In machine learning, the loss function for linear regression with L2 regularization (Ridge regression) is defined as $J(\\theta) = \\frac{1}{2} \\|X\\theta - y\\|^2 + \\frac{\\lambda}{2} \\|\\theta\\|^2$, where $X$ is the feature matrix, $y$ is the target vector, and $\\lambda > 0$ is the regularization parameter. What is the gradient $\\nabla_{\\theta} J(\\theta)$?