If we really want to test how forgiving the X86's memory model is, we can write multi-threaded code that skips any use of the std::sync::atomic ...
確定! 回上一頁