Thank You. My intention here is not to just find optima. Since in convex smooth settings we indeed study the second order conditions to understand the nature of optima whether its unique or not. I wanted to know if something analogous existed in non smooth settings.