Aidan Hobson Sayers

@balajisivaraman that sounds great - I'm certainly willing to review and merge if you make a PR :)

At minimum we should allow (say) 100MB in `/tmp`.

Got a link? Usually if you're generating source files you can use `OUT_DIR` to not pollute your source dir, e.g. https://github.com/sfackler/rust-phf#phf_codegen (was also talking about this to @tglman who creates...

> I assume cargobomb will work in that case?, as OUT_DIR is a place which must be writable for any compilation to work at all? Yup, `OUT_DIR` is under `target`...

Great to see your interest here! > 1. Remove as much as possible from the prepare-ex stage of a crater run: everything crate-specific will be prepared just before that crate...

Option 1 for now is totally fine - in time we'll move to it being team based, but I want to have appropriate guard rails in place first (e.g. making...

Very interesting, the rust issue appears to be this line: `thread 'main' has overflowed its stack`. I'm 99% sure I know what the issue is (a very deeply nested set...

Interesting! The ayzim problem will be because of an incredibly deeply nested expression/set of conditionals in a function, and I assumed that a 'callstack size exceeded' would be due to...

The old optimizer doesn't actually free memory, it just leaks it. So given something like `x+y` which becomes something like `BinaryOp("+", Name("x"), Name("y"))`, the old optimizer will do nothing, whereas...

There's very little recursion in the old optimizer, courtesy of the `traverse*` functions in [simpleast](https://github.com/kripken/emscripten/blob/cd7396e5c75f77e29991b9db64639363e6e887b7/tools/optimizer/simple_ast.cpp#L132) which keeps function node traversal to a fixed shallow depth. The one place that I...