Maxime Arthaud

Regarding free: It's not entirely clear to me either. It looks like this is an implementation detail in glibc for thread cancellation, see [man7.org](http://man7.org/linux/man-pages/man3/pthread_cancel.3.html). Apparently these exceptions have a specific...

Hi @PeterJCLaw, If you remove the `*` parameter in the first overload, this works properly. We don't handle correctly `*` parameters in overloads, this is a known issue.

Thanks @gh2375, I can reproduce, but only on Windows. A lot of code is pulled from the C++ standard library. The llvm bitcode after optimization is 2600 lines. I will...

Thank you @gh2375, this is very helpful. It looks like this is due to ABI requirements on Windows. The function `MakeIterator` returns an integer on Windows instead of returning a...

I tried the `-target` parameter in clang to change the target abi. It looks like it could work, but I found issues with it: * Sometimes, the libc headers cannot...

Yes it did :)

Question from @grievejia: could we fix this in typeshed directly?

Hi @OnlyOneCannolo, This is a bug in the APRON library. It has been fixed upstream. See https://gforge.inria.fr/scm/viewvc.php/apron?view=revision&revision=594 I opened an issue on the Ubuntu bug tracker a few months ago....

Great! Now we should wait for the release and for platforms to upgrade. We will need to rework the frontend llvm regression tests too. You just updated the llvm version...

This should happen whenever I get the time and approval from my company. I would recommend staying on LLVM 9 if possible since this patch hasn't been tested extensively.