Summary:
Previously, there were two variants of CLOSURE in C--:

    - Top-level CLOSURE(foo_closure, foo, lits...), which defines a new
      static closure and gives it a name, and

    - Array CLOSURE(foo, lits...), which was used for the static char
      and integer arrays.

They used the same name, were confusing, and didn't even generate
the correct internal label representation!  So now, we have two
new forms:

    - Top-level CLOSURE(foo, lits...) which automatically generates
      foo_closure (along with foo_info, which we were doing already)

    - Array ANONYMOUS_CLOSURE(foo, lits...) which doesn't generate
      a foo_closure identifier.

Part of remove HEAP_ALLOCED patch set (#8199)

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

Test Plan: validate

Reviewers: simonmar, austin

Subscribers: simonmar, ezyang, carter, thomie

Differential Revision: https://phabricator.haskell.org/D264

GHC Trac Issues: #8199

Summary:
The primary reason for doing this is assisting debuggability:
if static closures are all in the same section, they are
guaranteed to be adjacent to one another.  This will help
later when we add some code that takes section start/end and
uses this to sanity-check the sections.

Part of remove HEAP_ALLOCED patch set (#8199)

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

Test Plan: validate

Reviewers: simonmar, austin

Subscribers: simonmar, ezyang, carter, thomie

Differential Revision: https://phabricator.haskell.org/D263

GHC Trac Issues: #8199

Part of remove HEAP_ALLOCED patch set (#8199)

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

Test Plan: validate

Reviewers: simonmar, austin

Subscribers: simonmar, ezyang, carter, thomie

Differential Revision: https://phabricator.haskell.org/D262

GHC Trac Issues: #8199

Summary:
Most of the changes is adaptation of old Python 2 only code.
My priority was not breaking Python 2, and so I avoided bigger
changes to the driver. In particular, under Python 3 the output
is a str and buffering cannot be disabled.

To test, define PYTHON=python3 in testsuite/mk/boilerplate.mk.

Thanks to aspidites <emarshall85@gmail.com> who provided the initial patch.

Test Plan: validate under 2 and 3

Reviewers: hvr, simonmar, thomie, austin

Reviewed By: thomie, austin

Subscribers: aspidites, thomie, simonmar, ezyang, carter

Differential Revision: https://phabricator.haskell.org/D233

GHC Trac Issues: #9184

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

Summary: Using `dropWhileEndLE` tends to be faster and easier to read
than the `reverse . dropWhile p . reverse` idiom. This also cleans up
some other, nearby, messes. Fix #9616 (incorrect number formatting
potentially leading to incorrect numbers in output).

Test Plan: Run validate

Reviewers: thomie, rwbarton, nomeata, austin

Reviewed By: nomeata, austin

Subscribers: simonmar, ezyang, carter, thomie

Projects: #ghc

Differential Revision: https://phabricator.haskell.org/D259

GHC Trac Issues: #9623, #9616

Conflicts:
	compiler/basicTypes/OccName.lhs

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>

After 23bb9046, these were slightly busted for `c-mode`.

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

Summary:
When we run `./validate`, we are typically given an incredibly large
heap of information, a large majority of which isn't really
necessary. In particular, we don't really care about what `make` is
doing, nor `ghc` itself most of the time.

This reduces some of the output by making `./validate` quietier. By
running:

  $ ./validate --quiet

you'll enable `V=0` in the build, suppressing compiler messages, and
you will suppress `make` commands by running `make` in 'silent
mode'. It also runs the testsuite with `VERBOSE=2` to avoid extra
lines. This alone makes quite a difference for build log sizes.

Furthermore, by making the build logs less verbose, life is easier for
systems like Harbormaster and Travis-CI, which dislike dealing with
logs that are 10k lines or more.

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

Test Plan: iiam

Reviewers: hvr, nomeata, ezyang

Reviewed By: ezyang

Subscribers: simonmar, ezyang, carter, thomie

Projects: #ghc

Differential Revision: https://phabricator.haskell.org/D298

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

Closes #9561.

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

analogously to mapM. Fixes #9546.

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

There seem to be various issues with general fold/cons and even
cons/build rules, but it seems pretty clear to me that the simple
fold/cons/build rule is a good idea. It doesn't do much for nofib
allocation, but it seems to improve some other analyses and speed
several benchmarks up. Implements #9536.

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

This fixes #9502.

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

This fixes #9355.

David Feuer authored 10 years ago and Joachim Breitner committed 10 years ago

in order to make its RULES semantics preserving. This fixes #9495.

The code is smaller and simpler now.  Thanks to Richard for
raising the question.

to get David’s code improvements (#9654).

This reverts commit 39b5c1cb.

This completes the `Foldable` class by two important operations which
this way can be optimised for the underlying structure more easily.

A minor fix for the `containers` submodule was needed to due name clash

Addresses #9621

Reviewed By: ekmett, dfeuer, austin

Differential Revision: https://phabricator.haskell.org/D250

This was missed in D253 / a07ce165

Generalise `when`/`unless`from `Monad` to `Applicative` and `guard` from
`MonadPlus` to `Alternative` respectively.

This was made possible by the AMP and is somewhat related to #9586
(but generalising in the context of the AMP instead of the FTP)

Reviewed By: ekmett, austin

Differential Revision: https://phabricator.haskell.org/D253

The previous fix was incorrectly eta-reduced.

This is made possible by the AMP, as we don't need the `WrappedMonad`
helper for that anymore.

Approved-by: Edward Kmett <ekmett@gmail.com>

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

Don't export `getUs` and `getUniqueUs`. `UniqSM` has a `MonadUnique` instance:

    instance MonadUnique UniqSM where
        getUniqueSupplyM = getUs
        getUniqueM  = getUniqueUs
        getUniquesM = getUniquesUs

Commandline-fu used:

    git grep -l 'getUs\>' |
        grep -v compiler/basicTypes/UniqSupply.lhs |
        xargs sed -i 's/getUs/getUniqueSupplyM/g

    git grep -l 'getUniqueUs\>' |
        grep -v combiler/basicTypes/UniqSupply.lhs |
        xargs sed -i 's/getUniqueUs/getUniqueM/g'

Follow up on b522d3a3

Reviewed By: austin, hvr

Differential Revision: https://phabricator.haskell.org/D220

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