zturner added a comment. Well, it's not just an optimization. Threading code is hard to reason about, and the more complicated an implementation the more likely it is to have race conditions or other problems. So any opportunity to reduce the amount of manual thread management is a win in that sense.
Anyway, it was mostly a question / suggestion, not something I'm going to block over, so if everyone feels it's ok to go in this way then go for it. But if you read through the CL, it's not easy to be 100% confident that there are no race conditions. If all you've got is a single condition variable which wakes up and sends something to the standard library implementation, it's a lot easier to have that confidence. The optimization is just a side benefit. http://reviews.llvm.org/D13727 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
