> If it hadn't been you'd not be saying any of the above. Well, yes, that's kind of the point. If GCC adopts that interpretation and gets a speed boost out of

> If it hadn't been you'd not be saying any of the above. Well, yes, that's kind of the point. If GCC adopts that interpretation and gets a speed boost out of it, then there's pressure on Clang to do the same, unless they can convince developers that not making such aggressive optimizations is worth it. "Your code may be slower so this (arguably) edge case behaves better, but you get better error messages/compile times" is not going to be as easy to sell as "Your code will behave the same and perform as well as when compiled with GCC, and you get better error messages/compile times to boot". > "Clang produces slow code" might be due to it not deleting "unpossible" code, but screw that, you can always go delete it yourself if it was dead. The entire point of the dead code elimination pass is to do that for you. If anything, that's why the optimizer exists; so you don't have to perform optimizations manually.