[GHC] #14667: Compiling a function with a lot of alternatives bottlenecks on insertIntHeap

[GHC]

#14667: Compiling a function with a lot of alternatives bottlenecks on
insertIntHeap
-------------------------------------+-------------------------------------
        Reporter:  niteria           |                Owner:  niteria
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:
      Resolution:                    |             Keywords:
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Compile-time      |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by simonpj):

 The commit in comment:1 says "Applied on top of a fix for #14667 it gives
 a 30% improvement".  So that implies that you have a separate fix for this
 ticket.  Can you

 * Characterise what causes the perf problem that this ticket identifies?
 * Describe your proposed fix?

 If you want to revert Simon M's patch you'd better discuss it with him.
 Perhaps he had a case that was improved by it.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14667#comment:5>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler