Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Summary:
A while back when I was refactoring the package code, I tried to solve
a performance problem by introducing a fastpath for module lookups.  Well,
it turned out the performance problem was unrelated, but I kept the optimization
because it seemed vaguely useful.

In this commit, I remove the optimization because I don't really think it's
buying us much and it increased code complexity.

ToDo: Inline mkModuleToPkgConfGeneric into mkModuleToPkgConfAll

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Test Plan: validate

Reviewers: simonpj, austin

Reviewed By: austin

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D434

Summary: Instead, describe what it actually does.

Reviewers: austin, ekmett, simonpj, hvr

Reviewed By: austin, ekmett

Subscribers: simonpj, thomie, carter

Projects: #ghc

Differential Revision: https://phabricator.haskell.org/D448

Summary:
The derived Show instances for SrcSpan and SrcLoc are very verbose.

This patch replaces them with hand-made ones which use positional
syntax for the record constructors, rather than exhaustively listing
each one.

Test Plan: sh ./validate

Reviewers: austin

Reviewed By: austin

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D445

Summary: Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Test Plan: n/a

Reviewers: austin, Mikolaj

Reviewed By: austin, Mikolaj

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D446

Summary: Make listArray fuse with a list producer. Note: if code size increases too much, we can fix that with some `RULES`.

Reviewers: nomeata, hvr, austin, ekmett, simonmar, bgamari

Reviewed By: bgamari

Subscribers: bgamari, thomie, carter

Differential Revision: https://phabricator.haskell.org/D474

GHC Trac Issues: #9801

Summary:
It describes a work around Trac #3838, but it is already fixed and the
workaround removed, Trac #7653

Test Plan: not needed

Reviewers: hvr, Mikolaj, austin

Reviewed By: austin

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D478

Summary: This will *not* work on NixOS for example.

Reviewers: austin

Reviewed By: austin

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D479

GHC Trac Issues: #9057

Reviewers: austin

Reviewed By: austin

Subscribers: thomie, carter

Differential Revision: https://phabricator.haskell.org/D481

GHC Trac Issues: #4900

It turns out that emitInsoluble isn't the right thing, because the
insolubles should be returned in the wc_insol field of the
WantedConstraints.

updated to handle `-fwarn-trustworthy-safe`.

At optimization level `-O0` `wordToInteger` wasn't inlined, and this caused a
`__integer` literal to turn up in final Core, which would trigger the GHC panic

  ghc-stage1: panic! (the 'impossible' happened)
    (GHC version 7.9.20141114 for x86_64-unknown-linux):
	Can't use Integer in integer-*

By forcing inlining via `inline` this is avoided.
This should hopefully address #9800.

This pulls in the new `Generic`-based `-XDefaultSignature`-based default
implementation for `rnf`[1], and will be interesting to use in combination
with the soon to be merged `-XDeriveAnyClass` extension.

This requires updating several other submodules as well in order
to relax the upper bound on `deepseq` and/or in a few cases to
avoid relying on the default method implementation of `rnf`:

 - `Cabal`
 - `bytestring`
 - `containers`
 - `parallel`
 - `process`
 - `time`

 [1]: http://permalink.gmane.org/gmane.comp.lang.haskell.libraries/23031

This is only a temporary kludge until the issue workarounded by 452d6aa9
gets properly fixed

Summary:
Instead of recording exposed-modules and reexported-modules as seperate
fields in the installed package database, this commit merges them into
a single field (exposed-modules).  The motivation for this change is
in preparation for the inclusion of *signatures* into the installed
package database, which may also be reexported.  Merging the representation
means that we can treat reexports uniformly, no matter if they're a normal
module or a signature.

This commit adds a stub for signatures, but that code isn't wired up to
anything yet.

Contains Cabal submodule update to accommodate these changes.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Test Plan: validate

Reviewers: simonpj, duncan, austin

Subscribers: thomie, carter, simonmar

Differential Revision: https://phabricator.haskell.org/D421

This introduces ./validate failures for Windows right now, so in the
mean time let's just back this flag out as a default -Wall flag.

Signed-off-by: Austin Seipp <austin@well-typed.com>

The solver is now provided with a boolean flag, which is False when
invoked inside solveFlats and True when invoked on the unflattened
constraints at the end.

Thomas Miedema authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

[skip ci]

David Feuer authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

Implement an `amap`/`coerce` rule in `GHC.Arr` to match the
`map`/`coerce` rule in GHC.Base.

In order to do so, delay inlining `amap` until phase 1.

To prevent the inlining delay from causing major inefficiencies due to
missed list fusion, rewrite `amap` to avoid relying on list fusion. This
has the extra benefit of reducing the size of the compiled amap code by
skipping the impossible case of an array with a negative size.

Reviewed By: nomeata

Differential Revision: https://phabricator.haskell.org/D471

as type arguments, not value arguments

See Note [Unfolding while desugaring] for the rationale.

generate a worker function of type Void# -> T#, and redirect the wrapper
(via a compulsory unfolding) to the worker. Fixes #9732.

GMP 4.x didn't provide the `mp_bitcnt_t` typedef yet, so we locally
define one if GMP 4.x is detected.

David Feuer authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

Previously, `Array`s were simply converted to lists, and the list
methods used. That works acceptably well for `foldr` and `foldr1`, but
not so sensibly for most other things. Left folds ended up "twisted" the
way they are for lists, leading to surprising performance
characteristics.

Moreover, this implements `length` and `null` so they check the array
size directly.

Finally, a test is added to the testsuite ensuring the overridden
`Foldable` methods agree with their expected default semantics.

Addresses #9763

Reviewed By: hvr, austin

Differential Revision: https://phabricator.haskell.org/D459

David Feuer authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

This avoids duplication in `GHC.Base`; originally, we had

  mapM f = sequence . map f

This led to excessive allocation in `cryptarithm2`. Defining

  sequence = mapM id

does not appear to cause any `nofib` problems.

Reviewed By: hvr

Differential Revision: https://phabricator.haskell.org/D470

David Feuer authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

Make `words` a good producer and `unwords` a good consumer for list
fusion. Thus `unwords . words` will avoid producing an intermediate list
of words, although it will produce each individual word.

Make `unwords` slightly lazier, so that
`unwords (s : undefined) = s ++ undefined` instead of `= undefined`.

Reviewed By: austin

Differential Revision: https://phabricator.haskell.org/D375

David Feuer authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

Make the comment on the map/coerce rule refer to the right section in
the paper; give the full name of the papers, and name its authors.

[skip ci]

Differential Revision: https://phabricator.haskell.org/D472

Update submodule haskell2010, haskell98, hoop, hpc and stm to fix new
warnings.

This warns when a module marked as `-XTrustworthy` could have been
inferred as safe instead.