We now put the libffi objects into the RTS library, rather than trying
to mangle libffi into being a ghc package itself. It would be nicer to
 make it a separate library (but not a ghc package), but for now
 hopefully this will get the build going through on Windows again.

Now that we always link in an extra object, we may as well always
define rtsOptsEnabled in it, rather than having a default value in
rts/hooks/RtsOptsEnabled.c.

Without it we get warnings like:
    ld: warning: could not create compact unwind for .LFB3: non-standard register 5 being saved in prolog

ld gives loads of warnings like:
    ld: warning: text reloc in _base_GHCziArr_unsafeArray_info to _base_GHCziArr_unsafeArray_closure
when linking any program. We're not sure whether this is something we
ought to fix, but for now we use the -Wl,-read_only_relocs,suppress flag
to silence them.

In CorePrep we used to call the occurrence analyser to drop dead code
(see Note [Dead code in CorePrep]), but the occurrence analyser
sometimes introduces new let bindings for case binders.  This was
leading to the bug in #5433: the let binding introduced was for an
unlifted value, and the code generator is not expecting to see let
bindings of unlifted values (CorePrep is supposed to eliminate them).
We don't want this let binding anyway, so instead of using the
occurrence analyser here we have a simple custom dead-code analyser.

It broke during the ic_exports tidyup
(e.g. commit 5cd39aa3).

Add documentation describing all the variables that contain options
for Haskell compilations, what they mean and where they are (or can
be) defined.

In due course we should expand this to cover all the build system
variables, and move it to the wiki, but this is a start.

replacing them by the appropriate flag.

Suggested by Trac #5526.

in response to Neil's suggestion on Trac #4877

Also removed the type argument to Any in primops.txt.pp. I don't
see why we need it, and it now breaks haddocking GHC.Prim.

We now have addrToAny# rather than addrToHValue#, and both addrToAny#
and mkApUpd0# return "Any" rather than "a". This makes it a little
easier to see what's going on, and fixes a warning in ByteCodeLink.