ed added a comment. A general note I have regarding this change:
Now that we're introducing separate implementations for mutexes and condition variables, could we also consider letting `shared_mutex` and friends simply use `pthread_rwlock_*()`? We currently have it implemented as a wrapper on top of mutexes and condition variables, but the downside of this approach is that it potentially has more overhead, but also works around priority inheritance if supported by the implementation. Repository: rL LLVM http://reviews.llvm.org/D11781 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
