cmd/compile: optimize slice bounds checking with SUB/SUBconst comparisons

Optimize ARM64 code generation for slice bounds checking by recognizing
patterns where comparisons to zero involve SUB or SUBconst operations.
This change adds SSA opt rules to simplify:
 (CMPconst [0] (SUB x y)) => (CMP x y)

The optimizations apply to EQ, NE, ULE, and UGT comparisons, enabling
more efficient bounds checking for slice operations.

Code size improvement:
compile: .text:    9088004  ->  9065988 (-0.24%)
etcd:    .text:    10500276 -> 10497092 (-0.03%)
Change-Id: I467cb27674351652bcacc52b87e1f19677bd46a8
Reviewed-on: https://go-review.googlesource.com/c/go/+/679915
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Michael Knyszek <mknyszek@google.com>
Auto-Submit: Keith Randall <khr@golang.org>

1 files changed, 18 insertions, 0 deletions

diff --git a/test/codegen/slices.go b/test/codegen/slices.go
index 9e8990c586..30a131a5a5 100644
--- a/test/codegen/slices.go
+++ b/test/codegen/slices.go
@@ -429,3 +429,21 @@ func Slice0(p *struct{}, i int) []struct{} {
 	// amd64:-"MULQ"
 	return unsafe.Slice(p, i)
 }
+
+// --------------------------------------- //
+//   Code generation for slice bounds      //
+//   checking comparison                   //
+// --------------------------------------- //
+
+func SlicePut(a []byte, c uint8) []byte {
+	// arm64:`CBZ\tR1`
+	a[0] = c
+	// arm64:`CMP\t\$1, R1`
+	a = a[1:]
+	a[0] = c
+	// arm64:`CMP\t\$2, R1`
+	a = a[1:]
+	a[0] = c
+	a = a[1:]
+	return a
+}