Reply to d-bugmail@puremagic.com,

> Even if it's not that, it's a different issue, since the code here is
> correctly failing on a null pointer dereference.

I'm not shure that is where the seg-v is. But I'm to lazy to check right now.