[llvm] [AArch64] Extend usage of `XAR` instruction for fixed-length operations (PR #139460)

[David Green via llvm-commits]

================
@@ -4606,7 +4606,36 @@ bool AArch64DAGToDAGISel::trySelectXAR(SDNode *N) {
     return false;
   }
 
-  if (!Subtarget->hasSHA3())
+  // We have Neon SHA3 XAR operation for v2i64 but for types
+  // v4i32, v8i16, v16i8 we can use SVE operations when SVE2-SHA3
+  // is available.
+  EVT SVT;
+  switch (VT.getSimpleVT().SimpleTy) {
+  case MVT::v4i32:
+    SVT = MVT::nxv4i32;
+    break;
+  case MVT::v8i16:
+    SVT = MVT::nxv8i16;
+    break;
+  case MVT::v16i8:
+    SVT = MVT::nxv16i8;
+    break;
+  case MVT::v1i64:
+  case MVT::v2i32:
+  case MVT::v4i16:
+  case MVT::v8i8:
+    // Widen type to v2i64.
+    SVT = MVT::v2i64;
+    break;
+  default:
+    if (VT != MVT::v2i64)
----------------
davemgreen wrote:

Could v2i64 be combined into the v1i64 case so that they share the same code and the default is just return false? The v2i64 case could then be something like `SVT = Subtarget->hasSHA3() ? MVT::v2i64 : MVT::nxv2i64;` so it triggers in even more cases (when we have SVE2, but not SHA3).

https://github.com/llvm/llvm-project/pull/139460