Fix hanging when solver exits with a non-zero exit code

[atomb]

Previously, if the solver quit abruptly with a non-zero exit code (such as for a segmentation fault), Boogie would hang. These changes turn that into a well-behaved solver exception.

[keyboardDrummer]

I'd ask you to avoid calls to Console., but I see this file already contains them ^,^

[atomb]

Yeah, we should refactor it to use something more parameterized.

[keyboardDrummer]

Looks amazing. How did you find this?

Also, I added a commit to increment the version.

[atomb]

Looks amazing. How did you find this?

It was a rather frustrating bout of debugging, I must say! I've known that this was an issue for a while, and I expected that the SolverExited code would be run when the solver had a segfault, so I kept putting either breakpoints or print statements there and never having them executed. So, IIRC, I think I read a bunch of the documentation for Process and came across the EnableRaisingEvents property. I actually created this branch a month or so ago and held off creating a PR while I figured out how to create a test that intentionally crashed without needing to compile some C code in the test suite. And then I realized it didn't need to segfault, just return non-zero, so /usr/bin/false does the trick.

Also, I added a commit to increment the version.

Thanks!