[GHC] #14816: Missed Called Arity opportunity?

GHC ghc-devs at haskell.org
Tue Feb 20 13:49:22 UTC 2018


#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 nomeata):

 >  You can circumvent that if you make f an argument to go (reverse static
 arg transform, essentially):

 That’s smart! By making it an argument, you essentially tell GHC to apply
 inductive reasoning, and then the Demand Analyzer finds out that `f` is
 used at most once! Cool.

 And I presume if the recursion were non-linear, it would also do the right
 thing…

 So – can we just include the free variables in the strictness signature
 during fixpointing and get that optimization without the code
 transformation?

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14816#comment:8>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler


More information about the ghc-tickets mailing list