[wasm][mono] conflicting naming for wasm import/export

[lewing]

        [UnmanagedCallersOnly(EntryPoint = "test:variants/test#variant-zeros")]
        public static unsafe nint wasmExportVariantZeros(int p0, int p1, int p2, long p3, int p4, float p5, int p6, double p7) {

            [DllImport("test:variants/test", EntryPoint = "variant-zeros"), WasmImportLinkage]
            internal static extern void wasmImportVariantZeros(int p0, int p1, int p2, long p3, int p4, float p5, int p6, double p7, nint p8);

is generating code that doesn't match the assumptions wit-bindgen is making about the UCO signatures, we will need to verify correctness here

__attribute__((import_module("test:variants/test"),import_name("variant-zeros")))
extern void test_3A_variants_2F_test_23_variant_zeros (int32_t, int32_t, int32_t, int64_t, int32_t, float, int32_t, double, void *);


typedef void (*WasmInterpEntrySig_10) (int*, int*, int*, int*, int*, int*, int*, int*, int*, int*);
__attribute__((export_name("test:variants/test#variant-zeros")))
void * test_3A_variants_2F_test_23_variant_zeros (int32_t arg0, int32_t arg1, int32_t arg2, int64_t arg3, int32_t arg4, float arg5, int32_t arg6, double arg7) { 
  void * res;
  if (!(WasmInterpEntrySig_10)wasm_native_to_interp_ftndescs [10].func) {
   mono_wasm_marshal_get_managed_wrapper ("csharp-wasm", "VariantsWorld.wit.exports.test.variants.TestInterop", "wasmExportVariantZeros", 8);
  }
  ((WasmInterpEntrySig_10)wasm_native_to_interp_ftndescs [10].func) ((int*)&res, (int*)&arg0, (int*)&arg1, (int*)&arg2, (int*)&arg3, (int*)&arg4, (int*)&arg5, (int*)&arg6, (int*)&arg7, wasm_native_to_interp_ftndescs [10].arg);
  return res;
}

It looks like there is an import and an export with different signatures because the return value is being marshaled differently?

[dotnet-policy-service]

Tagging subscribers to 'arch-wasm': @lewing
See info in area-owners.md if you want to be subscribed.

[mkhamoyan]

@lewing I've reviewed our code, and the signature appears to be correct. However, I'm uncertain why wit-bindgen chooses to move the return type as a parameter. Is this behavior expected?

[lewing]

For some background I was running the wit-bindgen tests against the interpreter using bytecodealliance/wit-bindgen#958 and cargo test -p wit-bindgen-cli --no-default-features -F csharp-mono it looks at first glance like we're handling the nint return type differently than nativeaot but I haven't had time to look more closely

cc @lambdageek @maraf @kg

[lewing]

Looking closer this appears to be an issue in bindgen not mono