feat(antminer): add L11 support

[cryptographicturk]

Adds the Antminer L11 to the AntMiner model enum and hardware table.

Hardware shape

Verified against 35 live L11s (firmware FR-1.2(251206-L11&L11Pro), miner version 86.48-2.0.0), cross-checked from two independent endpoints on the same device:

Source	Boards	Chips/board	Fans
/cgi-bin/stats.cgi	chain_num: 3	asic_num: 88	fan_num: 4
cgminer 4028 stats	miner_count: 3	chain_acn1..3: 88	fan_num: 4

3 boards × 88 chips = 264, 4 fans.

Measured on the same miner before and after this change:

before:  model=ANTMINER L11 (Unknown)
         hardware: boards=None  board_count=None
         hashboards: 0
         expected_chips=None  total_chips=None

after:   model=L11
         hardware: boards=Some([Some(88), Some(88), Some(88)])  board_count=Some(3)
         hashboards: 3
           board[0] working_chips=Some(88) expected_chips=Some(88) active=Some(true)
           board[1] working_chips=Some(88) expected_chips=Some(88) active=Some(true)
           board[2] working_chips=Some(88) expected_chips=Some(88) active=Some(true)
         expected_chips=Some(264)  total_chips=Some(264)

An L9 on the same firmware family was used as a control and behaves identically, which is the intended outcome — the L11 now matches an already-supported sibling exactly.

Verification

• cargo fmt --all -- --check — clean
• cargo clippy --workspace --all-features — no new diagnostics
• cargo test -p asic-rs-makes-antminer — passing
• python scripts/docs/gen_supported_devices.py — regenerated; the docs change in this PR is entirely generator output
• Live check against multiple L11s on a production farm, with an L9 on the same firmware family as a control

Two pre-existing issues found while testing — not addressed here

Both affect the already-supported L9 identically, so they are not L11-specific and are deliberately left out of this PR. Happy to open separate issues or PRs if useful:

1. chain_rate{idx} is parsed as a string, but this firmware emits a JSON number. parse_hashboards does .and_then(|v| v.as_str()), so per-board hashrate is always None on 86.48-2.0.0. The values are present: "chain_rate1": 6.790325248.

2. temp_pcb{idx} does not exist on this firmware. It reports temp_in_pcb_1..3 / temp_out_pcb_1..3 (and temp_in_chip_* / temp_out_chip_*), so per-board board_temperature is always None. The 4-field temperature model added in 0.7.0 looks like a natural fit for these keys.

There is also a smaller one: get_system_info.cgi reports "Algorithm": "Scrypt" for both L11 and L9, but the AntMiner make reports SHA256 for both. Again pre-existing and not L11-specific.