[llvm] [X86] Use GFNI for LZCNT vXi8 ops (PR #141888)

[Simon Pilgrim via llvm-commits]

================
@@ -28998,6 +28998,35 @@ static SDValue LowerVectorCTLZ(SDValue Op, const SDLoc &DL,
   assert(Subtarget.hasSSSE3() && "Expected SSSE3 support for PSHUFB");
   return LowerVectorCTLZInRegLUT(Op, DL, Subtarget, DAG);
 }
+static SDValue LowerVectorCTLZ_GFNI(SDValue Op, SelectionDAG &DAG,
+                                    const X86Subtarget &Subtarget) {
+  SDLoc dl(Op);
+  MVT VT = Op.getSimpleValueType();
+  SDValue Input = Op.getOperand(0);
+
+  if (!VT.isVector() || VT.getVectorElementType() != MVT::i8)
+    return SDValue();
----------------
RKSimon wrote:

Isn't this already tested for in the call from LowerCTLZ? Maybe just assert?

https://github.com/llvm/llvm-project/pull/141888