buffer: make compare/equals faster

[tniessen]

This patch adds a V8 fast API implementation for the buffer.compare binding, which is used both by Buffer.prototype.equals and Buffer.prototype.compare. In particular, it significantly improves the performance of comparing buffers for equality.

Local benchmark:

                                                                      confidence improvement accuracy (*)    (**)   (***)
buffers/buffer-compare-instance-method.js n=1000000 args=1 size=16           ***    225.73 %      ±16.81% ±22.62% ±29.94%
buffers/buffer-compare-instance-method.js n=1000000 args=1 size=16386        ***     15.57 %       ±1.32%  ±1.76%  ±2.29%
buffers/buffer-compare-instance-method.js n=1000000 args=1 size=4096         ***     61.44 %       ±6.15%  ±8.27% ±10.93%
buffers/buffer-compare-instance-method.js n=1000000 args=1 size=512          ***    181.97 %       ±7.02%  ±9.45% ±12.52%
buffers/buffer-compare-instance-method.js n=1000000 args=2 size=16                    1.82 %       ±3.08%  ±4.11%  ±5.39%
buffers/buffer-compare-instance-method.js n=1000000 args=2 size=16386                -0.02 %       ±1.38%  ±1.84%  ±2.39%
buffers/buffer-compare-instance-method.js n=1000000 args=2 size=4096           *     -2.30 %       ±2.21%  ±2.93%  ±3.82%
buffers/buffer-compare-instance-method.js n=1000000 args=2 size=512            *     -3.92 %       ±3.02%  ±4.02%  ±5.24%
buffers/buffer-compare-instance-method.js n=1000000 args=5 size=16                   -2.86 %       ±3.69%  ±4.93%  ±6.48%
buffers/buffer-compare-instance-method.js n=1000000 args=5 size=16386                -0.17 %       ±2.94%  ±3.93%  ±5.13%
buffers/buffer-compare-instance-method.js n=1000000 args=5 size=4096                 -0.62 %       ±2.00%  ±2.67%  ±3.47%
buffers/buffer-compare-instance-method.js n=1000000 args=5 size=512                  -1.16 %       ±3.34%  ±4.45%  ±5.79%
buffers/buffer-compare-offset.js n=1000000 size=16 method='offset'                   -0.75 %       ±3.85%  ±5.13%  ±6.68%
buffers/buffer-compare-offset.js n=1000000 size=16 method='slice'            ***     55.75 %       ±3.72%  ±4.97%  ±6.51%
buffers/buffer-compare-offset.js n=1000000 size=16386 method='offset'                 0.60 %       ±2.39%  ±3.19%  ±4.15%
buffers/buffer-compare-offset.js n=1000000 size=16386 method='slice'         ***     54.36 %       ±4.39%  ±5.85%  ±7.62%
buffers/buffer-compare-offset.js n=1000000 size=4096 method='offset'                 -0.17 %       ±2.52%  ±3.37%  ±4.41%
buffers/buffer-compare-offset.js n=1000000 size=4096 method='slice'          ***     53.48 %       ±4.26%  ±5.69%  ±7.44%
buffers/buffer-compare-offset.js n=1000000 size=512 method='offset'                   1.06 %       ±1.97%  ±2.63%  ±3.45%
buffers/buffer-compare-offset.js n=1000000 size=512 method='slice'           ***     49.36 %       ±5.02%  ±6.73%  ±8.88%
buffers/buffer-compare.js n=1000000 size=16                                  ***    136.22 %       ±6.78%  ±9.07% ±11.90%
buffers/buffer-compare.js n=1000000 size=16386                                 *      4.93 %       ±4.56%  ±6.10%  ±8.01%
buffers/buffer-compare.js n=1000000 size=4096                                ***     36.27 %       ±5.07%  ±6.79%  ±8.92%
buffers/buffer-compare.js n=1000000 size=512                                 ***    113.32 %       ±6.16%  ±8.27% ±10.92%
buffers/buffer-equals.js n=1000000 difflen='false' size=0                             0.39 %       ±2.76%  ±3.69%  ±4.84%
buffers/buffer-equals.js n=1000000 difflen='false' size=16386                ***     18.69 %       ±5.35%  ±7.20%  ±9.52%
buffers/buffer-equals.js n=1000000 difflen='false' size=512                  ***    141.31 %      ±10.23% ±13.77% ±18.22%
buffers/buffer-equals.js n=1000000 difflen='true' size=0                              0.23 %       ±2.81%  ±3.74%  ±4.87%
buffers/buffer-equals.js n=1000000 difflen='true' size=16386                 ***     -4.23 %       ±1.99%  ±2.66%  ±3.48%
buffers/buffer-equals.js n=1000000 difflen='true' size=512                           -1.28 %       ±4.89%  ±6.50%  ±8.47%

Be aware that when doing many comparisons the risk of a false-positive result increases.
In this case, there are 30 comparisons, you can thus expect the following amount of false-positive results:
  1.50 false positives, when considering a   5% risk acceptance (*, **, ***),
  0.30 false positives, when considering a   1% risk acceptance (**, ***),
  0.03 false positives, when considering a 0.1% risk acceptance (***)

Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1566/

[anonrig]

legend

[VoltrexKeyva]

Amazing!

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/59238/

[nodejs-github-bot]

Landed in 075853e

[ronag]

@tniessen any chance we could do the same for indexOf?

[tniessen]

@ronag Let's start with the simplest signature: #53455