1. 07 Nov, 2018 7 commits
  2. 06 Nov, 2018 1 commit
  3. 05 Nov, 2018 1 commit
  4. 04 Nov, 2018 1 commit
    • Roland Senn's avatar
      Fix for Trac #15611: Scope errors lie about what modules are imported. · 1a3b9bd0
      Roland Senn authored
      Summary:
      For the error message:
          Not in scope X.Y
          Module X does not export Y
          No module named ‘X’ is imported:
      there are 2 cases, where we don't show the last "no module named is imported" line:
      1. If the module X has been imported.
      2. If the module X is the current module. There are 2 subcases:
      
         2.1 If the unknown module name is in a input source file,
             then we can use the getModule function to get the current module name.
      
         2.2 If the unknown module name has been entered by the user in GHCi,
             then the getModule function returns something like "interactive:Ghci1",
             and we have to check the current module in the last added entry of
             the HomePackageTable.
      
      Test Plan: make test TESTS="T15611a T15611b"
      
      Reviewers: monoidal, hvr, thomie, dfeuer, bgamari, DavidEichmann
      
      Reviewed By: monoidal, DavidEichmann
      
      Subscribers: rwbarton, carter
      
      GHC Trac Issues: #15611
      
      Differential Revision: https://phabricator.haskell.org/D5284
      1a3b9bd0
  5. 02 Nov, 2018 8 commits
  6. 01 Nov, 2018 8 commits
    • Matthías Páll Gissurarson's avatar
      Add built-in syntax suggestions, and refactor to allow library use · 1c92f193
      Matthías Páll Gissurarson authored
      Summary:
      This change to the valid hole fits adds built-in syntax candidates (like (:) and []),
      so that they are checked in addition to what is in scope.
      
      The rest is merely a refactor and export of the functions used to find the valid
      hole fits, since there was interest at ICFP to use the valid hole fit machinery for
      additional uses. By exporting the `tcFilterHoleFits` function, this can now be done
      without having to rely on parsing the actual error message.
      
      Test Plan: Test for built-in syntax included
      
      Reviewers: bgamari
      
      Reviewed By: bgamari
      
      Subscribers: RyanGlScott, rwbarton, carter
      
      Differential Revision: https://phabricator.haskell.org/D5227
      1c92f193
    • Alp Mestanogullari's avatar
      hadrian: build ghc-iserv-prof in addition to ghc-iserv · 695f1f2f
      Alp Mestanogullari authored
      As it is required for 10+ tests.
      
      Hadrian didn't give us a chance to build a given executable in vanilla
      and profiling, simultaneously, under two different names. This patch
      implements support for this in general and applies it to
      ghc-iserv[-prof].
      
      Test Plan: scc001 fails without this patch, passes with it.
      
      Reviewers: snowleopard, bgamari
      
      Reviewed By: bgamari
      
      Subscribers: simonpj, ndmitchell, simonmar, rwbarton, carter
      
      Differential Revision: https://phabricator.haskell.org/D5255
      695f1f2f
    • Ben Gamari's avatar
      users-guide: Fix formatting of eventlog format documentation · fcd919f9
      Ben Gamari authored
      Test Plan: Read it
      
      Reviewers: mpickering
      
      Reviewed By: mpickering
      
      Subscribers: rwbarton, carter
      
      Differential Revision: https://phabricator.haskell.org/D5264
      fcd919f9
    • Zhou Fangyi's avatar
      Data.Maybe: add callstack for fromJust (Trac #15559) · 614028e3
      Zhou Fangyi authored
      Per feature request, add `HasCallStack` to `fromJust` in `Data.Maybe`
      and use `error` instead of `errorWithoutStackTrace`. This allows
      `fromJust` to print call stacks when throwing the error.
      
      Also add a new test case for the behaviour, modify existing test cases
      for new signature
      
      Test Plan: New test cases
      
      Reviewers: hvr, bgamari
      
      Reviewed By: bgamari
      
      Subscribers: ulysses4ever, rwbarton, carter
      
      GHC Trac Issues: #15559
      
      Differential Revision: https://phabricator.haskell.org/D5256
      614028e3
    • Vladislav Zavialov's avatar
      Lower precedence for {-# UNPACK #-} · a78e23b8
      Vladislav Zavialov authored
      Test Plan: Validate
      
      Reviewers: goldfire, bgamari
      
      Subscribers: osa1, mpickering, rwbarton, carter
      
      Differential Revision: https://phabricator.haskell.org/D5221
      a78e23b8
    • Richard Eisenberg's avatar
      Don't lint erroneous programs. · 1f72a1c8
      Richard Eisenberg authored
      newFamInst lints its types. This is good. But it's not so good
      when there have been errors and thus recovery tycons are about.
      So we now don't.
      
      Fixes #15796.
      
      Test case: typecheck/should_fail/T15796
      1f72a1c8
    • Richard Eisenberg's avatar
      Actually fail in failIfEmitsConstraints · 74ed9c1c
      Richard Eisenberg authored
      The function TcHsType.failIfEmitsConstraints says that it fails.
      It even does so in its name. But it didn't! It *reported* constraints
      but didn't fail. Now it does.
      
      This is important in tcHsClsInstType; see the comments therein.
      
      This was discovered while looking at #15797, but that ticket
      requires visible kind application to exhibit the bug; the test
      case will come with the patch for #12045.
      74ed9c1c
    • Richard Eisenberg's avatar
      Fix embarrassing, egregious bug in roles of (->) · 255d2e32
      Richard Eisenberg authored
      Previously, I had inexplicably decided that (->)'s roles
      were all Representational. But, of course, its first two
      parameters are *dependent* RuntimeReps. All dependent parameters
      have a Nominal role, because all roles in kinds are Nominal.
      
      Fix is easy, but I have no idea how the world hasn't come
      crashing down before now.
      
      This was found while investigating #15801, which requires
      visible type application in types to observe. Hence, the test
      case will come with the main patch for #12045.
      255d2e32
  7. 31 Oct, 2018 1 commit
  8. 30 Oct, 2018 5 commits
  9. 29 Oct, 2018 8 commits