Require T: 'static for Store, upgrade to wasmtime_runtime_layer v34.0.0

[juntyr]

With bytecodealliance/wasmtime#10760, wasmtime now requires T: 'static for store. While we advocated to not have the bound in wasmi, it now seems like the ship has sailed and we have to adapt.

@alexcrichton Since you also tried to avoid a static bound for as long as possible, do you think it makes sense for our abstraction over several WASM runtimes to adopt the bound?

@Robbepop If we go forward here, it means you can bring back the static bound in wasmi as well - thank you for removing it a few versions ago to help our downstream crate, now the momentum is in favour of your original change.

@DouglasDwyer This will require a major release for wasm_runtime_layer (and wasm_component_layer). I've prepared updates for wasmtime_runtime_layer and js_wasm_runtime_layer in this PR, I think for wasmi and wasmer we just wait until their next release.

[juntyr]

The nightly error seems to come from a wasmer regression on nightly, see wasmerio/wasmer#5610

[Robbepop]

@juntyr thanks for the CC. I wonder why Wasmtime now requires the 'static. Funny coincidence that Wasmi briefly introduced T:'static some weeks prior.

I will not re-introduce T:'static in Wasmi for now, as its not required and the most recent changes even improved performance.

[alexcrichton]

I tried my best to avoid adding it to Wasmtime but 🤷. I am but a slave to rustc compiler error messages in the end. The tl;dr; is that this helps avoid a lot of unsafe transmutes within Wasmtime and while we think it may be a bug in the Rust compiler we value safety in the interim higher for now.