[GHC] #14610: newtype wrapping of a monadic stack kills performance

[GHC]

#14610: newtype wrapping of a monadic stack kills performance
-------------------------------------+-------------------------------------
        Reporter:  mrkkrp            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.6.1
       Component:  Compiler          |              Version:  8.2.2
      Resolution:                    |             Keywords:  JoinPoints
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Runtime           |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #14620            |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by simonpj):

 > So am I right to say “we forbid polymorphic return types because we
 cannot prove it to be sound, even though we know it would not actually
 crash if we did allow it”?

 Yes, that's right. But be specific about what you mean by "it" in "would
 not crash if we did allow it".  In comment:14 I show a particular
 transformation.  I believe it is (slightly) beneficial, and will not
 crash, but the result is not type-correct.

 If you can figure out how to express the proof that it won't crash, as a
 coercion perhaps, maybe you could somehow express that proof in Core.

 You don't need casts to expose the problem: #14620 is enough.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14610#comment:17>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler