Re: [GHC] #8763: forM_ [1..N] does not get fused (10 times slower than go function)

11 Apr 2018

      #8763: forM_ [1..N] does not get fused (10 times slower than go function)
-------------------------------------+-------------------------------------
        Reporter:  nh2               |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.6.1
       Component:  Compiler          |              Version:  7.6.3
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Runtime           |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #7206             |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by sgraf):

 Note that's not a problem for `[x..y]` (`eftInt`), because that doesn't
 need to consider counting down. It's not an issue for `[x,y..]`
 (`edtInt`), because although it calls `efdtInt{Up,Dn}` internally, it
 doesn't take part in fusion at all (is that an oversight or by design?).

-- 
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8763#comment:59
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler