AArch64: Improve immediate expansion [PR105928]

Support immediate expansion of immediates which can be created from 2 MOVKs
and a shifted ORR or BIC instruction.  Change aarch64_split_dimode_const_store
to apply if we save one instruction.

This reduces the number of 4-instruction immediates in SPECINT/FP by 5%.

gcc/ChangeLog:
	PR target/105928
	* config/aarch64/aarch64.cc (aarch64_internal_mov_immediate)
	Add support for immediates using shifted ORR/BIC.
	(aarch64_split_dimode_const_store): Apply if we save one instruction.
	* config/aarch64/aarch64.md (<LOGICAL:optab>_<SHIFT:optab><mode>3):
	Make pattern global.

gcc/testsuite:
	PR target/105928
	* gcc.target/aarch64/pr105928.c: Add new test.
	* gcc.target/aarch64/vect-cse-codegen.c: Fix test.

4 files changed, 77 insertions, 14 deletions

diff --git a/gcc/config/aarch64/aarch64.cc b/gcc/config/aarch64/aarch64.cc
index 7bb1161..219c4ee 100644
--- a/gcc/config/aarch64/aarch64.cc
+++ b/gcc/config/aarch64/aarch64.cc
@@ -5640,7 +5640,7 @@ aarch64_internal_mov_immediate (rtx dest, rtx imm, bool generate,
 				machine_mode mode)
 {
   int i;
-  unsigned HOST_WIDE_INT val, val2, mask;
+  unsigned HOST_WIDE_INT val, val2, val3, mask;
   int one_match, zero_match;
   int num_insns;
 
@@ -5722,6 +5722,35 @@ aarch64_internal_mov_immediate (rtx dest, rtx imm, bool generate,
 		}
 	      return 3;
 	    }
+
+      /* Try shifting and inserting the bottom 32-bits into the top bits.  */
+      val2 = val & 0xffffffff;
+      val3 = 0xffffffff;
+      val3 = val2 | (val3 << 32);
+      for (i = 17; i < 48; i++)
+	if ((val2 | (val2 << i)) == val)
+	  {
+	    if (generate)
+	      {
+		emit_insn (gen_rtx_SET (dest, GEN_INT (val2 & 0xffff)));
+		emit_insn (gen_insv_immdi (dest, GEN_INT (16),
+					   GEN_INT (val2 >> 16)));
+		emit_insn (gen_ior_ashldi3 (dest, dest, GEN_INT (i), dest));
+	      }
+	    return 3;
+	  }
+	else if ((val3 & ~(val3 << i)) == val)
+	  {
+	    if (generate)
+	      {
+		emit_insn (gen_rtx_SET (dest, GEN_INT (val3 | 0xffff0000)));
+		emit_insn (gen_insv_immdi (dest, GEN_INT (16),
+					   GEN_INT (val2 >> 16)));
+		emit_insn (gen_and_one_cmpl_ashldi3 (dest, dest, GEN_INT (i),
+						      dest));
+	      }
+	    return 3;
+	  }
     }
 
   /* Generate 2-4 instructions, skipping 16 bits of all zeroes or ones which
@@ -25540,8 +25569,6 @@ aarch64_split_dimode_const_store (rtx dst, rtx src)
   rtx lo = gen_lowpart (SImode, src);
   rtx hi = gen_highpart_mode (SImode, DImode, src);
 
-  bool size_p = optimize_function_for_size_p (cfun);
-
   if (!rtx_equal_p (lo, hi))
     return false;
 
@@ -25560,14 +25587,8 @@ aarch64_split_dimode_const_store (rtx dst, rtx src)
      MOV	w1, 49370
      MOVK	w1, 0x140, lsl 16
      STP	w1, w1, [x0]
-   So we want to perform this only when we save two instructions
-   or more.  When optimizing for size, however, accept any code size
-   savings we can.  */
-  if (size_p && orig_cost <= lo_cost)
-    return false;
-
-  if (!size_p
-      && (orig_cost <= lo_cost + 1))
+   So we want to perform this when we save at least one instruction.  */
+  if (orig_cost <= lo_cost)
     return false;
 
   rtx mem_lo = adjust_address (dst, SImode, 0);
diff --git a/gcc/config/aarch64/aarch64.md b/gcc/config/aarch64/aarch64.md
index 6f7827b..60133b5 100644
--- a/gcc/config/aarch64/aarch64.md
+++ b/gcc/config/aarch64/aarch64.md
@@ -4637,7 +4637,7 @@
   [(set_attr "type" "logics_shift_imm")]
 )
 
-(define_insn "*<LOGICAL:optab>_<SHIFT:optab><mode>3"
+(define_insn "<LOGICAL:optab>_<SHIFT:optab><mode>3"
   [(set (match_operand:GPI 0 "register_operand" "=r")
 	(LOGICAL:GPI (SHIFT:GPI
 		      (match_operand:GPI 1 "register_operand" "r")
diff --git a/gcc/testsuite/gcc.target/aarch64/pr105928.c b/gcc/testsuite/gcc.target/aarch64/pr105928.c
new file mode 100644
index 0000000..9aff892
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/pr105928.c
@@ -0,0 +1,43 @@
+/* { dg-do assemble } */
+/* { dg-options "-O2 --save-temps" } */
+
+long long f1 (void)
+{
+  return 0x80402010080400;
+}
+
+long long f2 (void)
+{
+  return 0x1234567812345678;
+}
+
+long long f3 (void)
+{
+  return 0x4567800012345678;
+}
+
+long long f4 (void)
+{
+  return 0x3ecccccd3ecccccd;
+}
+
+long long f5 (void)
+{
+  return 0x38e38e38e38e38e;
+}
+
+long long f6 (void)
+{
+  return 0x1745d1745d1745d;
+}
+
+void f7 (long long *p)
+{
+  *p = 0x1234567812345678;
+}
+
+/* { dg-final { scan-assembler-times {\tmovk\t} 7 } } */
+/* { dg-final { scan-assembler-times {\tmov\t} 7 } } */
+/* { dg-final { scan-assembler-times {\tbic\t} 2 } } */
+/* { dg-final { scan-assembler-times {\torr\t} 4 } } */
+/* { dg-final { scan-assembler-times {\tstp\t} 1 } } */
diff --git a/gcc/testsuite/gcc.target/aarch64/vect-cse-codegen.c b/gcc/testsuite/gcc.target/aarch64/vect-cse-codegen.c
index d025e98..2b8e643 100644
--- a/gcc/testsuite/gcc.target/aarch64/vect-cse-codegen.c
+++ b/gcc/testsuite/gcc.target/aarch64/vect-cse-codegen.c
@@ -72,8 +72,7 @@ test3 (uint32_t a, uint32x4_t b, uint32x4_t* rt)
 **	ushr	v[0-9]+.16b, v[0-9]+.16b, 7
 **	mov	x[0-9]+, 16512
 **	movk	x[0-9]+, 0x1020, lsl 16
-**	movk	x[0-9]+, 0x408, lsl 32
-**	movk	x[0-9]+, 0x102, lsl 48
+**	orr	x[0-9]+, x[0-9]+, x[0-9]+, lsl 28
 **	fmov	d[0-9]+, x[0-9]+
 **	pmull	v[0-9]+.1q, v[0-9]+.1d, v[0-9]+.1d
 **	dup	v[0-9]+.2d, v[0-9]+.d\[0\]