1. 03 Jul, 2015 2 commits
    • Ben Gamari's avatar
      Use `+RTS -G1` for more stable residency measurements (#9675) · b5e1944e
      Ben Gamari authored
      Reviewers: ezyang, austin, thomie
      Subscribers: thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D1006
      GHC Trac Issues: #10557
    • Peter Trommler's avatar
      Implement PowerPC 64-bit native code backend for Linux · d3c1dda6
      Peter Trommler authored
      Extend the PowerPC 32-bit native code generator for "64-bit
      PowerPC ELF Application Binary Interface Supplement 1.9" by
      Ian Lance Taylor and "Power Architecture 64-Bit ELF V2 ABI Specification --
      OpenPOWER ABI for Linux Supplement" by IBM.
      The latter ABI is mainly used on POWER7/7+ and POWER8
      Linux systems running in little-endian mode. The code generator
      supports both static and dynamic linking. PowerPC 64-bit
      code for ELF ABI 1.9 and 2 is mostly position independent
      anyway, and thus so is all the code emitted by the code
      generator. In other words, -fPIC does not make a difference.
      rts/stg/SMP.h support is implemented.
      Following the spirit of the introductory comment in
      PPC/CodeGen.hs, the rest of the code is a straightforward
      extension of the 32-bit implementation.
      * Code is generated only in the medium code model, which
        is also gcc's default
      * Local symbols are not accessed directly, which seems to
        also be the case for 32-bit
      * LLVM does not work, but this does not work on 32-bit either
      * Must use the system runtime linker in GHCi, because the
        GHC linker for "static" object files (rts/Linker.c) for
        PPC 64-bit is not implemented. The system runtime
        (dynamic) linker works.
      * The handling of the system stack (register 1) is not ELF-
        compliant so stack traces break. Instead of allocating a new
        stack frame, spill code should use the "official" spill area
        in the current stack frame and deallocation code should restore
        the back chain
      * DWARF support is missing
      Fixes #9863
      Test Plan: validate (on powerpc, too)
      Reviewers: simonmar, trofi, erikd, austin
      Reviewed By: trofi
      Subscribers: bgamari, arnons1, kgardas, thomie
      Differential Revision: https://phabricator.haskell.org/D629
      GHC Trac Issues: #9863
  2. 02 Jul, 2015 3 commits
  3. 30 Jun, 2015 5 commits
  4. 29 Jun, 2015 3 commits
    • Matthew Pickering's avatar
      Correct BangPat SrcSpan calculation · c6bb2fc5
      Matthew Pickering authored
      Previously when the split was performed in splitBang, `BangPat` was given the
      same SrcSpan as the whole of the LHS of the declaration. This patch
      correctly calculates the value.
      Reviewers: alanz, austin
      Reviewed By: alanz, austin
      Subscribers: thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D1020
      GHC Trac Issues: #10588
    • Joachim Breitner's avatar
      Update performance numbers due to #10482 · 9b5df2a4
      Joachim Breitner authored
      The fix in 0b7e538a has regressed these benchmarks. I have recentered
      them rather than marking them as broken(10482), because nobody
      systematically watches the broken test cases, and we want to catch
      future regressions (or improvements!). #10482 is currently still open,
      presumably because this needs investigating.
    • Simon Marlow's avatar
      Mask to avoid uncaught ^C exceptions · bb0e462b
      Simon Marlow authored
      Summary: It was possible to kill GHCi with a carefully-timed ^C
      Test Plan: The bug in #10017 exposed this
      Reviewers: bgamari, austin
      Reviewed By: austin
      Subscribers: thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D1015
      GHC Trac Issues: #10017
  5. 28 Jun, 2015 2 commits
    • Ben Gamari's avatar
      Add -fcross-module-specialise flag · 302d9377
      Ben Gamari authored
      As of 7.10.1 we specialize INLINEABLE identifiers defined in other
      modules. This can expose issues (compiler bugs or otherwise) in some cases
      (e.g. Trac #10491) and therefore we now provide a way for the user to disable
      this optimization.
      Test Plan: Successfully compile Splice.hs from Trac #10491.
      Reviewers: simonpj, austin
      Reviewed By: simonpj
      Subscribers: simonpj, thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D999
      GHC Trac Issues: #10491
    • Ben Gamari's avatar
      Add -fcross-module-specialise flag · 89834d6d
      Ben Gamari authored
      As of 7.10.1 we specialize INLINEABLE identifiers defined in other
      modules. This can expose issues (compiler bugs or otherwise) in some cases
      (e.g. Trac #10491) and therefore we now provide a way for the user to disable
      this optimization.
      Test Plan: Successfully compile Splice.hs from Trac #10491.
      Reviewers: simonpj, austin
      Reviewed By: simonpj
      Subscribers: simonpj, thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D999
      GHC Trac Issues: #10491
  6. 27 Jun, 2015 1 commit
  7. 26 Jun, 2015 24 commits
    • Simon Peyton Jones's avatar
      Test Trac #10524 · 0e1e7987
      Simon Peyton Jones authored
    • Simon Peyton Jones's avatar
      Improve error message for Typeable k (T k) · ceb3c844
      Simon Peyton Jones authored
      GHC can't yest build a TypeRep for a type involving kind variables.
      (We await kinds = types for that.)  But the error message was terrible,
      as fixing #10524 reminded me.
      This improves it a lot.
    • Simon Peyton Jones's avatar
      Better tracing and tiny refactoring · a64a26f0
      Simon Peyton Jones authored
    • Simon Peyton Jones's avatar
      Make fvType ignore kinds · 8e347839
      Simon Peyton Jones authored
      TcValidity.fvTypes works in partnership with sizeTypes, and
      hence should ignore kinds in exactly the same way.  It wasn't
      doing so, which meant that validDerivPred said "No" when it
      should have said "Yes". That led to the bug reported in
      Trac #10524 comment:7.
      The error message is pretty terrible
        No instance for (Typeable T)
      but I'll fix that next
    • Simon Peyton Jones's avatar
      Kill off sizePred · 614ba3c5
      Simon Peyton Jones authored
      It really isn't needed, and life is simpler without
    • Simon Peyton Jones's avatar
      closeOverKinds *before* oclose in coverage check · 7c07cf16
      Simon Peyton Jones authored
      Combining functional dependencies with kind-polymorphism is
      devilishly tricky!  It's all documented in
          Note [Closing over kinds in coverage]
      Fixes Trac #10564
    • Simon Peyton Jones's avatar
      Improve CPR behavior for strict constructors · 0696fc6d
      Simon Peyton Jones authored
      When working on Trac #10482 I noticed that we could give constructor
      arguments the CPR property if they are use strictly.
      This is documented carefully in
          Note [CPR in a product case alternative]
      and also
          Note [Initial CPR for strict binders]
      There are a bunch of intersting examples in
          Note [CPR examples]
      which I have added to the test suite as T10482a.
      I also added a test for #10482 itself.
    • Simon Peyton Jones's avatar
      Small doc fixes · caf9d427
      Simon Peyton Jones authored
    • Simon Peyton Jones's avatar
      Comments only · cbd9278c
      Simon Peyton Jones authored
    • thomie's avatar
      GHCi docs: layout rule is respected inside :{ :} · 7c8ffd3d
      thomie authored
      I don't know if or when this changed, but the documentation on :{ :}
      (multiline input) seems out of date. Layout rule works fine. I added a
      regression test.
      Reviewers: austin
      Subscribers: thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D1013
    • Simon Marlow's avatar
      Fix deadlock (#10545) · 111ba4be
      Simon Marlow authored
      yieldCapability() was not prepared to be called by a Task that is not
      either a worker or a bound Task.  This could happen if we ended up in
      yieldCapability via this call stack:
      and there were a few other ways this could happen via requestSync.
      The fix is to handle this case in yieldCapability(): when the Task is
      not a worker or a bound Task, we put it on the returning_workers
      queue, where it will be woken up again.
      Summary of changes:
      * `yieldCapability`: factored out subroutine waitForWorkerCapability`
      * `waitForReturnCapability` renamed to `waitForCapability`, and
        factored out subroutine `waitForReturnCapability`
      * `releaseCapabilityAndQueue` worker renamed to `enqueueWorker`, does
        not take a lock and no longer tests if `!isBoundTask()`
      * `yieldCapability` adjusted for refactorings, only change in behavior
        is when it is not a worker or bound task.
      Test Plan:
      * new test concurrent/should_run/performGC
      * validate
      Reviewers: niteria, austin, ezyang, bgamari
      Subscribers: thomie, bgamari
      Differential Revision: https://phabricator.haskell.org/D997
      GHC Trac Issues: #10545
    • Simon Marlow's avatar
      Fix for crash in setnumcapabilities001 · be0ce871
      Simon Marlow authored
      getNewNursery() was unconditionally incrementing next_nursery, which
      is normally fine but it broke an assumption in
      storageAddCapabilities().  This manifested as an occasional crash in
      the setnumcapabilities001 test.
    • Simon Peyton Jones's avatar
      Tiny refactor plus comments · 0aaea5b8
      Simon Peyton Jones authored
    • Simon Peyton Jones's avatar
      Error message wibbles from out-of-scope changes · b98ff25f
      Simon Peyton Jones authored
      The patch "Treat out-of-scope variables as holes" makes
      lots of error messages change a bit. This patch has all
      the change.
    • Simon Peyton Jones's avatar
      Treat out-of-scope variables as holes · fb7b6922
      Simon Peyton Jones authored
      This patch implements the idea in Trac #10569.
      * An out-of-scope variable is treated as a typed expression
      * That is, we don't report it in the type checker, not the
        renamer, and we when we do report it, we give its type.
      * Moreover, we can defer the error to runtime with
      In implementation terms:
      * The renamer turns an unbound variable into a HsUnboundVar
      * The type checker emits a Hole constraint for a
        HsUnboundVar, and turns it back into a HsVar
      It was a bit painful to implement because a whole raft of
      error messages change slightly.  But there was absolutely
      nothing hard in principle.
      Holes are reported with a bunch of possibly-useful context,
      notably the "relevant bindings".  I found that this was
      distracting clutter in the very common case of a mis-typed
      variable that is only accidentally not in scope, so I've
      arranged to print the context information only for true holes,
      that is ones starting with an underscore.
      Unbound data constructors use in patterns, like
        f (D x) = x
      are still reportd by the renamer, and abort compilation
      before type checking.
    • Simon Peyton Jones's avatar
      Get rid of irrelevant impredicative polymoprhism · 95fc6d59
      Simon Peyton Jones authored
      These tests aren't about impredicativity
    • Simon Peyton Jones's avatar
      Get rid of irrlevant result type signature · 2f16a3b8
      Simon Peyton Jones authored
      Result type-sigs are now illegal, but that's not what this test
      is about
    • Simon Peyton Jones's avatar
      Add module header to test · 97e313cd
      Simon Peyton Jones authored
      ..to avoid irrelevant 'main is not defined' error message
    • Simon Peyton Jones's avatar
      Don't float out alpha[sig] ~ Int · b69dc731
      Simon Peyton Jones authored
      This is just a small twiddle to TcSimplify.usefulToFloat
      See Note [Which equalities to float].
    • Simon Peyton Jones's avatar
      Fix addDataConStrictness · 9014a7ee
      Simon Peyton Jones authored
      See Note [Add demands for strict constructors].
      The new bit is the test for isAbsDmd in addDataConStrictness.
      There was a cryptic note that said that this function
      should add a seqDmd even for Absent arguments, but that
      is definitely a bad thing (as the Note now says), causing
      unused arguments to be passed to the worker.
      Easy fix!
    • Simon Peyton Jones's avatar
      Minor fix to free-vars in RnTypes · cc0dba1e
      Simon Peyton Jones authored
      A type wild-card should't appear in the "uses"
      free-variable set.
    • Simon Peyton Jones's avatar
      Allow recursive unwrapping of data families · 0b7e538a
      Simon Peyton Jones authored
      When doing strictness analysis, we need to look inside products.
      To avoid unpacking infinitely, we must be careful about
      infinite types.  That in turn is controlled by TyCon.checkRecTc.
      For data families like
         data instance T (a,b) = MkT a (T b)
      we want to unpack the thing recursively for types like
        T (Int, (Int, (Int, Int)))
      This patch elaborates the checkRecTc mechanism in TyCon, to
      maintain a *count* of how many times a TyCon has shown up,
      rather than just a boolean.
      A simple change, and a useful one.  Fixes Trac #10482.
    • Simon Peyton Jones's avatar
      Use a Representaional coercion for data families · ff8a6716
      Simon Peyton Jones authored
      When we have
        data instance T (a,b) = MkT a b
      we make a represntation type
        data TPair a b = MkT a b
      plus an axiom to connect the two
        ax a b :: T (a,b)  ~R  TPair a b
      Previously this was a Nominal equality, and that worked ok
      but seems illogical since Nominal equalities are between
      types that the programmer thinks of as being equal.  But
      TPair is not visible to the programmer; indeed we call it
      the "representation TyCon".  So a Representational equality
      seems more suitable here.
    • Simon Peyton Jones's avatar
      Improve pretty-printing for CoPat · 9195927d
      Simon Peyton Jones authored