Merge backend-hacking-branch onto HEAD.  Yay!

Work around stricter CPP in gcc 3.4.1

Fix to previous commit

wibble

avoid string gaps

Avoid string gaps

Avoid string gaps.

Avoid string gaps in CPP'd source.

Take the timestamp of an interpreted linkable from the timestamp of
the source file.  This works better when the local clock is out of
sync with the filesystem, and it's just as accurate.

Finally fix this problem:

make: *** No rule to make target `Text/Regex/Posix_hsc_make.c', needed by `depend'.  Stop.

caused by temporary droppings from hsc2hs.

Windows fixes.  There were some strange things going on with Windows
installation directories - I'm sure I've mucked up whatever was
happening with 'make install' for GHC on Windows, but there were
defintitely some bugs in here.  eg. we shouldn't have been setting
datadir to $(prefix)/imports unconditionally on Windows.

Make install now works for Happy, Alex and Haddock.  GHC probably
needs a bit of tweaking to get it working again.

Add $(Windows), which is "YES" if the host platform is a Windows
platform (mingw32 or cygwin32), and "NO" otherwise.

Fix problem with inline foreign-call changes yesterday.  Foreign call
args sometimes have to be modified using shimFCallArg - nowadays this
is done at code generation time, whereas it used to be done at
pretty-printing time.

Two changes:

  - newly created or unblocked threads go to the back of the run
    queue.  This is fairer because it avoids possible starvation.
    Fixes problem with conc004 test going much slower with optimisation.

  - The context switch flag is only reset after a yield.  Previously,
    it was being reset before we ran a thread, which meant that a tick
    that arrived during GC would be lost, and a thread doing lots of
    GC could hog the CPU.  Now, GC is effectively counted as part of
    the last running thread's timeslice.

MERGE TO STABLE.

Allow case-of-unsafe-ccall to compile to straight-line code, like it
used to.  This has already been fixed on the backend-hacking-branch,
but I'm doing it here so that it can be merged into the STABLE branch,
where it will help to work around a bug.

The bug is in CgExpr.lhs:primRetUnboxedTuple, which picks temporaries
to assign the result of a ccall to.  The Cg monad doesn't have a uniq
supply (in the HEAD), so we always pick the same temporaries.  This
leads to clashes in complex function with multiple ccalls.  Again,
this is already fixed in the backend-hacking-branch.  I don't see an
easy fix for this bug.

The compilation of case-of-unsafe-ccall doesn't suffer from this
problem, and it will help work around some cases of the bug, so I'm
going to merge this onto the STABLE branch after some testing.

Fix scoped type variable example.

Put a "-- " before the location info, since the rest of the output
looks like Haskell source.

DocBook XML is case-sensitive. Note that this is still not correct DocBook, tons
of indexterm elements within varlistentries are still misplaced.

xmlize sgml docbooks

Update from Robert Ennals: allow either the ctags or etags file to be
generated, so as to be more friendly to case-insensitive filesystems.

Bugfix. --auto-ghci-libs doesn't work if the package was previously installed.

Fixed dvips commandline

MERGE TO STABLE

Add case for foreign types

Fix foreign-type predicate

Comments only

Fix GHCi buglet when browsing module

To make Text.Regex.Posix work with Hugs, move cbits/regex/regex.h under
include, where Hugs can find it.

Make total_allocated be an ullong, to accommodate programs that do a lot
of allocation.

MERGE TO STABLE

Wibble to :i for (,); make algTyConRhs behave right

Add location information to :i command

-------------------------------
       Sort out the :i command for GHCi
	-------------------------------

The :info command has been broken in the HEAD for some time, since the new IfaceSyn
story.  This commit sorts it out, and makes it nicer than before. For example, when
you :i a record selector, you get a cut-down data type declaration, so you can see
the context.

do the file writes one after another, so if the underlying file system
happens to be case-insensitive we won't run into a file locking issue.

MERGE TO STABLE

Template Haskell improvements

a) Make '() and '[] work.
b) Add tupleTypeName, tupleDataName
b) Try to improve error message for (lack of) existential data constructors in TH

Comments only

oops, back out rev. 1.31

awaitEvent(): for Integrity, use socketpair() instead of pipe() to implement socket()-aborting mechanism

Add a (rather vague) FAQ entry about the readline build problem
reported in bug #988585.

back out rev 1.61; to hardwire path to gcc, use --with-gcc=<foo>

tidy up gcc warnings re: %-related format specifiers, mainly rewriting \%%\% to %%%%

Workaround for hbc: It compiles matching against 0-ary constructors in
list comprehensions into wrong code, e.g.

   [ () | True <- [False] ]

results in something like

   Error: No match in I4680Pinteractive

when issued in hbi.