Keith Winstein

Those test changes don't look like a problem, but we should probably update the testsuite in a separate PR... For the non-wasm2c tests, it looks like they just need to...

This appears to be accurately (but clumsily) printing "unimplemented" because reference types are unimplemented in the current wasm2c (but wasm2c uses the default validator settings). Will be fixed by #1887...

Thanks very much for this. AFAIK support for function-references remains a work in progress in WABT -- we're not running the function-references tests and I don't think we really support...

lgtm. An update that runs update-spec-tests.py is at #2001; this brings in the new tokens.wast test, which requires a change to the Wast lexer to get everything passing.

wasm2c currently aborts on any SIMD instructions or types. Probably it should configure the validator to disable SIMD (or we should implement SIMD in wasm2c).

@aardappel Would you be able to take a look at this (along with #1989, #1990, #1991)?

I can't claim to speak for everybody, but my own tentative view is that if you can test it once on an emulated big-endian machine and once on an actual...

Unfortunately I don't know the answer for s390x or i686 with sse2. Some of the others probably know the history much better (@binji @sbc100 @kripken). I think for now, my...

If we're going to auto-close the ARM Mac bugs, I would be a lot happier if we can get at least one person to test this and report that this...

Ok, good enough for me -- thanks for testing. Support for `-frounding-math` in LLVM on arm64 seems to be semi-tracked at https://github.com/llvm/llvm-project/issues/54975