1. 15 Jan, 2016 21 commits
  2. 14 Jan, 2016 12 commits
  3. 13 Jan, 2016 7 commits
    • Ömer Sinan Ağacan's avatar
      Minor improvement in CoreDump outputs: · c3133273
      Ömer Sinan Ağacan authored
      Don't add parens unnecessarily when arguments of the application are all
      hidden (because of parameters like -dsuppress-all,
      -dsuppress-type-applications etc.)
      Reviewers: bgamari, austin
      Subscribers: thomie
      Differential Revision: https://phabricator.haskell.org/D1771
    • Iavor S. Diatchki's avatar
      Handle over-applied custom type errors too. · 6ea24af9
      Iavor S. Diatchki authored
          type family F :: Type -> Type where
             F = TypeError (Text "Error")
      Now, if we see something like `F Int` we should still report the custom
      type error.
    • kseo's avatar
      Print a message when loading a .ghci file. · c3f92464
      kseo authored
      Test Plan: ./validate
      Reviewers: austin, thomie, bgamari
      Reviewed By: bgamari
      Subscribers: rwbarton, thomie
      Differential Revision: https://phabricator.haskell.org/D1756
      GHC Trac Issues: #11389
    • Ben Gamari's avatar
      Add test for Data.Typeable.typeOf · e782e882
      Ben Gamari authored
      Test Plan: Validate
      Reviewers: goldfire, austin
      Subscribers: thomie
      Differential Revision: https://phabricator.haskell.org/D1770
      GHC Trac Issues: #11120
    • Ben Gamari's avatar
      Add missing type representations · ac3cf68c
      Ben Gamari authored
      Previously we were missing `Typeable` representations for several
      wired-in types (and their promoted constructors). These include,
       * `Nat`
       * `Symbol`
       * `':`
       * `'[]`
      Moreover, some constructors were incorrectly identified as being defined
      in `GHC.Types` whereas they were in fact defined in `GHC.Prim`.
      Ultimately this is just a temporary band-aid as there is general
      agreement that we should eliminate the manual definition of these
      representations entirely.
      Test Plan: Validate
      Reviewers: austin, hvr
      Subscribers: thomie
      Differential Revision: https://phabricator.haskell.org/D1769
      GHC Trac Issues: #11120
    • thomie's avatar
      TemplateHaskell: revive isStrict, notStrict and unpacked · d44bc5c0
      thomie authored
      These 3 functions are useful to keep around a bit longer, to prevent
      breaking existing code that uses them.
      Related to #10697.
      Reviewers: austin, goldfire, RyanGlScott, bgamari
      Reviewed By: RyanGlScott, bgamari
      Differential Revision: https://phabricator.haskell.org/D1761
    • niteria's avatar
      Use implicit CallStacks for ASSERT when available · 88d6d5aa
      niteria authored
      This aids with debugging, since all you have to do to get more
      stack frames is add a constraint `(?callStack :: CallStack) =>`.
      Old output:
      ghc-stage2: panic! (the 'impossible' happened)
        (GHC version 8.1.20160107 for x86_64-unknown-linux):
              ASSERT failed!
        file compiler/types/TyCoRep.hs line 1800
        InScope []
        [Xuv :-> n_av5[sk]]
      New output:
      ghc-stage2: panic! (the 'impossible' happened)
        (GHC version 8.1.20160107 for x86_64-unknown-linux):
              ASSERT failed!
        CallStack (from ImplicitParams):
        assertPprPanic, called at compiler/types/TyCoRep.hs:1800:95 in
        InScope []
        [Xuv :-> n_av5[sk]]
      Test Plan:
      manual testing
      Reviewers: austin, gridaphobe, bgamari
      Reviewed By: gridaphobe, bgamari
      Subscribers: thomie
      Differential Revision: https://phabricator.haskell.org/D1751