oxide system hardware switch-port status shouldn't include ipv6_enabled

This may be here due to historical reasons / implementation details, but oxide system hardware switch-port is an Ethernet level command (L1/L2), so ipv6_enabled doesn't belong here unless it's representing an Ethertype filter.

Current output:

✕ treyaspelund@Tallon-IV 01:05:08 PM | ~/Downloads
‣ ./oxide --profile colo-admin system hardware switch-port status --port qsfp8 --switch-location switch1 --rack-id 48759651-d754-43e2-ad49-64dde3adc1c2
{
  "link": {
    "address": {  
      "a": [      
        168,      
        64,       
        37,       
        5,        
        15,      
        67, 
      ] 
    },
    "asic_id": 136,
    "autoneg": false,
    "enabled": true,
    "fec": "RS",
    "fsm_state": "LinkUp",
    "ipv6_enabled": false,     <<<<<<
    "kr": false,
    "link_id": 0,
    "link_state": "up",
    "media": "Optical",
    "port_id": "qsfp8",
    "prbs": "Mission",
    "presence": true,
    "speed": "Speed100G",
    "tofino_connector": 41
  },
  "monitors": {
    "receiver_power": [
      {
        "average": 1.2298001050949097
      },
      {
        "average": 0.8708999752998352
      },
      {
        "average": 0.9979000091552734
      },
      {
        "average": 1.1165000200271606
      }
    ],
    "transmitter_bias_current": [
      47.880001068115234,
      47.880001068115234,
      47.880001068115234,
      47.880001068115234
    ],
    "transmitter_power": [
      0.959600031375885,
      0.7152000069618225,
      0.8762000203132629,
      1.259200096130371
    ]
  }
}

Start by locating the Rust implementation and response serialization for oxide system hardware switch-port status, then inspect how the link output is assembled. Remove ipv6_enabled from this Ethernet-level status response and update or add coverage if that command has tests; done means the status output no longer includes the field.