Move re-initialization logic into Rust/wasm-bindgen

[logan-gatlin]

This PR moves (most of) the responsibility for re-initializing Rust workers away from our custom JS shim, and into Rust/wasm-bindgen.

Latest wasm-bindgen handles re-initialization automatically when building for the panic=unwind target. This means that most of the shim we provide for WASM modules is unnecessary. This PR creates a new minimal shim for panic=unwind builds.

Durable objects require slightly more work since their state must be reset on an abort. Previously, durable objects were heap allocated Rust structs managed by JS objects which kept pointers into Rust's memory. This is problematic when the instance resets, as these pointers become dangling.

The new design is to keep a global HashMap in Rust containing all live DO instances, keyed by a unique u32. In our JS shim, we maintain a list of live DO ids and the arguments used to construct them. When re-initialization happens, every entry in this list is re-initialized.

Notes:
This requires a new feature in wasm-bindgen allowing for a post-reinitialization hook, right now I am just pointing the submodule to my feature branch. Both Rust and wasm-bindgen needed to be bumped to their latest version so I could write panic=unwind tests. This makes the diff messy because many doc tests subsequently failed to build, and CI had to be modified to support this new setup. When the wasm-bindgen reinit hook is merged I will probably split this up into two PRs

[codspeed-hq]

Merging this PR will not alter performance

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

✅ 2 untouched benchmarks

---

_{Comparing logan/remove-reinit-for-unwind (ac104bc) with main (2d022ab)}

[guybedford]

Would be nice to see how this integrates with the upstream branch of supporting StaleError and if that is sufficient to support the instance checks.

We also need to ensure unwind test coverage here.

[logan-gatlin]

This PR broke doctests somehow, probably because I bumped the versions of a lot of things. Had to go in and add ignore and norun in places

[logan-gatlin]

I updated this to use the new fork of reinit-hooks wasm-bindgen/wasm-bindgen#5059