Robin Freyler

The `wasmi_v1` crate is the one with the strange performance. The `wasmi` crate is the root of the workspace and depends on `wasmi_v1`. As stated above I tried both: Putting...

> The `Cargo.toml` of the `wasmi` crate already defines `lto = "fat` and `codegen-units = 1` for the `bench` profile: > > https://github.com/paritytech/wasmi/blob/70faf3410c4f0017f70c59041fa6a1d608562969/Cargo.toml#L61-L63 Yes that is true. On my local...

> A lot of the difference seems to be inlining. This patch already improves performance 40% for me in the default profile: > Click to expand Hi @tavianator and thanks...

> We will build the runtime with `cgu=1` and we will be fine and don't need those. Don't think we want to sprinkling inline everywhere is a sane approach. Optimizer...

@raravena80 Sorry for the late reply. What information do you guys need about `wasmi` and is this still needed?

I am no expert in cloud computing myself so I'd need to know what the requirements of cloud computing are with respect to interpreters such as `wasmi` before I can...

Yes, `wasmi` is a Wasm interpreter and is able to execute Wasm modules. I am not entirely sure what exactly you mean with "dry running" but form what I understood...

At this point I cannot give a guarantee. Right now we are working on a big update for `wasmi` that will take some more time. It won't make sense to...

https://github.com/paritytech/wasmi/issues/358 is probably important for this. Can you confirm this @raravena80 ?

With https://github.com/paritytech/wasmi/pull/346 merged we are one step closer to implementing this.