[GHC] #13417: piResultTys1

[GHC]

#13417: piResultTys1
-------------------------------------+-------------------------------------
        Reporter:  ralf              |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  Compiler          |              Version:  8.0.2
      Resolution:                    |             Keywords:
Operating System:  Linux             |         Architecture:  x86_64
 Type of failure:  Compile-time      |  (amd64)
  crash or panic                     |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by simonpj):

 Here's a simpler test case
 {{{
 data T a = E7

 cons7 :: T a -> T b
 cons7 E7 = E7
 }}}
 Amazingly, this crashes 8.0.2, because of this egregious bug in
 `SimplUtils.mkCase1`:
 {{{
     check_eq (Var v)    (DataAlt con) []   = v == dataConWorkId con
                                              -- Optimisation only
 ...
     check_eq rhs        (DataAlt con) args = cheapEqExpr' tickishFloatable
 rhs $
                                              mkConApp con (arg_tys ++
                                                            varsToCoreExprs
 args)
 }}}
 The first of these equations ("optimisation only") is only equivalent to
 the second if `arg_tys` is empty, which it isn't in this case.

 This bug has been there since I introduced it in Nov 2011:
 {{{
 commit 02ac2974ce8e537372bff8d9e0a6efb461ed2c59
 Author: Simon Peyton Jones <simonpj at microsoft.com>
 Date:   Wed Nov 16 10:37:47 2011 +0000

     Fix CaseIdentity optimisaion

     In fixing one bug I'd introduced another;
        case x of { T -> T; F -> F }
     wasn't getting optmised!  Trivial to fix.
 }}}
 It's just rather hard to trigger.  Unrelated changes in HEAD mean that it
 no longer shows up.  But I'll fix it anyway!

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13417#comment:8>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler