Evgeniy Istomin

I can propose some optimizations to for-of loop: 1. Array(Iterator) 1.1. ArrayIterator's builtin functions remain untouched. 1.2. cache length property if (array is frozen || array's length is not mutated)...

I know that previous results are from interpreted mode, but here are results when same code is run in dynapogo mode (= with full JIT?): ``` Start 45390ms (for-of) 12ms...

Will IIFE work or i need named function to do that?

I still get 3-4x increase in dynapogo mode using flags Code: ```js { const a = Array.from({ length: 1e6 }, _ => 0); function test(a) { let start = (console.log('Start'),...

I still get 2.5x increase in dynapogo mode, tested on x64_test build: ``` ======================Interpreted====================== 754.0545997619629ms (for-of) 48.02579975128174ms (for-cached) 78.80779981613159ms (for-non-cached) =======================Dynapogo======================= 1784.9175000190735ms (for-of) 12.22700023651123ms (for-cached) 13.011600017547607ms (for-non-cached) ```

> Is it possible that the time you're recording includes time spent jitting a loop body? After addiction of -off:jitloopbody flag on run, which is run in interpreted mode, it...

You need to feed off the `arguments` variable of JsCallFunction with an array where first must be set to an object or invalid reference as it controls the `this` value...

I can reproduce this error with property "name" aswell, Here is the code: ```js do { delete isNaN.name; Object.getOwnPropertyNames(isNaN); } while (1); ``` stderr is same: ``` ASSERTION 9816: (...\ChakraCore\lib\Runtime\Types\SimpleTypeHandler.cpp,...

Sure, i'll open a PR in a couple of days.

I have opened this issue only because when I add your mod into my modpack I can reproduce this issue. Also I see this issue **when stage 3 of minecraft...