Don’t yse that method.
Preferred option: If you have Mosek 9.x available,aspecify it as solver under CVX 2.2
2nd choice option: Follow the instructions at CVXQUAD: How to use CVXQUAD's Pade Approximant instead of CVX's unreliable Successive Approximation for GP mode, log, exp, entr, rel_entr, kl_div, log_det, det_rootn, exponential cone. CVXQUAD's Quantum (Matrix) Entropy & Matrix Log related functions
3rd choice option:: Continue doing what you’re doing, using CVX’s unreliable Successive Approximation method. If you do, you’re on your own.