On 7/11/2014 14:46, Damien Kick wrote:
> Actually, I was expecting that any error case would throw an
> exception with the code as I've written it; i.e. we would crash in a
> call to std::terminate but not a segfault. I may be mistaken in that
> regard, too, though.
An uncaught exception should crash the program with SIGABRT, not
SIGSEGV. That said, I suppose it's possible that the SIGSEGV is
happening during stack unwinding, before the SIGABRT can be raised.
It's easy to rule that out by wrapping the worker's body in a try block.