[libcxx] Optimize `rng::generate_n` for segmented iterators (#165280)
Part of #102817.
This patch optimizes `rng::generate_n` for segmented iterators by
forwarding the implementation directly to `std::generate_n`.
- before
```
rng::generate_n(deque<int>)/32 21.7 ns 22.0 ns 32000000
rng::generate_n(deque<int>)/50 30.8 ns 30.7 ns 22400000
rng::generate_n(deque<int>)/1024 492 ns 488 ns 1120000
rng::generate_n(deque<int>)/8192 3938 ns 3924 ns 179200
```
- after
```
rng::generate_n(deque<int>)/32 11.0 ns 11.0 ns 64000000
rng::generate_n(deque<int>)/50 16.2 ns 16.1 ns 40727273
rng::generate_n(deque<int>)/1024 292 ns 286 ns 2240000
rng::generate_n(deque<int>)/8192 2291 ns 2302 ns 298667
```
NOKEYCHECK=True
GitOrigin-RevId: 0621fd0b8837192612d21785ad60664516513cea
3 files changed
