[kvdb-rocksdb] Add benchmark for point lookups

[dvdplm]

This is far from perfect but might be a start of something. The tricky part of benchmarking databases is to setup the data set properly. This PR doesn't do anything particularly fancy but it's worth mentioning the basics:

• each benchmark run adds 1M keys to the DB (single tx, so if you're low on memory be patient) ;
• empty/small DBs are much faster obviously so the first few benchmark runs are not very indicative of anything; I think at least 3-8M keys are needed before the values start to actually mean something;
• interestingly though, during my testing, performance seem to plateau and after a while, and adding more data doesn't seem to slow any of the tests down by much
• the benchmarks all count the number of allocations; this is why they use iter_custom() rather than iter().

Benchmark results:

get key                 time:   [2.5151 us 2.5616 us 2.6140 us]
                        change: [-4.3238% -1.1310% +2.0720%] (p = 0.50 > 0.05)
                        No change in performance detected.
Found 4 outliers among 100 measurements (4.00%)
  3 (3.00%) high mild
  1 (1.00%) high severe

[get key] total: iterations=4016151, allocations=6880994; allocations per iter=1.71

get key by prefix       time:   [15.011 us 15.252 us 15.511 us]
                        change: [-1.5796% +2.0998% +5.6256%] (p = 0.25 > 0.05)
                        No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
  5 (5.00%) high mild
  2 (2.00%) high severe

[get key by prefix] total: iterations=580293, allocations=1160586; allocations per iter=2.00

iterate over 1k keys    time:   [430.52 us 433.30 us 436.16 us]
                        change: [-11.873% -3.7270% +2.5027%] (p = 0.44 > 0.05)
                        No change in performance detected.
Found 7 outliers among 100 measurements (7.00%)
  3 (3.00%) high mild
  4 (4.00%) high severe

[iterate over 1k keys] total: iterations=23341, allocations=46705341; allocations per iter=2.00

single key from iterator
                        time:   [2.8204 us 2.8454 us 2.8730 us]
                        change: [-6.8037% -4.5874% -2.2547%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 3 outliers among 100 measurements (3.00%)
  2 (2.00%) high mild
  1 (1.00%) high severe

[single key from iterator] total: iterations=2785775, allocations=5571550; allocations per iter=2.00

[ordian]

Looks good as a starting point.

[dvdplm]

ping @niklasad1