throw new MyRuntimeException();
      }
    }
    Fallback fallback = new Fallback();
    SettableFuture<Object> input = SettableFuture.create();

    ListenableFuture<Object> output = catching(input, Throwable.class, fallback, directExecutor());
    fallback.output = output;

    input.setException(new MyException());
    assertTrue(output.isCancelled());
  }

        throw new MyRuntimeException();
      }
    }
    Fallback fallback = new Fallback();
    SettableFuture<Object> input = SettableFuture.create();

    ListenableFuture<Object> output = catching(input, Throwable.class, fallback, directExecutor());
    fallback.output = output;

    input.setException(new MyException());
    assertTrue(output.isCancelled());
  }

		add chain ip kube-proxy filter-output { type filter hook output priority -110 ; }
		add rule ip kube-proxy filter-output ct state new jump service-endpoints-check
		add rule ip kube-proxy filter-output ct state new jump firewall-check
		add chain ip kube-proxy filter-output-post-dnat { type filter hook output priority -90 ; }
		add rule ip kube-proxy filter-output-post-dnat ct state new jump cluster-ips-check

  return rewriter.create<arith::ConstantOp>(output.getLoc(), attr_type, attr);
}

Value ExpandedDynamicShape(PatternRewriter& rewriter, Value input,
                           DenseIntElementsAttr broadcast_dimensions,
                           Value output) {
  assert(mlir::cast<ShapedType>(output.getType()) &&
         "output type must be of ShapedType");

}

void CreateSubgraphTouchingRefVar(const Scope& s) {
  Output variable =
      ops::Variable(s.WithOpName("variable"), PartialTensorShape{}, DT_FLOAT);
  Output read = ops::Identity(s.WithOpName("read_ref_var"), variable);
  Output neg = ops::Negate(s.WithOpName("negate_ref"), read);
  Output add = ops::Add(s.WithOpName("add_ref"), neg, neg);

  Output constant =

    return op.emitError() << "output 'sparse_indices' should have same length "
                          << "as attribute 'sparse_types'";
  }
  if (op.getSparseShapes().size() != sparse_types_count) {
    return op.emitError() << "output 'sparse_shapes' should have same length "
                          << "as attribute 'sparse_types'";
  }

  // Validate ragged variadic output lengths.

            if (!s.ok()) {
              return;
            }
            node_map[n] = copy;
          }

          // Only handle input edges. Output edges will be added later as
          // its output nodes' input edges.
          for (auto e : n->in_edges()) {
            if (node_map.find(e->src()) == node_map.end()) {
              s = errors::Internal("Cannot find node image for ",

// is equal to the non-contracting dimension after a reshape
bool BroadcastDimsProductEqual(Value input, Value output,
                               size_t agg_start_idx) {
  ArrayRef<int64_t> input_shape =
      mlir::cast<ShapedType>(input.getType()).getShape();
  ArrayRef<int64_t> output_shape =
      mlir::cast<ShapedType>(output.getType()).getShape();

  int64_t agg_value = 1;

  %cst = "tfl.no_value"() {value = unit} : () -> none
  // expected-error @+1 {{expect output type has rank = 4, got output type tensor<64x84x32xf32>}}

  bool InferShapeForRestore(Operation* op);

  // Infers the shape IfOp outputs based on the shapes of the then and else
  // function result types.
  bool InferShapeForIf(IfOp op);

  // Infers the shape IfRegion outputs based on the shapes of the then and else
  // yields.
  bool InferShapeForIfRegion(IfRegionOp op);

  // Infers the shape CaseOp outputs based on the shapes of branch function
  // result types.