ieee-flpt.h is in ../../includes, not ../includes.

GHC gets upset if you have types like

	Eq a => a->a

where 'a' is *not* universally quantified.
By "upset" I mean that the typechecker generates
rather bogus code, that subsequently kills Lint.

Such types used to be rejected in the renamer,
but Jeff removed that in favour of an ambiguity
check in TcMonoType.  I remember agreeing to move
the renamer check to the type checker, and Jeff did this,
but the check in TcMonoType was only checking for
*ambiguity*, which isn't quite the same.

I've restored the missing check and commented it
better in TcMonoType.

Jeff: if this isn't right for you, let's dicuss.

Incidentally, I also generalise the ambiguity check
to detect
	forall a. ?x::a => Int
which is ambiguous.

I did a few formatting changes too.

Don't group variable bindings with the same name together.  They
should be flagged as duplicate definitions.

remove NonExhaustiveGuards exception

Guard failure will now throw a PatternMatchFail, the
NonExhaustiveGuards exception is to be removed.

Part of the -split-objs commit that I forgot yesterday.

In my commit of 24 May I got this boolean condition
back to front:

    tryWW non_rec fn_id rhs
      | not (isNeverInlinePrag inline_prag)
      =  -- Don't split things that will never be inlined

The 'not' is obviously wrong!  As a result virtually nothing
is being worker-wrapper'd

How this has survived for more than two weeks beats me.

Commiting version of the STG->GOO code generator that works with fib.

autoconf hackery for CTypes{,ISO} (continued)

autoconf hackery for CTypes{,ISO}

Remove out-of-date table that conflicts with the new one in the FFI
documentation.

Commented out the revision history.

Improved link to Foreign hslibs stuff.

Tidied up a table.

Better chapter name.

Added new FFI story from fptools/docs, and removed ccall docs.

Pls mrg

Make object file splitting simpler, in preparation for conversion to
the new driver.

The "inject split markers" phase is now omitted, instead we generate
the split markers directly.

Driver: also removed now-defunct -fpedantic-bottoms flag.

Changed ffi doc to worker-wrapper format.

remove unneeded file.

update interface file syntax

update expected output

Move read010 to rename/should_fail; this is really a renamer test.

DocBooked and moved here from hdirect/doc.

Fix printing of unfoldings in hi-files: lambda binders are now grouped
again like they used to be, rather than printed one-at-a-time.

Adding types to the names inside the GOO.

All needed for a langauge with unboxed types ...

More wibbles towards compiling data constructors and unboxing correctly.

Significant reworking of Java code generator, towards
getting unboxing working.

The n-th attempt to get GCD right, this time negative Int arguments in
combination with zero. Another bug in this area and I'll leap out of
the window immediately...  :-P

Synched with reality

Added tests for gcdInt/gcdInteger

Fixed braino.

The GCD saga continues: Fixed gcdInt/gcdInteger to work correctly for
non-positive values, i.e. call mpn_gcd_1 only with positive inputs,
and issue an error if both operands are zero. This should fix the bug
reported by Marc.

Unbreak the driver again

Class decls weren't taking functional dependencies into account when
checking for ambiguous type vars.  Fixed.

Updated DocBook notes to reflect heroic efforts of Messrs Panne and
Chakravarty.

Added new greencard package.

checkTyCon shouldn't look through newtypes.  Instead, we add a new
function checkRepTyCon which does, and use it in the few cases where
it is needed.

This fixes a nasty bug with overloaded literals of newtypes, and
probably a whole bunch of other lurking bugs too.

add new flag -fglasgow-exts-no-lang, which is the same as
-fglsagow-exts but doesn't include the implicit -package lang.  This
is needed only for compiling ghc/lib/std, where hslibs/lang may or may
not exist.

Fix bug in new version of hGetLine: it didn't handle unbuffered
handles properly.

The fix involves re-importing the old code for hGetLine :-(

While adding a syslib, setup include dirs in addition to cc injects, too.
The driver is really a piece from hell... :-(