arsnyder16

@kripken When i am running the fuzzer on this branch i am receiving ``` You found a bug in the fuzzer itself! It failed to generate a valid wasm file...

Thanks, I’ll put that branch and my changes together and rerun

i am still hitting the `bug in the fuzzer` but less frequent `10616326470480632236` `1630626949815546522` `3487257774816750057`

I think the fuzzer is finding something with this branch as well, looking into that now, not seeing it with main

I think the reason is with what i am testing is there are 130k functions, and so doing a big merge at the end is where i was seeing a...

@kripken Just to make sure we are talking about the same thing the "merge" is the following code. ``` // Combine all the info. std::map allCalls; for (auto& [_, info]...

Also since #4629 was rolled back, I attempted to fix that issue, i have a proposed solution, but i am struggling to understand the exact issue, so i am not...

@kripken What is the total runtime of those wasm files with and without my change? I realize there is still an issue to work out but curious if the net...

One theory is my library has some boost spirit which is very templated and can result in large class and function names which could play into the hashing theory side...

I agree about the iterations, my hesitation with just capping it is two fold. One since we can only apply all dead argument removal or return value removal in one...