Flags cleanup.

Basically the purpose of this commit is to move more of the compiler's
global state into DynFlags, which is moving in the direction we need
to go for the GHC API which can have multiple active sessions
supported by a single GHC instance.

Before:

$ grep 'global_var' */*hs | wc -l
     78

After:

$ grep 'global_var' */*hs | wc -l
     27

Well, it's an improvement.  Most of what's left won't really affect
our ability to host multiple sessions.

Lots of static flags have become dynamic flags (yay!).  Notably lots
of flags that we used to think of as "driver" flags, like -I and -L,
are now dynamic.  The most notable static flags left behind are the
"way" flags, eg. -prof.  It would be nice to fix this, but it isn't
urgent.

On the way, lots of cleanup has happened.  Everything related to
static and dynamic flags lives in StaticFlags and DynFlags
respectively, and they share a common command-line parser library in
CmdLineParser.  The flags related to modes (--makde, --interactive
etc.) are now private to the front end: in fact private to Main
itself, for now.

merge rev. 1.6.2.1, simplified slightly:

  Initialise a CostCentreStack by generating SIZEOF_CostCentreStack
  (gotten from the C compiler) zeros, padded to the nearest word.
  Improves on the previous fixes for unpredictable padding (see comment).

Add SIZEOF_CostCentreStack

Clarify -main-is documentation

Fix a few CCCS that should be W_[CCCS] that I just noticed.  I suspect
that heap-profiling STM code might run into difficulties in 6.4.

Fix +RTS -xc, which is broken in 6.4.

Rearrange the fields of CostCentreStack to leave no (or less) room for
arbitrary padding decisions by C compilers.

Re-clarify hs-boot docs

Follow-up fix to this:
  ghc --make should check for a module graph that
  contains a hs-boot file with no corresponding parent module

Namely, instead of complaining, just add the parent module.

	Merge to STABLE

Clarify hs-boot docs

minor to fix to (still probably broken) stglint

ghc --make should check for a module graph that
  contains a hs-boot file with no corresponding parent module

	Merge to STABLE

Buglet in compiling hs-boot files
	We should make GlobalIds not LocalIds

	Merge to STABLE

Comments

Add note about GLUT from Claus (actually I have very little idea if
this is correct or not - Sven, if you have a chance to take a look
that would be great.)

Localise the binder for the nested defn in LiberateCase,
to avoid name clashes between top-level names.
Discovered by Laszlo.

	Merge to STABLE

Further to previous patch: only create the user database if we're modifying

Be a bit less eager about creating the user database: now we only
create it when explicitly asked to use it with --user.

Fix InstalledPackageInfo link

Remove Main module from exported-modules field

----------------------------------
 	Two GADT error-reporting bugs
	----------------------------------

	Merge to STABLE

...missed file...

1.  Bug in kind-checking for GADTs; turned out to be in
    isOpenTypeKind on KindVars

....

[mingw only]setProgName():
  robustify by looking for the last '/' or '\\' in argv[0].
  My copy of 'gdb' likes to use a mixture of the two, which causes a
  fatal error when trying to debug libHSrts_p.a. Other process-invoking
  apps/shells might be equally wavering when it comes to the use of slashes.

Merge to STABLE.

all modules from GHC are listed in exposed-modules field

TARGET->HOST

----------------------------------
 	Two GADT error-reporting bugs
	----------------------------------

	Merge to STABLE

1.  Bug in kind-checking for GADTs; turned out to be in
    isOpenTypeKind on KindVars

2.  Missed check for the return type for GADTs

package.conf.in is changed to the new syntax. The list of exposed modules still
have only one module GHC

A very rough start at the GHC API, here so that others can look and
modify it.

Fix (and test) for SourceForge bug 1161624: erroneous rejection of

	foo = proc x -> arr (\y -> y-1) -< x

Now open a new level for the left side of -<, so that variables bound
in the proc are illegal, but variables bound in the expression are OK.

Note that the levels gimmick doesn't really implement holes in the
scope: it rules out nasty obfuscations like

	foo x = proc x -> arr (\y -> x-1) -< x

Also added the same treatment to the head of a `form', where it was
missing.

(for STABLE)

Without semicolon after "foundit:" at least mingw32-gcc raises "error: label at end of compound statement".

Fix obscure_ccall

Remove locale setting, it is no longer required

configure-bin.ac: I didn't know this file existed.
                  Add powerpc-linux and powerpc64-linux targets.

oops, forgot to clean runhaskell

----------------------------------
 	Attend to fixity of '->' in types
	----------------------------------

	Merge to STABLE

Another wibble to the infix-type-constructor story. Actually
this has been a bug for some time: function type constructors
were not being re-associated, because they are not HsOpAppTys.

----------------------------------
 	Fix a "class used as a type" crash
	----------------------------------

	Merge to STABLE

Thanks to Shae for finding this one.  tcfail134 tests.

Implement foreign import wrapper on x86_64

Define __CYGWIN32__ for nhc98 build of hsc2hs.

x86: For some reason the code for obscure_ccall_ret_code was allocated
dynamically and specified using literal bytes rather than inline
assembly.  Change it to use inline assembly.

Clean .hi-boot and .o-boot files when appropriate.  The standard
cleaning doesn't work in ghc/compiler because of $(odir).

Add x86_64 platforms