1. 07 Apr, 2020 3 commits
    • Ben Gamari's avatar
      simplifier: Kill off ufKeenessFactor · 3d2991f8
      Ben Gamari authored
      We used to have another factor, ufKeenessFactor, which would scale the
      discounts before they were subtracted from the size. This was justified
      with the following comment:
        -- We multiple the raw discounts (args_discount and result_discount)
        -- ty opt_UnfoldingKeenessFactor because the former have to do with
        --  *size* whereas the discounts imply that there's some extra
        --  *efficiency* to be gained (e.g. beta reductions, case reductions)
        -- by inlining.
      However, this is highly suspect since it means that we subtract a
      *scaled* size from an absolute size, resulting in crazy (e.g. negative)
      scores in some cases (#15304). We consequently killed off
      ufKeenessFactor and bumped up the ufUseThreshold to compensate.
      Adjustment of unfolding use threshold
      Since this removes a discount from our inlining heuristic, I revisited our
      default choice of -funfolding-use-threshold to minimize the change in
      overall inlining behavior. Specifically, I measured runtime allocations
      and executable size of nofib and the testsuite performance tests built
      using compilers (and core libraries) built with several values of
      This comes as a result of a quantitative comparison of testsuite
      performance and code size as a function of ufUseThreshold, comparing
      GHC trees using values of 50, 60, 70, 80, 90, and 100. The test set
      consisted of nofib and the testsuite performance tests.
      A full summary of these measurements are found in the description of
      Comparing executable sizes (relative to the base commit) across all
      nofib tests, we see that sizes are similar to the baseline:
                  gmean      min      max   median
      50         -6.36%   -7.04%   -4.82%   -6.46%
      60         -5.04%   -5.97%   -3.83%   -5.11%
      70         -2.90%   -3.84%   -2.31%   -2.92%
      80         -0.75%   -2.16%   -0.42%   -0.73%
      90         +0.24%   -0.41%   +0.55%   +0.26%
      100        +1.36%   +0.80%   +1.64%   +1.37%
      baseline   +0.00%   +0.00%   +0.00%   +0.00%
      Likewise, looking at runtime allocations we see that 80 gives slightly
      better optimisation than the baseline:
                  gmean      min      max   median
      50         +0.16%   -0.16%   +4.43%   +0.00%
      60         +0.09%   -0.00%   +3.10%   +0.00%
      70         +0.04%   -0.09%   +2.29%   +0.00%
      80         +0.02%   -1.17%   +2.29%   +0.00%
      90         -0.02%   -2.59%   +1.86%   +0.00%
      100        +0.00%   -2.59%   +7.51%   -0.00%
      baseline   +0.00%   +0.00%   +0.00%   +0.00%
      Finally, I had to add a NOINLINE in T4306 to ensure that `upd` is
      worker-wrappered as the test expects. This makes me wonder whether the
      inlining heuristic is now too liberal as `upd` is quite a large
      function. The same measure was taken in T12600.
                   Wall clock time compiling Cabal with -O0
      thresh       50     60     70     80     90      100    baseline
      build-Cabal  93.88  89.58  92.59  90.09  100.26  94.81  89.13
      Also, this change happens to avoid the spurious test output in
      `plugin-recomp-change` and `plugin-recomp-change-prof` (see #17308).
      Metric Decrease:
      Metric Increase:
    • Sebastian Graf's avatar
      Re-export GHC.Magic.noinline from base · bcd66859
      Sebastian Graf authored
    • Daniel Gröber (dxld)'s avatar
      rts: ProfHeap: Fix memory leak when not compiled with profiling · f38e8d61
      Daniel Gröber (dxld) authored
      If we're doing heap profiling on an unprofiled executable we keep
      allocating new space in initEra via nextEra on each profiler run but we
      don't have a corresponding freeEra call.
      We do free the last era in endHeapProfiling but previous eras will have
      been overwritten by initEra and will never get free()ed.
      Metric Decrease:
  2. 06 Apr, 2020 3 commits
    • Simon Peyton Jones's avatar
      Refactoring only · e850d14f
      Simon Peyton Jones authored
      This refactors DictBinds into a data type rather than a pair.
      No change in behaviour, just better code
    • Simon Peyton Jones's avatar
      Fix an tricky specialiser loop · cec2c71f
      Simon Peyton Jones authored
      Issue #17151 was a very tricky example of a bug in which the
      specialiser accidentally constructs a recurive dictionary,
      so that everything turns into bottom.
      I have fixed variants of this bug at least twice before:
      see Note [Avoiding loops].  It was a bit of a struggle
      to isolate the problem, greatly aided by the work that
      Alexey Kuleshevich did in distilling a test case.
      Once I'd understood the problem, it was not difficult to fix,
      though it did lead me a bit of refactoring in specImports.
    • Ömer Sinan Ağacan's avatar
      Don't override proc CafInfos in ticky builds · dcfe29c8
      Ömer Sinan Ağacan authored
      Fixes #17947
      When we have a ticky label for a proc, IdLabels for the ticky counter
      and proc entry share the same Name. This caused overriding proc CafInfos
      with the ticky CafInfos (i.e. NoCafRefs) during SRT analysis.
      We now ignore the ticky labels when building SRTMaps. This makes sense
      - When building the current module they don't need to be in SRTMaps as
        they're initialized as non-CAFFY (see mkRednCountsLabel), so they
        don't take part in the dependency analysis and they're never added to
        (Reminder: a "dependency" in the SRT analysis is a CAFFY dependency,
        non-CAFFY uses are not considered as dependencies for the algorithm)
      - They don't appear in the interfaces as they're not exported, so it
        doesn't matter for cross-module concerns whether they're in the SRTMap
        or not.
      See also the new Note [Ticky labels in SRT analysis].
  3. 04 Apr, 2020 1 commit
  4. 03 Apr, 2020 10 commits
    • Ömer Sinan Ağacan's avatar
      Revert accidental change in 9462452a · 40a85563
      Ömer Sinan Ağacan authored
      [ci skip]
    • Simon Peyton Jones's avatar
      Major improvements to the specialiser · 4291bdda
      Simon Peyton Jones authored
      This patch is joint work of Alexis King and Simon PJ.  It does some
      significant refactoring of the type-class specialiser.  Main highlights:
      * We can specialise functions with types like
           f :: Eq a => a -> Ord b => b => blah
        where the classes aren't all at the front (#16473).  Here we can
        correctly specialise 'f' based on a call like
           f @Int @Bool dEqInt x dOrdBool
        This change really happened in an earlier patch
           commit 2d0cf625
           Author: Sandy Maguire <sandy@sandymaguire.me>
           Date:   Thu May 16 12:12:10 2019 -0400
        work that this new patch builds directly on that work, and refactors
        it a bit.
      * We can specialise functions with implicit parameters (#17930)
           g :: (?foo :: Bool, Show a) => a -> String
        Previously we could not, but now they behave just like a non-class
        argument as in 'f' above.
      * We can specialise under-saturated calls, where some (but not all of
        the dictionary arguments are provided (#17966).  For example, we can
        specialise the above 'f' based on a call
           map (f @Int dEqInt) xs
        even though we don't (and can't) give Ord dictionary.
        This may sound exotic, but #17966 is a program from the wild, and
        showed significant perf loss for functions like f, if you need
        saturation of all dictionaries.
      * We fix a buglet in which a floated dictionary had a bogus demand
        (#17810), by using zapIdDemandInfo in the NonRec case of specBind.
      * A tiny side benefit: we can drop dead arguments to specialised
        functions; see Note [Drop dead args from specialisations]
      * Fixed a bug in deciding what dictionaries are "interesting"; see
        Note [Keep the old dictionaries interesting]
      This is all achieved by by building on Sandy Macguire's work in
      defining SpecArg, which mkCallUDs uses to describe the arguments of
      the call. Main changes:
      * Main work is in specHeader, which marched down the [InBndr] from the
        function definition and the [SpecArg] from the call site, together.
      * specCalls no longer has an arity check; the entire mechanism now
        handles unders-saturated calls fine.
      * mkCallUDs decides on an argument-by-argument basis whether to
        specialise a particular dictionary argument; this is new.
        See mk_spec_arg in mkCallUDs.
      It looks as if there are many more lines of code, but I think that
      all the extra lines are comments!
    • Andreas Klebinger's avatar
      Turn newlines into spaces for hadrian/ghci. · 1b7e8a94
      Andreas Klebinger authored
      The newlines break the command on windows.
    • Maxim Koltsov's avatar
    • Sylvain Henry's avatar
      Refactor CmmStatics · cc2918a0
      Sylvain Henry authored
      In !2959 we noticed that there was some redundant code (in GHC.Cmm.Utils
      and GHC.Cmm.StgToCmm.Utils) used to deal with `CmmStatics` datatype
      (before SRT generation) and `RawCmmStatics` datatype (after SRT
      This patch removes this redundant code by using a single GADT for
    • Sylvain Henry's avatar
      Move blob handling into StgToCmm · a485c3c4
      Sylvain Henry authored
      Move handling of big literal strings from CmmToAsm to StgToCmm. It
      avoids the use of `sdocWithDynFlags` (cf #10143). We might need to move
      this handling even higher in the pipeline in the future (cf #17960):
      this patch will make it easier.
    • Sylvain Henry's avatar
      Testsuite: measure compiler stats for T16190 · f7597aa0
      Sylvain Henry authored
      We were mistakenly measuring program stats
    • muesli4's avatar
      Add singleton to NonEmpty in libraries/base · a214d214
      muesli4 authored
      This adds a definition to construct a singleton non-empty list
      (Data.List.NonEmpty) according to issue #17851.
    • Andreas Klebinger's avatar
      Improve and refactor StgToCmm codegen for DataCons. · 9462452a
      Andreas Klebinger authored
      We now differentiate three cases of constructor bindings:
      1)Bindings which we can "replace" with a reference to
        an existing closure. Reference the replacement closure
        when accessing the binding.
      2)Bindings which we can "replace" as above. But we still
        generate a closure which will be referenced by modules
        importing this binding.
      3)For any other binding generate a closure. Then reference
      Before this patch 1) did only apply to local bindings and we
      didn't do 2) at all.
    • Zubin's avatar
      Add outputable instances for the types in GHC.Iface.Ext.Types, add -ddump-hie · ef7576c4
      Zubin authored
      flag to dump pretty printed contents of the .hie file
      Metric Increase:
      Because of the regression on i386:
      compile_time/bytes allocated increased from i386-linux-deb9 baseline @ HEAD~10:
          Expected    hie002 (normal) compile_time/bytes allocated: 583014888.0 +/-10%
          Lower bound hie002 (normal) compile_time/bytes allocated:   524713399
          Upper bound hie002 (normal) compile_time/bytes allocated:   641316377
          Actual      hie002 (normal) compile_time/bytes allocated:   877986292
          Deviation   hie002 (normal) compile_time/bytes allocated:        50.6 %
      *** unexpected stat test failure for hie002(normal)
  5. 02 Apr, 2020 8 commits
    • Ryan Scott's avatar
      Fix two ASSERT buglets in reifyDataCon · 30a63e79
      Ryan Scott authored
      Two `ASSERT`s in `reifyDataCon` were always using `arg_tys`, but
      `arg_tys` is not meaningful for GADT constructors. In fact, it's
      worse than non-meaningful, since using `arg_tys` when reifying a
      GADT constructor can lead to failed `ASSERT`ions, as #17305
      This patch applies the simplest possible fix to the immediate
      problem. The `ASSERT`s now use `r_arg_tys` instead of `arg_tys`, as
      the former makes sure to give something meaningful for GADT
      constructors. This makes the panic go away at the very least. There
      is still an underlying issue with the way the internals of
      `reifyDataCon` work, as described in
      ghc/ghc#17305 (comment 227023), but we
      leave that as future work, since fixing the underlying issue is
      much trickier (see
      ghc/ghc#17305 (comment 227087)).
    • Sylvain Henry's avatar
      Update Stack resolver for hadrian/build-stack · 49802002
      Sylvain Henry authored
      Broken by 57b888c0
    • Ryan Scott's avatar
      Make Hadrian build with Cabal-3.2 · 27740f24
      Ryan Scott authored
      GHC 8.10 ships with `Cabal-`, so it would be convenient to
      make Hadrian supporting building against 3.2.* instead of having to
      rebuild the entirety of `Cabal-`. There is one API change in
      `Cabal-3.2.*` that affects Hadrian: the `synopsis` and `description`
      functions now return `ShortText` instead of `String`. Since Hadrian
      manipulates these `String`s in various places, I found that the
      simplest fix was to use CPP to convert `ShortText` to `String`s
      where appropriate.
    • Ben Gamari's avatar
      Session: Memoize stderrSupportsAnsiColors · 88f38b03
      Ben Gamari authored
      Not only is this a reasonable efficiency measure but it avoids making
      reentrant calls into ncurses, which is not thread-safe. See #17922.
    • Ömer Sinan Ağacan's avatar
      Remove unused closure stg_IND_direct · 5beac042
      Ömer Sinan Ağacan authored
    • Ömer Sinan Ağacan's avatar
      Fix a pointer format string in RTS · 0a88dd11
      Ömer Sinan Ağacan authored
    • Sebastian Graf's avatar
      Preserve precise exceptions in strictness analysis · 42d68364
      Sebastian Graf authored
      Fix #13380 and #17676 by
      1. Changing `raiseIO#` to have `topDiv` instead of `botDiv`
      2. Give it special treatment in `Simplifier.Util.mkArgInfo`, treating it
         as if it still had `botDiv`, to recover dead code elimination.
      This is the first commit of the plan outlined in
      ghc/ghc!2525 (comment 260886).
    • Simon Peyton Jones's avatar
      Re-engineer the binder-swap transformation · b943b25d
      Simon Peyton Jones authored
      The binder-swap transformation is implemented by the occurrence
      analyser -- see Note [Binder swap] in OccurAnal. However it had
      a very nasty corner in it, for the case where the case scrutinee
      was a GlobalId.  This led to trouble and hacks, and ultimately
      to #16296.
      This patch re-engineers how the occurrence analyser implements
      the binder-swap, by actually carrying out a substitution rather
      than by adding a let-binding.  It's all described in
      Note [The binder-swap substitution].
      I did a few other things along the way
      * Fix a bug in StgCse, which could allow a loop breaker to be CSE'd
        away.  See Note [Care with loop breakers] in StgCse.  I think it can
        only show up if occurrence analyser sets up bad loop breakers, but
      * Better commenting in SimplUtils.prepareAlts
      * A little refactoring in CoreUnfold; nothing significant
        e.g. rename CoreUnfold.mkTopUnfolding to mkFinalUnfolding
      * Renamed CoreSyn.isFragileUnfolding to hasCoreUnfolding
      * Move mkRuleInfo to CoreFVs
      We observed respectively 4.6% and 5.9% allocation decreases for the following
      Metric Decrease:
  6. 01 Apr, 2020 5 commits
  7. 31 Mar, 2020 6 commits
  8. 30 Mar, 2020 1 commit
  9. 29 Mar, 2020 3 commits
    • Krzysztof Gogolewski's avatar
      Minor cleanup · 45eb9d8c
      Krzysztof Gogolewski authored
      - Simplify mkBuildExpr, the function newTyVars was called
        only on a one-element list.
      - TTG: use noExtCon in more places. This is more future-proof.
      - In zonkExpr, panic instead of printing a warning.
    • Ryan Scott's avatar
      Run checkNewDataCon before constraint-solving newtype constructors · a0d8e92e
      Ryan Scott authored
      Within `checkValidDataCon`, we used to run `checkValidType` on the
      argument types of a newtype constructor before running
      `checkNewDataCon`, which ensures that the user does not attempt
      non-sensical things such as newtypes with multiple arguments or
      constraints. This works out in most situations, but this falls over
      on a corner case revealed in #17955:
      newtype T = Coercible () T => T ()
      `checkValidType`, among other things, peforms an ambiguity check on
      the context of a data constructor, and that it turn invokes the
      constraint solver. It turns out that there is a special case in the
      constraint solver for representational equalities (read: `Coercible`
      constraints) that causes newtypes to be unwrapped (see
      `Note [Unwrap newtypes first]` in `TcCanonical`). This special case
      does not know how to cope with an ill formed newtype like `T`, so
      it ends up panicking.
      The solution is surprisingly simple: just invoke `checkNewDataCon`
      before `checkValidType` to ensure that the illicit newtype
      constructor context is detected before the constraint solver can
      run amok with it.
      Fixes #17955.
    • Ben Gamari's avatar
      gitlab-ci: Add FreeBSD release job · 64bf7f51
      Ben Gamari authored