state.operands.push_back(source->getResult(0));
    } else {
      state.operands.push_back(dst->getOperand(input - 1));
    }
  }
  state.attributes.assign(dst->getAttrs().begin(), dst->getAttrs().end());
  state.types.assign(dst->getResultTypes().begin(),
                     dst->getResultTypes().end());
  builder_.setInsertionPoint(dst);
  auto* new_dst = builder_.create(state);

  if (failed(::mlir::TF::UpdateDataFormat(data_format, this))) return failure();

  // Update convolution attributes.
  (*this)->setAttr("dilations", ShuffleArrayAttr(getDilations(), perm));
  (*this)->setAttr("strides", ShuffleArrayAttr(getStrides(), perm));
  (*this)->setAttr("explicit_paddings",
                   ShuffleArrayAttr(getExplicitPaddings(), perm, 2));

  return success();
}

					v1.ResourceEphemeralStorage: *resource.NewQuantity(5000, resource.BinarySI),
				},
			}
			// Since this test retroactively overrides the stub container manager,
			// we have to regenerate default status setters.
			kubelet.setNodeStatusFuncs = kubelet.defaultNodeStatusFuncs()

			kubeClient := testKubelet.fakeKubeClient
			existingNode := v1.Node{ObjectMeta: metav1.ObjectMeta{Name: testKubeletHostname}}

	}
	return storage.SelectionPredicate{
		Label:    labels.Everything().Add(*l),
		Field:    fields.Everything(),
		GetAttrs: getPodAttrs,
	}
}

func matchEverything() storage.SelectionPredicate {
	return storage.SelectionPredicate{
		Label: labels.Everything(),
		Field: fields.Everything(),
		GetAttrs: func(obj runtime.Object) (label labels.Set, field fields.Set, err error) {
			return nil, nil, nil
		},
	}
}

    p.printGenericOp(op, /*printOpName=*/false);
    return;
  }

  p << ' ';
  p.printOperands(op->getOperands());
  p.printOptionalAttrDict(op->getAttrs());
  // Now we can output only one type for all operands and the result.
  p << " : " << resultType;
}

Operation* getDefiningBroadcastArgsOp(Value operand) {
  auto* defining_op = operand.getDefiningOp();

	}
	for name, tc := range cases {
		t.Run(name, func(t *testing.T) {
			if tc.job == nil {
				_, _, err := GetAttrs(tc.nonJobObject)
				if diff := cmp.Diff(tc.wantErr, err.Error()); diff != "" {
					t.Errorf("Unexpected errors (-want,+got):\n%s", diff)
				}
			} else {
				gotLabels, _, err := GetAttrs(tc.job)
				if err != nil {
					t.Errorf("Error %s supposed to be nil", err.Error())
				}

    auto tf_fused_batch_norm_op_v3 = CreateTfOp<TF::FusedBatchNormV3Op>(
        rewriter, tf_fused_batch_norm_op, new_result_types,
        tf_fused_batch_norm_op.getOperands(),
        tf_fused_batch_norm_op->getAttrs());

    rewriter.replaceOp(tf_fused_batch_norm_op,
                       tf_fused_batch_norm_op_v3->getResults().drop_back());
    return success();
  }
};
}  // namespace.

      auto name = attr.getName();
      return name == "call_target_name" || name == "backend_config" ||
             name == "api_version" || name == "called_computations";
    };
    for (const auto& attr : op->getAttrs()) {
      if (!is_supported_attr_name(attr)) {
        return op.emitOpError()
               << attr.getName().getValue()
               << " is not a supported attribute for ApproxTopK";
      }
    }

                .str());
        // building the computation info
        auto flex_builder = std::make_unique<flexbuffers::Builder>();
        size_t map_start = flex_builder->StartMap();
        auto attrs = vhlo_op->getAttrs();
        std::vector<mlir::NamedAttribute> extra_attributes;

        for (size_t i = 0; i < attrs.size(); ++i) {
          auto name = attrs[i].getName().str();
          auto attr = attrs[i].getValue();