aarch64: Use RTL builtins for [su]mlal intrinsics

Rewrite [su]mlal Neon intrinsics to use RTL builtins rather than
inline assembly code, allowing for better scheduling and
optimization.

gcc/ChangeLog:

2021-01-26  Jonathan Wright  <jonathan.wright@arm.com>

	* config/aarch64/aarch64-simd-builtins.def: Add [su]mlal
	builtin generator macros.
	* config/aarch64/aarch64-simd.md (*aarch64_<su>mlal<mode>):
	Rename to...
	(aarch64_<su>mlal<mode>): This.
	* config/aarch64/arm_neon.h (vmlal_s8): Use RTL builtin
	instead of inline asm.
	(vmlal_s16): Likewise.
	(vmlal_s32): Likewise.
	(vmlal_u8): Likewise.
	(vmlal_u16): Likewise.
	(vmlal_u32): Likewise.

diff --git a/gcc/config/aarch64/aarch64-simd-builtins.def b/gcc/config/aarch64/aarch64-simd-builtins.def
index 32aee60..a71ae4d 100644 (file)
--- a/gcc/config/aarch64/aarch64-simd-builtins.def
+++ b/gcc/config/aarch64/aarch64-simd-builtins.def
@@ -192,6 +192,10 @@
   BUILTIN_VD_BHSI (TERNOP, smlsl, 0, NONE)
   BUILTIN_VD_BHSI (TERNOPU, umlsl, 0, NONE)
 
+  /* Implemented by aarch64_<su>mlal<mode>.  */
+  BUILTIN_VD_BHSI (TERNOP, smlal, 0, NONE)
+  BUILTIN_VD_BHSI (TERNOPU, umlal, 0, NONE)
+
   /* Implemented by aarch64_<su>mlsl_hi<mode>.  */
   BUILTIN_VQW (TERNOP, smlsl_hi, 0, NONE)
   BUILTIN_VQW (TERNOPU, umlsl_hi, 0, NONE)

diff --git a/gcc/config/aarch64/aarch64-simd.md b/gcc/config/aarch64/aarch64-simd.md
index 544bac7..db56b61 100644 (file)
--- a/gcc/config/aarch64/aarch64-simd.md
+++ b/gcc/config/aarch64/aarch64-simd.md
@@ -1825,17 +1825,17 @@
 }
 )
 
-(define_insn "*aarch64_<su>mlal<mode>"
+(define_insn "aarch64_<su>mlal<mode>"
   [(set (match_operand:<VWIDE> 0 "register_operand" "=w")
         (plus:<VWIDE>
           (mult:<VWIDE>
             (ANY_EXTEND:<VWIDE>
-              (match_operand:VD_BHSI 1 "register_operand" "w"))
+              (match_operand:VD_BHSI 2 "register_operand" "w"))
             (ANY_EXTEND:<VWIDE>
-              (match_operand:VD_BHSI 2 "register_operand" "w")))
-          (match_operand:<VWIDE> 3 "register_operand" "0")))]
+              (match_operand:VD_BHSI 3 "register_operand" "w")))
+          (match_operand:<VWIDE> 1 "register_operand" "0")))]
   "TARGET_SIMD"
-  "<su>mlal\t%0.<Vwtype>, %1.<Vtype>, %2.<Vtype>"
+  "<su>mlal\t%0.<Vwtype>, %2.<Vtype>, %3.<Vtype>"
   [(set_attr "type" "neon_mla_<Vetype>_long")]
 )
 

diff --git a/gcc/config/aarch64/arm_neon.h b/gcc/config/aarch64/arm_neon.h
index d891067..b5c1f06 100644 (file)
--- a/gcc/config/aarch64/arm_neon.h
+++ b/gcc/config/aarch64/arm_neon.h
@@ -7662,72 +7662,42 @@ __extension__ extern __inline int16x8_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 vmlal_s8 (int16x8_t __a, int8x8_t __b, int8x8_t __c)
 {
-  int16x8_t __result;
-  __asm__ ("smlal %0.8h,%2.8b,%3.8b"
-           : "=w"(__result)
-           : "0"(__a), "w"(__b), "w"(__c)
-           : /* No clobbers */);
-  return __result;
+  return __builtin_aarch64_smlalv8qi (__a, __b, __c);
 }
 
 __extension__ extern __inline int32x4_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 vmlal_s16 (int32x4_t __a, int16x4_t __b, int16x4_t __c)
 {
-  int32x4_t __result;
-  __asm__ ("smlal %0.4s,%2.4h,%3.4h"
-           : "=w"(__result)
-           : "0"(__a), "w"(__b), "w"(__c)
-           : /* No clobbers */);
-  return __result;
+  return __builtin_aarch64_smlalv4hi (__a, __b, __c);
 }
 
 __extension__ extern __inline int64x2_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 vmlal_s32 (int64x2_t __a, int32x2_t __b, int32x2_t __c)
 {
-  int64x2_t __result;
-  __asm__ ("smlal %0.2d,%2.2s,%3.2s"
-           : "=w"(__result)
-           : "0"(__a), "w"(__b), "w"(__c)
-           : /* No clobbers */);
-  return __result;
+  return __builtin_aarch64_smlalv2si (__a, __b, __c);
 }
 
 __extension__ extern __inline uint16x8_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 vmlal_u8 (uint16x8_t __a, uint8x8_t __b, uint8x8_t __c)
 {
-  uint16x8_t __result;
-  __asm__ ("umlal %0.8h,%2.8b,%3.8b"
-           : "=w"(__result)
-           : "0"(__a), "w"(__b), "w"(__c)
-           : /* No clobbers */);
-  return __result;
+  return __builtin_aarch64_umlalv8qi_uuuu (__a, __b, __c);
 }
 
 __extension__ extern __inline uint32x4_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 vmlal_u16 (uint32x4_t __a, uint16x4_t __b, uint16x4_t __c)
 {
-  uint32x4_t __result;
-  __asm__ ("umlal %0.4s,%2.4h,%3.4h"
-           : "=w"(__result)
-           : "0"(__a), "w"(__b), "w"(__c)
-           : /* No clobbers */);
-  return __result;
+  return __builtin_aarch64_umlalv4hi_uuuu (__a, __b, __c);
 }
 
 __extension__ extern __inline uint64x2_t
 __attribute__ ((__always_inline__, __gnu_inline__, __artificial__))
 vmlal_u32 (uint64x2_t __a, uint32x2_t __b, uint32x2_t __c)
 {
-  uint64x2_t __result;
-  __asm__ ("umlal %0.2d,%2.2s,%3.2s"
-           : "=w"(__result)
-           : "0"(__a), "w"(__b), "w"(__c)
-           : /* No clobbers */);
-  return __result;
+  return __builtin_aarch64_umlalv2si_uuuu (__a, __b, __c);
 }
 
 __extension__ extern __inline float32x4_t