I suspect "S/L" stands for substrate/latchup, which is what guardrings are supposed to help prevent.
I cannot explain your observed DRC issue except to wonder if the rule is from PO-to-active, or PO-to-contact. It could make a difference.
The better practice at my company is to always surround pads with guardrings. Usually the ESD device (associated with the pad) and the pad are both surrounded with the same ring.
Hastings ("The Art of Analog Layout") has quite a bit to say about guardrings and their role in latchup prevention.