Add test exception with message

[facug91]

Solves issue #130

[facug91]

I tested this on Ubuntu 20 and it worked fine. I seem to be missing something for Windows 😅. Sorry, I'll check this out later.

[sheredom]

Windows seems fine - but there is some UBSAN issue.

[facug91]

In my computer the sanitizer is failing even in master. In the CI it is failing with use-of-uninitialized-value, but I just added message_caught which is initialized with the value UTEST_NULL, and message_size which is initialized with strlen(e.what()) + 1. I can't find the problem.
I've been reading on stackoverflow that MemorySanitizer sometimes returns false positives for this, but I'm not sure this is the case. Can you help me please @sheredom? I am a bit lost.

[sheredom]

Trying to work this out at the minute in the other PR (with some choice language, its just for testing as I don't have Linux locally setup at present!) #132

[facug91]

I see in your commits in the other PR, that it seems to be a problem with the compiler. I made a commit to disable the test like you did, but only for test.cpp, which is the file that was failing here. If I see that it also fails in the others, I will disable it in those files as well. Thank you!

[sheredom]

Yeah @facug91 it seems like this is the same GLIBC cstdlib sanitizer bug - pretty hard to track down!

[sheredom]

Good work getting this through!