Likelihood ratio test for a multivariate normal mean

Suppose we have $n$ i.i.d. observations from multivariate $\text{Normal}_p(\uline{\mu},\Sigma)$. The likelihood ratio test of $H_0:\uline{\mu}=\uline{\mu}_0$ against $H_1:\uline{\mu}\ne\uline{\mu}_0$ uses: