We Make a std::shared_mutex 10 Times Faster Atomic operations and C++11 memory barriers and assembler instructions generated on x86_64 CPUs Click here to read the article