Chris Fallin

@afonso360 just noted on that issue -- I agree with the proposed fix to i128-constant semantics.

Thanks for letting us know @glaebhoerl! I haven't gotten to the point in the egraph exploration yet where I'm doing detailed performance evaluation and tuning, but when I do, I'll...

This would be a nice simplification, I think. One potential issue to check for would be whether `InstructionData` grows as a result of two branch targets rather than one. Otherwise...

I tend to agree, but downstream packagers want this -- @decathorpe in #3761 needs LICENSE files in each crate (or at least the published ones) for packaging in Fedora (IIRC?)....

@decathorpe understood that we need to include the license file, one way or another, in the tarball that is uploaded to crates.io. I think this is covered by what @alexcrichton...

This looks like rust-lang/cargo#8537, which is closed with a "we won't support this" disposition -- see [this comment](https://github.com/rust-lang/cargo/issues/8537#issuecomment-924329183) in particular. Their reasoning is that adding both keys (`license` and `license-file`)...

@decathorpe I've modified this PR to add LICENSE files just to the six (!) crates we have added recently without LICENSE files, and to add a check to the `publish.rs`...

@decathorpe could you respond to the question: > Unfortunately it looks like wasi-nn is missing a LICENSE file as well, and that's an upstream repo that would need a separate...

> So far my only complaint is that the 16-character limit on function names forces changing a lot of tests, when I would think it'd be better to just increase...

This should work: the `?` (try-operator) at the call to `build()` will bubble any errors up from `run_test()` to the toplevel fuzzing entry point, and then the `.expect("build with VeriWasm...