[MLIR][Standard] Fix use of `dyn_cast_or_null`

[frgossen]

Summary

The value may be a function argument in which case getDefiningOp will return a
nullptr.

[pifon2a]

Thank you, Frederik!

mlir/lib/Dialect/StandardOps/IR/Ops.cpp

struct ExtractElementFromTensorFromElements
     : public OpRewritePattern<ExtractElementOp> {
   using OpRewritePattern<ExtractElementOp>::OpRewritePattern;
 
   LogicalResult matchAndRewrite(ExtractElementOp extract,
                                 PatternRewriter &rewriter) const final {
     if (extract.indices().size() != 1)
       return failure();
 
-    auto tensor_from_elements =
-        dyn_cast<TensorFromElementsOp>(extract.aggregate().getDefiningOp());
+    auto tensor_from_elements = dyn_cast_or_null<TensorFromElementsOp>(
+        extract.aggregate().getDefiningOp());
     if (tensor_from_elements == nullptr)
       return failure();
 
     APInt index;
     if (!matchPattern(*extract.indices().begin(), m_ConstantInt(&index)))
       return failure();
     rewriter.replaceOp(extract,
                        tensor_from_elements.getOperand(index.getZExtValue()));