aarch64: Model zero-high-half semantics of XTN instruction in RTL

Modeling the zero-high-half semantics of the XTN narrowing
instruction in RTL indicates to the compiler that this is a totally
destructive operation. This enables more RTL simplifications and also
prevents some register allocation issues.

Add new tests to narrow_zero_high_half.c to verify the benefit of
this change.

gcc/ChangeLog:

2021-06-11  Jonathan Wright  <jonathan.wright@arm.com>

	* config/aarch64/aarch64-simd.md (aarch64_xtn<mode>_insn_le):
	Define - modeling zero-high-half semantics.
	(aarch64_xtn<mode>): Change to an expander that emits the
	appropriate instruction depending on endianness.
	(aarch64_xtn<mode>_insn_be): Define - modeling zero-high-half
	semantics.
	(aarch64_xtn2<mode>_le): Rename to...
	(aarch64_xtn2<mode>_insn_le): This.
	(aarch64_xtn2<mode>_be): Rename to...
	(aarch64_xtn2<mode>_insn_be): This.
	(vec_pack_trunc_<mode>): Emit truncation instruction instead
	of aarch64_xtn.
	* config/aarch64/iterators.md (Vnarrowd): Add Vnarrowd mode
	attribute iterator.

gcc/testsuite/ChangeLog:

	* gcc.target/aarch64/narrow_zero_high_half.c: Add new tests.

3 files changed, 88 insertions, 35 deletions

diff --git a/gcc/config/aarch64/aarch64-simd.md b/gcc/config/aarch64/aarch64-simd.md
index e750fae..b23556b 100644
--- a/gcc/config/aarch64/aarch64-simd.md
+++ b/gcc/config/aarch64/aarch64-simd.md
@@ -1690,17 +1690,48 @@
 
 ;; Narrowing operations.
 
-;; For doubles.
+(define_insn "aarch64_xtn<mode>_insn_le"
+  [(set (match_operand:<VNARROWQ2> 0 "register_operand" "=w")
+	(vec_concat:<VNARROWQ2>
+	  (truncate:<VNARROWQ> (match_operand:VQN 1 "register_operand" "w"))
+	  (match_operand:<VNARROWQ> 2 "aarch64_simd_or_scalar_imm_zero")))]
+  "TARGET_SIMD && !BYTES_BIG_ENDIAN"
+  "xtn\\t%0.<Vntype>, %1.<Vtype>"
+  [(set_attr "type" "neon_move_narrow_q")]
+)
 
-(define_insn "aarch64_xtn<mode>"
-  [(set (match_operand:<VNARROWQ> 0 "register_operand" "=w")
-	(truncate:<VNARROWQ> (match_operand:VQN 1 "register_operand" "w")))]
-  "TARGET_SIMD"
+(define_insn "aarch64_xtn<mode>_insn_be"
+  [(set (match_operand:<VNARROWQ2> 0 "register_operand" "=w")
+	(vec_concat:<VNARROWQ2>
+	  (match_operand:<VNARROWQ> 2 "aarch64_simd_or_scalar_imm_zero")
+	  (truncate:<VNARROWQ> (match_operand:VQN 1 "register_operand" "w"))))]
+  "TARGET_SIMD && BYTES_BIG_ENDIAN"
   "xtn\\t%0.<Vntype>, %1.<Vtype>"
   [(set_attr "type" "neon_move_narrow_q")]
 )
 
-(define_insn "aarch64_xtn2<mode>_le"
+(define_expand "aarch64_xtn<mode>"
+  [(set (match_operand:<VNARROWQ> 0 "register_operand")
+	(truncate:<VNARROWQ> (match_operand:VQN 1 "register_operand")))]
+  "TARGET_SIMD"
+  {
+    rtx tmp = gen_reg_rtx (<VNARROWQ2>mode);
+    if (BYTES_BIG_ENDIAN)
+      emit_insn (gen_aarch64_xtn<mode>_insn_be (tmp, operands[1],
+				CONST0_RTX (<VNARROWQ>mode)));
+    else
+      emit_insn (gen_aarch64_xtn<mode>_insn_le (tmp, operands[1],
+				CONST0_RTX (<VNARROWQ>mode)));
+
+    /* The intrinsic expects a narrow result, so emit a subreg that will get
+       optimized away as appropriate.  */
+    emit_move_insn (operands[0], lowpart_subreg (<VNARROWQ>mode, tmp,
+						 <VNARROWQ2>mode));
+    DONE;
+  }
+)
+
+(define_insn "aarch64_xtn2<mode>_insn_le"
   [(set (match_operand:<VNARROWQ2> 0 "register_operand" "=w")
 	(vec_concat:<VNARROWQ2>
 	  (match_operand:<VNARROWQ> 1 "register_operand" "0")
@@ -1710,7 +1741,7 @@
   [(set_attr "type" "neon_move_narrow_q")]
 )
 
-(define_insn "aarch64_xtn2<mode>_be"
+(define_insn "aarch64_xtn2<mode>_insn_be"
   [(set (match_operand:<VNARROWQ2> 0 "register_operand" "=w")
 	(vec_concat:<VNARROWQ2>
 	  (truncate:<VNARROWQ> (match_operand:VQN 2 "register_operand" "w"))
@@ -1727,15 +1758,17 @@
   "TARGET_SIMD"
   {
     if (BYTES_BIG_ENDIAN)
-      emit_insn (gen_aarch64_xtn2<mode>_be (operands[0], operands[1],
-					     operands[2]));
+      emit_insn (gen_aarch64_xtn2<mode>_insn_be (operands[0], operands[1],
+						 operands[2]));
     else
-      emit_insn (gen_aarch64_xtn2<mode>_le (operands[0], operands[1],
-					     operands[2]));
+      emit_insn (gen_aarch64_xtn2<mode>_insn_le (operands[0], operands[1],
+						 operands[2]));
     DONE;
   }
 )
 
+;; Packing doubles.
+
 (define_expand "vec_pack_trunc_<mode>"
  [(match_operand:<VNARROWD> 0 "register_operand")
   (match_operand:VDN 1 "register_operand")
@@ -1748,10 +1781,35 @@
 
   emit_insn (gen_move_lo_quad_<Vdbl> (tempreg, operands[lo]));
   emit_insn (gen_move_hi_quad_<Vdbl> (tempreg, operands[hi]));
-  emit_insn (gen_aarch64_xtn<Vdbl> (operands[0], tempreg));
+  emit_insn (gen_trunc<Vdbl><Vnarrowd>2 (operands[0], tempreg));
   DONE;
 })
 
+;; Packing quads.
+
+(define_expand "vec_pack_trunc_<mode>"
+ [(set (match_operand:<VNARROWQ2> 0 "register_operand")
+       (vec_concat:<VNARROWQ2>
+	 (truncate:<VNARROWQ> (match_operand:VQN 1 "register_operand"))
+	 (truncate:<VNARROWQ> (match_operand:VQN 2 "register_operand"))))]
+ "TARGET_SIMD"
+ {
+   rtx tmpreg = gen_reg_rtx (<VNARROWQ>mode);
+   int lo = BYTES_BIG_ENDIAN ? 2 : 1;
+   int hi = BYTES_BIG_ENDIAN ? 1 : 2;
+
+   emit_insn (gen_trunc<mode><Vnarrowq>2 (tmpreg, operands[lo]));
+
+   if (BYTES_BIG_ENDIAN)
+     emit_insn (gen_aarch64_xtn2<mode>_insn_be (operands[0], tmpreg,
+						operands[hi]));
+   else
+     emit_insn (gen_aarch64_xtn2<mode>_insn_le (operands[0], tmpreg,
+						operands[hi]));
+   DONE;
+ }
+)
+
 (define_insn "aarch64_shrn<mode>_insn_le"
   [(set (match_operand:<VNARROWQ2> 0 "register_operand" "=w")
 	(vec_concat:<VNARROWQ2>
@@ -1936,29 +1994,6 @@
   }
 )
 
-;; For quads.
-
-(define_expand "vec_pack_trunc_<mode>"
- [(set (match_operand:<VNARROWQ2> 0 "register_operand")
-       (vec_concat:<VNARROWQ2>
-	 (truncate:<VNARROWQ> (match_operand:VQN 1 "register_operand"))
-	 (truncate:<VNARROWQ> (match_operand:VQN 2 "register_operand"))))]
- "TARGET_SIMD"
- {
-   rtx tmpreg = gen_reg_rtx (<VNARROWQ>mode);
-   int lo = BYTES_BIG_ENDIAN ? 2 : 1;
-   int hi = BYTES_BIG_ENDIAN ? 1 : 2;
-
-   emit_insn (gen_aarch64_xtn<mode> (tmpreg, operands[lo]));
-
-   if (BYTES_BIG_ENDIAN)
-     emit_insn (gen_aarch64_xtn2<mode>_be (operands[0], tmpreg, operands[hi]));
-   else
-     emit_insn (gen_aarch64_xtn2<mode>_le (operands[0], tmpreg, operands[hi]));
-   DONE;
- }
-)
-
 ;; Widening operations.
 
 (define_insn "aarch64_simd_vec_unpack<su>_lo_<mode>"
diff --git a/gcc/config/aarch64/iterators.md b/gcc/config/aarch64/iterators.md
index e9047d0..caa42f8 100644
--- a/gcc/config/aarch64/iterators.md
+++ b/gcc/config/aarch64/iterators.md
@@ -1257,6 +1257,8 @@
 ;; Narrowed modes for VDN.
 (define_mode_attr VNARROWD [(V4HI "V8QI") (V2SI "V4HI")
 			    (DI   "V2SI")])
+(define_mode_attr Vnarrowd [(V4HI "v8qi") (V2SI "v4hi")
+			    (DI   "v2si")])
 
 ;; Narrowed double-modes for VQN (Used for XTN).
 (define_mode_attr VNARROWQ [(V8HI "V8QI") (V4SI "V4HI")
diff --git a/gcc/testsuite/gcc.target/aarch64/narrow_zero_high_half.c b/gcc/testsuite/gcc.target/aarch64/narrow_zero_high_half.c
index a79a4c3..451b011 100644
--- a/gcc/testsuite/gcc.target/aarch64/narrow_zero_high_half.c
+++ b/gcc/testsuite/gcc.target/aarch64/narrow_zero_high_half.c
@@ -48,6 +48,21 @@ TEST_SHIFT (vqrshrun_n, uint8x16_t, int16x8_t, s16, u8)
 TEST_SHIFT (vqrshrun_n, uint16x8_t, int32x4_t, s32, u16)
 TEST_SHIFT (vqrshrun_n, uint32x4_t, int64x2_t, s64, u32)
 
+#define TEST_UNARY(name, rettype, intype, fs, rs) \
+  rettype test_ ## name ## _ ## fs ## _zero_high \
+		(intype a) \
+	{ \
+		return vcombine_ ## rs (name ## _ ## fs (a), \
+					vdup_n_ ## rs (0)); \
+	}
+
+TEST_UNARY (vmovn, int8x16_t, int16x8_t, s16, s8)
+TEST_UNARY (vmovn, int16x8_t, int32x4_t, s32, s16)
+TEST_UNARY (vmovn, int32x4_t, int64x2_t, s64, s32)
+TEST_UNARY (vmovn, uint8x16_t, uint16x8_t, u16, u8)
+TEST_UNARY (vmovn, uint16x8_t, uint32x4_t, u32, u16)
+TEST_UNARY (vmovn, uint32x4_t, uint64x2_t, u64, u32)
+
 /* { dg-final { scan-assembler-not "dup\\t" } } */
 
 /* { dg-final { scan-assembler-times "\\tshrn\\tv" 6} }  */
@@ -58,3 +73,4 @@ TEST_SHIFT (vqrshrun_n, uint32x4_t, int64x2_t, s64, u32)
 /* { dg-final { scan-assembler-times "\\tuqrshrn\\tv" 3} }  */
 /* { dg-final { scan-assembler-times "\\tsqshrun\\tv" 3} }  */
 /* { dg-final { scan-assembler-times "\\tsqrshrun\\tv" 3} }  */
+/* { dg-final { scan-assembler-times "\\txtn\\tv" 6} }  */