Re: [GHC] #16089: seq is not cooperating with :sprint in GHCi as expected

25 Dec 2018

      #16089: seq is not cooperating with :sprint in GHCi as expected
-------------------------------------+-------------------------------------
        Reporter:  radrow            |                Owner:  (none)
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:
       Component:  GHCi              |              Version:  8.6.3
      Resolution:                    |             Keywords:  seq sprint
                                     |  strictness
Operating System:  Linux             |         Architecture:
 Type of failure:  Incorrect result  |  Unknown/Multiple
  at runtime                         |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #7253             |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by osa1):

 I did a little bit of digging. The only difference between `:sprint` and
 `:print` is that `:sprint` doesn't bind "suspensions". So sometimes you
 see a
 `_` with `:sprint`, but `:print` gives you a variable instead (see last
 two
 code snippets in comment:1).

 What is a "suspension"? `RtClosureInspect` has a simplistic view of
 Haskell
 runtime objects, and it treats these as "suspensions":

 - `BLACKHOLE`s that point to `TSO` or `BLOCKING_QUEUE`s (as expected,
 becuase
   these are thunks that are not evaluated yet)
 - Anything other than `MUT_VAR` and `CONSTR`s.

 When we do `:print x` what `RtClosureInspect` uses `ghc-heap` to inspect
 the
 value of `x` and sees an `AP` that looks like this:

 {{{
 APClosure
     { info = StgInfoTable
         { entry = Nothing
         , ptrs = 0
         , nptrs = 0
         , tipe = AP
         , srtlen = 0
         , code = Nothing
         }
     , arity = 709704
     , n_args = 0
     , fun = () <-- this is actually a `ForeignRef HValue`, but shown as ()
                    because we cna't show `ForeignRef`s
     , payload = []
     }
 }}}

 which is probably a bug in `ghc-heap` because even if `x` is really an
 `AP` the
 `arity` field makes no sense (it should be 0).

 For reference, the BCOs generated for `x = [True,False]`:

 {{{
 ProtoBCO x1_r1zt#0 []:
    GHC.Types.:
      @ GHC.Types.Bool GHC.Types.False (GHC.Types.[] @ GHC.Types.Bool)
    bitmap:  0 []
    PUSH_G   GHC.Types.[]
    PUSH_G   GHC.Types.False
    PACK     : 2
    ENTER

 ProtoBCO Ghci1.x#0 []:
    GHC.Types.: @ GHC.Types.Bool GHC.Types.True x1_r1zt
    bitmap:  0 []
    PUSH_G   x1_r1zt
    PUSH_G   GHC.Types.True
    PACK     : 2
    ENTER
 }}}

 BCO generated for `let x = [True,False]`:

 {{{
 ProtoBCO ExprTopLevel_E0#0 []:
    let sat_s1yG = ... in ...
    bitmap:  0 []
    PUSH_G   GHC.Types.[]
    PUSH_G   GHC.Types.False
    PACK     : 2
    PUSH_L   0
    PUSH_G   GHC.Types.True
    PACK     : 2
    PUSH_G   GHC.Types.[]
    PUSH_L   1
    PACK     : 2
    PUSH_L   0
    PUSH_APPLY_P
    PUSH_G   GHC.Base.returnIO
    SLIDE    3 3
    ENTER
 }}}

 These should really be identical. For this I filed #16096.

-- 
Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/16089#comment:10
GHC http://www.haskell.org/ghc/
The Glasgow Haskell Compiler