Tim Ledbetter

LibCore: Introduce `REPORT_TIME()` for per call timing information

1

This macro outputs the time taken to reach the end of the current scope to the debug console. It also shows the average time, total cumulative time and the total...

LibWeb: Use a faster string search algorithm for content filtering

1

This PR improves the performance of `ContentFilter` by using the [Aho-Corasick algorithm](), which allows multiple strings to be checked simultaneously. Previously, we were checking each string one at a time....

This reduces time spent in `StyleComputer::compute_property_values()` by ~10% when loading https://html.spec.whatwg.org/

This reduces the time spent in `SelectorEngine::matches_lang_pseudo_class()` from 1.90% to 0.41% on https://cloudflare.com

This reduces time spent in `~StyleProperty()` from 2.18% to 0.67% on https://cloudflare.com.

This headless mode will not automatically output anything and runs until `internals.signalTestIsDone()` is called. I'm using this to run Speedometer headlessly.

This filter primitive is used to erode or dilate an image. Motivated by seeing some debug spam on https://shopify.com. This change gives us +23 WPT subtest passes in `css/filter-effects`. It...

This test fails fairly often on CI. The most common failure seems to be because we are not invalidating the image state correctly when changing the `src` / `srcset` attribute...

This VERIFY was particularly expensive and removing it gives a significant speedup on multiple benchmarks on my machine. ``` Suite Test Speedup Old (Mean ± Range) New (Mean ± Range)...

This yields a ~3% performance increase across the board in JS benchmarks. ``` Suite Test Speedup Old (Mean ± Range) New (Mean ± Range) ---------- ----------------------------------- --------- --------------------------------- --------------------------------- Kraken...