This also updates a few submodules

This was done in d94de872 to avoid orphans
but since a94dc4c3 moved `Alternative`
into GHC.Base, this isn't needed anymore.

This is important, as otherwise this would require a non-neglectable amount
of `Control.Monad hiding ((<|>), empty)` imports in user code.

The Haddock submodule is updated as well

Test Plan: partial local ./validate --fast, let Harbormaster doublecheck it

Reviewed By: ekmett, austin

Differential Revision: https://phabricator.haskell.org/D248

A premature complaint was causing Trac #9634.  Acutally this
change also simplifies the lexer and eliminates duplication.
(The renamer was already making the check, as it happens.)

I'm not quite sure why these have improved following the previous
four commits, but I'm quite happy about it

Reponding to Trac #9612:

 * Track the CtOrigin of a Derived equality, arising from a
   functional dependency

 * And report it clearly in the error stream

This relies on a previous commit, in which I stop dropping Derived
insolubles on the floor.

This is preparing for a fix to Trac #9612. The idea is that insoluble
constraints are nice solid errors that we should not discard before
we have a chance to report them.  So TcRnTypes.dropDerivedWC now
keeps insoluble Derived constrains, and instead TcSimplify.solve_wanteds
filters them out

We get somewhat better error message for kind-equality failures too.

A slight downside is that to avoid *duplicate* kind-equality failures
when we float a kind-incompatible equality (e.g.  alpha:* ~ Int#),
I've disabled constraint-floating when there are insolubles.  But that
in turn makes a handful of error messages a little less informative;
good examples are mc21, mc22, mc25.  But I am re-jigging the
constraint floating machinery in another branch, which will make this
go back to the way it was before.

Fixes Trac #9576.  Turned out to be pretty easy.

Summary: Signed-off-by: Dave Laing <dave.laing.80@gmail.com>

Test Plan: Locally tested

Reviewers: thomie, austin

Reviewed By: thomie, austin

Subscribers: simonmar, ezyang, carter, thomie

Differential Revision: https://phabricator.haskell.org/D244

GHC Trac Issues: #9635

Test Plan: harbormaster

Reviewers: austin

Reviewed By: austin

Subscribers: simonmar, ezyang, carter, thomie

Differential Revision: https://phabricator.haskell.org/D243

Summary:
Introduced in 6c7b41cc.

I checked the rest of that commit, and this is all that was left to revert.

Test Plan: x

Reviewers: ezyang, austin

Reviewed By: ezyang, austin

Subscribers: simonmar, ezyang, carter, thomie

Differential Revision: https://phabricator.haskell.org/D241

Summary: More thorough version of a75383cd

Test Plan:
change of comments only

[skip ci]

Reviewers: austin, simonmar, ekmett

Reviewed By: austin, ekmett

Subscribers: simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D239

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

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

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

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

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

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

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

This removes a couple of `INLINE` pragmas from the generics helper
classes. With this change the compile times and memory usage should
go back to the previous GHC 7.8.3 situation.

This has been submitted upstream as https://github.com/kolmodin/binary/pull/62

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

This point the user that there is a relation between the
UndecibadleInstances flag and the type families, not just type
classes.

...several modules in `base` recently touched by me

Edward Z. Yang authored 10 years ago and Herbert Valerio Riedel committed 10 years ago

Summary:
The main change is that Cabal changed the representation of module
re-exports to distinguish reexports in source .cabal files versus
re-exports in installed package registraion files.

Cabal now also does the resolution of re-exports to specific installed
packages itself, so ghc-pkg no longer has to do this. This is a cleaner
design overall because re-export resolution can fail so it is better to
do it during package configuration rather than package registration.
It also simplifies the re-export representation that ghc-pkg has to use.

Add extra ghc-pkg sanity check for module re-exports and duplicates

For re-exports, check that the defining package exists and that it
exposes the defining module (or for self-rexport exposed or hidden
modules). Also check that the defining package is actually a direct
or indirect dependency of the package doing the re-exporting.

Also add a check for duplicate modules in a package, including
re-exported modules.

Test Plan:
So far the sanity checks are totally untested. Should add some test
case to make sure the sanity checks do catch things correctly, and
don't ban legal things.

Reviewers: austin, duncan

Subscribers: angerman, simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D183

GHC Trac Issues:

Summary:
```
git grep -l '#!' | xargs sed -i 's|#!.*/bin/\([^ ]*\)$|#!/usr/bin/env \1|'
```
and some manual tweaking

Test Plan: harbormaster

Reviewers: austin

Subscribers: hvr, simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D237

GHC Trac Issues: #9057

Summary:
```
git grep -l '\(#ifdef \|#if defined\)(\?__GLASGOW_HASKELL__)\?'
```

Test Plan: validate

Reviewers: rwbarton, hvr, austin

Reviewed By: rwbarton, hvr, austin

Subscribers: rwbarton, simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D218

Summary:
Since commits 8fe1f8 and bb6731 in the filepath packages (ticket #2034, closed
in 2010), takeDirectory "foo" returns ".", and not "", so this check is no
longer needed.

Other commits:
* Remove trailing whitespace
* Update comments for #2278

Test Plan: harbormaster

Reviewers: austin

Reviewed By: austin

Subscribers: simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D213

GHC Trac Issues: #2034

This patch fixes Trac #9583, a loop in the simplifier.

I thought this was going to be very complicated but it turned out to
be very simple!  The occurrence analyser does something called
"glomming" if the application of imported RULES means that something
that didn't look recursive becomes recursive.  See `Note [Glomming]`
in `OccurAnal`.  Under these circumstances we group all the top-level
bindings into a single massive `Rec`.

But, crucially, I failed to repeat the occurrence analysis on this
glommed set of bindings.  That means that we weren't establishing the
right loop breakers (indeed there were no loop breakers whatsoever),
and that led immediately to the loop. The only surprising this is that
it didn't happen before.

This would occur only on Windows during `./validate`

Refs #9586