Using Singular Value Decomposition to Compute Variance Covariance Matrix where n < p

Question

The problem is the same as the following:-

Using Singular Value Decomposition to Compute Variance Covariance Matrix from linear regression model

In R, this works.

vcov.matrix <- var.est * (v %*% d^(-2) %*% t(v))

But as mentioned "assumption is that X is full-rank and n≥p throughout. If this is not the case, you'll have to make minor modifications to the above"

In my case, n<p (no of variables more than data points) so I want to know what should be modifications to the formula?