Question about withMVar

[David Feuer]

It's pointing out that it's possible that after it takes the MVar some
other thread could put the MVar. That could cause it to block, or could
violate assumptions about how the MVar value changes. The basic rule is
that whenever you're using an MVar you need all the threads to agree to
some particular discipline. Most commonly, and best supported by the API,
is "take, then put". However, there are other patterns, the most important
of which is probably "put once", used to communicate thread completion.

On Sun, Jul 18, 2021, 9:49 AM Harendra Kumar <harendra.kumar at gmail.com>
wrote:

> The documentation of withMVar (in Control.Concurrent.MVar module of base
> package) says:
>
> {-|  'withMVar' is an exception-safe wrapper for operating on the contents  of an 'MVar'.  This operation is exception-safe: it will replace the  original contents of the 'MVar' if an exception is raised (see  "Control.Exception").  However, it is only atomic if there are no  other producers for this 'MVar'.-}withMVar :: MVar a -> (a -> IO b) -> IO bwithMVar m io =  mask $ \restore -> do    a <- takeMVar m    b <- restore (io a) `onException` putMVar m a    putMVar m a    return b
>
> Can someone shed some light on what is meant by the statement -
> "However, it is only atomic if there are no other producers for this 'MVar'."?
>
> I hope this is the right mailing list for this question.
>
> Thanks,
> Harendra
>
> _______________________________________________
> Libraries mailing list
> Libraries at haskell.org
> http://mail.haskell.org/cgi-bin/mailman/listinfo/libraries
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/libraries/attachments/20210718/521e5e63/attachment.html>