Helper changes for symbolic testing

[denis-bogdanas]

No description provided.

[denis-bogdanas]

@ehildenb Except one question above this is ready for review. Should pass the tests.

[ehildenb]

It looks pretty good to me, but can @daejunpark look and @virgil-serbanuta as well?

[virgil-serbanuta]

This seems reasonable, I have one small question.

[denis-bogdanas]

@ehildenb @daejunpark All comments addressed. I see it passes now. No idea why it was failing before. Changes are unrelated to failure. Ready for re-review.

[ehildenb]

Please don't merge the update submodule PR into this one @denis-bogdanas . Instead, you can just approve that PR and get it merged directly, then merge master into this one.

If other PRs update the submodules, then the submodule updater gets confused, it can't handle merge conflicts at all.

[denis-bogdanas]

This PR would pass tests with latest K, but K update PR fails for other reasons as you can see. That one must be fixed first.

[virgil-serbanuta]

I still think that it would be better to ask the backend to resolve that ceil, but I guess that this is also fine.

[denis-bogdanas]

@ehildenb We need updated K for this PR to pass. Other than that it looks ready.

[ehildenb]

We're waiting on a regression in the LLVM backend to update K. I'll keep you posted, thanks!

[denis-bogdanas]

@ehildenb ready for review