[mlir] Use std::nullopt instead of None (NFC)

This patch mechanically replaces None with std::nullopt where the
compiler would warn if None were deprecated.  The intent is to reduce
the amount of manual work required in migrating from Optional to
std::optional.

This is part of an effort to migrate from llvm::Optional to
std::optional:

https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716

mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorCodegen.cpp

@@ -122,7 +122,7 @@ static Optional<Value> sizeFromTensorAtDim(OpBuilder &builder, Location loc,
                                            Value adaptedValue, unsigned dim) {
   auto enc = getSparseTensorEncoding(tensorTp);
   if (!enc)
-    return llvm::None;
+    return std::nullopt;
 
   // Access into static dimension can query original type directly.
   // Note that this is typically already done by DimOp's folding.
@@ -200,7 +200,7 @@ static Optional<LogicalResult>
 convertSparseTensorType(Type type, SmallVectorImpl<Type> &fields) {
   auto enc = getSparseTensorEncoding(type);
   if (!enc)
-    return llvm::None;
+    return std::nullopt;
   // Construct the basic types.
   auto *context = type.getContext();
   RankedTensorType rType = type.cast<RankedTensorType>();
@@ -1105,7 +1105,7 @@ mlir::SparseTensorTypeToBufferConverter::SparseTensorTypeToBufferConverter() {
                               Location loc) -> Optional<Value> {
     if (!getSparseTensorEncoding(tp))
       // Not a sparse tensor.
-      return llvm::None;
+      return std::nullopt;
     // Sparse compiler knows how to cancel out these casts.
     return genTuple(builder, loc, tp, inputs);
   });