[llvm] [X86][SelectionDAG] - Add support for llvm.canonicalize intrinsic (PR #106370)

[Matt Arsenault via llvm-commits]

================
@@ -32664,6 +32679,24 @@ static SDValue LowerPREFETCH(SDValue Op, const X86Subtarget &Subtarget,
   return Op;
 }
 
+static SDValue LowerFCanonicalize(SDValue Op, SelectionDAG &DAG) {
+  SDNode *N = Op.getNode();
+  SDValue Operand = N->getOperand(0);
+  EVT VT = Operand.getValueType();
+  SDLoc dl(N);
+
+  SDValue One = DAG.getConstantFP(1.0, dl, VT);
+
+  // TODO: Fix Crash for bf16 when generating strict_fmul as it
+  // leads to a error : SoftPromoteHalfResult #0: t11: bf16,ch = strict_fmul t0,
+  // ConstantFP:bf16<APFloat(16256)>, t5 LLVM ERROR: Do not know how to soft
+  // promote this operator's result!
+  SDValue Chain = DAG.getEntryNode();
+  SDValue StrictFmul = DAG.getNode(ISD::STRICT_FMUL, dl, {VT, MVT::Other},
----------------
arsenm wrote:

No, it will only provide a reasonable default for other targets. Targets are still free to make it legal or custom lower as they choose 

https://github.com/llvm/llvm-project/pull/106370