Skip to content

benchmark: refactor path benchmarks

Rodrigo Muino Tomonari requested to merge github/fork/BridgeAR/refactor-path-benchs into master

So far the benchmarks created a highly specialized function which would inline exactly to the input. This changes it to provide a more realistic view to actual input by changing the input on each iteration. That prevents the function to be to specific.

It also reduces the number of iterations the benchmarks are run to reduce the overall runtime. A microbenchmark should already show a significant difference with lower iterations, otherwise the significance for real world applications is only limited.

This is pulled out from #25278.

Checklist
  • make -j4 test (UNIX), or vcbuild test (Windows) passes
  • tests and/or benchmarks are included
  • documentation is changed or added
  • commit message follows commit guidelines

Merge request reports