In BF-25887, we see a bad interaction between two tests (type6.js, type_bracket.js) on the selinux task.  The type6 test is known to cause issues for any test referencing MinKey (which are run in the same shell), and it's excluded from the parallel task for this reason. But, it was included in the selinux task, and I think by luck had not caused issues up until this point.

This ticket is to consider using the parallel denylist for the selinux task as well to avoid future issues like this.