new SUPER FUN blog post about threads, friends! Diving into concurrency: trying out mutexes and atomics http://jvns.ca/blog/2014/12/14/fun-with-threads/ …
-
@b0rk That the instruction after the lock gets charged all the cycles for the lock is an artifact of perf that's kinda bogus.
@bork The lock prevents later instructions from executing until the lock is retired, which makes the next instruction seem SUPER SLOW
-
-
@bork Since it's just sitting in a queue idling from when it's issued until the locked instruction retires it counts as taking a lot of time -
@danluu ah ha that explains everything! I feel like you explained this to me before already - 5 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.