Add sse4.1 fast path for u64

[dtolnay]

[xtqqczze]

The fast path is overall improvement for aarch64 as well, unsurprisingly

Benchmark results (aarch64)

Benchmarking u64[0]/itoa: Collecting 100 samples in estimated 5.0000 s (4.1B iteu64[0]/itoa             time:   [1.2231 ns 1.2238 ns 1.2246 ns]
                        change: [−28.457% −28.330% −28.184%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
  3 (3.00%) high mild
  6 (6.00%) high severe
Benchmarking u64[0]/std::fmt: Collecting 100 samples in estimated 5.0000 s (501Mu64[0]/std::fmt         time:   [9.9792 ns 9.9796 ns 9.9800 ns]
                        change: [−0.0531% +0.0025% +0.0578%] (p = 0.94 > 0.05)
                        No change in performance detected.
Found 14 outliers among 100 measurements (14.00%)
  7 (7.00%) high mild
  7 (7.00%) high severe

Benchmarking u64[half]/itoa: Collecting 100 samples in estimated 5.0000 s (1.9B u64[half]/itoa          time:   [2.7003 ns 2.7004 ns 2.7005 ns]
                        change: [−5.7593% −5.7075% −5.6534%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 10 outliers among 100 measurements (10.00%)
  3 (3.00%) high mild
  7 (7.00%) high severe
Benchmarking u64[half]/std::fmt: Collecting 100 samples in estimated 5.0001 s (3u64[half]/std::fmt      time:   [14.088 ns 14.089 ns 14.090 ns]
                        change: [−0.0510% +0.0054% +0.0652%] (p = 0.87 > 0.05)
                        No change in performance detected.
Found 13 outliers among 100 measurements (13.00%)
  4 (4.00%) high mild
  9 (9.00%) high severe

Benchmarking u64[max]/itoa: Collecting 100 samples in estimated 5.0000 s (1.4B iu64[max]/itoa           time:   [3.5462 ns 3.5472 ns 3.5490 ns]
                        change: [−24.522% −24.472% −24.422%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 9 outliers among 100 measurements (9.00%)
  4 (4.00%) high mild
  5 (5.00%) high severe
Benchmarking u64[max]/std::fmt: Collecting 100 samples in estimated 5.0001 s (26u64[max]/std::fmt       time:   [18.990 ns 19.074 ns 19.176 ns]
                        change: [+1.8795% +2.2892% +2.7503%] (p = 0.00 < 0.05)
                        Performance has regressed.
</details>