5 Commits (a01f8d7f28aadaebb516503e6fad31e45dc3da57)

Author	SHA1	Message	Date
Simon H	af6f1d309b	chore: run every benchmark in its own process (#18542) ... While benchmarking it became apparent that when running the full `pnpm bench:compare` suite results could be wildly different compared to only comparing single benchmarks. The reason (most likely) is that the heap/GC/JIT state from one benchmark contaminates the others. Therefore this adjusts the runners such that each benchmark entry is run in its own forked process, not just each branch. This should give better, more stable results. Also a little fix to compare results by name; I made the mistake of adding a new benchmark to main which wasn't present on other branches yet, which made the results really confusing.	3 weeks ago
Rich Harris	99b1467ba4	chore: generate CPU profiles when running benchmarks (#18043) ... This causes `pnpm bench` and `pnpm bench:compare` to generate `.cpuprofile` files for each benchmark, which should in theory make it easier to understand how different branches affect performance (and find opportunities for optimisation). These files can be opened directly in VS Code and other editors, or in [profiler.firefox.com](https://profiler.firefox.com), [speedscope.app](https://www.speedscope.app), Chrome's performance devtools and so on.	4 months ago
Rich Harris	0025d42366	chore: skippable benchmarks (#17506) ... * chore: add benchmark filtering mechanism * tweak * fix * try this * better logging for bench:compare * fix	7 months ago
Rich Harris	c7ce9fc004	fix benchmarks (#15560)	1 year ago
Rich Harris	331925356e	chore: add `pnpm bench:compare` script (#12099) ... * WIP compare script * WIP * fix * add script * update * fix * tweak * fix * update * add TODO * compare * tmp * gah * revert * fix * fix * prevent div/0 * reenable benchmarks --------- Co-authored-by: Dominic Gannaway <dg@domgan.com>	2 years ago