increase benchmark rounds to 10

[Uzlopak]

We need to increase the benchmark rounds to 10 to get realistic benchmarks. After 4 rounds it optimizes for maximum speed.

uzlopak@uzlopak-Lenovo-Legion-5-17ARH05H:~/Workspace/fastify/busboy$ npm run bench:dicer

> @fastify/busboy@1.0.0 bench:dicer
> node deps/dicer/bench/dicer-bench-multipart-parser.js

1515.15 mb/sec
1351.35 mb/sec
1333.33 mb/sec
2127.66 mb/sec
2127.66 mb/sec
2127.66 mb/sec
2127.66 mb/sec
2127.66 mb/sec
2173.91 mb/sec
2127.66 mb/sec

Checklist

• run npm run test and npm run benchmark
• tests and/or benchmarks are included
• documentation is changed or added
• commit message and code follows the Developer's Certification of Origin
  and the Code of conduct

[kibertoad]

Would be nice to move the whole benchmark thing over to some proper benchmarking engine, but this is a welcome improvement until that happens :)

[kibertoad]

linting is failing

[kibertoad]

I wonder if benchmark needs to be expanded to be run on less static data, otherwise its results are actually going to be less accurate due to optimizer tightly optimizing specifically for the one hardcoded path in the benchmark, e. g. completely disregarding potential megamorphism that would be present in real-world scenarios.