Add benchmarks for no-op submission

[ollie-etl]

This PR adds some baseline benchmarks for NoOp submission / completion. This allows users to very roughly estimate the maximum bound for the throughput achievable using tokio-uring.

2 Benchmarks are added: Criterion, measuring wall time, and Lai, measuring cpu cycles. To run lai benchmarks, cachegrind is required.

###Notes
I've no doubt these can be improved, and should someone more experienced in micro-benchmarking wish to point out the glaring inefficiencies, I'd be very happy.

There are 2 lines of commented out code. These trigger a panic, issue #145

Some results

Hardware: 6 core i7

Criterion

o_op/1                  time:   [145.16 ms 145.48 ms 145.82 ms]
                        thrpt:  [685.78 Kelem/s 687.38 Kelem/s 688.92 Kelem/s]
                 change:
                        time:   [-2.8876% -2.4510% -2.0114%] (p = 0.00 < 0.05)
                        thrpt:  [+2.0526% +2.5126% +2.9734%]
                        Performance has improved.
Found 1 outliers among 100 measurements (1.00%)
  1 (1.00%) high mild
  
no_op/32                time:   [27.023 ms 27.078 ms 27.148 ms]
                        thrpt:  [3.6835 Melem/s 3.6930 Melem/s 3.7005 Melem/s]
                 change:
                        time:   [-1.3035% -0.9492% -0.5849%] (p = 0.00 < 0.05)
                        thrpt:  [+0.5883% +0.9583% +1.3207%]
                        Change within noise threshold.
Found 5 outliers among 100 measurements (5.00%)
  4 (4.00%) high mild
  1 (1.00%) high severe
  
no_op/64                time:   [25.175 ms 25.221 ms 25.274 ms]
                        thrpt:  [3.9567 Melem/s 3.9650 Melem/s 3.9722 Melem/s]
                 change:
                        time:   [-0.4796% -0.0752% +0.3019%] (p = 0.72 > 0.05)
                        thrpt:  [-0.3010% +0.0753% +0.4819%]
                        No change in performance detected.
Found 4 outliers among 100 measurements (4.00%)
  2 (2.00%) high mild
  2 (2.00%) high severe
  
no_op/256               time:   [23.487 ms 23.539 ms 23.601 ms]
                        thrpt:  [4.2372 Melem/s 4.2482 Melem/s 4.2576 Melem/s]
                 change:
                        time:   [-2.9132% -2.5070% -2.0627%] (p = 0.00 < 0.05)
                        thrpt:  [+2.1062% +2.5715% +3.0006%]
                        Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
  5 (5.00%) high mild
  3 (3.00%) high severe

Lai

runtime_only
  Instructions:               28479 (No change)
  L1 Accesses:                40528 (No change)
  L2 Accesses:                  243 (No change)
  RAM Accesses:                 977 (No change)
  Estimated Cycles:           75938 (No change)

no_op_x1
  Instructions:           500699958 (+0.023513%)
  L1 Accesses:            754985269 (+0.015636%)
  L2 Accesses:              1575761 (No change)
  RAM Accesses:                1308 (No change)
  Estimated Cycles:       762909854 (+0.015474%)

no_op_x32
  Instructions:           129944394 (-0.008193%)
  L1 Accesses:            204870254 (-0.005207%)
  L2 Accesses:                49632 (No change)
  RAM Accesses:                1373 (No change)
  Estimated Cycles:       205166469 (-0.005200%)

no_op_x64
  Instructions:           123956020 (+0.003089%)
  L1 Accesses:            195968649 (+0.001997%)
  L2 Accesses:                39143 (No change)
  RAM Accesses:                1441 (No change)
  Estimated Cycles:       196214799 (+0.001994%)

no_op_x256
  Instructions:           119631265 (+0.000911%)
  L1 Accesses:            188921752 (+0.000577%)
  L2 Accesses:               649009 (No change)
  RAM Accesses:                1869 (No change)
  Estimated Cycles:       192232212 (+0.000567%)