GitLab

Import wibbles

In nubBy, put argument to eq in the standard order

Fix so that it compiles on both pre and post 5.00 series compilers.

Comment out an uncomfortable ASSERT for now.

Add more symbols for mingwin version to get it to work. This still needs
tidying up.

Remove hacked-in definition of system for mingwin; instead we now just link
against the new version of systemCmd from lib/std/cbits/system.c.

----------------
	Squash newtypes
	----------------

This commit squashes newtypes and their coerces, from the typechecker
onwards.  The original idea was that the coerces would not get in the
way of optimising transformations, but despite much effort they continue
to do so.   There's no very good reason to retain newtype information
beyond the typechecker, so now we don't.

Main points:

* The post-typechecker suite of Type-manipulating functions is in
types/Type.lhs, as before.   But now there's a new suite in types/TcType.lhs.
The difference is that in the former, newtype are transparent, while in
the latter they are opaque.  The typechecker should only import TcType,
not Type.

* The operations in TcType are all non-monadic, and most of them start with
"tc" (e.g. tcSplitTyConApp).  All the monadic operations (used exclusively
by the typechecker) are in a new module, typecheck/TcMType.lhs

* I've grouped newtypes with predicate types, thus:
	data Type = TyVarTy Tyvar | ....
		  | SourceTy SourceType

	data SourceType = NType TyCon [Type]
			| ClassP Class [Type]
			| IParam Type

[SourceType was called PredType.]  This is a little wierd in some ways,
because NTypes can't occur in qualified types.   However, the idea is that
a SourceType is a type that is opaque to the type checker, but transparent
to the rest of the compiler, and newtypes fit that as do implicit parameters
and dictionaries.

* Recursive newtypes still retain their coreces, exactly as before. If
they were transparent we'd get a recursive type, and that would make
various bits of the compiler diverge (e.g. things which do type comparison).

* I've removed types/Unify.lhs (non-monadic type unifier and matcher),
merging it into TcType.

Ditto typecheck/TcUnify.lhs (monadic unifier), merging it into TcMType.

---------------------------
	Add a new case optimisation
	---------------------------

I found that lib/std/PrelCError had a case-expression that was
generating terrible code.   Something like this

	x | p `is` 1 -> e1
	  | p `is` 2 -> e2
	...etc...

where @is@ was something like

	p `is` n = p /= (-1) && p == n

This gave rise to a horrible sequence of cases

	case p of
	  (-1) -> $j p
	  1    -> e1
	  DEFAULT -> $j p

and similarly in cascade for all the join points!


Solution: add the following transformation:

	case e of		=====>     case e of
	  C _ -> <expr>			     D v -> ....v....
	  D v -> ....v....		     DEFAULT -> <expr>
	  DEFAULT -> <expr>

The point is that we merge common RHSs, at least for the DEFAULT case.
[One could do something more elaborate but I've never seen it needed.]

This transformation is implemented in SimplUtils.mkCase


*** WARNING ***
	To make this transformation easy, I have switched the convention
	for DEFAULT clauses.  They must now occur FIRST in the list of
	alternatives for a Core case expression.  (The semantics is
	unchanged: they still are a catch-all case.)

	The reason is that DEFAULT clauses sometimes need special treatment,
	and it's a lot easier to find them at the front.

	The easiest way to be insensitive to this change is to use
	CoreUtils.findDefault to pull the default clause out.

I've made the (surprisingly few) changes consequent on this changed
of convention, but they aren't in this commit.  Instead they are part
of the big commit on newtypes I'm doing at the same time.

----------------------------------
	Fix a predicate-simplification bug
	----------------------------------

Fixes a bug pointed out by Marcin

    data R = R {f :: Int}
    foo:: (?x :: Int) => R -> R
    foo r = r {f = ?x}

    Test.hs:4:
	Could not deduce `?x :: Int' from the context ()
	arising from use of implicit parameter `?x' at Test.hs:4
	In the record update: r {f = ?x}
	In the definition of `foo': r {f = ?x}

The predicate simplifier was declining to 'inherit' an
implicit parameter.  This is right for a let-binding, but
wrong for an expression binding.  For example, a simple
expression type signature:

		(?x + 1) :: Int

This was rejected because the ?x constraint could not be
floated out -- but that's wrong for expressions.

One RCS  will do

With -no-hs-main, don't force PrelMain.o to be linked (=> Main.o)

Use libexecdir in scripts, not libdir
Note: hsc2hs has to be changed, too...

Resurrect non-M$ compilation

Add some extra tests and correct some error messages. This file is no longer
used for its original purpose, but is a useful example of how to start and
manage processes under Windows.

Add machinery to copy system.c to main/ from lib/std/cbits/ when building
GHC on Windows with GHC<5.01, so that the new implementation of system can
be used rather than the old.

Instead of using the old kludgedSystem on Windows, use the new system. This
makes the use of DOS built-ins such as copy work, which they didn't when the
command was run under sh (as the old kludgedSystem did).

Move include of config.h to top so it works for mkdependC.

Correct path to NCG.h

Add wsock32 to list of libraries needed for rts under Windows.

Fix paths to HsStd.h.

Fix typo.

For GHCen < 5.00, assume that Exception.throwTo is Exception.raiseInThread

Forced commit to note that the previous commit also adds flushing of
stdout/stderr before printing an exception.  This fixes the problem
reported by Leon Smith last week.

Fix Unix imports

Adapted to changes in build system

Make it compile

More windows wibbles

* Restore SysTools.system, which implements a kludged version
  of system for reasons that are explained at length in the comments
  [overlong command-lines fail if compiling GHC with pre-5.02 GHCs]

* Wibble in Makefile

Restore an #ifdef mingw32_TARGET_OS that shouldn't have been removed
(but do it better this time and don't use an arch-specific #ifdef).

- separate location installed binaries from non-binaries (previously
  installed package.conf was in $(libdir)/extra-bin).

- remove v_Path_Perl so that we don't accidentally wire-in the location of
  perl on Unix systems.  It wasn't actually used anywhere.

- minor non-functional cleanups and comment fixups.

- We still look for package.conf twice; I haven't done anything about
  this yet.

import Packages

override $(libexecdir) to be $(libdir)/bin (the installation machinery
will have to catch up with this too).

Some tidying up

* Remove CmStaticInfo
   - GhciMode moves to HscTypes
   - The package stuff moves to new module main/Packages.lhs

[put any package-related stuff in the new module]

* Add Outputable.docToSDoc

Add getExecDir to return current directory of executable on Windows (to find
config information).

Windows wibbles

- CURRENT_DIR isn't used any more
- TmpFiles doesn't exist

fix typo in Simon's commit (cGHC_RAWCPP ==> GHC_RAWCPP)

----------------------
	Installation packaging
	----------------------

GHC runs various system programs like
	cp, touch
	gcc, as, ld etc

On Windows we plan to deliver these programs along with GHC,
so we have to be careful about where to find them.

This commit isolates all these dependencies in a single module

	main/SysTools.lhs

Most of the #ifdefery for mingw has moved into this module.
There's some documentation in SysTools.lhs

Along the way I did lots of other cleanups.  In particular

  * There is no more 'globbing' needed when calling runSomething
  * All file removal goes via the standard Directory.removeFile
  * TmpFiles.hs has gone; absorbed into SysTools
  * Some DynFlag stuff has moved from DriverFlags to CmdLineOpts


Still to do:

  **	I'm a bit concerned that calling removeFile one at a time
	when deleting masses of split-object files is going to be
	rather slow

  **	GHC now expects to find split,mangle,unlit in
		libdir/extra-bin
	instead of just
		libdir

	So something needs to change in the Unix installation scripts

  **    The "ineffective C preprocessor" is a perversion and should die

just move some code around.

comment fixes