wasmtime: Implement Extended Constant Expressions proposal

[jameysharp]

This implements https://github.com/webassembly/extended-const for Wasmtime.

Currently the documentation comment for the wasm_extended_const configuration option says it's false by default, but Config::new sets it to true. One of those should be fixed and I'm not sure which. Other config-related things are still to-do as well, such as adding the appropriate CLI flag.

Also, I could use a little help understanding why one of the funcrefs spec tests fails:

Caused by:
    0: failed directive on tests/spec_testsuite/proposals/function-references/elem.wast:539:1
    1: assert_invalid: expected "constant expression required", got "failed to parse WebAssembly module
       
       Caused by:
           type mismatch: expected funcref, found i32 (at offset 0x1b)"

cc: @fitzgen

[github-actions]

Label Messager: wasmtime:config

It looks like you are changing Wasmtime's configuration options. Make sure to
complete this check list:

• If you added a new Config method, you wrote extensive documentation for
  it.

  Our documentation should be of the following form:

  Short, simple summary sentence.
  
  More details. These details can be multiple paragraphs. There should be
  information about not just the method, but its parameters and results as
  well.
  
  Is this method fallible? If so, when can it return an error?
  
  Can this method panic? If so, when does it panic?
  
  # Example
  
  Optional example here.
  

• If you added a new Config method, or modified an existing one, you
  ensured that this configuration is exercised by the fuzz targets.

  For example, if you expose a new strategy for allocating the next instance
  slot inside the pooling allocator, you should ensure that at least one of our
  fuzz targets exercises that new strategy.

  Often, all that is required of you is to ensure that there is a knob for this
  configuration option in wasmtime_fuzzing::Config (or one
  of its nested structs).

  Rarely, this may require authoring a new fuzz target to specifically test this
  configuration. See our docs on fuzzing for more details.

• If you are enabling a configuration option by default, make sure that it
  has been fuzzed for at least two weeks before turning it on by default.

---

To modify this label's message, edit the .github/label-messager/wasmtime-config.md file.

To add new label messages or remove existing label messages, edit the
.github/label-messager.json configuration file.

Learn more.

[alexcrichton]

For the failure here I think you'll need to set wasm_extended_const(false) in Config when running the tests. Spec proposals aren't always in-sync with one another and it takes quite some time to merge into the official spec, so tests often assert that a specific construct is invalid when another proposal specifically makes it valid, meaning that we have to be careful about enabled features when testing various proposals.

[fitzgen]

Yeah that failure is just because it is using an i32.add in a constant expression which used to be disallowed but is now allowed with the extended constants proposal.

We can ignore that particular failure by adding a clause here: https://github.com/bytecodealliance/wasmtime/blob/main/crates/wast/src/wast.rs#L544

Would be nice to ideally send a PR upstream to fix the issue (switch the add to some operator that isn't allowed even with the extended constants proposal) or at least make a local copy for the misc_testsuite that does that same thing.

[fitzgen]

Currently the documentation comment for the wasm_extended_const configuration option says it's false by default, but Config::new sets it to true. One of those should be fixed and I'm not sure which.

I think it is fine to have it enabled by default since it is a phase 4 proposal (basically done) and will have more than 2 weeks fuzzing before it hits a release.

[alexcrichton]

Oh I'll note that wasm-smith doesn't have support for extended consts right now I think, so AFAIK we won't fuzz this at all

[fitzgen]

Ah yes, we should add that support before enabling this proposal by default.