[GHC] #367: Infinite loops can hang Concurrent Haskell

[GHC]

#367: Infinite loops can hang Concurrent Haskell
------------------------------------------+---------------------------------
  Reporter:  simonpj                      |          Owner:  ezyang              
      Type:  bug                          |         Status:  new                 
  Priority:  lowest                       |      Milestone:  _|_                 
 Component:  Compiler                     |        Version:  6.4.1               
Resolution:  None                         |       Keywords:  scheduler allocation
        Os:  Unknown/Multiple             |   Architecture:  Unknown/Multiple    
   Failure:  Incorrect result at runtime  |     Difficulty:  Unknown             
  Testcase:                               |      Blockedby:                      
  Blocking:                               |        Related:                      
------------------------------------------+---------------------------------

Comment(by simonmar):

 What you've done here is omit the yield checks on (some) case
 alternatives.  Which is almost ok, because all loops will go through a
 function entry, except for let-no-escapes which use `altHeapCheck`.  So
 with this patch you won't catch non-allocating recursive let-no-escapes.

 To make this correct you need to ensure that let-no-escapes get a yield
 point too, but you could also omit the yield points from ordinary
 `altHeapCheck`s which should reduce the code size hit further.

 The confusion probably arose because of my naming scheme:
 `altHeapCheckReturnsTo` is the heap check for a case alternative where the
 case scrutinee made an external call, and hence had a "returns to"
 continuation that we can re-use for the heap check's continuation.

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/367#comment:26>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler