[Haskell-cafe] Re: Waiting for thread to finish

ChrisK haskell at list.mightyreason.com
Thu Dec 6 12:52:52 EST 2007

Jules Bean wrote:
> ChrisK wrote:
>> A safer gimmick...
>> Ben Franksen wrote:
>>> tickWhileDoing :: String -> IO a -> IO a
>>> tickWhileDoing msg act = do
>>>   hPutStr stderr msg >> hPutChar stderr ' ' >> hFlush stderr
>>>   start_time <- getCPUTime
>>>   tickerId <- forkIO ticker
>> ... an async exception here will leave the ticker runnning....
>>>   res <- act `finally` killThread tickerId
>> The best way to make this safe that I know of is:
>>>   res <- block $ do
>>>     tickerId <- forkIO ticker
>>>     unblock act `finally` killThread tickerId
> ...but with a change that Simon M just checked in to GHC head, this will
> now spawn 'ticker' in blocked state, so you won't be able to kill it.
> You would therefore want unblock $ forkIO ticker or forkIO $ unblock ticker
> I'm not sure if there is a strong reason to prefer one over the other.
> Jules

That is new. Ah, I see GHC.Conc.forkIO now has a note:

> GHC note: the new thread inherits the /blocked/ state of the parent 
> (see 'Control.Exception.block').

BUT...doesn't this change some of the semantics of old code that used forkIO ?

I wanted a way to control the blocked status of new threads, since this makes it
 easier to be _sure_ some race conditions will never happen.

And so my new preferred way of writing this is now:

> -- we are in parent's blocked state, so make the ticker explicit:
>   res <- bracket (forkIO (unblock ticker))
>                  killThread
>                  const act  -- act runs in parent's blocked state


More information about the Haskell-Cafe mailing list