winch(aarch64): Run wast test in CI

[saulecabrera]

Closes #9566

This commit takes inspiration from
https://github.com/bytecodealliance/wasmtime/pull/10738/files

During the 2025-05-07 Cranelift meeting we discussed potential avenues to start running wast tests for winch-aarch64. One potential idea was to declare an allowlist of tests that should be exectued and potentially ignore everyhing else. Although that idea works, I decided to diverge a bit from it, in favor of introducing a very strict classification of the state of tests for aarch64, namely, in this commit tests are classified as:

• Run and expected to pass
• Run and exepected to fail in a recoverable way
• Don't run since it's known that they produce an unrecoverable error e.g., segafault.

This approach is probably more verbose than the one discussed in the meeting, however, I think it's easier to have a global view of that status for aarch64 and potentially other backends in the future.

@alexcrichton curious to see what do you think.

[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.

[saulecabrera]

The failure in CI happens because wasmtime_test_util::wast::Compiler::should_fail gets invoked from the wasmtime_test macro, and as far as I understand, #[cfg(target_arch="...")] in macro code will resolve to the host, not to the target, which ends up returning true for should_fail, but since we're cross compiling (aarch64 -> x64), in reality the test passes and the #[should_panic] predicate is not met.

Prior to this change this was not an issue, since we were only taking x64 into account for Winch.

[alexcrichton]

Bah I was afraid of this happening at some point...

Could we perhaps delay this to happening at runtime? We've got the two structures there at test-run-time so we should be able to run the same code. I'd prefer to not catch panics so it means though that all tests would have to return Result<()> and the runtime result of should_fail would result in expecting an error or not, however.

[saulecabrera]

Yeah, I'll try that. I remember we had discussed this, I just wasn't sure if there was a strong preference regarding catching panics.

[alexcrichton]

Personally I'd say I have a weak preference, if that's the best solution maintenance-wise I'd go with catching panics

[saulecabrera]

Ok, ended up going with emitting the call to Compiler::should_fail at runtime. Would you mind taking another look?