Tests run

• Call one function - a benchmark timing how long it takes for the calling context to call a "thunk", a function that does nothing, and then return. The y axis is how long it takes to call the thunk N times. Run test
• Add two numbers - similar to "call one function" above, except this time the function we call simply adds its two arguments and then returns the result. Run test
• Call JS function - this benchmark times how long it takes a body of code to call out to JS, N times. Each benchmark will run once, and internally it will call a JS "thunk" (a function that does nothing) N times, timing how long it takes. Run test
• Call JS Add - same as the above JS function benchmark, except the function we're calling in JS adds two numbers and returns the result. Run test
• Calculate Fib(40) - calculates the 40th fibonacci number N times, measuring how long it takes Run test
• `structural` vs not - this is a benchmark of the #[wasm_bindgen(structural)] attribute and not. The baseline bar (labeled as JS) represents a function call in wasm through the prototype chain, using Function.call The wasm-bindgen bar (the second one) is structural Run test
• catch vs not - this is a benchmark of the #[wasm_bindgen(catch)] attribute and not. The baseline bar (labeled as JS) represents a function call in wasm which does not use catch. The second bar (labelled wasm-bindgen) is calling a function in wasm-bindgen using catch but the function doesn't actually throw. (same as baseline but catching an exception if one happens to happen). Run test
• Node.firstChild Run test
• Node.nodeType Run test
• Node.hasChildNodes Run test
• count all node types on page Run test