[GHC] #13390: Strict literal float-out during desugaring regresses T1969 at -O0

[GHC]

#13390: Strict literal float-out during desugaring regresses T1969 at -O0
-------------------------------------+-------------------------------------
        Reporter:  bgamari           |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.4.1
       Component:  Compiler          |              Version:  8.0.1
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
                                     |  Unknown/Multiple
 Type of failure:  None/Unknown      |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by dfeuer):

 bgamari asked me to summarize my current best guess very briefly, so here
 goes. I believe the basic problem when the string floating patch is
 applied is that we're seeing something like

 {{{#!hs
 \x -> c x
 }}}

 and deciding to inline `c` into the body of the lambda. In this case, at
 least, it would be better to eta reduce instead. I'm not sure that it
 would ever be advantageous to inline a function into such a removable
 lambda. I can't think of such a situation myself, but that doesn't mean
 much.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13390#comment:12>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler