Does this happen in HEAD with GHC’s own STG printer?  If so, could you file a Trac ticket – it’s clearly wrong.

But I do wonder if it could perhaps be something to do with your branch?

Thanks

Simon

From: Csaba Hruska <csaba.hruska@gmail.com>
Sent: 05 November 2018 16:33
To: Simon Peyton Jones <simonpj@microsoft.com>
Cc: ghc-devs@haskell.org
Subject: Re: StgRhsClosure freevar and argument name duplicates

Correction!

The problem happens in integer-gmp:
https://github.com/ghc/ghc/blob/master/libraries/integer-gmp/src/GHC/Integer/Type.hs#L761-L770

On Mon, Nov 5, 2018 at 5:27 PM Csaba Hruska <csaba.hruska@gmail.com> wrote:

An example for the duplication please check the divModInteger function from integer-simple GHC.Integer.Type.

The STG (GHC 8.2.2) generated from divModInteger :: Integer -> Integer -> (# Integer, Integer #) contains duplications in a closure binder list.

Using my custom STG printer it looks like:

module GHC.Integer.Type where

using GHC.Prim
using GHC.Tuple
using GHC.Types

GHC.Integer.Type.divModInteger {-083-} =
  closure (F:) (B:
  n.s84123 {-s84123-}
  d.s84124 {-s84124-}) {
  case GHC.Integer.Type.quotRemInteger {-084-}
         n.s84123 {-s84123-}
         d.s84124 {-s84124-}
  of qr.s84125 {-s84125-} {
    GHC.Prim.(#,#) {-86-} ipv.s84126 {-s84126-} ipv1.s84127 {-s84127-} ->
      let $j.s84128 {-s84128-} =
            closure (F:
            d.s84124 {-s84124-}
            ipv.s84126 {-s84126-}
            ipv1.s84127 {-s84127-}
            ipv.s84126 {-s84126-}
ipv1.s84127 {-s84127-}) (B:
            wild.s84129 {-s84129-}) {
            let $j1.s84130 {-s84130-} =
                  closure (F:
                  d.s84124 {-s84124-}
                  ipv.s84126 {-s84126-}
                  ipv1.s84127 {-s84127-}
                  ipv.s84126 {-s84126-}
                  ipv1.s84127 {-s84127-}
                  wild.s84129 {-s84129-}) (B:
                  wild1.s84131 {-s84131-}) {
                  case _stg_prim_negateInt#
                         wild.s84129 {-s84129-}
                  of sat.s84132 {-s84132-} {
                    DEFAULT ->
                      case _stg_prim_==#
                             wild1.s84131 {-s84131-}
                             sat.s84132 {-s84132-}
                      of sat.s84133 {-s84133-} {
                        DEFAULT ->
                          case _stg_prim_tagToEnum#
                                 sat.s84133 {-s84133-}
                          of wild2.s84134 {-s84134-} {
                            GHC.Types.False {-612-} ->
                              GHC.Prim.(#,#) {-86-}
                                ipv.s84126 {-s84126-}
                                ipv1.s84127 {-s84127-}
                            GHC.Types.True {-645-} ->
                              case GHC.Integer.Type.plusInteger {-066-}
                                     ipv1.s84127 {-s84127-}
                                     d.s84124 {-s84124-}
                              of r'.s84135 {-s84135-} {
                                DEFAULT ->
                                  case GHC.Integer.Type.plusInteger {-066-}
                                         ipv.s84126 {-s84126-}
                                         GHC.Integer.Type.lvl11 {-r50574-}
                                  of q'.s84136 {-s84136-} {
                                    DEFAULT ->
                                      GHC.Prim.(#,#) {-86-}
                                        q'.s84136 {-s84136-}
                                        r'.s84135 {-s84135-}
                                  }
                              }
                          }
                      }
                  }}

            in case ipv1.s84127 {-s84127-}
               of wild1.s84137 {-s84137-} {
                 GHC.Integer.Type.S# {-621-} i#.s84138 {-s84138-} ->
                   case _stg_prim_<#
                          i#.s84138 {-s84138-} 0#
                   of sat.s84140 {-s84140-} {
                     DEFAULT ->
                       case _stg_prim_>#
                              i#.s84138 {-s84138-} 0#
                       of sat.s84139 {-s84139-} {
                         DEFAULT ->
                           case _stg_prim_-#
                                  sat.s84139 {-s84139-}
                                  sat.s84140 {-s84140-}
                           of sat.s84141 {-s84141-} {
                             DEFAULT ->
                               $j1.s84130 {-s84130-}
                                 sat.s84141 {-s84141-}
                           }
                       }
                   }
GHC.Integer.Type.Jp# {-r5813-} dt.s84142 {-s84142-} ->
                   $j1.s84130 {-s84130-} 1#
                 GHC.Integer.Type.Jn# {-r5814-} dt.s84143 {-s84143-} ->
                   $j1.s84130 {-s84130-} -1#
               }}

      in case d.s84124 {-s84124-}
         of wild.s84144 {-s84144-} {
           GHC.Integer.Type.S# {-621-} i#.s84145 {-s84145-} ->
             case _stg_prim_<#
                    i#.s84145 {-s84145-} 0#
             of sat.s84147 {-s84147-} {
               DEFAULT ->
                 case _stg_prim_>#
                        i#.s84145 {-s84145-} 0#
                 of sat.s84146 {-s84146-} {
                   DEFAULT ->
                     case _stg_prim_-#
                            sat.s84146 {-s84146-}
                            sat.s84147 {-s84147-}
                     of sat.s84148 {-s84148-} {
                       DEFAULT ->
                         $j.s84128 {-s84128-}
                           sat.s84148 {-s84148-}
                     }
                 }
             }
GHC.Integer.Type.Jp# {-r5813-} dt.s84149 {-s84149-} ->
             $j.s84128 {-s84128-} 1#
           GHC.Integer.Type.Jn# {-r5814-} dt.s84150 {-s84150-} ->
             $j.s84128 {-s84128-} -1#
         }
  }}

On Mon, Nov 5, 2018 at 2:08 PM Simon Peyton Jones <simonpj@microsoft.com> wrote:

I don’t think there should be duplicates in either. Do you have a test case that shows duplicates?

Simon

From: ghc-devs <ghc-devs-bounces@haskell.org> On Behalf Of Csaba Hruska
Sent: 04 November 2018 11:22
To: ghc-devs@haskell.org
Subject: Re: StgRhsClosure freevar and argument name duplicates

Is it possible that GHC generates STG with invalid binding semantics for certain cases that the Cmm codegen fix or ignore?

This could explain my observations.

I've checked the Stg linter source (StgLint.hs ; GHC 8.2.2 and github master) and it does not check StgRhsClosure free var and binder list at all.

And the scope checker function (addInScopeVars) does not check for duplicates.

Any thoughts?

Cheers,

Csaba

On Sat, Nov 3, 2018 at 9:53 AM Csaba Hruska <csaba.hruska@gmail.com> wrote:

Hi,

Can StgRhsClosure's freevar list ([occ]) or argument list ([bndr]) contain duplicates?

Cheers,

Csaba

data GenStgRhs bndr occ
  = StgRhsClosure
        CostCentreStack         -- CCS to be attached (default is CurrentCCS)
        StgBinderInfo           -- Info about how this binder is used (see below)
[occ]                   -- non-global free vars; a list, rather than
                                -- a set, because order is important
        !UpdateFlag             -- ReEntrant | Updatable | SingleEntry
[bndr]                  -- arguments; if empty, then not a function;
                                -- as above, order is important.
        (GenStgExpr bndr occ)   -- body