[GHC] #13426: compile-time memory-usage regression for DynFlags between GHC 8.0 and GHC 8.2

[GHC]

#13426: compile-time memory-usage regression for DynFlags between GHC 8.0 and GHC
8.2
-------------------------------------+-------------------------------------
        Reporter:  hvr               |                Owner:  rwbarton
            Type:  bug               |               Status:  new
        Priority:  high              |            Milestone:  8.2.1
       Component:  Compiler          |              Version:  8.1
      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):

 Wow, I have no idea how you found that! Well done.

 >  Alternatively, one can insert a seqType into reallyRebuildCase:

 That would be ''much'' better.  That kills the leak when we build a
 `Case`, whereas the former alternative waits until the ''subsequent'' run
 of the simplifier before squashing it.

 Please add a `Note` to explain the `seq` (point to this ticket too).

 Does it need to be a `seqType` or will `seq` do?  I expect that `seq`
 would be enough, since the `OutType` returned by `contResultType` is (I
 think) a result of calling `simplType`.

 >  It must share structure with some other part of the program, but I'm
 not sure what.

 My guess: `exprType` often returns that type.  Still I'm surprised that it
 leads to a ''cumulative'' space leak (i.e. every growing).  Does it?

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13426#comment:16>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler