MartinNowack

Maybe we could do the following: First, renaming `--optimize` to `--optimize-module`. Second, add a meta switch (`--optimize`) that enables *all the good stuff* for performance. Any thoughts?

@illera88 How did you compile and run your application?

@ccadar Everything works as expected ;) The reason why this still passes is that we explicitly ignore this issue: https://github.com/klee/klee/blob/master/utils/sanitizers/lsan.txt

@kren1 Not for test-comp.

@operasfantom Sorry for the delay of the review of the patch. I went through your changes, they are of excellent quality! From a design point of view, we might be...

@operasfantom Thanks a lot work your work. For the library implementation - yes, similar to the POSIX handling would be great. For the `_ubsan_*` functions, do you have an idea...

> @MartinNowack I used to know 6 months ago 😄 > 😄 Good point! > As far as I remember, the behavior of `__ubsan_handle_pointer_overflow` had been changing. Although mainly new...

@operasfantom Sorry for my delay. The changes look great. To rather avoid any delay, I would rather like to have this PR in without any additional features added or major...

Great changes! I added one comment. The plan is to get this PR in before #1446 is merged. @operasfantom Thanks a lot for all your effort!

@operasfantom You mean those ones? > Some important notes with regard to the future of the process: > > UBSan runtime is always built now. At the same time, however,...