_scatter_shifted_offset_f]
+;; [vstrwq_scatter_shifted_offset_s vstrwq_scatter_shifted_offset_u]
+;; [vstrwq_scatter_shifted_offset_f]
+;; [vstrdq_scatter_shifted_offset_s vstrdq_scatter_shifted_offset_u]
+(define_insn "@mve_vstrq_scatter_shifted_offset_<mode>"
+ [(set (mem:BLK (scratch))
+ (unspec:BLK
+ [(match_operand:SI 0 "register_operand" "r")
+ (match_operand:<MVE_scatter_offset> 1 "s_register_operand" "w")
+ (match_operand:MVE_VLD_ST_scatter_shifted 2 "s_register_operand" "w")]
+ VSTRSSOQ))]
+ "(TARGET_HAVE_MVE && VALID_MVE_SI_MODE (<MODE>mode))
+ || (TARGET_HAVE_MVE_FLOAT && VALID_MVE_SF_MODE (<MODE>mode))"
+ "vstr<MVE_elem_ch>.<V_sz_elem>\t%q2, [%0, %q1, uxtw #<MVE_scatter_shift>]"
+ [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrq_scatter_shifted_offset_<mode>"))
+ (set_attr "length" "4")])
+
+;; Truncating vector scatter stores with shifted offset
;;
-(define_expand "mve_vstrhq_scatter_shifted_offset_p_<supf><mode>"
- [(match_operand:<MVE_H_ELEM> 0 "mve_scatter_memory")
- (match_operand:MVE_5 1 "s_register_operand")
- (match_operand:MVE_5 2 "s_register_operand")
- (match_operand:<MVE_VPRED> 3 "vpr_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRHSSOQ)]
+;; [vstrhq_scatter_shifted_offset_s32 vstrhq_scatter_shifted_offset_u32]
+(define_insn "mve_vstrq_truncate_scatter_shifted_offset_v4si"
+ [(set (mem:BLK (scratch))
+ (unspec:BLK
+ [(match_operand:SI 0 "register_operand" "r")
+ (match_operand:V4SI 1 "s_register_operand" "w")
+ (truncate:V4HI
+ (match_operand:V4SI 2 "s_register_operand" "w"))]
+ VSTRSSOQ_TRUNC))]
"TARGET_HAVE_MVE"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrhq_scatter_shifted_offset_p_<supf><mode>_insn (ind, operands[1],
- operands[2],
- operands[3]));
- DONE;
-})
+ "vstrh.32\t%q2, [%0, %q1, uxtw #1]"
+ [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrq_truncate_scatter_shifted_offset_v4si"))
+ (set_attr "length" "4")])
-(define_insn "mve_vstrhq_scatter_shifted_offset_p_<supf><mode>_insn"
+;; Predicated vector scatter stores with shifted offset
+;;
+;; [vstrhq_scatter_shifted_offset_p_s vstrhq_scatter_shifted_offset_p_u]
+;; [vstrhq_scatter_shifted_offset_p_f]
+;; [vstrwq_scatter_shifted_offset_p_s vstrwq_scatter_shifted_offset_p_u]
+;; [vstrwq_scatter_shifted_offset_p_f]
+;; [vstrdq_scatter_shifted_offset_p_s vstrdq_scatter_shifted_offset_p_u]
+;;
+(define_insn "@mve_vstrq_scatter_shifted_offset_p_<mode>"
[(set (mem:BLK (scratch))
(unspec:BLK
[(match_operand:SI 0 "register_operand" "r")
- (match_operand:MVE_5 1 "s_register_operand" "w")
- (match_operand:MVE_5 2 "s_register_operand" "w")
+ (match_operand:<MVE_scatter_offset> 1 "s_register_operand" "w")
+ (match_operand:MVE_VLD_ST_scatter_shifted 2 "s_register_operand" "w")
(match_operand:<MVE_VPRED> 3 "vpr_register_operand" "Up")]
- VSTRHSSOQ))]
- "TARGET_HAVE_MVE"
- "vpst\;vstrht.<V_sz_elem>\t%q2, [%0, %q1, uxtw #1]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrhq_scatter_shifted_offset_<supf><mode>_insn"))
+ VSTRSSOQ_P))]
+ "(TARGET_HAVE_MVE && VALID_MVE_SI_MODE (<MODE>mode))
+ || (TARGET_HAVE_MVE_FLOAT && VALID_MVE_SF_MODE (<MODE>mode))"
+ "vpst\;vstr<MVE_elem_ch>t.<V_sz_elem>\t%q2, [%0, %q1, uxtw #<MVE_scatter_shift>]"
+ [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrq_scatter_shifted_offset_<mode>"))
(set_attr "length" "8")])
+;; Predicated truncating vector scatter stores with shifted offset
;;
-;; [vstrhq_scatter_shifted_offset_s vstrhq_scatter_shifted_offset_u]
-;;
-(define_expand "mve_vstrhq_scatter_shifted_offset_<supf><mode>"
- [(match_operand:<MVE_H_ELEM> 0 "mve_scatter_memory")
- (match_operand:MVE_5 1 "s_register_operand")
- (match_operand:MVE_5 2 "s_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRHSSOQ)]
- "TARGET_HAVE_MVE"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrhq_scatter_shifted_offset_<supf><mode>_insn (ind, operands[1],
- operands[2]));
- DONE;
-})
-
-(define_insn "mve_vstrhq_scatter_shifted_offset_<supf><mode>_insn"
+;; [vstrhq_scatter_shifted_offset_p_s32 vstrhq_scatter_shifted_offset_p_u32]
+(define_insn "mve_vstrq_truncate_scatter_shifted_offset_p_v4si"
[(set (mem:BLK (scratch))
(unspec:BLK
[(match_operand:SI 0 "register_operand" "r")
- (match_operand:MVE_5 1 "s_register_operand" "w")
- (match_operand:MVE_5 2 "s_register_operand" "w")]
- VSTRHSSOQ))]
+ (match_operand:V4SI 1 "s_register_operand" "w")
+ (truncate:V4HI
+ (match_operand:V4SI 2 "s_register_operand" "w"))
+ (match_operand:<MVE_VPRED> 3 "vpr_register_operand" "Up")]
+ VSTRSSOQ_TRUNC_P))]
"TARGET_HAVE_MVE"
- "vstrh.<V_sz_elem>\t%q2, [%0, %q1, uxtw #1]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrhq_scatter_shifted_offset_<supf><mode>_insn"))
- (set_attr "length" "4")])
+ "vpst\;vstrht.32\t%q2, [%0, %q1, uxtw #1]"
+ [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrq_truncate_scatter_shifted_offset_v4si"))
+ (set_attr "length" "8")])
;;
;; [vstrdq_scatter_base_p_s vstrdq_scatter_base_p_u]
@@ -4182,131 +4194,6 @@ (define_insn "mve_vstrdq_scatter_base_<supf>v2di"
[(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrdq_scatter_base_<supf>v2di"))
(set_attr "length" "4")])
-;;
-;; [vstrdq_scatter_shifted_offset_p_s vstrdq_scatter_shifted_offset_p_u]
-;;
-(define_expand "mve_vstrdq_scatter_shifted_offset_p_<supf>v2di"
- [(match_operand:V2DI 0 "mve_scatter_memory")
- (match_operand:V2DI 1 "s_register_operand")
- (match_operand:V2DI 2 "s_register_operand")
- (match_operand:V2QI 3 "vpr_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRDSSOQ)]
- "TARGET_HAVE_MVE"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrdq_scatter_shifted_offset_p_<supf>v2di_insn (ind, operands[1],
- operands[2],
- operands[3]));
- DONE;
-})
-
-(define_insn "mve_vstrdq_scatter_shifted_offset_p_<supf>v2di_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V2DI 1 "s_register_operand" "w")
- (match_operand:V2DI 2 "s_register_operand" "w")
- (match_operand:V2QI 3 "vpr_register_operand" "Up")]
- VSTRDSSOQ))]
- "TARGET_HAVE_MVE"
- "vpst\;vstrdt.64\t%q2, [%0, %q1, uxtw #3]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrdq_scatter_shifted_offset_<supf>v2di_insn"))
- (set_attr "length" "8")])
-
-;;
-;; [vstrdq_scatter_shifted_offset_s vstrdq_scatter_shifted_offset_u]
-;;
-(define_expand "mve_vstrdq_scatter_shifted_offset_<supf>v2di"
- [(match_operand:V2DI 0 "mve_scatter_memory")
- (match_operand:V2DI 1 "s_register_operand")
- (match_operand:V2DI 2 "s_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRDSSOQ)]
- "TARGET_HAVE_MVE"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrdq_scatter_shifted_offset_<supf>v2di_insn (ind, operands[1],
- operands[2]));
- DONE;
-})
-
-(define_insn "mve_vstrdq_scatter_shifted_offset_<supf>v2di_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V2DI 1 "s_register_operand" "w")
- (match_operand:V2DI 2 "s_register_operand" "w")]
- VSTRDSSOQ))]
- "TARGET_HAVE_MVE"
- "vstrd.64\t%q2, [%0, %q1, uxtw #3]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrdq_scatter_shifted_offset_<supf>v2di_insn"))
- (set_attr "length" "4")])
-
-;;
-;; [vstrhq_scatter_shifted_offset_f]
-;;
-(define_expand "mve_vstrhq_scatter_shifted_offset_fv8hf"
- [(match_operand:V8HI 0 "memory_operand" "=Us")
- (match_operand:V8HI 1 "s_register_operand" "w")
- (match_operand:V8HF 2 "s_register_operand" "w")
- (unspec:V4SI [(const_int 0)] VSTRHQSSO_F)]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (gen_mve_vstrhq_scatter_shifted_offset_fv8hf_insn (ind, operands[1],
- operands[2]));
- DONE;
-})
-
-(define_insn "mve_vstrhq_scatter_shifted_offset_fv8hf_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V8HI 1 "s_register_operand" "w")
- (match_operand:V8HF 2 "s_register_operand" "w")]
- VSTRHQSSO_F))]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
- "vstrh.16\t%q2, [%0, %q1, uxtw #1]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrhq_scatter_shifted_offset_fv8hf_insn"))
- (set_attr "length" "4")])
-
-;;
-;; [vstrhq_scatter_shifted_offset_p_f]
-;;
-(define_expand "mve_vstrhq_scatter_shifted_offset_p_fv8hf"
- [(match_operand:V8HI 0 "memory_operand" "=Us")
- (match_operand:V8HI 1 "s_register_operand" "w")
- (match_operand:V8HF 2 "s_register_operand" "w")
- (match_operand:V8BI 3 "vpr_register_operand" "Up")
- (unspec:V4SI [(const_int 0)] VSTRHQSSO_F)]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrhq_scatter_shifted_offset_p_fv8hf_insn (ind, operands[1],
- operands[2],
- operands[3]));
- DONE;
-})
-
-(define_insn "mve_vstrhq_scatter_shifted_offset_p_fv8hf_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V8HI 1 "s_register_operand" "w")
- (match_operand:V8HF 2 "s_register_operand" "w")
- (match_operand:V8BI 3 "vpr_register_operand" "Up")]
- VSTRHQSSO_F))]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
- "vpst\;vstrht.16\t%q2, [%0, %q1, uxtw #1]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrhq_scatter_shifted_offset_fv8hf_insn"))
- (set_attr "length" "8")])
-
;;
;; [vstrwq_scatter_base_f]
;;
@@ -4354,131 +4241,6 @@ (define_insn "mve_vstrwq_scatter_base_p_fv4sf"
[(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrwq_scatter_base_fv4sf"))
(set_attr "length" "8")])
-;;
-;; [vstrwq_scatter_shifted_offset_f]
-;;
-(define_expand "mve_vstrwq_scatter_shifted_offset_fv4sf"
- [(match_operand:V4SI 0 "mve_scatter_memory")
- (match_operand:V4SI 1 "s_register_operand")
- (match_operand:V4SF 2 "s_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRWQSSO_F)]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (gen_mve_vstrwq_scatter_shifted_offset_fv4sf_insn (ind, operands[1],
- operands[2]));
- DONE;
-})
-
-(define_insn "mve_vstrwq_scatter_shifted_offset_fv4sf_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V4SI 1 "s_register_operand" "w")
- (match_operand:V4SF 2 "s_register_operand" "w")]
- VSTRWQSSO_F))]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
- "vstrw.32\t%q2, [%0, %q1, uxtw #2]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrwq_scatter_shifted_offset_fv4sf_insn"))
- (set_attr "length" "8")])
-
-;;
-;; [vstrwq_scatter_shifted_offset_p_f]
-;;
-(define_expand "mve_vstrwq_scatter_shifted_offset_p_fv4sf"
- [(match_operand:V4SI 0 "mve_scatter_memory")
- (match_operand:V4SI 1 "s_register_operand")
- (match_operand:V4SF 2 "s_register_operand")
- (match_operand:V4BI 3 "vpr_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRWQSSO_F)]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrwq_scatter_shifted_offset_p_fv4sf_insn (ind, operands[1],
- operands[2],
- operands[3]));
- DONE;
-})
-
-(define_insn "mve_vstrwq_scatter_shifted_offset_p_fv4sf_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V4SI 1 "s_register_operand" "w")
- (match_operand:V4SF 2 "s_register_operand" "w")
- (match_operand:V4BI 3 "vpr_register_operand" "Up")]
- VSTRWQSSO_F))]
- "TARGET_HAVE_MVE && TARGET_HAVE_MVE_FLOAT"
- "vpst\;vstrwt.32\t%q2, [%0, %q1, uxtw #2]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrwq_scatter_shifted_offset_fv4sf_insn"))
- (set_attr "length" "8")])
-
-;;
-;; [vstrwq_scatter_shifted_offset_p_s vstrwq_scatter_shifted_offset_p_u]
-;;
-(define_expand "mve_vstrwq_scatter_shifted_offset_p_<supf>v4si"
- [(match_operand:V4SI 0 "mve_scatter_memory")
- (match_operand:V4SI 1 "s_register_operand")
- (match_operand:V4SI 2 "s_register_operand")
- (match_operand:V4BI 3 "vpr_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRWSSOQ)]
- "TARGET_HAVE_MVE"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrwq_scatter_shifted_offset_p_<supf>v4si_insn (ind, operands[1],
- operands[2],
- operands[3]));
- DONE;
-})
-
-(define_insn "mve_vstrwq_scatter_shifted_offset_p_<supf>v4si_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V4SI 1 "s_register_operand" "w")
- (match_operand:V4SI 2 "s_register_operand" "w")
- (match_operand:V4BI 3 "vpr_register_operand" "Up")]
- VSTRWSSOQ))]
- "TARGET_HAVE_MVE"
- "vpst\;vstrwt.32\t%q2, [%0, %q1, uxtw #2]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrwq_scatter_shifted_offset_<supf>v4si_insn"))
- (set_attr "length" "8")])
-
-;;
-;; [vstrwq_scatter_shifted_offset_s vstrwq_scatter_shifted_offset_u]
-;;
-(define_expand "mve_vstrwq_scatter_shifted_offset_<supf>v4si"
- [(match_operand:V4SI 0 "mve_scatter_memory")
- (match_operand:V4SI 1 "s_register_operand")
- (match_operand:V4SI 2 "s_register_operand")
- (unspec:V4SI [(const_int 0)] VSTRWSSOQ)]
- "TARGET_HAVE_MVE"
-{
- rtx ind = XEXP (operands[0], 0);
- gcc_assert (REG_P (ind));
- emit_insn (
- gen_mve_vstrwq_scatter_shifted_offset_<supf>v4si_insn (ind, operands[1],
- operands[2]));
- DONE;
-})
-
-(define_insn "mve_vstrwq_scatter_shifted_offset_<supf>v4si_insn"
- [(set (mem:BLK (scratch))
- (unspec:BLK
- [(match_operand:SI 0 "register_operand" "r")
- (match_operand:V4SI 1 "s_register_operand" "w")
- (match_operand:V4SI 2 "s_register_operand" "w")]
- VSTRWSSOQ))]
- "TARGET_HAVE_MVE"
- "vstrw.32\t%q2, [%0, %q1, uxtw #2]"
- [(set (attr "mve_unpredicated_insn") (symbol_ref "CODE_FOR_mve_vstrwq_scatter_shifted_offset_<supf>v4si_insn"))
- (set_attr "length" "4")])
-
;;
;;
;; [vddupq_u_insn, vidupq_u_insn]
@@ -1185,13 +1185,11 @@ (define_c_enum "unspec" [
VSTRQ_TRUNC_P
VSTRDQSB_S
VSTRDQSB_U
- VSTRDQSSO_S
- VSTRDQSSO_U
- VSTRWQSSO_S
- VSTRWQSSO_U
- VSTRHQSSO_F
+ VSTRSSOQ
+ VSTRSSOQ_P
+ VSTRSSOQ_TRUNC
+ VSTRSSOQ_TRUNC_P
VSTRWQSB_F
- VSTRWQSSO_F
VDDUPQ
VDDUPQ_M
VDWDUPQ
@@ -1231,8 +1229,6 @@ (define_c_enum "unspec" [
VST2Q
VSHLCQ_M_U
VSHLCQ_M_S
- VSTRHQSSO_S
- VSTRHQSSO_U
VSTRHQ_S
SRSHRL
SRSHR