Around the time of the recent LLVM merge we’ve started seeing sporadic timeouts on a couple of the stdlib tests. I don’t think it was related to resilience (it was also re-enabled at around the same time) but I could be wrong. In any case I haven’t been able to reproduce it locally. Could be a miscompile in LLVM or SIL optimizer, or we’re relying on some UB that newer LLVM optimizes differently?