[Git][ghc/ghc][wip/T15304] 12 commits: Add outputable instances for the types in GHC.Iface.Ext.Types, add -ddump-hie

Ben Gamari gitlab at gitlab.haskell.org
Sat Apr 4 22:05:23 UTC 2020



Ben Gamari pushed to branch wip/T15304 at Glasgow Haskell Compiler / GHC


Commits:
ef7576c4 by Zubin Duggal at 2020-04-03T06:24:56-04:00
Add outputable instances for the types in GHC.Iface.Ext.Types, add -ddump-hie
flag to dump pretty printed contents of the .hie file

Metric Increase:
   hie002

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)

- - - - -
9462452a by Andreas Klebinger at 2020-04-03T06:25:33-04:00
Improve and refactor StgToCmm codegen for DataCons.

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
  it.

Before this patch 1) did only apply to local bindings and we
didn't do 2) at all.

- - - - -
a214d214 by Moritz Bruder at 2020-04-03T06:26:11-04:00
Add singleton to NonEmpty in libraries/base

This adds a definition to construct a singleton non-empty list
(Data.List.NonEmpty) according to issue #17851.

- - - - -
f7597aa0 by Sylvain Henry at 2020-04-03T06:26:54-04:00
Testsuite: measure compiler stats for T16190

We were mistakenly measuring program stats

- - - - -
a485c3c4 by Sylvain Henry at 2020-04-03T06:26:54-04:00
Move blob handling into StgToCmm

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.

- - - - -
cc2918a0 by Sylvain Henry at 2020-04-03T06:26:54-04:00
Refactor CmmStatics

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
generation).

This patch removes this redundant code by using a single GADT for
(Raw)CmmStatics.

- - - - -
9e60273d by Maxim Koltsov at 2020-04-03T06:27:32-04:00
Fix haddock formatting in Control.Monad.ST.Lazy.Imp.hs

- - - - -
1b7e8a94 by Andreas Klebinger at 2020-04-03T06:28:08-04:00
Turn newlines into spaces for hadrian/ghci.

The newlines break the command on windows.

- - - - -
4291bdda by Simon Peyton Jones at 2020-04-03T06:28:44-04:00
Major improvements to the specialiser

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 2d0cf6252957b8980d89481ecd0b79891da4b14b
     Author: Sandy Maguire <sandy at 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!

- - - - -
40a85563 by Ömer Sinan Ağacan at 2020-04-03T18:26:19+03:00
Revert accidental change in 9462452

[ci skip]

- - - - -
bd75e5da by Ryan Scott at 2020-04-04T07:07:58-04:00
Enable ImpredicativeTypes internally when typechecking selector bindings

This is necessary for certain record selectors with higher-rank
types, such as the examples in #18005. See
`Note [Impredicative record selectors]` in `TcTyDecls`.

Fixes #18005.

- - - - -
37bc7a6c by Ben Gamari at 2020-04-04T18:05:15-04:00
simplifier: Kill off ufKeenessFactor

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
-funfolding-use-threshold.

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
!2608

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
thresh
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
thresh
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:
    hie002
    T12234
    T13035
    T13719
    T14683
    T4801
    T5631
    T5642
    T9020
    T9872d
    T9961
Metric Increase:
    T12150
    T12425
    T13701
    T14697
    T15426
    T1969
    T3064
    T5837
    T9203
    T9872a
    T9872b
    T9872c
    T9872d
    haddock.Cabal
    haddock.base
    haddock.compiler

- - - - -


30 changed files:

- compiler/GHC/Cmm.hs
- compiler/GHC/Cmm/DebugBlock.hs
- compiler/GHC/Cmm/Info.hs
- compiler/GHC/Cmm/Info/Build.hs
- compiler/GHC/Cmm/Parser.y
- compiler/GHC/Cmm/Ppr/Decl.hs
- compiler/GHC/Cmm/Utils.hs
- compiler/GHC/CmmToAsm/PPC/CodeGen.hs
- compiler/GHC/CmmToAsm/PPC/Ppr.hs
- compiler/GHC/CmmToAsm/PPC/RegInfo.hs
- compiler/GHC/CmmToAsm/Ppr.hs
- compiler/GHC/CmmToAsm/SPARC/CodeGen.hs
- compiler/GHC/CmmToAsm/SPARC/CodeGen/Gen32.hs
- compiler/GHC/CmmToAsm/SPARC/Ppr.hs
- compiler/GHC/CmmToAsm/SPARC/ShortcutJump.hs
- compiler/GHC/CmmToAsm/X86/CodeGen.hs
- compiler/GHC/CmmToAsm/X86/Instr.hs
- compiler/GHC/CmmToAsm/X86/Ppr.hs
- compiler/GHC/CmmToC.hs
- compiler/GHC/CmmToLlvm.hs
- compiler/GHC/CmmToLlvm/Data.hs
- compiler/GHC/CmmToLlvm/Ppr.hs
- compiler/GHC/Core/Op/Specialise.hs
- compiler/GHC/Core/Subst.hs
- compiler/GHC/Core/Unfold.hs
- compiler/GHC/Driver/Flags.hs
- compiler/GHC/Driver/Main.hs
- compiler/GHC/Driver/Session.hs
- compiler/GHC/HsToCore/Binds.hs
- compiler/GHC/Iface/Ext/Debug.hs


The diff was not included because it is too large.


View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5d847cf0cda61a82643fde3ed98c5ba383341941...37bc7a6cbbfe885fe7c0f22dde6892c0f9a6ed11

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/5d847cf0cda61a82643fde3ed98c5ba383341941...37bc7a6cbbfe885fe7c0f22dde6892c0f9a6ed11
You're receiving this email because of your account on gitlab.haskell.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20200404/ae4ea03b/attachment.html>


More information about the ghc-commits mailing list