Digging through v6 RFC 4681, is there any specific reason/design choice that ND, specifically NS, only uses the source link-layer address option over just the MAC in the ethernet header? Best I can find is potential load balancing mechanism and a niche case where a sender can choose not to use that address in the option as the dst. MAC. Solicitations are going to have the src. MAC essentially every time. The v6 process appears to ignore the L2 header entirely and just processes ICMPv6 packet or am I overlooking a critical function?
