[GHC] #9321: Support for waiting on multiple MVars

[GHC]

#9321: Support for waiting on multiple MVars
-------------------------------------+-------------------------------------
              Reporter:  schyler     |             Owner:  simonmar
                  Type:  feature     |            Status:  new
  request                            |         Milestone:
              Priority:  normal      |           Version:  7.8.3
             Component:  Runtime     |          Keywords:  mvar
  System                             |  Operating System:  Unknown/Multiple
            Resolution:              |   Type of failure:  None/Unknown
Differential Revisions:              |         Test Case:
          Architecture:              |          Blocking:
  Unknown/Multiple                   |
            Difficulty:  Unknown     |
            Blocked By:              |
       Related Tickets:              |
-------------------------------------+-------------------------------------

Comment (by simonmar):

 I think it's really really hard to add support for multi-MVar operations.
 So I'm with Simon: we should put our efforts into improving STM instead.

 On the fairness point, MVar's fairness guarantee means that MVar actually
 performs quite badly under contention, due to the excessive number of
 context switches.  Have you measured your application using both MVar and
 STM?  STM transactions with only a few TVars perform quite well in my
 experience, and tend to outperform MVar under heavy contention.

 Nevertheless, if you want to go ahead with this, you'll need (a) a clear
 description of the desired semantics, and (b) a detailed implementation
 plan.  I think you'll encounter lots of problems trying to do both of
 these.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9321#comment:8>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler