CVXR: not recognizing a convex function

Question

I am trying to do some quadratic optimization with linear constraints, using CVXR. I have the following matrix in R:

HL <- matrix(c(
        s[2]*sqp[2]*sqq[2],0,0,0,0,0,0,0,0,0,0,0,
        0,s[3]*sqp[3]*sqq[3],0,0,0,0,0,0,0,0,0,0,
        0,0,s[4]*sqp[4]*sqq[4],0,0,0,0,0,0,0,0,0,
        0,0,0,s[1]*sqp[1]*sqq[1],0,0,0,0,0,0,0,0,
        0,0,0,0,s[3]*sqp[3]*sqq[3],0,0,0,0,0,0,0,
        0,0,0,0,0,s[4]*sqp[4]*sqq[4],0,0,0,0,0,0,
        0,0,0,0,0,0,s[1]*sqp[1]*sqq[1],0,0,0,0,0,
        0,0,0,0,0,0,0,s[2]*sqp[2]*sqq[2],0,0,0,0,
        0,0,0,0,0,0,0,0,s[4]*sqp[4]*sqq[4],0,0,0,
        0,0,0,0,0,0,0,0,0,s[1]*sqp[1]*sqq[1],0,0,
        0,0,0,0,0,0,0,0,0,0,s[2]*sqp[2]*sqq[2],0,
        0,0,0,0,0,0,0,0,0,0,0,s[3]*sqp[3]*sqq[3]
        ), nrow = num_vars, ncol = num_vars);

Then, I try to construct a DCP Ruleset-complying form: by first defining x = Variable(num_vars), and then writing fnc <- t(x) %*% HL %*% HL %*% x, or even something like y <- HL%*%x, and fnc <- t(y)%*%y. According to DCP Ruleset (see Quadratic forms section), this should be recognized as convex. However, when I load my source code into R, I get

Forming a non-convex expression (affine) * (affine)

How to fix this? I need CVXR to be able to recognize my function as "following DCP rules".