Replying to [comment:6 rrnewton]:
By the way, I started peeking at the benchmark code. I don't have a good sense currently of how many cycles each of the following takes in
#8972: Investigate adding fast compare-and-swap Int type/primops -------------------------------------+------------------------------------ Reporter: tibbe | Owner: tibbe Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.9 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: 8157, 7883 -------------------------------------+------------------------------------ Comment (by jberryman): Replying to [comment:7 jberryman]: practice:
* between a forkIO and the new thread waking up on another CPU.
.......
Oops sorry! Here I meant to write that I'd sort of benchmarked this, but just added a better test (fork a `putMVar`, then block on `takeMVar` in main thread) which I measure at between 5.4 - 5.8 μs . So as long as you can get your contention test runs into the milliseconds range those effects should be completely hidden. Feel free to email me if this is distracting from this ticket and you want to chat further. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8972#comment:8 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler