Fix sparc breakage caused by recent (DLLery) NCG hackery.  :-)

Make the _ret symbols static; there's no need to have these exported.

Fix type casts w/ CIndex assignments.  Should eliminate some warnings
in via-C compilations.

* A shell-classic: Inserted space before ] (the lexical syntax of
  shells is always resource of miracles...)

* Made searching for a catalog a little bit more cunning: If
  SGML_CATALOG_FILES is not set, try FPTOOLS_CATALOG_FILE, then
  /usr/share/sgml/CATALOG.docbkdsl, the SuSE-Linux default, don't know
  about Red Hat, Debian, etc.  This way "make dvi html pdf ps rtf"
  works out of the box, at least for me. :-)

Add parsec subdir.

'-#include <foo.h>' gobbled the following option. Fixed.

Teach the NCG about the dereferencing and naming conventions to be
used when compiling for a DLLised world.  Some cleanups on the way
too.  The scheme is that

* All CLabels which are in different DLLs from the current module
  will, via the renamer, already be such that labelDynamic returns
  True for them.

* Redo the StixPrim/StixMacro stuff so that all references to symbols
  in the RTS are via CLabels.  That means that the usual labelDynamic
  story can be used.

* When a label is printed in PprMach, labelDynamic is consulted, to
  generate the __imp_ prefix if necessary.

* In MachCode.stmt2Instrs, selectively ask derefDLL to walk trees
  before code generation and insert deferencing code around other-DLL
  symbols.

* When generating Stix for SRTs, add 1 to other-DLL refs.

* When generating static closures, insert a zero word before
  the _closure label.

New version of hGetLine that is roughly 4 times faster than the
original, and is tail-recursive to boot.

I'm not entirely happy with the code, but it needs to get some testing.

Add addrOf_ErrorHdrHook prototype

Make stackOverflow prototype available Haskell compilation

cosmetic only

Remove -noC, add -package concurrent to simpl006_HC_OPTS.

Upd_frame ==> upd_frame

$(SGML_DOC) can only support one document, so build the Build System
Guide by default.

doc-building fix.

Defer ambiguity test for class members, so that the test is done *after*
we know the (extended) functional dependencies (this is a patch that was
mostly carried over from hugs98, but a few lines were out of place).

Unbreak util package dependencies for non-M$ platforms

Upd_frame ==> upd_frame

(somehow this file got missed from yesterday's commit)

Put the -c flag in the right place.

Added a description of SGML comments.

Added FPTOOLS_CATALOG_FILE

Made it buildable again; notes on installing DocBook/Jade removed for the moment.

Added docbook-cheat-sheet

Added -package concurrent.

Now that -ddump-* dumps to stdout, the stderr/out files need
reorganising.

Added import PrelIOBase, which is now needed.

Add util packages to memo tests.

Package fix: don't have posix as a dependency of util on Windows (where
posix doesn't work).

Much pain in order to get the horrid heap of brokenness that is the DocBook
catalog system to work.

oops... better unbreak the driver

Fixed missing case in dependency analysis for an implicit param.

Adjust treatment of rules in SimplCore to enable a Core pass to alter
them if necessary.  Use tricks to ensure that the common case (no change)
is still efficient.

I lied earlier.  _ccall_GC_ should work now.

wibble

fix bug in module initialization

add bLOCK_SIZE

add CurrentTSO, CurrentNursery, bdescr->start and bdescr->free offsets.

forgot a file in the Upd_frame ==> upd_frame commit.

Change the order of building stubs vs. the main .hc file so that one
doesn't trample on the other.  This stuff is *so* horrible.

callback is a definite via-C test