Light-matter interaction is always fancinating but obsure. Particularly, when light marries to magnetism, the circumstance becomes even more horrible. If we look at the basic band theory, one immediately sees rare understanding on how photon controls magnetism, from an electronic point of view. Recently, some time-dependent dynamics simulations have been performed, but they lack of description on how geometric phase in momentum space of Bloch electrons control/manipulate magnetization. In a recent work, we developed a second order perturbation theory to add a brick into this big project. Two promising aspects are revealed: 1. How do the quantum metric and Berry curvature (real and imaginary parts of quantum geometry tensor) relates with magnetic torque; 2. For the nonequilibrium process, how much the orbital moment contributes to spin orientation (via spin orbit coupling), which is usually silent at the equilibrium state? Furthermore, we have to add a showcase to have our work published, which is based on a NiCl₂ layer that favors an in-plane magnetization. We adopt our theory to show that it can be switched to out-of-plane, under at least subnanosecond time scale. See our recent report in ACS Nano at https://pubs.acs.org/doi/10.1021/acsnano.4c06453.