[GHC] #8900: Strictness analysis regression

[GHC]

#8900: Strictness analysis regression
--------------------------------------------+------------------------------
        Reporter:  tibbe                    |            Owner:
            Type:  bug                      |           Status:  new
        Priority:  normal                   |        Milestone:
       Component:  Compiler                 |          Version:  7.8.1-rc2
      Resolution:                           |         Keywords:
Operating System:  MacOS X                  |     Architecture:  x86_64
 Type of failure:  Runtime performance bug  |  (amd64)
       Test Case:                           |       Difficulty:  Unknown
        Blocking:                           |       Blocked By:
                                            |  Related Tickets:
--------------------------------------------+------------------------------

Comment (by simonpj):

 OK I have restored the old behaviour.  (No need to merge.)

 I did a nofib run and found
  * No visible change in binary size
  * No change in allocation
  * Wibbles up and down in runtime but nothing consistent.  (On my machine
 running `k-nucleotide` varies by up to 10% runtime in successive runes,
 for example.)

 So I don't think there are significant losses, and Johan definitely has a
 gain to get.

 I have not actually tried Johan's example, but perhaps Johan can?

 Simon

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8900#comment:19>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler