[GHC] #14816: Missed Called Arity opportunity?

[GHC]

#14816: Missed Called Arity opportunity?
-------------------------------------+-------------------------------------
        Reporter:  dfeuer            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.6.1
       Component:  Compiler          |              Version:  8.2.2
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Runtime           |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by sgraf):

 I spent some time thinking about how to implement this, but given that
 `lazy_fvs` plays a largely different role for thunks (unleashes strictness
 through signatures and stores usage in `lazy_fvs`) vs. non-thunks (puts
 weak demands into `lazy_fvs`, e.g. `<L,U>` and `<L,C(C1(U))>` which won't
 change anymore during fixpointing), I don't think pursuing this is worth
 the trouble.

 I'm afraid we just have to live with free variables getting a second-class
 treatment compared to parameters for the time being. This is another good
 argument for eventually splitting the demand analyser into its three sub-
 analyses.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14816#comment:13>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler