diff mbox series

[17/20] arm: [MVE intrinsics] rework vaddlvq

Message ID 20230510133036.596530-17-christophe.lyon@arm.com
State Committed
Commit 218e01eeca8cdb596aa59a4652f17437d0dbb3ab
Headers
Series [01/20] arm: [MVE intrinsics] factorize vcmp |

Commit Message

Christophe Lyon May 10, 2023, 1:30 p.m. UTC 
  Implement vaddlvq using the new MVE builtins framework.

Since we kept v4si hardcoded in the builtin name, we need to
special-case it in unspec_mve_function_exact_insn_pred_p.

2022-10-25  Christophe Lyon  <christophe.lyon@arm.com>

	gcc/
	* config/arm/arm-mve-builtins-base.cc (vaddlvq): New.
	* config/arm/arm-mve-builtins-base.def (vaddlvq): New.
	* config/arm/arm-mve-builtins-base.h (vaddlvq): New.
	* config/arm/arm-mve-builtins-functions.h
	(unspec_mve_function_exact_insn_pred_p): Handle vaddlvq.
	* config/arm/arm_mve.h (vaddlvq): Remove.
	(vaddlvq_p): Remove.
	(vaddlvq_s32): Remove.
	(vaddlvq_u32): Remove.
	(vaddlvq_p_s32): Remove.
	(vaddlvq_p_u32): Remove.
	(__arm_vaddlvq_s32): Remove.
	(__arm_vaddlvq_u32): Remove.
	(__arm_vaddlvq_p_s32): Remove.
	(__arm_vaddlvq_p_u32): Remove.
	(__arm_vaddlvq): Remove.
	(__arm_vaddlvq_p): Remove.
---
 gcc/config/arm/arm-mve-builtins-base.cc     |  1 +
 gcc/config/arm/arm-mve-builtins-base.def    |  1 +
 gcc/config/arm/arm-mve-builtins-base.h      |  1 +
 gcc/config/arm/arm-mve-builtins-functions.h | 69 ++++++++++++++------
 gcc/config/arm/arm_mve.h                    | 72 ---------------------
 5 files changed, 51 insertions(+), 93 deletions(-)
diff mbox series

Patch

diff --git a/gcc/config/arm/arm-mve-builtins-base.cc b/gcc/config/arm/arm-mve-builtins-base.cc
index e87069b0467..fdc0ff50b96 100644
--- a/gcc/config/arm/arm-mve-builtins-base.cc
+++ b/gcc/config/arm/arm-mve-builtins-base.cc
@@ -244,6 +244,7 @@  namespace arm_mve {
 FUNCTION_WITHOUT_N (vabdq, VABDQ)
 FUNCTION (vabsq, unspec_based_mve_function_exact_insn, (ABS, ABS, ABS, -1, -1, -1, VABSQ_M_S, -1, VABSQ_M_F, -1, -1, -1))
 FUNCTION_WITH_RTX_M_N (vaddq, PLUS, VADDQ)
+FUNCTION_PRED_P_S_U (vaddlvq, VADDLVQ)
 FUNCTION_PRED_P_S_U (vaddvq, VADDVQ)
 FUNCTION_PRED_P_S_U (vaddvaq, VADDVAQ)
 FUNCTION_WITH_RTX_M (vandq, AND, VANDQ)
diff --git a/gcc/config/arm/arm-mve-builtins-base.def b/gcc/config/arm/arm-mve-builtins-base.def
index 413fe4a1ef0..dcfb426a7fb 100644
--- a/gcc/config/arm/arm-mve-builtins-base.def
+++ b/gcc/config/arm/arm-mve-builtins-base.def
@@ -20,6 +20,7 @@ 
 #define REQUIRES_FLOAT false
 DEF_MVE_FUNCTION (vabdq, binary, all_integer, mx_or_none)
 DEF_MVE_FUNCTION (vabsq, unary, all_signed, mx_or_none)
+DEF_MVE_FUNCTION (vaddlvq, unary_acc, integer_32, p_or_none)
 DEF_MVE_FUNCTION (vaddq, binary_opt_n, all_integer, mx_or_none)
 DEF_MVE_FUNCTION (vaddvaq, unary_int32_acc, all_integer, p_or_none)
 DEF_MVE_FUNCTION (vaddvq, unary_int32, all_integer, p_or_none)
diff --git a/gcc/config/arm/arm-mve-builtins-base.h b/gcc/config/arm/arm-mve-builtins-base.h
index 5338b777444..5de70d5e1d4 100644
--- a/gcc/config/arm/arm-mve-builtins-base.h
+++ b/gcc/config/arm/arm-mve-builtins-base.h
@@ -25,6 +25,7 @@  namespace functions {
 
 extern const function_base *const vabdq;
 extern const function_base *const vabsq;
+extern const function_base *const vaddlvq;
 extern const function_base *const vaddq;
 extern const function_base *const vaddvaq;
 extern const function_base *const vaddvq;
diff --git a/gcc/config/arm/arm-mve-builtins-functions.h b/gcc/config/arm/arm-mve-builtins-functions.h
index d069990dcab..ea926e42b81 100644
--- a/gcc/config/arm/arm-mve-builtins-functions.h
+++ b/gcc/config/arm/arm-mve-builtins-functions.h
@@ -408,32 +408,59 @@  public:
   expand (function_expander &e) const override
   {
     insn_code code;
-    switch (e.pred)
+
+    if ((m_unspec_for_sint == VADDLVQ_S)
+	|| m_unspec_for_sint == VADDLVAQ_S)
       {
-      case PRED_none:
-	if (e.type_suffix (0).integer_p)
-	  if (e.type_suffix (0).unsigned_p)
-	    code = code_for_mve_q (m_unspec_for_uint, m_unspec_for_uint, e.vector_mode (0));
-	  else
-	    code = code_for_mve_q (m_unspec_for_sint, m_unspec_for_sint, e.vector_mode (0));
-	else
-	  code = code_for_mve_q_f (m_unspec_for_fp, e.vector_mode (0));
+	switch (e.pred)
+	  {
+	  case PRED_none:
+	    if (e.type_suffix (0).unsigned_p)
+	      code = code_for_mve_q_v4si (m_unspec_for_uint, m_unspec_for_uint);
+	    else
+	      code = code_for_mve_q_v4si (m_unspec_for_sint, m_unspec_for_sint);
+	    return e.use_exact_insn (code);
 
-	return e.use_exact_insn (code);
+	  case PRED_p:
+	    if (e.type_suffix (0).unsigned_p)
+	      code = code_for_mve_q_p_v4si (m_unspec_for_p_uint, m_unspec_for_p_uint);
+	    else
+	      code = code_for_mve_q_p_v4si (m_unspec_for_p_sint, m_unspec_for_p_sint);
+	    return e.use_exact_insn (code);
 
-      case PRED_p:
-	if (e.type_suffix (0).integer_p)
-	  if (e.type_suffix (0).unsigned_p)
-	    code = code_for_mve_q_p (m_unspec_for_p_uint, m_unspec_for_p_uint, e.vector_mode (0));
-	  else
-	    code = code_for_mve_q_p (m_unspec_for_p_sint, m_unspec_for_p_sint, e.vector_mode (0));
-	else
-	  code = code_for_mve_q_p_f (m_unspec_for_p_fp, e.vector_mode (0));
+	  default:
+	    gcc_unreachable ();
+	  }
+      }
+    else
+      {
+	switch (e.pred)
+	  {
+	  case PRED_none:
+	    if (e.type_suffix (0).integer_p)
+	      if (e.type_suffix (0).unsigned_p)
+		code = code_for_mve_q (m_unspec_for_uint, m_unspec_for_uint, e.vector_mode (0));
+	      else
+		code = code_for_mve_q (m_unspec_for_sint, m_unspec_for_sint, e.vector_mode (0));
+	    else
+	      code = code_for_mve_q_f (m_unspec_for_fp, e.vector_mode (0));
 
-	return e.use_exact_insn (code);
+	    return e.use_exact_insn (code);
 
-      default:
-	gcc_unreachable ();
+	  case PRED_p:
+	    if (e.type_suffix (0).integer_p)
+	      if (e.type_suffix (0).unsigned_p)
+		code = code_for_mve_q_p (m_unspec_for_p_uint, m_unspec_for_p_uint, e.vector_mode (0));
+	      else
+		code = code_for_mve_q_p (m_unspec_for_p_sint, m_unspec_for_p_sint, e.vector_mode (0));
+	    else
+	      code = code_for_mve_q_p_f (m_unspec_for_p_fp, e.vector_mode (0));
+
+	    return e.use_exact_insn (code);
+
+	  default:
+	    gcc_unreachable ();
+	  }
       }
 
     gcc_unreachable ();
diff --git a/gcc/config/arm/arm_mve.h b/gcc/config/arm/arm_mve.h
index 74783570561..21d7768a732 100644
--- a/gcc/config/arm/arm_mve.h
+++ b/gcc/config/arm/arm_mve.h
@@ -42,11 +42,9 @@ 
 
 #ifndef __ARM_MVE_PRESERVE_USER_NAMESPACE
 #define vst4q(__addr, __value) __arm_vst4q(__addr, __value)
-#define vaddlvq(__a) __arm_vaddlvq(__a)
 #define vmovlbq(__a) __arm_vmovlbq(__a)
 #define vmovltq(__a) __arm_vmovltq(__a)
 #define vmvnq(__a) __arm_vmvnq(__a)
-#define vaddlvq_p(__a, __p) __arm_vaddlvq_p(__a, __p)
 #define vornq(__a, __b) __arm_vornq(__a, __b)
 #define vmulltq_int(__a, __b) __arm_vmulltq_int(__a, __b)
 #define vmullbq_int(__a, __b) __arm_vmullbq_int(__a, __b)
@@ -324,7 +322,6 @@ 
 #define vcvtq_f32_s32(__a) __arm_vcvtq_f32_s32(__a)
 #define vcvtq_f16_u16(__a) __arm_vcvtq_f16_u16(__a)
 #define vcvtq_f32_u32(__a) __arm_vcvtq_f32_u32(__a)
-#define vaddlvq_s32(__a) __arm_vaddlvq_s32(__a)
 #define vmovlbq_s8(__a) __arm_vmovlbq_s8(__a)
 #define vmovlbq_s16(__a) __arm_vmovlbq_s16(__a)
 #define vmovltq_s8(__a) __arm_vmovltq_s8(__a)
@@ -353,7 +350,6 @@ 
 #define vmovlbq_u16(__a) __arm_vmovlbq_u16(__a)
 #define vmvnq_n_u16( __imm) __arm_vmvnq_n_u16( __imm)
 #define vmvnq_n_u32( __imm) __arm_vmvnq_n_u32( __imm)
-#define vaddlvq_u32(__a) __arm_vaddlvq_u32(__a)
 #define vcvtq_u16_f16(__a) __arm_vcvtq_u16_f16(__a)
 #define vcvtq_u32_f32(__a) __arm_vcvtq_u32_f32(__a)
 #define vcvtpq_u16_f16(__a) __arm_vcvtpq_u16_f16(__a)
@@ -379,8 +375,6 @@ 
 #define vcvtq_n_s32_f32(__a,  __imm6) __arm_vcvtq_n_s32_f32(__a,  __imm6)
 #define vcvtq_n_u16_f16(__a,  __imm6) __arm_vcvtq_n_u16_f16(__a,  __imm6)
 #define vcvtq_n_u32_f32(__a,  __imm6) __arm_vcvtq_n_u32_f32(__a,  __imm6)
-#define vaddlvq_p_s32(__a, __p) __arm_vaddlvq_p_s32(__a, __p)
-#define vaddlvq_p_u32(__a, __p) __arm_vaddlvq_p_u32(__a, __p)
 #define vornq_u8(__a, __b) __arm_vornq_u8(__a, __b)
 #define vmulltq_int_u8(__a, __b) __arm_vmulltq_int_u8(__a, __b)
 #define vmullbq_int_u8(__a, __b) __arm_vmullbq_int_u8(__a, __b)
@@ -1499,13 +1493,6 @@  __arm_vst4q_u32 (uint32_t * __addr, uint32x4x4_t __value)
   __builtin_mve_vst4qv4si ((__builtin_neon_si *) __addr, __rv.__o);
 }
 
-__extension__ extern __inline int64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq_s32 (int32x4_t __a)
-{
-  return __builtin_mve_vaddlvq_sv4si (__a);
-}
-
 __extension__ extern __inline int16x8_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 __arm_vmovlbq_s8 (int8x16_t __a)
@@ -1632,13 +1619,6 @@  __arm_vmvnq_n_u32 (const int __imm)
   return __builtin_mve_vmvnq_n_uv4si (__imm);
 }
 
-__extension__ extern __inline uint64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq_u32 (uint32x4_t __a)
-{
-  return __builtin_mve_vaddlvq_uv4si (__a);
-}
-
 __extension__ extern __inline mve_pred16_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 __arm_vctp16q (uint32_t __a)
@@ -1674,20 +1654,6 @@  __arm_vpnot (mve_pred16_t __a)
   return __builtin_mve_vpnotv16bi (__a);
 }
 
-__extension__ extern __inline int64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq_p_s32 (int32x4_t __a, mve_pred16_t __p)
-{
-  return __builtin_mve_vaddlvq_p_sv4si (__a, __p);
-}
-
-__extension__ extern __inline uint64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq_p_u32 (uint32x4_t __a, mve_pred16_t __p)
-{
-  return __builtin_mve_vaddlvq_p_uv4si (__a, __p);
-}
-
 __extension__ extern __inline uint8x16_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 __arm_vornq_u8 (uint8x16_t __a, uint8x16_t __b)
@@ -9586,13 +9552,6 @@  __arm_vst4q (uint32_t * __addr, uint32x4x4_t __value)
  __arm_vst4q_u32 (__addr, __value);
 }
 
-__extension__ extern __inline int64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq (int32x4_t __a)
-{
- return __arm_vaddlvq_s32 (__a);
-}
-
 __extension__ extern __inline int16x8_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 __arm_vmovlbq (int8x16_t __a)
@@ -9691,27 +9650,6 @@  __arm_vmovlbq (uint16x8_t __a)
  return __arm_vmovlbq_u16 (__a);
 }
 
-__extension__ extern __inline uint64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq (uint32x4_t __a)
-{
- return __arm_vaddlvq_u32 (__a);
-}
-
-__extension__ extern __inline int64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq_p (int32x4_t __a, mve_pred16_t __p)
-{
- return __arm_vaddlvq_p_s32 (__a, __p);
-}
-
-__extension__ extern __inline uint64_t
-__attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
-__arm_vaddlvq_p (uint32x4_t __a, mve_pred16_t __p)
-{
- return __arm_vaddlvq_p_u32 (__a, __p);
-}
-
 __extension__ extern __inline uint8x16_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 __arm_vornq (uint8x16_t __a, uint8x16_t __b)
@@ -19212,16 +19150,6 @@  extern void *__ARM_undef;
   int (*)[__ARM_mve_type_int_n][__ARM_mve_type_int32x4_t]: __arm_vaddlvaq_p_s32 (__ARM_mve_coerce3(p0, int), __ARM_mve_coerce(__p1, int32x4_t), p2), \
   int (*)[__ARM_mve_type_int_n][__ARM_mve_type_uint32x4_t]: __arm_vaddlvaq_p_u32 (__ARM_mve_coerce3(p0, int), __ARM_mve_coerce(__p1, uint32x4_t), p2));})
 
-#define __arm_vaddlvq(p0) ({ __typeof(p0) __p0 = (p0); \
-  _Generic( (int (*)[__ARM_mve_typeid(__p0)])0, \
-  int (*)[__ARM_mve_type_int32x4_t]: __arm_vaddlvq_s32 (__ARM_mve_coerce(__p0, int32x4_t)), \
-  int (*)[__ARM_mve_type_uint32x4_t]: __arm_vaddlvq_u32 (__ARM_mve_coerce(__p0, uint32x4_t)));})
-
-#define __arm_vaddlvq_p(p0,p1) ({ __typeof(p0) __p0 = (p0); \
-  _Generic( (int (*)[__ARM_mve_typeid(__p0)])0, \
-  int (*)[__ARM_mve_type_int32x4_t]: __arm_vaddlvq_p_s32 (__ARM_mve_coerce(__p0, int32x4_t), p1), \
-  int (*)[__ARM_mve_type_uint32x4_t]: __arm_vaddlvq_p_u32 (__ARM_mve_coerce(__p0, uint32x4_t), p1));})
-
 #define __arm_vmladavaq(p0,p1,p2) ({ __typeof(p0) __p0 = (p0); \
   __typeof(p1) __p1 = (p1); \
   __typeof(p2) __p2 = (p2); \