Yices invocation almost always fails

[atomb]

A recent change in What4 led to the Yices backend erroneously thinking that Yices didn't support bit vectors. This causes some of the saw-script tests to fail. It looks as though @robdockins has fixed the issue, though, so we just need to update the crucible submodule.

[brianhuffman]

Do any of the integration tests in intTests use w4_unint_yices? If not, we should add one. I did run all the integration tests before I merged the last update-submodule-dependencies pull request.

[atomb]

Huh. Strange. Several of the intTests tests were failing on Jenkins overnight. I'll check locally.

[brianhuffman]

I'll try it again too. It's possible that I accidentally had the wrong submodule version checked out when I ran the tests.

[brianhuffman]

Yes, I'm seeing the failures now. I must have made some mistake yesterday when running the tests.

[brianhuffman]

I think what happened is that I actually ran the right tests, looked at the end of the output, and didn't realize that there were any failures. I opened #603 in the hope that this won't be such an easy mistake to make in the future.

[brianhuffman]

Fixed as of 4b0019b.