[GHC] #8900: Strictness analysis regression

[GHC]

#8900: Strictness analysis regression
--------------------------------------------+------------------------------
        Reporter:  tibbe                    |            Owner:
            Type:  bug                      |           Status:  closed
        Priority:  normal                   |        Milestone:
       Component:  Compiler                 |          Version:  7.8.1-rc2
      Resolution:  fixed                    |         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 tibbe):

 Replying to [comment:21 simonpj]:
 > Comment 9 you said "While the extra case is definitely a regression, it
 doesn't seem to be the cause of the time difference". and in comment 14
 "without the extra case there is still a small difference".
 >
 > Is that still true? If so there might still be something to track down.

 There's still a (quite small) difference in mutator time. I haven't had
 time to investigate it.

 The Cmm output by 7.6.3 (attachment:HashMapInsert-7.6.3.dump-opt-cmm) and
 HEAD (attachment:HashMapInsert-HEAD.dump-opt-cmm) look quite a bit
 different, at least superficially. For example, 7.6.3 has the core loop
 split into two functions, `$wpoly_go_info` and `s30b_ret`, while HEAD has
 just one function, `$wpoly_go_entry`.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8900#comment:22>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler