JDevlieghere added a comment. In D81612#2086939 <https://reviews.llvm.org/D81612#2086939>, @labath wrote:
> I'm pretty indifferent about this functionality -- I don't think it hurts, > but it also doesn't seem like a pressing problem that needs addressing. Yeah. I agree it's not a problem we face right now, but the reproducer scenario showed that we can get into a situation where we're not testing what we thing we are. I took me and Jim a little time to understand why that was happening and I think this is a small price to pay to avoid that in the future and guarantee some sanity. > Regarding the implementation, be aware that assertion failures during test > teardown are reproted pretty weirdly -- IIRC at this point the test has > already been declared "successful", and these failures manifest as "CLEANUP > ERROR"s somewhere. And I have a feeling they don't actually fail the > check-lldb command. I don't know if this is due to something we've done, or > if it's just how the python unittest framework works... Okay, I was on the fence between using the unittest2 assert method or the built in Python assertion. The latter might be a better fit. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D81612/new/ https://reviews.llvm.org/D81612 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
