gc4.1 tarball importgc4_1

72 files changed, 20252 insertions, 0 deletions

diff --git a/Makefile b/Makefile
new file mode 100644
index 00000000..feb83e60
--- /dev/null
+++ b/Makefile
@@ -0,0 +1,205 @@
+# Primary targets:
+# gc.a - builds basic library
+# c++ - adds C++ interface to library and include directory
+# cords - adds cords (heavyweight strings) to library and include directory
+# test - prints porting information, then builds basic version of gc.a, and runs
+#        some tests of collector and cords.  Does not add cords or c++ interface to gc.a
+# cord/de - builds dumb editor based on cords.
+CC= cc
+CXX=g++
+# Needed only for "make c++", which adds the c++ interface
+
+CFLAGS= -O -DALL_INTERIOR_POINTERS -DSILENT
+# Setjmp_test may yield overly optimistic results when compiled
+# without optimization.
+# -DSILENT disables statistics printing, and improves performance.
+# -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
+#   altered stubborn objects, at substantial performance cost.
+# -DFIND_LEAK causes the collector to assume that all inaccessible
+#   objects should have been explicitly deallocated, and reports exceptions
+# -DSOLARIS_THREADS enables support for Solaris (thr_) threads.
+#   (Clients should also define SOLARIS_THREADS and then include
+#   gc.h before performing thr_ or GC_ operations.)
+# -DALL_INTERIOR_POINTERS allows all pointers to the interior
+#   of objects to be recognized.  (See gc_private.h for consequences.)
+# -DSMALL_CONFIG tries to tune the collector for small heap sizes,
+#   usually causing it to use less space in such situations.
+#   Incremental collection no longer works in this case.
+# -DDONT_ADD_BYTE_AT_END is meaningful only with
+#   -DALL_INTERIOR_POINTERS.  Normally -DALL_INTERIOR_POINTERS
+#   causes all objects to be padded so that pointers just past the end of
+#   an object can be recognized.  This can be expensive.  (The padding
+#   is normally more than one byte due to alignment constraints.)
+#   -DDONT_ADD_BYTE_AT_END disables the padding.
+
+AR= ar
+RANLIB= ranlib
+
+
+# Redefining srcdir allows object code for the nonPCR version of the collector
+# to be generated in different directories
+srcdir = .
+VPATH = $(srcdir)
+
+OBJS= alloc.o reclaim.o allchblk.o misc.o mach_dep.o os_dep.o mark_rts.o headers.o mark.o obj_map.o blacklst.o finalize.o new_hblk.o dyn_load.o dbg_mlc.o malloc.o stubborn.o checksums.o solaris_threads.o typd_mlc.o
+
+CSRCS= reclaim.c allchblk.c misc.c alloc.c mach_dep.c os_dep.c mark_rts.c headers.c mark.c obj_map.c pcr_interface.c blacklst.c finalize.c new_hblk.c real_malloc.c dyn_load.c dbg_mlc.c malloc.c stubborn.c checksums.c solaris_threads.c typd_mlc.c
+
+CORD_SRCS=  cord/cordbscs.c cord/cordxtra.c cord/cordprnt.c cord/de.c cord/cordtest.c cord/cord.h cord/ec.h cord/cord_pos.h cord/de_win.c cord/de_win.h cord/de_cmds.h cord/de_win.ICO cord/de_win.RC
+
+CORD_OBJS=  cord/cordbscs.o cord/cordxtra.o cord/cordprnt.o
+
+SRCS= $(CSRCS) mips_mach_dep.s rs6000_mach_dep.s alpha_mach_dep.s sparc_mach_dep.s gc.h gc_typed.h gc_hdrs.h gc_priv.h gc_private.h config.h gc_mark.h gc_inl.h gc_inline.h gc.man if_mach.c if_not_there.c gc_c++.cc gc_c++.h $(CORD_SRCS)
+
+OTHER_FILES= Makefile PCR-Makefile OS2_MAKEFILE NT_MAKEFILE \
+           README test.c setjmp_t.c SMakefile.amiga SCoptions.amiga \
+           README.amiga README.win32 cord/README include/gc.h \
+           include/gc_typed.h README.QUICK callprocs pc_excludes \
+           barrett_diagram README.OS2
+
+CORD_INCLUDE_FILES= $(srcdir)/gc.h $(srcdir)/cord/cord.h $(srcdir)/cord/ec.h \
+           $(srcdir)/cord/cord_pos.h
+
+# Libraries needed for curses applications.  Only needed for de.
+CURSES= -lcurses -ltermlib
+
+# The following is irrelevant on most systems.  But a few
+# versions of make otherwise fork the shell specified in
+# the SHELL environment variable.
+SHELL= /bin/sh
+
+SPECIALCFLAGS = 
+# Alternative flags to the C compiler for mach_dep.c.
+# Mach_dep.c often doesn't like optimization, and it's
+# not time-critical anyway.
+# Set SPECIALCFLAGS to -q nodirect_code on Encore.
+
+ALPHACFLAGS = -non_shared
+# Extra flags for linking compilation on DEC Alpha
+
+all: gc.a gctest
+
+pcr: PCR-Makefile gc_private.h gc_hdrs.h gc.h config.h mach_dep.o $(SRCS)
+	make -f PCR-Makefile depend
+	make -f PCR-Makefile
+
+$(OBJS) test.o: $(srcdir)/gc_priv.h $(srcdir)/gc_hdrs.h $(srcdir)/gc.h \
+    $(srcdir)/config.h $(srcdir)/gc_typed.h Makefile
+# The dependency on Makefile is needed.  Changing
+# options such as -DSILENT affects the size of GC_arrays,
+# invalidating all .o files that rely on gc_priv.h
+
+mark.o typd_mlc.o finalize.o: $(srcdir)/gc_mark.h
+
+gc.a: $(OBJS)
+	$(AR) ru gc.a $(OBJS)
+	$(RANLIB) gc.a || cat /dev/null
+#	ignore ranlib failure; that usually means it doesn't exist, and isn't needed
+
+cords: $(CORD_OBJS) cord/cordtest
+	$(AR) ru gc.a $(CORD_OBJS)
+	$(RANLIB) gc.a || cat /dev/null
+	cp $(srcdir)/cord/cord.h include/cord.h
+	cp $(srcdir)/cord/ec.h include/ec.h
+	cp $(srcdir)/cord/cord_pos.h include/cord_pos.h
+
+gc_c++.o: $(srcdir)/gc_c++.cc $(srcdir)/gc_c++.h
+	$(CXX) -c -O $(srcdir)/gc_c++.cc
+	
+c++: gc_c++.o $(srcdir)/gc_c++.h
+	$(AR) ru gc.a gc_c++.o
+	$(RANLIB) gc.a || cat /dev/null
+	cp $(srcdir)/gc_c++.h include/gc_c++.h 
+
+mach_dep.o: $(srcdir)/mach_dep.c $(srcdir)/mips_mach_dep.s $(srcdir)/rs6000_mach_dep.s if_mach if_not_there
+	rm -f mach_dep.o
+	./if_mach MIPS "" as -o mach_dep.o $(srcdir)/mips_mach_dep.s
+	./if_mach RS6000 "" as -o mach_dep.o $(srcdir)/rs6000_mach_dep.s
+	./if_mach ALPHA "" as -o mach_dep.o $(srcdir)/alpha_mach_dep.s
+	./if_mach SPARC SUNOS5 as -o mach_dep.o $(srcdir)/sparc_mach_dep.s
+	./if_not_there mach_dep.o $(CC) -c $(SPECIALCFLAGS) $(srcdir)/mach_dep.c
+
+mark_rts.o: $(srcdir)/mark_rts.c if_mach if_not_there
+	rm -f mark_rts.o
+	./if_mach ALPHA "" $(CC) -c $(CFLAGS) -Wo,-notail $(srcdir)/mark_rts.c
+	./if_not_there mark_rts.o $(CC) -c $(CFLAGS) $(srcdir)/mark_rts.c
+#	work-around for DEC optimizer tail recursion elimination bug
+
+cord/cordbscs.o: $(srcdir)/cord/cordbscs.c $(CORD_INCLUDE_FILES)
+	$(CC) $(CFLAGS) -c $(srcdir)/cord/cordbscs.c
+	mv cordbscs.o cord/cordbscs.o
+#  not all compilers understand -o filename
+
+cord/cordxtra.o: $(srcdir)/cord/cordxtra.c $(CORD_INCLUDE_FILES)
+	$(CC) $(CFLAGS) -c $(srcdir)/cord/cordxtra.c
+	mv cordxtra.o cord/cordxtra.o
+
+cord/cordprnt.o: $(srcdir)/cord/cordprnt.c $(CORD_INCLUDE_FILES)
+	$(CC) $(CFLAGS) -c $(srcdir)/cord/cordprnt.c
+	mv cordprnt.o cord/cordprnt.o
+
+cord/cordtest: $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a
+	rm -f cord/cordtest
+	./if_mach SPARC SUNOS5 $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -lthread
+	./if_not_there cord/cord_test $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a
+
+cord/de: $(srcdir)/cord/de.c $(srcdir)/cord/cordbscs.o $(srcdir)/cord/cordxtra.o gc.a
+	rm -f cord/de
+	./if_mach SPARC SUNOS5 $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c $(srcdir)/cord/cordbscs.o $(srcdir)/cord/cordxtra.o gc.a $(CURSES) -lthread
+	./if_mach RS6000 "" $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c $(srcdir)/cord/cordbscs.o $(srcdir)/cord/cordxtra.o gc.a -lcurses
+	./if_not_there cord/de $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c $(srcdir)/cord/cordbscs.o $(srcdir)/cord/cordxtra.o gc.a $(CURSES)
+
+if_mach: $(srcdir)/if_mach.c $(srcdir)/config.h
+	$(CC) $(CFLAGS) -o if_mach $(srcdir)/if_mach.c
+
+if_not_there: $(srcdir)/if_not_there.c
+	$(CC) $(CFLAGS) -o if_not_there $(srcdir)/if_not_there.c
+
+clean: 
+	rm -f gc.a test.o gctest output-local output-diff $(OBJS) \
+	      setjmp_test  mon.out gmon.out a.out core if_not_there if_mach \
+	      $(CORD_OBJS) cord/cordtest cord/de
+	-rm -f *~
+
+gctest: test.o gc.a if_mach if_not_there
+	rm -f gctest
+	./if_mach ALPHA "" $(CC) $(CFLAGS) -o gctest $(ALPHACFLAGS) test.o gc.a
+	./if_mach SPARC SUNOS5 $(CC) $(CFLAGS) -o gctest $(CFLAGS) test.o gc.a -lthread
+	./if_not_there gctest $(CC) $(CFLAGS) -o gctest test.o gc.a
+
+# If an optimized setjmp_test generates a segmentation fault,
+# odds are your compiler is broken.  Gctest may still work.
+# Try compiling setjmp_t.c unoptimized.
+setjmp_test: $(srcdir)/setjmp_t.c $(srcdir)/gc.h if_mach if_not_there
+	rm -f setjmp_test
+	./if_mach ALPHA "" $(CC) $(CFLAGS) -o setjmp_test $(ALPHACFLAGS) $(srcdir)/setjmp_t.c
+	./if_not_there setjmp_test $(CC) $(CFLAGS) -o setjmp_test $(srcdir)/setjmp_t.c
+
+test: setjmp_test gctest
+	./setjmp_test
+	./gctest
+	make cord/cordtest
+	cord/cordtest
+
+gc.tar: $(SRCS) $(OTHER_FILES)
+	tar cvf gc.tar $(SRCS) $(OTHER_FILES)
+	
+pc_gc.tar: $(SRCS) $(OTHER_FILES)
+	tar cvfX pc_gc.tar pc_excludes $(SRCS) $(OTHER_FILES)
+
+floppy: pc_gc.tar
+	-mmd a:/cord
+	-mmd a:/include
+	mkdir /tmp/pc_gc
+	cat pc_gc.tar | (cd /tmp/pc_gc; tar xvf -)
+	-mcopy -tmn /tmp/pc_gc/* a:
+	-mcopy -tmn /tmp/pc_gc/cord/* a:/cord
+	-mcopy -mn /tmp/pc_gc/cord/de_win.ICO a:/cord
+	-mcopy -tmn /tmp/pc_gc/include/* a:/cord
+	rm -r /tmp/pc_gc
+
+gc.tar.Z: gc.tar
+	compress gc.tar
+
+lint: $(CSRCS) test.c
+	lint -DLINT $(CSRCS) test.c | egrep -v "possible pointer alignment problem|abort|exit|sbrk|mprotect|syscall"
diff --git a/NT_MAKEFILE b/NT_MAKEFILE
new file mode 100644
index 00000000..2817aa5b
--- /dev/null
+++ b/NT_MAKEFILE
@@ -0,0 +1,37 @@
+# Makefile for Windows NT.  Assumes Microsoft compiler, and a single thread.
+# DLLs are included in the root set under NT, but not under win32S.
+# Use "nmake nodebug=1 all" for optimized versions of library, gctest and editor.
+
+!include <ntwin32.mak>
+
+# We also haven't figured out how to do partial links or build static libraries.  Hence a
+# client currently needs to link against all of the following:
+
+OBJS= alloc.obj reclaim.obj allchblk.obj misc.obj mach_dep.obj os_dep.obj mark_rts.obj headers.obj mark.obj obj_map.obj blacklst.obj finalize.obj new_hblk.obj dbg_mlc.obj malloc.obj stubborn.obj dyn_load.obj typd_mlc.obj
+
+all: gctest.exe cord\de.exe
+
+.c.obj:
+	$(cc) $(cdebug) $(cflags) $(cvars) -DSMALL_CONFIG -DSILENT -DALL_INTERIOR_POINTERS $*.c /Fo$*.obj
+
+$(OBJS) test.obj: gc_priv.h gc_hdrs.h gc.h
+
+gc.lib: $(OBJS)
+	lib32 /MACHINE:i386 /out:gc.lib $(OBJS)
+
+gctest.exe: test.obj gc.lib
+#	The following works for win32 debugging.  For win32s debugging use debugtype:coff
+#	and add mapsympe line.
+	$(link) -debug:full -debugtype:cv $(guiflags) -stack:131072 -out:$*.exe test.obj $(conlibs) gc.lib
+#	mapsympe -n -o gctest.sym gctest.exe
+
+cord\de_win.rbj: cord\de_win.res
+	cvtres -$(CPU) cord\de_win.res -o cord\de_win.rbj
+
+cord\de.obj cord\de_win.obj: cord\cord.h cord\cord_pos.h cord\de_win.h cord\de_cmds.h
+
+cord\de_win.res: cord\de_win.rc cord\de_win.h cord\de_cmds.h
+	$(rc) $(rcvars) -r -fo cord\de_win.res $(cvars) cord\de_win.rc
+
+cord\de.exe: cord\cordbscs.obj cord\cordxtra.obj cord\de.obj cord\de_win.obj cord\de_win.rbj gc.lib
+	$(link) -debug:full -debugtype:cv $(guiflags) -stack:16384 -out:cord\de.exe  cord\cordbscs.obj cord\cordxtra.obj cord\de.obj cord\de_win.obj cord\de_win.rbj gc.lib $(guilibs)
\ No newline at end of file
diff --git a/OS2_MAKEFILE b/OS2_MAKEFILE
new file mode 100644
index 00000000..6e0a0ac7
--- /dev/null
+++ b/OS2_MAKEFILE
@@ -0,0 +1,39 @@
+# Makefile for OS/2.  Assumes IBM's compiler, static linking, and a single thread.
+# Adding dynamic linking support seems easy, but takes a little bit of work.
+# Adding thread support may be nontrivial, since we haven't yet figured out how to
+# look at another thread's registers.
+
+# We also haven't figured out how to do partial links or build static libraries.  Hence a
+# client currently needs to link against all of the following:
+
+OBJS= alloc.obj reclaim.obj allchblk.obj misc.obj mach_dep.obj os_dep.obj mark_rts.obj headers.obj mark.obj obj_map.obj blacklst.obj finalize.obj new_hblk.obj dbg_mlc.obj malloc.obj stubborn.obj typd_mlc.obj
+
+CORDOBJS= cord\cordbscs.obj cord\cordxtra.obj cord\cordprnt.obj
+
+CC= icc
+CFLAGS= /O /Q /DSILENT /DSMALL_CONFIG /DALL_INTERIOR_POINTERS
+# Use /Ti instead of /O for debugging
+# Setjmp_test may yield overly optimistic results when compiled
+# without optimization.
+
+all: $(OBJS) gctest.exe cord\cordtest.exe
+
+$(OBJS) test.obj: gc_priv.h gc_hdrs.h gc.h
+
+mach_dep.obj: mach_dep.c
+	$(CC) $(CFLAGS) /C mach_dep.c
+
+gctest.exe: test.obj $(OBJS)
+	$(CC) $(CFLAGS) /B"/STACK:524288" /Fegctest test.obj $(OBJS)
+
+cord\cordbscs.obj: cord\cordbscs.c cord\cord.h cord\cord_pos.h
+	$(CC) $(CFLAGS) /C /Focord\cordbscs cord\cordbscs.c
+
+cord\cordxtra.obj: cord\cordxtra.c cord\cord.h cord\cord_pos.h cord\ec.h
+	$(CC) $(CFLAGS) /C /Focord\cordxtra cord\cordxtra.c
+
+cord\cordprnt.obj: cord\cordprnt.c cord\cord.h cord\cord_pos.h cord\ec.h
+	$(CC) $(CFLAGS) /C /Focord\cordprnt cord\cordprnt.c
+
+cord\cordtest.exe: cord\cordtest.c cord\cord.h cord\cord_pos.h cord\ec.h $(CORDOBJS)
+	$(CC) $(CFLAGS) /B"/STACK:65536" /Fecord\cordtest cord\cordtest.c $(OBJS) $(CORDOBJS)
\ No newline at end of file
diff --git a/PCR-Makefile b/PCR-Makefile
new file mode 100644
index 00000000..637ceb7e
--- /dev/null
+++ b/PCR-Makefile
@@ -0,0 +1,46 @@
+OBJS= alloc.o reclaim.o allchblk.o misc.o mach_dep.o os_dep.o mark_rts.o headers.o mark.o obj_map.o pcr_interface.o blacklst.o finalize.o new_hblk.o real_malloc.o dynamic_load.o dbg_mlc.o malloc.o stubborn.o
+
+CSRCS= reclaim.c allchblk.c misc.c alloc.c mach_dep.c os_dep.c mark_rts.c headers.c mark.c obj_map.c pcr_interface.c blacklst.c finalize.c new_hblk.c real_malloc.c dynamic_load.c debug_mlc.c malloc.c stubborn.c
+
+SHELL= /bin/sh
+
+# Fix to point to local pcr installation directory.
+PCRDIR= /project/ppcr/dev
+CC= gcc
+CFLAGS= -g -DPCR -I$(PCRDIR) -I$(PCRDIR)/ansi -I$(PCRDIR)/posix
+
+# We assume that mach_dep.o has already been built by top level makefile.  It doesn't
+# care about pcr vs UNIX, and we don't want to repeat that cruft.
+
+default: gc.o
+
+all: gc.o test.o gcpcr
+
+gcpcr: gc.o test.o $(PCRDIR)/base/pcr.o $(PCRDIR)/base/PCR_BaseMain.o
+	$(CC) -o gcpcr $(PCRDIR)/base/pcr.o $(PCRDIR)/base/PCR_BaseMain.o gc.o test.o -ldl
+
+gc.o: $(OBJS)
+	-ld -r -o gc.o $(OBJS)
+
+#
+# Dependency construction
+#
+# NOTE: the makefile must include "# DO NOT DELETE THIS LINE" after the
+#   last target.  "make depend" will replace everything following that line
+#   by a newly-constructed list of dependencies.
+#
+depend: $(CSRCS)
+	rm -f makedep eddep ; \
+    $(CC) -M $(CFLAGS) $(CSRCS) \
+			| sed -e '/:$$/d' > makedep ; \
+    echo '/^# DO NOT DELETE THIS LINE/+1,$$d' >eddep ; \
+	echo '$$r makedep' >>eddep ; \
+	echo 'w' >>eddep ; \
+	cp PCR-Makefile PCR-Makefile.bak ; \
+	ex - PCR-Makefile < eddep ; \
+	rm -f eddep makedep
+	touch depend
+	
+# DO NOT DELETE THIS LINE
+
+
diff --git a/README b/README
new file mode 100644
index 00000000..8cb1c44f
--- /dev/null
+++ b/README
@@ -0,0 +1,815 @@
+Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+
+THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+
+Permission is hereby granted to use or copy this program
+for any purpose,  provided the above notices are retained on all copies.
+Permission to modify the code and to distribute modified code is granted,
+provided the above notices are retained, and a notice that the code was
+modified is included with the above copyright notice.
+
+This is version 4.1 of a conservative garbage collector for C and C++.
+
+HISTORY -
+
+  Early versions of this collector were developed as a part of research
+projects supported in part by the National Science Foundation
+and the Defense Advance Research Projects Agency.
+Much of the code was rewritten by Hans-J. Boehm at Xerox PARC.
+The SPARC specific code was contributed by Mark Weiser
+(weiser@parc.xerox.com).  The Encore Multimax modifications were supplied by
+Kevin Kenny (kenny@m.cs.uiuc.edu).  The adaptation to the RT is largely due
+to Vernon Lee (scorpion@rice.edu), on machines made available by IBM.
+Much of the HP specific code and a number of good suggestions for improving the
+generic code are due to Walter Underwood (wunder@hp-ses.sde.hp.com).
+Robert Brazile (brazile@diamond.bbn.com) originally supplied the ULTRIX code.
+Al Dosser (dosser@src.dec.com) and Regis Cridlig (Regis.Cridlig@cl.cam.ac.uk)
+subsequently provided updates and information on variation between ULTRIX
+systems.  Parag Patel (parag@netcom.com) supplied the A/UX code.
+Jesper Peterson(jep@mtiame.mtia.oz.au) supplied the Amiga port.
+Thomas Funke (thf@zelator.in-berlin.de(?)) supplied the NeXT port.
+Bill Janssen (janssen@parc.xerox.com) supplied the SunOS dynamic loader
+specific code. Manuel Serrano (serrano@cornas.inria.fr) supplied linux and
+Sony News specific code.  Al Dosser provided Alpha/OSF/1 code.  He and
+Dave Detlefs(detlefs@src.dec.com) also provided several generic bug fixes.
+Alistair G. Crooks(agc@uts.amdahl.com) supplied the NetBSD and 386BSD ports.
+Jeffrey Hsu (hsu@soda.berkeley.edu) provided the FreeBSD port.
+Brent Benson (brent@jade.ssd.csd.harris.com) ported the collector to
+a Motorola 88K processor running CX/UX (Harris NightHawk).
+Ari Huttunen (Ari.Huttunen@hut.fi) generalized the OS/2 port to
+nonIBM development environments (a nontrivial task).
+David Chase, then at Olivetti Research, suggested several improvements.
+Scott Schwartz (schwartz@groucho.cse.psu.edu) supplied some of the
+code to save and print call stacks for leak detection on a SPARC.
+Jesse Hull and John Ellis supplied the C++ interface code.
+Zhong Shao performed much of the experimentation that led to the
+current typed allocation facility.  (His dynamic type inference code hasn't
+made it into the released version of the collector, yet.)
+(Blame for misinstallation of these modifications goes to the first author,
+however.)
+
+    This is intended to be a general purpose, garbage collecting storage
+allocator.  The algorithms used are described in:
+
+Boehm, H., and M. Weiser, "Garbage Collection in an Uncooperative Environment",
+Software Practice & Experience, September 1988, pp. 807-820.
+
+Boehm, H., A. Demers, and S. Shenker, "Mostly Parallel Garbage Collection",
+Proceedings of the ACM SIGPLAN '91 Conference on Programming Language Design
+and Implementation, SIGPLAN Notices 26, 6 (June 1991), pp. 157-164.
+
+Boehm, H., "Space Efficient Conservative Garbage Collection", Proceedings
+of the ACM SIGPLAN '91 Conference on Programming Language Design and
+Implementation, SIGPLAN Notices 28, 6 (June 1993), pp. 197-206.
+
+  Unlike the collector described in the second reference, this collector
+operates either with the mutator stopped during the entire collection
+(default) or incrementally during allocations.  (The latter is supported
+on only a few machines.)  It does not rely on threads, but is intended
+to be thread-safe.
+
+  Some of the ideas underlying the collector have previously been explored
+by others.  (Doug McIlroy wrote a vaguely similar collector that is part of
+version 8 UNIX (tm).)  However none of this work appears to have been widely
+disseminated.
+
+  Rudimentary tools for use of the collector as a leak detector are included, as
+is a fairly sophisticated string package "cord" that makes use of the collector.
+(See cord/README.)
+
+
+GENERAL DESCRIPTION
+
+  This is a garbage colecting storage allocator that is intended to be
+used as a plug-in replacement for C's malloc.
+
+  Since the collector does not require pointers to be tagged, it does not
+attempt to ensure that all inaccessible storage is reclaimed.  However,
+in our experience, it is typically more successful at reclaiming unused
+memory than most C programs using explicit deallocation.  Unlike manually
+introduced leaks, the amount of unreclaimed memory typically stays
+bounded.
+
+  In the following, an "object" is defined to be a region of memory allocated
+by the routines described below.  
+
+  Any objects not intended to be collected must be pointed to either
+from other such accessible objects, or from the registers,
+stack, data, or statically allocated bss segments.  Pointers from
+the stack or registers may point to anywhere inside an object.
+However, it is usually assumed that all pointers originating in the
+heap point to the beginning of an object.  (This does
+not disallow interior pointers; it simply requires that there must be a
+pointer to the beginning of every accessible object, in addition to any
+interior pointers.)  There are two facilities for altering this behavior.
+The macro ALL_INTERIOR_POINTERS may be defined in gc_private.h to
+cause any pointer into an object (or one past the end) to retain the
+object.  A routine GC_register_displacement is provided to allow for
+more controlled interior pointer use in the heap.  Defining
+ALL_INTERIOR_POINTERS is somewhat dangerous, in that it can result
+in unnecessary memroy retention.  However this is much less of a
+problem than with older collector versions.  The routine
+GC_register_displacement is described in gc.h.
+
+  Note that pointers inside memory allocated by the standard "malloc" are not
+seen by the garbage collector.  Thus objects pointed to only from such a
+region may be prematurely deallocated.  It is thus suggested that the
+standard "malloc" be used only for memory regions, such as I/O buffers, that
+are guaranteed not to contain pointers.  Pointers in C language automatic,
+static, or register variables, are correctly recognized.  (Note that
+GC_malloc_uncollectable has semantics similar to standard malloc,
+but allocates objects that are traced by the collector.)
+
+  The collector does not generally know how to find pointers in data
+areas that are associated with dynamic libraries.  This is easy to
+remedy IF you know how to find those data areas on your operating
+system (see GC_add_roots).  Code for doing this under SunOS and IRIX 5.X is
+included (see dynamic_load.c).
+
+  Note that the garbage collector does not need to be informed of shared
+read-only data.  However if the shared library mechanism can introduce
+discontiguous data areas that may contain pointers, then the collector does
+need to be informed.
+
+  Signal processing for most signals is normally deferred during collection,
+and during uninterruptible parts of the allocation process.  Unlike
+standard ANSI C mallocs, it is intended to be safe to invoke malloc
+from a signal handler while another malloc is in progress, provided
+the original malloc is not restarted.  (Empirically, many UNIX
+applications already asssume this.)  Even this modest level of signal-
+safety may be too expensive on some systems.  If so, ENABLE_SIGNALS
+and DISABLE_SIGNALS may be redefined to the empty statement in gc_private.h.
+
+  The allocator/collector can also be configured for thread-safe operation.
+(Full signal safety can also be acheived, but only at the cost of two system
+calls per malloc, which is usually unacceptable.)
+
+INSTALLATION AND PORTABILITY
+
+  As distributed, the macro SILENT is defined in Makefile.
+In the event of problems, this can be removed to obtain a moderate
+amount of descriptive output for each collection.
+(The given statistics exhibit a few peculiarities.
+Things don't appear to add up for a variety of reasons, most notably
+fragmentation losses.  These are probably much more significant for the
+contrived program "test.c" than for your application.)
+
+  Note that typing "make test" will automatically build the collector
+and then run setjmp_test and gctest. Setjmp_test will give you information
+about configuring the collector, which is useful primarily if you have
+a machine that's not already supported.  Gctest is a somewhat superficial
+test of collector functionality.  Failure is indicated by a core dump or
+a message to the effect that the collector is broken.  Gctest takes about 
+35 seconds to run on a SPARCstation 2. On a slower machine,
+expect it to take a while.  It may use up to 8 MB of memory.  (The
+multi-threaded version will use more.)  "Make test" will also, as
+its last step, attempt to build and test the "cord" string library.
+This will fail without an ANSI C compiler.
+
+  The Makefile will generate a library gc.a which you should link against.
+Typing "make cords" will add the cord library to gc.a.
+Note that this requires an ANSI C compiler.
+
+  It is suggested that if you need to replace a piece of the collector
+(e.g. GC_mark_rts.c) you simply list your version ahead of gc.a on the
+ld command line, rather than replacing the one in gc.a.  (This will
+generate numerous warnings under some versions of AIX, but it still
+works.)
+
+  All include files that need to be used by clients will be put in the
+include subdirectory.  (Normally this is just gc.h.  "Make cords" adds
+"cord.h" and "ec.h".)
+
+  The collector currently is designed to run essentially unmodified on
+the following machines (most of the operating systems mentioned are
+trademarks of their respective holders):
+
+	    Sun 3
+	    Sun 4 under SunOS 4.X or Solaris2.X (with or without threads)
+	    Vax under 4.3BSD, Ultrix
+	    Intel 386 or 486 under many operating systems, but not MSDOS.
+	    	(Win32S is somewhat supported, so it is possible to
+	    	build applications for Windows 3.1)
+	    Sequent Symmetry  (single threaded)
+	    Encore Multimax   (single threaded)
+	    MIPS M/120 (and presumably M/2000) (RISC/os 4.0 with BSD libraries)
+	    IBM PC/RT  (Berkeley UNIX)
+	    IBM RS/6000
+	    HP9000/300
+	    HP9000/700
+	    DECstations under Ultrix
+	    DEC Alpha running OSF/1
+	    SGI workstations under IRIX
+	    Sony News
+	    Apple MacIntosh under A/UX
+	    Commodore Amiga (see README.amiga)
+	    NeXT machines
+
+  In a few cases (Amiga, OS/2, Win32) a separate makefile is supplied.
+
+  Dynamic libraries are completely supported only under SunOS
+(and even that support is not functional on the last Sun 3 release),
+IRIX 5, Win32 (not Win32S) and OSF/1 on DEC AXP machines.
+On other machines we recommend that you do one of the following:
+
+  1) Add dynamic library support (and send us the code).
+  2) Use static versions of the libraries.
+  3) Arrange for dynamic libraries to use the standard malloc.
+     This is still dangerous if the library stores a pointer to a
+     garbage collected object.  But nearly all standard interfaces
+     prohibit this, because they deal correctly with pointers
+     to stack allocated objects.  (Strtok is an exception.  Don't
+     use it.)
+
+  In all cases we assume that pointer alignment is consistent with that
+enforced by the standard C compilers.  If you use a nonstandard compiler
+you may have to adjust the alignment parameters defined in gc_private.h.
+
+  A port to a machine that is not byte addressed, or does not use 32 bit
+addresses will require a major effort.  (Parts of the code try to anticipate
+64 bit addresses.  Others will need to be rewritten, since different data
+structures are needed.)  A port to MSDOS is hopeless, unless you are willing
+to assume an 80386 or better, and that only flat 32 bit pointers will ever be
+used.
+
+  For machines not already mentioned, or for nonstandard compilers, the
+following are likely to require change:
+
+1.  The parameters at the top of gc_private.h.
+      The parameters that will usually require adjustment are
+   STACKBOTTOM,  ALIGNMENT and DATASTART.  Setjmp_test
+   prints its guesses of the first two.
+      DATASTART should be an expression for computing the
+   address of the beginning of the data segment.  This can often be
+   &etext.  But some memory management units require that there be
+   some unmapped space between the text and the data segment.  Thus
+   it may be more complicated.   On UNIX systems, this is rarely
+   documented.  But the adb "$m" command may be helpful.  (Note
+   that DATASTART will usually be a function of &etext.  Thus a
+   single experiment is usually insufficient.)
+     STACKBOTTOM is used to initialize GC_stackbottom, which
+   should be a sufficient approximation to the coldest stack address.
+   On some machines, it is difficult to obtain such a value that is
+   valid across a variety of MMUs, OS releases, etc.  A number of
+   alternatives exist for using the collector in spite of this.  See the
+   discussion in config.h.h immediately preceding the various
+   definitions of STACKBOTTOM.
+   
+2.  mach_dep.c.
+      The most important routine here is one to mark from registers.
+    The distributed file includes a generic hack (based on setjmp) that
+    happens to work on many machines, and may work on yours.  Try
+    compiling and running setjmp_test.c to see whether it has a chance of
+    working.  (This is not correct C, so don't blame your compiler if it
+    doesn't work.  Based on limited experience, register window machines
+    are likely to cause trouble.  If your version of setjmp claims that
+    all accessible variables, including registers, have the value they
+    had at the time of the longjmp, it also will not work.  Vanilla 4.2 BSD
+    makes such a claim.  SunOS does not.)
+      If your compiler does not allow in-line assembly code, or if you prefer
+    not to use such a facility, mach_dep.c may be replaced by a .s file
+    (as we did for the MIPS machine and the PC/RT).
+
+3.  mark_roots.c.
+      These are the top level mark routines that determine which sections
+    of memory the collector should mark from.  This is normally not
+    architecture specific (aside from the macros defined in gc_private.h and
+    referenced here), but it can be programming language and compiler
+    specific.  The supplied routine should work for most C compilers
+    running under UNIX.  Calls to GC_add_roots may sometimes be used
+    for similar effect.
+
+4.  The sigsetmask call does not appear to exist under early system V UNIX.
+    It is used by the collector to block and unblock signals at times at
+    which an asynchronous allocation inside a signal handler could not
+    be tolerated.  Under system V, it is possible to remove these calls,
+    provided no storage allocation is done by signal handlers.  The
+    alternative is to issue a sequence of system V system calls, one per
+    signal that is actually used.  This may be a bit slow.
+
+  For a different versions of Berkeley UN*X or different machines using the
+Motorola 68000, Vax, SPARC, 80386, NS 32000, PC/RT, or MIPS architecture,
+it should frequently suffice to change definitions in gc_private.h.
+
+
+THE C INTERFACE TO THE ALLOCATOR
+
+  The following routines are intended to be directly called by the user.
+Note that usually only GC_malloc is necessary.  GC_clear_roots and GC_add_roots
+calls may be required if the collector has to trace from nonstandard places
+(e.g. from dynamic library data areas on a machine on which the 
+collector doesn't already understand them.)  On some machines, it may
+be desirable to set GC_stacktop to a good approximation of the stack base. 
+(This enhances code portability on HP PA machines, since there is no
+good way for the collector to compute this value.)  Client code may include
+"gc.h", which defines all of the following, plus a few others.
+
+1)  GC_malloc(nbytes)
+    - allocate an object of size nbytes.  Unlike malloc, the object is
+      cleared before being returned to the user.  Gc_malloc will
+      invoke the garbage collector when it determines this to be appropriate.
+      GC_malloc may return 0 if it is unable to acquire sufficient
+      space from the operating system.  This is the most probable
+      consequence of running out of space.  Other possible consequences
+      are that a function call will fail due to lack of stack space,
+      or that the collector will fail in other ways because it cannot
+      maintain its internal data structures, or that a crucial system
+      process will fail and take down the machine.  Most of these
+      possibilities are independent of the malloc implementation.
+
+2)  GC_malloc_atomic(nbytes)
+    - allocate an object of size nbytes that is guaranteed not to contain any
+      pointers.  The returned object is not guaranteed to be cleeared.
+      (Can always be replaced by GC_malloc, but results in faster collection
+      times.  The collector will probably run faster if large character
+      arrays, etc. are allocated with GC_malloc_atomic than if they are
+      statically allocated.)
+
+3)  GC_realloc(object, new_size)
+    - change the size of object to be new_size.  Returns a pointer to the
+      new object, which may, or may not, be the same as the pointer to
+      the old object.  The new object is taken to be atomic iff the old one
+      was.  If the new object is composite and larger than the original object,
+      then the newly added bytes are cleared (we hope).  This is very likely
+      to allocate a new object, unless MERGE_SIZES is defined in gc_private.h.
+      Even then, it is likely to recycle the old object only if the object
+      is grown in small additive increments (which, we claim, is generally bad
+      coding practice.)
+
+4)  GC_free(object)
+    - explicitly deallocate an object returned by GC_malloc or
+      GC_malloc_atomic.  Not necessary, but can be used to minimize
+      collections if performance is critical.
+
+5)  GC_expand_hp(number_of_4K_blocks)
+    - Explicitly increase the heap size.  (This is normally done automatically
+      if a garbage collection failed to GC_reclaim enough memory.  Explicit
+      calls to GC_expand_hp may prevent unnecessarily frequent collections at
+      program startup.)
+      
+6)  GC_clear_roots()
+    - Reset the collectors idea of where static variables containing pointers
+      may be located to the empty set of locations.  No statically allocated
+      variables will be traced from after this call, unless there are
+      intervening GC_add_roots calls.  The collector will still trace from
+      registers and the program stack.
+	  
+7)  GC_add_roots(low_address, high_address_plus_1)
+    - Add [low_address, high_address) as an area that may contain root pointers
+      and should be traced by the collector.  The static data and bss segments
+      are considered by default, and should not be added unless GC_clear_roots
+      has been called.  The number of root areas is currently limited to 50.
+      This is intended as a way to register data areas for dynamic libraries,
+      or to replace the entire data ans bss segments by smaller areas that are
+      known to contain all the roots. 
+
+8) Several routines to allow for registration of finalization code.
+   User supplied finalization code may be invoked when an object becomes
+   unreachable.  To call (*f)(obj, x) when obj becomes inaccessible, use
+	GC_register_finalizer(obj, f, x, 0, 0);
+   For more sophisticated uses, and for finalization ordering issues,
+   see gc.h.
+
+  The global variable GC_free_space_divisor may be adjusted up from its
+default value of 4 to use less space and more collection time, or down for
+the opposite effect.  Setting it to 1 or 0 will effectively disable collections
+and cause all allocations to simply grow the heap.
+
+  The variable GC_non_gc_bytes, which is normally 0, may be changed to reflect
+the amount of memory allocated by the above routines that should not be
+considered as a candidate for collection.  Careless use may, of course, result
+in excessive memory consumption.
+
+  Some additional tuning is possible through the parameters defined
+near the top of gc_private.h.
+  
+  If only GC_malloc is intended to be used, it might be appropriate to define:
+
+#define malloc(n) GC_malloc(n)
+#define calloc(m,n) GC_malloc((m)*(n))
+
+  For small pieces of VERY allocation intensive code, gc_inl.h
+includes some allocation macros that may be used in place of GC_malloc
+and friends.
+
+  All externally visible names in the garbage collector start with "GC_".
+To avoid name conflicts, client code should avoid this prefix, except when
+accessing garbage collector routines or variables.
+
+  Thre are provisions for allocation with explicit type information.
+This is rarely necessary.  Details can be found in gc_typed.h.
+
+
+USE AS LEAK DETECTOR:
+
+  The collector may be used to track down leaks in C programs that are
+intended to run with malloc/free (e.g. code with extreme real-time or
+portability constraints).  To do so define FIND_LEAK somewhere in
+gc_priv.h.  This will cause the collector to invoke the report_leak
+routine defined near the top of reclaim.c whenever an inaccessible
+object is found that has not been explicitly freed.
+  Productive use of this facility normally involves redefining report_leak
+to do something more intelligent.  This typically requires annotating
+objects with additional information (e.g. creation time stack trace) that
+identifies their origin.  Such code is typically not very portable, and is
+not included here, except on SPARC machines.
+  If all objects are allocated with GC_DEBUG_MALLOC (see next section),
+then the default version of report_leak will report the source file
+and line number at which the leaked object was allocated.  This may
+sometimes be sufficient.  (On SPARC/SUNOS4 machines, it will also report
+a cryptic stack trace.  This can often be turned into a sympolic stack
+trace by invoking program "foo" with "callprocs foo".  Callprocs is
+a short shell script that invokes adb to expand program counter values
+to symbolic addresses.  It was largely supplied by Scott Schwartz.)
+  Note that the debugging facilities described in the next section can
+sometimes be slightly LESS effective in leak finding mode, since in
+leak finding mode, GC_debug_free actually results in reuse of the object.
+(Otherwise the object is simply marked invalid.)
+
+DEBUGGING FACILITIES:
+
+  The routines GC_debug_malloc, GC_debug_malloc_atomic, GC_debug_realloc,
+and GC_debug_free provide an alternate interface to the collector, which
+provides some help with memory overwrite errors, and the like.
+Objects allocated in this way are annotated with additional
+information.  Some of this information is checked during garbage
+collections, and detected inconsistencies are reported to stderr.
+
+  Simple cases of writing past the end of an allocated object should
+be caught if the object is explicitly deallocated, or if the
+collector is invoked while the object is live.  The first deallocation
+of an object will clear the debugging info associated with an
+object, so accidentally repeated calls to GC_debug_free will report the
+deallocation of an object without debugging information.  Out of
+memory errors will be reported to stderr, in addition to returning
+NIL.
+
+  GC_debug_malloc checking  during garbage collection is enabled
+with the first call to GC_debug_malloc.  This will result in some
+slowdown during collections.  If frequent heap checks are desired,
+this can be acheived by explicitly invoking GC_gcollect, e.g. from
+the debugger.
+
+  GC_debug_malloc allocated objects should not be passed to GC_realloc
+or GC_free, and conversely.  It is however acceptable to allocate only
+some objects with GC_debug_malloc, and to use GC_malloc for other objects,
+provided the two pools are kept distinct.  In this case, there is a very
+low probablility that GC_malloc allocated objects may be misidentified as
+having been overwritten.  This should happen with probability at most
+one in 2**32.  This probability is zero if GC_debug_malloc is never called.
+
+  GC_debug_malloc, GC_malloc_atomic, and GC_debug_realloc take two
+additional trailing arguments, a string and an integer.  These are not
+interpreted by the allocator.  They are stored in the object (the string is
+not copied).  If an error involving the object is detected, they are printed.
+
+  The macros GC_MALLOC, GC_MALLOC_ATOMIC, GC_REALLOC, GC_FREE, and
+GC_REGISTER_FINALIZER are also provided.  These require the same arguments
+as the corresponding (nondebugging) routines.  If gc.h is included
+with GC_DEBUG defined, they call the debugging versions of these
+functions, passing the current file name and line number as the two
+extra arguments, where appropriate.  If gc.h is included without GC_DEBUG
+defined, then all these macros will instead be defined to their nondebugging
+equivalents.  (GC_REGISTER_FINALIZER is necessary, since pointers to
+objects with debugging information are really pointers to a displacement
+of 16 bytes form the object beginning, and some translation is necessary
+when finalization routines are invoked.  For details, about what's stored
+in the header, see the definition of the type oh in debug_malloc.c)
+
+INCREMENTAL/GENERATIONAL COLLECTION:
+
+The collector normally interrupts client code for the duration of 
+a garbage collection mark phase.  This may be unacceptable if interactive
+response is needed for programs with large heaps.  The collector
+can also run in a "generational" mode, in which it usually attempts to
+collect only objects allocated since the last garbage collection.
+Furthermore, in this mode, garbage collections run mostly incrementally,
+with a small amount of work performed in response to each of a large number of
+GC_malloc requests.
+
+This mode is enabled by a call to GC_enable_incremental().
+
+Incremental and generational collection is effective in reducing
+pause times only if the collector has some way to tell which objects
+or pages have been recently modified.  The collector uses two sources
+of information:
+
+1. Information provided by the VM system.  This may be provided in
+one of several forms.  Under Solaris 2.X (and potentially under other
+similar systems) information on dirty pages can be read from the
+/proc file system.  Under other systems (currently SunOS4.X) it is
+possible to write-protect the heap, and catch the resulting faults.
+On these systems we require that system calls writing to the heap
+(other than read) be handled specially by client code.
+See os_dep.c for details.
+
+2. Information supplied by the programmer.  We define "stubborn"
+objects to be objects that are rarely changed.  Such an object
+can be allocated (and enabled for writing) with GC_malloc_stubborn.
+Once it has been initialized, the collector should be informed with
+a call to GC_end_stubborn_change.  Subsequent writes that store
+pointers into the object must be preceded by a call to
+GC_change_stubborn.
+
+This mechanism performs best for objects that are written only for
+initialization, and such that only one stubborn object is writable
+at once.  It is typically not worth using for short-lived
+objects.  Stubborn objects are treated less efficiently than pointerfree
+(atomic) objects.
+
+A rough rule of thumb is that, in the absence of VM information, garbage
+collection pauses are proportional to the amount of pointerful storage
+plus the amount of modified "stubborn" storage that is reachable during
+the collection.  
+
+Initial allocation of stubborn objects takes longer than allocation
+of other objects, since other data structures need to be maintained.
+
+We recommend against random use of stubborn objects in client
+code, since bugs caused by inappropriate writes to stubborn objects
+are likely to be very infrequently observed and hard to trace.  
+However, their use may be appropriate in a few carefully written
+library routines that do not make the objects themselves available
+for writing by client code.
+
+
+BUGS:
+
+  Any memory that does not have a recognizable pointer to it will be
+reclaimed.  Exclusive-or'ing forward and backward links in a list
+doesn't cut it.
+  Some C optimizers may lose the last undisguised pointer to a memory
+object as a consequence of clever optimizations.  This has almost
+never been observed in practice.  Send mail to boehm@parc.xerox.com
+for suggestions on how to fix your compiler.
+  This is not a real-time collector.  In the standard configuration,
+percentage of time required for collection should be constant across
+heap sizes.  But collection pauses will increase for larger heaps.
+(On SPARCstation 2s collection times will be on the order of 300 msecs
+per MB of accessible memory that needs to be scanned.  Your mileage
+may vary.)  The incremental/generational collection facility helps,
+but is portable only if "stubborn" allocation is used.
+  Please address bug reports to boehm@parc.xerox.com.  If you are
+contemplating a major addition, you might also send mail to ask whether
+it's already been done.
+
+RECENT VERSIONS:
+
+  Version 1.3 and immediately preceding versions contained spurious
+assembly language assignments to TMP_SP.  Only the assignment in the PC/RT
+code is necessary.  On other machines, with certain compiler options,
+the assignments can lead to an unsaved register being overwritten.
+Known to cause problems under SunOS 3.5 WITHOUT the -O option.  (With
+-O the compiler recognizes it as dead code.  It probably shouldn't,
+but that's another story.)
+
+  Version 1.4 and earlier versions used compile time determined values
+for the stack base.  This no longer works on Sun 3s, since Sun 3/80s use
+a different stack base.  We now use a straightforward heuristic on all
+machines on which it is known to work (incl. Sun 3s) and compile-time
+determined values for the rest.  There should really be library calls
+to determine such values.
+
+  Version 1.5 and earlier did not ensure 8 byte alignment for objects
+allocated on a sparc based machine.
+
+  Version 1.8 added ULTRIX support in gc_private.h.
+  
+  Version 1.9 fixed a major bug in gc_realloc.
+  
+  Version 2.0 introduced a consistent naming convention for collector
+routines and added support for registering dynamic library data segments
+in the standard mark_roots.c.  Most of the data structures were revamped.
+The treatment of interior pointers was completely changed.  Finalization
+was added.  Support for locking was added.  Object kinds were added.
+We added a black listing facility to avoid allocating at addresses known
+to occur as integers somewhere in the address space.  Much of this
+was accomplished by adapting ideas and code from the PCR collector.
+The test program was changed and expanded.
+
+  Version 2.1 was the first stable version since 1.9, and added support
+for PPCR.
+
+  Version 2.2 added debugging allocation, and fixed various bugs.  Among them:
+- GC_realloc could fail to extend the size of the object for certain large object sizes.
+- A blatant subscript range error in GC_printf, which unfortunately
+  wasn't excercised on machines with sufficient stack alignment constraints.
+- GC_register_displacement did the wrong thing if it was called after
+  any allocation had taken place.
+- The leak finding code would eventually break after 2048 byte
+  byte objects leaked.
+- interface.c didn't compile.
+- The heap size remained much too small for large stacks.
+- The stack clearing code behaved badly for large stacks, and perhaps
+  on HP/PA machines.
+
+  Version 2.3 added ALL_INTERIOR_POINTERS and fixed the following bugs:
+- Missing declaration of etext in the A/UX version.
+- Some PCR root-finding problems.
+- Blacklisting was not 100% effective, because the plausible future
+  heap bounds were being miscalculated.
+- GC_realloc didn't handle out-of-memory correctly.
+- GC_base could return a nonzero value for addresses inside free blocks.
+- test.c wasn't really thread safe, and could erroneously report failure
+  in a multithreaded environment.  (The locking primitives need to be
+  replaced for other threads packages.)
+- GC_CONS was thoroughly broken.
+- On a SPARC with dynamic linking, signals stayed diabled while the
+  client code was running.
+  (Thanks to Manuel Serrano at INRIA for reporting the last two.)
+  
+  Version 2.4 added GC_free_space_divisor as a tuning knob, added
+  support for OS/2 and linux, and fixed the following bugs:
+- On machines with unaligned pointers (e.g. Sun 3), every 128th word could
+  fail to be considered for marking.
+- Dynamic_load.c erroneously added 4 bytes to the length of the data and
+  bss sections of the dynamic library.  This could result in a bad memory
+  reference if the actual length was a multiple of a page.  (Observed on
+  Sun 3.  Can probably also happen on a Sun 4.)
+  (Thanks to Robert Brazile for pointing out that the Sun 3 version
+  was broken.  Dynamic library handling is still broken on Sun 3s
+  under 4.1.1U1, but apparently not 4.1.1.  If you have such a machine,
+  use -Bstatic.)
+  
+  Version 2.5 fixed the following bugs:
+- Removed an explicit call to exit(1)
+- Fixed calls to GC_printf and GC_err_printf, so the correct number of
+  arguments are always supplied.  The OS/2 C compiler gets confused if
+  the number of actuals and the number of formals differ.  (ANSI C
+  doesn't require this to work.  The ANSI sanctioned way of doing things
+  causes too many compatibility problems.)
+  
+  Version 3.0  added generational/incremental collection and stubborn
+  objects.
+
+  Version 3.1 added the following features:
+- A workaround for a SunOS 4.X SPARC C compiler
+  misfeature that caused problems when the collector was turned into
+  a dynamic library.  
+- A fix for a bug in GC_base that could result in a memory fault.
+- A fix for a performance bug (and several other misfeatures) pointed
+  out by Dave Detelfs and Al Dosser.
+- Use of dirty bit information for static data under Solaris 2.X.
+- DEC Alpha/OSF1 support (thanks to Al Dosser).
+- Incremental collection on more platforms.
+- A more refined heap expansion policy.  Less space usage by default.
+- Various minor enhancements to reduce space usage, and to reduce
+  the amount of memory scanned by the collector.
+- Uncollectable allocation without per object overhead.
+- More conscientious handling of out-of-memory conditions.
+- Fixed a bug in debugging stubborn allocation.
+- Fixed a bug that resulted in occasional erroneous reporting of smashed
+  objects with debugging allocation.
+- Fixed bogus leak reports of size 4096 blocks with FIND_LEAK.
+
+  Version 3.2 fixed a serious and not entirely repeatable bug in
+  the incremental collector.  It appeared only when dirty bit info
+  on the roots was available, which is normally only under Solaris.
+  It also added GC_general_register_disappearing_link, and some
+  testing code.  Interface.c disappeared.
+
+  Version 3.3 fixes several bugs and adds new ports:
+- PCR-specific bugs.
+- Missing locking in GC_free, redundant FASTUNLOCK
+  in GC_malloc_stubborn, and 2 bugs in
+  GC_unregister_disappearing_link.
+  All of the above were pointed out by Neil Sharman
+  (neil@cs.mu.oz.au).
+- Common symbols allocated by the SunOS4.X dynamic loader
+  were not included in the root set.
+- Bug in GC_finalize (reported by Brian Beuning and Al Dosser)
+- Merged Amiga port from Jesper Peterson (untested)
+- Merged NeXT port from Thomas Funke (significantly
+  modified and untested)
+
+  Version 3.4:
+- Fixed a performance bug in GC_realloc.
+- Updated the amiga port.
+- Added NetBSD and 386BSD ports.
+- Added cord library.
+- Added trivial performance enhancement for
+  ALL_INTERIOR_POINTERS.  (Don't scan last word.)
+  
+  Version 3.5
+- Minor collections now mark from roots only once, if that
+  doesn't cause an excessive pause.
+- The stack clearing heuristic was refined to prevent anomalies
+  with very heavily recursive programs and sparse stacks.
+- Fixed a bug that prevented mark stack growth in some cases.
+  GC_objects_are_marked should be set to TRUE after a call
+  to GC_push_roots and as part of GC_push_marked, since
+  both can now set mark bits.  I think this is only a performance
+  bug, but I wouldn't bet on it.  It's certainly very hard to argue
+  that the old version was correct.
+- Fixed an incremental collection bug that prevented it from
+  working at all when HBLKSIZE != getpagesize()
+- Changed dynamic_loading.c to include gc_private.h before testing
+  DYNAMIC_LOADING.  SunOS dynamic library scanning
+  must have been broken in 3.4.
+- Object size rounding now adapts to program behavior.
+- Added a workaround (provided by Manuel Serrano and
+  colleagues) to a long-standing SunOS 4.X (and 3.X?) ld bug
+  that I had incorrectly assumed to have been squished.
+  The collector was broken if the text segment size was within
+  32 bytes of a multiple of 8K bytes, and if the beginning of
+  the data segment contained interesting roots.  The workaround
+  assumes a demand-loadable executable.  The original may have
+  have "worked" in some other cases.
+- Added dynamic library support under IRIX5.
+- Added support for EMX under OS/2 (thanks to Ari Huttunen).
+  
+Version 3.6:
+- fixed a bug in the mark stack growth code that was introduced
+  in 3.4.
+- fixed Makefile to work around DEC AXP compiler tail recursion
+  bug.
+
+Version 3.7:
+- Added a workaround for an HP/UX compiler bug.
+- Fixed another stack clearing performance bug.  Reworked
+  that code once more.
+  
+Version 4.0:
+- Added support for Solaris threads (which was possible
+  only be reimplementing some fraction of Solaris threads,
+  since Sun doesn't currently make the thread debugging
+  interface available).
+- Added non-threads win32 and win32S support.
+- (Grudgingly, with suitable muttering of obscenities) renamed
+  files so that the collector distribution could live on a FAT
+  file system.  Files that are guaranteed to be useless on
+  a PC still have long names.  Gc_inline.h and gc_private.h
+  still exist, but now just include  gc_inl.h and gc_priv.h.
+- Fixed a really obscure bug in finalization that could cause
+  undetected mark stack overflows.  (I would be surprised if
+  any real code ever tickled this one.)
+- Changed finalization code to dynamically resize the hash
+  tables it maintains.  (This probably does not matter for well-
+  -written code.  It no doubt does for C++ code that overuses
+  destructors.)
+- Added typed allocation primitves.  Rewrote the marker to
+  accommodate them with more reasonable efficiency.  This
+  change should also speed up marking for GC_malloc allocated
+  objects a little.  See gc_typed.h for new primitives.
+- Improved debugging facilities slightly.  Allocation time
+  stack traces are now kept by default on SPARC/SUNOS4.
+  (Thanks to Scott Schwartz.)
+- Added better support for small heap applications.
+- Significantly extended cord package.  Fixed a bug in the
+  implementation of lazily read files.  Printf and friends now
+  have cord variants.  Cord traversals are a bit faster.
+- Made ALL_INTERIOR_POINTERS recognition the default.
+- Fixed de so that it can run in constant space, independent
+  of file size.  Added simple string searching to cords and de.
+- Added the Hull-Ellis C++ interface.
+- Added dynamic library support for OSF/1.
+  (Thanks to Al Dosser and Tim Bingham at DEC.)
+- Changed argument to GC_expand_hp to be expressed
+  in units of bytes instead of heap blocks.  (Necessary
+  since the heap block size now varies depending on
+  configuration.  The old version was never very clean.)
+- Added GC_get_heap_size().  The previous "equivalent"
+  was broken.
+- Restructured the Makefile a bit.  
+
+Since version 4.0:
+- Changed finalization implementation to guarantee that
+  finalization procedures are called outside of the allocation
+  lock, making direct use of the interface a little less dangerous.
+  MAY BREAK EXISTING CLIENTS that assume finalizers
+  are protected by a lock.  Since there seem to be few multithreaded
+  clients that use finalization, this is hopefully not much of
+  a problem.
+- Fixed a gross bug in CORD_prev.
+- Fixed a bug in blacklst.c that could result in unbounded
+  heap growth during startup on machines that do not clear
+  memory obtained from the OS (e.g. win32S).
+- Ported de editor to win32/win32S.  (This is now the only
+  version with a mouse-sensitive UI.)
+- Added GC_malloc_ignore_off_page to allocate large arrays
+  in the presence of ALL_INTERIOR_POINTERS.
+- Changed GC_call_with_alloc_lock to not disable signals in
+  the single-threaded case.
+- Reduced retry count in GC_collect_or_expand for garbage
+  collecting when out of memory.
+- Made uncollectable allocations bypass black-listing, as they
+  should.
+- Fixed a bug in typed_test in test.c that could cause (legitimate)
+  GC crashes.
+- Fixed some potential synchronization problems in finalize.c
+- Fixed a real locking problem in typd_mlc.c.
+- Worked around an AIX 3.2 compiler feature that results in
+  out of bounds memory references.
+- Partially worked around an IRIX5.2 beta problem (which may
+  or may not persist to the final release).
+- Fixed a bug in the heap integrity checking code that could
+  result in explicitly deallocated objects being identified as
+  smashed.  Fixed a bug in the dbg_mlc stack saving code
+  that caused old argument pointers to be considered live.
+- Fixed a bug in CORD_ncmp (and hence CORD_str).
+- Repaired the OS2 port, which had suffered from bit rot
+  in 4.0.  Worked around what appears to be CSet/2 V1.0
+  optimizer bug.
+- Fixed a Makefile bug for target "c++".
diff --git a/README.OS2 b/README.OS2
new file mode 100644
index 00000000..5345bbd0
--- /dev/null
+++ b/README.OS2
@@ -0,0 +1,6 @@
+The code assumes static linking, and a single thread.  The editor de has
+not been ported.  The cord test program has.  The supplied OS2_MAKEFILE
+assumes the IBM C Set/2 environment, but the code shouldn't.
+
+Since we haven't figured out hoe to do perform partial links or to build static
+libraries, clients currently need to link against a long list of executables.
diff --git a/README.QUICK b/README.QUICK
new file mode 100644
index 00000000..98947660
--- /dev/null
+++ b/README.QUICK
@@ -0,0 +1,39 @@
+Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+
+THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+
+Permission is hereby granted to use or copy this program
+for any purpose,  provided the above notices are retained on all copies.
+Permission to modify the code and to distribute modified code is granted,
+provided the above notices are retained, and a notice that the code was
+modified is included with the above copyright notice.
+
+
+For more details and the names of other contributors, see the
+README file and gc.h.  This file describes typical use of
+the collector on a machine that is already supported.
+
+INSTALLATION:
+Under UN*X, type "make test".  Under OS/2 or Windows NT, copy the
+appropriate makefile to MAKEFILE, read it, and type "nmake test".
+Read the machine specific README if one exists.  The only way to
+develop code with the collector for Windows 3.1 is to develop under
+Windows NT, and then to use win32S.
+
+If you wish to use the cord (structured string) library type
+"make cords". (This requires an ANSI C compiler.  You may need
+to redefine CC in the Makefile.)
+
+If you wish to use the collector from C++, type
+"make c++".  These add further files to gc.a and to the include
+subdirectory.  See cord/cord.h and gc_c++.h.
+
+TYPICAL USE:
+Include "gc.h" from this directory.  Link against the appropriate library
+("gc.a" under UN*X).  Replace calls to malloc by calls to GC_MALLOC,
+and calls to realloc by calls to GC_REALLOC.  If the object is known
+to never contain pointers, use GC_MALLOC_ATOMIC instead of
+GC_MALLOC.
+
diff --git a/README.amiga b/README.amiga
new file mode 100644
index 00000000..cfb1fe81
--- /dev/null
+++ b/README.amiga
@@ -0,0 +1,83 @@
+
+ADDITIONAL NOTES FOR AMIGA PORT
+
+These notes assume some familiarity with Amiga internals.
+
+WHY I PORTED TO THE AMIGA
+
+The sole reason why I made this port was as a first step in getting
+the Sather(*) language on the Amiga. A port of this language will
+be done as soon as the Sather 1.0 sources are made available to me.
+Given this motivation, the garbage collection (GC) port is rather
+minimal.
+
+(*) For information on Sather read the comp.lang.sather newsgroup.
+
+LIMITATIONS
+
+This port assumes that the startup code linked with target programs
+is that supplied with SAS/C versions 6.0 or later. This allows
+assumptions to be made about where to find the stack base pointer
+and data segments when programs are run from WorkBench, as opposed
+to running from the CLI. The compiler dependent code is all in the
+GC_get_stack_base() and GC_register_data_segments() functions, but
+may spread as I add Amiga specific features.
+
+Given that SAS/C was assumed, the port is set up to be built with
+"smake" using the "SMakefile". Compiler options in "SCoptions" can
+be set with "scopts" program. Both "smake" and "scopts" are part of
+the SAS/C commercial development system.
+
+In keeping with the porting philosophy outlined above, this port
+will not behave well with Amiga specific code. Especially not inter-
+process comms via messages, and setting up public structures like
+Intuition objects or anything else in the system lists. For the
+time being the use of this library is limited to single threaded
+ANSI/POSIX  compliant or near-complient code. (ie. Stick to stdio
+for now). Given this limitation there is currently no mechanism for
+allocating "CHIP" or "PUBLIC" memory under the garbage collector.
+I'll add this after giving it considerable thought. The major
+problem is the entire physical address space may have to me scanned,
+since there is no telling who we may have passed memory to.
+
+If you allocate your own stack in client code, you will have to
+assign the pointer plus stack size to GC_stackbottom.
+
+The initial stack size of the target program can be compiled in by
+setting the __stack symbol (see SAS documentaion). It can be over-
+ridden from the CLI by running the AmigaDOS "stack" program, or from
+the WorkBench by setting the stack size in the tool types window.
+
+SAS/C COMPILER OPTIONS (SCoptions)
+
+You may wish to check the "CPU" code option is appropriate for your
+intended target system.
+
+Under no circumstances set the "StackExtend" code option in either
+compiling the library or *ANY* client code.
+
+All benign compiler warnings have been suppressed. These mainly
+involve lack of prototypes in the code, and dead assignments
+detected by the optimizer.
+
+THE GOOD NEWS
+
+The library as it stands is compatible with the GigaMem commercial
+virtual memory software, and probably similar PD software.
+
+The performance of "gctest" on an Amiga 2630 (68030 @ 25Mhz)
+compares favourably with an HP9000 with similar architecture (a 325
+with a 68030 I think).
+
+-----------------------------------------------------------------------
+
+The Amiga port has been brought to you by:
+
+Jesper Peterson.
+
+jep@mtiame.mtia.oz.au		(preferred, but 1 week turnaround)
+jep@orca1.vic.design.telecom.au (that's orca<one>, 1 day turnaround)
+
+At least one of these addresses should be around for a while, even
+though I don't work for either of the companies involved.
+
diff --git a/README.win32 b/README.win32
new file mode 100644
index 00000000..1eb77668
--- /dev/null
+++ b/README.win32
@@ -0,0 +1,39 @@
+The collector currently does not handle multiple threads.  There
+is good reason to believe this is fixable.  (SRC M3 works with
+NT threads.)
+
+The collector has only been compiled under Windows NT, with the
+Microsoft tools.
+
+It runs under both win32s and win32, but with different semantics.
+Under win32, all writable pages outside of the heaps and stack are
+scanned for roots.  Thus the collector sees pointers in DLL data
+segments.  Under win32s, only the main data segment is scanned.
+Thus all accessible objects should be excessible from local variables
+or variables in the main data segment.  Alternatively, other data
+segments (e.g. in DLLs) may be registered with the collector by
+calling GC_init() and then GC_register_root_section(a), where
+a is the address of some variable inside the data segment.  (Duplicate
+registrations are ignored, but not terribly quickly.)
+
+(There are two reasons for this.  We didn't want to see many 16:16
+pointers.  And the VirtualQuery call has different semantics under
+the two systems.)
+
+The collector test program "gctest" is linked as a GUI application,
+but does not open any windows.  Its output appears in the file
+"gc.log".  It may be started from the file manager.  The hour glass
+cursor will appear as long as it's running.
+
+The cord test program has not been ported (but should port
+easily).  A toy editor (cord/de.exe) based on cords (heavyweight
+strings represented as trees) has been ported and is included.
+It runs fine under either win32 or win32S.  It serves as an example
+of a true Windows application, except that it was written by a
+nonexpert Windows programmer.  (There are some peculiarities
+in the way files are displayed.  The <cr> is displayed explicitly
+for standard DOS text files.  As in the UNIX version, control
+characters are displayed explicitly, but in this case as red text.
+This may be suboptimal for some tastes and/or sets of default
+window colors.)
+
diff --git a/SCoptions.amiga b/SCoptions.amiga
new file mode 100644
index 00000000..9207e13e
--- /dev/null
+++ b/SCoptions.amiga
@@ -0,0 +1,15 @@
+CPU=68030
+NOSTACKCHECK
+ERRORREXX
+OPTIMIZE
+MAPHUNK
+NOVERSION
+NOICONS
+OPTIMIZERTIME
+DEFINE SILENT
+IGNORE=105
+IGNORE=304
+IGNORE=154
+IGNORE=85
+IGNORE=100
+IGNORE=161
diff --git a/SMakefile.amiga b/SMakefile.amiga
new file mode 100644
index 00000000..0727f423
--- /dev/null
+++ b/SMakefile.amiga
@@ -0,0 +1,45 @@
+OBJS= alloc.o reclaim.o allochblk.o misc.o mach_dep.o os_dep.o mark_roots.o headers.o mark.o obj_map.o black_list.o finalize.o new_hblk.o real_malloc.o dynamic_load.o debug_malloc.o malloc.o stubborn.o checksums.o
+
+INC=  gc_private.h gc_headers.h gc.h config.h
+
+all: gctest setjmp_test
+
+alloc.o : alloc.c $(INC)
+reclaim.o : reclaim.c $(INC)
+allochblk.o : allochblk.c $(INC)
+misc.o : misc.c $(INC)
+os_dep.o : os_dep.c $(INC)
+mark_roots.o : mark_roots.c $(INC)
+headers.o : headers.c $(INC)
+mark.o : mark.c $(INC)
+obj_map.o : obj_map.c $(INC)
+black_list.o : black_list.c $(INC)
+finalize.o : finalize.c $(INC)
+new_hblk.o : new_hblk.c $(INC)
+real_malloc.o : real_malloc.c $(INC)
+dynamic_load.o : dynamic_load.c $(INC)
+debug_malloc.o : debug_malloc.c $(INC)
+malloc.o : malloc.c $(INC)
+stubborn.o : stubborn.c $(INC)
+checksums.o : checksums.c $(INC)
+test.o : test.c $(INC)
+
+mach_dep.o : mach_dep.c $(INC)
+	sc noopt mach_dep.c			# optimizer mangles reg save hack
+
+gc.lib: $(OBJS)
+	oml gc.lib r $(OBJS)
+
+clean:
+	delete gc.lib gctest setjmp_test \#?.o
+
+gctest: gc.lib test.o
+	slink LIB:c.o test.o to $@ lib gc.lib LIB:sc.lib LIB:scm.lib
+
+setjmp_test: setjmp_test.c gc.h
+	sc setjmp_test.c
+	slink LIB:c.o $@.o to $@ lib LIB:sc.lib
+
+test: setjmp_test gctest
+	setjmp_test
+	gctest
diff --git a/allchblk.c b/allchblk.c
new file mode 100644
index 00000000..b8b9f890
--- /dev/null
+++ b/allchblk.c
@@ -0,0 +1,359 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 1:55 pm PDT */
+
+#define DEBUG
+#undef DEBUG
+#include <stdio.h>
+#include "gc_priv.h"
+
+
+/**/
+/* allocate/free routines for heap blocks
+/* Note that everything called from outside the garbage collector
+/* should be prepared to abort at any point as the result of a signal.
+/**/
+
+/*
+ * Free heap blocks are kept on a list sorted by address.
+ * The hb_hdr.hbh_sz field of a free heap block contains the length
+ * (in bytes) of the entire block.
+ * Neighbors are coalesced.
+ */
+ 
+# define MAX_BLACK_LIST_ALLOC (2*HBLKSIZE)
+		/* largest block we will allocate starting on a black   */
+		/* listed block.  Must be >= HBLKSIZE.			*/
+
+struct hblk * GC_hblkfreelist = 0;
+
+struct hblk *GC_savhbp = (struct hblk *)0;  /* heap block preceding next */
+					 /* block to be examined by   */
+					 /* GC_allochblk.                */
+
+void GC_print_hblkfreelist()
+{
+    struct hblk * h = GC_hblkfreelist;
+    word total_free = 0;
+    hdr * hhdr = HDR(h);
+    word sz;
+    
+    while (h != 0) {
+        sz = hhdr -> hb_sz;
+    	GC_printf2("0x%lx size %lu ", (unsigned long)h, (unsigned long)sz);
+    	total_free += sz;
+        if (GC_is_black_listed(h, HBLKSIZE) != 0) {
+             GC_printf0("start black listed\n");
+        } else if (GC_is_black_listed(h, hhdr -> hb_sz) != 0) {
+             GC_printf0("partially black listed\n");
+        } else {
+             GC_printf0("not black listed\n");
+        }
+        h = hhdr -> hb_next;
+        hhdr = HDR(h);
+    }
+    GC_printf1("Total of %lu bytes on free list\n", (unsigned long)total_free);
+}
+
+/* Initialize hdr for a block containing the indicated size and 	*/
+/* kind of objects.							*/
+/* Return FALSE on failure.						*/
+static bool setup_header(hhdr, sz, kind, flags)
+register hdr * hhdr;
+word sz;	/* object size in words */
+int kind;
+unsigned char flags;
+{
+    register word descr;
+    
+    /* Add description of valid object pointers */
+      if (!GC_add_map_entry(sz)) return(FALSE);
+      hhdr -> hb_map = GC_obj_map[sz > MAXOBJSZ? 0 : sz];
+      
+    /* Set size, kind and mark proc fields */
+      hhdr -> hb_sz = sz;
+      hhdr -> hb_obj_kind = kind;
+      hhdr -> hb_flags = flags;
+      descr = GC_obj_kinds[kind].ok_descriptor;
+      if (GC_obj_kinds[kind].ok_relocate_descr) descr += WORDS_TO_BYTES(sz);
+      hhdr -> hb_descr = descr;
+      
+    /* Clear mark bits */
+      GC_clear_hdr_marks(hhdr);
+      
+    hhdr -> hb_last_reclaimed = (unsigned short)GC_gc_no;
+    return(TRUE);
+}
+
+/*
+ * Allocate (and return pointer to) a heap block
+ *   for objects of size sz words.
+ *
+ * NOTE: We set obj_map field in header correctly.
+ *       Caller is resposnsible for building an object freelist in block.
+ *
+ * We clear the block if it is destined for large objects, and if
+ * kind requires that newly allocated objects be cleared.
+ */
+struct hblk *
+GC_allochblk(sz, kind, flags)
+word sz;
+int kind;
+unsigned char flags;
+{
+    register struct hblk *thishbp;
+    register hdr * thishdr;		/* Header corr. to thishbp */
+    register struct hblk *hbp;
+    register hdr * hhdr;		/* Header corr. to hbp */
+    struct hblk *prevhbp;
+    register hdr * phdr;		/* Header corr. to prevhbp */
+    signed_word size_needed;    /* number of bytes in requested objects */
+    signed_word size_avail;	/* bytes available in this block	*/
+    bool first_time = TRUE;
+
+    size_needed = HBLKSIZE * OBJ_SZ_TO_BLOCKS(sz);
+
+    /* search for a big enough block in free list */
+	hbp = GC_savhbp;
+	hhdr = HDR(hbp);
+	for(;;) {
+
+	    prevhbp = hbp;
+	    phdr = hhdr;
+	    hbp = (prevhbp == 0? GC_hblkfreelist : phdr->hb_next);
+	    hhdr = HDR(hbp);
+
+	    if( prevhbp == GC_savhbp && !first_time) {
+	        return(0);
+	    }
+
+	    first_time = FALSE;
+
+	    if( hbp == 0 ) continue;
+
+	    size_avail = hhdr->hb_sz;
+	    if (size_avail < size_needed) continue;
+	    /* If the next heap block is obviously better, go on.	*/
+	    /* This prevents us from disassembling a single large block */
+	    /* to get tiny blocks.					*/
+	    {
+	      signed_word next_size;
+	      
+	      thishbp = hhdr -> hb_next;
+	      if (thishbp == 0) thishbp = GC_hblkfreelist; 
+	      thishdr = HDR(thishbp);
+	      next_size = (signed_word)(thishdr -> hb_sz);
+	      if (next_size < size_avail
+	          && next_size >= size_needed
+	          && !GC_is_black_listed(thishbp, (word)size_needed)) {
+	          continue;
+	      }
+	    }
+	    if ( kind != UNCOLLECTABLE &&
+	         (kind != PTRFREE || size_needed > MAX_BLACK_LIST_ALLOC)) {
+	      struct hblk * lasthbp = hbp;
+	      ptr_t search_end = (ptr_t)hbp + size_avail - size_needed;
+	      signed_word eff_size_needed = ((flags & IGNORE_OFF_PAGE)?
+	      					HBLKSIZE
+	      					: size_needed);
+	      
+	      
+	      while ((ptr_t)lasthbp <= search_end
+	             && (thishbp = GC_is_black_listed(lasthbp,
+	             				      (word)eff_size_needed))) {
+	        lasthbp = thishbp;
+	      }
+	      size_avail -= (ptr_t)lasthbp - (ptr_t)hbp;
+	      thishbp = lasthbp;
+	      if (size_avail >= size_needed && thishbp != hbp
+	          && GC_install_header(thishbp)) {
+	          /* Split the block at thishbp */
+	              thishdr = HDR(thishbp);
+	              /* GC_invalidate_map not needed, since we will	*/
+	              /* allocate this block.				*/
+		      thishdr -> hb_next = hhdr -> hb_next;
+		      thishdr -> hb_sz = size_avail;
+		      hhdr -> hb_sz = (ptr_t)thishbp - (ptr_t)hbp;
+		      hhdr -> hb_next = thishbp;
+		  /* Advance to thishbp */
+		      prevhbp = hbp;
+		      phdr = hhdr;
+		      hbp = thishbp;
+		      hhdr = thishdr;
+	      } else if (size_avail == 0
+	      		 && size_needed == HBLKSIZE
+	      		 && prevhbp != 0) {
+#		ifndef FIND_LEAK
+	      	  static unsigned count = 0;
+	      	  
+	      	  /* The block is completely blacklisted.  We need 	*/
+	      	  /* to drop some such blocks, since otherwise we spend */
+	      	  /* all our time traversing them if pointerfree	*/
+	      	  /* blocks are unpopular.				*/
+	          /* A dropped block will be reconsidered at next GC.	*/
+	          if ((++count & 3) == 0) {
+	            /* Allocate and drop the block */
+	              if (GC_install_counts(hbp, hhdr->hb_sz)) {
+	                phdr -> hb_next = hhdr -> hb_next;
+	                (void) setup_header(
+	                	  hhdr,
+	              		  BYTES_TO_WORDS(hhdr->hb_sz - HDR_BYTES),
+	              		  PTRFREE, 0); /* Cant fail */
+	              	if (GC_debugging_started) {
+	              	    BZERO(hbp + HDR_BYTES, hhdr->hb_sz - HDR_BYTES);
+	              	}
+	                if (GC_savhbp == hbp) GC_savhbp = prevhbp;
+	              }
+	            /* Restore hbp to point at free block */
+	              hbp = prevhbp;
+	              hhdr = phdr;
+	              if (hbp == GC_savhbp) first_time = TRUE;
+	          }
+#		endif
+	      }
+	    }
+	    if( size_avail >= size_needed ) {
+		/* found a big enough block       */
+		/* let thishbp --> the block      */
+		/* set prevhbp, hbp to bracket it */
+		    thishbp = hbp;
+		    thishdr = hhdr;
+		    if( size_avail == size_needed ) {
+			hbp = hhdr->hb_next;
+			hhdr = HDR(hbp);
+		    } else {
+			hbp = (struct hblk *)
+			    (((word)thishbp) + size_needed);
+			if (!GC_install_header(hbp)) continue;
+			hhdr = HDR(hbp);
+			GC_invalidate_map(hhdr);
+			hhdr->hb_next = thishdr->hb_next;
+			hhdr->hb_sz = size_avail - size_needed;
+		    }
+		/* remove *thishbp from hblk freelist */
+		    if( prevhbp == 0 ) {
+			GC_hblkfreelist = hbp;
+		    } else {
+			phdr->hb_next = hbp;
+		    }
+		/* save current list search position */
+		    GC_savhbp = hbp;
+		break;
+	    }
+	}
+	
+    /* Notify virtual dirty bit implementation that we are about to write. */
+    	GC_write_hint(thishbp);
+    
+    /* Add it to map of valid blocks */
+    	if (!GC_install_counts(thishbp, (word)size_needed)) return(0);
+    	/* This leaks memory under very rare conditions. */
+    		
+    /* Set up header */
+        if (!setup_header(thishdr, sz, kind, flags)) {
+            GC_remove_counts(thishbp, (word)size_needed);
+            return(0); /* ditto */
+        }
+        
+    /* Clear block if necessary */
+	if (GC_debugging_started
+	    || sz > MAXOBJSZ && GC_obj_kinds[kind].ok_init) {
+	    BZERO(thishbp + HDR_BYTES,  size_needed - HDR_BYTES);
+	}
+    
+    return( thishbp );
+}
+ 
+struct hblk * GC_freehblk_ptr = 0;  /* Search position hint for GC_freehblk */
+
+/*
+ * Free a heap block.
+ *
+ * Coalesce the block with its neighbors if possible.
+ *
+ * All mark words are assumed to be cleared.
+ */
+void
+GC_freehblk(p)
+register struct hblk *p;
+{
+register hdr *phdr;	/* Header corresponding to p */
+register struct hblk *hbp, *prevhbp;
+register hdr *hhdr, *prevhdr;
+register signed_word size;
+
+    /* GC_savhbp may become invalid due to coalescing.  Clear it. */
+	GC_savhbp = (struct hblk *)0;
+
+    phdr = HDR(p);
+    size = phdr->hb_sz;
+    size = HBLKSIZE * OBJ_SZ_TO_BLOCKS(size);
+    GC_remove_counts(p, (word)size);
+    phdr->hb_sz = size;
+    GC_invalidate_map(phdr);
+    prevhbp = 0;
+    
+    /* The following optimization was suggested by David Detlefs.	*/
+    /* Note that the header cannot be NIL, since there cannot be an	*/
+    /* intervening  call to GC_freehblk without resetting		*/
+    /* GC_freehblk_ptr.							*/
+    if (GC_freehblk_ptr != 0 &&
+    	HDR(GC_freehblk_ptr)->hb_map == GC_invalid_map &&
+    	(ptr_t)GC_freehblk_ptr < (ptr_t)p) {
+      hbp = GC_freehblk_ptr;
+    } else {
+      hbp = GC_hblkfreelist;
+    };
+    hhdr = HDR(hbp);
+
+    while( (hbp != 0) && (hbp < p) ) {
+	prevhbp = hbp;
+	prevhdr = hhdr;
+	hbp = hhdr->hb_next;
+	hhdr = HDR(hbp);
+    }
+    GC_freehblk_ptr = prevhbp;
+    
+    /* Check for duplicate deallocation in the easy case */
+      if (hbp != 0 && (ptr_t)p + size > (ptr_t)hbp
+        || prevhbp != 0 && (ptr_t)prevhbp + prevhdr->hb_sz > (ptr_t)p) {
+        GC_printf1("Duplicate large block deallocation of 0x%lx\n",
+        	   (unsigned long) p);
+        GC_printf2("Surrounding free blocks are 0x%lx and 0x%lx\n",
+           	   (unsigned long) prevhbp, (unsigned long) hbp);
+      }
+
+    /* Coalesce with successor, if possible */
+      if( (((word)p)+size) == ((word)hbp) ) {
+	phdr->hb_next = hhdr->hb_next;
+	phdr->hb_sz += hhdr->hb_sz;
+	GC_remove_header(hbp);
+      } else {
+	phdr->hb_next = hbp;
+      }
+
+    
+    if( prevhbp == 0 ) {
+	GC_hblkfreelist = p;
+    } else if( (((word)prevhbp) + prevhdr->hb_sz)
+      	       == ((word)p) ) {
+      /* Coalesce with predecessor */
+	prevhdr->hb_next = phdr->hb_next;
+	prevhdr->hb_sz += phdr->hb_sz;
+	GC_remove_header(p);
+    } else {
+	prevhdr->hb_next = p;
+    }
+}
+
diff --git a/alloc.c b/alloc.c
new file mode 100644
index 00000000..33629ab6
--- /dev/null
+++ b/alloc.c
@@ -0,0 +1,634 @@
+/*
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ */
+/* Boehm, May 19, 1994 2:02 pm PDT */
+
+
+# include <stdio.h>
+# include <signal.h>
+# include <sys/types.h>
+# include "gc_priv.h"
+
+/*
+ * Separate free lists are maintained for different sized objects
+ * up to MAXOBJSZ.
+ * The call GC_allocobj(i,k) ensures that the freelist for
+ * kind k objects of size i points to a non-empty
+ * free list. It returns a pointer to the first entry on the free list.
+ * In a single-threaded world, GC_allocobj may be called to allocate
+ * an object of (small) size i as follows:
+ *
+ *            opp = &(GC_objfreelist[i]);
+ *            if (*opp == 0) GC_allocobj(i, NORMAL);
+ *            ptr = *opp;
+ *            *opp = obj_link(ptr);
+ *
+ * Note that this is very fast if the free list is non-empty; it should
+ * only involve the execution of 4 or 5 simple instructions.
+ * All composite objects on freelists are cleared, except for
+ * their first word.
+ */
+
+/*
+ *  The allocator uses GC_allochblk to allocate large chunks of objects.
+ * These chunks all start on addresses which are multiples of
+ * HBLKSZ.   Each allocated chunk has an associated header,
+ * which can be located quickly based on the address of the chunk.
+ * (See headers.c for details.) 
+ * This makes it possible to check quickly whether an
+ * arbitrary address corresponds to an object administered by the
+ * allocator.
+ */
+
+word GC_non_gc_bytes = 0;  /* Number of bytes not intended to be collected */
+
+word GC_gc_no = 0;
+
+int GC_incremental = 0;    /* By default, stop the world.	*/
+
+int GC_full_freq = 4;	   /* Every 5th collection is a full	*/
+			   /* collection.			*/
+
+char * GC_copyright[] =
+{"Copyright 1988,1989 Hans-J. Boehm and Alan J. Demers",
+"Copyright (c) 1991-1993 by Xerox Corporation.  All rights reserved.",
+"THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY",
+" EXPRESSED OR IMPLIED.  ANY USE IS AT YOUR OWN RISK."};
+
+
+/* some more variables */
+
+extern signed_word GC_mem_found;  /* Number of reclaimed longwords	*/
+				  /* after garbage collection      	*/
+
+bool GC_dont_expand = 0;
+
+word GC_free_space_divisor = 4;
+
+/* Return the minimum number of words that must be allocated between	*/
+/* collections to amortize the collection cost.				*/
+static word min_words_allocd()
+{
+    int dummy;
+#   ifdef THREADS
+ 	/* We punt, for now. */
+ 	register signed_word stack_size = 10000;
+#   else
+        register signed_word stack_size = (ptr_t)(&dummy) - GC_stackbottom;
+#   endif
+    register word total_root_size;  /* includes double stack size,	*/
+    				    /* since the stack is expensive	*/
+    				    /* to scan.				*/
+    
+    if (stack_size < 0) stack_size = -stack_size;
+    total_root_size = 2 * stack_size + GC_root_size;
+    if (GC_incremental) {
+        return(BYTES_TO_WORDS(GC_heapsize + total_root_size)
+               / (2 * GC_free_space_divisor));
+    } else {
+        return(BYTES_TO_WORDS(GC_heapsize + total_root_size)
+               / GC_free_space_divisor);
+    }
+}
+
+/* Return the number of words allocated, adjusted for explicit storage	*/
+/* management, etc..  This number is used in deciding when to trigger	*/
+/* collections.								*/
+word GC_adj_words_allocd()
+{
+    register signed_word result;
+    register signed_word expl_managed =
+    		BYTES_TO_WORDS((long)GC_non_gc_bytes
+    				- (long)GC_non_gc_bytes_at_gc);
+    
+    /* Don't count what was explicitly freed, or newly allocated for	*/
+    /* explicit management.  Note that deallocating an explicitly	*/
+    /* managed object should not alter result, assuming the client	*/
+    /* is playing by the rules.						*/
+    result = (signed_word)GC_words_allocd
+    	     - (signed_word)GC_mem_freed - expl_managed;
+    if (result > (signed_word)GC_words_allocd) result = GC_words_allocd;
+    	/* probably client bug or unfortunate scheduling */
+    result += GC_words_wasted;
+     	/* This doesn't reflect useful work.  But if there is lots of	*/
+     	/* new fragmentation, the same is probably true of the heap,	*/
+     	/* and the collection will be correspondingly cheaper.		*/
+    if (result < (signed_word)(GC_words_allocd >> 2)) {
+    	/* Always count at least 1/8 of the allocations.  We don't want	*/
+    	/* to collect too infrequently, since that would inhibit	*/
+    	/* coalescing of free storage blocks.				*/
+    	/* This also makes us partially robust against client bugs.	*/
+        return(GC_words_allocd >> 3);
+    } else {
+        return(result);
+    }
+}
+
+
+/* Clear up a few frames worth of garbage left at the top of the stack.	*/
+/* This is used to prevent us from accidentally treating garbade left	*/
+/* on the stack by other parts of the collector as roots.  This 	*/
+/* differs from the code in misc.c, which actually tries to keep the	*/
+/* stack clear of long-lived, client-generated garbage.			*/
+void GC_clear_a_few_frames()
+{
+#   define NWORDS 64
+    word frames[NWORDS];
+    register int i;
+    
+    for (i = 0; i < NWORDS; i++) frames[i] = 0;
+}
+
+/* Have we allocated enough to amortize a collection? */
+bool GC_should_collect()
+{
+    return(GC_adj_words_allocd() >= min_words_allocd());
+}
+
+/* 
+ * Initiate a garbage collection if appropriate.
+ * Choose judiciously
+ * between partial, full, and stop-world collections.
+ * Assumes lock held, signals disabled.
+ */
+void GC_maybe_gc()
+{
+    static int n_partial_gcs = 0;
+    if (GC_should_collect()) {
+        if (!GC_incremental) {
+            GC_gcollect_inner();
+            n_partial_gcs = 0;
+        } else if (n_partial_gcs >= GC_full_freq) {
+            GC_initiate_full();
+            n_partial_gcs = 0;
+        } else {
+            /* We try to mark with the world stopped.	*/
+            /* If we run out of time, this turns into	*/
+            /* incremental marking.			*/
+            if (GC_stopped_mark(FALSE)) GC_finish_collection();
+            n_partial_gcs++;
+        }
+    }
+}
+
+/*
+ * Stop the world garbage collection.  Assumes lock held, signals disabled.
+ */
+void GC_gcollect_inner()
+{
+#   ifdef PRINTSTATS
+	GC_printf2(
+	   "Initiating full world-stop collection %lu after %ld allocd bytes\n",
+	   (unsigned long) GC_gc_no+1,
+	   (long)WORDS_TO_BYTES(GC_words_allocd));
+#   endif
+    GC_promote_black_lists();
+    /* GC_reclaim_or_delete_all();  -- not needed: no intervening allocation */
+    GC_clear_marks();
+    (void) GC_stopped_mark(TRUE);
+    GC_finish_collection();
+}
+
+/*
+ * Perform n units of garbage collection work.  A unit is intended to touch
+ * roughly a GC_RATE pages.  Every once in a while, we do more than that.
+ */
+# define GC_RATE 8
+
+int GC_deficit = 0;	/* The number of extra calls to GC_mark_some	*/
+			/* that we have made.				*/
+			/* Negative values are equivalent to 0.		*/
+extern bool GC_collection_in_progress();
+
+void GC_collect_a_little(n)
+int n;
+{
+    register int i;
+    
+    if (GC_collection_in_progress()) {
+    	for (i = GC_deficit; i < GC_RATE*n; i++) {
+    	    if (GC_mark_some()) {
+    	        /* Need to finish a collection */
+    	        (void) GC_stopped_mark(TRUE);
+    	        GC_finish_collection();
+    	        break;
+    	    }
+    	}
+    	if (GC_deficit > 0) GC_deficit -= GC_RATE*n;
+    } else {
+        GC_maybe_gc();
+    }
+}
+
+/*
+ * Assumes lock is held, signals are disabled.
+ * We stop the world.
+ * If final is TRUE, then we finish the collection, no matter how long
+ * it takes.
+ * Otherwise we may fail and return FALSE if this takes too long.
+ * Increment GC_gc_no if we succeed.
+ */
+bool GC_stopped_mark(final)
+bool final;
+{
+    CLOCK_TYPE start_time;
+    CLOCK_TYPE current_time;
+    unsigned long time_diff;
+    register int i;
+	
+    GET_TIME(start_time);
+    STOP_WORLD();
+#   ifdef PRINTSTATS
+	GC_printf1("--> Marking for collection %lu ",
+	           (unsigned long) GC_gc_no + 1);
+	GC_printf2("after %lu allocd bytes + %lu wasted bytes\n",
+	   	   (unsigned long) WORDS_TO_BYTES(GC_words_allocd),
+	   	   (unsigned long) WORDS_TO_BYTES(GC_words_wasted));
+#   endif
+
+    /* Mark from all roots.  */
+        /* Minimize junk left in my registers and on the stack */
+            GC_clear_a_few_frames();
+            GC_noop(0,0,0,0,0,0);
+	GC_initiate_partial();
+	for(i = 0;;i++) {
+	    if (GC_mark_some()) break;
+	    if (final) continue;
+	    if ((i & 3) == 0) {
+	        GET_TIME(current_time);
+	        time_diff = MS_TIME_DIFF(current_time,start_time);
+	        if (time_diff >= TIME_LIMIT) {
+	            START_WORLD();
+#   		    ifdef PRINTSTATS
+		    	GC_printf0("Abandoning stopped marking after ");
+			GC_printf2("%lu iterations and %lu msecs\n",
+				   (unsigned long)i,
+			    	   (unsigned long)time_diff);
+#		    endif
+		    GC_deficit = i;  /* Give the mutator a chance. */
+	            return(FALSE);
+	        }
+	    }
+	}
+	
+    GC_gc_no++;
+#   ifdef PRINTSTATS
+      GC_printf2("Collection %lu reclaimed %ld bytes",
+		  (unsigned long) GC_gc_no - 1,
+	   	  (long)WORDS_TO_BYTES(GC_mem_found));
+      GC_printf1(" ---> heapsize = %lu bytes\n",
+      	        (unsigned long) GC_heapsize);
+      /* Printf arguments may be pushed in funny places.  Clear the	*/
+      /* space.								*/
+      GC_printf0("");
+#   endif      
+
+    /* Check all debugged objects for consistency */
+        if (GC_debugging_started) {
+            (*GC_check_heap)();
+        }
+    
+#   ifdef PRINTTIMES
+	GET_TIME(current_time);
+	GC_printf1("World-stopped marking took %lu msecs\n",
+	           MS_TIME_DIFF(current_time,start_time));
+#   endif
+    START_WORLD();
+    return(TRUE);
+}
+
+
+/* Finish up a collection.  Assumes lock is held, signals are disabled,	*/
+/* but the world is otherwise running.					*/
+void GC_finish_collection()
+{
+#   ifdef PRINTTIMES
+	CLOCK_TYPE start_time;
+	CLOCK_TYPE finalize_time;
+	CLOCK_TYPE done_time;
+	
+	GET_TIME(start_time);
+	finalize_time = start_time;
+#   endif
+
+#   ifdef GATHERSTATS
+        GC_mem_found = 0;
+#   endif
+#   ifdef FIND_LEAK
+      /* Mark all objects on the free list.  All objects should be */
+      /* marked when we're done.				   */
+	{
+	  register word size;		/* current object size		*/
+	  register ptr_t p;	/* pointer to current object	*/
+	  register struct hblk * h;	/* pointer to block containing *p */
+	  register hdr * hhdr;
+	  register int word_no;           /* "index" of *p in *q          */
+	  int kind;
+
+	  for (kind = 0; kind < GC_n_kinds; kind++) {
+	    for (size = 1; size <= MAXOBJSZ; size++) {
+	      for (p= GC_obj_kinds[kind].ok_freelist[size];
+	           p != 0; p=obj_link(p)){
+		h = HBLKPTR(p);
+		hhdr = HDR(h);
+		word_no = (((word *)p) - ((word *)h));
+		set_mark_bit_from_hdr(hhdr, word_no);
+	      }
+	    }
+	  }
+	}
+      /* Check that everything is marked */
+	GC_start_reclaim(TRUE);
+#   else
+
+      GC_finalize();
+#     ifdef STUBBORN_ALLOC
+        GC_clean_changing_list();
+#     endif
+
+#     ifdef PRINTTIMES
+	GET_TIME(finalize_time);
+#     endif
+
+      /* Clear free list mark bits, in case they got accidentally marked   */
+      /* Note: HBLKPTR(p) == pointer to head of block containing *p        */
+      /* Also subtract memory remaining from GC_mem_found count.           */
+      /* Note that composite objects on free list are cleared.             */
+      /* Thus accidentally marking a free list is not a problem;  only     */
+      /* objects on the list itself will be marked, and that's fixed here. */
+      {
+	register word size;		/* current object size		*/
+	register ptr_t p;	/* pointer to current object	*/
+	register struct hblk * h;	/* pointer to block containing *p */
+	register hdr * hhdr;
+	register int word_no;           /* "index" of *p in *q          */
+	int kind;
+
+	for (kind = 0; kind < GC_n_kinds; kind++) {
+	  for (size = 1; size <= MAXOBJSZ; size++) {
+	    for (p= GC_obj_kinds[kind].ok_freelist[size];
+	         p != 0; p=obj_link(p)){
+		h = HBLKPTR(p);
+		hhdr = HDR(h);
+		word_no = (((word *)p) - ((word *)h));
+		clear_mark_bit_from_hdr(hhdr, word_no);
+#		ifdef GATHERSTATS
+		    GC_mem_found -= size;
+#		endif
+	    }
+	  }
+	}
+      }
+
+
+#     ifdef PRINTSTATS
+	GC_printf1("Bytes recovered before sweep - f.l. count = %ld\n",
+	          (long)WORDS_TO_BYTES(GC_mem_found));
+#     endif
+
+    /* Reconstruct free lists to contain everything not marked */
+      GC_start_reclaim(FALSE);
+    
+#   endif /* !FIND_LEAK */
+
+#   ifdef PRINTSTATS
+	GC_printf2(
+		  "Immediately reclaimed %ld bytes in heap of size %lu bytes\n",
+	          (long)WORDS_TO_BYTES(GC_mem_found),
+	          (unsigned long)GC_heapsize);
+	GC_printf2("%lu (atomic) + %lu (composite) bytes in use\n",
+	           (unsigned long)WORDS_TO_BYTES(GC_atomic_in_use),
+	           (unsigned long)WORDS_TO_BYTES(GC_composite_in_use));
+#   endif
+
+    /* Reset or increment counters for next cycle */
+      GC_words_allocd_before_gc += GC_words_allocd;
+      GC_non_gc_bytes_at_gc = GC_non_gc_bytes;
+      GC_words_allocd = 0;
+      GC_words_wasted = 0;
+      GC_mem_freed = 0;
+      
+#   ifdef PRINTTIMES
+	GET_TIME(done_time);
+	GC_printf2("Finalize + initiate sweep took %lu + %lu msecs\n",
+	           MS_TIME_DIFF(finalize_time,start_time),
+	           MS_TIME_DIFF(done_time,finalize_time));
+#   endif
+}
+
+/* Externally callable routine to invoke full, stop-world collection */
+void GC_gcollect()
+{
+    DCL_LOCK_STATE;
+    
+    GC_invoke_finalizers();
+    DISABLE_SIGNALS();
+    LOCK();
+    if (!GC_is_initialized) GC_init_inner();
+    /* Minimize junk left in my registers */
+      GC_noop(0,0,0,0,0,0);
+    GC_gcollect_inner();
+    UNLOCK();
+    ENABLE_SIGNALS();
+    GC_invoke_finalizers();
+}
+
+word GC_n_heap_sects = 0;	/* Number of sections currently in heap. */
+
+/*
+ * Use the chunk of memory starting at p of syze bytes as part of the heap.
+ * Assumes p is HBLKSIZE aligned, and bytes is a multiple of HBLKSIZE.
+ */
+void GC_add_to_heap(p, bytes)
+struct hblk *p;
+word bytes;
+{
+    word words;
+    
+    if (GC_n_heap_sects >= MAX_HEAP_SECTS) {
+    	ABORT("Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS");
+    }
+    if (!GC_install_header(p)) {
+    	/* This is extremely unlikely. Can't add it.  This will		*/
+    	/* almost certainly result in a	0 return from the allocator,	*/
+    	/* which is entirely appropriate.				*/
+    	return;
+    }
+    GC_heap_sects[GC_n_heap_sects].hs_start = (ptr_t)p;
+    GC_heap_sects[GC_n_heap_sects].hs_bytes = bytes;
+    GC_n_heap_sects++;
+    words = BYTES_TO_WORDS(bytes - HDR_BYTES);
+    HDR(p) -> hb_sz = words;
+    GC_freehblk(p);
+    GC_heapsize += bytes;
+    if ((ptr_t)p <= GC_least_plausible_heap_addr
+        || GC_least_plausible_heap_addr == 0) {
+        GC_least_plausible_heap_addr = (ptr_t)p - sizeof(word);
+        	/* Making it a little smaller than necessary prevents	*/
+        	/* us from getting a false hit from the variable	*/
+        	/* itself.  There's some unintentional reflection	*/
+        	/* here.						*/
+    }
+    if ((ptr_t)p + bytes >= GC_greatest_plausible_heap_addr) {
+        GC_greatest_plausible_heap_addr = (ptr_t)p + bytes;
+    }
+}
+
+ptr_t GC_least_plausible_heap_addr = (ptr_t)ONES;
+ptr_t GC_greatest_plausible_heap_addr = 0;
+
+ptr_t GC_max(x,y)
+ptr_t x, y;
+{
+    return(x > y? x : y);
+}
+
+ptr_t GC_min(x,y)
+ptr_t x, y;
+{
+    return(x < y? x : y);
+}
+
+/*
+ * this explicitly increases the size of the heap.  It is used
+ * internally, but may also be invoked from GC_expand_hp by the user.
+ * The argument is in units of HBLKSIZE.
+ * Tiny values of n are rounded up.
+ * Returns FALSE on failure.
+ */
+bool GC_expand_hp_inner(n)
+word n;
+{
+    word bytes;
+    struct hblk * space;
+    word expansion_slop;	/* Number of bytes by which we expect the */
+    				/* heap to expand soon.			  */
+
+    if (n < MINHINCR) n = MINHINCR;
+    bytes = n * HBLKSIZE;
+    space = GET_MEM(bytes);
+    if( space == 0 ) {
+	return(FALSE);
+    }
+#   ifdef PRINTSTATS
+	GC_printf2("Increasing heap size by %lu after %lu allocated bytes\n",
+	           (unsigned long)bytes,
+	           (unsigned long)WORDS_TO_BYTES(GC_words_allocd));
+# 	ifdef UNDEFINED
+	  GC_printf1("Root size = %lu\n", GC_root_size);
+	  GC_print_block_list(); GC_print_hblkfreelist();
+	  GC_printf0("\n");
+#	endif
+#   endif
+    expansion_slop = 8 * WORDS_TO_BYTES(min_words_allocd());
+    if (5 * HBLKSIZE * MAXHINCR > expansion_slop) {
+        expansion_slop = 5 * HBLKSIZE * MAXHINCR;
+    }
+    if (GC_last_heap_addr == 0 && !((word)space & SIGNB)
+        || GC_last_heap_addr != 0 && GC_last_heap_addr < (ptr_t)space) {
+        /* Assume the heap is growing up */
+        GC_greatest_plausible_heap_addr =
+            GC_max(GC_greatest_plausible_heap_addr,
+                   (ptr_t)space + bytes + expansion_slop);
+    } else {
+        /* Heap is growing down */
+        GC_least_plausible_heap_addr =
+            GC_min(GC_least_plausible_heap_addr,
+                   (ptr_t)space - expansion_slop);
+    }
+    GC_prev_heap_addr = GC_last_heap_addr;
+    GC_last_heap_addr = (ptr_t)space;
+    GC_add_to_heap(space, bytes);
+    return(TRUE);
+}
+
+/* Really returns a bool, but it's externally visible, so that's clumsy. */
+/* Arguments is in bytes.						*/
+int GC_expand_hp(bytes)
+size_t bytes;
+{
+    int result;
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    if (!GC_is_initialized) GC_init_inner();
+    result = (int)GC_expand_hp_inner(divHBLKSZ((word)bytes));
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return(result);
+}
+
+bool GC_collect_or_expand(needed_blocks)
+word needed_blocks;
+{
+    static int count = 0;  /* How many failures? */
+    
+    if (!GC_incremental && !GC_dont_gc && GC_should_collect()) {
+      GC_gcollect_inner();
+    } else {
+      word blocks_to_get = GC_heapsize/(HBLKSIZE*GC_free_space_divisor)
+      			   + needed_blocks;
+      
+      if (blocks_to_get > MAXHINCR) {
+          if (needed_blocks > MAXHINCR) {
+              blocks_to_get = needed_blocks;
+          } else {
+              blocks_to_get = MAXHINCR;
+          }
+      }
+      if (!GC_expand_hp_inner(blocks_to_get)
+        && !GC_expand_hp_inner(needed_blocks)) {
+      	if (count++ < 5) {
+      	    WARN("Out of Memory!  Trying to continue ...\n");
+	    GC_gcollect_inner();
+	} else {
+	    WARN("Out of Memory!  Returning NIL!\n");
+	    return(FALSE);
+	}
+      }
+    }
+    return(TRUE);
+}
+
+/*
+ * Make sure the object free list for sz is not empty.
+ * Return a pointer to the first object on the free list.
+ * The object MUST BE REMOVED FROM THE FREE LIST BY THE CALLER.
+ * Assumes we hold the allocator lock and signals are disabled.
+ *
+ */
+ptr_t GC_allocobj(sz, kind)
+word sz;
+int kind;
+{
+    register ptr_t * flh = &(GC_obj_kinds[kind].ok_freelist[sz]);
+    
+    if (sz == 0) return(0);
+
+    while (*flh == 0) {
+      /* Do our share of marking work */
+        if(GC_incremental && !GC_dont_gc) GC_collect_a_little(1);
+      /* Sweep blocks for objects of this size */
+          GC_continue_reclaim(sz, kind);
+      if (*flh == 0) {
+        GC_new_hblk(sz, kind);
+      }
+      if (*flh == 0) {
+        if (!GC_collect_or_expand((word)1)) return(0);
+      }
+    }
+    
+    return(*flh);
+}
diff --git a/alpha_mach_dep.s b/alpha_mach_dep.s
new file mode 100644
index 00000000..265c3141
--- /dev/null
+++ b/alpha_mach_dep.s
@@ -0,0 +1,58 @@
+ # $Id: alpha_mach_dep.s,v 1.2 1993/01/18 22:54:51 dosser Exp $
+
+# define call_push(x)    lda   $16, 0(x);    jsr   $26, GC_push_one
+
+        .text
+        .align  4
+        .globl  GC_push_regs
+        .ent    GC_push_regs 2
+GC_push_regs:
+        ldgp    $gp, 0($27)
+        lda     $sp, -32($sp)
+        stq     $26, 8($sp)
+        .mask   0x04000000, -8
+        .frame  $sp, 16, $26, 0
+
+ #      call_push($0)    # expression eval and int func result
+
+ #      call_push($1)    # temp regs - not preserved cross calls
+ #      call_push($2)
+ #      call_push($3)
+ #      call_push($4)
+ #      call_push($5)
+ #      call_push($6)
+ #      call_push($7)
+ #      call_push($8)
+
+        call_push($9)    # Saved regs
+        call_push($10)
+        call_push($11)
+        call_push($12)
+        call_push($13)
+        call_push($14)
+
+        call_push($15)   # frame ptr or saved reg
+
+ #      call_push($16)   # argument regs - not preserved cross calls
+ #      call_push($17)
+ #      call_push($18)
+ #      call_push($19)
+ #      call_push($20)
+ #      call_push($21)
+
+ #      call_push($22)   # temp regs - not preserved cross calls
+ #      call_push($23)
+ #      call_push($24)
+ #      call_push($25)
+
+ #      call_push($26)   # return address - expression eval
+ #      call_push($27)   # procedure value or temporary reg
+ #      call_push($28)   # assembler temp - not presrved
+        call_push($29)   # Global Pointer
+ #      call_push($30)   # Stack Pointer
+
+        ldgp    $gp, 0($26)
+        ldq     $26, 8($sp)
+        lda     $sp, 32($sp)
+        ret     $31, ($26), 1
+        .end    GC_push_regs
diff --git a/barrett_diagram b/barrett_diagram
new file mode 100644
index 00000000..27e80dc1
--- /dev/null
+++ b/barrett_diagram
@@ -0,0 +1,106 @@
+This is an ASCII diagram of the data structure used to check pointer
+validity.  It was provided by Dave Barrett <barrett@asgard.cs.colorado.edu>,
+and should be of use to others attempting to understand the code.
+The data structure in GC4.X is essentially the same.   -HB
+
+
+
+
+		Data Structure used by GC_base in gc3.7:
+			      21-Apr-94
+			 
+			
+
+
+    63                  LOG_TOP_SZ[11]  LOG_BOTTOM_SZ[10]   LOG_HBLKSIZE[13]
+   +------------------+----------------+------------------+------------------+
+ p:|                  |   TL_HASH(hi)  |                  |   HBLKDISPL(p)   |
+   +------------------+----------------+------------------+------------------+
+    \-----------------------HBLKPTR(p)-------------------/
+    \------------hi-------------------/ 
+                      \______ ________/ \________ _______/ \________ _______/
+                             V                   V                  V
+                             |                   |                  |
+           GC_top_index[]    |                   |                  | 
+ ---      +--------------+   |                   |                  |  
+  ^       |              |   |                   |                  |   
+  |       |              |   |                   |                  |   
+ TOP      +--------------+<--+                   |                  |      
+ _SZ   +-<|      []      | *                     |                  |     
+(items)|  +--------------+  if 0 < bi< HBLKSIZE  |                  |    
+  |    |  |              | then large object     |                  |    
+  |    |  |              | starts at the bi'th   |                  |    
+  v    |  |              | HBLK before p.        |             i    |    
+ ---   |  +--------------+                       |          (word-  |    
+       v                                         |         aligned) |    
+   bi= |GET_BI(p){->hash_link}->key==hi          |                  |   
+       v                                         |                  |    
+       |   (bottom_index)  \ scratch_alloc'd     |                  |    
+       |   ( struct  bi )  / by get_index()      |                  |    
+ ---   +->+--------------+                       |                  |    
+  ^       |              |                       |                  |
+  ^       |              |                       |                  |
+ BOTTOM   |              |   ha=GET_HDR_ADDR(p)  |                  |
+_SZ(items)+--------------+<----------------------+          +-------+
+  |   +--<|   index[]    |                                  |         
+  |   |   +--------------+                      GC_obj_map: v              
+  |   |   |              |              from      / +-+-+-----+-+-+-+-+  --- 
+  v   |   |              |              GC_add   < 0| | |     | | | | |   ^  
+ ---  |   +--------------+             _map_entry \ +-+-+-----+-+-+-+-+   |  
+      |   |   asc_link   |                          +-+-+-----+-+-+-+-+ MAXOBJSZ
+      |   +--------------+                      +-->| | |  j  | | | | |  +1   
+      |   |     key      |                      |   +-+-+-----+-+-+-+-+   |  
+      |   +--------------+                      |   +-+-+-----+-+-+-+-+   | 
+      |   |  hash_link   |                      |   | | |     | | | | |   v 
+      |   +--------------+                      |   +-+-+-----+-+-+-+-+  ---
+      |                                         |   |<--MAX_OFFSET--->|   
+      |                                         |         (bytes)
+HDR(p)| GC_find_header(p)                       |   |<--MAP_ENTRIES-->| 
+      |                           \ from        |    =HBLKSIZE/WORDSZ   
+      |    (hdr) (struct hblkhdr) / alloc_hdr() |    (1024 on Alpha)
+      +-->+----------------------+              |    (8/16 bits each)
+GET_HDR(p)| word   hb_sz (words) |              |          
+          +----------------------+              |     
+          | struct hblk *hb_next |              |
+          +----------------------+              |       
+          |mark_proc hb_mark_proc|              |
+          +----------------------+              |
+          | char * hb_map        |>-------------+
+          +----------------------+           
+          | ushort hb_obj_kind   |           
+          +----------------------+           
+          |   hb_last_reclaimed  |           
+ ---      +----------------------+                
+  ^       |                      |
+ MARK_BITS|       hb_marks[]     | *if hdr is free, hb_sz + DISCARD_WORDS
+_SZ(words)|                      |  is the size of a heap chunk (struct hblk)
+  v       |                      |  of at least MININCR*HBLKSIZE bytes (below),
+ ---      +----------------------+  otherwise, size of each object in chunk.
+
+Dynamic data structures above are interleaved throughout the heap in blocks of 
+size MININCR * HBLKSIZE bytes as done by gc_scratch_alloc which cannot be
+freed; free lists are used (e.g. alloc_hdr).  HBLKs's below are collected.
+
+	      (struct hblk)      
+ ---      +----------------------+ < HBLKSIZE ---         ---          DISCARD_
+  ^       |garbage[DISCARD_WORDS]|   aligned   ^           ^ HDR_BYTES WORDS
+  |       |                      |             |           v (bytes)   (words)
+  |       +-----hb_body----------+ < WORDSZ    |          ---   ---   
+  |       |                      |   aligned   |           ^     ^
+  |       |      Object 0        |             |           hb_sz |
+  |       |                      |           i |(word-    (words)|
+  |       |                      |      (bytes)|aligned)   v     |
+  |       + - - - - - - - - - - -+ ---         |          ---    |
+  |       |                      |  ^          |           ^     |
+  n *     |                      |  j (words)  |          hb_sz BODY_SZ 
+ HBLKSIZE |      Object 1        |  v          v           |   (words)
+ (bytes)  |                      |---------------          v   MAX_OFFSET
+  |       + - - - - - - - - - - -+                        ---  (bytes)
+  |       |                      | !All_INTERIOR_PTRS      ^     |
+  |       |                      | sets j only for       hb_sz   |
+  |       |      Object N        | valid object offsets.   |     |
+  v       |                      | All objects WORDSZ      v     v
+ ---      +----------------------+ aligned.               ---   ---
+
+DISCARD_WORDS is normally zero.  Indeed the collector has not been tested
+with another value in ages.
diff --git a/blacklst.c b/blacklst.c
new file mode 100644
index 00000000..9c2fac8f
--- /dev/null
+++ b/blacklst.c
@@ -0,0 +1,181 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 1:56 pm PDT */
+# include "gc_priv.h"
+
+/*
+ * We maintain several hash tables of hblks that have had false hits.
+ * Each contains one bit per hash bucket;  If any page in the bucket
+ * has had a false hit, we assume that all of them have.
+ * See the definition of page_hash_table in gc_private.h.
+ * False hits from the stack(s) are much more dangerous than false hits
+ * from elsewhere, since the former can pin a large object that spans the
+ * block, eventhough it does not start on the dangerous block.
+ */
+ 
+/*
+ * Externally callable routines are:
+ 
+ * GC_add_to_black_list_normal
+ * GC_add_to_black_list_stack
+ * GC_promote_black_lists
+ * GC_is_black_listed
+ *
+ * All require that the allocator lock is held.
+ */
+
+/* Pointers to individual tables.  We replace one table by another by 	*/
+/* switching these pointers. 						*/
+word * GC_old_normal_bl;
+		/* Nonstack false references seen at last full		*/
+		/* collection.						*/
+word * GC_incomplete_normal_bl;
+		/* Nonstack false references seen since last		*/
+		/* full collection.					*/
+word * GC_old_stack_bl;
+word * GC_incomplete_stack_bl;
+
+void GC_clear_bl();
+
+void GC_bl_init()
+{
+# ifndef ALL_INTERIOR_POINTERS
+    GC_old_normal_bl = (word *)
+    			 GC_scratch_alloc((word)(sizeof (page_hash_table)));
+    GC_incomplete_normal_bl = (word *)GC_scratch_alloc
+    					((word)(sizeof(page_hash_table)));
+    if (GC_old_normal_bl == 0 || GC_incomplete_normal_bl == 0) {
+        GC_err_printf0("Insufficient memory for black list\n");
+        EXIT();
+    }
+    GC_clear_bl(GC_old_normal_bl);
+    GC_clear_bl(GC_incomplete_normal_bl);
+# endif
+    GC_old_stack_bl = (word *)GC_scratch_alloc((word)(sizeof(page_hash_table)));
+    GC_incomplete_stack_bl = (word *)GC_scratch_alloc
+    					((word)(sizeof(page_hash_table)));
+    if (GC_old_stack_bl == 0 || GC_incomplete_stack_bl == 0) {
+        GC_err_printf0("Insufficient memory for black list\n");
+        EXIT();
+    }
+    GC_clear_bl(GC_old_stack_bl);
+    GC_clear_bl(GC_incomplete_stack_bl);
+}
+		
+void GC_clear_bl(doomed)
+word *doomed;
+{
+    BZERO(doomed, sizeof(page_hash_table));
+}
+
+/* Signal the completion of a collection.  Turn the incomplete black	*/
+/* lists into new black lists, etc.					*/			 
+void GC_promote_black_lists()
+{
+    word * very_old_normal_bl = GC_old_normal_bl;
+    word * very_old_stack_bl = GC_old_stack_bl;
+    
+    GC_old_normal_bl = GC_incomplete_normal_bl;
+    GC_old_stack_bl = GC_incomplete_stack_bl;
+#   ifndef ALL_INTERIOR_POINTERS
+      GC_clear_bl(very_old_normal_bl);
+#   endif
+    GC_clear_bl(very_old_stack_bl);
+    GC_incomplete_normal_bl = very_old_normal_bl;
+    GC_incomplete_stack_bl = very_old_stack_bl;
+}
+
+# ifndef ALL_INTERIOR_POINTERS
+/* P is not a valid pointer reference, but it falls inside	*/
+/* the plausible heap bounds.					*/
+/* Add it to the normal incomplete black list if appropriate.	*/
+void GC_add_to_black_list_normal(p)
+word p;
+{
+    if (!(GC_modws_valid_offsets[p & (sizeof(word)-1)])) return;
+    {
+        register int index = PHT_HASH(p);
+        
+        if (HDR(p) == 0 || get_pht_entry_from_index(GC_old_normal_bl, index)) {
+#   	    ifdef PRINTBLACKLIST
+		if (!get_pht_entry_from_index(GC_incomplete_normal_bl, index)) {
+	    	  GC_printf1("Black listing (normal) 0x%lx\n",
+	    	  	     (unsigned long) p);
+	    	}
+#           endif
+            set_pht_entry_from_index(GC_incomplete_normal_bl, index);
+        } /* else this is probably just an interior pointer to an allocated */
+          /* object, and isn't worth black listing.			    */
+    }
+}
+# endif
+
+/* And the same for false pointers from the stack. */
+void GC_add_to_black_list_stack(p)
+word p;
+{
+    register int index = PHT_HASH(p);
+        
+    if (HDR(p) == 0 || get_pht_entry_from_index(GC_old_stack_bl, index)) {
+#   	ifdef PRINTBLACKLIST
+	    if (!get_pht_entry_from_index(GC_incomplete_stack_bl, index)) {
+	    	  GC_printf1("Black listing (stack) 0x%lx\n",
+	    	             (unsigned long)p);
+	    }
+#       endif
+	set_pht_entry_from_index(GC_incomplete_stack_bl, index);
+    }
+}
+
+/*
+ * Is the block starting at h of size len bytes black listed?   If so,
+ * return the address of the next plausible r such that (r, len) might not
+ * be black listed.  (R may not actually be in the heap.  We guarantee only
+ * that every smaller value of r after h is also black listed.)
+ * If (h,len) is not black listed, return 0.
+ * Knows about the structure of the black list hash tables.
+ */
+struct hblk * GC_is_black_listed(h, len)
+struct hblk * h;
+word len;
+{
+    register int index = PHT_HASH((word)h);
+    register word i;
+    word nblocks = divHBLKSZ(len);
+
+#   ifndef ALL_INTERIOR_POINTERS
+      if (get_pht_entry_from_index(GC_old_normal_bl, index)
+          || get_pht_entry_from_index(GC_incomplete_normal_bl, index)) {
+        return(h+1);
+      }
+#   endif
+    
+    for (i = 0; ; ) {
+        if (GC_old_stack_bl[divWORDSZ(index)] == 0
+            && GC_incomplete_stack_bl[divWORDSZ(index)] == 0) {
+            /* An easy case */
+            i += WORDSZ - modWORDSZ(index);
+        } else {
+          if (get_pht_entry_from_index(GC_old_stack_bl, index)
+              || get_pht_entry_from_index(GC_incomplete_stack_bl, index)) {
+            return(h+i+1);
+          }
+          i++;
+        }
+        if (i >= nblocks) break;
+        index = PHT_HASH((word)(h+i));
+    }
+    return(0);
+}
+
diff --git a/callprocs b/callprocs
new file mode 100755
index 00000000..4f105cc2
--- /dev/null
+++ b/callprocs
@@ -0,0 +1,3 @@
+#!/bin/sh
+GC_DEBUG=1
+$* 2>&1 | awk '{print "0x3e=c\""$0"\""};/^\t##PC##=/ {if ($2 != 0) {print $2"?i"}}' | adb $1 | sed "s/^		>/>/"
diff --git a/checksums.c b/checksums.c
new file mode 100644
index 00000000..2cc37e41
--- /dev/null
+++ b/checksums.c
@@ -0,0 +1,151 @@
+/*
+ * Copyright (c) 1992-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:07 pm PDT */
+# ifdef CHECKSUMS
+
+# include "gc_priv.h"
+
+/* This is debugging code intended to verify the results of dirty bit	*/
+/* computations. Works only in a single threaded environment.		*/
+/* We assume that stubborn objects are changed only when they are 	*/
+/* enabled for writing.  (Certain kinds of writing are actually		*/
+/* safe under other conditions.)					*/
+# define NSUMS 2000
+
+# define OFFSET 100000
+
+typedef struct {
+	bool new_valid;
+	word old_sum;
+	word new_sum;
+	struct hblk * block;	/* Block to which this refers + OFFSET  */
+				/* to hide it from colector.		*/
+} page_entry;
+
+page_entry GC_sums [NSUMS];
+
+word GC_checksum(h)
+struct hblk *h;
+{
+    register word *p = (word *)h;
+    register word *lim = (word *)(h+1);
+    register word result = 0;
+    
+    while (p < lim) {
+        result += *p++;
+    }
+    return(result);
+}
+
+# ifdef STUBBORN_ALLOC
+/* Check whether a stubborn object from the given block appears on	*/
+/* the appropriate free list.						*/
+bool GC_on_free_list(h)
+struct hblk *h;
+{
+    register hdr * hhdr = HDR(h);
+    register int sz = hhdr -> hb_sz;
+    ptr_t p;
+    
+    if (sz > MAXOBJSZ) return(FALSE);
+    for (p = GC_sobjfreelist[sz]; p != 0; p = obj_link(p)) {
+        if (HBLKPTR(p) == h) return(TRUE);
+    }
+    return(FALSE);
+}
+# endif
+ 
+int GC_n_dirty_errors;
+int GC_n_changed_errors;
+int GC_n_clean;
+int GC_n_dirty;
+
+void GC_update_check_page(h, index)
+struct hblk *h;
+int index;
+{
+    page_entry *pe = GC_sums + index;
+    register hdr * hhdr = HDR(h);
+    
+    if (pe -> block != 0 && pe -> block != h + OFFSET) ABORT("goofed");
+    pe -> old_sum = pe -> new_sum;
+    pe -> new_sum = GC_checksum(h);
+    if (GC_page_was_dirty(h)) {
+    	GC_n_dirty++;
+    } else {
+    	GC_n_clean++;
+    }
+    if (pe -> new_valid && pe -> old_sum != pe -> new_sum) {
+    	if (!GC_page_was_dirty(h)) {
+    	    /* Set breakpoint here */GC_n_dirty_errors++;
+    	}
+#	ifdef STUBBORN_ALLOC
+    	  if (!IS_FORWARDING_ADDR_OR_NIL(hhdr)
+    	    && hhdr -> hb_map != GC_invalid_map
+    	    && hhdr -> hb_obj_kind == STUBBORN
+    	    && !GC_page_was_changed(h)
+    	    && !GC_on_free_list(h)) {
+    	    /* if GC_on_free_list(h) then reclaim may have touched it	*/
+    	    /* without any allocations taking place.			*/
+    	    /* Set breakpoint here */GC_n_changed_errors++;
+    	  }
+#	endif
+    }
+    pe -> new_valid = TRUE;
+    pe -> block = h + OFFSET;
+}
+
+/* Should be called immediately after GC_read_dirty and GC_read_changed. */
+void GC_check_dirty()
+{
+    register int index;
+    register int i;
+    register struct hblk *h;
+    register ptr_t start;
+    
+    GC_n_dirty_errors = 0;
+    GC_n_changed_errors = 0;
+    GC_n_clean = 0;
+    GC_n_dirty = 0;
+    
+    index = 0;
+    for (i = 0; i < GC_n_heap_sects; i++) {
+    	start = GC_heap_sects[i].hs_start;
+        for (h = (struct hblk *)start;
+             h < (struct hblk *)(start + GC_heap_sects[i].hs_bytes);
+             h++) {
+             GC_update_check_page(h, index);
+             index++;
+             if (index >= NSUMS) goto out;
+        }
+    }
+out:
+    GC_printf2("Checked %lu clean and %lu dirty pages\n",
+    	      (unsigned long) GC_n_clean, (unsigned long) GC_n_dirty);
+    if (GC_n_dirty_errors > 0) {
+        GC_printf1("Found %lu dirty bit errors\n",
+        	   (unsigned long)GC_n_dirty_errors);
+    }
+    if (GC_n_changed_errors > 0) {
+    	GC_printf1("Found %lu changed bit errors\n",
+        	   (unsigned long)GC_n_changed_errors);
+    }
+}
+
+# else
+
+extern int GC_quiet;
+	/* ANSI C doesn't allow translation units to be empty.	*/
+	/* So we guarantee this one is nonempty.		*/
+
+# endif /* CHECKSUMS */
diff --git a/config.h b/config.h
new file mode 100644
index 00000000..4e096102
--- /dev/null
+++ b/config.h
@@ -0,0 +1,541 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:11 pm PDT */
+ 
+#ifndef CONFIG_H
+
+# define CONFIG_H
+
+/* Machine dependent parameters.  Some tuning parameters can be found	*/
+/* near the top of gc_private.h.					*/
+
+/* Machine specific parts contributed by various people.  See README file. */
+
+/* Determine the machine type: */
+# if defined(sun) && defined(mc68000)
+#    define M68K
+#    define SUNOS4
+#    define mach_type_known
+# endif
+# if defined(hp9000s300)
+#    define M68K
+#    define HP
+#    define mach_type_known
+# endif
+# if defined(vax)
+#    define VAX
+#    ifdef ultrix
+#	define ULTRIX
+#    else
+#	define BSD
+#    endif
+#    define mach_type_known
+# endif
+# if defined(mips)
+#    define MIPS
+#    ifdef ultrix
+#	define ULTRIX
+#    else
+#	ifdef _SYSTYPE_SVR4
+#	  define IRIX5
+#	else
+#	  define RISCOS  /* or IRIX 4.X */
+#	endif
+#    endif
+#    define mach_type_known
+# endif
+# if defined(sequent) && defined(i386)
+#    define I386
+#    define SEQUENT
+#    define mach_type_known
+# endif
+# if defined(sun) && defined(i386)
+#    define I386
+#    define SUNOS5
+#    define mach_type_known
+# endif
+# if defined(__OS2__) && defined(__32BIT__)
+#    define I386
+#    define OS2
+#    define mach_type_known
+# endif
+# if defined(ibm032)
+#   define RT
+#   define mach_type_known
+# endif
+# if defined(sun) && defined(sparc)
+#   define SPARC
+    /* Test for SunOS 5.x */
+#     include <errno.h>
+#     ifdef ECHRNG
+#       define SUNOS5
+#     else
+#	define SUNOS4
+#     endif
+#   define mach_type_known
+# endif
+# if defined(_IBMR2)
+#   define RS6000
+#   define mach_type_known
+# endif
+# if defined(SCO)
+#   define I386
+#   define SCO
+#   define mach_type_known
+/*	--> incompletely implemented */
+# endif
+# if defined(_AUX_SOURCE)
+#   define M68K
+#   define SYSV
+#   define mach_type_known
+# endif
+# if defined(_PA_RISC1_0) || defined(_PA_RISC1_1)
+#   define HP_PA
+#   define mach_type_known
+# endif
+# if defined(linux) && defined(i386)
+#    define I386
+#    define LINUX
+#    define mach_type_known
+# endif
+# if defined(__alpha)
+#   define ALPHA
+#   define mach_type_known
+# endif
+# if defined(_AMIGA)
+#   define AMIGA
+#   define M68K
+#   define mach_type_known
+# endif
+# if defined(NeXT) && defined(mc68000)
+#   define M68K
+#   define NEXT
+#   define mach_type_known
+# endif
+# if defined(__FreeBSD__) && defined(i386)
+#   define I386
+#   define FREEBSD
+#   define mach_type_known
+# endif
+# if defined(__NetBSD__) && defined(i386)
+#   define I386
+#   define NETBSD
+#   define mach_type_known
+# endif
+# if defined(bsdi) && defined(i386)
+#    define I386
+#    define BSDI
+#    define mach_type_known
+# endif
+# if !defined(mach_type_known) && defined(__386BSD__)
+#   define I386
+#   define THREE86BSD
+#   define mach_type_known
+# endif
+# if defined(_CX_UX) && defined(_M88K)
+#   define M88K
+#   define CX_UX
+#   define mach_type_known
+# endif
+# if defined(_MSDOS) && (_M_IX86 == 300) || (_M_IX86 == 400)
+#   define I386
+#   define MSWIN32	/* or Win32s */
+#   define mach_type_known
+# endif
+
+/* Feel free to add more clauses here */
+
+/* Or manually define the machine type here.  A machine type is 	*/
+/* characterized by the architecture.  Some				*/
+/* machine types are further subdivided by OS.				*/
+/* the macros ULTRIX, RISCOS, and BSD to distinguish.			*/
+/* Note that SGI IRIX is treated identically to RISCOS.			*/
+/* SYSV on an M68K actually means A/UX.					*/
+/* The distinction in these cases is usually the stack starting address */
+# ifndef mach_type_known
+	--> unknown machine type
+# endif
+		    /* Mapping is: M68K       ==> Motorola 680X0	*/
+		    /*		   (SUNOS4,HP,NEXT, and SYSV (A/UX),	*/
+		    /*		   and AMIGA variants)			*/
+		    /*             I386       ==> Intel 386	 	*/
+		    /*		    (SEQUENT, OS2, SCO, LINUX, NETBSD,	*/
+		    /*		     FREEBSD, THREE86BSD, MSWIN32,	*/
+		    /* 		     BSDI, SUNOS5 variants)		*/
+                    /*             NS32K      ==> Encore Multimax 	*/
+                    /*             MIPS       ==> R2000 or R3000	*/
+                    /*			(RISCOS, ULTRIX variants)	*/
+                    /*		   VAX	      ==> DEC VAX		*/
+                    /*			(BSD, ULTRIX variants)		*/
+                    /*		   RS6000     ==> IBM RS/6000 AIX3.1	*/
+                    /*		   RT	      ==> IBM PC/RT		*/
+                    /*		   HP_PA      ==> HP9000/700 & /800	*/
+                    /*				  HP/UX			*/
+		    /*		   SPARC      ==> SPARC under SunOS	*/
+		    /*			(SUNOS4, SUNOS5 variants)	*/
+		    /* 		   ALPHA      ==> DEC Alpha OSF/1	*/
+		    /* 		   M88K       ==> Motorola 88XX0        */
+		    /* 		        (CX/UX so far)			*/
+
+
+/*
+ * For each architecture and OS, the following need to be defined:
+ *
+ * CPP_WORD_SZ is a simple integer constant representing the word size.
+ * in bits.  We assume byte addressibility, where a byte has 8 bits.
+ * We also assume CPP_WORD_SZ is either 32 or 64.
+ * (We care about the length of pointers, not hardware
+ * bus widths.  Thus a 64 bit processor with a C compiler that uses
+ * 32 bit pointers should use CPP_WORD_SZ of 32, not 64. Default is 32.)
+ *
+ * MACH_TYPE is a string representation of the machine type.
+ * OS_TYPE is analogous for the OS.
+ *
+ * ALIGNMENT is the largest N, such that
+ * all pointer are guaranteed to be aligned on N byte boundaries.
+ * defining it to be 1 will always work, but perform poorly.
+ *
+ * DATASTART is the beginning of the data segment.
+ * On UNIX systems, the collector will scan the area between DATASTART
+ * and &end for root pointers.
+ *
+ * STACKBOTTOM is the cool end of the stack, which is usually the
+ * highest address in the stack.
+ * Under PCR or OS/2, we have other ways of finding thread stacks.
+ * For each machine, the following should:
+ * 1) define STACK_GROWS_UP if the stack grows toward higher addresses, and
+ * 2) define exactly one of
+ *	STACKBOTTOM (should be defined to be an expression)
+ *	HEURISTIC1
+ *	HEURISTIC2
+ * If either of the last two macros are defined, then STACKBOTTOM is computed
+ * during collector startup using one of the following two heuristics:
+ * HEURISTIC1:  Take an address inside GC_init's frame, and round it up to
+ *		the next multiple of 16 MB.
+ * HEURISTIC2:  Take an address inside GC_init's frame, increment it repeatedly
+ *		in small steps (decrement if STACK_GROWS_UP), and read the value
+ *		at each location.  Remember the value when the first
+ *		Segmentation violation or Bus error is signalled.  Round that
+ *		to the nearest plausible page boundary, and use that instead
+ *		of STACKBOTTOM.
+ *
+ * If no expression for STACKBOTTOM can be found, and neither of the above
+ * heuristics are usable, the collector can still be used with all of the above
+ * undefined, provided one of the following is done:
+ * 1) GC_mark_roots can be changed to somehow mark from the correct stack(s)
+ *    without reference to STACKBOTTOM.  This is appropriate for use in
+ *    conjunction with thread packages, since there will be multiple stacks.
+ *    (Allocating thread stacks in the heap, and treating them as ordinary
+ *    heap data objects is also possible as a last resort.  However, this is
+ *    likely to introduce significant amounts of excess storage retention
+ *    unless the dead parts of the thread stacks are periodically cleared.)
+ * 2) Client code may set GC_stackbottom before calling any GC_ routines.
+ *    If the author of the client code controls the main program, this is
+ *    easily accomplished by introducing a new main program, setting
+ *    GC_stackbottom to the address of a local variable, and then calling
+ *    the original main program.  The new main program would read something
+ *    like:
+ *
+ *		# include "gc_private.h"
+ *
+ *		main(argc, argv, envp)
+ *		int argc;
+ *		char **argv, **envp;
+ *		{
+ *		    int dummy;
+ *
+ *		    GC_stackbottom = (ptr_t)(&dummy);
+ *		    return(real_main(argc, argv, envp));
+ *		}
+ *
+ *
+ * Each architecture may also define the style of virtual dirty bit
+ * implementation to be used:
+ *   MPROTECT_VDB: Write protect the heap and catch faults.
+ *   PROC_VDB: Use the SVR4 /proc primitives to read dirty bits.
+ *
+ * An architecture may define DYNAMIC_LOADING if dynamic_load.c
+ * defined GC_register_dynamic_libraries() for the architecture.
+ */
+
+
+# ifdef M68K
+#   define MACH_TYPE "M68K"
+#   define ALIGNMENT 2
+#   ifdef SUNOS4
+#	define OS_TYPE "SUNOS4"
+	extern char etext;
+#	define DATASTART ((ptr_t)((((word) (&etext)) + 0x1ffff) & ~0x1ffff))
+#	define HEURISTIC1	/* differs	*/
+#	define DYNAMIC_LOADING
+#   endif
+#   ifdef HP
+#	define OS_TYPE "HP"
+	extern char etext;
+#       define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
+#       define STACKBOTTOM ((ptr_t) 0xffeffffc)
+			      /* empirically determined.  seems to work. */
+#   endif
+#   ifdef SYSV
+#	define OS_TYPE "SYSV"
+	extern etext;
+#   	define DATASTART ((ptr_t)((((word) (&etext)) + 0x3fffff) \
+				   & ~0x3fffff) \
+				  +((word)&etext & 0x1fff))
+	/* This only works for shared-text binaries with magic number 0413.
+	   The other sorts of SysV binaries put the data at the end of the text,
+	   in which case the default of &etext would work.  Unfortunately,
+	   handling both would require having the magic-number available.
+	   	   		-- Parag
+	   */
+#	define STACKBOTTOM ((ptr_t)0xFFFFFFFE)
+			/* The stack starts at the top of memory, but   */
+			/* 0x0 cannot be used as setjump_test complains */
+			/* that the stack direction is incorrect.  Two  */
+			/* bytes down from 0x0 should be safe enough.   */
+			/* 		--Parag				*/
+#   endif
+#   ifdef AMIGA
+#	define OS_TYPE "AMIGA"
+ 	    	/* STACKBOTTOM and DATASTART handled specially	*/
+ 	    	/* in os_dep.c					*/
+#   endif
+#   ifdef NEXT
+#	define OS_TYPE "NEXT"
+#	define DATASTART ((ptr_t) get_etext())
+#	define STACKBOTTOM ((ptr_t) 0x4000000)
+#   endif
+# endif
+
+# ifdef VAX
+#   define MACH_TYPE "VAX"
+#   define ALIGNMENT 4	/* Pointers are longword aligned by 4.2 C compiler */
+    extern char etext;
+#   define DATASTART ((ptr_t)(&etext))
+#   ifdef BSD
+#	define OS_TYPE "BSD"
+#	define HEURISTIC1
+			/* HEURISTIC2 may be OK, but it's hard to test. */
+#   endif
+#   ifdef ULTRIX
+#	define OS_TYPE "ULTRIX"
+#	define STACKBOTTOM ((ptr_t) 0x7fffc800)
+#   endif
+# endif
+
+# ifdef RT
+#   define MACH_TYPE "RT"
+#   define ALIGNMENT 4
+#   define DATASTART ((ptr_t) 0x10000000)
+#   define STACKBOTTOM ((ptr_t) 0x1fffd800)
+# endif
+
+# ifdef SPARC
+#   define MACH_TYPE "SPARC"
+#   define ALIGNMENT 4	/* Required by hardware	*/
+    extern int etext;
+#   ifdef SUNOS5
+#	define OS_TYPE "SUNOS5"
+#       define DATASTART ((ptr_t)((((word) (&etext)) + 0x10003) & ~0x3))
+#	define PROC_VDB
+#   endif
+#   ifdef SUNOS4
+#	define OS_TYPE "SUNOS4"
+	/* [If you have a weak stomach, don't read this.]		*/
+	/* We would like to use:					*/
+/* #       define DATASTART ((ptr_t)((((word) (&etext)) + 0x1fff) & ~0x1fff)) */
+	/* This fails occasionally, due to an ancient, but very 	*/
+	/* persistent ld bug.  &etext is set 32 bytes too high.		*/
+	/* We instead read the text segment size from the a.out		*/
+	/* header, which happens to be mapped into our address space	*/
+	/* at the start of the text segment.  The detective work here	*/
+	/* was done by Robert Ehrlich, Manuel Serrano, and Bernard	*/
+	/* Serpette of INRIA.						*/
+	/* This assumes ZMAGIC, i.e. demand-loadable executables.	*/
+#       define DATASTART ((ptr_t)(*(int *)0x2004+0x2000))
+#	define MPROTECT_VDB
+#   endif
+#   define HEURISTIC1
+#   define DYNAMIC_LOADING
+# endif
+
+# ifdef I386
+#   define MACH_TYPE "I386"
+#   define ALIGNMENT 4	/* Appears to hold for all "32 bit" compilers */
+#   ifdef SEQUENT
+#	define OS_TYPE "SEQUENT"
+	extern int etext;
+#       define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
+#       define STACKBOTTOM ((ptr_t) 0x3ffff000) 
+#   endif
+#   ifdef SUNOS5
+#	define OS_TYPE "SUNOS5"
+  	extern int etext;
+#       define DATASTART ((ptr_t)((((word) (&etext)) + 0x1003) & ~0x3))
+	extern int _start();
+#	define STACKBOTTOM ((ptr_t)(&_start))
+#	define PROC_VDB
+#   endif
+#   ifdef SCO
+#	define OS_TYPE "SCO"
+#   	define DATASTART ((ptr_t)((((word) (&etext)) + 0x3fffff) \
+				  & ~0x3fffff) \
+				 +((word)&etext & 0xfff))
+#	define STACKBOTTOM ((ptr_t) 0x7ffffffc)
+#   endif
+#   ifdef LINUX
+#	define OS_TYPE "LINUX"
+	extern int etext;
+#       define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
+#	define STACKBOTTOM ((ptr_t)0xc0000000)
+#   endif
+#   ifdef OS2
+#	define OS_TYPE "OS2"
+ 	    	/* STACKBOTTOM and DATASTART are handled specially in 	*/
+		/* os_dep.c. OS2 actually has the right			*/
+		/* system call!						*/
+#   endif
+#   ifdef MSWIN32
+#	define OS_TYPE "MSWIN32"
+		/* STACKBOTTOM and DATASTART are handled specially in 	*/
+		/* os_dep.c.						*/
+#   endif
+#   ifdef FREEBSD
+#	define OS_TYPE "FREEBSD"
+#	define MPROTECT_VDB
+#   endif
+#   ifdef NETBSD
+#	define OS_TYPE "NETBSD"
+#   endif
+#   ifdef THREE86BSD
+#	define OS_TYPE "THREE86BSD"
+#   endif
+#   ifdef BSDI
+#	define OS_TYPE "BSDI"
+#   endif
+#   if defined(FREEBSD) || defined(NETBSD) \
+        || defined(THREE86BSD) || defined(BSDI)
+#	define HEURISTIC2
+	extern char etext;
+#	define DATASTART ((ptr_t)(&etext))
+#   endif
+# endif
+
+# ifdef NS32K
+#   define MACH_TYPE "NS32K"
+#   define ALIGNMENT 4
+    extern char **environ;
+#   define DATASTART ((ptr_t)(&environ))
+			      /* hideous kludge: environ is the first   */
+			      /* word in crt0.o, and delimits the start */
+			      /* of the data segment, no matter which   */
+			      /* ld options were passed through.        */
+#   define STACKBOTTOM ((ptr_t) 0xfffff000) /* for Encore */
+# endif
+
+# ifdef MIPS
+#   define MACH_TYPE "MIPS"
+#   define ALIGNMENT 4	/* Required by hardware	*/
+#   define DATASTART 0x10000000
+			      /* Could probably be slightly higher since */
+			      /* startup code allocates lots of junk     */
+#   define HEURISTIC2
+#   ifdef ULTRIX
+#	define OS_TYPE "ULTRIX"
+#   endif
+#   ifdef RISCOS
+#	define OS_TYPE "RISCOS"
+#   endif
+#   ifdef IRIX5
+#	define OS_TYPE "IRIX5"
+#	define MPROTECT_VDB
+#	define DYNAMIC_LOADING
+#   endif
+# endif
+
+# ifdef RS6000
+#   define MACH_TYPE "RS6000"
+#   define ALIGNMENT 4
+#   define DATASTART ((ptr_t)0x20000000)
+#   define STACKBOTTOM ((ptr_t)0x2ff80000)
+# endif
+
+# ifdef HP_PA
+#   define MACH_TYPE "HP_PA"
+#   define ALIGNMENT 4
+    extern int __data_start;
+#   define DATASTART ((ptr_t)(&__data_start))
+#   define HEURISTIC2
+#   define STACK_GROWS_UP
+# endif
+
+# ifdef ALPHA
+#   define MACH_TYPE "ALPHA"
+#   define ALIGNMENT 8
+#   define DATASTART ((ptr_t) 0x140000000)
+#   define HEURISTIC2
+#   define CPP_WORDSZ 64
+#   define MPROTECT_VDB
+# endif
+
+# ifdef M88K
+#   define MACH_TYPE "M88K"
+#   define ALIGNMENT 4
+#   define DATASTART ((((word)&etext + 0x3fffff) & ~0x3fffff) + 0x10000)
+#   define STACKBOTTOM ((char*)0xf0000000) /* determined empirically */
+# endif
+
+# ifndef STACK_GROWS_UP
+#   define STACK_GROWS_DOWN
+# endif
+
+# ifndef CPP_WORDSZ
+#   define CPP_WORDSZ 32
+# endif
+
+# ifndef OS_TYPE
+#   define OS_TYPE ""
+# endif
+
+# if CPP_WORDSZ != 32 && CPP_WORDSZ != 64
+   -> bad word size
+# endif
+
+# ifdef PCR
+#   undef DYNAMIC_LOADING
+#   undef STACKBOTTOM
+#   undef HEURISTIC1
+#   undef HEURISTIC2
+#   undef PROC_VDB
+#   undef MPROTECT_VDB
+#   define PCR_VDB
+# endif
+
+# ifdef SRC_M3
+/* Postponed for now. */
+#   undef PROC_VDB
+#   undef MPROTECT_VDB
+# endif
+
+# ifdef SMALL_CONFIG
+/* Presumably not worth the space it takes. */
+#   undef PROC_VDB
+#   undef MPROTECT_VDB
+# endif
+
+# if !defined(PCR_VDB) && !defined(PROC_VDB) && !defined(MPROTECT_VDB)
+#   define DEFAULT_VDB
+# endif
+
+# endif
diff --git a/cord/README b/cord/README
new file mode 100644
index 00000000..865725ee
--- /dev/null
+++ b/cord/README
@@ -0,0 +1,31 @@
+Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+
+THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+
+Permission is hereby granted to use or copy this program
+for any purpose,  provided the above notices are retained on all copies.
+Permission to modify the code and to distribute modified code is granted,
+provided the above notices are retained, and a notice that the code was
+modified is included with the above copyright notice.
+
+Please send bug reports to Hans-J. Boehm (boehm@parc.xerox.com).
+
+This is a string packages that uses a tree-based representation.
+See gc.h for a description of the functions provided.  Ec.h describes
+"extensible cords", which are essentially output streams that write
+to a cord.  These allow for efficient construction of cords without
+requiring a bound on the size of a cord.
+
+de.c is a very dumb text editor that illustrates the use of cords.
+It maintains a list of file versions.  Each version is simply a
+cord representing the file contents.  Nonetheless, standard
+editing operations are efficient, even on very large files.
+(Its 3 line "user manual" can be obtained by invoking it without
+arguments.  Note that ^R^N and ^R^P move the cursor by
+almost a screen.  It does not understand tabs, which will show
+up as highlighred "I"s.  Use the UNIX "expand" program first.)
+To build the editor, type "make cord/de" in the gc directory.
+
+This package assumes an ANSI C compiler such as gcc.  It will
+not compile with an old-style K&R compiler.
diff --git a/cord/cord.h b/cord/cord.h
new file mode 100644
index 00000000..cdf5e03c
--- /dev/null
+++ b/cord/cord.h
@@ -0,0 +1,297 @@
+/* 
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ * Author: Hans-J. Boehm (boehm@parc.xerox.com)
+ */
+/* Boehm, May 19, 1994 2:22 pm PDT */
+ 
+/*
+ * Cords are immutable character strings.  A number of operations
+ * on long cords are much more efficient than their strings.h counterpart.
+ * In particular, concatenation takes constant time independent of the length
+ * of the arguments.  (Cords are represented as trees, with internal
+ * nodes representing concatenation and leaves consisting of either C
+ * strings or a functional description of the string.)
+ *
+ * The following are reasonable applications of cords.  They would perform
+ * unacceptably if C strings were used:
+ * - A compiler that produces assembly language output by repeatedly
+ *   concatenating instructions onto a cord representing the output file.
+ * - A text editor that converts the input file to a cord, and then
+ *   performs editing operations by producing a new cord representing
+ *   the file after echa character change (and keeping the old ones in an
+ *   edit history)
+ *
+ * For optimal performance, cords should be built by
+ * concatenating short sections.
+ * This interface is designed for maximum compatibility with C strings.
+ * ASCII NUL characters may be embedded in cords using CORD_from_fn.
+ * This is handled correctly, but CORD_to_char_star will produce a string
+ * with embedded NULs when given such a cord. 
+ */
+# ifndef CORD_H
+
+# define CORD_H
+# include <stddef.h>
+# include <stdio.h>
+/* Cords have type const char *.  This is cheating quite a bit, and not	*/
+/* 100% portable.  But it means that nonempty character string		*/
+/* constants may be used as cords directly, provided the string is	*/
+/* never modified in place.  The empty cord is represented by, and	*/
+/* can be written as, 0.						*/
+
+typedef const char * CORD;
+
+/* An empty cord is always represented as nil 	*/
+# define CORD_EMPTY 0
+
+/* Is a nonempty cord represented as a C string? */
+#define IS_STRING(s) (*(s) != '\0')
+
+/* Concatenate two cords.  If the arguments are C strings, they may 	*/
+/* not be subsequently altered.						*/
+CORD CORD_cat(CORD x, CORD y);
+
+/* Concatenate a cord and a C string with known length.  Except for the	*/
+/* empty string case, this is a special case of CORD_cat.  Since the	*/
+/* length is known, it can be faster.					*/
+CORD CORD_cat_char_star(CORD x, const char * y, size_t leny);
+
+/* Compute the length of a cord */
+size_t CORD_len(CORD x);
+
+/* Cords may be represented by functions defining the ith character */
+typedef char (* CORD_fn)(size_t i, void * client_data);
+
+/* Turn a functional description into a cord. 	*/
+CORD CORD_from_fn(CORD_fn fn, void * client_data, size_t len);
+
+/* Return the substring (subcord really) of x with length at most n,	*/
+/* starting at position i.  (The initial character has position 0.)	*/
+CORD CORD_substr(CORD x, size_t i, size_t n);
+
+/* Return the argument, but rebalanced to allow more efficient   	*/
+/* character retrieval, substring operations, and comparisons.		*/
+/* This is useful only for cords that were built using repeated 	*/
+/* concatenation.  Guarantees log time access to the result, unless	*/
+/* x was obtained through a large number of repeated substring ops	*/
+/* or the embedded functional descriptions take longer to evaluate.	*/
+/* May reallocate significant parts of the cord.  The argument is not	*/
+/* modified; only the result is balanced.				*/
+CORD CORD_balance(CORD x);
+
+/* The following traverse a cord by applying a function to each 	*/
+/* character.  This is occasionally appropriate, especially where	*/
+/* speed is crucial.  But, since C doesn't have nested functions,	*/
+/* clients of this sort of traversal are clumsy to write.  Consider	*/
+/* the functions that operate on cord positions instead.		*/
+
+/* Function to iteratively apply to individual characters in cord.	*/
+typedef int (* CORD_iter_fn)(char c, void * client_data);
+
+/* Function to apply to substrings of a cord.  Each substring is a 	*/
+/* a C character string, not a general cord.				*/
+typedef int (* CORD_batched_iter_fn)(const char * s, void * client_data);
+# define CORD_NO_FN ((CORD_batched_iter_fn)0)
+
+/* Apply f1 to each character in the cord, in ascending order,		*/
+/* starting at position i. If						*/
+/* f2 is not CORD_NO_FN, then multiple calls to f1 may be replaced by	*/
+/* a single call to f2.  The parameter f2 is provided only to allow	*/
+/* some optimization by the client.  This terminates when the right	*/
+/* end of this string is reached, or when f1 or f2 return != 0.  In the	*/
+/* latter case CORD_iter returns != 0.  Otherwise it returns 0.		*/
+/* The specified value of i must be < CORD_len(x).			*/
+int CORD_iter5(CORD x, size_t i, CORD_iter_fn f1,
+	       CORD_batched_iter_fn f2, void * client_data);
+
+/* A simpler version that starts at 0, and without f2:	*/
+int CORD_iter(CORD x, CORD_iter_fn f1, void * client_data);
+# define CORD_iter(x, f1, cd) CORD_iter5(x, 0, f1, CORD_NO_FN, cd)
+
+/* Similar to CORD_iter5, but end-to-beginning.	No provisions for	*/
+/* CORD_batched_iter_fn.						*/
+int CORD_riter4(CORD x, size_t i, CORD_iter_fn f1, void * client_data);
+
+/* A simpler version that starts at the end:	*/
+int CORD_riter(CORD x, CORD_iter_fn f1, void * client_data);
+
+/* Functions that operate on cord positions.  The easy way to traverse	*/
+/* cords.  A cord position is logically a pair consisting of a cord	*/
+/* and an index into that cord.  But it is much faster to retrieve a	*/
+/* charcter based on a position than on an index.  Unfortunately,	*/
+/* positions are big (order of a few 100 bytes), so allocate them with	*/
+/* caution.								*/
+/* Things in cord_pos.h should be treated as opaque, except as		*/
+/* described below.  Also note that					*/
+/* CORD_pos_fetch, CORD_next and CORD_prev have both macro and function	*/
+/* definitions.  The former may evaluate their argument more than once. */
+# include "cord_pos.h"
+
+/*
+	Visible definitions from above:
+	
+	typedef <OPAQUE but fairly big> CORD_pos[1];
+	
+	/* Extract the cord from a position:
+	CORD CORD_pos_to_cord(CORD_pos p);
+	
+	/* Extract the current index from a position:
+	size_t CORD_pos_to_index(CORD_pos p);
+	
+	/* Fetch the character located at the given position:
+	char CORD_pos_fetch(register CORD_pos p);
+	
+	/* Initialize the position to refer to the give cord and index.
+	/* Note that this is the most expensive function on positions:
+	void CORD_set_pos(CORD_pos p, CORD x, size_t i);
+	
+	/* Advance the position to the next character.
+	/* P must be initialized and valid.
+	/* Invalidates p if past end:
+	void CORD_next(CORD_pos p);
+	
+	/* Move the position to the preceding character.
+	/* P must be initialized and valid.
+	/* Invalidates p if past beginning:
+	void CORD_prev(CORD_pos p);
+	
+	/* Is the position valid, i.e. inside the cord?
+	int CORD_pos_valid(CORD_pos p);
+*/
+# define CORD_FOR(pos, cord) \
+    for (CORD_set_pos(pos, cord, 0); CORD_pos_valid(pos); CORD_next(pos))
+
+			
+/* An out of memory handler to call.  May be supplied by client.	*/
+/* Must not return.							*/
+extern void (* CORD_oom_fn)(void);
+
+/* Dump the representation of x to stdout in an implementation defined	*/
+/* manner.  Intended for debugging only.				*/
+void CORD_dump(CORD x);
+
+/* The following could easily be implemented by the client.  They are	*/
+/* provided in cord_xtra.c for convenience.				*/
+
+/* Concatenate a character to the end of a cord.	*/
+CORD CORD_cat_char(CORD x, char c);
+
+/* Return the character in CORD_substr(x, i, 1)  	*/
+char CORD_fetch(CORD x, size_t i);
+
+/* Return < 0, 0, or > 0, depending on whether x < y, x = y, x > y	*/
+int CORD_cmp(CORD x, CORD y);
+
+/* A generalization that takes both starting positions for the 		*/
+/* comparison, and a limit on the number of characters to be compared.	*/
+int CORD_ncmp(CORD x, size_t x_start, CORD y, size_t y_start, size_t len);
+
+/* Find the first occurrence of s in x at position start or later.	*/
+/* Return the position of the first character of s in x, or		*/
+/* CORD_NOT_FOUND if there is none.					*/
+size_t CORD_str(CORD x, size_t start, CORD s);
+
+/* Return a cord consisting of i copies of (possibly NUL) c.  Dangerous	*/
+/* in conjunction with CORD_to_char_star.				*/
+/* The resulting representation takes constant space, independent of i.	*/
+CORD CORD_chars(char c, size_t i);
+# define CORD_nul(i) CORD_chars('\0', (i))
+
+/* Turn a file into cord.  The file must be seekable.  Its contents	*/
+/* must remain constant.  The file may be accessed as an immediate	*/
+/* result of this call and/or as a result of subsequent accesses to 	*/
+/* the cord.  Short files are likely to be immediately read, but	*/
+/* long files are likely to be read on demand, possibly relying on 	*/
+/* stdio for buffering.							*/
+/* We must have exclusive access to the descriptor f, i.e. we may	*/
+/* read it at any time, and expect the file pointer to be		*/
+/* where we left it.  Normally this should be invoked as		*/
+/* CORD_from_file(fopen(...))						*/
+/* CORD_from_file arranges to close the file descriptor when it is no	*/
+/* longer needed (e.g. when the result becomes inaccessible).		*/ 
+/* The file f must be such that ftell reflects the actual character	*/
+/* position in the file, i.e. the number of characters that can be 	*/
+/* or were read with fread.  On UNIX systems this is always true.  On	*/
+/* MS Windows systems, f must be opened in binary mode.			*/
+CORD CORD_from_file(FILE * f);
+
+/* Equivalent to the above, except that the entire file will be read	*/
+/* and the file pointer will be closed immediately.			*/
+/* The binary mode restriction from above does not apply.		*/
+CORD CORD_from_file_eager(FILE * f);
+
+/* Equivalent to the above, except that the file will be read on demand.*/
+/* The binary mode restriction applies.					*/
+CORD CORD_from_file_lazy(FILE * f);
+
+/* Turn a cord into a C string.	The result shares no structure with	*/
+/* x, and is thus modifiable.						*/
+char * CORD_to_char_star(CORD x);
+
+/* Write a cord to a file, starting at the current position.  No	*/
+/* trailing NULs are newlines are added.				*/
+/* Returns EOF if a write error occurs, 1 otherwise.			*/
+int CORD_put(CORD x, FILE * f);
+
+/* "Not found" result for the following two functions.			*/
+# define CORD_NOT_FOUND ((size_t)(-1))
+
+/* A vague analog of strchr.  Returns the position (an integer, not	*/
+/* a pointer) of the first occurrence of (char) c inside x at position 	*/
+/* i or later. The value i must be < CORD_len(x).			*/
+size_t CORD_chr(CORD x, size_t i, int c);
+
+/* A vague analog of strrchr.  Returns index of the last occurrence	*/
+/* of (char) c inside x at position i or earlier. The value i		*/
+/* must be < CORD_len(x).						*/
+size_t CORD_rchr(CORD x, size_t i, int c);
+
+
+/* The following are also not primitive, but are implemented in 	*/
+/* cordprnt.c.  They provide functionality similar to the ANSI C	*/
+/* functions with corresponding names, but with the following		*/
+/* additions and changes:						*/
+/* 1. A %r conversion specification specifies a CORD argument.  Field	*/
+/*    width, precision, etc. have the same semantics as for %s.		*/
+/*    (Note that %c,%C, and %S were already taken.)			*/
+/* 2. The format string is represented as a CORD.		        */
+/* 3. CORD_sprintf and CORD_vsprintf assign the result through the 1st	*/ 	/*    argument.	Unlike their ANSI C versions, there is no need to guess	*/
+/*    the correct buffer size.						*/
+/* 4. Most of the conversions are implement through the native 		*/
+/*    vsprintf.  Hence they are usually no faster, and 			*/
+/*    idiosyncracies of the native printf are preserved.  However,	*/
+/*    CORD arguments to CORD_sprintf and CORD_vsprintf are NOT copied;	*/
+/*    the result shares the original structure.  This may make them	*/
+/*    very efficient in some unusual applications.			*/
+/*    The format string is copied.					*/
+/* All functions return the number of characters generated or -1 on	*/
+/* error.  This complies with the ANSI standard, but is inconsistent	*/
+/* with some older implementations of sprintf.				*/
+
+/* The implementation of these is probably less portable than the rest	*/
+/* of this package.							*/
+
+#ifndef CORD_NO_IO
+
+#include <stdarg.h>
+
+int CORD_sprintf(CORD * out, CORD format, ...);
+int CORD_vsprintf(CORD * out, CORD format, va_list args);
+int CORD_fprintf(FILE * f, CORD format, ...);
+int CORD_vfprintf(FILE * f, CORD format, va_list args);
+int CORD_printf(CORD format, ...);
+int CORD_vprintf(CORD format, va_list args);
+
+#endif /* CORD_NO_IO */
+
+# endif /* CORD_H */
diff --git a/cord/cord_pos.h b/cord/cord_pos.h
new file mode 100644
index 00000000..a07d07f6
--- /dev/null
+++ b/cord/cord_pos.h
@@ -0,0 +1,118 @@
+/* 
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:23 pm PDT */
+# ifndef CORD_POSITION_H
+
+/* The representation of CORD_position.  This is private to the	*/
+/* implementation, but the ise is known to clients.  Also	*/
+/* the implementation of some exported macros relies on it.	*/
+/* Don't use anything defined here and not in cord.h.		*/
+
+# define MAX_DEPTH 48
+	/* The maximum depth of a balanced cord + 1.		*/
+	/* We don't let cords get deeper than MAX_DEPTH.	*/
+
+struct CORD_pe {
+    CORD pe_cord;
+    size_t pe_start_pos;
+};
+
+/* A structure describing an entry on the path from the root 	*/
+/* to current position.						*/
+typedef struct CORD_pos {
+    size_t cur_pos;
+    int path_len;
+#	define CORD_POS_INVALID (0x55555555)
+		/* path_len == INVALID <==> position invalid */
+    const char *cur_leaf;	/* Current leaf, if it is a string.	*/
+    				/* If the current leaf is a function,	*/
+    				/* then this may point to function_buf	*/
+    				/* containing the next few characters.	*/
+    				/* Always points to a valid string	*/
+    				/* containing the current character 	*/
+    				/* unless cur_end is 0.			*/
+    size_t cur_start;	/* Start position of cur_leaf	*/
+    size_t cur_end;	/* Ending position of cur_leaf	*/
+    			/* 0 if cur_leaf is invalid.	*/
+    struct CORD_pe path[MAX_DEPTH + 1];
+    	/* path[path_len] is the leaf corresponding to cur_pos	*/
+    	/* path[0].pe_cord is the cord we point to.		*/
+#   define FUNCTION_BUF_SZ 8
+    char function_buf[FUNCTION_BUF_SZ];	/* Space for next few chars	*/
+    					/* from function node.		*/
+} CORD_pos[1];
+
+/* Extract the cord from a position:	*/
+CORD CORD_pos_to_cord(CORD_pos p);
+	
+/* Extract the current index from a position:	*/
+size_t CORD_pos_to_index(CORD_pos p);
+	
+/* Fetch the character located at the given position:	*/
+char CORD_pos_fetch(CORD_pos p);
+	
+/* Initialize the position to refer to the give cord and index.	*/
+/* Note that this is the most expensive function on positions:	*/
+void CORD_set_pos(CORD_pos p, CORD x, size_t i);
+	
+/* Advance the position to the next character.	*/
+/* P must be initialized and valid.		*/
+/* Invalidates p if past end:			*/
+void CORD_next(CORD_pos p);
+
+/* Move the position to the preceding character.	*/
+/* P must be initialized and valid.			*/
+/* Invalidates p if past beginning:			*/
+void CORD_prev(CORD_pos p);
+	
+/* Is the position valid, i.e. inside the cord?		*/
+int CORD_pos_valid(CORD_pos p);
+
+char CORD__pos_fetch(CORD_pos);
+void CORD__next(CORD_pos);
+void CORD__prev(CORD_pos);
+
+#define CORD_pos_fetch(p)	\
+    (((p)[0].cur_end != 0)? \
+     	(p)[0].cur_leaf[(p)[0].cur_pos - (p)[0].cur_start] \
+     	: CORD__pos_fetch(p))
+
+#define CORD_next(p)	\
+    (((p)[0].cur_pos + 1 < (p)[0].cur_end)? \
+    	(p)[0].cur_pos++ \
+    	: (CORD__next(p), 0))
+
+#define CORD_prev(p)	\
+    (((p)[0].cur_end != 0 && (p)[0].cur_pos > (p)[0].cur_start)? \
+    	(p)[0].cur_pos-- \
+    	: (CORD__prev(p), 0))
+
+#define CORD_pos_to_index(p) ((p)[0].cur_pos)
+
+#define CORD_pos_to_cord(p) ((p)[0].path[0].pe_cord)
+
+#define CORD_pos_valid(p) ((p)[0].path_len != CORD_POS_INVALID)
+
+/* Some grubby stuff for performance-critical friends:	*/
+#define CORD_pos_chars_left(p) ((long)((p)[0].cur_end) - (long)((p)[0].cur_pos))
+	/* Number of characters in cache.  <= 0 ==> none	*/
+
+#define CORD_pos_advance(p,n) ((p)[0].cur_pos += (n) - 1, CORD_next(p))
+	/* Advance position by n characters	*/
+	/* 0 < n < CORD_pos_chars_left(p)	*/
+
+#define CORD_pos_cur_char_addr(p) \
+	(p)[0].cur_leaf + ((p)[0].cur_pos - (p)[0].cur_start)
+	/* address of current character in cache.	*/
+
+#endif
diff --git a/cord/cordbscs.c b/cord/cordbscs.c
new file mode 100644
index 00000000..d828155b
--- /dev/null
+++ b/cord/cordbscs.c
@@ -0,0 +1,913 @@
+/*
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ * Author: Hans-J. Boehm (boehm@parc.xerox.com)
+ */
+/* Boehm, May 19, 1994 2:18 pm PDT */
+# include "../gc.h"
+# include "cord.h"
+# include <stdlib.h>
+# include <stdio.h>
+# include <string.h>
+
+/* An implementation of the cord primitives.  These are the only 	*/
+/* Functions that understand the representation.  We perform only	*/
+/* minimal checks on arguments to these functions.  Out of bounds	*/
+/* arguments to the iteration functions may result in client functions	*/
+/* invoked on garbage data.  In most cases, client functions should be	*/
+/* programmed defensively enough that this does not result in memory	*/
+/* smashes.								*/ 
+
+typedef void (* oom_fn)(void);
+
+oom_fn CORD_oom_fn = (oom_fn) 0;
+
+# define OUT_OF_MEMORY {  if (CORD_oom_fn != (oom_fn) 0) (*CORD_oom_fn)(); \
+			  ABORT("Out of memory\n"); }
+# define ABORT(msg) { fprintf(stderr, "%s\n", msg); abort(); }
+
+typedef unsigned long word;
+
+typedef union {
+    struct Concatenation {
+    	char null;
+	char header;
+	char depth;	/* concatenation nesting depth. */
+	unsigned char left_len;
+			/* Length of left child if it is sufficiently	*/
+			/* short; 0 otherwise.				*/
+#	    define MAX_LEFT_LEN 255
+	word len;
+	CORD left;	/* length(left) > 0	*/
+	CORD right;	/* length(right) > 0	*/
+    } concatenation;
+    struct Function {
+	char null;
+	char header;
+	char depth;	/* always 0	*/
+	char left_len;	/* always 0	*/
+	word len;
+	CORD_fn fn;
+	void * client_data;
+    } function;
+    struct Generic {
+    	char null;
+	char header;
+	char depth;
+	char left_len;
+	word len;
+    } generic;
+    char string[1];
+} CordRep;
+
+# define CONCAT_HDR 1
+	
+# define FN_HDR 4
+# define SUBSTR_HDR 6
+	/* Substring nodes are a special case of function nodes.  	*/
+	/* The client_data field is known to point to a substr_args	*/
+	/* structure, and the function is either CORD_apply_access_fn 	*/
+	/* or CORD_index_access_fn.					*/
+
+/* The following may be applied only to function and concatenation nodes: */
+#define IS_CONCATENATION(s)  (((CordRep *)s)->generic.header == CONCAT_HDR)
+
+#define IS_FUNCTION(s)  ((((CordRep *)s)->generic.header & FN_HDR) != 0)
+
+#define IS_SUBSTR(s) (((CordRep *)s)->generic.header == SUBSTR_HDR)
+
+#define LEN(s) (((CordRep *)s) -> generic.len)
+#define DEPTH(s) (((CordRep *)s) -> generic.depth)
+#define GEN_LEN(s) (IS_STRING(s) ? strlen(s) : LEN(s))
+
+#define LEFT_LEN(c) ((c) -> left_len != 0? \
+				(c) -> left_len \
+				: (IS_STRING((c) -> left) ? \
+					(c) -> len - GEN_LEN((c) -> right) \
+					: LEN((c) -> left)))
+
+#define SHORT_LIMIT (sizeof(CordRep) - 1)
+	/* Cords shorter than this are C strings */
+
+
+/* Dump the internal representation of x to stdout, with initial 	*/
+/* indentation level n.							*/
+void CORD_dump_inner(CORD x, unsigned n)
+{
+    register size_t i;
+    
+    for (i = 0; i < (size_t)n; i++) {
+        fputs("  ", stdout);
+    }
+    if (x == 0) {
+      	fputs("NIL\n", stdout);
+    } else if (IS_STRING(x)) {
+        for (i = 0; i <= SHORT_LIMIT; i++) {
+            if (x[i] == '\0') break;
+            putchar(x[i]);
+        }
+        if (x[i] != '\0') fputs("...", stdout);
+        putchar('\n');
+    } else if (IS_CONCATENATION(x)) {
+        register struct Concatenation * conc =
+        			&(((CordRep *)x) -> concatenation);
+        printf("Concatenation: %p (len: %d, depth: %d)\n",
+               x, (int)(conc -> len), (int)(conc -> depth));
+        CORD_dump_inner(conc -> left, n+1);
+        CORD_dump_inner(conc -> right, n+1);
+    } else /* function */{
+        register struct Function * func =
+        			&(((CordRep *)x) -> function);
+        if (IS_SUBSTR(x)) printf("(Substring) ");
+        printf("Function: %p (len: %d): ", x, (int)(func -> len));
+        for (i = 0; i < 20 && i < func -> len; i++) {
+            putchar((*(func -> fn))(i, func -> client_data));
+        }
+        if (i < func -> len) fputs("...", stdout);
+        putchar('\n');
+    }
+}
+
+/* Dump the internal representation of x to stdout	*/
+void CORD_dump(CORD x)
+{
+    CORD_dump_inner(x, 0);
+    fflush(stdout);
+}
+
+CORD CORD_cat_char_star(CORD x, const char * y, size_t leny)
+{
+    register size_t result_len;
+    register size_t lenx;
+    register int depth;
+    
+    if (x == CORD_EMPTY) return(y);
+    if (leny == 0) return(x);
+    if (IS_STRING(x)) {
+        lenx = strlen(x);
+        result_len = lenx + leny;
+        if (result_len <= SHORT_LIMIT) {
+            register char * result = GC_MALLOC_ATOMIC(result_len+1);
+        
+            if (result == 0) OUT_OF_MEMORY;
+            memcpy(result, x, lenx);
+            memcpy(result + lenx, y, leny);
+            result[result_len] = '\0';
+            return((CORD) result);
+        } else {
+            depth = 1;
+        }
+    } else {
+    	register CORD right;
+    	register CORD left;
+    	register char * new_right;
+    	register size_t right_len;
+    	
+    	lenx = LEN(x);
+    	
+        if (leny <= SHORT_LIMIT/2
+    	    && IS_CONCATENATION(x)
+            && IS_STRING(right = ((CordRep *)x) -> concatenation.right)) {
+            /* Merge y into right part of x. */
+            if (!IS_STRING(left = ((CordRep *)x) -> concatenation.left)) {
+            	right_len = lenx - LEN(left);
+            } else if (((CordRep *)x) -> concatenation.left_len != 0) {
+                right_len = lenx - ((CordRep *)x) -> concatenation.left_len;
+            } else {
+            	right_len = strlen(right);
+            }
+            result_len = right_len + leny;  /* length of new_right */
+            if (result_len <= SHORT_LIMIT) {
+            	new_right = GC_MALLOC_ATOMIC(result_len + 1);
+            	memcpy(new_right, right, right_len);
+            	memcpy(new_right + right_len, y, leny);
+            	new_right[result_len] = '\0';
+            	y = new_right;
+            	leny = result_len;
+            	x = left;
+            	lenx -= right_len;
+            	/* Now fall through to concatenate the two pieces: */
+            }
+            if (IS_STRING(x)) {
+                depth = 1;
+            } else {
+                depth = DEPTH(x) + 1;
+            }
+        } else {
+            depth = DEPTH(x) + 1;
+        }
+        result_len = lenx + leny;
+    }
+    {
+      /* The general case; lenx, result_len is known: */
+    	register struct Concatenation * result;
+    	
+    	result = GC_NEW(struct Concatenation);
+    	if (result == 0) OUT_OF_MEMORY;
+    	result->header = CONCAT_HDR;
+    	result->depth = depth;
+    	if (lenx <= MAX_LEFT_LEN) result->left_len = lenx;
+    	result->len = result_len;
+    	result->left = x;
+    	result->right = y;
+    	if (depth > MAX_DEPTH) {
+    	    return(CORD_balance((CORD)result));
+    	} else {
+    	    return((CORD) result);
+    	}
+    }
+}
+
+
+CORD CORD_cat(CORD x, CORD y)
+{
+    register size_t result_len;
+    register int depth;
+    register size_t lenx;
+    
+    if (x == CORD_EMPTY) return(y);
+    if (y == CORD_EMPTY) return(x);
+    if (IS_STRING(y)) {
+        return(CORD_cat_char_star(x, y, strlen(y)));
+    } else if (IS_STRING(x)) {
+        lenx = strlen(x);
+        depth = DEPTH(y) + 1;
+    } else {
+        register int depthy = DEPTH(y);
+        
+        lenx = LEN(x);
+        depth = DEPTH(x) + 1;
+        if (depthy >= depth) depth = depthy + 1;
+    }
+    result_len = lenx + LEN(y);
+    {
+    	register struct Concatenation * result;
+    	
+    	result = GC_NEW(struct Concatenation);
+    	if (result == 0) OUT_OF_MEMORY;
+    	result->header = CONCAT_HDR;
+    	result->depth = depth;
+    	if (lenx <= MAX_LEFT_LEN) result->left_len = lenx;
+    	result->len = result_len;
+    	result->left = x;
+    	result->right = y;
+    	return((CORD) result);
+    }
+}
+
+
+
+CORD CORD_from_fn(CORD_fn fn, void * client_data, size_t len)
+{
+    if (len <= 0) return(0);
+    if (len <= SHORT_LIMIT) {
+        register char * result;
+        register size_t i;
+        char buf[SHORT_LIMIT+1];
+        register char c;
+        
+        for (i = 0; i < len; i++) {
+            c = (*fn)(i, client_data);
+            if (c == '\0') goto gen_case;
+            buf[i] = c;
+        }
+        buf[i] = '\0';
+        result = GC_MALLOC_ATOMIC(len+1);
+        if (result == 0) OUT_OF_MEMORY;
+        strcpy(result, buf);
+        result[len] = '\0';
+        return((CORD) result);
+    }
+  gen_case:
+    {
+    	register struct Function * result;
+    	
+    	result = GC_NEW(struct Function);
+    	if (result == 0) OUT_OF_MEMORY;
+    	result->header = FN_HDR;
+    	/* depth is already 0 */
+    	result->len = len;
+    	result->fn = fn;
+    	result->client_data = client_data;
+    	return((CORD) result);
+    }
+}
+
+size_t CORD_len(CORD x)
+{
+    if (x == 0) {
+     	return(0);
+    } else {
+	return(GEN_LEN(x));
+    }
+}
+
+struct substr_args {
+    CordRep * sa_cord;
+    size_t sa_index;
+};
+
+char CORD_index_access_fn(size_t i, void * client_data)
+{
+    register struct substr_args *descr = (struct substr_args *)client_data;
+    
+    return(((char *)(descr->sa_cord))[i + descr->sa_index]);
+}
+
+char CORD_apply_access_fn(size_t i, void * client_data)
+{
+    register struct substr_args *descr = (struct substr_args *)client_data;
+    register struct Function * fn_cord = &(descr->sa_cord->function);
+    
+    return((*(fn_cord->fn))(i + descr->sa_index, fn_cord->client_data));
+}
+
+/* A version of CORD_substr that simply returns a function node, thus	*/
+/* postponing its work.	The fourth argument is a function that may	*/
+/* be used for efficient access to the ith character.			*/
+/* Assumes i >= 0 and i + n < length(x).				*/
+CORD CORD_substr_closure(CORD x, size_t i, size_t n, CORD_fn f)
+{
+    register struct substr_args * sa = GC_NEW(struct substr_args);
+    CORD result;
+    
+    if (sa == 0) OUT_OF_MEMORY;
+    sa->sa_cord = (CordRep *)x;
+    sa->sa_index = i;
+    result = CORD_from_fn(f, (void *)sa, n);
+    ((CordRep *)result) -> function.header = SUBSTR_HDR;
+    return (result);
+}
+
+# define SUBSTR_LIMIT (10 * SHORT_LIMIT)
+	/* Substrings of function nodes and flat strings shorter than 	*/
+	/* this are flat strings.  Othewise we use a functional 	*/
+	/* representation, which is significantly slower to access.	*/
+
+/* A version of CORD_substr that assumes i >= 0, n > 0, and i + n < length(x).*/
+CORD CORD_substr_checked(CORD x, size_t i, size_t n)
+{
+    if (IS_STRING(x)) {
+        if (n > SUBSTR_LIMIT) {
+            return(CORD_substr_closure(x, i, n, CORD_index_access_fn));
+        } else {
+            register char * result = GC_MALLOC_ATOMIC(n+1);
+            register char * p = result;
+            
+            if (result == 0) OUT_OF_MEMORY;
+            strncpy(result, x+i, n);
+            result[n] = '\0';
+            return(result);
+        }
+    } else if (IS_CONCATENATION(x)) {
+    	register struct Concatenation * conc
+    			= &(((CordRep *)x) -> concatenation);
+    	register size_t left_len;
+    	register size_t right_len;
+    	
+    	left_len = LEFT_LEN(conc);
+    	right_len = conc -> len - left_len;
+    	if (i >= left_len) {
+    	    if (n == right_len) return(conc -> right);
+    	    return(CORD_substr_checked(conc -> right, i - left_len, n));
+    	} else if (i+n <= left_len) {
+    	    if (n == left_len) return(conc -> left);
+    	    return(CORD_substr_checked(conc -> left, i, n));
+    	} else {
+    	    /* Need at least one character from each side. */
+    	    register CORD left_part;
+    	    register CORD right_part;
+    	    register size_t left_part_len = left_len - i;
+     	
+    	    if (i == 0) {
+    	        left_part = conc -> left;
+    	    } else {
+    	        left_part = CORD_substr_checked(conc -> left, i, left_part_len);
+    	    }
+    	    if (i + n == right_len + left_len) {
+    	         right_part = conc -> right;
+    	    } else {
+    	         right_part = CORD_substr_checked(conc -> right, 0,
+    	    				          n - left_part_len);
+    	    }
+    	    return(CORD_cat(left_part, right_part));
+    	}
+    } else /* function */ {
+        if (n > SUBSTR_LIMIT) {
+            if (IS_SUBSTR(x)) {
+            	/* Avoid nesting substring nodes.	*/
+            	register struct Function * f = &(((CordRep *)x) -> function);
+            	register struct substr_args *descr =
+            			(struct substr_args *)(f -> client_data);
+            	
+            	return(CORD_substr_closure((CORD)descr->sa_cord,
+            				   i + descr->sa_index,
+            				   n, f -> fn));
+            } else {
+                return(CORD_substr_closure(x, i, n, CORD_apply_access_fn));
+            }
+        } else {
+            char * result;
+            register struct Function * f = &(((CordRep *)x) -> function);
+            char buf[SUBSTR_LIMIT+1];
+            register char * p = buf;
+            register char c;
+            register int j;
+            register int lim = i + n;
+            
+            for (j = i; j < lim; j++) {
+            	c = (*(f -> fn))(j, f -> client_data);
+            	if (c == '\0') {
+            	    return(CORD_substr_closure(x, i, n, CORD_apply_access_fn));
+            	}
+            	*p++ = c;
+            }
+            *p = '\0';
+            result = GC_MALLOC_ATOMIC(n+1);
+            if (result == 0) OUT_OF_MEMORY;
+            strcpy(result, buf);
+            return(result);
+        }
+    }
+}
+
+CORD CORD_substr(CORD x, size_t i, size_t n)
+{
+    register size_t len = CORD_len(x);
+    
+    if (i >= len || n <= 0) return(0);
+    	/* n < 0 is impossible in a correct C implementation, but	*/
+    	/* quite possible  under SunOS 4.X.				*/
+    if (i + n > len) n = len - i;
+    if (i < 0) ABORT("CORD_substr: second arg. negative");
+    	/* Possible only if both client and C implementation are buggy.	*/
+    	/* But empirically this happens frequently.			*/
+    return(CORD_substr_checked(x, i, n));
+}
+
+/* See cord.h for definition.  We assume i is in range.	*/
+int CORD_iter5(CORD x, size_t i, CORD_iter_fn f1,
+			 CORD_batched_iter_fn f2, void * client_data)
+{
+    if (x == 0) return(0);
+    if (IS_STRING(x)) {
+    	register const char *p = x+i;
+    	
+    	if (*p == '\0') ABORT("2nd arg to CORD_iter5 too big");
+        if (f2 != CORD_NO_FN) {
+            return((*f2)(p, client_data));
+        } else {
+	    while (*p) {
+                if ((*f1)(*p, client_data)) return(1);
+                p++;
+	    }
+	    return(0);
+        }
+    } else if (IS_CONCATENATION(x)) {
+    	register struct Concatenation * conc
+    			= &(((CordRep *)x) -> concatenation);
+    	
+    	
+    	if (i > 0) {
+    	    register size_t left_len = LEFT_LEN(conc);
+    	    
+    	    if (i >= left_len) {
+    	        return(CORD_iter5(conc -> right, i - left_len, f1, f2,
+    	        		  client_data));
+    	    }
+    	}
+    	if (CORD_iter5(conc -> left, i, f1, f2, client_data)) {
+    	    return(1);
+    	}
+    	return(CORD_iter5(conc -> right, 0, f1, f2, client_data));
+    } else /* function */ {
+        register struct Function * f = &(((CordRep *)x) -> function);
+        register size_t j;
+        register size_t lim = f -> len;
+        
+        for (j = i; j < lim; j++) {
+            if ((*f1)((*(f -> fn))(j, f -> client_data), client_data)) {
+                return(1);
+            }
+        }
+        return(0);
+    }
+}
+			
+#undef CORD_iter
+int CORD_iter(CORD x, CORD_iter_fn f1, void * client_data)
+{
+    return(CORD_iter5(x, 0, f1, CORD_NO_FN, client_data));
+}
+
+int CORD_riter4(CORD x, size_t i, CORD_iter_fn f1, void * client_data)
+{
+    if (x == 0) return(0);
+    if (IS_STRING(x)) {
+	register const char *p = x + i;
+	register char c;
+               
+	while (p >= x) {
+	    c = *p;
+	    if (c == '\0') ABORT("2nd arg to CORD_riter4 too big");
+            if ((*f1)(c, client_data)) return(1);
+            p--;
+	}
+	return(0);
+    } else if (IS_CONCATENATION(x)) {
+    	register struct Concatenation * conc
+    			= &(((CordRep *)x) -> concatenation);
+    	register CORD left_part = conc -> left;
+    	register size_t left_len;
+    	
+    	left_len = LEFT_LEN(conc);
+    	if (i >= left_len) {
+    	    if (CORD_riter4(conc -> right, i - left_len, f1, client_data)) {
+    	    	return(1);
+    	    }
+    	    return(CORD_riter4(left_part, left_len - 1, f1, client_data));
+    	} else {
+    	    return(CORD_riter4(left_part, i, f1, client_data));
+    	}
+    } else /* function */ {
+        register struct Function * f = &(((CordRep *)x) -> function);
+        register size_t j;
+        
+        for (j = i; j >= 0; j--) {
+            if ((*f1)((*(f -> fn))(j, f -> client_data), client_data)) {
+                return(1);
+            }
+        }
+        return(0);
+    }
+}
+
+int CORD_riter(CORD x, CORD_iter_fn f1, void * client_data)
+{
+    return(CORD_riter4(x, CORD_len(x) - 1, f1, client_data));
+}
+
+/*
+ * The following functions are concerned with balancing cords.
+ * Strategy:
+ * Scan the cord from left to right, keeping the cord scanned so far
+ * as a forest of balanced trees of exponentialy decreasing length.
+ * When a new subtree needs to be added to the forest, we concatenate all
+ * shorter ones to the new tree in the appropriate order, and then insert
+ * the result into the forest.
+ * Crucial invariants:
+ * 1. The concatenation of the forest (in decreasing order) with the
+ *     unscanned part of the rope is equal to the rope being balanced.
+ * 2. All trees in the forest are balanced.
+ * 3. forest[i] has depth at most i.
+ */
+
+typedef struct {
+    CORD c;
+    size_t len;		/* Actual ength of c 	*/
+} ForestElement;
+
+static size_t min_len [ MAX_DEPTH ];
+
+static int min_len_init = 0;
+
+int CORD_max_len;
+
+typedef ForestElement Forest [ MAX_DEPTH ];
+			/* forest[i].min_length = fib(i+1)	*/
+			/* The string is the concatenation	*/
+			/* of the forest in order of DECREASING */
+			/* indices.				*/
+
+void CORD_init_min_len()
+{
+    register int i;
+    register size_t last, previous, current;
+        
+    min_len[0] = previous = 1;
+    min_len[1] = last = 2;
+    for (i = 2; i < MAX_DEPTH; i++) {
+    	current = last + previous;
+    	if (current < last) /* overflow */ current = last;
+    	min_len[i] = current;
+    	previous = last;
+    	last = current;
+    }
+    CORD_max_len = last - 1;
+    min_len_init = 1;
+}
+
+
+void CORD_init_forest(ForestElement * forest, size_t max_len)
+{
+    register int i;
+    
+    for (i = 0; i < MAX_DEPTH; i++) {
+    	forest[i].c = 0;
+    	if (min_len[i] > max_len) return;
+    }
+    ABORT("Cord too long");
+}
+
+/* Add a leaf to the appropriate level in the forest, cleaning		*/
+/* out lower levels as necessary.					*/
+/* Also works if x is a balanced tree of concatenations; however	*/
+/* in this case an extra concatenation node may be inserted above x;	*/
+/* This node should not be counted in the statement of the invariants.	*/
+void CORD_add_forest(ForestElement * forest, CORD x, size_t len)
+{
+    register int i = 0;
+    register CORD sum = CORD_EMPTY;
+    register size_t sum_len = 0;
+    
+    while (len > min_len[i + 1]) {
+    	if (forest[i].c != 0) {
+    	    sum = CORD_cat(forest[i].c, sum);
+    	    sum_len += forest[i].len;
+    	    forest[i].c = 0;
+    	}
+        i++;
+    }
+    /* Sum has depth at most 1 greter than what would be required 	*/
+    /* for balance.							*/
+    sum = CORD_cat(sum, x);
+    sum_len += len;
+    /* If x was a leaf, then sum is now balanced.  To see this		*/
+    /* consider the two cases in whichforest[i-1] either is or is 	*/
+    /* not empty.							*/
+    while (sum_len >= min_len[i]) {
+    	if (forest[i].c != 0) {
+    	    sum = CORD_cat(forest[i].c, sum);
+    	    sum_len += forest[i].len;
+    	    /* This is again balanced, since sum was balanced, and has	*/
+    	    /* allowable depth that differs from i by at most 1.	*/
+    	    forest[i].c = 0;
+    	}
+        i++;
+    }
+    i--;
+    forest[i].c = sum;
+    forest[i].len = sum_len;
+}
+
+CORD CORD_concat_forest(ForestElement * forest, size_t expected_len)
+{
+    register int i = 0;
+    CORD sum = 0;
+    size_t sum_len = 0;
+    
+    while (sum_len != expected_len) {
+    	if (forest[i].c != 0) {
+    	    sum = CORD_cat(forest[i].c, sum);
+    	    sum_len += forest[i].len;
+    	}
+        i++;
+    }
+    return(sum);
+}
+
+/* Insert the frontier of x into forest.  Balanced subtrees are	*/
+/* treated as leaves.  This potentially adds one to the depth	*/
+/* of the final tree.						*/
+void CORD_balance_insert(CORD x, size_t len, ForestElement * forest)
+{
+    register int depth;
+    
+    if (IS_STRING(x)) {
+        CORD_add_forest(forest, x, len);
+    } else if (IS_CONCATENATION(x)
+               && ((depth = DEPTH(x)) >= MAX_DEPTH
+                   || len < min_len[depth])) {
+    	register struct Concatenation * conc
+    			= &(((CordRep *)x) -> concatenation);
+    	size_t left_len = LEFT_LEN(conc);
+    	
+    	CORD_balance_insert(conc -> left, left_len, forest);
+    	CORD_balance_insert(conc -> right, len - left_len, forest);
+    } else /* function or balanced */ {
+    	CORD_add_forest(forest, x, len);
+    }
+}
+
+
+CORD CORD_balance(CORD x)
+{
+    Forest forest;
+    register size_t len;
+    
+    if (x == 0) return(0);
+    if (IS_STRING(x)) return(x);
+    if (!min_len_init) CORD_init_min_len();
+    len = LEN(x);
+    CORD_init_forest(forest, len);
+    CORD_balance_insert(x, len, forest);
+    return(CORD_concat_forest(forest, len));
+}
+
+
+/* Position primitives	*/
+
+/* Private routines to deal with the hard cases only: */
+
+/* P contains a prefix of the  path to cur_pos.	Extend it to a full	*/
+/* path and set up leaf info.						*/
+/* Return 0 if past the end of cord, 1 o.w.				*/
+void CORD__extend_path(register CORD_pos p)
+{
+     register struct CORD_pe * current_pe = &(p[0].path[p[0].path_len]);
+     register CORD top = current_pe -> pe_cord;
+     register size_t pos = p[0].cur_pos;
+     register size_t top_pos = current_pe -> pe_start_pos;
+     register size_t top_len = GEN_LEN(top);
+     
+     /* Fill in the rest of the path. */
+       while(!IS_STRING(top) && IS_CONCATENATION(top)) {
+     	 register struct Concatenation * conc =
+     	 		&(((CordRep *)top) -> concatenation);
+     	 register size_t left_len;
+     	 
+     	 left_len = LEFT_LEN(conc);
+     	 current_pe++;
+     	 if (pos >= top_pos + left_len) {
+     	     current_pe -> pe_cord = top = conc -> right;
+     	     current_pe -> pe_start_pos = top_pos = top_pos + left_len;
+     	     top_len -= left_len;
+     	 } else {
+     	     current_pe -> pe_cord = top = conc -> left;
+     	     current_pe -> pe_start_pos = top_pos;
+     	     top_len = left_len;
+     	 }
+     	 p[0].path_len++;
+       }
+     /* Fill in leaf description for fast access. */
+       if (IS_STRING(top)) {
+         p[0].cur_leaf = top;
+         p[0].cur_start = top_pos;
+         p[0].cur_end = top_pos + top_len;
+       } else {
+         p[0].cur_end = 0;
+       }
+       if (pos >= top_pos + top_len) p[0].path_len = CORD_POS_INVALID;
+}
+
+char CORD__pos_fetch(register CORD_pos p)
+{
+    /* Leaf is a function node */
+    struct CORD_pe * pe = &((p)[0].path[(p)[0].path_len]);
+    CORD leaf = pe -> pe_cord;
+    register struct Function * f = &(((CordRep *)leaf) -> function);
+    
+    if (!IS_FUNCTION(leaf)) ABORT("CORD_pos_fetch: bad leaf");
+    return ((*(f -> fn))(p[0].cur_pos - pe -> pe_start_pos, f -> client_data));
+}
+
+void CORD__next(register CORD_pos p)
+{
+    register size_t cur_pos = p[0].cur_pos + 1;
+    register struct CORD_pe * current_pe = &((p)[0].path[(p)[0].path_len]);
+    register CORD leaf = current_pe -> pe_cord;
+    
+    /* Leaf is not a string or we're at end of leaf */
+    p[0].cur_pos = cur_pos;
+    if (!IS_STRING(leaf)) {
+    	/* Function leaf	*/
+    	register struct Function * f = &(((CordRep *)leaf) -> function);
+    	register size_t start_pos = current_pe -> pe_start_pos;
+    	register size_t end_pos = start_pos + f -> len;
+    	
+    	if (cur_pos < end_pos) {
+    	  /* Fill cache and return. */
+    	    register size_t i;
+    	    register size_t limit = cur_pos + FUNCTION_BUF_SZ;
+    	    register CORD_fn fn = f -> fn;
+    	    register void * client_data = f -> client_data;
+    	    
+    	    if (limit > end_pos) {
+    	        limit = end_pos;
+    	    }
+    	    for (i = cur_pos; i < limit; i++) {
+    	        p[0].function_buf[i - cur_pos] =
+    	        	(*fn)(i - start_pos, client_data);
+    	    }
+    	    p[0].cur_start = cur_pos;
+    	    p[0].cur_leaf = p[0].function_buf;
+    	    p[0].cur_end = limit;
+    	    return;
+    	}
+    }
+    /* End of leaf	*/
+    /* Pop the stack until we find two concatenation nodes with the 	*/
+    /* same start position: this implies we were in left part.		*/
+    {
+    	while (p[0].path_len > 0
+    	       && current_pe[0].pe_start_pos != current_pe[-1].pe_start_pos) {
+    	    p[0].path_len--;
+    	    current_pe--;
+    	}
+    	if (p[0].path_len == 0) {
+	    p[0].path_len = CORD_POS_INVALID;
+            return;
+	}
+    }
+    p[0].path_len--;
+    CORD__extend_path(p);
+}
+
+void CORD__prev(register CORD_pos p)
+{
+    register struct CORD_pe * pe = &(p[0].path[p[0].path_len]);
+    
+    if (p[0].cur_pos == 0) {
+        p[0].path_len = CORD_POS_INVALID;
+        return;
+    }
+    p[0].cur_pos--;
+    if (p[0].cur_pos >= pe -> pe_start_pos) return;
+    
+    /* Beginning of leaf	*/
+    
+    /* Pop the stack until we find two concatenation nodes with the 	*/
+    /* different start position: this implies we were in right part.	*/
+    {
+    	register struct CORD_pe * current_pe = &((p)[0].path[(p)[0].path_len]);
+    	
+    	while (p[0].path_len > 0
+    	       && current_pe[0].pe_start_pos == current_pe[-1].pe_start_pos) {
+    	    p[0].path_len--;
+    	    current_pe--;
+    	}
+    }
+    p[0].path_len--;
+    CORD__extend_path(p);
+}
+
+#undef CORD_pos_fetch
+#undef CORD_next
+#undef CORD_prev
+#undef CORD_pos_to_index
+#undef CORD_pos_to_cord
+#undef CORD_pos_valid
+
+char CORD_pos_fetch(register CORD_pos p)
+{
+    if (p[0].cur_start <= p[0].cur_pos && p[0].cur_pos < p[0].cur_end) {
+    	return(p[0].cur_leaf[p[0].cur_pos - p[0].cur_start]);
+    } else {
+        return(CORD__pos_fetch(p));
+    }
+}
+
+void CORD_next(CORD_pos p)
+{
+    if (p[0].cur_pos < p[0].cur_end - 1) {
+    	p[0].cur_pos++;
+    } else {
+    	CORD__next(p);
+    }
+}
+
+void CORD_prev(CORD_pos p)
+{
+    if (p[0].cur_end != 0 && p[0].cur_pos > p[0].cur_start) {
+    	p[0].cur_pos--;
+    } else {
+    	CORD__prev(p);
+    }
+}
+
+size_t CORD_pos_to_index(CORD_pos p)
+{
+    return(p[0].cur_pos);
+}
+
+CORD CORD_pos_to_cord(CORD_pos p)
+{
+    return(p[0].path[0].pe_cord);
+}
+
+int CORD_pos_valid(CORD_pos p)
+{
+    return(p[0].path_len != CORD_POS_INVALID);
+}
+
+void CORD_set_pos(CORD_pos p, CORD x, size_t i)
+{
+    if (x == CORD_EMPTY) {
+    	p[0].path_len = CORD_POS_INVALID;
+    	return;
+    }
+    p[0].path[0].pe_cord = x;
+    p[0].path[0].pe_start_pos = 0;
+    p[0].path_len = 0;
+    p[0].cur_pos = i;
+    CORD__extend_path(p);
+}
diff --git a/cord/cordprnt.c b/cord/cordprnt.c
new file mode 100644
index 00000000..1b043152
--- /dev/null
+++ b/cord/cordprnt.c
@@ -0,0 +1,388 @@
+/* 
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* An sprintf implementation that understands cords.  This is probably	*/
+/* not terribly portable.  It assumes an ANSI stdarg.h.  It further	*/
+/* assumes that I can make copies of va_list variables, and read 	*/
+/* arguments repeatedly by applyting va_arg to the copies.  This	*/
+/* could be avoided at some performance cost.				*/
+/* We also assume that unsigned and signed integers of various kinds	*/
+/* have the same sizes, and can be cast back and forth.			*/
+/* We assume that void * and char * have the same size.			*/
+/* All this cruft is needed because we want to rely on the underlying	*/
+/* sprintf implementation whenever possible.				*/
+/* Boehm, May 19, 1994 2:19 pm PDT */
+
+#include "cord.h"
+#include "ec.h"
+#include <stdio.h>
+#include <stdarg.h>
+#include <string.h>
+#include "../gc.h"
+
+#define CONV_SPEC_LEN 50	/* Maximum length of a single	*/
+				/* conversion specification.	*/
+#define CONV_RESULT_LEN 50	/* Maximum length of any 	*/
+				/* conversion with default	*/
+				/* width and prec.		*/
+
+
+static int ec_len(CORD_ec x)
+{
+    return(CORD_len(x[0].ec_cord) + (x[0].ec_bufptr - x[0].ec_buf));
+}
+
+/* Possible nonumeric precision values.	*/
+# define NONE -1
+# define VARIABLE -2
+/* Copy the conversion specification from CORD_pos into the buffer buf	*/
+/* Return negative on error.						*/
+/* Source initially points one past the leading %.			*/
+/* It is left pointing at the conversion type.				*/
+/* Assign field width and precision to *width and *prec.		*/
+/* If width or prec is *, VARIABLE is assigned.				*/
+/* Set *left to 1 if left adjustment flag is present.			*/
+/* Set *long_arg to 1 if long flag ('l' or 'L') is present, or to	*/
+/* -1 if 'h' is present.						*/
+static int extract_conv_spec(CORD_pos source, char *buf,
+			     int * width, int *prec, int *left, int * long_arg)
+{
+    register int result = 0;
+    register int current_number = 0;
+    register int saw_period = 0;
+    register int saw_number;
+    register int chars_so_far = 0;
+    register char current;
+    
+    *width = NONE;
+    buf[chars_so_far++] = '%';
+    while(CORD_pos_valid(source)) {
+        if (chars_so_far >= CONV_SPEC_LEN) return(-1);
+        current = CORD_pos_fetch(source);
+        buf[chars_so_far++] = current;
+        switch(current) {
+	  case '*':
+	    saw_number = 1;
+	    current_number = VARIABLE;
+	    break;
+          case '0':
+            if (!saw_number) {
+                /* Zero fill flag; ignore */
+                break;
+            } /* otherwise fall through: */
+          case '1':
+	  case '2':
+	  case '3':
+	  case '4':
+	  case '5':
+          case '6':
+	  case '7':
+	  case '8':
+	  case '9':
+	    saw_number = 1;
+	    current_number *= 10;
+	    current_number += current - '0';
+	    break;
+	  case '.':
+	    saw_period = 1;
+	    if(saw_number) {
+	        *width = current_number;
+	        saw_number = 0;
+	    }
+	    current_number = 0;
+	    break;
+	  case 'l':
+	  case 'L':
+	    *long_arg = 1;
+	    current_number = 0;
+	    break;
+	  case 'h':
+	    *long_arg = -1;
+	    current_number = 0;
+	    break;
+	  case ' ':
+	  case '+':
+	  case '#':
+	    current_number = 0;
+	    break;
+	  case '-':
+	    *left = 1;
+	    current_number = 0;
+	    break;
+	  case 'd':
+	  case 'i':
+	  case 'o':
+	  case 'u':
+	  case 'x':
+	  case 'X':
+	  case 'f':
+	  case 'e':
+	  case 'E':
+	  case 'g':
+	  case 'G':
+	  case 'c':
+	  case 'C':
+	  case 's':
+	  case 'S':
+	  case 'p':
+	  case 'n':
+	  case 'r':
+	    goto done;          
+          default:
+            return(-1);
+        }
+        CORD_next(source);
+    }
+    return(-1);
+  done:
+    if (saw_number) {
+    	if (saw_period) {
+    	    *prec = current_number;
+    	} else {
+    	    *prec = NONE;
+    	    *width = current_number;
+    	}
+    } else {
+    	*prec = NONE;
+    }
+    buf[chars_so_far] = '\0';
+    return(result);
+}
+
+int CORD_vsprintf(CORD * out, CORD format, va_list args)
+{
+    CORD_ec result;
+    register int count;
+    register char current;
+    CORD_pos pos;
+    char conv_spec[CONV_SPEC_LEN + 1];
+    
+    CORD_ec_init(result);
+    for (CORD_set_pos(pos, format, 0); CORD_pos_valid(pos); CORD_next(pos)) {
+       	current = CORD_pos_fetch(pos);
+       	if (current == '%') {
+            CORD_next(pos);
+            if (!CORD_pos_valid(pos)) return(-1);
+            current = CORD_pos_fetch(pos);
+            if (current == '%') {
+               	CORD_ec_append(result, current);
+            } else {
+             	int width, prec;
+             	int left_adj = 0;
+             	int long_arg = 0;
+		CORD arg;
+		size_t len;
+               
+              	if (extract_conv_spec(pos, conv_spec,
+              			      &width, &prec,
+              			      &left_adj, &long_arg) < 0) {
+              	    return(-1);
+              	}
+              	current = CORD_pos_fetch(pos);
+            	switch(current) {
+            	    case 'n':
+            	    	/* Assign length to next arg */
+            	    	if (long_arg == 0) {
+            	    	    int * pos_ptr;
+            	    	    pos_ptr = va_arg(args, int *);
+            	    	    *pos_ptr = ec_len(result);
+            	    	} else if (long_arg > 0) {
+            	    	    long * pos_ptr;
+            	    	    pos_ptr = va_arg(args, long *);
+            	    	    *pos_ptr = ec_len(result);
+            	    	} else {
+            	    	    short * pos_ptr;
+            	    	    pos_ptr = va_arg(args, short *);
+            	    	    *pos_ptr = ec_len(result);
+            	    	}
+            	    	goto done;
+            	    case 'r':
+            	    	/* Append cord and any padding	*/
+            	    	if (width == VARIABLE) width = va_arg(args, int);
+            	    	if (prec == VARIABLE) prec = va_arg(args, int);
+			arg = va_arg(args, CORD);
+			len = CORD_len(arg);
+			if (prec != NONE && len > prec) {
+			  if (prec < 0) return(-1);
+			  arg = CORD_substr(arg, 0, prec);
+			  len = prec;
+			}
+			if (width != NONE && len < width) {
+			  char * blanks = GC_MALLOC_ATOMIC(width-len+1);
+
+			  memset(blanks, ' ', width-len);
+			  blanks[width-len] = '\0';
+			  if (left_adj) {
+			    arg = CORD_cat(arg, blanks);
+			  } else {
+			    arg = CORD_cat(blanks, arg);
+			  }
+			}
+			CORD_ec_append_cord(result, arg);
+            	    	goto done;
+		    case 'c':
+			if (width == NONE && prec == NONE) {
+			    register char c = va_arg(args, char);
+
+			    CORD_ec_append(result, c);
+			    goto done;
+			}
+			break;
+		    case 's':
+		        if (width == NONE && prec == NONE) {
+			    char * str = va_arg(args, char *);
+			    register char c;
+
+			    while (c = *str++) {
+			        CORD_ec_append(result, c);
+			    }
+			    goto done;
+			}
+			break;
+            	    default:
+            	        break;
+            	}
+            	/* Use standard sprintf to perform conversion */
+            	{
+            	    register char * buf;
+            	    int needed_sz;
+            	    va_list vsprintf_args = args;
+            	    	/* The above does not appear to be sanctioned	*/
+            	    	/* by the ANSI C standard.			*/
+            	    int max_size = 0;
+            	    	
+            	    if (width == VARIABLE) width = va_arg(args, int);
+            	    if (prec == VARIABLE) prec = va_arg(args, int);
+            	    if (width != NONE) max_size = width;
+            	    if (prec != NONE && prec > max_size) max_size = prec;
+            	    max_size += CONV_RESULT_LEN;
+            	    if (max_size >= CORD_BUFSZ) {
+            	        buf = GC_MALLOC_ATOMIC(max_size + 1);
+            	    } else {
+            	        if (CORD_BUFSZ - (result[0].ec_bufptr-result[0].ec_buf)
+            	            < max_size) {
+            	            CORD_ec_flush_buf(result);
+            	        }
+            	        buf = result[0].ec_bufptr;
+            	    }
+            	    switch(current) {
+            	        case 'd':
+            	        case 'i':
+            	        case 'o':
+            	        case 'u':
+            	        case 'x':
+            	        case 'X':
+            	        case 'c':
+            	            if (long_arg <= 0) {
+            	              (void) va_arg(args, int);
+            	            } else if (long_arg > 0) {
+            	              (void) va_arg(args, long);
+            	            }
+            	            break;
+            	        case 's':
+            	        case 'p':
+            	            (void) va_arg(args, char *);
+            	            break;
+            	        case 'f':
+            	        case 'e':
+            	        case 'E':
+            	        case 'g':
+            	        case 'G':
+            	            (void) va_arg(args, double);
+            	            break;
+            	        default:
+            	            return(-1);
+            	    }
+            	    len = (size_t)vsprintf(buf, conv_spec, vsprintf_args);
+            	    if ((char *)len == buf) {
+            	    	/* old style vsprintf */
+            	    	len = strlen(buf);
+            	    } else if (len < 0) {
+            	        return(-1);
+            	    }
+            	    if (buf != result[0].ec_bufptr) {
+            	        register char c;
+
+			while (c = *buf++) {
+			    CORD_ec_append(result, c);
+		        }
+		    } else {
+		        result[0].ec_bufptr = buf + len;
+		    }
+            	}
+              done:;
+            }
+        } else {
+            CORD_ec_append(result, current);
+        }
+    }
+    count = ec_len(result);
+    *out = CORD_balance(CORD_ec_to_cord(result));
+    return(count);
+}
+
+int CORD_sprintf(CORD * out, CORD format, ...)
+{
+    va_list args;
+    int result;
+    
+    va_start(args, format);
+    result = CORD_vsprintf(out, format, args);
+    va_end(args);
+    return(result);
+}
+
+int CORD_fprintf(FILE * f, CORD format, ...)
+{
+    va_list args;
+    int result;
+    CORD out;
+    
+    va_start(args, format);
+    result = CORD_vsprintf(&out, format, args);
+    va_end(args);
+    if (result > 0) CORD_put(out, f);
+    return(result);
+}
+
+int CORD_vfprintf(FILE * f, CORD format, va_list args)
+{
+    int result;
+    CORD out;
+    
+    result = CORD_vsprintf(&out, format, args);
+    if (result > 0) CORD_put(out, f);
+    return(result);
+}
+
+int CORD_printf(CORD format, ...)
+{
+    va_list args;
+    int result;
+    CORD out;
+    
+    va_start(args, format);
+    result = CORD_vsprintf(&out, format, args);
+    va_end(args);
+    if (result > 0) CORD_put(out, stdout);
+    return(result);
+}
+
+int CORD_vprintf(CORD format, va_list args)
+{
+    int result;
+    CORD out;
+    
+    result = CORD_vsprintf(&out, format, args);
+    if (result > 0) CORD_put(out, stdout);
+    return(result);
+}
diff --git a/cord/cordtest.c b/cord/cordtest.c
new file mode 100644
index 00000000..cf1c4a45
--- /dev/null
+++ b/cord/cordtest.c
@@ -0,0 +1,218 @@
+/* 
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:21 pm PDT */
+# include "cord.h"
+# include <stdio.h>
+/* This is a very incomplete test of the cord package.  It knows about	*/
+/* a few internals of the package (e.g. when C strings are returned)	*/
+/* that real clients shouldn't rely on.					*/
+
+# define ABORT(string) \
+{ int x = 0; fprintf(stderr, "FAILED: %s\n", string); x = 1 / x; abort(); }
+
+int count;
+
+int test_fn(char c, void * client_data)
+{
+    if (client_data != (void *)13) ABORT("bad client data");
+    if (count < 64*1024+1) {
+        if ((count & 1) == 0) {
+            if (c != 'b') ABORT("bad char");
+        } else {
+            if (c != 'a') ABORT("bad char");
+        }
+        count++;
+        return(0);
+    } else {
+        if (c != 'c') ABORT("bad char");
+        count++;
+        return(1);
+    }
+}
+
+char id_cord_fn(size_t i, void * client_data)
+{
+    return((char)i);
+}
+
+test_basics()
+{
+    CORD x = "ab";
+    register int i;
+    char c;
+    CORD y;
+    CORD_pos p;
+    
+    x = CORD_cat(x,x);
+    if (!IS_STRING(x)) ABORT("short cord should usually be a string");
+    if (strcmp(x, "abab") != 0) ABORT("bad CORD_cat result");
+    
+    for (i = 1; i < 16; i++) {
+        x = CORD_cat(x,x);
+    }
+    x = CORD_cat(x,"c");
+    if (CORD_len(x) != 128*1024+1) ABORT("bad length");
+    
+    count = 0;
+    if (CORD_iter5(x, 64*1024-1, test_fn, CORD_NO_FN, (void *)13) == 0) {
+        ABORT("CORD_iter5 failed");
+    }
+    if (count != 64*1024 + 2) ABORT("CORD_iter5 failed");
+    
+    count = 0;
+    CORD_set_pos(p, x, 64*1024-1);
+    while(CORD_pos_valid(p)) {
+       	(void) test_fn(CORD_pos_fetch(p), (void *)13);
+	CORD_next(p);
+    }
+    if (count != 64*1024 + 2) ABORT("Position based iteration failed");
+    
+    y = CORD_substr(x, 1023, 5);
+    if (!IS_STRING(y)) ABORT("short cord should usually be a string");
+    if (strcmp(y, "babab") != 0) ABORT("bad CORD_substr result");
+    
+    y = CORD_substr(x, 1024, 8);
+    if (!IS_STRING(y)) ABORT("short cord should usually be a string");
+    if (strcmp(y, "abababab") != 0) ABORT("bad CORD_substr result");
+    
+    y = CORD_substr(x, 128*1024-1, 8);
+    if (!IS_STRING(y)) ABORT("short cord should usually be a string");
+    if (strcmp(y, "bc") != 0) ABORT("bad CORD_substr result");
+    
+    x = CORD_balance(x);
+    if (CORD_len(x) != 128*1024+1) ABORT("bad length");
+    
+    count = 0;
+    if (CORD_iter5(x, 64*1024-1, test_fn, CORD_NO_FN, (void *)13) == 0) {
+        ABORT("CORD_iter5 failed");
+    }
+    if (count != 64*1024 + 2) ABORT("CORD_iter5 failed");
+    
+    y = CORD_substr(x, 1023, 5);
+    if (!IS_STRING(y)) ABORT("short cord should usually be a string");
+    if (strcmp(y, "babab") != 0) ABORT("bad CORD_substr result");
+    y = CORD_from_fn(id_cord_fn, 0, 13);
+    i = 0;
+    CORD_set_pos(p, y, i);
+    while(CORD_pos_valid(p)) {
+        c = CORD_pos_fetch(p);
+       	if(c != i) ABORT("Traversal of function node failed");
+	CORD_next(p); i++;
+    }
+    if (i != 13) ABORT("Bad apparent length for function node");
+}
+
+test_extras()
+{
+#   ifdef __OS2__
+#	define FNAME1 "tmp1"
+#	define FNAME2 "tmp2"
+#   else
+#	define FNAME1 "/tmp/cord_test"
+#	define FNAME2 "/tmp/cord_test2"
+#   endif
+    register int i;
+    CORD y = "abcdefghijklmnopqrstuvwxyz0123456789";
+    CORD x = "{}";
+    CORD w, z;
+    FILE *f;
+    FILE *f1a, *f1b, *f2;
+    
+    for (i = 1; i < 100; i++) {
+        x = CORD_cat(x, y);
+    }
+    z = CORD_balance(x);
+    if (CORD_cmp(x,z) != 0) ABORT("balanced string comparison wrong");
+    if (CORD_cmp(x,CORD_cat(z, CORD_nul(13))) >= 0) ABORT("comparison 2");
+    if (CORD_cmp(CORD_cat(x, CORD_nul(13)), z) <= 0) ABORT("comparison 3");
+    if (CORD_cmp(x,CORD_cat(z, "13")) >= 0) ABORT("comparison 4");
+    if ((f = fopen(FNAME1, "w")) == 0) ABORT("open failed");
+    if (CORD_put(z,f) == EOF) ABORT("CORD_put failed");
+    if (fclose(f) == EOF) ABORT("fclose failed");
+    w = CORD_from_file(f1a = fopen(FNAME1, "rb"));
+    if (CORD_len(w) != CORD_len(z)) ABORT("file length wrong");
+    if (CORD_cmp(w,z) != 0) ABORT("file comparison wrong");
+    if (CORD_cmp(CORD_substr(w, 50*36+2, 36), y) != 0)
+    	ABORT("file substr wrong");
+    z = CORD_from_file_lazy(f1b = fopen(FNAME1, "rb"));
+    if (CORD_cmp(w,z) != 0) ABORT("File conversions differ");
+    if (CORD_chr(w, 0, '9') != 37) ABORT("CORD_chr failed 1");
+    if (CORD_chr(w, 3, 'a') != 38) ABORT("CORD_chr failed 2");
+    if (CORD_rchr(w, CORD_len(w) - 1, '}') != 1) ABORT("CORD_rchr failed");
+    x = y;
+    for (i = 1; i < 14; i++) {
+        x = CORD_cat(x,x);
+    }
+    if ((f = fopen(FNAME2, "w")) == 0) ABORT("2nd open failed");
+    if (CORD_put(x,f) == EOF) ABORT("CORD_put failed");
+    if (fclose(f) == EOF) ABORT("fclose failed");
+    w = CORD_from_file(f2 = fopen(FNAME2, "rb"));
+    if (CORD_len(w) != CORD_len(x)) ABORT("file length wrong");
+    if (CORD_cmp(w,x) != 0) ABORT("file comparison wrong");
+    if (CORD_cmp(CORD_substr(w, 1000*36, 36), y) != 0)
+    	ABORT("file substr wrong");
+    if (strcmp(CORD_to_char_star(CORD_substr(w, 1000*36, 36)), y) != 0)
+    	ABORT("char * file substr wrong");
+    if (strcmp(CORD_substr(w, 1000*36, 2), "ab") != 0)
+    	ABORT("short file substr wrong");
+    if (CORD_str(x,1,"9a") != 35) ABORT("CORD_str failed 1");
+    if (CORD_str(x,0,"9abcdefghijk") != 35) ABORT("CORD_str failed 2");
+    if (CORD_str(x,0,"9abcdefghijx") != CORD_NOT_FOUND)
+    	ABORT("CORD_str failed 3");
+    if (CORD_str(x,0,"9>") != CORD_NOT_FOUND) ABORT("CORD_str failed 4");
+    if (remove(FNAME1) != 0) {
+    	/* On some systems, e.g. OS2, this may fail if f1 is still open. */
+    	if ((fclose(f1a) == EOF) & (fclose(f1b) == EOF))
+    		ABORT("fclose(f1) failed");
+    	if (remove(FNAME1) != 0) ABORT("remove 1 failed");
+    }
+    if (remove(FNAME2) != 0) {
+    	if (fclose(f2) == EOF) ABORT("fclose(f2) failed");
+    	if (remove(FNAME2) != 0) ABORT("remove 2 failed");
+    }
+}
+
+test_printf()
+{
+    CORD result;
+    char result2[200];
+    long l;
+    short s;
+    CORD x;
+    
+    if (CORD_sprintf(&result, "%7.2f%ln", 3.14159, &l) != 7)
+    	ABORT("CORD_sprintf failed 1");
+    if (CORD_cmp(result, "   3.14") != 0)ABORT("CORD_sprintf goofed 1");
+    if (l != 7) ABORT("CORD_sprintf goofed 2");
+    if (CORD_sprintf(&result, "%-7.2s%hn%c%s", "abcd", &s, 'x', "yz") != 10)
+    	ABORT("CORD_sprintf failed 2");
+    if (CORD_cmp(result, "ab     xyz") != 0)ABORT("CORD_sprintf goofed 3");
+    if (s != 7) ABORT("CORD_sprintf goofed 4");
+    x = "abcdefghij";
+    x = CORD_cat(x,x);
+    x = CORD_cat(x,x);
+    x = CORD_cat(x,x);
+    if (CORD_sprintf(&result, "->%-120.78r!\n", x) != 124)
+    	ABORT("CORD_sprintf failed 3");
+    (void) sprintf(result2, "->%-120.78s!\n", CORD_to_char_star(x));
+    if (CORD_cmp(result, result2) != 0)ABORT("CORD_sprintf goofed 5");
+}
+
+main()
+{
+    test_basics();
+    test_extras();
+    test_printf();
+    CORD_fprintf(stderr, "SUCCEEDED\n");
+    return(0);
+}
diff --git a/cord/cordxtra.c b/cord/cordxtra.c
new file mode 100644
index 00000000..4aaaf6e7
--- /dev/null
+++ b/cord/cordxtra.c
@@ -0,0 +1,566 @@
+/*
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ * Author: Hans-J. Boehm (boehm@parc.xerox.com)
+ */
+/*
+ * These are functions on cords that do not need to understand their
+ * implementation.  They serve also serve as example client code for
+ * cord_basics.
+ */
+/* Boehm, May 19, 1994 2:18 pm PDT */
+# include <stdio.h>
+# include <string.h>
+# include <stdlib.h>
+# include "cord.h"
+# include "ec.h"
+# define I_HIDE_POINTERS	/* So we get access to allocation lock.	*/
+				/* We use this for lazy file reading, 	*/
+				/* so that we remain independent 	*/
+				/* of the threads primitives.		*/
+# include "../gc.h"
+
+/* The standard says these are in stdio.h, but they aren't always: */
+# ifndef SEEK_SET
+#   define SEEK_SET 0
+# endif
+# ifndef SEEK_END
+#   define SEEK_END 2
+# endif
+
+# define BUFSZ 2048	/* Size of stack allocated buffers when		*/
+			/* we want large buffers.			*/
+
+typedef void (* oom_fn)(void);
+
+# define OUT_OF_MEMORY {  if (CORD_oom_fn != (oom_fn) 0) (*CORD_oom_fn)(); \
+			  ABORT("Out of memory\n"); }
+# define ABORT(msg) { fprintf(stderr, "%s\n", msg); abort(); }
+
+CORD CORD_cat_char(CORD x, char c)
+{
+    register char * string;
+    
+    if (c == '\0') return(CORD_cat(x, CORD_nul(1)));
+    string = GC_MALLOC_ATOMIC(2);
+    if (string == 0) OUT_OF_MEMORY;
+    string[0] = c;
+    string[1] = '\0';
+    return(CORD_cat_char_star(x, string, 1));
+}
+
+typedef struct {
+	size_t len;
+	size_t count;
+	char * buf;
+} CORD_fill_data;
+
+int CORD_fill_proc(char c, void * client_data)
+{
+    register CORD_fill_data * d = (CORD_fill_data *)client_data;
+    register size_t count = d -> count;
+    
+    (d -> buf)[count] = c;
+    d -> count = ++count;
+    if (count >= d -> len) {
+    	return(1);
+    } else {
+    	return(0);
+    }
+}
+
+int CORD_batched_fill_proc(const char * s, void * client_data)
+{
+    register CORD_fill_data * d = (CORD_fill_data *)client_data;
+    register size_t count = d -> count;
+    register size_t max = d -> len;
+    register char * buf = d -> buf;
+    register const char * t = s;
+    
+    while(((d -> buf)[count] = *t++) != '\0') {
+        count++;
+        if (count >= max) {
+            d -> count = count;
+            return(1);
+        }
+    }
+    d -> count = count;
+    return(0);
+}
+
+/* Fill buf with between min and max characters starting at i.  	*/
+/* Assumes len characters are available.				*/ 
+void CORD_fill_buf(CORD x, size_t i, size_t len, char * buf)
+{
+    CORD_fill_data fd;
+    
+    fd.len = len;
+    fd.buf = buf;
+    fd.count = 0;
+    (void)CORD_iter5(x, i, CORD_fill_proc, CORD_batched_fill_proc, &fd);
+}
+
+int CORD_cmp(CORD x, CORD y)
+{
+    CORD_pos xpos;
+    CORD_pos ypos;
+    register size_t avail, yavail;
+    
+    if (y == CORD_EMPTY) return(x != CORD_EMPTY);
+    if (x == CORD_EMPTY) return(-1);
+    if (IS_STRING(y) && IS_STRING(x)) return(strcmp(x,y));
+    CORD_set_pos(xpos, x, 0);
+    CORD_set_pos(ypos, y, 0);
+    for(;;) {
+        if (!CORD_pos_valid(xpos)) {
+            if (CORD_pos_valid(ypos)) {
+            	return(-1);
+            } else {
+                return(0);
+            }
+        }
+        if (!CORD_pos_valid(ypos)) {
+            return(1);
+        }
+        if ((avail = CORD_pos_chars_left(xpos)) <= 0
+            || (yavail = CORD_pos_chars_left(ypos)) <= 0) {
+            register char xcurrent = CORD_pos_fetch(xpos);
+            register char ycurrent = CORD_pos_fetch(ypos);
+            if (xcurrent != ycurrent) return(xcurrent - ycurrent);
+            CORD_next(xpos);
+            CORD_next(ypos);
+        } else {
+            /* process as many characters as we can	*/
+            register int result;
+            
+            if (avail > yavail) avail = yavail;
+            result = strncmp(CORD_pos_cur_char_addr(xpos),
+            		     CORD_pos_cur_char_addr(ypos), avail);
+            if (result != 0) return(result);
+            CORD_pos_advance(xpos, avail);
+            CORD_pos_advance(ypos, avail);
+        }
+    }
+}
+
+int CORD_ncmp(CORD x, size_t x_start, CORD y, size_t y_start, size_t len)
+{
+    CORD_pos xpos;
+    CORD_pos ypos;
+    register size_t count;
+    register long avail, yavail;
+    
+    CORD_set_pos(xpos, x, x_start);
+    CORD_set_pos(ypos, y, y_start);
+    for(count = 0; count < len;) {
+        if (!CORD_pos_valid(xpos)) {
+            if (CORD_pos_valid(ypos)) {
+            	return(-1);
+            } else {
+                return(0);
+            }
+        }
+        if (!CORD_pos_valid(ypos)) {
+            return(1);
+        }
+        if ((avail = CORD_pos_chars_left(xpos)) <= 0
+            || (yavail = CORD_pos_chars_left(ypos)) <= 0) {
+            register char xcurrent = CORD_pos_fetch(xpos);
+            register char ycurrent = CORD_pos_fetch(ypos);
+            if (xcurrent != ycurrent) return(xcurrent - ycurrent);
+            CORD_next(xpos);
+            CORD_next(ypos);
+            count++;
+        } else {
+            /* process as many characters as we can	*/
+            register int result;
+            
+            if (avail > yavail) avail = yavail;
+            count += avail;
+            if (count > len) avail -= (count - len);
+            result = strncmp(CORD_pos_cur_char_addr(xpos),
+            		     CORD_pos_cur_char_addr(ypos), (size_t)avail);
+            if (result != 0) return(result);
+            CORD_pos_advance(xpos, (size_t)avail);
+            CORD_pos_advance(ypos, (size_t)avail);
+        }
+    }
+    return(0);
+}
+
+char * CORD_to_char_star(CORD x)
+{
+    register size_t len;
+    char * result;
+    
+    if (x == 0) return("");
+    len = CORD_len(x);
+    result = (char *)GC_MALLOC_ATOMIC(len + 1);
+    if (result == 0) OUT_OF_MEMORY;
+    CORD_fill_buf(x, 0, len, result);
+    result[len] = '\0';
+    return(result);
+}
+
+char CORD_fetch(CORD x, size_t i)
+{
+    CORD_pos xpos;
+    
+    CORD_set_pos(xpos, x, i);
+    if (!CORD_pos_valid(xpos)) ABORT("bad index?");
+    return(CORD_pos_fetch(xpos));
+}
+
+
+int CORD_put_proc(char c, void * client_data)
+{
+    register FILE * f = (FILE *)client_data;
+    
+    return(putc(c, f) == EOF);
+}
+
+int CORD_batched_put_proc(const char * s, void * client_data)
+{
+    register FILE * f = (FILE *)client_data;
+    
+    return(fputs(s, f) == EOF);
+}
+    
+
+int CORD_put(CORD x, FILE * f)
+{
+    if (CORD_iter5(x, 0, CORD_put_proc, CORD_batched_put_proc, f)) {
+        return(EOF);
+    } else {
+    	return(1);
+    }
+}
+
+typedef struct {
+    size_t pos;		/* Current position in the cord */
+    char target;	/* Character we're looking for	*/
+} chr_data;
+
+int CORD_chr_proc(char c, void * client_data)
+{
+    register chr_data * d = (chr_data *)client_data;
+    
+    if (c == d -> target) return(1);
+    (d -> pos) ++;
+    return(0);
+}
+
+int CORD_rchr_proc(char c, void * client_data)
+{
+    register chr_data * d = (chr_data *)client_data;
+    
+    if (c == d -> target) return(1);
+    (d -> pos) --;
+    return(0);
+}
+
+int CORD_batched_chr_proc(const char *s, void * client_data)
+{
+    register chr_data * d = (chr_data *)client_data;
+    register char * occ = strchr(s, d -> target);
+    
+    if (occ == 0) {
+      	d -> pos += strlen(s);
+      	return(0);
+    } else {
+    	d -> pos += occ - s;
+    	return(1);
+    }
+}
+
+size_t CORD_chr(CORD x, size_t i, int c)
+{
+    chr_data d;
+    
+    d.pos = i;
+    d.target = c;
+    if (CORD_iter5(x, i, CORD_chr_proc, CORD_batched_chr_proc, &d)) {
+        return(d.pos);
+    } else {
+    	return(CORD_NOT_FOUND);
+    }
+}
+
+size_t CORD_rchr(CORD x, size_t i, int c)
+{
+    chr_data d;
+    
+    d.pos = i;
+    d.target = c;
+    if (CORD_riter4(x, i, CORD_rchr_proc, &d)) {
+        return(d.pos);
+    } else {
+    	return(CORD_NOT_FOUND);
+    }
+}
+
+/* Find the first occurrence of s in x at position start or later.	*/
+/* This uses an asymptotically poor algorithm, which should typically 	*/
+/* perform acceptably.  We compare the first few characters directly,	*/
+/* and call CORD_ncmp whenever there is a partial match.		*/
+/* This has the advantage that we allocate very little, or not at all.	*/
+/* It's very fast if there are few close misses.			*/
+size_t CORD_str(CORD x, size_t start, CORD s)
+{
+    CORD_pos xpos;
+    size_t xlen = CORD_len(x);
+    size_t slen;
+    register size_t start_len;
+    const char * s_start;
+    unsigned long s_buf = 0;	/* The first few characters of s	*/
+    unsigned long x_buf = 0;	/* Start of candidate substring.	*/
+    				/* Initialized only to make compilers	*/
+    				/* happy.				*/
+    unsigned long mask = 0;
+    register size_t i;
+    register size_t match_pos;
+    
+    if (s == CORD_EMPTY) return(start);
+    if (IS_STRING(s)) {
+        s_start = s;
+        slen = strlen(s);
+    } else {
+        s_start = CORD_to_char_star(CORD_substr(s, 0, sizeof(unsigned long)));
+        slen = CORD_len(s);
+    }
+    if (xlen < start || xlen - start < slen) return(CORD_NOT_FOUND);
+    start_len = slen;
+    if (start_len > sizeof(unsigned long)) start_len = sizeof(unsigned long);
+    CORD_set_pos(xpos, x, start);
+    for (i = 0; i < start_len; i++) {
+        mask <<= 8;
+        mask |= 0xff;
+        s_buf <<= 8;
+        s_buf |= s_start[i];
+        x_buf <<= 8;
+        x_buf |= CORD_pos_fetch(xpos);
+        CORD_next(xpos);
+    }
+    for (match_pos = start; match_pos < xlen - slen; match_pos++) {
+    	if ((x_buf & mask) == s_buf) {
+    	    if (slen == start_len ||
+    	     	CORD_ncmp(x, match_pos + start_len,
+    	     	 	  s, start_len, slen - start_len) == 0) {
+    	        return(match_pos);
+    	    }
+    	}
+    	x_buf <<= 8;
+        x_buf |= CORD_pos_fetch(xpos);
+        CORD_next(xpos);
+    }
+    return(CORD_NOT_FOUND);
+}
+
+void CORD_ec_flush_buf(CORD_ec x)
+{
+    register size_t len = x[0].ec_bufptr - x[0].ec_buf;
+    char * s;
+
+    if (len == 0) return;
+    s = GC_MALLOC_ATOMIC(len+1);
+    memcpy(s, x[0].ec_buf, len);
+    s[len] = '\0';
+    x[0].ec_cord = CORD_cat_char_star(x[0].ec_cord, s, len);
+    x[0].ec_bufptr = x[0].ec_buf;
+}
+
+void CORD_ec_append_cord(CORD_ec x, CORD s)
+{
+    CORD_ec_flush_buf(x);
+    x[0].ec_cord = CORD_cat(x[0].ec_cord, s);
+}
+
+/*ARGSUSED*/
+char CORD_nul_func(size_t i, void * client_data)
+{
+    return((char)(unsigned long)client_data);
+}
+
+
+CORD CORD_chars(char c, size_t i)
+{
+    return(CORD_from_fn(CORD_nul_func, (void *)(unsigned long)c, i));
+}
+
+CORD CORD_from_file_eager(FILE * f)
+{
+    register int c;
+    CORD_ec ecord;
+    
+    CORD_ec_init(ecord);
+    for(;;) {
+        c = getc(f);
+        if (c == 0) {
+          /* Append the right number of NULs	*/
+          /* Note that any string of NULs is rpresented in 4 words,	*/
+          /* independent of its length.					*/
+            register size_t count = 1;
+            
+            CORD_ec_flush_buf(ecord);
+            while ((c = getc(f)) == 0) count++;
+            ecord[0].ec_cord = CORD_cat(ecord[0].ec_cord, CORD_nul(count));
+        }
+        if (c == EOF) break;
+        CORD_ec_append(ecord, c);
+    }
+    (void) fclose(f);
+    return(CORD_balance(CORD_ec_to_cord(ecord)));
+}
+
+/* The state maintained for a lazily read file consists primarily	*/
+/* of a large direct-mapped cache of previously read values.		*/
+/* We could rely more on stdio buffering.  That would have 2		*/
+/* disadvantages:							*/
+/*  	1) Empirically, not all fseek implementations preserve the	*/
+/*	   buffer whenever they could.					*/
+/*	2) It would fail if 2 different sections of a long cord		*/
+/*	   were being read alternately.					*/
+/* We do use the stdio buffer for read ahead.				*/
+/* To guarantee thread safety in the presence of atomic pointer		*/
+/* writes, cache lines are always replaced, and never modified in	*/
+/* place.								*/
+
+# define LOG_CACHE_SZ 14
+# define CACHE_SZ (1 << LOG_CACHE_SZ)
+# define LOG_LINE_SZ 9
+# define LINE_SZ (1 << LOG_LINE_SZ)
+
+typedef struct {
+    size_t tag;
+    char data[LINE_SZ];
+    	/* data[i%LINE_SZ] = ith char in file if tag = i/LINE_SZ	*/
+} cache_line;
+
+typedef struct {
+    FILE * lf_file;
+    size_t lf_current;	/* Current file pointer value */
+    cache_line * volatile lf_cache[CACHE_SZ/LINE_SZ];
+} lf_state;
+
+# define MOD_CACHE_SZ(n) ((n) & (CACHE_SZ - 1))
+# define DIV_CACHE_SZ(n) ((n) >> LOG_CACHE_SZ)
+# define MOD_LINE_SZ(n) ((n) & (LINE_SZ - 1))
+# define DIV_LINE_SZ(n) ((n) >> LOG_LINE_SZ)
+# define LINE_START(n) ((n) & ~(LINE_SZ - 1))
+
+typedef struct {
+    lf_state * state;
+    size_t file_pos;	/* Position of needed character. */
+    cache_line * new_cache;
+} refill_data;
+
+/* Executed with allocation lock. */
+static char refill_cache(client_data)
+refill_data * client_data;
+{
+    register lf_state * state = client_data -> state;
+    register size_t file_pos = client_data -> file_pos;
+    FILE *f = state -> lf_file;
+    size_t line_start = LINE_START(file_pos);
+    size_t line_no = DIV_LINE_SZ(MOD_CACHE_SZ(file_pos));
+    cache_line * new_cache = client_data -> new_cache;
+    
+    if (line_start != state -> lf_current
+        && fseek(f, line_start, SEEK_SET) != 0) {
+    	    ABORT("fseek failed");
+    }
+    if (fread(new_cache -> data, sizeof(char), LINE_SZ, f)
+    	<= file_pos - line_start) {
+    	ABORT("fread failed");
+    }
+    new_cache -> tag = DIV_LINE_SZ(file_pos);
+    /* Store barrier goes here. */
+    state -> lf_cache[line_no] = new_cache;
+    state -> lf_current = line_start + LINE_SZ;
+    return(new_cache->data[MOD_LINE_SZ(file_pos)]);
+}
+
+char CORD_lf_func(size_t i, void * client_data)
+{
+    register lf_state * state = (lf_state *)client_data;
+    register cache_line * cl = state -> lf_cache[DIV_LINE_SZ(MOD_CACHE_SZ(i))];
+    
+    if (cl == 0 || cl -> tag != DIV_LINE_SZ(i)) {
+    	/* Cache miss */
+    	refill_data rd;
+    	
+        rd.state = state;
+        rd.file_pos =  i;
+        rd.new_cache = GC_NEW_ATOMIC(cache_line);
+        if (rd.new_cache == 0) OUT_OF_MEMORY;
+        return((char)(GC_word)
+        	  GC_call_with_alloc_lock((GC_fn_type) refill_cache, &rd));
+    }
+    return(cl -> data[MOD_LINE_SZ(i)]);
+}    
+
+/*ARGSUSED*/
+void CORD_lf_close_proc(void * obj, void * client_data)  
+{
+    if (fclose(((lf_state *)obj) -> lf_file) != 0) {
+    	ABORT("CORD_lf_close_proc: fclose failed");
+    }
+}			
+
+CORD CORD_from_file_lazy_inner(FILE * f, size_t len)
+{
+    register lf_state * state = GC_NEW(lf_state);
+    register int i;
+    
+    if (state == 0) OUT_OF_MEMORY;
+    state -> lf_file = f;
+    for (i = 0; i < CACHE_SZ/LINE_SZ; i++) {
+        state -> lf_cache[i] = 0;
+    }
+    state -> lf_current = 0;
+    GC_register_finalizer(state, CORD_lf_close_proc, 0, 0, 0);
+    return(CORD_from_fn(CORD_lf_func, state, len));
+}
+
+CORD CORD_from_file_lazy(FILE * f)
+{
+    register size_t len;
+    
+    if (fseek(f, 0l, SEEK_END) != 0) {
+        ABORT("Bad fd argument - fseek failed");
+    }
+    if ((len = ftell(f)) < 0) {
+        ABORT("Bad fd argument - ftell failed");
+    }
+    rewind(f);
+    return(CORD_from_file_lazy_inner(f, len));
+}
+
+# define LAZY_THRESHOLD (128*1024 + 1)
+
+CORD CORD_from_file(FILE * f)
+{
+    register size_t len;
+    
+    if (fseek(f, 0l, SEEK_END) != 0) {
+        ABORT("Bad fd argument - fseek failed");
+    }
+    if ((len = ftell(f)) < 0) {
+        ABORT("Bad fd argument - ftell failed");
+    }
+    rewind(f);
+    if (len < LAZY_THRESHOLD) {
+        return(CORD_from_file_eager(f));
+    } else {
+        return(CORD_from_file_lazy_inner(f, len));
+    }
+}
diff --git a/cord/de.c b/cord/de.c
new file mode 100644
index 00000000..c2cad50a
--- /dev/null
+++ b/cord/de.c
@@ -0,0 +1,543 @@
+/*
+ * Copyright (c) 1993-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ * Author: Hans-J. Boehm (boehm@parc.xerox.com)
+ */
+/*
+ * A really simple-minded text editor based on cords.
+ * Things it does right:
+ * 	No size bounds.
+ *	Inbounded undo.
+ *	Shouldn't crash no matter what file you invoke it on (e.g. /vmunix)
+ *		(Make sure /vmunix is not writable before you try this.)
+ *	Scrolls horizontally.
+ * Things it does wrong:
+ *	It doesn't handle tabs reasonably (use "expand" first).
+ *	The command set is MUCH too small.
+ *	The redisplay algorithm doesn't let curses do the scrolling.
+ *	The rule for moving the window over the file is suboptimal.
+ */
+/* Boehm, May 19, 1994 2:20 pm PDT */
+#include <stdio.h>
+#include "../gc.h"
+#include "cord.h"
+#ifdef WIN32
+#  include <windows.h>
+#  include "de_win.h"
+#else
+#  include <curses.h>
+#  define de_error(s) { fprintf(stderr, s); sleep(2); }
+#endif
+#include "de_cmds.h"
+
+
+/* List of line number to position mappings, in descending order. */
+/* There may be holes.						  */
+typedef struct LineMapRep {
+    int line;
+    size_t pos;
+    struct LineMapRep * previous;
+} * line_map;
+
+/* List of file versions, one per edit operation */
+typedef struct HistoryRep {
+    CORD file_contents;
+    struct HistoryRep * previous;
+    line_map map;	/* Invalid for first record "now" */
+} * history;
+
+history now = 0;
+CORD current;		/* == now -> file_contents.	*/
+size_t current_len;	/* Current file length.		*/
+line_map current_map = 0;	/* Current line no. to pos. map	 */
+size_t current_map_size = 0;	/* Number of current_map entries.	*/
+				/* Not always accurate, but reset	*/
+				/* by prune_map.			*/
+# define MAX_MAP_SIZE 3000
+
+/* Current display position */
+int dis_line = 0;
+int dis_col = 0;
+
+# define ALL -1
+# define NONE - 2
+int need_redisplay = 0;	/* Line that needs to be redisplayed.	*/
+
+
+/* Current cursor position. Always within file. */
+int line = 0; 
+int col = 0;
+size_t file_pos = 0;	/* Character position corresponding to cursor.	*/
+
+/* Invalidate line map for lines > i */
+void invalidate_map(int i)
+{
+    while(current_map -> line > i) {
+        current_map = current_map -> previous;
+        current_map_size--;
+    }
+}
+
+/* Reduce the number of map entries to save space for huge files. */
+/* This also affects maps in histories.				  */
+void prune_map()
+{
+    line_map map = current_map;
+    int start_line = map -> line;
+    
+    current_map_size = 0;
+    for(; map != 0; map = map -> previous) {
+    	current_map_size++;
+    	if (map -> line < start_line - LINES && map -> previous != 0) {
+    	    map -> previous = map -> previous -> previous;
+    	}
+    }
+}
+/* Add mapping entry */
+void add_map(int line, size_t pos)
+{
+    line_map new_map = GC_NEW(struct LineMapRep);
+    
+    if (current_map_size >= MAX_MAP_SIZE) prune_map();
+    new_map -> line = line;
+    new_map -> pos = pos;
+    new_map -> previous = current_map;
+    current_map = new_map;
+    current_map_size++;
+}
+
+
+
+/* Return position of column *c of ith line in   */
+/* current file. Adjust *c to be within the line.*/
+/* A 0 pointer is taken as 0 column.		 */
+/* Returns CORD_NOT_FOUND if i is too big.	 */
+/* Assumes i > dis_line.			 */
+size_t line_pos(int i, int *c)
+{
+    int j;
+    size_t cur;
+    size_t next;
+    line_map map = current_map;
+    
+    while (map -> line > i) map = map -> previous;
+    if (map -> line < i - 2) /* rebuild */ invalidate_map(i);
+    for (j = map -> line, cur = map -> pos; j < i;) {
+	cur = CORD_chr(current, cur, '\n');
+        if (cur == current_len-1) return(CORD_NOT_FOUND);
+        cur++;
+        if (++j > current_map -> line) add_map(j, cur);
+    }
+    if (c != 0) {
+        next = CORD_chr(current, cur, '\n');
+        if (next == CORD_NOT_FOUND) next = current_len - 1;
+        if (next < cur + *c) {
+            *c = next - cur;
+        }
+        cur += *c;
+    }
+    return(cur);
+}
+
+void add_hist(CORD s)
+{
+    history new_file = GC_NEW(struct HistoryRep);
+    
+    new_file -> file_contents = current = s;
+    current_len = CORD_len(s);
+    new_file -> previous = now;
+    if (now != 0) now -> map = current_map;
+    now = new_file;
+}
+
+void del_hist(void)
+{
+    now = now -> previous;
+    current = now -> file_contents;
+    current_map = now -> map;
+    current_len = CORD_len(current);
+}
+
+/* Current screen_contents; a dynamically allocated array of CORDs	*/
+CORD * screen = 0;
+int screen_size = 0;
+
+# ifndef WIN32
+/* Replace a line in the curses stdscr.	All control characters are	*/
+/* displayed as upper case characters in standout mode.  This isn't	*/
+/* terribly appropriate for tabs.					*/
+void replace_line(int i, CORD s)
+{
+    register int c;
+    CORD_pos p;
+    
+    if (screen == 0 || LINES > screen_size) {
+        screen_size = LINES;
+    	screen = (CORD *)GC_MALLOC(screen_size * sizeof(CORD));
+    }
+    if (CORD_cmp(screen[i], s) != 0) {
+        move(i,0); clrtoeol();
+        /* A gross workaround for an apparent curses bug: */
+            if (i == LINES-1) s = CORD_substr(s, 0, CORD_len(s) - 1);
+        CORD_FOR (p, s) {
+            c = CORD_pos_fetch(p) & 0x7f;
+            if (iscntrl(c)) {
+            	standout(); addch(c + 0x40); standend();
+            } else {
+    	        addch(c);
+    	    }
+    	}
+    	screen[i] = s;
+    }
+}
+#else
+# define replace_line(i,s) invalidate_line(i)
+#endif
+
+/* Return up to COLS characters of the line of s starting at pos,	*/
+/* returning only characters after the given column.			*/
+CORD retrieve_line(CORD s, size_t pos, unsigned column)
+{
+    CORD candidate = CORD_substr(s, pos, column + COLS);
+    			/* avoids scanning very long lines	*/
+    int eol = CORD_chr(candidate, 0, '\n');
+    int len;
+    
+    if (eol == CORD_NOT_FOUND) eol = CORD_len(candidate);
+    len = (int)eol - (int)column;
+    if (len < 0) len = 0;
+    return(CORD_substr(s, pos + column, len));
+}
+
+# ifdef WIN32
+#   define refresh();
+
+    CORD retrieve_screen_line(int i)
+    {
+    	register size_t pos;
+    	
+    	invalidate_map(dis_line + LINES);	/* Prune search */
+    	pos = line_pos(dis_line + i, 0);
+    	if (pos == CORD_NOT_FOUND) return(CORD_EMPTY);
+    	return(retrieve_line(current, pos, dis_col));
+    }
+# endif
+
+/* Display the visible section of the current file	 */
+void redisplay(void)
+{
+    register int i;
+    
+    invalidate_map(dis_line + LINES);	/* Prune search */
+    for (i = 0; i < LINES; i++) {
+        if (need_redisplay == ALL || need_redisplay == i) {
+            register size_t pos = line_pos(dis_line + i, 0);
+            
+            if (pos == CORD_NOT_FOUND) break;
+            replace_line(i, retrieve_line(current, pos, dis_col));
+            if (need_redisplay == i) goto done;
+        }
+    }
+    for (; i < LINES; i++) replace_line(i, CORD_EMPTY);
+done:
+    refresh();
+    need_redisplay = NONE;
+}
+
+int dis_granularity;
+
+/* Update dis_line, dis_col, and dis_pos to make cursor visible.	*/
+/* Assumes line, col, dis_line, dis_pos are in bounds.			*/
+void normalize_display()
+{
+    int old_line = dis_line;
+    int old_col = dis_col;
+    
+    dis_granularity = 1;
+    if (LINES > 15 && COLS > 15) dis_granularity = 5;
+    while (dis_line > line) dis_line -= dis_granularity;
+    while (dis_col > col) dis_col -= dis_granularity;
+    while (line >= dis_line + LINES) dis_line += dis_granularity;
+    while (col >= dis_col + COLS) dis_col += dis_granularity;
+    if (old_line != dis_line || old_col != dis_col) {
+        need_redisplay = ALL;
+    }
+}
+
+# ifndef WIN32
+#   define move_cursor(x,y) move(y,x)
+# endif
+
+/* Adjust display so that cursor is visible; move cursor into position	*/
+/* Update screen if necessary.						*/
+void fix_cursor(void)
+{
+    normalize_display();
+    if (need_redisplay != NONE) redisplay();
+    move_cursor(col - dis_col, line - dis_line);
+    refresh();
+#   ifndef WIN32
+      fflush(stdout);
+#   endif
+}
+
+/* Make sure line, col, and dis_pos are somewhere inside file.	*/
+/* Recompute file_pos.	Assumes dis_pos is accurate or past eof	*/
+void fix_pos()
+{
+    int my_col = col;
+    
+    if ((size_t)line > current_len) line = current_len;
+    file_pos = line_pos(line, &my_col);
+    if (file_pos == CORD_NOT_FOUND) {
+        for (line = current_map -> line, file_pos = current_map -> pos;
+             file_pos < current_len;
+             line++, file_pos = CORD_chr(current, file_pos, '\n') + 1);
+    	line--;
+        file_pos = line_pos(line, &col);
+    } else {
+    	col = my_col;
+    }
+}
+
+#ifndef WIN32
+/*
+ * beep() is part of some curses packages and not others.
+ * We try to match the type of the builtin one, if any.
+ */
+#ifdef __STDC__
+    int beep(void)
+#else
+    int beep()
+#endif
+{
+    putc('\007', stderr);
+    return(0);
+}
+#else
+#  define beep() Beep(1000 /* Hz */, 300 /* msecs */) 
+#endif
+
+#   define NO_PREFIX -1
+#   define BARE_PREFIX -2
+int repeat_count = NO_PREFIX;	/* Current command prefix. */
+
+int locate_mode = 0;			/* Currently between 2 ^Ls	*/
+CORD locate_string = CORD_EMPTY;	/* Current search string.	*/
+
+char * arg_file_name;
+
+#ifdef WIN32
+/* Change the current position to whatever is currently displayed at	*/
+/* the given SCREEN coordinates.					*/
+void set_position(int c, int l)
+{
+    line = l + dis_line;
+    col = c + dis_col;
+    fix_pos();
+    move_cursor(col - dis_col, line - dis_line);
+}
+#endif /* WIN32 */
+
+/* Perform the command associated with character c.  C may be an	*/
+/* integer > 256 denoting a windows command, one of the above control	*/
+/* characters, or another ASCII character to be used as either a 	*/
+/* character to be inserted, a repeat count, or a search string, 	*/
+/* depending on the current state.					*/
+void do_command(int c)
+{
+    int i;
+    int need_fix_pos;
+    FILE * out;
+    
+    if ( c == '\r') c = '\n';
+    if (locate_mode) {
+        size_t new_pos;
+          
+        if (c == LOCATE) {
+              locate_mode = 0;
+              locate_string = CORD_EMPTY;
+              return;
+        }
+        locate_string = CORD_cat_char(locate_string, (char)c);
+        new_pos = CORD_str(current, file_pos - CORD_len(locate_string) + 1,
+          		   locate_string);
+        if (new_pos != CORD_NOT_FOUND) {
+            need_redisplay = ALL;
+            new_pos += CORD_len(locate_string);
+            for (;;) {
+              	file_pos = line_pos(line + 1, 0);
+              	if (file_pos > new_pos) break;
+              	line++;
+            }
+            col = new_pos - line_pos(line, 0);
+            file_pos = new_pos;
+            fix_cursor();
+        } else {
+            locate_string = CORD_substr(locate_string, 0,
+              				CORD_len(locate_string) - 1);
+            beep();
+        }
+        return;
+    }
+    if (c == REPEAT) {
+      	repeat_count = BARE_PREFIX; return;
+    } else if (c < 0x100 && isdigit(c)){
+        if (repeat_count == BARE_PREFIX) {
+          repeat_count = c - '0'; return;
+        } else if (repeat_count != NO_PREFIX) {
+          repeat_count = 10 * repeat_count + c - '0'; return;
+        }
+    }
+    if (repeat_count == NO_PREFIX) repeat_count = 1;
+    if (repeat_count == BARE_PREFIX && (c == UP || c == DOWN)) {
+      	repeat_count = LINES - dis_granularity;
+    }
+    if (repeat_count == BARE_PREFIX) repeat_count = 8;
+    need_fix_pos = 0;
+    for (i = 0; i < repeat_count; i++) {
+        switch(c) {
+          case LOCATE:
+            locate_mode = 1;
+            break;
+          case TOP:
+            line = col = file_pos = 0;
+            break;
+     	  case UP:
+     	    if (line != 0) {
+     	        line--;
+     	        need_fix_pos = 1;
+     	    }
+     	    break;
+     	  case DOWN:
+     	    line++;
+     	    need_fix_pos = 1;
+     	    break;
+     	  case LEFT:
+     	    if (col != 0) {
+     	        col--; file_pos--;
+     	    }
+     	    break;
+     	  case RIGHT:
+     	    if (CORD_fetch(current, file_pos) == '\n') break;
+     	    col++; file_pos++;
+     	    break;
+     	  case UNDO:
+     	    del_hist();
+     	    need_redisplay = ALL; need_fix_pos = 1;
+     	    break;
+     	  case BS:
+     	    if (col == 0) {
+     	        beep();
+     	        break;
+     	    }
+     	    col--; file_pos--;
+     	    /* fall through: */
+     	  case DEL:
+     	    if (file_pos == current_len-1) break;
+     	    	/* Can't delete trailing newline */
+     	    if (CORD_fetch(current, file_pos) == '\n') {
+     	        need_redisplay = ALL; need_fix_pos = 1;
+     	    } else {
+     	        need_redisplay = line - dis_line;
+     	    }
+     	    add_hist(CORD_cat(
+     	    		CORD_substr(current, 0, file_pos),
+     	    		CORD_substr(current, file_pos+1, current_len)));
+     	    invalidate_map(line);
+     	    break;
+     	  case WRITE:
+    	    if ((out = fopen(arg_file_name, "wb")) == NULL
+  	        || CORD_put(current, out) == EOF) {
+        	de_error("Write failed\n");
+        	need_redisplay = ALL;
+            } else {
+                fclose(out);
+            }
+            break;
+     	  default:
+     	    {
+     	        CORD left_part = CORD_substr(current, 0, file_pos);
+     	        CORD right_part = CORD_substr(current, file_pos, current_len);
+     	        
+     	        add_hist(CORD_cat(CORD_cat_char(left_part, (char)c),
+     	        		  right_part));
+     	        invalidate_map(line);
+     	        if (c == '\n') {
+     	            col = 0; line++; file_pos++;
+     	            need_redisplay = ALL;
+     	        } else {
+     	            col++; file_pos++;
+     	            need_redisplay = line - dis_line;
+     	    	}
+     	        break;
+     	    }
+        }
+    }
+    if (need_fix_pos) fix_pos();
+    fix_cursor();
+    repeat_count = NO_PREFIX;
+}
+
+/* OS independent initialization */
+void generic_init(void)
+{
+    FILE * f;
+    CORD initial;
+    
+    if ((f = fopen(arg_file_name, "rb")) == NULL) {
+     	initial = "\n";
+    } else {
+        initial = CORD_from_file(f);
+        if (initial == CORD_EMPTY
+            || CORD_fetch(initial, CORD_len(initial)-1) != '\n') {
+            initial = CORD_cat(initial, "\n");
+        }
+    }
+    add_map(0,0);
+    add_hist(initial);
+    now -> map = current_map;
+    now -> previous = now;  /* Can't back up further: beginning of the world */
+    need_redisplay = ALL;
+    fix_cursor();
+}
+
+#ifndef WIN32
+
+main(argc, argv)
+int argc;
+char ** argv;
+{
+    int c;
+    CORD initial;
+    
+    if (argc != 2) goto usage;
+    arg_file_name = argv[1];
+    setvbuf(stdout, GC_MALLOC_ATOMIC(8192), _IOFBF, 8192);
+    initscr();
+    noecho(); nonl(); cbreak();
+    generic_init();
+    while ((c = getchar()) != QUIT) {
+    	do_command(c);
+    }
+done:
+    endwin();
+    exit(0);
+usage:
+    fprintf(stderr, "Usage: %s file\n", argv[0]);
+    fprintf(stderr, "Cursor keys: ^B(left) ^F(right) ^P(up) ^N(down)\n");
+    fprintf(stderr, "Undo: ^U    Write: ^W   Quit:^D  Repeat count: ^R[n]\n");
+    fprintf(stderr, "Top: ^T   Locate (search, find): ^L text ^L\n");
+    exit(1);
+}
+
+#endif  /* !WIN32 */
diff --git a/cord/de_cmds.h b/cord/de_cmds.h
new file mode 100644
index 00000000..f42ddcf2
--- /dev/null
+++ b/cord/de_cmds.h
@@ -0,0 +1,33 @@
+/*
+ * Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:24 pm PDT */
+
+#ifndef DE_CMDS_H
+
+# define DE_CMDS_H
+
+# define UP     16	/* ^P */
+# define DOWN   14	/* ^N */
+# define LEFT   2	/* ^B */
+# define RIGHT  6	/* ^F */
+# define DEL	127	/* ^? */
+# define BS     8	/* ^H */
+# define UNDO   21	/* ^U */
+# define WRITE  23	/* ^W */
+# define QUIT   4	/* ^D */
+# define REPEAT 18	/* ^R */
+# define LOCATE 12	/* ^L */
+# define TOP    20	/* ^T */
+
+#endif
+
diff --git a/cord/de_win.ICO b/cord/de_win.ICO
new file mode 100755
index 00000000..b20ac3ee
--- /dev/null
+++ b/cord/de_win.ICO
diff --git a/cord/de_win.RC b/cord/de_win.RC
new file mode 100644
index 00000000..554a3004
--- /dev/null
+++ b/cord/de_win.RC
@@ -0,0 +1,78 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to copy this garbage collector for any purpose,
+ * provided the above notices are retained on all copies.
+ */
+/* Boehm, May 13, 1994 9:50 am PDT */
+
+#include "windows.h"
+#include "de_cmds.h"
+#include "de_win.h"
+
+
+
+ABOUTBOX DIALOG 19, 21, 163, 47
+STYLE DS_MODALFRAME | WS_POPUP | WS_CAPTION | WS_SYSMENU
+CAPTION "About Demonstration Text Editor"
+BEGIN
+	ICON "DE", -1, 8, 8, 13, 13, WS_CHILD | WS_VISIBLE
+	LTEXT "Demonstration Text Editor", -1, 44, 8, 118, 8, WS_CHILD | WS_VISIBLE | WS_GROUP
+	LTEXT "Version 4.1", -1, 44, 16, 60, 8, WS_CHILD | WS_VISIBLE | WS_GROUP
+	PUSHBUTTON "OK", IDOK, 118, 27, 24, 14, WS_CHILD | WS_VISIBLE | WS_TABSTOP
+END
+
+
+DE MENU 
+BEGIN
+	POPUP "&File"
+	BEGIN
+		MENUITEM "&Save\t^W", IDM_FILESAVE
+		MENUITEM "E&xit\t^D", IDM_FILEEXIT
+	END
+
+	POPUP "&Edit"
+	BEGIN
+	    MENUITEM "Page &Down\t^R^N", IDM_EDITPDOWN
+	    MENUITEM "Page &Up\t^R^P", IDM_EDITPUP
+		MENUITEM "U&ndo\t^U", IDM_EDITUNDO
+		MENUITEM "&Locate\t^L ... ^L", IDM_EDITLOCATE
+		MENUITEM "D&own\t^N", IDM_EDITDOWN
+	    MENUITEM "U&p\t^P", IDM_EDITUP
+	    MENUITEM "Le&ft\t^B", IDM_EDITLEFT
+	    MENUITEM "&Right\t^F", IDM_EDITRIGHT
+	    MENUITEM "Delete &Backward\tBS", IDM_EDITBS
+	    MENUITEM "Delete F&orward\tDEL", IDM_EDITDEL
+	    MENUITEM "&Top\t^T", IDM_EDITTOP
+	END
+	
+	POPUP "&Help"
+	BEGIN
+		MENUITEM "&Contents", IDM_HELPCONTENTS
+		MENUITEM "&About...", IDM_HELPABOUT
+	END
+	
+	MENUITEM "Page_&Down", IDM_EDITPDOWN
+	MENUITEM "Page_&Up", IDM_EDITPUP
+END
+
+
+DE ACCELERATORS 
+BEGIN
+    "^R", IDM_EDITREPEAT
+    "^N", IDM_EDITDOWN
+    "^P", IDM_EDITUP
+    "^L", IDM_EDITLOCATE
+    "^B", IDM_EDITLEFT
+    "^F", IDM_EDITRIGHT
+    "^T", IDM_EDITTOP
+	VK_DELETE, IDM_EDITDEL, VIRTKEY
+	VK_BACK, IDM_EDITBS, VIRTKEY
+END
+
+
+DE ICON cord\de_win.ICO
+
diff --git a/cord/de_win.c b/cord/de_win.c
new file mode 100644
index 00000000..13567517
--- /dev/null
+++ b/cord/de_win.c
@@ -0,0 +1,365 @@
+/*
+ * Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:21 pm PDT */
+
+/*
+ * The MS Windows specific part of de.  
+ * This started as the generic Windows application template
+ * made available by Rob Haack (rhaack@polaris.unm.edu), but
+ * significant parts didn't survive to the final version.
+ *
+ * This was written by a nonexpert windows programmer.
+ */
+
+
+#include "windows.h"
+#include "gc.h"
+#include "cord.h"
+#include "de_cmds.h"
+#include "de_win.h"
+
+int LINES = 0;
+int COLS = 0;
+
+char       szAppName[]     = "DE";
+char       FullAppName[]   = "Demonstration Editor";
+
+HWND        hwnd;
+
+void de_error(char *s)
+{
+    MessageBox( hwnd, (LPSTR) s,
+                (LPSTR) FullAppName,
+                MB_ICONINFORMATION | MB_OK );
+    InvalidateRect(hwnd, NULL, TRUE);
+}
+
+int APIENTRY WinMain (HINSTANCE hInstance, HINSTANCE hPrevInstance,
+                      LPSTR command_line, int nCmdShow)
+{
+   MSG         msg;
+   WNDCLASS    wndclass;
+   HANDLE      hAccel;
+
+   if (!hPrevInstance)
+   {
+      wndclass.style          = CS_HREDRAW | CS_VREDRAW;
+      wndclass.lpfnWndProc    = WndProc;
+      wndclass.cbClsExtra     = 0;
+      wndclass.cbWndExtra     = DLGWINDOWEXTRA;
+      wndclass.hInstance      = hInstance;
+      wndclass.hIcon          = LoadIcon (hInstance, szAppName);
+      wndclass.hCursor        = LoadCursor (NULL, IDC_ARROW);
+      wndclass.hbrBackground  = GetStockObject(WHITE_BRUSH);
+      wndclass.lpszMenuName   = "DE";
+      wndclass.lpszClassName  = szAppName;
+
+      if (RegisterClass (&wndclass) == 0) {
+          char buf[50];
+   	
+   	  sprintf(buf, "RegisterClass: error code: 0x%X", GetLastError());
+   	  de_error(buf);
+   	  return(0);
+      }
+   }
+   
+   /* Empirically, the command line does not include the command name ...
+   if (command_line != 0) {
+       while (isspace(*command_line)) command_line++;
+       while (*command_line != 0 && !isspace(*command_line)) command_line++;
+       while (isspace(*command_line)) command_line++;
+   } */
+   
+   if (command_line == 0 || *command_line == 0) {
+        de_error("File name argument required");
+        return( 0 );
+   } else {
+        char *p = command_line;
+        
+        while (*p != 0 && !isspace(*p)) p++;
+   	arg_file_name = CORD_to_char_star(
+   			    CORD_substr(command_line, 0, p - command_line));
+   }
+
+   hwnd = CreateWindow (szAppName,
+   			FullAppName,
+   			WS_OVERLAPPEDWINDOW | WS_CAPTION, /* Window style */
+   			CW_USEDEFAULT, 0, /* default pos. */,
+   			CW_USEDEFAULT, 0, /* default width, height */,
+   			NULL,	/* No parent */
+   			NULL, 	/* Window class menu */
+   			hInstance, NULL);
+   if (hwnd == NULL) {
+   	char buf[50];
+   	
+   	sprintf(buf, "CreateWindow: error code: 0x%X", GetLastError());
+   	de_error(buf);
+   	return(0);
+   }
+
+   ShowWindow (hwnd, nCmdShow);
+
+   hAccel = LoadAccelerators( hInstance, szAppName );
+   
+   while (GetMessage (&msg, NULL, 0, 0))
+   {
+      if( !TranslateAccelerator( hwnd, hAccel, &msg ) )
+      {
+         TranslateMessage (&msg);
+         DispatchMessage (&msg);
+      }
+   }
+   return msg.wParam;
+}
+
+/* Return the argument with all control characters replaced by blanks.	*/
+char * plain_chars(char * text, size_t len)
+{
+    char * result = GC_MALLOC_ATOMIC(len + 1);
+    register size_t i;
+    
+    for (i = 0; i < len; i++) {
+       if (iscntrl(text[i])) {
+           result[i] = ' ';
+       } else {
+           result[i] = text[i];
+       }
+    }
+    result[len] = '\0';
+    return(result);
+}
+
+/* Return the argument with all non-control-characters replaced by 	*/
+/* blank, and all control characters c replaced by c + 32.		*/
+char * control_chars(char * text, size_t len)
+{
+    char * result = GC_MALLOC_ATOMIC(len + 1);
+    register size_t i;
+    
+    for (i = 0; i < len; i++) {
+       if (iscntrl(text[i])) {
+           result[i] = text[i] + 0x40;
+       } else {
+           result[i] = ' ';
+       }
+    }
+    result[len] = '\0';
+    return(result);
+}
+
+int char_width;
+int char_height;
+
+void get_line_rect(int line, int win_width, RECT * rectp)
+{
+    rectp -> top = line * char_height;
+    rectp -> bottom = rectp->top + char_height;
+    rectp -> left = 0;
+    rectp -> right = win_width;
+}
+
+int caret_visible = 0;	/* Caret is currently visible.	*/
+
+int screen_was_painted = 0;/* Screen has been painted at least once.	*/
+
+void update_cursor(void);
+
+LRESULT CALLBACK WndProc (HWND hwnd, UINT message,
+                          WPARAM wParam, LPARAM lParam)
+{
+   static FARPROC lpfnAboutBox;
+   static HANDLE  hInstance;
+   HDC dc;
+   PAINTSTRUCT ps;
+   RECT client_area;
+   RECT this_line;
+   RECT dummy;
+   TEXTMETRIC tm;
+   register int i;
+   int id;
+
+   switch (message)
+   {
+      case WM_CREATE:
+           hInstance = ( (LPCREATESTRUCT) lParam)->hInstance;
+           lpfnAboutBox = MakeProcInstance( (FARPROC) AboutBox, hInstance );
+           dc = GetDC(hwnd);
+           SelectObject(dc, GetStockObject(SYSTEM_FIXED_FONT));
+           GetTextMetrics(dc, &tm);
+           ReleaseDC(hwnd, dc);
+           char_width = tm.tmAveCharWidth;
+           char_height = tm.tmHeight + tm.tmExternalLeading;
+           GetClientRect(hwnd, &client_area);
+      	   COLS = (client_area.right - client_area.left)/char_width;
+      	   LINES = (client_area.bottom - client_area.top)/char_height;
+      	   generic_init();
+           return(0);
+
+      case WM_CHAR:
+      	   if (wParam == QUIT) {
+      	       SendMessage( hwnd, WM_CLOSE, 0, 0L );
+      	   } else {
+      	       do_command(wParam);
+      	   }
+      	   return(0);
+      
+      case WM_SETFOCUS:
+      	   CreateCaret(hwnd, NULL, char_width, char_height);
+      	   ShowCaret(hwnd);
+      	   caret_visible = 1;
+      	   update_cursor();
+      	   return(0);
+      	   
+      case WM_KILLFOCUS:
+      	   HideCaret(hwnd);
+      	   DestroyCaret();
+      	   caret_visible = 0;
+      	   return(0);
+      	   
+      case WM_LBUTTONUP:
+      	   {
+      	       unsigned xpos = LOWORD(lParam);	/* From left	*/
+      	       unsigned ypos = HIWORD(lParam);	/* from top */
+      	       
+      	       set_position( xpos/char_width, ypos/char_height );
+      	       return(0);
+      	   }
+      	   
+      case WM_COMMAND:
+      	   id = LOWORD(wParam);
+      	   if (id & EDIT_CMD_FLAG) {
+               if (id & REPEAT_FLAG) do_command(REPEAT);
+               do_command(CHAR_CMD(id));
+               return( 0 );
+           } else {
+             switch(id) {
+               case IDM_FILEEXIT:
+                  SendMessage( hwnd, WM_CLOSE, 0, 0L );
+                  return( 0 );
+
+               case IDM_HELPABOUT:
+                  if( DialogBox( hInstance, "ABOUTBOX",
+                                 hwnd, lpfnAboutBox ) );
+                     InvalidateRect( hwnd, NULL, TRUE );
+                  return( 0 );
+	       case IDM_HELPCONTENTS:
+	     	  de_error(
+	     	       "Cursor keys: ^B(left) ^F(right) ^P(up) ^N(down)\n"
+	     	       "Undo: ^U    Write: ^W   Quit:^D  Repeat count: ^R[n]\n"
+	     	       "Top: ^T   Locate (search, find): ^L text ^L\n");
+	     	  return( 0 );
+	     }
+	   }
+           break;
+
+      case WM_CLOSE:
+           DestroyWindow( hwnd );
+           return 0;
+
+      case WM_DESTROY:
+           PostQuitMessage (0);
+           return 0;
+      
+      case WM_PAINT:
+      	   dc = BeginPaint(hwnd, &ps);
+      	   GetClientRect(hwnd, &client_area);
+      	   COLS = (client_area.right - client_area.left)/char_width;
+      	   LINES = (client_area.bottom - client_area.top)/char_height;
+      	   SelectObject(dc, GetStockObject(SYSTEM_FIXED_FONT));
+      	   for (i = 0; i < LINES; i++) {
+      	       get_line_rect(i, client_area.right, &this_line);
+      	       if (IntersectRect(&dummy, &this_line, &ps.rcPaint)) {
+      	           CORD raw_line = retrieve_screen_line(i);
+      	           size_t len = CORD_len(raw_line);
+      	           char * text = CORD_to_char_star(raw_line);
+      	           		/* May contain embedded NULLs	*/
+      	           char * plain = plain_chars(text, len);
+      	           char * blanks = CORD_to_char_star(CORD_chars(' ',
+      	           				                COLS - len));
+      	           char * control = control_chars(text, len);
+#		   define RED RGB(255,0,0)
+      	           
+      	           SetBkMode(dc, OPAQUE);
+      	           SetTextColor(dc, GetSysColor(COLOR_WINDOWTEXT));
+      	           
+      	           TextOut(dc, this_line.left, this_line.top,
+      	           	   plain, len);
+      	           TextOut(dc, this_line.left + len * char_width, this_line.top,
+      	           	   blanks, COLS - len);
+      	           SetBkMode(dc, TRANSPARENT);
+      	           SetTextColor(dc, RED);
+      	           TextOut(dc, this_line.left, this_line.top,
+      	           	   control, strlen(control));
+      	       }
+      	   }
+      	   EndPaint(hwnd, &ps);
+      	   screen_was_painted = 1;
+      	   return 0;
+   }
+   return DefWindowProc (hwnd, message, wParam, lParam);
+}
+
+int last_col;
+int last_line;
+
+void move_cursor(int c, int l)
+{
+    last_col = c;
+    last_line = l;
+    
+    if (caret_visible) update_cursor();
+}
+
+void update_cursor(void)
+{
+    SetCaretPos(last_col * char_width, last_line * char_height);
+    ShowCaret(hwnd);
+}
+
+void invalidate_line(int i)
+{
+    RECT line;
+    
+    if (!screen_was_painted) return;
+    	/* Invalidating a rectangle before painting seems result in a	*/
+    	/* major performance problem.					*/
+    get_line_rect(i, COLS*char_width, &line);
+    InvalidateRect(hwnd, &line, FALSE);
+}
+
+LRESULT CALLBACK AboutBox( HWND hDlg, UINT message,
+                           WPARAM wParam, LPARAM lParam )
+{
+   switch( message )
+   {
+      case WM_INITDIALOG:
+           SetFocus( GetDlgItem( hDlg, IDOK ) );
+           break;
+
+      case WM_COMMAND:
+           switch( wParam )
+           {
+              case IDOK:
+                   EndDialog( hDlg, TRUE );
+                   break;
+           }
+           break;
+
+      case WM_CLOSE:
+           EndDialog( hDlg, TRUE );
+           return TRUE;
+
+   }
+   return FALSE;
+}
+
diff --git a/cord/de_win.h b/cord/de_win.h
new file mode 100644
index 00000000..57a47b45
--- /dev/null
+++ b/cord/de_win.h
@@ -0,0 +1,103 @@
+/*
+ * Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:25 pm PDT */
+
+/* cord.h, de_cmds.h, and windows.h should be included before this. */
+
+
+# define OTHER_FLAG	0x100
+# define EDIT_CMD_FLAG	0x200
+# define REPEAT_FLAG	0x400
+
+# define CHAR_CMD(i) ((i) & 0xff)
+
+/* MENU: DE */
+#define IDM_FILESAVE		(EDIT_CMD_FLAG + WRITE)
+#define IDM_FILEEXIT		(OTHER_FLAG + 1)
+#define IDM_HELPABOUT		(OTHER_FLAG + 2)
+#define IDM_HELPCONTENTS	(OTHER_FLAG + 3)
+
+#define IDM_EDITPDOWN		(REPEAT_FLAG + EDIT_CMD_FLAG + DOWN)
+#define IDM_EDITPUP		(REPEAT_FLAG + EDIT_CMD_FLAG + UP)
+#define IDM_EDITUNDO		(EDIT_CMD_FLAG + UNDO)
+#define IDM_EDITLOCATE		(EDIT_CMD_FLAG + LOCATE)
+#define IDM_EDITDOWN		(EDIT_CMD_FLAG + DOWN)
+#define IDM_EDITUP		(EDIT_CMD_FLAG + UP)
+#define IDM_EDITLEFT		(EDIT_CMD_FLAG + LEFT)
+#define IDM_EDITRIGHT		(EDIT_CMD_FLAG + RIGHT)
+#define IDM_EDITBS		(EDIT_CMD_FLAG + BS)
+#define IDM_EDITDEL		(EDIT_CMD_FLAG + DEL)
+#define IDM_EDITREPEAT		(EDIT_CMD_FLAG + REPEAT)
+#define IDM_EDITTOP		(EDIT_CMD_FLAG + TOP)
+
+
+
+
+/* Windows UI stuff	*/
+
+LRESULT CALLBACK WndProc (HWND hwnd, UINT message,
+			  UINT wParam, LONG lParam);
+
+LRESULT CALLBACK AboutBox( HWND hDlg, UINT message,
+			   UINT wParam, LONG lParam );
+
+
+/* Screen dimensions.  Maintained by de_win.c.	*/
+extern int LINES;
+extern int COLS;
+
+/* File being edited.	*/
+extern char * arg_file_name;
+
+/* Current display position in file.  Maintained by de.c	*/
+extern int dis_line;
+extern int dis_col;
+
+/* Current cursor position in file.				*/
+extern int line;
+extern int col;
+
+/*
+ *  Calls from de_win.c to de.c
+ */
+  
+CORD retrieve_screen_line(int i);
+			/* Get the contents of i'th screen line.	*/
+			/* Relies on COLS.				*/
+
+void set_position(int x, int y);
+			/* Set column, row.  Upper left of window = (0,0). */
+
+void do_command(int);
+			/* Execute an editor command.			*/
+			/* Agument is a command character or one	*/
+			/* of the IDM_ commands.			*/
+
+void generic_init(void);
+			/* OS independent initialization */
+
+
+/*
+ * Calls from de.c to de_win.c
+ */
+ 
+void move_cursor(int column, int line);
+			/* Physically move the cursor on the display,	*/
+			/* so that it appears at			*/
+			/* (column, line).				*/
+
+void invalidate_line(int line);
+			/* Invalidate line i on the screen.	*/
+
+void de_error(char *s);
+			/* Display error message.	*/
\ No newline at end of file
diff --git a/cord/ec.h b/cord/ec.h
new file mode 100644
index 00000000..c829b83a
--- /dev/null
+++ b/cord/ec.h
@@ -0,0 +1,70 @@
+# ifndef EC_H
+# define EC_H
+
+# ifndef CORD_H
+#  include "cord.h"
+# endif
+
+/* Extensible cords are strings that may be destructively appended to.	*/
+/* They allow fast construction of cords from characters that are	*/
+/* being read from a stream.						*/
+/*
+ * A client might look like:
+ *
+ *	{
+ *	    CORD_ec x;
+ *	    CORD result;
+ *	    char c;
+ *	    FILE *f;
+ *
+ *	    ...
+ *	    CORD_ec_init(x);
+ *	    while(...) {
+ *		c = getc(f);
+ *		...
+ *		CORD_ec_append(x, c);
+ *	    }
+ *	    result = CORD_balance(CORD_ec_to_cord(x));
+ *
+ * If a C string is desired as the final result, the call to CORD_balance
+ * may be replaced by a call to CORD_to_char_star.
+ */
+
+# ifndef CORD_BUFSZ
+#   define CORD_BUFSZ 128
+# endif
+
+typedef struct CORD_ec_struct {
+    CORD ec_cord;
+    char * ec_bufptr;
+    char ec_buf[CORD_BUFSZ+1];
+} CORD_ec[1];
+
+/* This structure represents the concatenation of ec_cord with		*/
+/* ec_buf[0 ... (ec_bufptr-ec_buf-1)]					*/
+
+/* Flush the buffer part of the extended chord into ec_cord.	*/
+/* Note that this is almost the only real function, and it is	*/
+/* implemented in 6 lines in cordxtra.c				*/
+void CORD_ec_flush_buf(CORD_ec x);
+      
+/* Convert an extensible cord to a cord. */
+# define CORD_ec_to_cord(x) (CORD_ec_flush_buf(x), (x)[0].ec_cord)
+
+/* Initialize an extensible cord. */
+# define CORD_ec_init(x) ((x)[0].ec_cord = 0, (x)[0].ec_bufptr = (x)[0].ec_buf)
+
+/* Append a character to an extensible cord.	*/
+# define CORD_ec_append(x, c) \
+    {  \
+	if ((x)[0].ec_bufptr == (x)[0].ec_buf + CORD_BUFSZ) { \
+	  	CORD_ec_flush_buf(x); \
+	} \
+	*((x)[0].ec_bufptr)++ = (c); \
+    }
+
+/* Append a cord to an extensible cord.  Structure remains shared with 	*/
+/* original.								*/
+void CORD_ec_append_cord(CORD_ec x, CORD s);
+
+# endif /* EC_H */
diff --git a/dbg_mlc.c b/dbg_mlc.c
new file mode 100644
index 00000000..87275d66
--- /dev/null
+++ b/dbg_mlc.c
@@ -0,0 +1,542 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:07 pm PDT */
+# include "gc_priv.h"
+
+/* Do we want to and know how to save the call stack at the time of	*/
+/* an allocation?  How much space do we want to use in each object?	*/
+
+# if defined(SPARC) && defined(SUNOS4)
+#   include <machine/frame.h>
+#   define SAVE_CALL_CHAIN
+#   define NFRAMES 5	/* Number of frames to save. */
+#   define NARGS 2	/* Mumber of arguments to save for each call. */
+#   if NARGS > 6
+	--> We only know how to to get the first 6 arguments
+#   endif
+# endif
+
+# define START_FLAG ((word)0xfedcedcb)
+# define END_FLAG ((word)0xbcdecdef)
+	/* Stored both one past the end of user object, and one before	*/
+	/* the end of the object as seen by the allocator.		*/
+
+#ifdef SAVE_CALL_CHAIN
+    struct callinfo {
+	word ci_pc;
+	word ci_arg[NARGS];	/* bit-wise complement to avoid retention */
+    };
+#endif
+
+/* Object header */
+typedef struct {
+    char * oh_string;		/* object descriptor string	*/
+    word oh_int;		/* object descriptor integers	*/
+#   ifdef SAVE_CALL_CHAIN
+      struct callinfo oh_ci[NFRAMES];
+#   endif
+    word oh_sz;			/* Original malloc arg.		*/
+    word oh_sf;			/* start flag */
+} oh;
+/* The size of the above structure is assumed not to dealign things,	*/
+/* and to be a multiple of the word length.				*/
+
+#define DEBUG_BYTES (sizeof (oh) + sizeof (word))
+#undef ROUNDED_UP_WORDS
+#define ROUNDED_UP_WORDS(n) BYTES_TO_WORDS((n) + WORDS_TO_BYTES(1) - 1)
+
+#if defined(SPARC) && defined(SUNOS4)
+/* Fill in the pc and argument information for up to NFRAMES of my	*/
+/* callers.  Ignore my frame and my callers frame.			*/
+void GC_save_callers (info) 
+struct callinfo info[NFRAMES];
+{
+  struct frame *frame;
+  struct frame *fp;
+  int nframes = 0;
+  word GC_save_regs_in_stack();
+
+  frame = (struct frame *) GC_save_regs_in_stack ();
+  
+  for (fp = frame -> fr_savfp; fp != 0 && nframes < NFRAMES;
+       fp = fp -> fr_savfp, nframes++) {
+      register int i;
+      
+      info[nframes].ci_pc = fp->fr_savpc;
+      for (i = 0; i < NARGS; i++) {
+	info[nframes].ci_arg[i] = ~(fp->fr_arg[i]);
+      }
+  }
+  if (nframes < NFRAMES) info[nframes].ci_pc = 0;
+}
+
+void GC_print_callers (info)
+struct callinfo info[NFRAMES];
+{
+    register int i,j;
+    
+    GC_err_printf0("\tCall chain at allocation:\n");
+    for (i = 0; i < NFRAMES; i++) {
+     	if (info[i].ci_pc == 0) break;
+     	GC_err_printf1("\t##PC##= 0x%X\n\t\targs: ", info[i].ci_pc);
+     	for (j = 0; j < NARGS; j++) {
+     	    if (j != 0) GC_err_printf0(", ");
+     	    GC_err_printf2("%d (0x%X)", ~(info[i].ci_arg[j]),
+     	    				~(info[i].ci_arg[j]));
+     	}
+     	GC_err_printf0("\n");
+    }
+}
+
+#endif /* SPARC & SUNOS4 */
+
+#ifdef SAVE_CALL_CHAIN
+#   define ADD_CALL_CHAIN(base) GC_save_callers(((oh *)(base)) -> oh_ci)
+#   define PRINT_CALL_CHAIN(base) GC_print_callers(((oh *)(base)) -> oh_ci)
+#else
+#   define ADD_CALL_CHAIN(base)
+#   define PRINT_CALL_CHAIN(base)
+#endif
+
+/* Check whether object with base pointer p has debugging info	*/ 
+/* p is assumed to point to a legitimate object in our part	*/
+/* of the heap.							*/
+bool GC_has_debug_info(p)
+ptr_t p;
+{
+    register oh * ohdr = (oh *)p;
+    register ptr_t body = (ptr_t)(ohdr + 1);
+    register word sz = GC_size((ptr_t) ohdr);
+    
+    if (HBLKPTR((ptr_t)ohdr) != HBLKPTR((ptr_t)body)
+        || sz < sizeof (oh)) {
+        return(FALSE);
+    }
+    if (ohdr -> oh_sz == sz) {
+    	/* Object may have had debug info, but has been deallocated	*/
+    	return(FALSE);
+    }
+    if (ohdr -> oh_sf == (START_FLAG ^ (word)body)) return(TRUE);
+    if (((word *)ohdr)[BYTES_TO_WORDS(sz)-1] == (END_FLAG ^ (word)body)) {
+        return(TRUE);
+    }
+    return(FALSE);
+}
+
+/* Store debugging info into p.  Return displaced pointer. */
+/* Assumes we don't hold allocation lock.		   */
+ptr_t GC_store_debug_info(p, sz, string, integer)
+register ptr_t p;	/* base pointer */
+word sz; 	/* bytes */
+char * string;
+word integer;
+{
+    register word * result = (word *)((oh *)p + 1);
+    DCL_LOCK_STATE;
+    
+    /* There is some argument that we should dissble signals here.	*/
+    /* But that's expensive.  And this way things should only appear	*/
+    /* inconsistent while we're in the handler.				*/
+    LOCK();
+    ((oh *)p) -> oh_string = string;
+    ((oh *)p) -> oh_int = integer;
+    ((oh *)p) -> oh_sz = sz;
+    ((oh *)p) -> oh_sf = START_FLAG ^ (word)result;
+    ((word *)p)[BYTES_TO_WORDS(GC_size(p))-1] =
+         result[ROUNDED_UP_WORDS(sz)] = END_FLAG ^ (word)result;
+    UNLOCK();
+    return((ptr_t)result);
+}
+
+/* Check the object with debugging info at p 		*/
+/* return NIL if it's OK.  Else return clobbered	*/
+/* address.						*/
+ptr_t GC_check_annotated_obj(ohdr)
+register oh * ohdr;
+{
+    register ptr_t body = (ptr_t)(ohdr + 1);
+    register word gc_sz = GC_size((ptr_t)ohdr);
+    if (ohdr -> oh_sz + DEBUG_BYTES > gc_sz) {
+        return((ptr_t)(&(ohdr -> oh_sz)));
+    }
+    if (ohdr -> oh_sf != (START_FLAG ^ (word)body)) {
+        return((ptr_t)(&(ohdr -> oh_sf)));
+    }
+    if (((word *)ohdr)[BYTES_TO_WORDS(gc_sz)-1] != (END_FLAG ^ (word)body)) {
+        return((ptr_t)((word *)ohdr + BYTES_TO_WORDS(gc_sz)-1));
+    }
+    if (((word *)body)[ROUNDED_UP_WORDS(ohdr -> oh_sz)]
+        != (END_FLAG ^ (word)body)) {
+        return((ptr_t)((word *)body + ROUNDED_UP_WORDS(ohdr -> oh_sz)));
+    }
+    return(0);
+}
+
+void GC_print_obj(p)
+ptr_t p;
+{
+    register oh * ohdr = (oh *)GC_base(p);
+    
+    GC_err_printf1("0x%lx (", (unsigned long)ohdr + sizeof(oh));
+    GC_err_puts(ohdr -> oh_string);
+    GC_err_printf2(":%ld, sz=%ld)\n", (unsigned long)(ohdr -> oh_int),
+        			      (unsigned long)(ohdr -> oh_sz));
+    PRINT_CALL_CHAIN(ohdr);
+}
+void GC_print_smashed_obj(p, clobbered_addr)
+ptr_t p, clobbered_addr;
+{
+    register oh * ohdr = (oh *)GC_base(p);
+    
+    GC_err_printf2("0x%lx in object at 0x%lx(", (unsigned long)clobbered_addr,
+    					        (unsigned long)p);
+    if (clobbered_addr <= (ptr_t)(&(ohdr -> oh_sz))
+        || ohdr -> oh_string == 0) {
+        GC_err_printf1("<smashed>, appr. sz = %ld)\n",
+        	       BYTES_TO_WORDS(GC_size((ptr_t)ohdr)));
+    } else {
+        if (ohdr -> oh_string[0] == '\0') {
+            GC_err_puts("EMPTY(smashed?)");
+        } else {
+            GC_err_puts(ohdr -> oh_string);
+        }
+        GC_err_printf2(":%ld, sz=%ld)\n", (unsigned long)(ohdr -> oh_int),
+        			          (unsigned long)(ohdr -> oh_sz));
+    }
+}
+
+void GC_check_heap_proc();
+
+void GC_start_debugging()
+{
+    GC_check_heap = GC_check_heap_proc;
+    GC_debugging_started = TRUE;
+    GC_register_displacement((word)sizeof(oh));
+}
+
+# ifdef __STDC__
+    extern_ptr_t GC_debug_malloc(size_t lb, char * s, int i)
+# else
+    extern_ptr_t GC_debug_malloc(lb, s, i)
+    size_t lb;
+    char * s;
+    int i;
+# endif
+{
+    extern_ptr_t result = GC_malloc(lb + DEBUG_BYTES);
+    
+    if (result == 0) {
+        GC_err_printf1("GC_debug_malloc(%ld) returning NIL (",
+        	       (unsigned long) lb);
+        GC_err_puts(s);
+        GC_err_printf1(":%ld)\n", (unsigned long)i);
+        return(0);
+    }
+    if (!GC_debugging_started) {
+    	GC_start_debugging();
+    }
+    ADD_CALL_CHAIN(result);
+    return (GC_store_debug_info(result, (word)lb, s, (word)i));
+}
+
+#ifdef STUBBORN_ALLOC
+# ifdef __STDC__
+    extern_ptr_t GC_debug_malloc_stubborn(size_t lb, char * s, int i)
+# else
+    extern_ptr_t GC_debug_malloc_stubborn(lb, s, i)
+    size_t lb;
+    char * s;
+    int i;
+# endif
+{
+    extern_ptr_t result = GC_malloc_stubborn(lb + DEBUG_BYTES);
+    
+    if (result == 0) {
+        GC_err_printf1("GC_debug_malloc(%ld) returning NIL (",
+        	       (unsigned long) lb);
+        GC_err_puts(s);
+        GC_err_printf1(":%ld)\n", (unsigned long)i);
+        return(0);
+    }
+    if (!GC_debugging_started) {
+    	GC_start_debugging();
+    }
+    ADD_CALL_CHAIN(result);
+    return (GC_store_debug_info(result, (word)lb, s, (word)i));
+}
+
+void GC_debug_change_stubborn(p)
+extern_ptr_t p;
+{
+    register extern_ptr_t q = GC_base(p);
+    register hdr * hhdr;
+    
+    if (q == 0) {
+        GC_err_printf1("Bad argument: 0x%lx to GC_debug_change_stubborn\n",
+        	       (unsigned long) p);
+        ABORT("GC_debug_change_stubborn: bad arg");
+    }
+    hhdr = HDR(q);
+    if (hhdr -> hb_obj_kind != STUBBORN) {
+        GC_err_printf1("GC_debug_change_stubborn arg not stubborn: 0x%lx\n",
+        	       (unsigned long) p);
+        ABORT("GC_debug_change_stubborn: arg not stubborn");
+    }
+    GC_change_stubborn(q);
+}
+
+void GC_debug_end_stubborn_change(p)
+extern_ptr_t p;
+{
+    register extern_ptr_t q = GC_base(p);
+    register hdr * hhdr;
+    
+    if (q == 0) {
+        GC_err_printf1("Bad argument: 0x%lx to GC_debug_end_stubborn_change\n",
+        	       (unsigned long) p);
+        ABORT("GC_debug_end_stubborn_change: bad arg");
+    }
+    hhdr = HDR(q);
+    if (hhdr -> hb_obj_kind != STUBBORN) {
+        GC_err_printf1("debug_end_stubborn_change arg not stubborn: 0x%lx\n",
+        	       (unsigned long) p);
+        ABORT("GC_debug_end_stubborn_change: arg not stubborn");
+    }
+    GC_end_stubborn_change(q);
+}
+
+#endif /* STUBBORN_ALLOC */
+
+# ifdef __STDC__
+    extern_ptr_t GC_debug_malloc_atomic(size_t lb, char * s, int i)
+# else
+    extern_ptr_t GC_debug_malloc_atomic(lb, s, i)
+    size_t lb;
+    char * s;
+    int i;
+# endif
+{
+    extern_ptr_t result = GC_malloc_atomic(lb + DEBUG_BYTES);
+    
+    if (result == 0) {
+        GC_err_printf1("GC_debug_malloc_atomic(%ld) returning NIL (",
+        	      (unsigned long) lb);
+        GC_err_puts(s);
+        GC_err_printf1(":%ld)\n", (unsigned long)i);
+        return(0);
+    }
+    if (!GC_debugging_started) {
+        GC_start_debugging();
+    }
+    ADD_CALL_CHAIN(result);
+    return (GC_store_debug_info(result, (word)lb, s, (word)i));
+}
+
+# ifdef __STDC__
+    extern_ptr_t GC_debug_malloc_uncollectable(size_t lb, char * s, int i)
+# else
+    extern_ptr_t GC_debug_malloc_uncollectable(lb, s, i)
+    size_t lb;
+    char * s;
+    int i;
+# endif
+{
+    extern_ptr_t result = GC_malloc_uncollectable(lb + DEBUG_BYTES);
+    
+    if (result == 0) {
+        GC_err_printf1("GC_debug_malloc_uncollectable(%ld) returning NIL (",
+        	      (unsigned long) lb);
+        GC_err_puts(s);
+        GC_err_printf1(":%ld)\n", (unsigned long)i);
+        return(0);
+    }
+    if (!GC_debugging_started) {
+        GC_start_debugging();
+    }
+    ADD_CALL_CHAIN(result);
+    return (GC_store_debug_info(result, (word)lb, s, (word)i));
+}
+
+
+# ifdef __STDC__
+    void GC_debug_free(extern_ptr_t p)
+# else
+    void GC_debug_free(p)
+    extern_ptr_t p;
+# endif
+{
+    register extern_ptr_t base = GC_base(p);
+    register ptr_t clobbered;
+    
+    if (base == 0) {
+        GC_err_printf1("Attempt to free invalid pointer %lx\n",
+        	       (unsigned long)p);
+        if (p != 0) ABORT("free(invalid pointer)");
+    }
+    if ((ptr_t)p - (ptr_t)base != sizeof(oh)) {
+        GC_err_printf1(
+        	  "GC_debug_free called on pointer %lx wo debugging info\n",
+        	  (unsigned long)p);
+    } else {
+      clobbered = GC_check_annotated_obj((oh *)base);
+      if (clobbered != 0) {
+        if (((oh *)base) -> oh_sz == GC_size(base)) {
+            GC_err_printf0(
+                  "GC_debug_free: found previously deallocated (?) object at ");
+        } else {
+            GC_err_printf0("GC_debug_free: found smashed object at ");
+        }
+        GC_print_smashed_obj(p, clobbered);
+      }
+      /* Invalidate size */
+      ((oh *)base) -> oh_sz = GC_size(base);
+    }
+#   ifdef FIND_LEAK
+        GC_free(base);
+#   endif
+}
+
+# ifdef __STDC__
+    extern_ptr_t GC_debug_realloc(extern_ptr_t p, size_t lb, char *s, int i)
+# else
+    extern_ptr_t GC_debug_realloc(p, lb, s, i)
+    extern_ptr_t p;
+    size_t lb;
+    char *s;
+    int i;
+# endif
+{
+    register extern_ptr_t base = GC_base(p);
+    register ptr_t clobbered;
+    register extern_ptr_t result = GC_debug_malloc(lb, s, i);
+    register size_t copy_sz = lb;
+    register size_t old_sz;
+    register hdr * hhdr;
+    
+    if (p == 0) return(GC_debug_malloc(lb, s, i));
+    if (base == 0) {
+        GC_err_printf1(
+              "Attempt to free invalid pointer %lx\n", (unsigned long)p);
+        ABORT("realloc(invalid pointer)");
+    }
+    if ((ptr_t)p - (ptr_t)base != sizeof(oh)) {
+        GC_err_printf1(
+        	"GC_debug_realloc called on pointer %lx wo debugging info\n",
+        	(unsigned long)p);
+        return(GC_realloc(p, lb));
+    }
+    hhdr = HDR(base);
+    switch (hhdr -> hb_obj_kind) {
+#    ifdef STUBBORN_ALLOC
+      case STUBBORN:
+        result = GC_debug_malloc_stubborn(lb, s, i);
+        break;
+#    endif
+      case NORMAL:
+        result = GC_debug_malloc(lb, s, i);
+        break;
+      case PTRFREE:
+        result = GC_debug_malloc_atomic(lb, s, i);
+        break;
+      default:
+        GC_err_printf0("GC_debug_realloc: encountered bad kind\n");
+        ABORT("bad kind");
+    }
+    clobbered = GC_check_annotated_obj((oh *)base);
+    if (clobbered != 0) {
+        GC_err_printf0("GC_debug_realloc: found smashed object at ");
+        GC_print_smashed_obj(p, clobbered);
+    }
+    old_sz = ((oh *)base) -> oh_sz;
+    if (old_sz < copy_sz) copy_sz = old_sz;
+    if (result == 0) return(0);
+    BCOPY(p, result,  copy_sz);
+    return(result);
+}
+
+/* Check all marked objects in the given block for validity */
+/*ARGSUSED*/
+void GC_check_heap_block(hbp, dummy)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+word dummy;
+{
+    register struct hblkhdr * hhdr = HDR(hbp);
+    register word sz = hhdr -> hb_sz;
+    register int word_no;
+    register word *p, *plim;
+    
+    p = (word *)(hbp->hb_body);
+    word_no = HDR_WORDS;
+    plim = (word *)((((word)hbp) + HBLKSIZE)
+		   - WORDS_TO_BYTES(sz));
+
+    /* go through all words in block */
+	do {
+	    if( mark_bit_from_hdr(hhdr, word_no)
+	        && GC_has_debug_info((ptr_t)p)) {
+	        ptr_t clobbered = GC_check_annotated_obj((oh *)p);
+	        
+	        if (clobbered != 0) {
+	            GC_err_printf0(
+	                "GC_check_heap_block: found smashed object at ");
+        	    GC_print_smashed_obj((ptr_t)p, clobbered);
+	        }
+	    }
+	    word_no += sz;
+	    p += sz;
+	} while( p <= plim );
+}
+
+
+/* This assumes that all accessible objects are marked, and that	*/
+/* I hold the allocation lock.	Normally called by collector.		*/
+void GC_check_heap_proc()
+{
+    GC_apply_to_all_blocks(GC_check_heap_block, (word)0);
+}
+
+struct closure {
+    GC_finalization_proc cl_fn;
+    extern_ptr_t cl_data;
+};
+
+# ifdef __STDC__
+    void * GC_make_closure(GC_finalization_proc fn, void * data)
+# else
+    extern_ptr_t GC_make_closure(fn, data)
+    GC_finalization_proc fn;
+    extern_ptr_t data;
+# endif
+{
+    struct closure * result =
+    		(struct closure *) GC_malloc(sizeof (struct closure));
+    
+    result -> cl_fn = fn;
+    result -> cl_data = data;
+    return((extern_ptr_t)result);
+}
+
+# ifdef __STDC__
+    void GC_debug_invoke_finalizer(void * obj, void * data)
+# else
+    void GC_debug_invoke_finalizer(obj, data)
+    char * obj;
+    char * data;
+# endif
+{
+    register struct closure * cl = (struct closure *) data;
+    
+    (*(cl -> cl_fn))((extern_ptr_t)((char *)obj + sizeof(oh)), cl -> cl_data);
+} 
+
diff --git a/dyn_load.c b/dyn_load.c
new file mode 100644
index 00000000..28817b0c
--- /dev/null
+++ b/dyn_load.c
@@ -0,0 +1,530 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ * Original author: Bill Janssen
+ * Heavily modified by Hans Boehm and others
+ */
+/* Boehm, May 19, 1994 1:57 pm PDT */
+
+/*
+ * This is incredibly OS specific code for tracking down data sections in
+ * dynamic libraries.  There appears to be no way of doing this quickly
+ * without groveling through undocumented data structures.  We would argue
+ * that this is a bug in the design of the dlopen interface.  THIS CODE
+ * MAY BREAK IN FUTURE OS RELEASES.  If this matters to you, don't hesitate
+ * to let your vendor know ...
+ *
+ * None of this is safe with dlclose and incremental collection.
+ * But then not much of anything is safe in the presence of dlclose.
+ */
+#include <sys/types.h>
+#include "gc_priv.h"
+
+#if (defined(DYNAMIC_LOADING) || defined(MSWIN32)) && !defined(PCR)
+#if !defined(SUNOS4) && !defined(SUNOS5) && !defined(IRIX5) && !defined(MSWIN32)
+ --> We only know how to find data segments of dynamic libraries under SunOS,
+ --> IRIX5 and Win32.  Additional SVR4 variants might not be too hard to add.
+#endif
+
+#include <stdio.h>
+#ifdef SUNOS5
+#   include <sys/elf.h>
+#   include <dlfcn.h>
+#   include <link.h>
+#endif
+#ifdef SUNOS4
+#   include <dlfcn.h>
+#   include <link.h>
+#   include <a.out.h>
+  /* struct link_map field overrides */
+#   define l_next	lm_next
+#   define l_addr	lm_addr
+#   define l_name	lm_name
+#endif
+
+
+#ifdef SUNOS5
+
+#ifdef LINT
+    Elf32_Dyn _DYNAMIC;
+#endif
+
+static struct link_map *
+GC_FirstDLOpenedLinkMap()
+{
+    extern Elf32_Dyn _DYNAMIC;
+    Elf32_Dyn *dp;
+    struct r_debug *r;
+    static struct link_map * cachedResult = 0;
+
+    if( &_DYNAMIC == 0) {
+        return(0);
+    }
+    if( cachedResult == 0 ) {
+        int tag;
+        for( dp = ((Elf32_Dyn *)(&_DYNAMIC)); (tag = dp->d_tag) != 0; dp++ ) {
+            if( tag == DT_DEBUG ) {
+                struct link_map *lm
+                        = ((struct r_debug *)(dp->d_un.d_ptr))->r_map;
+                if( lm != 0 ) cachedResult = lm->l_next; /* might be NIL */
+                break;
+            }
+        }
+    }
+    return cachedResult;
+}
+
+#endif
+
+#ifdef SUNOS4
+
+#ifdef LINT
+    struct link_dynamic _DYNAMIC;
+#endif
+
+static struct link_map *
+GC_FirstDLOpenedLinkMap()
+{
+    extern struct link_dynamic _DYNAMIC;
+
+    if( &_DYNAMIC == 0) {
+        return(0);
+    }
+    return(_DYNAMIC.ld_un.ld_1->ld_loaded);
+}
+
+/* Return the address of the ld.so allocated common symbol	*/
+/* with the least address, or 0 if none.			*/
+static ptr_t GC_first_common()
+{
+    ptr_t result = 0;
+    extern struct link_dynamic _DYNAMIC;
+    struct rtc_symb * curr_symbol;
+    
+    if( &_DYNAMIC == 0) {
+        return(0);
+    }
+    curr_symbol = _DYNAMIC.ldd -> ldd_cp;
+    for (; curr_symbol != 0; curr_symbol = curr_symbol -> rtc_next) {
+        if (result == 0
+            || (ptr_t)(curr_symbol -> rtc_sp -> n_value) < result) {
+            result = (ptr_t)(curr_symbol -> rtc_sp -> n_value);
+        }
+    }
+    return(result);
+}
+
+#endif
+
+# if defined(SUNOS4) || defined(SUNOS5)
+/* Add dynamic library data sections to the root set.		*/
+# if !defined(PCR) && !defined(SOLARIS_THREADS) && defined(THREADS)
+#   ifndef SRC_M3
+	--> fix mutual exclusion with dlopen
+#   endif  /* We assume M3 programs don't call dlopen for now */
+# endif
+
+# ifdef SOLARIS_THREADS
+  /* Redefine dlopen to guarantee mutual exclusion with	*/
+  /* GC_register_dynamic_libraries.			*/
+  /* assumes that dlopen doesn't need to call GC_malloc	*/
+  /* and friends.					*/
+# include <thread.h>
+# include <synch.h>
+  
+void * GC_dlopen(const char *path, int mode)
+{
+    void * result;
+    
+    mutex_lock(&GC_allocate_ml);
+    result = dlopen(path, mode);
+    mutex_unlock(&GC_allocate_ml);
+    return(result);
+}
+# endif
+
+void GC_register_dynamic_libraries()
+{
+  struct link_map *lm = GC_FirstDLOpenedLinkMap();
+  
+
+  for (lm = GC_FirstDLOpenedLinkMap();
+       lm != (struct link_map *) 0;  lm = lm->l_next)
+    {
+#     ifdef SUNOS4
+	struct exec *e;
+	 
+        e = (struct exec *) lm->lm_addr;
+        GC_add_roots_inner(
+      		    ((char *) (N_DATOFF(*e) + lm->lm_addr)),
+		    ((char *) (N_BSSADDR(*e) + e->a_bss + lm->lm_addr)));
+#     endif
+#     ifdef SUNOS5
+	Elf32_Ehdr * e;
+        Elf32_Phdr * p;
+        unsigned long offset;
+        char * start;
+        register int i;
+        
+	e = (Elf32_Ehdr *) lm->l_addr;
+        p = ((Elf32_Phdr *)(((char *)(e)) + e->e_phoff));
+        offset = ((unsigned long)(lm->l_addr));
+        for( i = 0; i < (int)(e->e_phnum); ((i++),(p++)) ) {
+          switch( p->p_type ) {
+            case PT_LOAD:
+              {
+                if( !(p->p_flags & PF_W) ) break;
+                start = ((char *)(p->p_vaddr)) + offset;
+                GC_add_roots_inner(
+                  start,
+                  start + p->p_memsz
+                );
+              }
+              break;
+            default:
+              break;
+          }
+	}
+#     endif
+    }
+#   ifdef SUNOS4
+      {
+      	static ptr_t common_start = 0;
+      	ptr_t common_end;
+      	extern ptr_t GC_find_limit();
+      	
+      	if (common_start == 0) common_start = GC_first_common();
+      	if (common_start != 0) {
+      	    common_end = GC_find_limit(common_start, TRUE);
+      	    GC_add_roots_inner((char *)common_start, (char *)common_end);
+      	}
+      }
+#   endif
+}
+
+# endif /* SUNOS */
+
+#ifdef IRIX5
+
+#include <sys/procfs.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <elf.h>
+
+extern void * GC_roots_present();
+
+extern ptr_t GC_scratch_end_ptr;   /* End of GC_scratch_alloc arena	*/
+
+/* We use /proc to track down all parts of the address space that are	*/
+/* mapped by the process, and throw out regions we know we shouldn't	*/
+/* worry about.  This may also work under other SVR4 variants.		*/
+void GC_register_dynamic_libraries()
+{
+    static int fd = -1;
+    char buf[30];
+    static prmap_t * addr_map = 0;
+    static int current_sz = 0;	/* Number of records currently in addr_map */
+    static int needed_sz;	/* Required size of addr_map		*/
+    register int i;
+    register long flags;
+    register ptr_t start;
+    register ptr_t limit;
+    ptr_t heap_end = (ptr_t)DATASTART;
+
+    if (fd < 0) {
+      sprintf(buf, "/proc/%d", getpid());
+      fd = open(buf, O_RDONLY);
+      if (fd < 0) {
+    	ABORT("/proc open failed");
+      }
+    }
+    if (ioctl(fd, PIOCNMAP, &needed_sz) < 0) {
+    	ABORT("/proc PIOCNMAP ioctl failed");
+    }
+    if (needed_sz >= current_sz) {
+        current_sz = needed_sz * 2 + 1;
+        		/* Expansion, plus room for 0 record */
+        addr_map = (prmap_t *)GC_scratch_alloc(current_sz * sizeof(prmap_t));
+    }
+    if (ioctl(fd, PIOCMAP, addr_map) < 0) {
+    	ABORT("/proc PIOCMAP ioctl failed");
+    };
+    if (GC_n_heap_sects > 0) {
+    	heap_end = GC_heap_sects[GC_n_heap_sects-1].hs_start
+    			+ GC_heap_sects[GC_n_heap_sects-1].hs_bytes;
+    	if (heap_end < GC_scratch_end_ptr) heap_end = GC_scratch_end_ptr; 
+    }
+    for (i = 0; i < needed_sz; i++) {
+        flags = addr_map[i].pr_mflags;
+        if ((flags & (MA_BREAK | MA_STACK | MA_PHYS)) != 0) goto irrelevant;
+        if ((flags & (MA_READ | MA_WRITE)) != (MA_READ | MA_WRITE))
+            goto irrelevant;
+          /* The latter test is empirically useless.  Other than the	*/
+          /* main data and stack segments, everything appears to be	*/
+          /* mapped readable, writable, executable, and shared(!!).	*/
+          /* This makes no sense to me.	- HB				*/
+        start = (ptr_t)(addr_map[i].pr_vaddr);
+        if (GC_roots_present(start)) goto irrelevant;
+        if (start < heap_end && start >= (ptr_t)DATASTART)
+        	goto irrelevant;
+        limit = start + addr_map[i].pr_size;
+	if (addr_map[i].pr_off == 0 && strncmp(start, ELFMAG, 4) == 0) {
+	    /* Discard text segments, i.e. 0-offset mappings against	*/
+	    /* executable files which appear to have ELF headers.	*/
+	    caddr_t arg;
+	    int obj;
+#	    define MAP_IRR_SZ 10
+	    static ptr_t map_irr[MAP_IRR_SZ];
+	    				/* Known irrelevant map entries	*/
+	    static int n_irr = 0;
+	    struct stat buf;
+	    register int i;
+	    
+	    for (i = 0; i < n_irr; i++) {
+	        if (map_irr[i] == start) goto irrelevant;
+	    }
+	    arg = (caddr_t)start;
+	    obj = ioctl(fd, PIOCOPENM, &arg);
+	    if (obj >= 0) {
+	        fstat(obj, &buf);
+	        close(obj);
+	        if ((buf.st_mode & 0111) != 0) {
+	            if (n_irr < MAP_IRR_SZ) {
+	                map_irr[n_irr++] = start;
+	            }
+	            goto irrelevant;
+	        }
+	    }
+	}
+        GC_add_roots_inner(start, limit);
+      irrelevant: ;
+    }
+}
+
+#endif  /* IRIX5 */
+
+# ifdef MSWIN32
+
+# define WIN32_LEAN_AND_MEAN
+# define NOSERVICE
+# include <windows.h>
+# include <stdlib.h>
+
+  /* We traverse the entire address space and register all segments 	*/
+  /* that could possibly have been written to.				*/
+  DWORD GC_allocation_granularity;
+  
+  extern bool GC_is_heap_base (ptr_t p);
+  
+  void GC_cond_add_roots(char *base, char * limit)
+  {
+    char dummy;
+    char * stack_top
+           = (char *) ((word)(&dummy) & ~(GC_allocation_granularity-1));
+    if (base == limit) return;
+    if (limit > stack_top && base < GC_stackbottom) {
+    	/* Part of the stack; ignore it. */
+    	return;
+    }
+    GC_add_roots_inner(base, limit);
+  }
+  
+  extern bool GC_win32s;
+  
+  void GC_register_dynamic_libraries()
+  {
+    MEMORY_BASIC_INFORMATION buf;
+    SYSTEM_INFO sysinfo;
+    DWORD result;
+    DWORD protect;
+    LPVOID p;
+    char * base;
+    char * limit, * new_limit;
+    
+    if (GC_win32s) return;
+    GetSystemInfo(&sysinfo);
+    base = limit = p = sysinfo.lpMinimumApplicationAddress;
+    GC_allocation_granularity = sysinfo.dwAllocationGranularity;
+    while (p < sysinfo.lpMaximumApplicationAddress) {
+        result = VirtualQuery(p, &buf, sizeof(buf));
+        if (result != sizeof(buf)) {
+            ABORT("Weird VirtualQuery result");
+        }
+        new_limit = (char *)p + buf.RegionSize;
+        protect = buf.Protect;
+        if (buf.State == MEM_COMMIT
+            && (protect == PAGE_EXECUTE_READWRITE
+                || protect == PAGE_READWRITE)
+            && !GC_is_heap_base(buf.AllocationBase)) {
+            if ((char *)p == limit) {
+                limit = new_limit;
+            } else {
+                GC_cond_add_roots(base, limit);
+                base = p;
+                limit = new_limit;
+            }
+        }
+        if (p > (LPVOID)new_limit /* overflow */) break;
+        p = (LPVOID)new_limit;
+    }
+    GC_cond_add_roots(base, limit);
+  }
+
+#endif /* MSWIN32 */
+
+#if defined(ALPHA)
+void GC_register_dynamic_libraries()
+{
+  int status;
+  ldr_process_t mypid;
+
+  /* module */
+    ldr_module_t moduleid = LDR_NULL_MODULE;
+    ldr_module_info_t moduleinfo;
+    size_t moduleinfosize = sizeof(moduleinfo);
+    size_t modulereturnsize;    
+
+  /* region */
+    ldr_region_t region; 
+    ldr_region_info_t regioninfo;
+    size_t regioninfosize = sizeof(regioninfo);
+    size_t regionreturnsize;
+
+  /* Obtain id of this process */
+    mypid = ldr_my_process();
+  
+  /* For each module */
+    while (TRUE) {
+
+      /* Get the next (first) module */
+        status = ldr_next_module(mypid, &moduleid);
+
+      /* Any more modules? */
+        if (moduleid == LDR_NULL_MODULE)
+            break;    /* No more modules */
+
+      /* Check status AFTER checking moduleid because */
+      /* of a bug in the non-shared ldr_next_module stub */
+        if (status != 0 ) {
+            GC_printf("dynamic_load: status = %ld\n", (long)status);
+            {
+                extern char *sys_errlist[];
+                extern int sys_nerr;
+                extern int errno;
+                if (errno <= sys_nerr) {
+                    GC_printf("dynamic_load: %s\n", sys_errlist[errno]);
+               } else {
+                    GC_printf("dynamic_load: %d\n", errno);
+                }
+        }
+            ABORT("ldr_next_module failed");
+         }
+
+      /* Get the module information */
+        status = ldr_inq_module(mypid, moduleid, &moduleinfo,
+                                moduleinfosize, &modulereturnsize); 
+        if (status != 0 )
+            ABORT("ldr_inq_module failed");
+
+      /* is module for the main program (i.e. nonshared portion)? */
+          if (moduleinfo.lmi_flags & LDR_MAIN)
+              continue;    /* skip the main module */
+
+#     ifdef VERBOSE
+          GC_printf("---Module---\n");
+          GC_printf("Module ID            = %16ld\n", moduleinfo.lmi_modid);
+          GC_printf("Count of regions     = %16d\n", moduleinfo.lmi_nregion);
+          GC_printf("flags for module     = %16lx\n", moduleinfo.lmi_flags); 
+          GC_printf("pathname of module   = \"%s\"\n", moduleinfo.lmi_name);
+#     endif
+
+      /* For each region in this module */
+        for (region = 0; region < moduleinfo.lmi_nregion; region++) {
+
+          /* Get the region information */
+            status = ldr_inq_region(mypid, moduleid, region, &regioninfo,
+                                    regioninfosize, &regionreturnsize);
+            if (status != 0 )
+                ABORT("ldr_inq_region failed");
+
+          /* only process writable (data) regions */
+            if (! (regioninfo.lri_prot & LDR_W))
+                continue;
+
+#         ifdef VERBOSE
+              GC_printf("--- Region ---\n");
+              GC_printf("Region number    = %16ld\n",
+              	        regioninfo.lri_region_no);
+              GC_printf("Protection flags = %016x\n",  regioninfo.lri_prot);
+              GC_printf("Virtual address  = %16p\n",   regioninfo.lri_vaddr);
+              GC_printf("Mapped address   = %16p\n",   regioninfo.lri_mapaddr);
+              GC_printf("Region size      = %16ld\n",  regioninfo.lri_size);
+              GC_printf("Region name      = \"%s\"\n", regioninfo.lri_name);
+#         endif
+
+          /* register region as a garbage collection root */
+            GC_add_roots_inner (
+                (char *)regioninfo.lri_mapaddr,
+                (char *)regioninfo.lri_mapaddr + regioninfo.lri_size);
+
+        }
+    }
+}
+#endif
+
+
+#else /* !DYNAMIC_LOADING */
+
+#ifdef PCR
+
+#   include "il/PCR_IL.h"
+#   include "th/PCR_ThCtl.h"
+#   include "mm/PCR_MM.h"
+
+void GC_register_dynamic_libraries()
+{
+    /* Add new static data areas of dynamically loaded modules.	*/
+        {
+          PCR_IL_LoadedFile * p = PCR_IL_GetLastLoadedFile();
+          PCR_IL_LoadedSegment * q;
+          
+          /* Skip uncommited files */
+          while (p != NIL && !(p -> lf_commitPoint)) {
+              /* The loading of this file has not yet been committed	*/
+              /* Hence its description could be inconsistent.  		*/
+              /* Furthermore, it hasn't yet been run.  Hence its data	*/
+              /* segments can't possibly reference heap allocated	*/
+              /* objects.						*/
+              p = p -> lf_prev;
+          }
+          for (; p != NIL; p = p -> lf_prev) {
+            for (q = p -> lf_ls; q != NIL; q = q -> ls_next) {
+              if ((q -> ls_flags & PCR_IL_SegFlags_Traced_MASK)
+                  == PCR_IL_SegFlags_Traced_on) {
+                GC_add_roots_inner
+                	((char *)(q -> ls_addr), 
+                	 (char *)(q -> ls_addr) + q -> ls_bytes);
+              }
+            }
+          }
+        }
+}
+
+
+#else /* !PCR */
+
+void GC_register_dynamic_libraries(){}
+
+int GC_no_dynamic_loading;
+
+#endif /* !PCR */
+#endif /* !DYNAMIC_LOADING */
diff --git a/finalize.c b/finalize.c
new file mode 100644
index 00000000..45339b3c
--- /dev/null
+++ b/finalize.c
@@ -0,0 +1,523 @@
+/*
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:08 pm PDT */
+# define I_HIDE_POINTERS
+# include "gc.h"
+# include "gc_priv.h"
+# include "gc_mark.h"
+
+# define HASH3(addr,size,log_size) \
+    ((((word)(addr) >> 3) ^ ((word)(addr) >> (3+(log_size)))) \
+    & ((size) - 1))
+#define HASH2(addr,log_size) HASH3(addr, 1 << log_size, log_size)
+
+struct hash_chain_entry {
+    word hidden_key;
+    struct hash_chain_entry * next;
+};
+
+unsigned GC_finalization_failures = 0;
+	/* Number of finalization requests that failed for lack of memory. */
+
+static struct disappearing_link {
+    struct hash_chain_entry prolog;
+#   define dl_hidden_link prolog.hidden_key
+				/* Field to be cleared.		*/
+#   define dl_next(x) (struct disappearing_link *)((x) -> prolog.next)
+#   define dl_set_next(x,y) (x) -> prolog.next = (struct hash_chain_entry *)(y)
+
+    word dl_hidden_obj;		/* Pointer to object base	*/
+} **dl_head = 0;
+
+static signed_word log_dl_table_size = -1;
+			/* Binary log of				*/
+			/* current size of array pointed to by dl_head.	*/
+			/* -1 ==> size is 0.				*/
+
+word GC_dl_entries = 0;	/* Number of entries currently in disappearing	*/
+			/* link table.					*/
+
+static struct finalizable_object {
+    struct hash_chain_entry prolog;
+#   define fo_hidden_base prolog.hidden_key
+				/* Pointer to object base.	*/
+#   define fo_next(x) (struct finalizable_object *)((x) -> prolog.next)
+#   define fo_set_next(x,y) (x) -> prolog.next = (struct hash_chain_entry *)(y)
+    GC_finalization_proc fo_fn;	/* Finalizer.			*/
+    ptr_t fo_client_data;
+    word fo_object_size;	/* In bytes.			*/
+} **fo_head = 0;
+
+struct finalizable_object * GC_finalize_now = 0;
+	/* LIst of objects that should be finalized now.	*/
+
+static signed_word log_fo_table_size = -1;
+
+word GC_fo_entries = 0;
+
+# ifdef SRC_M3
+void GC_push_finalizer_structures()
+{
+    GC_push_all((ptr_t)(&dl_head), (ptr_t)(&dl_head) + sizeof(word));
+    GC_push_all((ptr_t)(&fo_head), (ptr_t)(&fo_head) + sizeof(word));
+}
+# endif
+
+# define ALLOC(x, t) t *x = GC_NEW(t)
+
+/* Double the size of a hash table. *size_ptr is the log of its current	*/
+/* size.  May be a noop.						*/
+/* *table is a pointer to an array of hash headers.  If we succeed, we	*/
+/* update both *table and *log_size_ptr.				*/
+/* Lock is held.  Signals are disabled.					*/
+void GC_grow_table(table, log_size_ptr)
+struct hash_chain_entry ***table;
+signed_word * log_size_ptr;
+{
+    register word i;
+    register struct hash_chain_entry *p;
+    int log_old_size = *log_size_ptr;
+    register int log_new_size = log_old_size + 1;
+    word old_size = ((log_old_size == -1)? 0: (1 << log_old_size));
+    register word new_size = 1 << log_new_size;
+    struct hash_chain_entry **new_table = (struct hash_chain_entry **)
+    	GC_malloc_ignore_off_page_inner(
+    		(size_t)new_size * sizeof(struct hash_chain_entry *));
+    
+    if (new_table == 0) {
+    	if (table == 0) {
+    	    ABORT("Insufficient space for initial table allocation");
+    	} else {
+    	    return;
+    	}
+    }
+    for (i = 0; i < old_size; i++) {
+      p = (*table)[i];
+      while (p != 0) {
+        register ptr_t real_key = (ptr_t)REVEAL_POINTER(p -> hidden_key);
+        register struct hash_chain_entry *next = p -> next;
+        register int new_hash = HASH3(real_key, new_size, log_new_size);
+        
+        p -> next = new_table[new_hash];
+        new_table[new_hash] = p;
+        p = next;
+      }
+    }
+    *log_size_ptr = log_new_size;
+    *table = new_table;
+}
+
+
+int GC_register_disappearing_link(link)
+extern_ptr_t * link;
+{
+    ptr_t base;
+    
+    base = (ptr_t)GC_base((extern_ptr_t)link);
+    if (base == 0)
+    	ABORT("Bad arg to GC_register_disappearing_link");
+    return(GC_general_register_disappearing_link(link, base));
+}
+
+int GC_general_register_disappearing_link(link, obj)
+extern_ptr_t * link;
+extern_ptr_t obj;
+{
+    struct disappearing_link *curr_dl;
+    int index;
+    struct disappearing_link * new_dl;
+    DCL_LOCK_STATE;
+    
+    if ((word)link & (ALIGNMENT-1))
+    	ABORT("Bad arg to GC_general_register_disappearing_link");
+#   ifdef THREADS
+    	DISABLE_SIGNALS();
+    	LOCK();
+#   endif
+    if (log_dl_table_size == -1
+        || GC_dl_entries > ((word)1 << log_dl_table_size)) {
+#	ifndef THREADS
+	    DISABLE_SIGNALS();
+#	endif
+    	GC_grow_table((struct hash_chain_entry ***)(&dl_head),
+    		      &log_dl_table_size);
+#	ifdef PRINTSTATS
+	    GC_printf1("Grew dl table to %lu entries\n",
+	    		(unsigned long)(1 << log_dl_table_size));
+#	endif
+#	ifndef THREADS
+	    ENABLE_SIGNALS();
+#	endif
+    }
+    index = HASH2(link, log_dl_table_size);
+    curr_dl = dl_head[index];
+    for (curr_dl = dl_head[index]; curr_dl != 0; curr_dl = dl_next(curr_dl)) {
+        if (curr_dl -> dl_hidden_link == HIDE_POINTER(link)) {
+            curr_dl -> dl_hidden_obj = HIDE_POINTER(obj);
+#	    ifdef THREADS
+                UNLOCK();
+    	        ENABLE_SIGNALS();
+#	    endif
+            return(1);
+        }
+    }
+#   ifdef THREADS
+      new_dl = (struct disappearing_link *)
+    	GC_generic_malloc_inner(sizeof(struct disappearing_link),NORMAL);
+#   else
+      new_dl = GC_NEW(struct disappearing_link);
+#   endif
+    if (new_dl != 0) {
+        new_dl -> dl_hidden_obj = HIDE_POINTER(obj);
+        new_dl -> dl_hidden_link = HIDE_POINTER(link);
+        dl_set_next(new_dl, dl_head[index]);
+        dl_head[index] = new_dl;
+        GC_dl_entries++;
+    } else {
+        GC_finalization_failures++;
+    }
+#   ifdef THREADS
+        UNLOCK();
+        ENABLE_SIGNALS();
+#   endif
+    return(0);
+}
+
+int GC_unregister_disappearing_link(link)
+extern_ptr_t * link;
+{
+    struct disappearing_link *curr_dl, *prev_dl;
+    int index;
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    index = HASH2(link, log_dl_table_size);
+    if (((unsigned long)link & (ALIGNMENT-1))) goto out;
+    prev_dl = 0; curr_dl = dl_head[index];
+    while (curr_dl != 0) {
+        if (curr_dl -> dl_hidden_link == HIDE_POINTER(link)) {
+            if (prev_dl == 0) {
+                dl_head[index] = dl_next(curr_dl);
+            } else {
+                dl_set_next(prev_dl, dl_next(curr_dl));
+            }
+            GC_dl_entries--;
+            UNLOCK();
+    	    ENABLE_SIGNALS();
+            GC_free((extern_ptr_t)curr_dl);
+            return(1);
+        }
+        prev_dl = curr_dl;
+        curr_dl = dl_next(curr_dl);
+    }
+out:
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return(0);
+}
+
+/* Register a finalization function.  See gc.h for details.	*/
+/* in the nonthreads case, we try to avoid disabling signals,	*/
+/* since it can be expensive.  Threads packages typically	*/
+/* make it cheaper.						*/
+void GC_register_finalizer(obj, fn, cd, ofn, ocd)
+extern_ptr_t obj;
+GC_finalization_proc fn;
+extern_ptr_t cd;
+GC_finalization_proc * ofn;
+extern_ptr_t * ocd;
+{
+    ptr_t base;
+    struct finalizable_object * curr_fo, * prev_fo;
+    int index;
+    struct finalizable_object *new_fo;
+    DCL_LOCK_STATE;
+
+#   ifdef THREADS
+	DISABLE_SIGNALS();
+	LOCK();
+#   endif
+    if (log_fo_table_size == -1
+        || GC_fo_entries > ((word)1 << log_fo_table_size)) {
+#	ifndef THREADS
+    	    DISABLE_SIGNALS();
+#	endif
+    	GC_grow_table((struct hash_chain_entry ***)(&fo_head),
+    		      &log_fo_table_size);
+#	ifdef PRINTSTATS
+	    GC_printf1("Grew fo table to %lu entries\n",
+	    		(unsigned long)(1 << log_fo_table_size));
+#	endif
+#	ifndef THREADS
+	    ENABLE_SIGNALS();
+#	endif
+    }
+    /* in the THREADS case signals are disabled and we hold allocation	*/
+    /* lock; otherwise neither is true.  Proceed carefully.		*/
+    base = (ptr_t)obj;
+    index = HASH2(base, log_fo_table_size);
+    prev_fo = 0; curr_fo = fo_head[index];
+    while (curr_fo != 0) {
+        if (curr_fo -> fo_hidden_base == HIDE_POINTER(base)) {
+            /* Interruption by a signal in the middle of this	*/
+            /* should be safe.  The client may see only *ocd	*/
+            /* updated, but we'll declare that to be his	*/
+            /* problem.						*/
+            if (ocd) *ocd = (extern_ptr_t) curr_fo -> fo_client_data;
+            if (ofn) *ofn = curr_fo -> fo_fn;
+            /* Delete the structure for base. */
+                if (prev_fo == 0) {
+                  fo_head[index] = fo_next(curr_fo);
+                } else {
+                  fo_set_next(prev_fo, fo_next(curr_fo));
+                }
+            if (fn == 0) {
+                GC_fo_entries--;
+                  /* May not happen if we get a signal.  But a high	*/
+                  /* estimate will only make the table larger than	*/
+                  /* necessary.						*/
+#		ifndef THREADS
+                  GC_free((extern_ptr_t)curr_fo);
+#		endif
+            } else {
+                curr_fo -> fo_fn = fn;
+                curr_fo -> fo_client_data = (ptr_t)cd;
+		/* Reinsert it.  We deleted it first to maintain	*/
+		/* consistency in the event of a signal.		*/
+		if (prev_fo == 0) {
+                  fo_head[index] = curr_fo;
+                } else {
+                  fo_set_next(prev_fo, curr_fo);
+                }
+            }
+#	    ifdef THREADS
+                UNLOCK();
+    	    	ENABLE_SIGNALS();
+#	    endif
+            return;
+        }
+        prev_fo = curr_fo;
+        curr_fo = fo_next(curr_fo);
+    }
+    if (ofn) *ofn = 0;
+    if (ocd) *ocd = 0;
+    if (fn == 0) {
+#	ifdef THREADS
+            UNLOCK();
+    	    ENABLE_SIGNALS();
+#	endif
+        return;
+    }
+#   ifdef THREADS
+      new_fo = (struct finalizable_object *)
+    	GC_generic_malloc_inner(sizeof(struct finalizable_object),NORMAL);
+#   else
+      new_fo = GC_NEW(struct finalizable_object);
+#   endif
+    if (new_fo != 0) {
+        new_fo -> fo_hidden_base = (word)HIDE_POINTER(base);
+	new_fo -> fo_fn = fn;
+	new_fo -> fo_client_data = (ptr_t)cd;
+	new_fo -> fo_object_size = GC_size(base);
+	fo_set_next(new_fo, fo_head[index]);
+	GC_fo_entries++;
+	fo_head[index] = new_fo;
+    } else {
+     	GC_finalization_failures++;
+    }
+#   ifdef THREADS
+        UNLOCK();
+    	ENABLE_SIGNALS();
+#   endif
+}
+
+/* Called with world stopped.  Cause disappearing links to disappear,	*/
+/* and invoke finalizers.						*/
+void GC_finalize()
+{
+    struct disappearing_link * curr_dl, * prev_dl, * next_dl;
+    struct finalizable_object * curr_fo, * prev_fo, * next_fo;
+    ptr_t real_ptr, real_link;
+    register int i;
+    int dl_size = 1 << log_dl_table_size;
+    int fo_size = 1 << log_fo_table_size;
+    
+  /* Make disappearing links disappear */
+    for (i = 0; i < dl_size; i++) {
+      curr_dl = dl_head[i];
+      prev_dl = 0;
+      while (curr_dl != 0) {
+        real_ptr = (ptr_t)REVEAL_POINTER(curr_dl -> dl_hidden_obj);
+        real_link = (ptr_t)REVEAL_POINTER(curr_dl -> dl_hidden_link);
+        if (!GC_is_marked(real_ptr)) {
+            *(word *)real_link = 0;
+            next_dl = dl_next(curr_dl);
+            if (prev_dl == 0) {
+                dl_head[i] = next_dl;
+            } else {
+                dl_set_next(prev_dl, next_dl);
+            }
+            GC_clear_mark_bit((ptr_t)curr_dl);
+            GC_dl_entries--;
+            curr_dl = next_dl;
+        } else {
+            prev_dl = curr_dl;
+            curr_dl = dl_next(curr_dl);
+        }
+      }
+    }
+  /* Mark all objects reachable via chains of 1 or more pointers	*/
+  /* from finalizable objects.						*/
+#   ifdef PRINTSTATS
+        if (GC_mark_state != MS_NONE) ABORT("Bad mark state");
+#   endif
+    for (i = 0; i < fo_size; i++) {
+      for (curr_fo = fo_head[i]; curr_fo != 0; curr_fo = fo_next(curr_fo)) {
+        real_ptr = (ptr_t)REVEAL_POINTER(curr_fo -> fo_hidden_base);
+        if (!GC_is_marked(real_ptr)) {
+            hdr * hhdr = HDR(real_ptr);
+            
+            PUSH_OBJ((word *)real_ptr, hhdr, GC_mark_stack_top,
+	             &(GC_mark_stack[GC_mark_stack_size]));
+            while (!GC_mark_stack_empty()) GC_mark_from_mark_stack();
+            if (GC_mark_state != MS_NONE) {
+                /* Mark stack overflowed. Very unlikely. */
+#		ifdef PRINTSTATS
+		    if (GC_mark_state != MS_INVALID) ABORT("Bad mark state");
+		    GC_printf0("Mark stack overflowed in finalization!!\n");
+#		endif
+		/* Make mark bits consistent again.  Forget about	*/
+		/* finalizing this object for now.			*/
+		    GC_set_mark_bit(real_ptr);
+		    while (!GC_mark_some());
+            }
+            /* 
+            if (GC_is_marked(real_ptr)) {
+                --> Report finalization cycle here, if desired
+            }
+            */
+        }
+        
+      }
+    }
+  /* Enqueue for finalization all objects that are still		*/
+  /* unreachable.							*/
+    for (i = 0; i < fo_size; i++) {
+      curr_fo = fo_head[i];
+      prev_fo = 0;
+      while (curr_fo != 0) {
+        real_ptr = (ptr_t)REVEAL_POINTER(curr_fo -> fo_hidden_base);
+        if (!GC_is_marked(real_ptr)) {
+            GC_set_mark_bit(real_ptr);
+            /* Delete from hash table */
+              next_fo = fo_next(curr_fo);
+              if (prev_fo == 0) {
+                fo_head[i] = next_fo;
+              } else {
+                fo_set_next(prev_fo, next_fo);
+              }
+              GC_fo_entries--;
+            /* Add to list of objects awaiting finalization.	*/
+              fo_set_next(curr_fo, GC_finalize_now);
+              GC_finalize_now = curr_fo;
+#	    ifdef PRINTSTATS
+              if (!GC_is_marked((ptr_t)curr_fo)) {
+                ABORT("GC_finalize: found accessible unmarked object\n");
+              }
+#	    endif
+            curr_fo = next_fo;
+        } else {
+            prev_fo = curr_fo;
+            curr_fo = fo_next(curr_fo);
+        }
+      }
+    }
+  /* Remove dangling disappearing links. */
+    for (i = 0; i < dl_size; i++) {
+      curr_dl = dl_head[i];
+      prev_dl = 0;
+      while (curr_dl != 0) {
+        real_link = GC_base((ptr_t)REVEAL_POINTER(curr_dl -> dl_hidden_link));
+        if (real_link != 0 && !GC_is_marked(real_link)) {
+            next_dl = dl_next(curr_dl);
+            if (prev_dl == 0) {
+                dl_head[i] = next_dl;
+            } else {
+                dl_set_next(prev_dl, next_dl);
+            }
+            GC_clear_mark_bit((ptr_t)curr_dl);
+            GC_dl_entries--;
+            curr_dl = next_dl;
+        } else {
+            prev_dl = curr_dl;
+            curr_dl = dl_next(curr_dl);
+        }
+      }
+    }
+}
+
+/* Invoke finalizers for all objects that are ready to be finalized.	*/
+/* Should be called without allocation lock.				*/
+void GC_invoke_finalizers()
+{
+    ptr_t real_ptr;
+    register struct finalizable_object * curr_fo;
+    DCL_LOCK_STATE;
+    
+    while (GC_finalize_now != 0) {
+#	ifdef THREADS
+	    DISABLE_SIGNALS();
+	    LOCK();
+#	endif
+    	curr_fo = GC_finalize_now;
+#	ifdef THREADS
+ 	    if (curr_fo != 0) GC_finalize_now = fo_next(curr_fo);
+	    UNLOCK();
+	    ENABLE_SIGNALS();
+	    if (curr_fo == 0) break;
+#	else
+	    GC_finalize_now = fo_next(curr_fo);
+#	endif
+    	real_ptr = (ptr_t)REVEAL_POINTER(curr_fo -> fo_hidden_base);
+    	(*(curr_fo -> fo_fn))(real_ptr, curr_fo -> fo_client_data);
+#	ifndef THREADS
+    	    GC_free((extern_ptr_t)curr_fo);
+#	endif
+    }
+}
+
+# ifdef __STDC__
+    extern_ptr_t GC_call_with_alloc_lock(GC_fn_type fn, extern_ptr_t client_data)
+# else
+    extern_ptr_t GC_call_with_alloc_lock(fn, client_data)
+    GC_fn_type fn;
+    extern_ptr_t client_data;
+# endif
+{
+    extern_ptr_t result;
+    DCL_LOCK_STATE;
+    
+#   ifdef THREADS
+      DISABLE_SIGNALS();
+      LOCK();
+#   endif
+    result = (*fn)(client_data);
+#   ifdef THREADS
+      UNLOCK();
+      ENABLE_SIGNALS();
+#   endif
+    return(result);
+}
+
diff --git a/gc.h b/gc.h
new file mode 100644
index 00000000..65a26093
--- /dev/null
+++ b/gc.h
@@ -0,0 +1,449 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:13 pm PDT */
+ 
+#ifndef _GC_H
+
+# define _GC_H
+
+# include <stddef.h>
+
+/* Define word and signed_word to be unsigned and signed types of the 	*/
+/* size as char * or void *.  There seems to be no way to do this	*/
+/* even semi-portably.  The following is probably no better/worse 	*/
+/* than almost anything else.						*/
+/* The ANSI standard suggests that size_t and ptr_diff_t might be 	*/
+/* better choices.  But those appear to have incorrect definitions	*/
+/* on may systems.  Notably "typedef int size_t" seems to be both	*/
+/* frequent and WRONG.							*/
+typedef unsigned long GC_word;
+typedef long GC_signed_word;
+
+/* Public read-only variables */
+
+extern GC_word GC_gc_no;/* Counter incremented per collection.  	*/
+			/* Includes empty GCs at startup.		*/
+			
+
+/* Public R/W variables */
+
+extern int GC_quiet;	/* Disable statistics output.  Only matters if	*/
+			/* collector has been compiled with statistics	*/
+			/* enabled.  This involves a performance cost,	*/
+			/* and is thus not the default.			*/
+
+extern int GC_dont_gc;	/* Dont collect unless explicitly requested, e.g. */
+			/* beacuse it's not safe.			  */
+
+extern int GC_dont_expand;
+			/* Dont expand heap unless explicitly requested */
+			/* or forced to.				*/
+
+extern int GC_full_freq;    /* Number of partial collections between	*/
+			    /* full collections.  Matters only if	*/
+			    /* GC_incremental is set.			*/
+			
+extern GC_word GC_non_gc_bytes;
+			/* Bytes not considered candidates for collection. */
+			/* Used only to control scheduling of collections. */
+
+extern GC_word GC_free_space_divisor;
+			/* We try to make sure that we allocate at 	*/
+			/* least N/GC_free_space_divisor bytes between	*/
+			/* collections, where N is the heap size plus	*/
+			/* a rough estimate of the root set size.	*/
+			/* Initially, GC_free_space_divisor = 4.	*/
+			/* Increasing its value will use less space	*/
+			/* but more collection time.  Decreasing it	*/
+			/* will appreciably decrease collection time	*/
+			/* at the expense of space.			*/
+			/* GC_free_space_divisor = 1 will effectively	*/
+			/* disable collections.				*/
+
+			
+/* Public procedures */
+/*
+ * general purpose allocation routines, with roughly malloc calling conv.
+ * The atomic versions promise that no relevant pointers are contained
+ * in the object.  The nonatomic versions guarantee that the new object
+ * is cleared.  GC_malloc_stubborn promises that no changes to the object
+ * will occur after GC_end_stubborn_change has been called on the
+ * result of GC_malloc_stubborn. GC_malloc_uncollectable allocates an object
+ * that is scanned for pointers to collectable objects, but is not itself
+ * collectable.  GC_malloc_uncollectable and GC_free called on the resulting
+ * object implicitly update GC_non_gc_bytes appropriately.
+ */
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_malloc(size_t size_in_bytes);
+  extern void * GC_malloc_atomic(size_t size_in_bytes);
+  extern void * GC_malloc_uncollectable(size_t size_in_bytes);
+  extern void * GC_malloc_stubborn(size_t size_in_bytes);
+# else
+  extern char * GC_malloc(/* size_in_bytes */);
+  extern char * GC_malloc_atomic(/* size_in_bytes */);
+  extern char * GC_malloc_uncollectable(/* size_in_bytes */);
+  extern char * GC_malloc_stubborn(/* size_in_bytes */);
+# endif
+
+/* Explicitly deallocate an object.  Dangerous if used incorrectly.     */
+/* Requires a pointer to the base of an object.				*/
+/* If the argument is stubborn, it should not be changeable when freed. */
+/* An object should not be enable for finalization when it is 		*/
+/* explicitly deallocated.						*/
+/* GC_free(0) is a no-op, as required by ANSI C for free.		*/
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void GC_free(void * object_addr);
+# else
+  extern void GC_free(/* object_addr */);
+# endif
+
+/*
+ * Stubborn objects may be changed only if the collector is explicitly informed.
+ * The collector is implicitly informed of coming change when such
+ * an object is first allocated.  The following routines inform the
+ * collector that an object will no longer be changed, or that it will
+ * once again be changed.  Only nonNIL pointer stores into the object
+ * are considered to be changes.  The argument to GC_end_stubborn_change
+ * must be exacly the value returned by GC_malloc_stubborn or passed to
+ * GC_change_stubborn.  (In the second case it may be an interior pointer
+ * within 512 bytes of the beginning of the objects.)
+ * There is a performance penalty for allowing more than
+ * one stubborn object to be changed at once, but it is acceptable to
+ * do so.  The same applies to dropping stubborn objects that are still
+ * changeable.
+ */
+void GC_change_stubborn(/* p */);
+void GC_end_stubborn_change(/* p */);
+
+/* Return a pointer to the base (lowest address) of an object given	*/
+/* a pointer to a location within the object.				*/
+/* Return 0 if displaced_pointer doesn't point to within a valid	*/
+/* object.								*/
+# if defined(__STDC__) || defined(__cplusplus)
+  void * GC_base(void * displaced_pointer);
+# else
+  char * GC_base(/* char * displaced_pointer */);
+# endif
+
+/* Given a pointer to the base of an object, return its size in bytes.	*/
+/* The returned size may be slightly larger than what was originally	*/
+/* requested.								*/
+# if defined(__STDC__) || defined(__cplusplus)
+  size_t GC_size(void * object_addr);
+# else
+  size_t GC_size(/* char * object_addr */);
+# endif
+
+/* For compatibility with C library.  This is occasionally faster than	*/
+/* a malloc followed by a bcopy.  But if you rely on that, either here	*/
+/* or with the standard C library, your code is broken.  In my		*/
+/* opinion, it shouldn't have been invented, but now we're stuck. -HB	*/
+/* The resulting object has the same kind as the original.		*/
+/* If the argument is stubborn, the result will have changes enabled.	*/
+/* It is an error to have changes enabled for the original object.	*/
+/* Follows ANSI comventions for NULL old_object.			*/
+# if defined(__STDC__) || defined(__cplusplus)
+    extern void * GC_realloc(void * old_object, size_t new_size_in_bytes);
+# else
+    extern char * GC_realloc(/* old_object, new_size_in_bytes */);
+# endif
+
+
+/* Explicitly increase the heap size.	*/
+/* Returns 0 on failure, 1 on success.  */
+extern int GC_expand_hp(/* number_of_bytes */);
+
+/* Clear the set of root segments */
+extern void GC_clear_roots();
+
+/* Add a root segment */
+extern void GC_add_roots(/* low_address, high_address_plus_1 */);
+
+/* Add a displacement to the set of those considered valid by the	*/
+/* collector.  GC_register_displacement(n) means that if p was returned */
+/* by GC_malloc, then (char *)p + n will be considered to be a valid	*/
+/* pointer to n.  N must be small and less than the size of p.		*/
+/* (All pointers to the interior of objects from the stack are		*/
+/* considered valid in any case.  This applies to heap objects and	*/
+/* static data.)							*/
+/* Preferably, this should be called before any other GC procedures.	*/
+/* Calling it later adds to the probability of excess memory		*/
+/* retention.								*/
+/* This is a no-op if the collector was compiled with recognition of	*/
+/* arbitrary interior pointers enabled, which is now the default.	*/
+void GC_register_displacement(/* n */);
+
+/* Explicitly trigger a collection. 	*/
+void GC_gcollect();
+
+/* Return the number of bytes in the heap.  Excludes collector private	*/
+/* data structures.  Includes empty blocks and fragmentation loss.	*/
+/* Includes some pages that were allocated but never written.		*/
+size_t GC_get_heap_size();
+
+/* Enable incremental/generational collection.	*/
+/* Not advisable unless dirty bits are 		*/
+/* available or most heap objects are		*/
+/* pointerfree(atomic) or immutable.		*/
+/* Don't use in leak finding mode.		*/
+/* Ignored if GC_dont_gc is true.		*/
+void GC_enable_incremental();
+
+/* Allocate an object of size lb bytes.  The client guarantees that	*/
+/* as long as the object is live, it will be referenced by a pointer	*/
+/* that points to somewhere within the first 256 bytes of the object.	*/
+/* (This should normally be declared volatile to prevent the compiler	*/
+/* from invalidating this assertion.)  This routine is only useful	*/
+/* if a large array is being allocated.  It reduces the chance of 	*/
+/* accidentally retaining such an array as a result of scanning an	*/
+/* integer that happens to be an address inside the array.  (Actually,	*/
+/* it reduces the chance of the allocator not finding space for such	*/
+/* an array, since it will try hard to avoid introducing such a false	*/
+/* reference.)  On a SunOS 4.X or MS Windows system this is recommended */
+/* for arrays likely to be larger than 100K or so.  For other systems,	*/
+/* or if the collector is not configured to recognize all interior	*/
+/* pointers, the threshold is normally much higher.			*/
+# if defined(__STDC__) || defined(__cplusplus)
+  void * GC_malloc_ignore_off_page(size_t lb);
+# else
+  char * GC_malloc_ignore_off_page(/* size_t lb */);
+# endif
+
+/* Debugging (annotated) allocation.  GC_gcollect will check 		*/
+/* objects allocated in this way for overwrites, etc.			*/
+# if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_debug_malloc(size_t size_in_bytes,
+  				char * descr_string, int descr_int);
+  extern void * GC_debug_malloc_atomic(size_t size_in_bytes,
+  				       char * descr_string, int descr_int);
+  extern void * GC_debug_malloc_uncollectable(size_t size_in_bytes,
+  				           char * descr_string, int descr_int);
+  extern void * GC_debug_malloc_stubborn(size_t size_in_bytes,
+  				         char * descr_string, int descr_int);
+  extern void GC_debug_free(void * object_addr);
+  extern void * GC_debug_realloc(void * old_object,
+  			 	 size_t new_size_in_bytes,
+  			 	 char * descr_string, int descr_int);
+# else
+  extern char * GC_debug_malloc(/* size_in_bytes, descr_string, descr_int */);
+  extern char * GC_debug_malloc_atomic(/* size_in_bytes, descr_string,
+  					  descr_int */);
+  extern char * GC_debug_malloc_uncollectable(/* size_in_bytes, descr_string,
+  					  descr_int */);
+  extern char * GC_debug_malloc_stubborn(/* size_in_bytes, descr_string,
+  					  descr_int */);
+  extern void GC_debug_free(/* object_addr */);
+  extern char * GC_debug_realloc(/* old_object, new_size_in_bytes,
+  			            descr_string, descr_int */);
+# endif
+void GC_debug_change_stubborn(/* p */);
+void GC_debug_end_stubborn_change(/* p */);
+# ifdef GC_DEBUG
+#   define GC_MALLOC(sz) GC_debug_malloc(sz, __FILE__, __LINE__)
+#   define GC_MALLOC_ATOMIC(sz) GC_debug_malloc_atomic(sz, __FILE__, __LINE__)
+#   define GC_MALLOC_UNCOLLECTABLE(sz) GC_debug_malloc_uncollectable(sz, \
+							__FILE__, __LINE__)
+#   define GC_REALLOC(old, sz) GC_debug_realloc(old, sz, __FILE__, \
+							       __LINE__)
+#   define GC_FREE(p) GC_debug_free(p)
+#   define GC_REGISTER_FINALIZER(p, f, d, of, od) \
+	GC_register_finalizer(GC_base(p), GC_debug_invoke_finalizer, \
+			      GC_make_closure(f,d), of, od)
+#   define GC_MALLOC_STUBBORN(sz) GC_debug_malloc_stubborn(sz, __FILE__, \
+							       __LINE__)
+#   define GC_CHANGE_STUBBORN(p) GC_debug_change_stubborn(p)
+#   define GC_END_STUBBORN_CHANGE(p) GC_debug_end_stubborn_change(p)
+# else
+#   define GC_MALLOC(sz) GC_malloc(sz)
+#   define GC_MALLOC_ATOMIC(sz) GC_malloc_atomic(sz)
+#   define GC_MALLOC_UNCOLLECTABLE(sz) GC_malloc_uncollectable(sz)
+#   define GC_REALLOC(old, sz) GC_realloc(old, sz)
+#   define GC_FREE(p) GC_free(p)
+#   define GC_REGISTER_FINALIZER(p, f, d, of, od) \
+	GC_register_finalizer(p, f, d, of, od)
+#   define GC_MALLOC_STUBBORN(sz) GC_malloc_stubborn(sz)
+#   define GC_CHANGE_STUBBORN(p) GC_change_stubborn(p)
+#   define GC_END_STUBBORN_CHANGE(p) GC_end_stubborn_change(p)
+# endif
+/* The following are included because they are often convenient, and	*/
+/* reduce the chance for a misspecifed size argument.  But calls may	*/
+/* expand to something syntactically incorrect if t is a complicated	*/
+/* type expression.  							*/
+# define GC_NEW(t) (t *)GC_MALLOC(sizeof (t))
+# define GC_NEW_ATOMIC(t) (t *)GC_MALLOC_ATOMIC(sizeof (t))
+# define GC_NEW_STUBBORN(t) (t *)GC_MALLOC_STUBBORN(sizeof (t))
+# define GC_NEW_UNCOLLECTABLE(t) (t *)GC_MALLOC_UNCOLLECTABLE(sizeof (t))
+
+/* Finalization.  Some of these primitives are grossly unsafe.		*/
+/* The idea is to make them both cheap, and sufficient to build		*/
+/* a safer layer, closer to PCedar finalization.			*/
+/* The interface represents my conclusions from a long discussion	*/
+/* with Alan Demers, Dan Greene, Carl Hauser, Barry Hayes, 		*/
+/* Christian Jacobi, and Russ Atkinson.  It's not perfect, and		*/
+/* probably nobody else agrees with it.	    Hans-J. Boehm  3/13/92	*/
+# if defined(__STDC__) || defined(__cplusplus)
+  typedef void (*GC_finalization_proc)(void * obj, void * client_data);
+# else
+  typedef void (*GC_finalization_proc)(/* void * obj, void * client_data */);
+# endif
+	
+void GC_register_finalizer(/* void * obj,
+			      GC_finalization_proc fn, void * cd,
+			      GC_finalization_proc *ofn, void ** ocd */);
+	/* When obj is no longer accessible, invoke		*/
+	/* (*fn)(obj, cd).  If a and b are inaccessible, and	*/
+	/* a points to b (after disappearing links have been	*/
+	/* made to disappear), then only a will be		*/
+	/* finalized.  (If this does not create any new		*/
+	/* pointers to b, then b will be finalized after the	*/
+	/* next collection.)  Any finalizable object that	*/
+	/* is reachable from itself by following one or more	*/
+	/* pointers will not be finalized (or collected).	*/
+	/* Thus cycles involving finalizable objects should	*/
+	/* be avoided, or broken by disappearing links.		*/
+	/* Fn should terminate as quickly as possible, and	*/
+	/* defer extended computation.				*/
+	/* All but the last finalizer registered for an object  */
+	/* is ignored.						*/
+	/* Finalization may be removed by passing 0 as fn.	*/
+	/* The old finalizer and client data are stored in	*/
+	/* *ofn and *ocd.					*/ 
+	/* Fn is never invoked on an accessible object,		*/
+	/* provided hidden pointers are converted to real 	*/
+	/* pointers only if the allocation lock is held, and	*/
+	/* such conversions are not performed by finalization	*/
+	/* routines.						*/
+	/* If GC_register_finalizer is aborted as a result of	*/
+	/* a signal, the object may be left with no		*/
+	/* finalization, even if neither the old nor new	*/
+	/* finalizer were NULL.					*/
+	/* Obj should be the nonNULL starting address of an 	*/
+	/* object allocated by GC_malloc or friends.		*/	
+
+/* The following routine may be used to break cycles between	*/
+/* finalizable objects, thus causing cyclic finalizable		*/
+/* objects to be finalized in the correct order.  Standard	*/
+/* use involves calling GC_register_disappearing_link(&p),	*/
+/* where p is a pointer that is not followed by finalization	*/
+/* code, and should not be considered in determining 		*/
+/* finalization order.						*/ 
+int GC_register_disappearing_link(/* void ** link */);
+	/* Link should point to a field of a heap allocated 	*/
+	/* object obj.  *link will be cleared when obj is	*/
+	/* found to be inaccessible.  This happens BEFORE any	*/
+	/* finalization code is invoked, and BEFORE any		*/
+	/* decisions about finalization order are made.		*/
+	/* This is useful in telling the finalizer that 	*/
+	/* some pointers are not essential for proper		*/
+	/* finalization.  This may avoid finalization cycles.	*/
+	/* Note that obj may be resurrected by another		*/
+	/* finalizer, and thus the clearing of *link may	*/
+	/* be visible to non-finalization code.  		*/
+	/* There's an argument that an arbitrary action should  */
+	/* be allowed here, instead of just clearing a pointer. */
+	/* But this causes problems if that action alters, or 	*/
+	/* examines connectivity.				*/
+	/* Returns 1 if link was already registered, 0		*/
+	/* otherwise.						*/
+	/* Only exists for backward compatibility.  See below:	*/
+int GC_general_register_disappearing_link(/* void ** link, void * obj */);
+	/* A slight generalization of the above. *link is	*/
+	/* cleared when obj first becomes inaccessible.  This	*/
+	/* can be used to implement weak pointers easily and	*/
+	/* safely. Typically link will point to a location	*/
+	/* holding a disguised pointer to obj.  In this way	*/
+	/* soft pointers are broken before any object		*/
+	/* reachable from them are finalized.  Each link	*/
+	/* May be registered only once, i.e. with one obj	*/
+	/* value.  This was added after a long email discussion */
+	/* with John Ellis.					*/
+	/* Obj must be a pointer to the first word of an object */
+	/* we allocated.  It is unsafe to explicitly deallocate */
+	/* the object containing link.  Explicitly deallocating */
+	/* obj may or may not cause link to eventually be	*/
+	/* cleared.						*/
+int GC_unregister_disappearing_link(/* void ** link */);
+	/* Returns 0 if link was not actually registered.	*/
+	/* Undoes a registration by either of the above two	*/
+	/* routines.						*/
+
+/* Auxiliary fns to make finalization work correctly with displaced	*/
+/* pointers introduced by the debugging allocators.			*/
+# if defined(__STDC__) || defined(__cplusplus)
+    void * GC_make_closure(GC_finalization_proc fn, void * data);
+    void GC_debug_invoke_finalizer(void * obj, void * data);
+# else
+    char * GC_make_closure(/* GC_finalization_proc fn, char * data */);
+    void GC_debug_invoke_finalizer(/* void * obj, void * data */);
+# endif
+
+	
+/* The following is intended to be used by a higher level	*/
+/* (e.g. cedar-like) finalization facility.  It is expected	*/
+/* that finalization code will arrange for hidden pointers to	*/
+/* disappear.  Otherwise objects can be accessed after they	*/
+/* have been collected.						*/
+# ifdef I_HIDE_POINTERS
+#   if defined(__STDC__) || defined(__cplusplus)
+#     define HIDE_POINTER(p) (~(size_t)(p))
+#     define REVEAL_POINTER(p) ((void *)(HIDE_POINTER(p)))
+#   else
+#     define HIDE_POINTER(p) (~(unsigned long)(p))
+#     define REVEAL_POINTER(p) ((char *)(HIDE_POINTER(p)))
+#   endif
+    /* Converting a hidden pointer to a real pointer requires verifying	*/
+    /* that the object still exists.  This involves acquiring the  	*/
+    /* allocator lock to avoid a race with the collector.		*/
+
+#   if defined(__STDC__) || defined(__cplusplus)
+        typedef void * (*GC_fn_type)();
+        void * GC_call_with_alloc_lock(GC_fn_type fn, void * client_data);
+#   else
+        typedef char * (*GC_fn_type)();
+        char * GC_call_with_alloc_lock(/* GC_fn_type fn, char * client_data */);
+#   endif
+# endif
+
+#ifdef SOLARIS_THREADS
+/* We need to intercept calls to many of the threads primitives, so 	*/
+/* that we can locate thread stacks and stop the world.			*/
+/* Note also that the collector cannot see thread specific data.	*/
+/* Thread specific data should generally consist of pointers to		*/
+/* uncollectable objects, which are deallocated using the destructor	*/
+/* facility in thr_keycreate.						*/
+# include <thread.h>
+  int GC_thr_create(void *stack_base, size_t stack_size,
+                    void *(*start_routine)(void *), void *arg, long flags,
+                    thread_t *new_thread);
+  int GC_thr_join(thread_t wait_for, thread_t *departed, void **status);
+  int GC_thr_suspend(thread_t target_thread);
+  int GC_thr_continue(thread_t target_thread);
+  void * GC_dlopen(const char *path, int mode);
+
+# define thr_create GC_thr_create
+# define thr_join GC_thr_join
+# define thr_suspend GC_thr_suspend
+# define thr_continue GC_thr_continue
+# define dlopen GC_dlopen
+
+/* This returns a list of objects, linked through their first		*/
+/* word.  Its use can greatly reduce lock contention problems, since	*/
+/* the allocation lock can be acquired and released many fewer times.	*/
+void * GC_malloc_many(size_t lb);
+#define GC_NEXT(p) (*(void **)(p)) 	/* Retrieve the next element	*/
+					/* in returned list.		*/
+
+#endif /* SOLARIS_THREADS */
+
+#endif /* _GC_H */
diff --git a/gc.man b/gc.man
new file mode 100644
index 00000000..73f8318b
--- /dev/null
+++ b/gc.man
@@ -0,0 +1,63 @@
+.TH GC_MALLOC 1L "20 April 1994"
+.SH NAME
+GC_malloc, GC_malloc_atomic, GC_free, GC_realloc, GC_enable_incremental, GC_register_finalizer \- Garbage collecting malloc replacement
+.SH SYNOPSIS
+#include "gc.h"
+.br
+# define malloc(n) GC_malloc(n)
+.br
+... malloc(...) ...
+.br
+.sp
+cc ... gc.a
+.LP
+.SH DESCRIPTION
+.I GC_malloc
+and
+.I GC_free
+are plug-in replacements for standard malloc and free.  However,
+.I
+GC_malloc
+will attempt to reclaim inaccessible space automaticaly by invoking a conservative garbage collector at appropriate points.  The collector traverses all data structures accessible by following pointers from the machines registers, stack(s), data, and bss segments.  Inaccessible structures will be reclaimed.  A machine word is considered to be a valid pointer if it is an address inside an object allocated by
+.I
+GC_malloc
+or friends.
+.LP
+Unlike the standard implementations of malloc,
+.I
+GC_malloc
+clears the newly allocated storage.
+.I
+GC_malloc_atomic
+does not.  Furthermore, it informs the collector that the resulting object will never contain any pointers, and should therefore not be scanned by the collector.
+.I
+GC_free
+can be used to deallocate objects, but its use is optional, and discouraged.
+.I
+GC_realloc
+has the standard realloc semantics.  It preserves pointer-free-ness.
+.I
+GC_register_finalizer
+allows for registration of functions that are invoked when an object becomes inaccessible.
+.LP
+It is also possible to use the collector to find storage leaks in programs destined to be run with standard malloc/free.  The collector can be compiled for thread-safe operation.  Unlike standard malloc, it is safe to call malloc after a previous malloc call was interrupted by a signal, provided the original malloc call is not resumed.
+.LP
+Debugging versions of many of the above routines are provided as macros.  Their names are identical to the above, but consist of all capital letters.  If GC_DEBUG is defined before gc.h is included, these routines do additional checking, and allow the leak detecting version of the collector to produce slightly more useful output.  Without GC_DEBUG defined, they behave exactly like the lower-case versions.
+.LP
+On some machines, collection will be performed incrementally after a call to
+.I
+GC_enable_incremental.
+This may temporarily write protect pages in the heap.  See the README file for more information on how this interacts with system calls that write to the heap.
+.LP
+Other facilities not discussed here include a C++ interface, limited facilities to support incremental collection on machines without appropriate VM support, provisions for providing more explicit object layout information to the garbage collector, more direct support for ``weak'' pointers, etc.
+.LP
+.SH "SEE ALSO"
+The README and gc.h files in the distribution.  More detailed definitions of the functions exported by the collector are given there.  (The above list is not complete.)
+.LP
+Boehm, H., and M. Weiser, "Garbage Collection in an Uncooperative Environment",
+\fISoftware Practice & Experience\fP, September 1988, pp. 807-820.
+.LP
+The malloc(3) man page.
+.LP
+.SH AUTHOR
+Hans-J. Boehm (boehm@parc.xerox.com).  Some of the code was written by others, most notably Alan Demers.
diff --git a/gc_c++.cc b/gc_c++.cc
new file mode 100644
index 00000000..6654241a
--- /dev/null
+++ b/gc_c++.cc
@@ -0,0 +1,33 @@
+/*************************************************************************
+
+
+Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
+ 
+THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ 
+Permission is hereby granted to copy this code for any purpose,
+provided the above notices are retained on all copies.
+
+This implementation module for gc_c++.h provides an implementation of
+the global operators "new" and "delete" that calls the Boehm
+allocator.  All objects allocated by this implementation will be
+non-collectable but part of the root set of the collector.
+
+You should ensure (using implementation-dependent techniques) that the
+linker finds this module before the library that defines the default
+built-in "new" and "delete".
+
+
+**************************************************************************/
+
+#include "gc_c++.h"
+
+void* operator new( size_t size ) {
+    return GC_MALLOC_UNCOLLECTABLE( size ); }
+  
+void operator delete( void* obj ) {
+    return GC_FREE( obj ); }
+  
+
+
diff --git a/gc_c++.h b/gc_c++.h
new file mode 100644
index 00000000..26019076
--- /dev/null
+++ b/gc_c++.h
@@ -0,0 +1,161 @@
+
+/****************************************************************************
+
+Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
+ 
+THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ 
+Permission is hereby granted to use or copy this program
+for any purpose,  provided the above notices are retained on all copies.
+Permission to modify the code and to distribute modified code is granted,
+provided the above notices are retained, and a notice that the code was
+modified is included with the above copyright notice.
+
+C++ Interface to the Boehm Collector
+
+    Jesse Hull and John Ellis
+    Last modified on Tue Feb 15 14:43:02 PST 1994 by ellis
+
+This interface provides access to the Boehm collector (versions 3.6
+and later).  It is intended to provide facilities similar to those
+described in the Ellis-Detlefs proposal for C++ garbage collection.
+
+To make a class collectable, derive it from the base class "gc":
+
+    class MyClass: gc {...}
+
+Then, "new MyClass" will allocate intances that will be automatically
+garbage collected.
+
+Collected objects can be explicitly deleted with "delete", e.g.
+
+    MyClass* m = ...;
+    delete m;
+
+This will free the object's storage immediately.
+
+Collected instances of non-class types can be allocated using
+placement syntax with the argument "GC":
+
+    typedef int A[ 10 ];
+    A* a = new (GC) A;
+
+The built-in "operator new" continues to allocate non-collectible
+objects that the programmer must explicitly delete.  Collected object
+may freely point at non-collected objects, and vice versa.
+
+Object clean-up (finalization) can be specified using class
+"gc_cleanup".  When an object derived from "gc_cleanup" is discovered
+to be inaccessible by the collector, or when it is explicitly deleted,
+its destructors will be invoked first.
+
+Clean-up functions for non-class types can be specified as additional
+placement arguments:
+
+    A* a = new (GC, MyCleanup) A;
+
+An object is considered "accessible" by the collector if it can be
+reached by a path of pointers from static variables, automatic
+variables of active functions, or from another object with clean-up
+enabled.  This implies that if object A and B both have clean-up
+enabled, and A points at B, B will be considered accessible, and A's
+clean-up will be be invoked before B's.  If A points at B and B points
+back to A, forming a cycle, that's considered a storage leak, and
+neither will ever become inaccessible.  See the C interface gc.h for
+low-level facilities for handling such cycles of objects with cleanup.
+
+****************************************************************************/
+
+#ifndef GC_CPP_H
+#define GC_CPP_H
+
+extern "C" {
+#include "gc.h"
+}
+
+enum GCPlacement {GC, NoGC};
+
+class gc {
+public:
+    void* operator new( size_t size );
+    void* operator new( size_t size, GCPlacement gcp );
+    void operator delete( void* obj ); };
+    /*
+    Intances of classes derived from "gc" will be allocated in the 
+    collected heap by default, unless an explicit NoGC placement is
+    specified. */
+
+class gc_cleanup: public gc {
+public:
+    gc_cleanup();
+    virtual ~gc_cleanup();
+private:
+    static void cleanup( void* obj, void* clientData ); };
+    /*
+    Instances of classes derived from "gc_cleanup" will be allocated
+    in the collected heap by default.  Further, when the collector
+    discovers an instance is inaccessible (see above) or when the
+    instance is explicitly deleted, its destructors will be invoked.
+    NOTE: Only one instance of "gc_cleanup" should occur in the
+    inheritance heirarchy -- i.e. it should always be a virtual
+    base. */
+
+void* operator new( 
+    size_t size, 
+    GCPlacement gcp,
+    void (*cleanup)( void*, void* ) = 0,
+    void* clientData = 0 );
+    /*
+    If "gcp = GC", then this "operator new" allocates in the collected
+    heap, otherwise in the non-collected heap.  When the allocated
+    object "obj" becomes inaccessible, the collector will invoke the
+    function "cleanup( obj, clientData )".  It is an error to specify
+    a non-null "cleanup" when "gcp = NoGC". */
+
+/****************************************************************************
+
+Inline implementation
+
+****************************************************************************/
+
+inline void* gc::operator new( size_t size ) {
+    return GC_MALLOC( size ); };
+
+inline void* gc::operator new( size_t size, GCPlacement gcp ) {
+    if (gcp == GC) 
+        return GC_MALLOC( size );
+    else
+        return GC_MALLOC_UNCOLLECTABLE( size ); }
+
+inline void gc::operator delete( void* obj ) {
+    GC_FREE( obj ); }; 
+
+inline gc_cleanup::gc_cleanup() {
+    GC_REGISTER_FINALIZER( GC_base( this ), cleanup, this, 0, 0 ); }
+
+inline void gc_cleanup::cleanup( void* obj, void* realThis ) {
+    ((gc_cleanup*) realThis)->~gc_cleanup(); }
+
+inline gc_cleanup::~gc_cleanup() {
+    GC_REGISTER_FINALIZER( this, 0, 0, 0, 0 ); }
+
+inline void* operator new( 
+    size_t size, 
+    GCPlacement gcp,
+    void (*cleanup)( void*, void* ) = 0,
+    void* clientData = 0 )
+{
+    void* obj;
+
+    if (gcp == GC) {
+        obj = GC_MALLOC( size );
+        if (cleanup != 0) 
+            GC_REGISTER_FINALIZER( obj, cleanup, clientData, 0, 0 ); }
+    else {
+        obj = GC_MALLOC_UNCOLLECTABLE( size ); };
+    return obj; }
+        
+
+#endif
+
diff --git a/gc_hdrs.h b/gc_hdrs.h
new file mode 100644
index 00000000..c4fd5577
--- /dev/null
+++ b/gc_hdrs.h
@@ -0,0 +1,133 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:16 pm PDT */
+# ifndef GC_HEADERS_H
+# define GC_HEADERS_H
+typedef struct hblkhdr hdr;
+
+# if CPP_WORDSZ != 32 && CPP_WORDSZ < 36
+	--> Get a real machine.
+# endif
+
+/*
+ * The 2 level tree data structure that is used to find block headers.
+ * If there are more than 32 bits in a pointer, the top level is a hash
+ * table.
+ */
+
+# if CPP_WORDSZ > 32
+#   define HASH_TL
+# endif
+
+/* Define appropriate out-degrees for each of the two tree levels	*/
+# ifdef SMALL_CONFIG
+#   define LOG_BOTTOM_SZ 11
+	/* Keep top index size reasonable with smaller blocks. */
+# else
+#   define LOG_BOTTOM_SZ 10
+# endif
+# ifndef HASH_TL
+#   define LOG_TOP_SZ (WORDSZ - LOG_BOTTOM_SZ - LOG_HBLKSIZE)
+# else
+#   define LOG_TOP_SZ 11
+# endif
+# define TOP_SZ (1 << LOG_TOP_SZ)
+# define BOTTOM_SZ (1 << LOG_BOTTOM_SZ)
+
+typedef struct bi {
+    hdr * index[BOTTOM_SZ];
+	/*
+ 	 * The bottom level index contains one of three kinds of values:
+	 * 0 means we're not responsible for this block.
+	 * 1 < (long)X <= MAX_JUMP means the block starts at least
+	 *        X * HBLKSIZE bytes before the current address.
+	 * A valid pointer points to a hdr structure. (The above can't be
+	 * valid pointers due to the GET_MEM return convention.)
+	 */
+    struct bi * asc_link;	/* All indices are linked in	*/
+    				/* ascending order.		*/
+    word key;			/* high order address bits.	*/
+# ifdef HASH_TL
+    struct bi * hash_link;	/* Hash chain link.		*/
+# endif
+} bottom_index;
+
+/* extern bottom_index GC_all_nils; - really part of GC_arrays */
+
+/* extern bottom_index * GC_top_index []; - really part of GC_arrays */
+				/* Each entry points to a bottom_index.	*/
+				/* On a 32 bit machine, it points to 	*/
+				/* the index for a set of high order	*/
+				/* bits equal to the index.  For longer	*/
+				/* addresses, we hash the high order	*/
+				/* bits to compute the index in 	*/
+				/* GC_top_index, and each entry points	*/
+				/* to a hash chain.			*/
+				/* The last entry in each chain is	*/
+				/* GC_all_nils.				*/
+
+
+# define MAX_JUMP (HBLKSIZE - 1)
+
+# ifndef HASH_TL
+#   define BI(p) (GC_top_index \
+		[(word)(p) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE)])
+#   define HDR_INNER(p) (BI(p)->index \
+		[((word)(p) >> LOG_HBLKSIZE) & (BOTTOM_SZ - 1)])
+#   ifdef SMALL_CONFIG
+#	define HDR(p) GC_find_header((ptr_t)(p))
+#   else
+#	define HDR(p) HDR_INNER(p)
+#   endif
+#   define GET_BI(p, bottom_indx) (bottom_indx) = BI(p)
+#   define GET_HDR(p, hhdr) (hhdr) = HDR(p)
+#   define SET_HDR(p, hhdr) HDR_INNER(p) = (hhdr)
+#   define GET_HDR_ADDR(p, ha) (ha) = &(HDR_INNER(p))
+# else /* hash */
+/*  Hash function for tree top level */
+#   define TL_HASH(hi) ((hi) & (TOP_SZ - 1))
+/*  Set bottom_indx to point to the bottom index for address p */
+#   define GET_BI(p, bottom_indx) \
+	{ \
+	    register word hi = \
+	        (word)(p) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE); \
+	    register bottom_index * _bi = GC_top_index[TL_HASH(hi)]; \
+	    \
+	    while (_bi -> key != hi && _bi != &GC_all_nils) \
+	    	_bi = _bi -> hash_link; \
+	    (bottom_indx) = _bi; \
+	}
+#   define GET_HDR_ADDR(p, ha) \
+	{ \
+	    register bottom_index * bi; \
+	    \
+	    GET_BI(p, bi);	\
+	    (ha) = &(bi->index[((unsigned long)(p)>>LOG_HBLKSIZE) \
+	        	  	& (BOTTOM_SZ - 1)]); \
+	}
+#   define GET_HDR(p, hhdr) { register hdr ** _ha; GET_HDR_ADDR(p, _ha); \
+			      (hhdr) = *_ha; }
+#   define SET_HDR(p, hhdr) { register hdr ** _ha; GET_HDR_ADDR(p, _ha); \
+			      *_ha = (hhdr); }
+#   define HDR(p) GC_find_header((ptr_t)(p))
+# endif
+			    
+/* Is the result a forwarding address to someplace closer to the	*/
+/* beginning of the block or NIL?					*/
+# define IS_FORWARDING_ADDR_OR_NIL(hhdr) ((unsigned long) (hhdr) <= MAX_JUMP)
+
+/* Get an HBLKSIZE aligned address closer to the beginning of the block */
+/* h.  Assumes hhdr == HDR(h) and IS_FORWARDING_ADDR(hhdr).		*/
+# define FORWARDED_ADDR(h, hhdr) ((struct hblk *)(h) - (unsigned long)(hhdr))
+# endif /*  GC_HEADERS_H */
diff --git a/gc_inl.h b/gc_inl.h
new file mode 100644
index 00000000..1f9a9a0d
--- /dev/null
+++ b/gc_inl.h
@@ -0,0 +1,95 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:12 pm PDT */
+ 
+# ifndef GC_PRIVATE_H
+#   include "gc_priv.h"
+# endif
+
+/* Allocate n words (NOT BYTES).  X is made to point to the result.	*/
+/* It is assumed that n < MAXOBJSZ, and					*/
+/* that n > 0.  On machines requiring double word alignment of some	*/
+/* data, we also assume that n is 1 or even.  This bypasses the		*/
+/* MERGE_SIZES mechanism.  In order to minimize the number of distinct	*/
+/* free lists that are maintained, the caller should ensure that a 	*/
+/* small number of distinct values of n are used.  (The MERGE_SIZES	*/
+/* mechanism normally does this by ensuring that only the leading three	*/
+/* bits of n may be nonzero.  See misc.c for details.)  We really 	*/
+/* recommend this only in cases in which n is a constant, and no	*/
+/* locking is required.							*/
+/* In that case it may allow the compiler to perform substantial	*/
+/* additional optimizations.						*/
+# define GC_MALLOC_WORDS(result,n) \
+{	\
+    register ptr_t op;	\
+    register ptr_t *opp;	\
+    DCL_LOCK_STATE;	\
+	\
+    opp = &(GC_objfreelist[n]);	\
+    FASTLOCK();	\
+    if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {	\
+        FASTUNLOCK();	\
+        (result) = GC_generic_malloc_words_small((n), NORMAL);	\
+    } else { 	\
+        *opp = obj_link(op);	\
+        obj_link(op) = 0;	\
+        GC_words_allocd += (n);	\
+        FASTUNLOCK();	\
+        (result) = (extern_ptr_t) op;	\
+    }	\
+}
+
+
+/* The same for atomic objects:	*/
+# define GC_MALLOC_ATOMIC_WORDS(result,n) \
+{	\
+    register ptr_t op;	\
+    register ptr_t *opp;	\
+    DCL_LOCK_STATE;	\
+	\
+    opp = &(GC_aobjfreelist[n]);	\
+    FASTLOCK();	\
+    if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {	\
+        FASTUNLOCK();	\
+        (result) = GC_generic_malloc_words_small((n), PTRFREE);	\
+    } else { 	\
+        *opp = obj_link(op);	\
+        obj_link(op) = 0;	\
+        GC_words_allocd += (n);	\
+        FASTUNLOCK();	\
+        (result) = (extern_ptr_t) op;	\
+    }	\
+}
+
+/* And once more for two word initialized objects: */
+# define GC_CONS(result, first, second) \
+{	\
+    register ptr_t op;	\
+    register ptr_t *opp;	\
+    DCL_LOCK_STATE;	\
+	\
+    opp = &(GC_objfreelist[2]);	\
+    FASTLOCK();	\
+    if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {	\
+        FASTUNLOCK();	\
+        op = GC_generic_malloc_words_small(2, NORMAL);	\
+    } else {	\
+        *opp = obj_link(op);	\
+        GC_words_allocd += 2;	\
+        FASTUNLOCK();	\
+    } \
+    ((word *)op)[0] = (word)(first);	\
+    ((word *)op)[1] = (word)(second);	\
+    (result) = (extern_ptr_t) op;	\
+}
diff --git a/gc_inline.h b/gc_inline.h
new file mode 100644
index 00000000..db62d1d5
--- /dev/null
+++ b/gc_inline.h
@@ -0,0 +1 @@
+# include "gc_inl.h"
diff --git a/gc_mark.h b/gc_mark.h
new file mode 100644
index 00000000..b1a7c37e
--- /dev/null
+++ b/gc_mark.h
@@ -0,0 +1,209 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ */
+/* Boehm, May 19, 1994 2:15 pm PDT */
+
+/*
+ * Declarations of mark stack.  Needed by marker and client supplied mark
+ * routines.  To be included after gc_priv.h.
+ */
+#ifndef GC_MARK_H
+# define GC_MARK_H
+
+/* A client supplied mark procedure.  Returns new mark stack pointer.	*/
+/* Not currently used for predefined object kinds.			*/
+/* Primary effect should be to push new entries on the mark stack.	*/
+/* Mark stack pointer values are passed and returned explicitly.	*/
+/* Global variables decribing mark stack are not necessarily valid.	*/
+/* (This usually saves a few cycles by keeping things in registers.)	*/
+/* Assumed to scan about PROC_BYTES on average.  If it needs to do	*/
+/* much more work than that, it should do it in smaller pieces by	*/
+/* pushing itself back on the mark stack.				*/
+/* Note that it should always do some work (defined as marking some	*/
+/* objects) before pushing more than one entry on the mark stack.	*/
+/* This is required to ensure termination in the event of mark stack	*/
+/* overflows.								*/
+/* This procedure is always called with at least one empty entry on the */
+/* mark stack.								*/
+/* Boehm, March 15, 1994 2:38 pm PST */
+# define PROC_BYTES 100
+typedef struct ms_entry * (*mark_proc)(/* word * addr, mark_stack_ptr,
+					  mark_stack_limit, env */);
+					  
+# define LOG_MAX_MARK_PROCS 6
+# define MAX_MARK_PROCS (1 << LOG_MAX_MARK_PROCS)
+extern mark_proc GC_mark_procs[MAX_MARK_PROCS];
+extern word GC_n_mark_procs;
+
+/* Object descriptors on mark stack or in objects.  Low order two	*/
+/* bits are tags distinguishing among the following 4 possibilities	*/
+/* for the high order 30 bits.						*/
+#define DS_TAG_BITS 2
+#define DS_TAGS   ((1 << DS_TAG_BITS) - 1)
+#define DS_LENGTH 0	/* The entire word is a length in bytes that	*/
+			/* must be a multiple of 4.			*/
+#define DS_BITMAP 1	/* 30 bits are a bitmap describing pointer	*/
+			/* fields.  The msb is 1 iff the first word	*/
+			/* is a pointer.				*/
+			/* (This unconventional ordering sometimes	*/
+			/* makes the marker slightly faster.)		*/
+			/* Zeroes indicate definite nonpointers.  Ones	*/
+			/* indicate possible pointers.			*/
+			/* Only usable if pointers are word aligned.	*/
+#   define BITMAP_BITS (WORDSZ - DS_TAG_BITS)
+#define DS_PROC   2
+			/* The objects referenced by this object can be */
+			/* pushed on the mark stack by invoking		*/
+			/* PROC(descr).  ENV(descr) is passed as the	*/
+			/* last argument.				*/
+#   define PROC(descr) \
+		(GC_mark_procs[((descr) >> DS_TAG_BITS) & (MAX_MARK_PROCS-1)])
+#   define ENV(descr) \
+		((descr) >> (DS_TAG_BITS + LOG_MAX_MARK_PROCS))
+#   define MAX_ENV \
+  	      (((word)1 << (WORDSZ - DS_TAG_BITS - LOG_MAX_MARK_PROCS)) - 1)
+#   define MAKE_PROC(proc_index, env) \
+	    (((((env) << LOG_MAX_MARK_PROCS) | (proc_index)) << DS_TAG_BITS) \
+	    | DS_PROC)
+#define DS_PER_OBJECT 3	/* The real descriptor is at the		*/
+			/* byte displacement from the beginning of the	*/
+			/* object given by descr & ~DS_TAGS		*/
+			
+typedef struct ms_entry {
+    word * mse_start;   /* First word of object */
+    word mse_descr;	/* Descriptor; low order two bits are tags,	*/
+    			/* identifying the upper 30 bits as one of the	*/
+    			/* following:					*/
+} mse;
+
+extern word GC_mark_stack_size;
+
+extern mse * GC_mark_stack_top;
+
+extern mse * GC_mark_stack;
+
+word GC_find_start();
+
+mse * GC_signal_mark_stack_overflow();
+
+# ifdef GATHERSTATS
+#   define ADD_TO_ATOMIC(sz) GC_atomic_in_use += (sz)
+#   define ADD_TO_COMPOSITE(sz) GC_composite_in_use += (sz)
+# else
+#   define ADD_TO_ATOMIC(sz)
+#   define ADD_TO_COMPOSITE(sz)
+# endif
+
+/* Push the object obj with corresponding heap block header hhdr onto 	*/
+/* the mark stack.							*/
+# define PUSH_OBJ(obj, hhdr, mark_stack_top, mark_stack_limit) \
+{ \
+    register word _descr = (hhdr) -> hb_descr; \
+        \
+    if (_descr == 0) { \
+    	ADD_TO_ATOMIC((hhdr) -> hb_sz); \
+    } else { \
+        ADD_TO_COMPOSITE((hhdr) -> hb_sz); \
+        mark_stack_top++; \
+        if (mark_stack_top >= mark_stack_limit) { \
+          mark_stack_top = GC_signal_mark_stack_overflow(mark_stack_top); \
+        } \
+        mark_stack_top -> mse_start = (obj); \
+        mark_stack_top -> mse_descr = _descr; \
+    } \
+}
+
+/* Push the contenst of current onto the mark stack if it is a valid	*/
+/* ptr to a currently unmarked object.  Mark it.			*/
+# define PUSH_CONTENTS(current, mark_stack_top, mark_stack_limit) \
+{ \
+    register int displ;  /* Displacement in block; first bytes, then words */ \
+    register hdr * hhdr; \
+    register map_entry_type map_entry; \
+    \
+    GET_HDR(current,hhdr); \
+    if (IS_FORWARDING_ADDR_OR_NIL(hhdr)) { \
+         current = GC_find_start(current, hhdr); \
+         if (current == 0) continue; \
+         hhdr = HDR(current); \
+    } \
+    displ = HBLKDISPL(current); \
+    map_entry = MAP_ENTRY((hhdr -> hb_map), displ); \
+    if (map_entry == OBJ_INVALID) { \
+        GC_ADD_TO_BLACK_LIST_NORMAL(current); continue; \
+    } \
+    displ = BYTES_TO_WORDS(displ); \
+    displ -= map_entry; \
+	\
+    { \
+        register word * mark_word_addr = hhdr -> hb_marks + divWORDSZ(displ); \
+        register word mark_word = *mark_word_addr; \
+        register word mark_bit = (word)1 << modWORDSZ(displ); \
+          \
+        if (mark_word & mark_bit) { \
+	      /* Mark bit is already set */ \
+	      continue; \
+        } \
+        *mark_word_addr = mark_word | mark_bit; \
+    } \
+    PUSH_OBJ(((word *)(HBLKPTR(current)) + displ), hhdr, \
+    	     mark_stack_top, mark_stack_limit) \
+}
+
+extern bool GC_mark_stack_too_small;
+				/* We need a larger mark stack.  May be	*/
+				/* set by client supplied mark routines.*/
+
+typedef int mark_state_t;	/* Current state of marking, as follows:*/
+				/* Used to remember where we are during */
+				/* concurrent marking.			*/
+
+				/* We say something is dirty if it was	*/
+				/* written since the last time we	*/
+				/* retrieved dirty bits.  We say it's 	*/
+				/* grungy if it was marked dirty in the	*/
+				/* last set of bits we retrieved.	*/
+				
+				/* Invariant I: all roots and marked	*/
+				/* objects p are either dirty, or point */
+				/* objects q that are either marked or	*/
+				/* a pointer to q appears in a range	*/
+				/* on the mark stack.			*/
+
+# define MS_NONE 0		/* No marking in progress. I holds.	*/
+				/* Mark stack is empty.			*/
+
+# define MS_PUSH_RESCUERS 1	/* Rescuing objects are currently 	*/
+				/* being pushed.  I holds, except	*/
+				/* that grungy roots may point to 	*/
+				/* unmarked objects, as may marked	*/
+				/* grungy objects above scan_ptr.	*/
+
+# define MS_PUSH_UNCOLLECTABLE 2
+				/* I holds, except that marked 		*/
+				/* uncollectable objects above scan_ptr */
+				/* may point to unmarked objects.	*/
+				/* Roots may point to unmarked objects	*/
+
+# define MS_ROOTS_PUSHED 3	/* I holds, mark stack may be nonempty  */
+
+# define MS_PARTIALLY_INVALID 4	/* I may not hold, e.g. because of M.S. */
+				/* overflow.  However marked heap	*/
+				/* objects below scan_ptr point to	*/
+				/* marked or stacked objects.		*/
+
+# define MS_INVALID 5		/* I may not hold.			*/
+
+extern mark_state_t GC_mark_state;
+
+#endif  /* GC_MARK_H */
diff --git a/gc_priv.h b/gc_priv.h
new file mode 100644
index 00000000..501e6f3b
--- /dev/null
+++ b/gc_priv.h
@@ -0,0 +1,1170 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:17 pm PDT */
+ 
+
+# ifndef GC_PRIVATE_H
+# define GC_PRIVATE_H
+
+# ifndef GC_H
+#   include "gc.h"
+# endif
+
+typedef GC_word word;
+typedef GC_signed_word signed_word;
+
+# ifndef CONFIG_H
+#   include "config.h"
+# endif
+
+# ifndef HEADERS_H
+#   include "gc_hdrs.h"
+# endif
+
+# ifndef bool
+    typedef int bool;
+# endif
+# define TRUE 1
+# define FALSE 0
+
+typedef char * ptr_t;	/* A generic pointer to which we can add	*/
+			/* byte displacments.				*/
+			/* Prefereably identical to caddr_t, if it 	*/
+			/* exists.					*/
+			
+#if defined(__STDC__)
+#   include <stdlib.h>
+#   if !(defined( sony_news ) )
+#       include <stddef.h>
+#   endif
+    typedef void * extern_ptr_t;
+#   define VOLATILE volatile
+#else
+#   ifdef MSWIN32
+#   	include <stdlib.h>
+#   endif
+    typedef char * extern_ptr_t;
+#   define VOLATILE
+#endif
+
+#ifdef AMIGA
+#   define GC_FAR __far
+#else
+#   define GC_FAR
+#endif
+
+/*********************************/
+/*                               */
+/* Definitions for conservative  */
+/* collector                     */
+/*                               */
+/*********************************/
+
+/*********************************/
+/*                               */
+/* Easily changeable parameters  */
+/*                               */
+/*********************************/
+
+#define STUBBORN_ALLOC	/* Define stubborn allocation primitives	*/
+#if defined(SRC_M3) || defined(SMALL_CONFIG)
+# undef STUBBORN_ALLOC
+#endif
+
+
+/* #define ALL_INTERIOR_POINTERS */
+		    /* Forces all pointers into the interior of an 	*/
+		    /* object to be considered valid.  Also causes the	*/
+		    /* sizes of all objects to be inflated by at least 	*/
+		    /* one byte.  This should suffice to guarantee	*/
+		    /* that in the presence of a compiler that does	*/
+		    /* not perform garbage-collector-unsafe		*/
+		    /* optimizations, all portable, strictly ANSI	*/
+		    /* conforming C programs should be safely usable	*/
+		    /* with malloc replaced by GC_malloc and free	*/
+		    /* calls removed.  There are several disadvantages: */
+		    /* 1. There are probably no interesting, portable,	*/
+		    /*    strictly ANSI	conforming C programs.		*/
+		    /* 2. This option makes it hard for the collector	*/
+		    /*    to allocate space that is not ``pointed to''  */
+		    /*    by integers, etc.  Under SunOS 4.X with a 	*/
+		    /*    statically linked libc, we empiricaly		*/
+		    /*    observed that it would be difficult to 	*/
+		    /*	  allocate individual objects larger than 100K.	*/
+		    /* 	  Even if only smaller objects are allocated,	*/
+		    /*    more swap space is likely to be needed.       */
+		    /*    Fortunately, much of this will never be	*/
+		    /*    touched.					*/
+		    /* If you can easily avoid using this option, do.	*/
+		    /* If not, try to keep individual objects small.	*/
+		    
+#define PRINTSTATS  /* Print garbage collection statistics          	*/
+		    /* For less verbose output, undefine in reclaim.c 	*/
+
+#define PRINTTIMES  /* Print the amount of time consumed by each garbage   */
+		    /* collection.                                         */
+
+#define PRINTBLOCKS /* Print object sizes associated with heap blocks,     */
+		    /* whether the objects are atomic or composite, and    */
+		    /* whether or not the block was found to be empty      */
+		    /* duing the reclaim phase.  Typically generates       */
+		    /* about one screenful per garbage collection.         */
+#undef PRINTBLOCKS
+
+#define PRINTBLACKLIST 	/* Print black listed blocks, i.e. values that     */
+			/* cause the allocator to avoid allocating certain */
+			/* blocks in order to avoid introducing "false	   */
+			/* hits".					   */
+#undef PRINTBLACKLIST
+
+#ifdef SILENT
+#  ifdef PRINTSTATS
+#    undef PRINTSTATS
+#  endif
+#  ifdef PRINTTIMES
+#    undef PRINTTIMES
+#  endif
+#  ifdef PRINTNBLOCKS
+#    undef PRINTNBLOCKS
+#  endif
+#endif
+
+#if defined(PRINTSTATS) && !defined(GATHERSTATS)
+#   define GATHERSTATS
+#endif
+
+# if defined(SOLARIS_THREADS) && !defined(SUNOS5)
+--> inconsistent configuration
+# endif
+# if defined(PCR) || defined(SRC_M3) || defined(SOLARIS_THREADS)
+#   define THREADS
+# endif
+
+#ifdef SPARC
+#   define ALIGN_DOUBLE  /* Align objects of size > 1 word on 2 word   */
+			 /* boundaries.  Wasteful of memory, but       */
+			 /* apparently required by SPARC architecture. */
+#   define ASM_CLEAR_CODE	/* Stack clearing is crucial, and we 	*/
+				/* include assembly code to do it well.	*/
+#endif
+
+#define MERGE_SIZES /* Round up some object sizes, so that fewer distinct */
+		    /* free lists are actually maintained.  This applies  */
+		    /* only to the top level routines in misc.c, not to   */
+		    /* user generated code that calls GC_allocobj and     */
+		    /* GC_allocaobj directly.                             */
+		    /* Slows down average programs slightly.  May however */
+		    /* substantially reduce fragmentation if allocation   */
+		    /* request sizes are widely scattered.                */
+		    /* May save significant amounts of space for obj_map  */
+		    /* entries.						  */
+
+/* ALIGN_DOUBLE requires MERGE_SIZES at present. */
+# if defined(ALIGN_DOUBLE) && !defined(MERGE_SIZES)
+#   define MERGE_SIZES
+# endif
+
+#if defined(ALL_INTERIOR_POINTERS) && !defined(DONT_ADD_BYTE_AT_END)
+# define ADD_BYTE_AT_END
+#endif
+
+
+# define MINHINCR 16       /* Minimum heap increment, in blocks of HBLKSIZE  */
+# define MAXHINCR 512      /* Maximum heap increment, in blocks              */
+
+# define TIME_LIMIT 50	   /* We try to keep pause times from exceeding	 */
+			   /* this by much. In milliseconds.		 */
+
+/*********************************/
+/*                               */
+/* OS interface routines	 */
+/*                               */
+/*********************************/
+
+#include <time.h>
+#if !defined(__STDC__) && defined(SPARC) && defined(SUNOS4)
+   clock_t clock();	/* Not in time.h, where it belongs	*/
+#endif
+#if !defined(CLOCKS_PER_SEC)
+#   define CLOCKS_PER_SEC 1000000
+/*
+ * This is technically a bug in the implementation.  ANSI requires that
+ * CLOCKS_PER_SEC be defined.  But at least under SunOS4.1.1, it isn't.
+ * Also note that the combination of ANSI C and POSIX is incredibly gross
+ * here. The type clock_t is used by both clock() and times().  But on
+ * some machines thes use different notions of a clock tick,  CLOCKS_PER_SEC
+ * seems to apply only to clock.  Hence we use it here.  On many machines,
+ * including SunOS, clock actually uses units of microseconds (which are
+ * not really clock ticks).
+ */
+#endif
+#define CLOCK_TYPE clock_t
+#define GET_TIME(x) x = clock()
+#define MS_TIME_DIFF(a,b) ((unsigned long) \
+		(1000.0*(double)((a)-(b))/(double)CLOCKS_PER_SEC))
+
+/* We use bzero and bcopy internally.  They may not be available.	*/
+# if defined(SPARC) && defined(SUNOS4)
+#   define BCOPY_EXISTS
+# endif
+# if defined(M68K) && defined(AMIGA)
+#   define BCOPY_EXISTS
+# endif
+# if defined(M68K) && defined(NEXT)
+#   define BCOPY_EXISTS
+# endif
+# if defined(VAX)
+#   define BCOPY_EXISTS
+# endif
+# if defined(AMIGA)
+#   include <string.h>
+#   define BCOPY_EXISTS
+# endif
+
+# ifndef BCOPY_EXISTS
+#   include <string.h>
+#   define BCOPY(x,y,n) memcpy(y, x, (size_t)(n))
+#   define BZERO(x,n)  memset(x, 0, (size_t)(n))
+# else
+#   define BCOPY(x,y,n) bcopy((char *)(x),(char *)(y),(int)(n))
+#   define BZERO(x,n) bzero((char *)(x),(int)(n))
+# endif
+
+/* HBLKSIZE aligned allocation.  0 is taken to mean failure 	*/
+/* space is assumed to be cleared.				*/
+# ifdef PCR
+    char * real_malloc();
+#   define GET_MEM(bytes) HBLKPTR(real_malloc((size_t)bytes + HBLKSIZE) \
+				  + HBLKSIZE-1)
+# else
+#   ifdef OS2
+      void * os2_alloc(size_t bytes);
+#     define GET_MEM(bytes) HBLKPTR((ptr_t)os2_alloc((size_t)bytes + HBLKSIZE) \
+                                    + HBLKSIZE-1)
+#   else
+#     if defined(AMIGA) || defined(NEXT)
+#       define GET_MEM(bytes) HBLKPTR(calloc(1, (size_t)bytes + HBLKSIZE) \
+				+ HBLKSIZE-1)
+#     else
+#	ifdef MSWIN32
+          extern ptr_t GC_win32_get_mem();
+#         define GET_MEM(bytes) (struct hblk *)GC_win32_get_mem(bytes)
+#	else
+          extern ptr_t GC_unix_get_mem();
+#         define GET_MEM(bytes) (struct hblk *)GC_unix_get_mem(bytes)
+#	endif
+#     endif
+#   endif
+# endif
+
+/*
+ * Mutual exclusion between allocator/collector routines.
+ * Needed if there is more than one allocator thread.
+ * FASTLOCK() is assumed to try to acquire the lock in a cheap and
+ * dirty way that is acceptable for a few instructions, e.g. by
+ * inhibiting preemption.  This is assumed to have succeeded only
+ * if a subsequent call to FASTLOCK_SUCCEEDED() returns TRUE.
+ * FASTUNLOCK() is called whether or not FASTLOCK_SUCCEEDED().
+ * If signals cannot be tolerated with the FASTLOCK held, then
+ * FASTLOCK should disable signals.  The code executed under
+ * FASTLOCK is otherwise immune to interruption, provided it is
+ * not restarted.
+ * DCL_LOCK_STATE declares any local variables needed by LOCK and UNLOCK
+ * and/or DISABLE_SIGNALS and ENABLE_SIGNALS and/or FASTLOCK.
+ * (There is currently no equivalent for FASTLOCK.)
+ */  
+# ifdef THREADS
+#  ifdef PCR_OBSOLETE	/* Faster, but broken with multiple lwp's	*/
+#    include  "th/PCR_Th.h"
+#    include  "th/PCR_ThCrSec.h"
+     extern struct PCR_Th_MLRep GC_allocate_ml;
+#    define DCL_LOCK_STATE  PCR_sigset_t GC_old_sig_mask
+#    define LOCK() PCR_Th_ML_Acquire(&GC_allocate_ml) 
+#    define UNLOCK() PCR_Th_ML_Release(&GC_allocate_ml)
+#    define FASTLOCK() PCR_ThCrSec_EnterSys()
+     /* Here we cheat (a lot): */
+#        define FASTLOCK_SUCCEEDED() (*(int *)(&GC_allocate_ml) == 0)
+		/* TRUE if nobody currently holds the lock */
+#    define FASTUNLOCK() PCR_ThCrSec_ExitSys()
+#  endif
+#  ifdef PCR
+#    include <base/PCR_Base.h>
+#    include <th/PCR_Th.h>
+     extern PCR_Th_ML GC_allocate_ml;
+#    define DCL_LOCK_STATE  PCR_ERes GC_fastLockRes; PCR_sigset_t GC_old_sig_mas
+k
+#    define LOCK() PCR_Th_ML_Acquire(&GC_allocate_ml)
+#    define UNLOCK() PCR_Th_ML_Release(&GC_allocate_ml)
+#    define FASTLOCK() (GC_fastLockRes = PCR_Th_ML_Try(&GC_allocate_ml))
+#    define FASTLOCK_SUCCEEDED() (GC_fastLockRes == PCR_ERes_okay)
+#    define FASTUNLOCK()  {\
+        if( FASTLOCK_SUCCEEDED() ) PCR_Th_ML_Release(&GC_allocate_ml); }
+#  endif
+#  ifdef SRC_M3
+     extern word RT0u__inCritical;
+#    define LOCK() RT0u__inCritical++
+#    define UNLOCK() RT0u__inCritical--
+#  endif
+#  ifdef SOLARIS_THREADS
+#    include <thread.h>
+#    include <signal.h>
+     extern mutex_t GC_allocate_ml;
+#    define LOCK() mutex_lock(&GC_allocate_ml);
+#    define UNLOCK() mutex_unlock(&GC_allocate_ml);
+#  endif
+# else
+#    define LOCK()
+#    define UNLOCK()
+# endif
+
+# ifndef DCL_LOCK_STATE
+#   define DCL_LOCK_STATE
+# endif
+# ifndef FASTLOCK
+#   define FASTLOCK() LOCK()
+#   define FASTLOCK_SUCCEEDED() TRUE
+#   define FASTUNLOCK() UNLOCK()
+# endif
+
+/* Delay any interrupts or signals that may abort this thread.  Data	*/
+/* structures are in a consistent state outside this pair of calls.	*/
+/* ANSI C allows both to be empty (though the standard isn't very	*/
+/* clear on that point).  Standard malloc implementations are usually	*/
+/* neither interruptable nor thread-safe, and thus correspond to	*/
+/* empty definitions.							*/
+# ifdef PCR
+#   define DISABLE_SIGNALS() \
+		 PCR_Th_SetSigMask(PCR_allSigsBlocked,&GC_old_sig_mask)
+#   define ENABLE_SIGNALS() \
+		PCR_Th_SetSigMask(&GC_old_sig_mask, NIL)
+# else
+#   if defined(SRC_M3) || defined(AMIGA) || defined(SOLARIS_THREADS) || defined(MSWIN32)
+			/* Also useful for debugging, and unusually	*/
+	 		/* correct client code.				*/
+	/* Should probably use thr_sigsetmask for SOLARIS_THREADS. */
+#     define DISABLE_SIGNALS()
+#     define ENABLE_SIGNALS()
+#   else
+#     define DISABLE_SIGNALS() GC_disable_signals()
+	void GC_disable_signals();
+#     define ENABLE_SIGNALS() GC_enable_signals()
+	void GC_enable_signals();
+#   endif
+# endif
+
+/*
+ * Stop and restart mutator threads.
+ */
+# ifdef PCR
+#     include "th/PCR_ThCtl.h"
+#     define STOP_WORLD() \
+ 	PCR_ThCtl_SetExclusiveMode(PCR_ThCtl_ExclusiveMode_stopNormal, \
+ 				   PCR_allSigsBlocked, \
+ 				   PCR_waitForever)
+#     define START_WORLD() \
+	PCR_ThCtl_SetExclusiveMode(PCR_ThCtl_ExclusiveMode_null, \
+ 				   PCR_allSigsBlocked, \
+ 				   PCR_waitForever);
+# else
+#   ifdef SOLARIS_THREADS
+#     define STOP_WORLD() GC_stop_world()
+#     define START_WORLD() GC_start_world()
+#   else
+#     define STOP_WORLD()
+#     define START_WORLD()
+#   endif
+# endif
+
+/* Abandon ship */
+# ifdef PCR
+    void PCR_Base_Panic(const char *fmt, ...);
+#   define ABORT(s) PCR_Base_Panic(s)
+# else
+#   ifdef SMALL_CONFIG
+#	define ABORT(msg) abort();
+#   else
+	void GC_abort();
+#       define ABORT(msg) GC_abort(msg);
+#   endif
+# endif
+
+/* Exit abnormally, but without making a mess (e.g. out of memory) */
+# ifdef PCR
+    void PCR_Base_Exit(int status);
+#   define EXIT() PCR_Base_Exit(1)
+# else
+#   define EXIT() (void)exit(1)
+# endif
+
+/* Print warning message, e.g. almost out of memory.	*/
+# define WARN(s) GC_printf0(s)
+
+/*********************************/
+/*                               */
+/* Word-size-dependent defines   */
+/*                               */
+/*********************************/
+
+#if CPP_WORDSZ == 32
+#  define WORDS_TO_BYTES(x)   ((x)<<2)
+#  define BYTES_TO_WORDS(x)   ((x)>>2)
+#  define LOGWL               ((word)5)    /* log[2] of CPP_WORDSZ */
+#  define modWORDSZ(n) ((n) & 0x1f)          /* n mod size of word	    */
+#endif
+
+#if CPP_WORDSZ == 64
+#  define WORDS_TO_BYTES(x)   ((x)<<3)
+#  define BYTES_TO_WORDS(x)   ((x)>>3)
+#  define LOGWL               ((word)6)    /* log[2] of CPP_WORDSZ */
+#  define modWORDSZ(n) ((n) & 0x3f)          /* n mod size of word	    */
+#endif
+
+#define WORDSZ ((word)CPP_WORDSZ)
+#define SIGNB  ((word)1 << (WORDSZ-1))
+#define BYTES_PER_WORD      ((word)(sizeof (word)))
+#define ONES                ((word)(-1))
+#define divWORDSZ(n) ((n) >> LOGWL)	   /* divide n by size of word      */
+
+/*********************/
+/*                   */
+/*  Size Parameters  */
+/*                   */
+/*********************/
+
+/*  heap block size, bytes. Should be power of 2 */
+
+#ifdef SMALL_CONFIG
+#   define CPP_LOG_HBLKSIZE 10
+#else
+# if CPP_WORDSZ == 32
+#   define CPP_LOG_HBLKSIZE 12
+# else
+#   define CPP_LOG_HBLKSIZE 13
+# endif
+#endif
+#define LOG_HBLKSIZE   ((word)CPP_LOG_HBLKSIZE)
+#define CPP_HBLKSIZE (1 << CPP_LOG_HBLKSIZE)
+#define HBLKSIZE ((word)CPP_HBLKSIZE)
+
+
+/*  max size objects supported by freelist (larger objects may be   */
+/*  allocated, but less efficiently)                                */
+
+#define CPP_MAXOBJSZ    BYTES_TO_WORDS(CPP_HBLKSIZE/2)
+#define MAXOBJSZ ((word)CPP_MAXOBJSZ)
+		
+# define divHBLKSZ(n) ((n) >> LOG_HBLKSIZE)
+
+# define HBLK_PTR_DIFF(p,q) divHBLKSZ((ptr_t)p - (ptr_t)q)
+	/* Equivalent to subtracting 2 hblk pointers.	*/
+	/* We do it this way because a compiler should	*/
+	/* find it hard to use an integer division	*/
+	/* instead of a shift.  The bundled SunOS 4.1	*/
+	/* o.w. sometimes pessimizes the subtraction to	*/
+	/* involve a call to .div.			*/
+ 
+# define modHBLKSZ(n) ((n) & (HBLKSIZE-1))
+ 
+# define HBLKPTR(objptr) ((struct hblk *)(((word) (objptr)) & ~(HBLKSIZE-1)))
+
+# define HBLKDISPL(objptr) (((word) (objptr)) & (HBLKSIZE-1))
+
+/* Round up byte allocation requests to integral number of words, etc. */
+# ifdef ADD_BYTE_AT_END
+#   define ROUNDED_UP_WORDS(n) BYTES_TO_WORDS((n) + WORDS_TO_BYTES(1))
+#   define SMALL_OBJ(bytes) ((bytes) < WORDS_TO_BYTES(MAXOBJSZ))
+#   define ADD_SLOP(bytes) ((bytes)+1)
+# else
+#   define ROUNDED_UP_WORDS(n) BYTES_TO_WORDS((n) + (WORDS_TO_BYTES(1) - 1))
+#   define SMALL_OBJ(bytes) ((bytes) <= WORDS_TO_BYTES(MAXOBJSZ))
+#   define ADD_SLOP(bytes) (bytes)
+# endif
+
+
+/*
+ * Hash table representation of sets of pages.  This assumes it is
+ * OK to add spurious entries to sets.
+ * Used by black-listing code, and perhaps by dirty bit maintenance code.
+ */
+ 
+# define LOG_PHT_ENTRIES  14	/* Collisions are likely if heap grows	*/
+				/* to more than 16K hblks = 64MB.	*/
+				/* Each hash table occupies 2K bytes.   */
+# define PHT_ENTRIES ((word)1 << LOG_PHT_ENTRIES)
+# define PHT_SIZE (PHT_ENTRIES >> LOGWL)
+typedef word page_hash_table[PHT_SIZE];
+
+# define PHT_HASH(addr) ((((word)(addr)) >> LOG_HBLKSIZE) & (PHT_ENTRIES - 1))
+
+# define get_pht_entry_from_index(bl, index) \
+		(((bl)[divWORDSZ(index)] >> modWORDSZ(index)) & 1)
+# define set_pht_entry_from_index(bl, index) \
+		(bl)[divWORDSZ(index)] |= (word)1 << modWORDSZ(index)
+# define clear_pht_entry_from_index(bl, index) \
+		(bl)[divWORDSZ(index)] &= ~((word)1 << modWORDSZ(index))
+	
+
+
+/********************************************/
+/*                                          */
+/*    H e a p   B l o c k s                 */
+/*                                          */
+/********************************************/
+
+/*  heap block header */
+#define HBLKMASK   (HBLKSIZE-1)
+
+#define BITS_PER_HBLK (HBLKSIZE * 8)
+
+#define MARK_BITS_PER_HBLK (BITS_PER_HBLK/CPP_WORDSZ)
+	   /* upper bound                                    */
+	   /* We allocate 1 bit/word.  Only the first word   */
+	   /* in each object is actually marked.             */
+
+# ifdef ALIGN_DOUBLE
+#   define MARK_BITS_SZ (((MARK_BITS_PER_HBLK + 2*CPP_WORDSZ - 1) \
+			  / (2*CPP_WORDSZ))*2)
+# else
+#   define MARK_BITS_SZ ((MARK_BITS_PER_HBLK + CPP_WORDSZ - 1)/CPP_WORDSZ)
+# endif
+	   /* Upper bound on number of mark words per heap block  */
+
+struct hblkhdr {
+    word hb_sz;  /* If in use, size in words, of objects in the block. */
+		 /* if free, the size in bytes of the whole block      */
+    struct hblk * hb_next; 	/* Link field for hblk free list	 */
+    				/* and for lists of chunks waiting to be */
+    				/* reclaimed.				 */
+    word hb_descr;   		/* object descriptor for marking.  See	*/
+    				/* mark.h.				*/
+    char* hb_map;	/* A pointer to a pointer validity map of the block. */
+    		      	/* See GC_obj_map.				     */
+    		     	/* Valid for all blocks with headers.		     */
+    		     	/* Free blocks point to GC_invalid_map.		     */
+    unsigned char hb_obj_kind;
+    			 /* Kind of objects in the block.  Each kind 	*/
+    			 /* identifies a mark procedure and a set of 	*/
+    			 /* list headers.  Sometimes called regions.	*/
+    unsigned char hb_flags;
+#	define IGNORE_OFF_PAGE	1	/* Ignore pointers that do not	*/
+					/* point to the first page of 	*/
+					/* this object.			*/
+    unsigned short hb_last_reclaimed;
+    				/* Value of GC_gc_no when block was	*/
+    				/* last allocated or swept. May wrap.   */
+    word hb_marks[MARK_BITS_SZ];
+			    /* Bit i in the array refers to the             */
+			    /* object starting at the ith word (header      */
+			    /* INCLUDED) in the heap block.                 */
+			    /* The lsb of word 0 is numbered 0.		    */
+};
+
+/*  heap block body */
+
+# define DISCARD_WORDS 0
+	/* Number of words to be dropped at the beginning of each block	*/
+	/* Must be a multiple of WORDSZ.  May reasonably be nonzero	*/
+	/* on machines that don't guarantee longword alignment of	*/
+	/* pointers, so that the number of false hits is minimized.	*/
+	/* 0 and WORDSZ are probably the only reasonable values.	*/
+
+# define BODY_SZ ((HBLKSIZE-WORDS_TO_BYTES(DISCARD_WORDS))/sizeof(word))
+
+struct hblk {
+#   if (DISCARD_WORDS != 0)
+        word garbage[DISCARD_WORDS];
+#   endif
+    word hb_body[BODY_SZ];
+};
+
+# define HDR_WORDS ((word)DISCARD_WORDS)
+# define HDR_BYTES ((word)WORDS_TO_BYTES(DISCARD_WORDS))
+
+# define OBJ_SZ_TO_BLOCKS(sz) \
+    divHBLKSZ(HDR_BYTES + WORDS_TO_BYTES(sz) + HBLKSIZE-1)
+    /* Size of block (in units of HBLKSIZE) needed to hold objects of	*/
+    /* given sz (in words).						*/
+
+/* Object free list link */
+# define obj_link(p) (*(ptr_t *)(p))
+
+/*  lists of all heap blocks and free lists	*/
+/* These are grouped together in a struct	*/
+/* so that they can be easily skipped by the	*/
+/* GC_mark routine.				*/
+/* The ordering is weird to make GC_malloc	*/
+/* faster by keeping the important fields	*/
+/* sufficiently close together that a		*/
+/* single load of a base register will do.	*/
+/* Scalars that could easily appear to		*/
+/* be pointers are also put here.		*/
+
+struct _GC_arrays {
+  word _heapsize;
+  ptr_t _last_heap_addr;
+  ptr_t _prev_heap_addr;
+  word _words_allocd_before_gc;
+		/* Number of words allocated before this	*/
+		/* collection cycle.				*/
+# ifdef GATHERSTATS
+    word _composite_in_use;
+   		/* Number of words in accessible composite	*/
+		/* objects.					*/
+    word _atomic_in_use;
+   		/* Number of words in accessible atomic		*/
+		/* objects.					*/
+# endif
+  word _words_allocd;
+  	/* Number of words allocated during this collection cycle */
+  word _words_wasted;
+  	/* Number of words wasted due to internal fragmentation	 */
+  	/* in large objects allocated since last gc. Approximate.*/
+  word _non_gc_bytes_at_gc;
+  	/* Number of explicitly managed bytes of storage 	*/
+  	/* at last collection.					*/
+  word _mem_freed;
+  	/* Number of explicitly deallocated words of memory	*/
+  	/* since last collection.				*/
+  	
+  ptr_t _objfreelist[MAXOBJSZ+1];
+			  /* free list for objects */
+# ifdef MERGE_SIZES
+    unsigned _size_map[WORDS_TO_BYTES(MAXOBJSZ+1)];
+    	/* Number of words to allocate for a given allocation request in */
+    	/* bytes.							 */
+# endif 
+  ptr_t _aobjfreelist[MAXOBJSZ+1];
+			  /* free list for atomic objs 	*/
+
+  ptr_t _uobjfreelist[MAXOBJSZ+1];
+			  /* uncollectable but traced objs 	*/
+
+# ifdef STUBBORN_ALLOC
+    ptr_t _sobjfreelist[MAXOBJSZ+1];
+# endif
+  			  /* free list for immutable objects	*/
+  ptr_t _obj_map[MAXOBJSZ+1];
+                       /* If not NIL, then a pointer to a map of valid  */
+    		       /* object addresses. hbh_map[sz][i] is j if the  */
+    		       /* address block_start+i is a valid pointer      */
+    		       /* to an object at				*/
+    		       /* block_start+i&~3 - WORDS_TO_BYTES(j).		*/
+    		       /* (If ALL_INTERIOR_POINTERS is defined, then	*/
+    		       /* instead ((short *)(hbh_map[sz])[i] is j if	*/
+    		       /* block_start+WORDS_TO_BYTES(i) is in the	*/
+    		       /* interior of an object starting at		*/
+    		       /* block_start+WORDS_TO_BYTES(i-j)).		*/
+    		       /* It is OBJ_INVALID if				*/
+    		       /* block_start+WORDS_TO_BYTES(i) is not		*/
+    		       /* valid as a pointer to an object.              */
+    		       /* We assume that all values of j <= OBJ_INVALID */
+    		       /* The zeroth entry corresponds to large objects.*/
+#   ifdef ALL_INTERIOR_POINTERS
+#	define map_entry_type short
+#       define OBJ_INVALID 0x7fff
+#	define MAP_ENTRY(map, bytes) \
+		(((map_entry_type *)(map))[BYTES_TO_WORDS(bytes)])
+#	define MAP_ENTRIES BYTES_TO_WORDS(HBLKSIZE)
+#	define MAP_SIZE (MAP_ENTRIES * sizeof(map_entry_type))
+#	define OFFSET_VALID(displ) TRUE
+#	define CPP_MAX_OFFSET (HBLKSIZE - HDR_BYTES - 1)
+#	define MAX_OFFSET ((word)CPP_MAX_OFFSET)
+#   else
+#	define map_entry_type char
+#       define OBJ_INVALID 0x7f
+#	define MAP_ENTRY(map, bytes) \
+		(map)[bytes]
+#	define MAP_ENTRIES HBLKSIZE
+#	define MAP_SIZE MAP_ENTRIES
+#	define CPP_MAX_OFFSET (WORDS_TO_BYTES(OBJ_INVALID) - 1)	
+#	define MAX_OFFSET ((word)CPP_MAX_OFFSET)
+# 	define VALID_OFFSET_SZ \
+	  (CPP_MAX_OFFSET > WORDS_TO_BYTES(CPP_MAXOBJSZ)? \
+	   CPP_MAX_OFFSET+1 \
+	   : WORDS_TO_BYTES(CPP_MAXOBJSZ)+1)
+  	char _valid_offsets[VALID_OFFSET_SZ];
+				/* GC_valid_offsets[i] == TRUE ==> i 	*/
+				/* is registered as a displacement.	*/
+#	define OFFSET_VALID(displ) GC_valid_offsets[displ]
+  	char _modws_valid_offsets[sizeof(word)];
+				/* GC_valid_offsets[i] ==>		  */
+				/* GC_modws_valid_offsets[i%sizeof(word)] */
+#   endif
+  struct hblk * _reclaim_list[MAXOBJSZ+1];
+  struct hblk * _areclaim_list[MAXOBJSZ+1];
+  struct hblk * _ureclaim_list[MAXOBJSZ+1];
+# ifdef STUBBORN_ALLOC
+      struct hblk * _sreclaim_list[MAXOBJSZ+1];
+      page_hash_table _changed_pages;
+        /* Stubborn object pages that were changes since last call to	*/
+	/* GC_read_changed.						*/
+      page_hash_table _prev_changed_pages;
+        /* Stubborn object pages that were changes before last call to	*/
+	/* GC_read_changed.						*/
+# endif
+# if defined(PROC_VDB) || defined(MPROTECT_VDB)
+      page_hash_table _grungy_pages; /* Pages that were dirty at last 	   */
+				     /* GC_read_dirty.			   */
+# endif
+# define MAX_HEAP_SECTS 256	/* Separately added heap sections. */
+  struct HeapSect {
+      ptr_t hs_start; word hs_bytes;
+  } _heap_sects[MAX_HEAP_SECTS];
+# ifdef MSWIN32
+    ptr_t _heap_bases[MAX_HEAP_SECTS];
+    		/* Start address of memory regions obtained from kernel. */
+# endif
+  /* Block header index; see gc_headers.h */
+  bottom_index _all_nils;
+  bottom_index * _top_index [TOP_SZ];
+};
+
+extern GC_FAR struct _GC_arrays GC_arrays; 
+
+# define GC_objfreelist GC_arrays._objfreelist
+# define GC_aobjfreelist GC_arrays._aobjfreelist
+# define GC_uobjfreelist GC_arrays._uobjfreelist
+# define GC_sobjfreelist GC_arrays._sobjfreelist
+# define GC_valid_offsets GC_arrays._valid_offsets
+# define GC_modws_valid_offsets GC_arrays._modws_valid_offsets
+# define GC_reclaim_list GC_arrays._reclaim_list
+# define GC_areclaim_list GC_arrays._areclaim_list
+# define GC_ureclaim_list GC_arrays._ureclaim_list
+# ifdef STUBBORN_ALLOC
+#    define GC_sreclaim_list GC_arrays._sreclaim_list
+#    define GC_changed_pages GC_arrays._changed_pages
+#    define GC_prev_changed_pages GC_arrays._prev_changed_pages
+# endif
+# define GC_obj_map GC_arrays._obj_map
+# define GC_last_heap_addr GC_arrays._last_heap_addr
+# define GC_prev_heap_addr GC_arrays._prev_heap_addr
+# define GC_words_allocd GC_arrays._words_allocd
+# define GC_words_wasted GC_arrays._words_wasted
+# define GC_non_gc_bytes_at_gc GC_arrays._non_gc_bytes_at_gc
+# define GC_mem_freed GC_arrays._mem_freed
+# define GC_heapsize GC_arrays._heapsize
+# define GC_words_allocd_before_gc GC_arrays._words_allocd_before_gc
+# define GC_heap_sects GC_arrays._heap_sects
+# ifdef MSWIN32
+#   define GC_heap_bases GC_arrays._heap_bases
+# endif
+# define GC_all_nils GC_arrays._all_nils
+# define GC_top_index GC_arrays._top_index
+# if defined(PROC_VDB) || defined(MPROTECT_VDB)
+#   define GC_grungy_pages GC_arrays._grungy_pages
+# endif
+# ifdef GATHERSTATS
+#   define GC_composite_in_use GC_arrays._composite_in_use
+#   define GC_atomic_in_use GC_arrays._atomic_in_use
+# endif
+# ifdef MERGE_SIZES
+#   define GC_size_map GC_arrays._size_map
+# endif
+
+# define beginGC_arrays ((ptr_t)(&GC_arrays))
+# define endGC_arrays (((ptr_t)(&GC_arrays)) + (sizeof GC_arrays))
+
+
+# define MAXOBJKINDS 16
+
+/* Object kinds: */
+extern struct obj_kind {
+   ptr_t *ok_freelist;	/* Array of free listheaders for this kind of object */
+   			/* Point either to GC_arrays or to storage allocated */
+   			/* with GC_scratch_alloc.			     */
+   struct hblk **ok_reclaim_list;
+   			/* List headers for lists of blocks waiting to be */
+   			/* swept.					  */
+   word ok_descriptor;  /* Descriptor template for objects in this	*/
+   			/* block.					*/
+   bool ok_relocate_descr;
+   			/* Add object size in bytes to descriptor 	*/
+   			/* template to obtain descriptor.  Otherwise	*/
+   			/* template is used as is.			*/
+   bool ok_init;     /* Clear objects before putting them on the free list. */
+} GC_obj_kinds[MAXOBJKINDS];
+/* Predefined kinds: */
+# define PTRFREE 0
+# define NORMAL  1
+# define UNCOLLECTABLE 2
+# define STUBBORN 3
+
+extern int GC_n_kinds;
+
+extern word GC_n_heap_sects;	/* Number of separately added heap	*/
+				/* sections.				*/
+
+# ifdef MSWIN32
+extern word GC_n_heap_bases;	/* See GC_heap_bases.	*/
+# endif
+
+extern char * GC_invalid_map;
+			/* Pointer to the nowhere valid hblk map */
+			/* Blocks pointing to this map are free. */
+
+extern struct hblk * GC_hblkfreelist;
+				/* List of completely empty heap blocks	*/
+				/* Linked through hb_next field of 	*/
+				/* header structure associated with	*/
+				/* block.				*/
+
+extern bool GC_is_initialized;		/* GC_init() has been run.	*/
+
+extern bool GC_objects_are_marked;	/* There are marked objects in  */
+					/* the heap.			*/
+
+extern int GC_incremental;  /* Using incremental/generational collection. */
+
+extern bool GC_dirty_maintained;/* Dirty bits are being maintained, 	*/
+				/* either for incremental collection,	*/
+				/* or to limit the root set.		*/
+
+# ifndef PCR
+    extern ptr_t GC_stackbottom;	/* Cool end of user stack	*/
+# endif
+
+extern word GC_root_size;	/* Total size of registered root sections */
+
+extern bool GC_debugging_started;	/* GC_debug_malloc has been called. */ 
+
+extern ptr_t GC_least_plausible_heap_addr;
+extern ptr_t GC_greatest_plausible_heap_addr;
+			/* Bounds on the heap.  Guaranteed valid	*/
+			/* Likely to include future heap expansion.	*/
+			
+/* Operations */
+# ifndef abs
+#   define abs(x)  ((x) < 0? (-(x)) : (x))
+# endif
+
+
+/*  Marks are in a reserved area in                          */
+/*  each heap block.  Each word has one mark bit associated  */
+/*  with it. Only those corresponding to the beginning of an */
+/*  object are used.                                         */
+
+
+/* Mark bit perations */
+
+/*
+ * Retrieve, set, clear the mark bit corresponding
+ * to the nth word in a given heap block.
+ *
+ * (Recall that bit n corresponds to object beginning at word n
+ * relative to the beginning of the block, including unused words)
+ */
+
+# define mark_bit_from_hdr(hhdr,n) (((hhdr)->hb_marks[divWORDSZ(n)] \
+			    >> (modWORDSZ(n))) & (word)1)
+# define set_mark_bit_from_hdr(hhdr,n) (hhdr)->hb_marks[divWORDSZ(n)] \
+				|= (word)1 << modWORDSZ(n)
+
+# define clear_mark_bit_from_hdr(hhdr,n) (hhdr)->hb_marks[divWORDSZ(n)] \
+				&= ~((word)1 << modWORDSZ(n))
+
+/* Important internal collector routines */
+
+void GC_apply_to_all_blocks(/*fn, client_data*/);
+			/* Invoke fn(hbp, client_data) for each 	*/
+			/* allocated heap block.			*/
+struct hblk * GC_next_block(/* struct hblk * h */);
+void GC_mark_init();
+void GC_clear_marks();	/* Clear mark bits for all heap objects. */
+void GC_mark_from_mark_stack(); /* Mark from everything on the mark stack. */
+				/* Return after about one pages worth of   */
+				/* work.				   */
+bool GC_mark_stack_empty();
+bool GC_mark_some();	/* Perform about one pages worth of marking	*/
+			/* work of whatever kind is needed.  Returns	*/
+			/* quickly if no collection is in progress.	*/
+			/* Return TRUE if mark phase finished.		*/
+void GC_initiate_full();	/* initiate full collection.		*/
+void GC_initiate_partial();	/* initiate partial collection.		*/			
+void GC_push_all(/*b,t*/);	/* Push everything in a range 		*/
+				/* onto mark stack.			*/
+void GC_push_dirty(/*b,t*/);      /* Push all possibly changed	 	*/
+				  /* subintervals of [b,t) onto		*/
+				  /* mark stack.			*/
+#ifndef SMALL_CONFIG
+  void GC_push_conditional(/* ptr_t b, ptr_t t, bool all*/);
+#else
+# define GC_push_conditional(b, t, all) GC_push_all(b, t)
+#endif
+                                /* Do either of the above, depending	*/
+				/* on the third arg.			*/
+void GC_push_all_stack(/*b,t*/);    /* As above, but consider		*/
+				    /*  interior pointers as valid  	*/
+void GC_push_roots(/* bool all */); /* Push all or dirty roots.	*/
+extern void (*GC_push_other_roots)();
+			/* Push system or application specific roots	*/
+			/* onto the mark stack.  In some environments	*/
+			/* (e.g. threads environments) this is		*/
+			/* predfined to be non-zero.  A client supplied */
+			/* replacement should also call the original	*/
+			/* function.					*/
+void GC_push_regs();	/* Push register contents onto mark stack.	*/
+void GC_remark();	/* Mark from all marked objects.  Used	*/
+		 	/* only if we had to drop something.	*/
+void GC_push_one(/*p*/);	/* If p points to an object, mark it	*/
+				/* and push contents on the mark stack	*/
+void GC_push_one_checked(/*p*/); /* Ditto, omits plausibility test	*/
+void GC_push_marked(/* struct hblk h, hdr * hhdr */);
+		/* Push contents of all marked objects in h onto	*/
+		/* mark stack.						*/
+#ifdef SMALL_CONFIG
+# define GC_push_next_marked_dirty(h) GC_push_next_marked(h)
+#else
+  struct hblk * GC_push_next_marked_dirty(/* h */);
+		/* Invoke GC_push_marked on next dirty block above h.	*/
+		/* Return a pointer just past the end of this block.	*/
+#endif /* !SMALL_CONFIG */
+struct hblk * GC_push_next_marked(/* h */);
+		/* Ditto, but also mark from clean pages.	*/
+struct hblk * GC_push_next_marked_uncollectable(/* h */);
+		/* Ditto, but mark only from uncollectable pages.	*/
+bool GC_stopped_mark(); /* Stop world and mark from all roots	*/
+			/* and rescuers.			*/
+void GC_clear_hdr_marks(/* hhdr */);  /* Clear the mark bits in a header */
+void GC_add_roots_inner();
+void GC_register_dynamic_libraries();
+		/* Add dynamic library data sections to the root set. */
+
+/* Machine dependent startup routines */
+ptr_t GC_get_stack_base();
+void GC_register_data_segments();
+
+/* Black listing: */
+void GC_bl_init(); 	
+# ifndef ALL_INTERIOR_POINTERS
+    void GC_add_to_black_list_normal(/* bits */);
+			/* Register bits as a possible future false	*/
+			/* reference from the heap or static data	*/
+#   define GC_ADD_TO_BLACK_LIST_NORMAL(bits) GC_add_to_black_list_normal(bits)
+# else
+#   define GC_ADD_TO_BLACK_LIST_NORMAL(bits) GC_add_to_black_list_stack(bits)
+# endif
+
+void GC_add_to_black_list_stack(/* bits */);
+struct hblk * GC_is_black_listed(/* h, len */);
+			/* If there are likely to be false references	*/
+			/* to a block starting at h of the indicated    */
+			/* length, then return the next plausible	*/
+			/* starting location for h that might avoid	*/
+			/* these false references.			*/
+void GC_promote_black_lists();
+			/* Declare an end to a black listing phase.	*/
+		 	
+ptr_t GC_scratch_alloc(/*bytes*/);
+				/* GC internal memory allocation for	*/
+				/* small objects.  Deallocation is not  */
+				/* possible.				*/
+	
+/* Heap block layout maps: */			
+void GC_invalidate_map(/* hdr */);
+				/* Remove the object map associated	*/
+				/* with the block.  This identifies	*/
+				/* the block as invalid to the mark	*/
+				/* routines.				*/
+bool GC_add_map_entry(/*sz*/);
+				/* Add a heap block map for objects of	*/
+				/* size sz to obj_map.			*/
+				/* Return FALSE on failure.		*/
+void GC_register_displacement_inner(/*offset*/);
+				/* Version of GC_register_displacement	*/
+				/* that assumes lock is already held	*/
+				/* and signals are already disabled.	*/
+
+/*  hblk allocation: */		
+void GC_new_hblk(/*size_in_words, kind*/);
+				/* Allocate a new heap block, and build */
+				/* a free list in it.			*/				
+struct hblk * GC_allochblk(/*size_in_words, kind*/);
+				/* Allocate a heap block, clear it if	*/
+				/* for composite objects, inform	*/
+				/* the marker that block is valid	*/
+				/* for objects of indicated size.	*/
+				/* sz < 0 ==> atomic.			*/ 
+void GC_freehblk();		/* Deallocate a heap block and mark it  */
+				/* as invalid.				*/
+				
+/*  Misc GC: */
+void GC_init_inner();
+bool GC_expand_hp_inner();
+void GC_start_reclaim(/*abort_if_found*/);
+				/* Restore unmarked objects to free	*/
+				/* lists, or (if abort_if_found is	*/
+				/* TRUE) report them.			*/
+				/* Sweeping of small object pages is	*/
+				/* largely deferred.			*/
+void GC_continue_reclaim(/*size, kind*/);
+				/* Sweep pages of the given size and	*/
+				/* kind, as long as possible, and	*/
+				/* as long as the corr. free list is    */
+				/* empty.				*/
+void GC_reclaim_or_delete_all();
+				/* Arrange for all reclaim lists to be	*/
+				/* empty.  Judiciously choose between	*/
+				/* sweeping and discarding each page.	*/
+bool GC_block_empty(/* hhdr */); /* Block completely unmarked? 	*/
+void GC_gcollect_inner();
+				/* Collect; caller must have acquired	*/
+				/* lock and disabled signals.		*/
+				/* FALSE return indicates nothing was	*/
+				/* done due to insufficient allocation. */
+void GC_finish_collection();	/* Finish collection.  Mark bits are	*/
+				/* consistent and lock is still held.	*/
+bool GC_collect_or_expand(/* needed_blocks */);
+				/* Collect or expand heap in an attempt */
+				/* make the indicated number of free	*/
+				/* blocks available.  Should be called	*/
+				/* until it fails by returning FALSE.	*/
+void GC_init();			/* Initialize collector.		*/
+void GC_collect_a_little(/* n */);
+				/* Do n units worth of garbage 		*/
+				/* collection work, if appropriate.	*/
+				/* A unit is an amount appropriate for  */
+				/* HBLKSIZE bytes of allocation.	*/
+ptr_t GC_generic_malloc(/* bytes, kind */);
+				/* Allocate an object of the given	*/
+				/* kind.  By default, there are only	*/
+				/* two kinds: composite, and atomic.	*/
+				/* We claim it's possible for clever	*/
+				/* client code that understands GC	*/
+				/* internals to add more, e.g. to	*/
+				/* communicate object layout info	*/
+				/* to the collector.			*/
+ptr_t GC_generic_malloc_inner(/* bytes, kind */);
+				/* Ditto, but I already hold lock, etc.	*/
+ptr_t GC_generic_malloc_words_small(/*words, kind*/);
+				/* As above, but size in units of words */
+				/* Bypasses MERGE_SIZES.  Assumes	*/
+				/* words <= MAXOBJSZ.			*/
+ptr_t GC_malloc_ignore_off_page_inner(/* bytes */);
+				/* Allocate an object, where		*/
+				/* the client guarantees that there	*/
+				/* will always be a pointer to the 	*/
+				/* beginning of the object while the	*/
+				/* object is live.			*/
+ptr_t GC_allocobj(/* sz_inn_words, kind */);
+				/* Make the indicated 			*/
+				/* free list nonempty, and return its	*/
+				/* head.				*/
+
+void GC_init_headers();
+bool GC_install_header(/*h*/);
+				/* Install a header for block h.	*/
+				/* Return FALSE on failure.		*/
+bool GC_install_counts(/*h, sz*/);
+				/* Set up forwarding counts for block	*/
+				/* h of size sz.			*/
+				/* Return FALSE on failure.		*/
+void GC_remove_header(/*h*/);
+				/* Remove the header for block h.	*/
+void GC_remove_counts(/*h, sz*/);
+				/* Remove forwarding counts for h.	*/
+hdr * GC_find_header(/*p*/);	/* Debugging only.			*/
+
+void GC_finalize();	/* Perform all indicated finalization actions	*/
+			/* on unmarked objects.				*/
+			/* Unreachable finalizable objects are enqueued	*/
+			/* for processing by GC_invoke_finalizers.	*/
+			/* Invoked with lock.				*/
+void GC_invoke_finalizers(); 	/* Run eligible finalizers.	*/
+				/* Invoked without lock.	*/	
+			
+void GC_add_to_heap(/*p, bytes*/);
+			/* Add a HBLKSIZE aligned chunk to the heap.	*/
+
+void GC_print_obj(/* ptr_t p */);
+			/* P points to somewhere inside an object with	*/
+			/* debugging info.  Print a human readable	*/
+			/* description of the object to stderr.		*/
+extern void (*GC_check_heap)();
+			/* Check that all objects in the heap with 	*/
+			/* debugging info are intact.  Print 		*/
+			/* descriptions of any that are not.		*/
+			
+/* Virtual dirty bit implementation:		*/
+/* Each implementation exports the following:	*/
+void GC_read_dirty();	/* Retrieve dirty bits.	*/
+bool GC_page_was_dirty(/* struct hblk * h  */);
+			/* Read retrieved dirty bits.	*/
+bool GC_page_was_ever_dirty(/* struct hblk * h  */);
+			/* Could the page contain valid heap pointers?	*/
+void GC_is_fresh(/* struct hblk * h, word number_of_blocks  */);
+			/* Assert the region currently contains no	*/
+			/* valid pointers.				*/
+void GC_write_hint(/* struct hblk * h  */);
+			/* h is about to be written.	*/
+void GC_dirty_init();
+
+/* Slow/general mark bit manipulation: */
+bool GC_is_marked();
+void GC_clear_mark_bit();
+void GC_set_mark_bit();
+
+/* Stubborn objects: */
+void GC_read_changed();	/* Analogous to GC_read_dirty */
+bool GC_page_was_changed(/* h */);	/* Analogous to GC_page_was_dirty */
+void GC_clean_changing_list();	/* Collect obsolete changing list entries */
+void GC_stubborn_init();
+
+/* Debugging print routines: */
+void GC_print_block_list();
+void GC_print_hblkfreelist();
+
+/* Make arguments appear live to compiler */
+void GC_noop();
+
+/* Logging and diagnostic output: 	*/
+void GC_printf(/* format, a, b, c, d, e, f */);
+			/* A version of printf that doesn't allocate,	*/
+			/* is restricted to long arguments, and		*/
+			/* (unfortunately) doesn't use varargs for	*/
+			/* portability.  Restricted to 6 args and	*/
+			/* 1K total output length.			*/
+			/* (We use sprintf.  Hopefully that doesn't	*/
+			/* allocate for long arguments.)  		*/
+# define GC_printf0(f) GC_printf(f, 0l, 0l, 0l, 0l, 0l, 0l)
+# define GC_printf1(f,a) GC_printf(f, (long)a, 0l, 0l, 0l, 0l, 0l)
+# define GC_printf2(f,a,b) GC_printf(f, (long)a, (long)b, 0l, 0l, 0l, 0l)
+# define GC_printf3(f,a,b,c) GC_printf(f, (long)a, (long)b, (long)c, 0l, 0l, 0l)
+# define GC_printf4(f,a,b,c,d) GC_printf(f, (long)a, (long)b, (long)c, \
+					    (long)d, 0l, 0l)
+# define GC_printf5(f,a,b,c,d,e) GC_printf(f, (long)a, (long)b, (long)c, \
+					      (long)d, (long)e, 0l)
+# define GC_printf6(f,a,b,c,d,e,g) GC_printf(f, (long)a, (long)b, (long)c, \
+						(long)d, (long)e, (long)g)
+
+void GC_err_printf(/* format, a, b, c, d, e, f */);
+# define GC_err_printf0(f) GC_err_puts(f)
+# define GC_err_printf1(f,a) GC_err_printf(f, (long)a, 0l, 0l, 0l, 0l, 0l)
+# define GC_err_printf2(f,a,b) GC_err_printf(f, (long)a, (long)b, 0l, 0l, 0l, 0l)
+# define GC_err_printf3(f,a,b,c) GC_err_printf(f, (long)a, (long)b, (long)c, \
+						  0l, 0l, 0l)
+# define GC_err_printf4(f,a,b,c,d) GC_err_printf(f, (long)a, (long)b, \
+						    (long)c, (long)d, 0l, 0l)
+# define GC_err_printf5(f,a,b,c,d,e) GC_err_printf(f, (long)a, (long)b, \
+						      (long)c, (long)d, \
+						      (long)e, 0l)
+# define GC_err_printf6(f,a,b,c,d,e,g) GC_err_printf(f, (long)a, (long)b, \
+							(long)c, (long)d, \
+							(long)e, (long)g)
+			/* Ditto, writes to stderr.			*/
+			
+void GC_err_puts(/* char *s */);
+			/* Write s to stderr, don't buffer, don't add	*/
+			/* newlines, don't ...				*/
+
+# endif /* GC_PRIVATE_H */
diff --git a/gc_private.h b/gc_private.h
new file mode 100644
index 00000000..3dd7c855
--- /dev/null
+++ b/gc_private.h
@@ -0,0 +1 @@
+# include "gc_priv.h"
diff --git a/gc_typed.h b/gc_typed.h
new file mode 100644
index 00000000..f7cc2f22
--- /dev/null
+++ b/gc_typed.h
@@ -0,0 +1,85 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/*
+ * Some simple primitives for allocation with explicit type information.
+ * Facilities for dynamic type inference may be added later.
+ * Should be used only for extremely performance critical applications,
+ * or if conservative collector leakage is otherwise a problem (unlikely).
+ * Note that this is implemented completely separately from the rest
+ * of the collector, and is not linked in unless referenced.
+ */
+/* Boehm, May 19, 1994 2:13 pm PDT */
+
+#ifndef _GC_TYPED_H
+# define _GC_TYPED_H
+# ifndef _GC_H
+#   include "gc.h"
+# endif
+
+typedef GC_word * GC_bitmap;
+	/* The least significant bit of the first word is one if	*/
+	/* the first word in the object may be a pointer.		*/
+	
+# define GC_get_bit(bm, index) \
+		(((bm)[divWORDSZ(index)] >> modWORDSZ(index)) & 1)
+# define GC_set_bit(bm, index) \
+		(bm)[divWORDSZ(index)] |= (word)1 << modWORDSZ(index)
+
+typedef GC_word GC_descr;
+
+#if defined(__STDC__) || defined(__cplusplus)
+  extern GC_descr GC_make_descriptor(GC_bitmap bm, size_t len);
+#else
+  extern GC_descr GC_make_descriptor(/* GC_bitmap bm, size_t len */);
+#endif
+		/* Return a type descriptor for the object whose layout	*/
+		/* is described by the argument.			*/
+		/* The least significant bit of the first word is one	*/
+		/* if the first word in the object may be a pointer.	*/
+		/* The second argument specifies the number of		*/
+		/* meaningful bits in the bitmap.  The actual object 	*/
+		/* may be larger (but not smaller).  Any additional	*/
+		/* words in the object are assumed not to contain 	*/
+		/* pointers.						*/
+		/* Returns a conservative approximation in the		*/
+		/* (unlikely) case of insufficient memory to build	*/
+		/* the descriptor.  Calls to GC_make_descriptor		*/
+		/* may consume some amount of a finite resource.  This	*/
+		/* is intended to be called once per type, not once	*/
+		/* per allocation.					*/
+
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_malloc_explicitly_typed(size_t size_in_bytes, GC_descr d);
+#else
+  extern char * GC_malloc_explicitly_typed(/* size_in_bytes, descriptor */);
+#endif
+		/* Allocate an object whose layout is described by d.	*/
+		/* The resulting object MAY NOT BE PASSED TO REALLOC.	*/
+		
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_calloc_explicitly_typed(size_t nelements,
+  					   size_t element_size_in_bytes,
+  					   GC_descr d);
+#else
+  char * GC_calloc_explicitly_typed(/* nelements, size_in_bytes, descriptor */);
+  	/* Allocate an array of nelements elements, each of the	*/
+  	/* given size, and with the given descriptor.		*/
+  	/* The elemnt size must be a multiple of the byte	*/
+  	/* alignment required for pointers.  E.g. on a 32-bit	*/
+  	/* machine with 16-bit aligned pointers, size_in_bytes	*/
+  	/* must be a multiple of 2.				*/
+#endif
+
+#endif /* _GC_TYPED_H */
+
diff --git a/headers.c b/headers.c
new file mode 100644
index 00000000..2efa27a8
--- /dev/null
+++ b/headers.c
@@ -0,0 +1,269 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:08 pm PDT */
+ 
+/*
+ * This implements:
+ * 1. allocation of heap block headers
+ * 2. A map from addresses to heap block addresses to heap block headers
+ *
+ * Access speed is crucial.  We implement an index structure based on a 2
+ * level tree.
+ */
+ 
+# include "gc_priv.h"
+
+bottom_index * GC_all_bottom_indices = 0;
+ 
+/* Non-macro version of header location routine */
+hdr * GC_find_header(h)
+ptr_t h;
+{
+#   ifdef HASH_TL
+	register hdr * result;
+	GET_HDR(h, result);
+	return(result);
+#   else
+	return(HDR_INNER(h));
+#   endif
+}
+ 
+/* Routines to dynamically allocate collector data structures that will */
+/* never be freed.							 */
+ 
+static ptr_t scratch_free_ptr = 0;
+ 
+ptr_t GC_scratch_end_ptr = 0;
+ 
+ptr_t GC_scratch_alloc(bytes)
+register word bytes;
+{
+    register ptr_t result = scratch_free_ptr;
+    scratch_free_ptr += bytes;
+    if (scratch_free_ptr <= GC_scratch_end_ptr) {
+        return(result);
+    }
+    {
+        word bytes_to_get = MINHINCR * HBLKSIZE;
+         
+        if (bytes_to_get <= bytes) {
+          /* Undo the damage, and get memory directly */
+            scratch_free_ptr -= bytes;
+            return((ptr_t)GET_MEM(bytes));
+        }
+        result = (ptr_t)GET_MEM(bytes_to_get);
+        if (result == 0) {
+#	    ifdef PRINTSTATS
+                GC_printf0("Out of memory - trying to allocate less\n");
+#	    endif
+            scratch_free_ptr -= bytes;
+            return((ptr_t)GET_MEM(bytes));
+        }
+        scratch_free_ptr = result;
+        GC_scratch_end_ptr = scratch_free_ptr + bytes_to_get;
+        return(GC_scratch_alloc(bytes));
+    }
+}
+
+static hdr * hdr_free_list = 0;
+
+/* Return an uninitialized header */
+static hdr * alloc_hdr()
+{
+    register hdr * result;
+    
+    if (hdr_free_list == 0) {
+        result = (hdr *) GC_scratch_alloc((word)(sizeof(hdr)));
+    } else {
+        result = hdr_free_list;
+        hdr_free_list = (hdr *) (result -> hb_next);
+    }
+    return(result);
+}
+
+static void free_hdr(hhdr)
+hdr * hhdr;
+{
+    hhdr -> hb_next = (struct hblk *) hdr_free_list;
+    hdr_free_list = hhdr;
+}
+ 
+void GC_init_headers()
+{
+    register int i;
+     
+    for (i = 0; i < TOP_SZ; i++) {
+        GC_top_index[i] = &GC_all_nils;
+    }
+}
+
+/* Make sure that there is a bottom level index block for address addr  */
+/* Return FALSE on failure.						*/
+static bool get_index(addr)
+register word addr;
+{
+    register word hi =
+    		(word)(addr) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE);
+    register bottom_index * r;
+    register bottom_index * p;
+    register bottom_index ** prev;
+#   ifdef HASH_TL
+      register i = TL_HASH(hi);
+      register bottom_index * old;
+      
+      old = p = GC_top_index[i];
+      while(p != &GC_all_nils) {
+          if (p -> key == hi) return(TRUE);
+          p = p -> hash_link;
+      }
+      r = (bottom_index*)GC_scratch_alloc((word)(sizeof (bottom_index)));
+      if (r == 0) return(FALSE);
+      BZERO(r, sizeof (bottom_index));
+      r -> hash_link = old;
+      GC_top_index[i] = r;
+#   else
+      if (GC_top_index[hi] != &GC_all_nils) return(TRUE);
+      r = (bottom_index*)GC_scratch_alloc((word)(sizeof (bottom_index)));
+      if (r == 0) return(FALSE);
+      GC_top_index[hi] = r;
+      BZERO(r, sizeof (bottom_index));
+# endif
+    r -> key = hi;
+    /* Add it to the list of bottom indices */
+      prev = &GC_all_bottom_indices;
+      while ((p = *prev) != 0 && p -> key < hi) prev = &(p -> asc_link);
+      r -> asc_link = p;
+      *prev = r;
+    return(TRUE);
+}
+
+/* Install a header for block h.  */
+/* The header is uninitialized.	  */
+/* Returns FALSE on failure.	  */
+bool GC_install_header(h)
+register struct hblk * h;
+{
+    hdr * result;
+    
+    if (!get_index((word) h)) return(FALSE);
+    result = alloc_hdr();
+    SET_HDR(h, result);
+    return(result != 0);
+}
+
+/* Set up forwarding counts for block h of size sz */
+bool GC_install_counts(h, sz)
+register struct hblk * h;
+register word sz; /* bytes */
+{
+    register struct hblk * hbp;
+    register int i;
+    
+    for (hbp = h; (char *)hbp < (char *)h + sz; hbp += BOTTOM_SZ) {
+        if (!get_index((word) hbp)) return(FALSE);
+    }
+    if (!get_index((word)h + sz - 1)) return(FALSE);
+    for (hbp = h + 1; (char *)hbp < (char *)h + sz; hbp += 1) {
+        i = HBLK_PTR_DIFF(hbp, h);
+        SET_HDR(hbp, (hdr *)(i > MAX_JUMP? MAX_JUMP : i));
+    }
+    return(TRUE);
+}
+
+/* Remove the header for block h */
+void GC_remove_header(h)
+register struct hblk * h;
+{
+    hdr ** ha;
+    
+    GET_HDR_ADDR(h, ha);
+    free_hdr(*ha);
+    *ha = 0;
+}
+
+/* Remove forwarding counts for h */
+void GC_remove_counts(h, sz)
+register struct hblk * h;
+register word sz; /* bytes */
+{
+    register struct hblk * hbp;
+    
+    for (hbp = h+1; (char *)hbp < (char *)h + sz; hbp += 1) {
+        SET_HDR(hbp, 0);
+    }
+}
+
+/* Apply fn to all allocated blocks */
+/*VARARGS1*/
+void GC_apply_to_all_blocks(fn, client_data)
+void (*fn)(/* struct hblk *h, word client_data */);
+word client_data;
+{
+    register int j;
+    register bottom_index * index_p;
+    
+    for (index_p = GC_all_bottom_indices; index_p != 0;
+         index_p = index_p -> asc_link) {
+        for (j = BOTTOM_SZ-1; j >= 0;) {
+            if (!IS_FORWARDING_ADDR_OR_NIL(index_p->index[j])) {
+                if (index_p->index[j]->hb_map != GC_invalid_map) {
+                    (*fn)(((struct hblk *)
+                  	      (((index_p->key << LOG_BOTTOM_SZ) + (word)j)
+                  	       << LOG_HBLKSIZE)),
+                          client_data);
+                }
+                j--;
+             } else if (index_p->index[j] == 0) {
+                j--;
+             } else {
+                j -= (int)(index_p->index[j]);
+             }
+         }
+     }
+}
+
+/* Get the next valid block whose address is at least h	*/
+/* Return 0 if there is none.				*/
+struct hblk * GC_next_block(h)
+struct hblk * h;
+{
+    register bottom_index * bi;
+    register word j = ((word)h >> LOG_HBLKSIZE) & (BOTTOM_SZ-1);
+    
+    GET_BI(h, bi);
+    if (bi == &GC_all_nils) {
+        register word hi = (word)h >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE);
+        bi = GC_all_bottom_indices;
+        while (bi != 0 && bi -> key < hi) bi = bi -> asc_link;
+        j = 0;
+    }
+    while(bi != 0) {
+        while (j < BOTTOM_SZ) {
+            if (IS_FORWARDING_ADDR_OR_NIL(bi -> index[j])) {
+                j++;
+            } else {
+                if (bi->index[j]->hb_map != GC_invalid_map) {
+                    return((struct hblk *)
+                  	      (((bi -> key << LOG_BOTTOM_SZ) + j)
+                  	       << LOG_HBLKSIZE));
+                } else {
+                    j += divHBLKSZ(bi->index[j] -> hb_sz);
+                }
+            }
+        }
+        j = 0;
+        bi = bi -> asc_link;
+    }
+    return(0);
+}
diff --git a/if_mach.c b/if_mach.c
new file mode 100644
index 00000000..7359a4f7
--- /dev/null
+++ b/if_mach.c
@@ -0,0 +1,22 @@
+/* Conditionally execute a command based on machine and OS from config.h */
+# include "config.h"
+# include <stdio.h>
+
+int main(argc, argv, envp)
+int argc;
+char ** argv;
+char ** envp;
+{
+    if (argc < 4) goto Usage;
+    if (strcmp(MACH_TYPE, argv[1]) != 0) return(0);
+    if (strcmp(OS_TYPE, "") != 0 && strcmp(argv[2], "") != 0
+        && strcmp(OS_TYPE, argv[2]) != 0) return(0);
+    execvp(argv[3], argv+3);
+    
+Usage:
+    fprintf(stderr, "Usage: %s mach_type os_type command\n", argv[0]);
+    fprintf(stderr, "Currently mach_type = %s, os_type = %s\n",
+    	    MACH_TYPE, OS_TYPE);
+    return(1);
+}
+
diff --git a/if_not_there.c b/if_not_there.c
new file mode 100644
index 00000000..806eed62
--- /dev/null
+++ b/if_not_there.c
@@ -0,0 +1,24 @@
+/* Conditionally execute a command based if the file argv[1] doesn't exist */
+/* Except for execvp, we stick to ANSI C.				   */
+# include "config.h"
+# include <stdio.h>
+
+int main(argc, argv, envp)
+int argc;
+char ** argv;
+char ** envp;
+{
+    FILE * f;
+    if (argc < 3) goto Usage;
+    if ((f = fopen(argv[1], "rb")) != 0
+        || (f = fopen(argv[1], "r")) != 0) {
+        fclose(f);
+        return(0);
+    }
+    execvp(argv[2], argv+2);
+    
+Usage:
+    fprintf(stderr, "Usage: %s file_name command\n", argv[0]);
+    return(1);
+}
+
diff --git a/include/gc.h b/include/gc.h
new file mode 100644
index 00000000..8c3560dd
--- /dev/null
+++ b/include/gc.h
@@ -0,0 +1,379 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to copy this garbage collector for any purpose,
+ * provided the above notices are retained on all copies.
+ */
+ 
+#ifndef GC_H
+
+# define GC_H
+
+# include <stddef.h>
+
+/* Define word and signed_word to be unsigned and signed types of the 	*/
+/* size as char * or void *.  There seems to be no way to do this	*/
+/* even semi-portably.  The following is probably no better/worse 	*/
+/* than almost anything else.						*/
+/* The ANSI standard suggests that size_t and ptr_diff_t might be 	*/
+/* better choices.  But those appear to have incorrect definitions	*/
+/* on may systems.  Notably "typedef int size_t" seems to be both	*/
+/* frequent and WRONG.							*/
+typedef unsigned long GC_word;
+typedef long GC_signed_word;
+
+/* Public read-only variables */
+
+extern GC_word GC_heapsize;       /* Heap size in bytes */
+
+extern GC_word GC_gc_no;/* Counter incremented per collection.  	*/
+			/* Includes empty GCs at startup.		*/
+			
+extern int GC_incremental;  /* Using incremental/generational collection. */
+
+
+/* Public R/W variables */
+
+extern int GC_quiet;	/* Disable statistics output.  Only matters if	*/
+			/* collector has been compiled with statistics	*/
+			/* enabled.  This involves a performance cost,	*/
+			/* and is thus not the default.			*/
+
+extern int GC_dont_gc;	/* Dont collect unless explicitly requested, e.g. */
+			/* beacuse it's not safe.			  */
+
+extern int GC_dont_expand;
+			/* Dont expand heap unless explicitly requested */
+			/* or forced to.				*/
+
+extern int GC_full_freq;    /* Number of partial collections between	*/
+			    /* full collections.  Matters only if	*/
+			    /* GC_incremental is set.			*/
+			
+extern GC_word GC_non_gc_bytes;
+			/* Bytes not considered candidates for collection. */
+			/* Used only to control scheduling of collections. */
+
+extern GC_word GC_free_space_divisor;
+			/* We try to make sure that we allocate at 	*/
+			/* least N/GC_free_space_divisor bytes between	*/
+			/* collections, where N is the heap size plus	*/
+			/* a rough estimate of the root set size.	*/
+			/* Initially, GC_free_space_divisor = 4.	*/
+			/* Increasing its value will use less space	*/
+			/* but more collection time.  Decreasing it	*/
+			/* will appreciably decrease collection time	*/
+			/* at the expens of space.			*/
+			/* GC_free_space_divisor = 1 will effectively	*/
+			/* disable collections.				*/
+			
+/* Public procedures */
+/*
+ * general purpose allocation routines, with roughly malloc calling conv.
+ * The atomic versions promise that no relevant pointers are contained
+ * in the object.  The nonatomic versions guarantee that the new object
+ * is cleared.  GC_malloc_stubborn promises that no changes to the object
+ * will occur after GC_end_stubborn_change has been called on the
+ * result of GC_malloc_stubborn. GC_malloc_uncollectable allocates an object
+ * that is scanned for pointers to collectable objects, but is not itself
+ * collectable.  GC_malloc_uncollectable and GC_free called on the resulting
+ * object implicitly update GC_non_gc_bytes appropriately.
+ */
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_malloc(size_t size_in_bytes);
+  extern void * GC_malloc_atomic(size_t size_in_bytes);
+  extern void * GC_malloc_uncollectable(size_t size_in_bytes);
+  extern void * GC_malloc_stubborn(size_t size_in_bytes);
+# else
+  extern char * GC_malloc(/* size_in_bytes */);
+  extern char * GC_malloc_atomic(/* size_in_bytes */);
+  extern char * GC_malloc_uncollectable(/* size_in_bytes */);
+  extern char * GC_malloc_stubborn(/* size_in_bytes */);
+# endif
+
+/* Explicitly deallocate an object.  Dangerous if used incorrectly.     */
+/* Requires a pointer to the base of an object.				*/
+/* If the argument is stubborn, it should not be changeable when freed. */
+/* An object should not be enable for finalization when it is 		*/
+/* explicitly deallocated.						*/
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void GC_free(void * object_addr);
+# else
+  extern void GC_free(/* object_addr */);
+# endif
+
+/*
+ * Stubborn objects may be changed only if the collector is explicitly informed.
+ * The collector is implicitly informed of coming change when such
+ * an object is first allocated.  The following routines inform the
+ * collector that an object will no longer be changed, or that it will
+ * once again be changed.  Only nonNIL pointer stores into the object
+ * are considered to be changes.  The argument to GC_end_stubborn_change
+ * must be exacly the value returned by GC_malloc_stubborn or passed to
+ * GC_change_stubborn.  (In the second case it may be an interior pointer
+ * within 512 bytes of the beginning of the objects.)
+ * There is a performance penalty for allowing more than
+ * one stubborn object to be changed at once, but it is acceptable to
+ * do so.  The same applies to dropping stubborn objects that are still
+ * changeable.
+ */
+void GC_change_stubborn(/* p */);
+void GC_end_stubborn_change(/* p */);
+
+/* Return a pointer to the base (lowest address) of an object given	*/
+/* a pointer to a location within the object.				*/
+/* Return 0 if displaced_pointer doesn't point to within a valid	*/
+/* object.								*/
+# if defined(__STDC__) || defined(__cplusplus)
+  void * GC_base(void * displaced_pointer);
+# else
+  char * GC_base(/* char * displaced_pointer */);
+# endif
+
+/* Given a pointer to the base of an object, return its size in bytes.	*/
+/* The returned size may be slightly larger than what was originally	*/
+/* requested.								*/
+# if defined(__STDC__) || defined(__cplusplus)
+  size_t GC_size(void * object_addr);
+# else
+  size_t GC_size(/* char * object_addr */);
+# endif
+
+/* For compatibility with C library.  This is occasionally faster than	*/
+/* a malloc followed by a bcopy.  But if you rely on that, either here	*/
+/* or with the standard C library, your code is broken.  In my		*/
+/* opinion, it shouldn't have been invented, but now we're stuck. -HB	*/
+/* The resulting object has the same kind as the original.		*/
+/* If the argument is stubborn, the result will have changes enabled.	*/
+/* It is an error to have changes enabled for the original object.	*/
+# if defined(__STDC__) || defined(__cplusplus)
+    extern void * GC_realloc(void * old_object, size_t new_size_in_bytes);
+# else
+    extern char * GC_realloc(/* old_object, new_size_in_bytes */);
+# endif
+
+
+/* Explicitly increase the heap size.	*/
+/* Returns 0 on failure, 1 on success.  */
+extern int GC_expand_hp(/* number_of_4K_blocks */);
+
+/* Clear the set of root segments */
+extern void GC_clear_roots();
+
+/* Add a root segment */
+extern void GC_add_roots(/* low_address, high_address_plus_1 */);
+
+/* Add a displacement to the set of those considered valid by the	*/
+/* collector.  GC_register_displacement(n) means that if p was returned */
+/* by GC_malloc, then (char *)p + n will be considered to be a valid	*/
+/* pointer to n.  N must be small and less than the size of p.		*/
+/* (All pointers to the interior of objects from the stack are		*/
+/* considered valid in any case.  This applies to heap objects and	*/
+/* static data.)							*/
+/* Preferably, this should be called before any other GC procedures.	*/
+/* Calling it later adds to the probability of excess memory		*/
+/* retention.								*/
+void GC_register_displacement(/* n */);
+
+/* Explicitly trigger a collection. 	*/
+void GC_gcollect();
+
+/* Enable incremental/generational collection.	*/
+/* Not advisable unless dirty bits are 		*/
+/* available or most heap objects are		*/
+/* pointerfree(atomic) or immutable.		*/
+/* Don't use in leak finding mode.		*/
+void GC_enable_incremental();
+
+/* Debugging (annotated) allocation.  GC_gcollect will check 		*/
+/* objects allocated in this way for overwrites, etc.			*/
+# if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_debug_malloc(size_t size_in_bytes,
+  				char * descr_string, int descr_int);
+  extern void * GC_debug_malloc_atomic(size_t size_in_bytes,
+  				       char * descr_string, int descr_int);
+  extern void * GC_debug_malloc_uncollectable(size_t size_in_bytes,
+  				           char * descr_string, int descr_int);
+  extern void * GC_debug_malloc_stubborn(size_t size_in_bytes,
+  				         char * descr_string, int descr_int);
+  extern void GC_debug_free(void * object_addr);
+  extern void * GC_debug_realloc(void * old_object,
+  			 	 size_t new_size_in_bytes,
+  			 	 char * descr_string, int descr_int);
+# else
+  extern char * GC_debug_malloc(/* size_in_bytes, descr_string, descr_int */);
+  extern char * GC_debug_malloc_atomic(/* size_in_bytes, descr_string,
+  					  descr_int */);
+  extern char * GC_debug_malloc_uncollectable(/* size_in_bytes, descr_string,
+  					  descr_int */);
+  extern char * GC_debug_malloc_stubborn(/* size_in_bytes, descr_string,
+  					  descr_int */);
+  extern void GC_debug_free(/* object_addr */);
+  extern char * GC_debug_realloc(/* old_object, new_size_in_bytes,
+  			            descr_string, descr_int */);
+# endif
+void GC_debug_change_stubborn(/* p */);
+void GC_debug_end_stubborn_change(/* p */);
+# ifdef GC_DEBUG
+#   define GC_MALLOC(sz) GC_debug_malloc(sz, __FILE__, __LINE__)
+#   define GC_MALLOC_ATOMIC(sz) GC_debug_malloc_atomic(sz, __FILE__, __LINE__)
+#   define GC_MALLOC_UNCOLLECTABLE(sz) GC_debug_malloc_uncollectable(sz, \
+							__FILE__, __LINE__)
+#   define GC_REALLOC(old, sz) GC_debug_realloc(old, sz, __FILE__, \
+							       __LINE__)
+#   define GC_FREE(p) GC_debug_free(p)
+#   define GC_REGISTER_FINALIZER(p, f, d, of, od) \
+	GC_register_finalizer(GC_base(p), GC_debug_invoke_finalizer, \
+			      GC_make_closure(f,d), of, od)
+#   define GC_MALLOC_STUBBORN(sz) GC_debug_malloc_stubborn(sz, __FILE__, \
+							       __LINE__)
+#   define GC_CHANGE_STUBBORN(p) GC_debug_change_stubborn(p)
+#   define GC_END_STUBBORN_CHANGE(p) GC_debug_end_stubborn_change(p)
+# else
+#   define GC_MALLOC(sz) GC_malloc(sz)
+#   define GC_MALLOC_ATOMIC(sz) GC_malloc_atomic(sz)
+#   define GC_MALLOC_UNCOLLECTABLE(sz) GC_malloc_uncollectable(sz)
+#   define GC_REALLOC(old, sz) GC_realloc(old, sz)
+#   define GC_FREE(p) GC_free(p)
+#   define GC_REGISTER_FINALIZER(p, f, d, of, od) \
+	GC_register_finalizer(p, f, d, of, od)
+#   define GC_MALLOC_STUBBORN(sz) GC_malloc_stubborn(sz)
+#   define GC_CHANGE_STUBBORN(p) GC_change_stubborn(p)
+#   define GC_END_STUBBORN_CHANGE(p) GC_end_stubborn_change(p)
+# endif
+/* The following are included because they are often convenient, and	*/
+/* reduce the chance for a misspecifed size argument.  But calls may	*/
+/* expand to something syntactically incorrect if t is a complicated	*/
+/* type expression.  							*/
+# define GC_NEW(t) (t *)GC_MALLOC(sizeof (t))
+# define GC_NEW_ATOMIC(t) (t *)GC_MALLOC_ATOMIC(sizeof (t))
+# define GC_NEW_STUBBORN(t) (t *)GC_MALLOC_STUBBORN(sizeof (t))
+# define GC_NEW_UNCOLLECTABLE(t) (t *)GC_NEW_UNCOLLECTABLE(sizeof (t))
+
+/* Finalization.  Some of these primitives are grossly unsafe.		*/
+/* The idea is to make them both cheap, and sufficient to build		*/
+/* a safer layer, closer to PCedar finalization.			*/
+/* The interface represents my conclusions from a long discussion	*/
+/* with Alan Demers, Dan Greene, Carl Hauser, Barry Hayes, 		*/
+/* Christian Jacobi, and Russ Atkinson.  It's not perfect, and		*/
+/* probably nobody else agrees with it.	    Hans-J. Boehm  3/13/92	*/
+# if defined(__STDC__) || defined(__cplusplus)
+  typedef void (*GC_finalization_proc)(void * obj, void * client_data);
+# else
+  typedef void (*GC_finalization_proc)(/* void * obj, void * client_data */);
+# endif
+	
+void GC_register_finalizer(/* void * obj,
+			      GC_finalization_proc fn, void * cd,
+			      GC_finalization_proc *ofn, void ** ocd */);
+	/* When obj is no longer accessible, invoke		*/
+	/* (*fn)(obj, cd).  If a and b are inaccessible, and	*/
+	/* a points to b (after disappearing links have been	*/
+	/* made to disappear), then only a will be		*/
+	/* finalized.  (If this does not create any new		*/
+	/* pointers to b, then b will be finalized after the	*/
+	/* next collection.)  Any finalizable object that	*/
+	/* is reachable from itself by following one or more	*/
+	/* pointers will not be finalized (or collected).	*/
+	/* Thus cycles involving finalizable objects should	*/
+	/* be avoided, or broken by disappearing links.		*/
+	/* fn is invoked with the allocation lock held.  It may */
+	/* not allocate.  (Any storage it might need		*/
+	/* should be preallocated and passed as part of cd.) 	*/
+	/* fn should terminate as quickly as possible, and	*/
+	/* defer extended computation.				*/
+	/* All but the last finalizer registered for an object  */
+	/* is ignored.						*/
+	/* Finalization may be removed by passing 0 as fn.	*/
+	/* The old finalizer and client data are stored in	*/
+	/* *ofn and *ocd.					*/ 
+	/* Fn is never invoked on an accessible object,		*/
+	/* provided hidden pointers are converted to real 	*/
+	/* pointers only if the allocation lock is held, and	*/
+	/* such conversions are not performed by finalization	*/
+	/* routines.						*/
+
+/* The following routine may be used to break cycles between	*/
+/* finalizable objects, thus causing cyclic finalizable		*/
+/* objects to be finalized in the correct order.  Standard	*/
+/* use involves calling GC_register_disappearing_link(&p),	*/
+/* where p is a pointer that is not followed by finalization	*/
+/* code, and should not be considered in determining 		*/
+/* finalization order.						*/ 
+int GC_register_disappearing_link(/* void ** link */);
+	/* Link should point to a field of a heap allocated 	*/
+	/* object obj.  *link will be cleared when obj is	*/
+	/* found to be inaccessible.  This happens BEFORE any	*/
+	/* finalization code is invoked, and BEFORE any		*/
+	/* decisions about finalization order are made.		*/
+	/* This is useful in telling the finalizer that 	*/
+	/* some pointers are not essential for proper		*/
+	/* finalization.  This may avoid finalization cycles.	*/
+	/* Note that obj may be resurrected by another		*/
+	/* finalizer, and thus the clearing of *link may	*/
+	/* be visible to non-finalization code.  		*/
+	/* There's an argument that an arbitrary action should  */
+	/* be allowed here, instead of just clearing a pointer. */
+	/* But this causes problems if that action alters, or 	*/
+	/* examines connectivity.				*/
+	/* Returns 1 if link was already registered, 0		*/
+	/* otherwise.						*/
+	/* Only exists for backward compatibility.  See below:	*/
+int GC_general_register_disappearing_link(/* void ** link, void * obj */);
+	/* A slight generalization of the above. *link is	*/
+	/* cleared when obj first becomes inaccessible.  This	*/
+	/* can be used to implement weak pointers easily and	*/
+	/* safely. Typically link will point to a location	*/
+	/* holding a disguised pointer to obj.  In this way	*/
+	/* soft pointers are broken before any object		*/
+	/* reachable from them are finalized.  Each link	*/
+	/* May be registered only once, i.e. with one obj	*/
+	/* value.  This was added after a long email discussion */
+	/* with John Ellis.					*/
+int GC_unregister_disappearing_link(/* void ** link */);
+	/* Returns 0 if link was not actually registered.	*/
+	/* Undoes a registration by either of the above two	*/
+	/* routines.						*/
+
+/* Auxiliary fns to make finalization work correctly with displaced	*/
+/* pointers introduced by the debugging allocators.			*/
+# if defined(__STDC__) || defined(__cplusplus)
+    void * GC_make_closure(GC_finalization_proc fn, void * data);
+    void GC_debug_invoke_finalizer(void * obj, void * data);
+# else
+    char * GC_make_closure(/* GC_finalization_proc fn, char * data */);
+    void GC_debug_invoke_finalizer(/* void * obj, void * data */);
+# endif
+
+	
+/* The following is intended to be used by a higher level	*/
+/* (e.g. cedar-like) finalization facility.  It is expected	*/
+/* that finalization code will arrange for hidden pointers to	*/
+/* disappear.  Otherwise objects can be accessed after they	*/
+/* have been collected.						*/
+# ifdef I_HIDE_POINTERS
+#   if defined(__STDC__) || defined(__cplusplus)
+#     define HIDE_POINTER(p) (~(size_t)(p))
+#     define REVEAL_POINTER(p) ((void *)(HIDE_POINTER(p)))
+#   else
+#     define HIDE_POINTER(p) (~(unsigned long)(p))
+#     define REVEAL_POINTER(p) ((char *)(HIDE_POINTER(p)))
+#   endif
+    /* Converting a hidden pointer to a real pointer requires verifying	*/
+    /* that the object still exists.  This involves acquiring the  	*/
+    /* allocator lock to avoid a race with the collector.		*/
+
+#   if defined(__STDC__) || defined(__cplusplus)
+        typedef void * (*GC_fn_type)();
+        void * GC_call_with_alloc_lock(GC_fn_type fn, void * client_data);
+#   else
+        typedef char * (*GC_fn_type)();
+        char * GC_call_with_alloc_lock(/* GC_fn_type fn, char * client_data */);
+#   endif
+# endif
+
+#endif
diff --git a/include/gc_typed.h b/include/gc_typed.h
new file mode 100644
index 00000000..401fd062
--- /dev/null
+++ b/include/gc_typed.h
@@ -0,0 +1,67 @@
+/*
+ * Some simple primitives for allocation with explicit type information.
+ * Facilities for dynamic type inference may be added later.
+ * Should be used only for extremely performance critical applications,
+ * or if conservative collector leakage is otherwise a problem (unlikely).
+ * Note that this is implemented completely separately from the rest
+ * of the collector, and is not linked in unless referenced.
+ */
+/* Boehm, March 31, 1994 4:43 pm PST */
+
+#ifndef _GC_TYPED_H
+# define _GC_TYPED_H
+# ifndef _GC_H
+#   include "gc.h"
+# endif
+
+typedef GC_word * GC_bitmap;
+	/* The least significant bit of the first word is one if	*/
+	/* the first word in the object may be a pointer.		*/
+	
+# define GC_get_bit(bm, index) \
+		(((bm)[divWORDSZ(index)] >> modWORDSZ(index)) & 1)
+# define GC_set_bit(bm, index) \
+		(bm)[divWORDSZ(index)] |= (word)1 << modWORDSZ(index)
+
+typedef GC_word GC_descr;
+
+#if defined(__STDC__) || defined(__cplusplus)
+  extern GC_descr GC_make_decriptor(GC_bitmap bm, size_t len);
+#else
+  extern GC_descr GC_make_decriptor(/* GC_bitmap bm, size_t len */);
+#endif
+		/* Return a type descriptor for the object whose layout	*/
+		/* is described by the argument.			*/
+		/* The least significant bit of the first word is one	*/
+		/* if the first word in the object may be a pointer.	*/
+		/* The second argument specifies the number of		*/
+		/* meaningful bits in the bitmap.  The actual object 	*/
+		/* may be larger (but not smaller).  Any additional	*/
+		/* words in the object are assumed not to contain 	*/
+		/* pointers.						*/
+		/* Returns (GC_descr)(-1) on failure (no memory).	*/
+
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_malloc_explicitly_typed(size_t size_in_bytes, GC_descr d);
+#else
+  extern char * GC_malloc_explicitly_typed(/* size_in_bytes, descriptor */);
+#endif
+		/* Allocate an object whose layout is described by d.	*/
+		/* The resulting object MAY NOT BE PASSED TO REALLOC.	*/
+		
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_calloc_explicitly_typed(size_t nelements,
+  					   size_t element_size_in_bytes,
+  					   GC_descr d);
+#else
+  char * GC_calloc_explicitly_typed(/* nelements, size_in_bytes, descriptor */);
+  	/* Allocate an array of nelements elements, each of the	*/
+  	/* given size, and with the given descriptor.		*/
+  	/* The elemnt size must be a multiple of the byte	*/
+  	/* alignment required for pointers.  E.g. on a 32-bit	*/
+  	/* machine with 16-bit aligned pointers, size_in_bytes	*/
+  	/* must be a multiple of 2.				*/
+#endif
+
+#endif /* _GC_TYPED_H */
+
diff --git a/mach_dep.c b/mach_dep.c
new file mode 100644
index 00000000..cd441f97
--- /dev/null
+++ b/mach_dep.c
@@ -0,0 +1,330 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 1:58 pm PDT */
+# include "gc_priv.h"
+# include <stdio.h>
+# include <setjmp.h>
+# if defined(OS2) || defined(CX_UX)
+#   define _setjmp(b) setjmp(b)
+#   define _longjmp(b,v) longjmp(b,v)
+# endif
+
+
+/* Routine to mark from registers that are preserved by the C compiler. */
+/* This must be ported to every new architecture.  There is a generic   */
+/* version at the end, that is likely, but not guaranteed to work       */
+/* on your architecture.  Run the test_setjmp program to see whether    */
+/* there is any chance it will work.                                    */
+
+#ifdef AMIGA
+__asm GC_push_regs(
+	  register __a2 word a2,
+	  register __a3 word a3,
+	  register __a4 word a4,
+	  register __a5 word a5,
+	  register __a6 word a6,
+	  register __d2 const word d2,
+	  register __d3 const word d3,
+	  register __d4 const word d4,
+	  register __d5 const word d5,
+	  register __d6 const word d6,
+	  register __d7 const word d7)
+#else
+  void GC_push_regs()
+#endif
+{
+#       ifdef RT
+	  register long TMP_SP; /* must be bound to r11 */
+#       endif
+#       ifdef VAX
+	  /* VAX - generic code below does not work under 4.2 */
+	  /* r1 through r5 are caller save, and therefore     */
+	  /* on the stack or dead.                            */
+	  asm("pushl r11");     asm("calls $1,_GC_push_one");
+	  asm("pushl r10"); 	asm("calls $1,_GC_push_one");
+	  asm("pushl r9");	asm("calls $1,_GC_push_one");
+	  asm("pushl r8");	asm("calls $1,_GC_push_one");
+	  asm("pushl r7");	asm("calls $1,_GC_push_one");
+	  asm("pushl r6");	asm("calls $1,_GC_push_one");
+#       endif
+#       if defined(M68K) && (defined(SUNOS4) || defined(NEXT))
+	/*  M68K SUNOS - could be replaced by generic code */
+	  /* a0, a1 and d1 are caller save          */
+	  /*  and therefore are on stack or dead.   */
+	
+	  asm("subqw #0x4,sp");		/* allocate word on top of stack */
+
+	  asm("movl a2,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl a3,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl a4,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl a5,sp@");	asm("jbsr _GC_push_one");
+	  /* Skip frame pointer and stack pointer */
+	  asm("movl d1,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl d2,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl d3,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl d4,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl d5,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl d6,sp@");	asm("jbsr _GC_push_one");
+	  asm("movl d7,sp@");	asm("jbsr _GC_push_one");
+
+	  asm("addqw #0x4,sp");		/* put stack back where it was	*/
+#       endif
+
+#       if defined(M68K) && defined(HP)
+	/*  M68K HP - could be replaced by generic code */
+	  /* a0, a1 and d1 are caller save.  */
+	
+	  asm("subq.w &0x4,%sp");	/* allocate word on top of stack */
+
+	  asm("mov.l %a2,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %a3,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %a4,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %a5,(%sp)"); asm("jsr _GC_push_one");
+	  /* Skip frame pointer and stack pointer */
+	  asm("mov.l %d1,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %d2,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %d3,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %d4,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %d5,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %d6,(%sp)"); asm("jsr _GC_push_one");
+	  asm("mov.l %d7,(%sp)"); asm("jsr _GC_push_one");
+
+	  asm("addq.w &0x4,%sp");	/* put stack back where it was	*/
+#       endif /* M68K HP */
+
+#       ifdef AMIGA
+	/*  AMIGA - could be replaced by generic code 			*/
+	/*	  SAS/C optimizer mangles this so compile with "noopt"	*/
+	  /* a0, a1, d0 and d1 are caller save */
+	  GC_push_one(a2);
+	  GC_push_one(a3);
+	  GC_push_one(a4);
+	  GC_push_one(a5);
+	  GC_push_one(a6);
+	  /* Skip stack pointer */
+	  GC_push_one(d2);
+	  GC_push_one(d3);
+	  GC_push_one(d4);
+	  GC_push_one(d5);
+	  GC_push_one(d6);
+	  GC_push_one(d7);
+#       endif
+
+#       if defined(I386) &&!defined(OS2) &&!defined(SUNOS5) &&!defined(MSWIN32)
+	/* I386 code, generic code does not appear to work */
+	/* It does appear to work under OS2, and asms dont */
+	  asm("pushl %eax");  asm("call _GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %ecx");  asm("call _GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %edx");  asm("call _GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %esi");  asm("call _GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %edi");  asm("call _GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %ebx");  asm("call _GC_push_one"); asm("addl $4,%esp");
+#       endif
+
+#       if defined(I386) && defined(MSWIN32)
+	/* I386 code, Microsoft variant		*/
+	  __asm  push eax
+	  __asm  call GC_push_one
+	  __asm  add esp,4
+	  __asm  push ecx
+	  __asm  call GC_push_one
+	  __asm  add esp,4
+	  __asm  push edx
+	  __asm  call GC_push_one
+	  __asm  add esp,4
+	  __asm  push esi
+	  __asm  call GC_push_one
+	  __asm  add esp,4
+	  __asm  push edi
+	  __asm  call GC_push_one
+	  __asm  add esp,4
+	  __asm  push ebx
+	  __asm  call GC_push_one
+	  __asm  add esp,4
+#       endif
+
+#       if defined(I386) && defined(SUNOS5)
+	/* I386 code, SVR4 variant, generic code does not appear to work */
+	  asm("pushl %eax");  asm("call GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %ecx");  asm("call GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %edx");  asm("call GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %esi");  asm("call GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %edi");  asm("call GC_push_one"); asm("addl $4,%esp");
+	  asm("pushl %ebx");  asm("call GC_push_one"); asm("addl $4,%esp");
+#       endif
+
+#       ifdef NS32K
+	  asm ("movd r3, tos"); asm ("bsr ?_GC_push_one"); asm ("adjspb $-4");
+	  asm ("movd r4, tos"); asm ("bsr ?_GC_push_one"); asm ("adjspb $-4");
+	  asm ("movd r5, tos"); asm ("bsr ?_GC_push_one"); asm ("adjspb $-4");
+	  asm ("movd r6, tos"); asm ("bsr ?_GC_push_one"); asm ("adjspb $-4");
+	  asm ("movd r7, tos"); asm ("bsr ?_GC_push_one"); asm ("adjspb $-4");
+#       endif
+
+#       ifdef SPARC
+	  {
+	      word GC_save_regs_in_stack();
+	      
+	      /* generic code will not work */
+	      (void)GC_save_regs_in_stack();
+	  }
+#       endif
+
+#	ifdef RT
+	    GC_push_one(TMP_SP);    /* GC_push_one from r11 */
+
+	    asm("cas r11, r6, r0"); GC_push_one(TMP_SP);	/* r6 */
+	    asm("cas r11, r7, r0"); GC_push_one(TMP_SP);	/* through */
+	    asm("cas r11, r8, r0"); GC_push_one(TMP_SP);	/* r10 */
+	    asm("cas r11, r9, r0"); GC_push_one(TMP_SP);
+	    asm("cas r11, r10, r0"); GC_push_one(TMP_SP);
+
+	    asm("cas r11, r12, r0"); GC_push_one(TMP_SP); /* r12 */
+	    asm("cas r11, r13, r0"); GC_push_one(TMP_SP); /* through */
+	    asm("cas r11, r14, r0"); GC_push_one(TMP_SP); /* r15 */
+	    asm("cas r11, r15, r0"); GC_push_one(TMP_SP);
+#       endif
+
+#       if defined(M68K) && defined(SYSV)
+  	/*  Once again similar to SUN and HP, though setjmp appears to work.
+  		--Parag
+  	 */
+#        ifdef __GNUC__
+  	  asm("subqw #0x4,%sp");	/* allocate word on top of stack */
+  
+  	  asm("movl %a2,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %a3,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %a4,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %a5,%sp@");	asm("jbsr GC_push_one");
+  	  /* Skip frame pointer and stack pointer */
+  	  asm("movl %d1,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %d2,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %d3,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %d4,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %d5,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %d6,%sp@");	asm("jbsr GC_push_one");
+  	  asm("movl %d7,%sp@");	asm("jbsr GC_push_one");
+  
+  	  asm("addqw #0x4,%sp");	/* put stack back where it was	*/
+#        else /* !__GNUC__*/
+  	  asm("subq.w &0x4,%sp");	/* allocate word on top of stack */
+  
+  	  asm("mov.l %a2,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %a3,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %a4,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %a5,(%sp)"); asm("jsr GC_push_one");
+  	  /* Skip frame pointer and stack pointer */
+  	  asm("mov.l %d1,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %d2,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %d3,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %d4,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %d5,(%sp)"); asm("jsr GC_push_one");
+   	  asm("mov.l %d6,(%sp)"); asm("jsr GC_push_one");
+  	  asm("mov.l %d7,(%sp)"); asm("jsr GC_push_one");
+  
+  	  asm("addq.w &0x4,%sp");	/* put stack back where it was	*/
+#        endif /* !__GNUC__ */
+#       endif /* M68K/SYSV */
+
+
+#     if defined(HP_PA) || defined(M88K) || (defined(I386) && defined(OS2))
+	/* Generic code                          */
+	/* The idea is due to Parag Patel at HP. */
+	/* We're not sure whether he would like  */
+	/* to be he acknowledged for it or not.  */
+	{
+	    static jmp_buf regs;
+	    register word * i = (word *) regs;
+	    register ptr_t lim = (ptr_t)(regs) + (sizeof regs);
+
+	    /* Setjmp on Sun 3s doesn't clear all of the buffer.  */
+	    /* That tends to preserve garbage.  Clear it.         */
+		for (; (char *)i < lim; i++) {
+		    *i = 0;
+		}
+	    (void) _setjmp(regs);
+	    GC_push_all_stack((ptr_t)regs, lim);
+	}
+#     endif
+
+      /* other machines... */
+#       if !(defined M68K) && !(defined VAX) && !(defined RT) 
+#	if !(defined SPARC) && !(defined I386) && !(defined NS32K)
+#	if !defined(HP_PA) && !defined(M88K)
+	    --> bad news <--
+#       endif
+#       endif
+#       endif
+}
+
+/* On register window machines, we need a way to force registers into 	*/
+/* the stack.	Return sp.						*/
+# ifdef SPARC
+    asm("	.seg 	\"text\"");
+#   ifdef SUNOS5
+      asm("	.globl	GC_save_regs_in_stack");
+      asm("GC_save_regs_in_stack:");
+#   else
+      asm("	.globl	_GC_save_regs_in_stack");
+      asm("_GC_save_regs_in_stack:");
+#   endif
+    asm("	ta	0x3   ! ST_FLUSH_WINDOWS");
+    asm("	mov	%sp,%o0");
+    asm("	retl");
+    asm("	nop");
+    
+#   ifdef LINT
+	word GC_save_regs_in_stack() { return(0 /* sp really */);}
+#   endif
+# endif
+
+
+/* GC_clear_stack_inner(arg, limit) clears stack area up to limit and	*/
+/* returns arg.  Stack clearing is crucial on SPARC, so we supply	*/
+/* an assembly version that's more careful.  Assumes limit is hotter	*/
+/* than sp, and limit is 8 byte aligned.				*/
+#if defined(ASM_CLEAR_CODE) && !defined(THREADS)
+#ifndef SPARC
+	--> fix it
+#endif
+# ifdef SUNOS4
+    asm(".globl _GC_clear_stack_inner");
+    asm("_GC_clear_stack_inner:");
+# else
+    asm(".globl GC_clear_stack_inner");
+    asm("GC_clear_stack_inner:");
+# endif
+  asm("mov %sp,%o2");		/* Save sp	*/
+  asm("add %sp,-8,%o3");	/* p = sp-8	*/
+  asm("clr %g1");		/* [g0,g1] = 0	*/
+  asm("add %o1,-0x60,%sp");	/* Move sp out of the way,	*/
+  				/* so that traps still work.	*/
+  				/* Includes some extra words	*/
+  				/* so we can be sloppy below.	*/
+  asm("loop:");
+  asm("std %g0,[%o3]");		/* *(long long *)p = 0	*/
+  asm("cmp %o3,%o1");
+  asm("bgu loop	");		/* if (p > limit) goto loop	*/
+    asm("add %o3,-8,%o3");	/* p -= 8 (delay slot) */
+  asm("retl");
+    asm("mov %o2,%sp");		/* Restore sp., delay slot	*/
+  /* First argument = %o0 = return value */
+  
+# ifdef LINT
+    /*ARGSUSED*/
+    ptr_t GC_clear_stack_inner(arg, limit)
+    ptr_t arg; word limit;
+    { return(arg); }
+# endif
+#endif  
diff --git a/malloc.c b/malloc.c
new file mode 100644
index 00000000..770826eb
--- /dev/null
+++ b/malloc.c
@@ -0,0 +1,541 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:03 pm PDT */
+ 
+#include <stdio.h>
+#include "gc_priv.h"
+
+extern ptr_t GC_clear_stack();	/* in misc.c, behaves like identity */
+void GC_extend_size_map();	/* in misc.c. */
+
+/* allocate lb bytes for an object of kind.	*/
+/* Should not be used to directly to allocate	*/
+/* objects such as STUBBORN objects that	*/
+/* require special handling on allocation.	*/
+/* First a version that assumes we already	*/
+/* hold lock:					*/
+ptr_t GC_generic_malloc_inner(lb, k)
+register word lb;
+register int k;
+{
+register word lw;
+register ptr_t op;
+register ptr_t *opp;
+
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+	  if (lw == 0) lw = 1;
+#       endif
+	opp = &(GC_obj_kinds[k].ok_freelist[lw]);
+        if( (op = *opp) == 0 ) {
+#	    ifdef MERGE_SIZES
+	      if (GC_size_map[lb] == 0) {
+	        if (!GC_is_initialized)  GC_init_inner();
+	        if (GC_size_map[lb] == 0) GC_extend_size_map(lb);
+	        return(GC_generic_malloc_inner(lb, k));
+	      }
+#	    else
+	      if (!GC_is_initialized) {
+	        GC_init_inner();
+	        return(GC_generic_malloc_inner(lb, k));
+	      }
+#	    endif
+	    op = GC_allocobj(lw, k);
+	    if (op == 0) goto out;
+        }
+        /* Here everything is in a consistent state.	*/
+        /* We assume the following assignment is	*/
+        /* atomic.  If we get aborted			*/
+        /* after the assignment, we lose an object,	*/
+        /* but that's benign.				*/
+        /* Volatile declarations may need to be added	*/
+        /* to prevent the compiler from breaking things.*/
+        *opp = obj_link(op);
+        obj_link(op) = 0;
+    } else {
+	register struct hblk * h;
+	register word n_blocks = divHBLKSZ(ADD_SLOP(lb)
+					   + HDR_BYTES + HBLKSIZE-1);
+	
+	if (!GC_is_initialized) GC_init_inner();
+	/* Do our share of marking work */
+          if(GC_incremental && !GC_dont_gc) GC_collect_a_little((int)n_blocks);
+	lw = ROUNDED_UP_WORDS(lb);
+	while ((h = GC_allochblk(lw, k, 0)) == 0
+		&& GC_collect_or_expand(n_blocks));
+	if (h == 0) {
+	    op = 0;
+	} else {
+	    op = (ptr_t) (h -> hb_body);
+	    GC_words_wasted += BYTES_TO_WORDS(n_blocks * HBLKSIZE) - lw;
+	}
+    }
+    GC_words_allocd += lw;
+    
+out:
+    return((ptr_t)op);
+}
+
+/* Allocate a composite object of size n bytes.  The caller guarantees	*/
+/* that pointers past the first page are not relevant.  Caller holds	*/
+/* allocation lock.							*/
+ptr_t GC_malloc_ignore_off_page_inner(lb)
+register size_t lb;
+{
+# ifdef ALL_INTERIOR_POINTERS
+    register struct hblk * h;
+    register word n_blocks;
+    register word lw;
+    register ptr_t op;
+
+    if (lb <= HBLKSIZE)
+        return(GC_generic_malloc_inner((word)lb, NORMAL));
+    n_blocks = divHBLKSZ(ADD_SLOP(lb) + HDR_BYTES + HBLKSIZE-1);
+    if (!GC_is_initialized) GC_init_inner();
+    /* Do our share of marking work */
+    if(GC_incremental && !GC_dont_gc) GC_collect_a_little((int)n_blocks);
+    lw = ROUNDED_UP_WORDS(lb);
+    while ((h = GC_allochblk(lw, NORMAL, IGNORE_OFF_PAGE)) == 0
+	   && GC_collect_or_expand(n_blocks));
+    if (h == 0) {
+	op = 0;
+    } else {
+	op = (ptr_t) (h -> hb_body);
+	GC_words_wasted += BYTES_TO_WORDS(n_blocks * HBLKSIZE) - lw;
+    }
+    GC_words_allocd += lw;
+    return((ptr_t)op);
+# else
+    return(GC_generic_malloc_inner((word)lb, NORMAL));
+# endif
+}
+
+# if defined(__STDC__) || defined(__cplusplus)
+  void * GC_malloc_ignore_off_page(size_t lb)
+# else
+  char * GC_malloc_ignore_off_page(lb)
+  register size_t lb;
+# endif
+{
+    register extern_ptr_t result;
+    DCL_LOCK_STATE;
+    
+    GC_invoke_finalizers();
+    DISABLE_SIGNALS();
+    LOCK();
+    result = GC_malloc_ignore_off_page_inner(lb);
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return(result);
+}
+
+ptr_t GC_generic_malloc(lb, k)
+register word lb;
+register int k;
+{
+    ptr_t result;
+    DCL_LOCK_STATE;
+
+    GC_invoke_finalizers();
+    DISABLE_SIGNALS();
+    LOCK();
+    result = GC_generic_malloc_inner(lb, k);
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return(result);
+}   
+
+
+/* Analogous to the above, but assumes a small object size, and 	*/
+/* bypasses MERGE_SIZES mechanism.  Used by gc_inline.h.		*/
+ptr_t GC_generic_malloc_words_small(lw, k)
+register word lw;
+register int k;
+{
+register ptr_t op;
+register ptr_t *opp;
+DCL_LOCK_STATE;
+
+    GC_invoke_finalizers();
+    DISABLE_SIGNALS();
+    LOCK();
+    opp = &(GC_obj_kinds[k].ok_freelist[lw]);
+    if( (op = *opp) == 0 ) {
+        if (!GC_is_initialized) {
+            GC_init_inner();
+        }
+	op = GC_clear_stack(GC_allocobj(lw, k));
+	if (op == 0) goto out;
+    }
+    *opp = obj_link(op);
+    obj_link(op) = 0;
+    GC_words_allocd += lw;
+    
+out:
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return((ptr_t)op);
+}
+
+#if defined(THREADS) && !defined(SRC_M3)
+/* Return a list of 1 or more objects of the indicated size, linked	*/
+/* through the first word in the object.  This has the advantage that	*/
+/* it acquires the allocation lock only once, and may greatly reduce	*/
+/* time wasted contending for the allocation lock.  Typical usage would */
+/* be in a thread that requires many items of the same size.  It would	*/
+/* keep its own free list in thread-local storage, and call		*/
+/* GC_malloc_many or friends to replenish it.  (We do not round up	*/
+/* object sizes, since a call indicates the intention to consume many	*/
+/* objects of exactly this size.)					*/
+/* Note that the client should usually clear the link field.		*/
+ptr_t GC_generic_malloc_many(lb, k)
+register word lb;
+register int k;
+{
+ptr_t op;
+register ptr_t p;
+ptr_t *opp;
+word lw;
+register word my_words_allocd;
+DCL_LOCK_STATE;
+
+    if (!SMALL_OBJ(lb)) {
+        op = GC_generic_malloc(lb, k);
+        obj_link(op) = 0;
+        return(op);
+    }
+    lw = ROUNDED_UP_WORDS(lb);
+    GC_invoke_finalizers();
+    DISABLE_SIGNALS();
+    LOCK();
+    opp = &(GC_obj_kinds[k].ok_freelist[lw]);
+    if( (op = *opp) == 0 ) {
+        if (!GC_is_initialized) {
+            GC_init_inner();
+        }
+	op = GC_clear_stack(GC_allocobj(lw, k));
+	if (op == 0) goto out;
+    }
+    *opp = 0;
+    my_words_allocd = 0;
+    for (p = op; p != 0; p = obj_link(p)) {
+        my_words_allocd += lw;
+        if (my_words_allocd >= BODY_SZ) {
+            *opp = obj_link(p);
+            obj_link(p) = 0;
+            break;
+        }
+    }
+    GC_words_allocd += my_words_allocd;
+    
+out:
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return(op);
+
+}
+
+void * GC_malloc_many(size_t lb)
+{
+    return(GC_generic_malloc_many(lb, NORMAL));
+}
+
+/* Note that the "atomic" version of this would be unsafe, since the	*/
+/* links would not be seen by the collector.				*/
+# endif
+
+#define GENERAL_MALLOC(lb,k) \
+    (extern_ptr_t)GC_clear_stack(GC_generic_malloc((word)lb, k))
+/* We make the GC_clear_stack_call a tail call, hoping to get more of	*/
+/* the stack.								*/
+
+/* Allocate lb bytes of atomic (pointerfree) data */
+# ifdef __STDC__
+    extern_ptr_t GC_malloc_atomic(size_t lb)
+# else
+    extern_ptr_t GC_malloc_atomic(lb)
+    size_t lb;
+# endif
+{
+register ptr_t op;
+register ptr_t * opp;
+register word lw;
+DCL_LOCK_STATE;
+
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+#       endif
+	opp = &(GC_aobjfreelist[lw]);
+	FASTLOCK();
+        if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {
+            FASTUNLOCK();
+            return(GENERAL_MALLOC((word)lb, PTRFREE));
+        }
+        /* See above comment on signals.	*/
+        *opp = obj_link(op);
+        GC_words_allocd += lw;
+        FASTUNLOCK();
+        return((extern_ptr_t) op);
+   } else {
+       return(GENERAL_MALLOC((word)lb, PTRFREE));
+   }
+}
+
+/* Allocate lb bytes of composite (pointerful) data */
+# ifdef __STDC__
+    extern_ptr_t GC_malloc(size_t lb)
+# else
+    extern_ptr_t GC_malloc(lb)
+    size_t lb;
+# endif
+{
+register ptr_t op;
+register ptr_t *opp;
+register word lw;
+DCL_LOCK_STATE;
+
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+#       endif
+	opp = &(GC_objfreelist[lw]);
+	FASTLOCK();
+        if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {
+            FASTUNLOCK();
+            return(GENERAL_MALLOC((word)lb, NORMAL));
+        }
+        /* See above comment on signals.	*/
+        *opp = obj_link(op);
+        obj_link(op) = 0;
+        GC_words_allocd += lw;
+        FASTUNLOCK();
+        return((extern_ptr_t) op);
+   } else {
+       return(GENERAL_MALLOC((word)lb, NORMAL));
+   }
+}
+
+/* Allocate lb bytes of pointerful, traced, but not collectable data */
+# ifdef __STDC__
+    extern_ptr_t GC_malloc_uncollectable(size_t lb)
+# else
+    extern_ptr_t GC_malloc_uncollectable(lb)
+    size_t lb;
+# endif
+{
+register ptr_t op;
+register ptr_t *opp;
+register word lw;
+DCL_LOCK_STATE;
+
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+#	  ifdef ADD_BYTE_AT_END
+	    lb--; /* We don't need the extra byte, since this won't be	*/
+	    	  /* collected anyway.					*/
+#	  endif
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+#       endif
+	opp = &(GC_uobjfreelist[lw]);
+	FASTLOCK();
+        if( FASTLOCK_SUCCEEDED() && (op = *opp) != 0 ) {
+            /* See above comment on signals.	*/
+            *opp = obj_link(op);
+            obj_link(op) = 0;
+            GC_words_allocd += lw;
+            GC_set_mark_bit(op);
+            GC_non_gc_bytes += WORDS_TO_BYTES(lw);
+            FASTUNLOCK();
+            return((extern_ptr_t) op);
+        }
+        FASTUNLOCK();
+        op = (ptr_t)GC_generic_malloc((word)lb, UNCOLLECTABLE);
+    } else {
+	op = (ptr_t)GC_generic_malloc((word)lb, UNCOLLECTABLE);
+    }
+    /* We don't need the lock here, since we have an undisguised 	*/
+    /* pointer.  We do need to hold the lock while we adjust		*/
+    /* mark bits.							*/
+    {
+	register struct hblk * h;
+	
+	h = HBLKPTR(op);
+	lw = HDR(h) -> hb_sz;
+	
+	DISABLE_SIGNALS();
+	LOCK();
+	GC_set_mark_bit(op);
+	GC_non_gc_bytes += WORDS_TO_BYTES(lw);
+	UNLOCK();
+	ENABLE_SIGNALS();
+	return((extern_ptr_t) op);
+    }
+}
+
+extern_ptr_t GC_generic_or_special_malloc(lb,knd)
+word lb;
+int knd;
+{
+    switch(knd) {
+#     ifdef STUBBORN_ALLOC
+	case STUBBORN:
+	    return(GC_malloc_stubborn((size_t)lb));
+#     endif
+	case PTRFREE:
+	    return(GC_malloc_atomic((size_t)lb));
+	case NORMAL:
+	    return(GC_malloc((size_t)lb));
+	case UNCOLLECTABLE:
+	    return(GC_malloc_uncollectable((size_t)lb));
+	default:
+	    return(GC_generic_malloc(lb,knd));
+    }
+}
+
+
+/* Change the size of the block pointed to by p to contain at least   */
+/* lb bytes.  The object may be (and quite likely will be) moved.     */
+/* The kind (e.g. atomic) is the same as that of the old.	      */
+/* Shrinking of large blocks is not implemented well.                 */
+# ifdef __STDC__
+    extern_ptr_t GC_realloc(extern_ptr_t p, size_t lb)
+# else
+    extern_ptr_t GC_realloc(p,lb)
+    extern_ptr_t p;
+    size_t lb;
+# endif
+{
+register struct hblk * h;
+register hdr * hhdr;
+register word sz;	 /* Current size in bytes	*/
+register word orig_sz;	 /* Original sz in bytes	*/
+int obj_kind;
+
+    if (p == 0) return(GC_malloc(lb));	/* Required by ANSI */
+    h = HBLKPTR(p);
+    hhdr = HDR(h);
+    sz = hhdr -> hb_sz;
+    obj_kind = hhdr -> hb_obj_kind;
+    sz = WORDS_TO_BYTES(sz);
+    orig_sz = sz;
+
+    if (sz > WORDS_TO_BYTES(MAXOBJSZ)) {
+	/* Round it up to the next whole heap block */
+	  
+	  sz = (sz+HDR_BYTES+HBLKSIZE-1)
+		& (~HBLKMASK);
+	  sz -= HDR_BYTES;
+	  hhdr -> hb_sz = BYTES_TO_WORDS(sz);
+	  if (obj_kind == UNCOLLECTABLE) GC_non_gc_bytes += (sz - orig_sz);
+	  /* Extra area is already cleared by allochblk. */
+    }
+    if (ADD_SLOP(lb) <= sz) {
+	if (lb >= (sz >> 1)) {
+#	    ifdef STUBBORN_ALLOC
+	        if (obj_kind == STUBBORN) GC_change_stubborn(p);
+#	    endif
+	    if (orig_sz > lb) {
+	      /* Clear unneeded part of object to avoid bogus pointer */
+	      /* tracing.					      */
+	      /* Safe for stubborn objects.			      */
+	        BZERO(((ptr_t)p) + lb, orig_sz - lb);
+	    }
+	    return(p);
+	} else {
+	    /* shrink */
+	      extern_ptr_t result =
+	      		GC_generic_or_special_malloc((word)lb, obj_kind);
+
+	      if (result == 0) return(0);
+	          /* Could also return original object.  But this 	*/
+	          /* gives the client warning of imminent disaster.	*/
+	      BCOPY(p, result, lb);
+	      GC_free(p);
+	      return(result);
+	}
+    } else {
+	/* grow */
+	  extern_ptr_t result =
+	  	GC_generic_or_special_malloc((word)lb, obj_kind);
+
+	  if (result == 0) return(0);
+	  BCOPY(p, result, sz);
+	  GC_free(p);
+	  return(result);
+    }
+}
+
+/* Explicitly deallocate an object p.				*/
+# ifdef __STDC__
+    void GC_free(extern_ptr_t p)
+# else
+    void GC_free(p)
+    extern_ptr_t p;
+# endif
+{
+    register struct hblk *h;
+    register hdr *hhdr;
+    register signed_word sz;
+    register ptr_t * flh;
+    register int knd;
+    register struct obj_kind * ok;
+    DCL_LOCK_STATE;
+
+    if (p == 0) return;
+    	/* Required by ANSI.  It's not my fault ...	*/
+    h = HBLKPTR(p);
+    hhdr = HDR(h);
+    knd = hhdr -> hb_obj_kind;
+    sz = hhdr -> hb_sz;
+    ok = &GC_obj_kinds[knd];
+    if (sz <= MAXOBJSZ) {
+#	ifdef THREADS
+	    DISABLE_SIGNALS();
+	    LOCK();
+#	endif
+	GC_mem_freed += sz;
+	/* A signal here can make GC_mem_freed and GC_non_gc_bytes	*/
+	/* inconsistent.  We claim this is benign.			*/
+	if (knd == UNCOLLECTABLE) GC_non_gc_bytes -= sz;
+	if (ok -> ok_init) {
+	    BZERO((word *)p + 1, WORDS_TO_BYTES(sz-1));
+	}
+	flh = &(ok -> ok_freelist[sz]);
+	obj_link(p) = *flh;
+	*flh = (ptr_t)p;
+#	ifdef THREADS
+	    UNLOCK();
+	    ENABLE_SIGNALS();
+#	endif
+    } else {
+    	DISABLE_SIGNALS();
+        LOCK();
+        GC_mem_freed += sz;
+	if (knd == UNCOLLECTABLE) GC_non_gc_bytes -= sz;
+        GC_freehblk(h);
+        UNLOCK();
+        ENABLE_SIGNALS();
+    }
+}
+
diff --git a/mark.c b/mark.c
new file mode 100644
index 00000000..b73ff0e4
--- /dev/null
+++ b/mark.c
@@ -0,0 +1,1026 @@
+
+/*
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ */
+
+
+# include <stdio.h>
+# include "gc_priv.h"
+# include "gc_mark.h"
+
+/* We put this here to minimize the risk of inlining. */
+/*VARARGS*/
+void GC_noop() {}
+
+mark_proc GC_mark_procs[MAX_MARK_PROCS] = {0};
+word GC_n_mark_procs = 0;
+
+/* Initialize GC_obj_kinds properly and standard free lists properly.  	*/
+/* This must be done statically since they may be accessed before 	*/
+/* GC_init is called.							*/
+/* It's done here, since we need to deal with mark descriptors.		*/
+struct obj_kind GC_obj_kinds[MAXOBJKINDS] = {
+/* PTRFREE */ { &GC_aobjfreelist[0], &GC_areclaim_list[0],
+		0 | DS_LENGTH, FALSE, FALSE },
+/* NORMAL  */ { &GC_objfreelist[0], &GC_reclaim_list[0],
+#		ifdef ADD_BYTE_AT_END
+		(word)(WORDS_TO_BYTES(-1)) | DS_LENGTH,
+#		else
+		0 | DS_LENGTH,
+#		endif
+		TRUE /* add length to descr */, TRUE },
+/* UNCOLLECTABLE */
+	      { &GC_uobjfreelist[0], &GC_ureclaim_list[0],
+		0 | DS_LENGTH, TRUE /* add length to descr */, TRUE },
+# ifdef STUBBORN_ALLOC
+/*STUBBORN*/ { &GC_sobjfreelist[0], &GC_sreclaim_list[0],
+		0 | DS_LENGTH, TRUE /* add length to descr */, TRUE },
+# endif
+};
+
+# ifdef STUBBORN_ALLOC
+  int GC_n_kinds = 4;
+# else
+  int GC_n_kinds = 3;
+# endif
+
+
+# define INITIAL_MARK_STACK_SIZE (1*HBLKSIZE)
+		/* INITIAL_MARK_STACK_SIZE * sizeof(mse) should be a 	*/
+		/* multiple of HBLKSIZE.				*/
+
+/*
+ * Limits of stack for GC_mark routine.
+ * All ranges between GC_mark_stack(incl.) and GC_mark_stack_top(incl.) still
+ * need to be marked from.
+ */
+
+word GC_n_rescuing_pages;	/* Number of dirty pages we marked from */
+				/* excludes ptrfree pages, etc.		*/
+
+mse * GC_mark_stack;
+
+word GC_mark_stack_size = 0;
+ 
+mse * GC_mark_stack_top;
+
+static struct hblk * scan_ptr;
+
+mark_state_t GC_mark_state = MS_NONE;
+
+bool GC_mark_stack_too_small = FALSE;
+
+bool GC_objects_are_marked = FALSE;	/* Are there collectable marked	*/
+					/* objects in the heap?		*/
+
+bool GC_collection_in_progress()
+{
+    return(GC_mark_state != MS_NONE);
+}
+
+/* clear all mark bits in the header */
+void GC_clear_hdr_marks(hhdr)
+register hdr * hhdr;
+{
+    BZERO(hhdr -> hb_marks, MARK_BITS_SZ*sizeof(word));
+}
+
+/*
+ * Clear all mark bits associated with block h.
+ */
+/*ARGSUSED*/
+static void clear_marks_for_block(h, dummy)
+struct hblk *h;
+word dummy;
+{
+    register hdr * hhdr = HDR(h);
+    
+    if (hhdr -> hb_obj_kind == UNCOLLECTABLE) return;
+        /* Mark bit for these is cleared only once the object is 	*/
+        /* explicitly deallocated.  This either frees the block, or	*/
+        /* the bit is cleared once the object is on the free list.	*/
+    GC_clear_hdr_marks(hhdr);
+}
+
+/* Slow but general routines for setting/clearing/asking about mark bits */
+void GC_set_mark_bit(p)
+ptr_t p;
+{
+    register struct hblk *h = HBLKPTR(p);
+    register hdr * hhdr = HDR(h);
+    register int word_no = (word *)p - (word *)h;
+    
+    set_mark_bit_from_hdr(hhdr, word_no);
+}
+
+void GC_clear_mark_bit(p)
+ptr_t p;
+{
+    register struct hblk *h = HBLKPTR(p);
+    register hdr * hhdr = HDR(h);
+    register int word_no = (word *)p - (word *)h;
+    
+    clear_mark_bit_from_hdr(hhdr, word_no);
+}
+
+bool GC_is_marked(p)
+ptr_t p;
+{
+    register struct hblk *h = HBLKPTR(p);
+    register hdr * hhdr = HDR(h);
+    register int word_no = (word *)p - (word *)h;
+    
+    return(mark_bit_from_hdr(hhdr, word_no));
+}
+
+
+/*
+ * Clear mark bits in all allocated heap blocks.  This invalidates
+ * the marker invariant, and sets GC_mark_state to reflect this.
+ * (This implicitly starts marking to reestablish the
+ */
+void GC_clear_marks()
+{
+    GC_apply_to_all_blocks(clear_marks_for_block, (word)0);
+    GC_objects_are_marked = FALSE;
+    GC_mark_state = MS_INVALID;
+    scan_ptr = 0;
+#   ifdef GATHERSTATS
+	/* Counters reflect currently marked objects: reset here */
+        GC_composite_in_use = 0;
+        GC_atomic_in_use = 0;
+#   endif
+
+}
+
+/* Initiate full marking.	*/
+void GC_initiate_full()
+{
+#   ifdef PRINTSTATS
+	GC_printf2("***>Full mark for collection %lu after %ld allocd bytes\n",
+		  (unsigned long) GC_gc_no+1,
+	   	  (long)WORDS_TO_BYTES(GC_words_allocd));
+#   endif
+    GC_promote_black_lists();
+    GC_reclaim_or_delete_all();
+    GC_clear_marks();
+    GC_read_dirty();
+#   ifdef STUBBORN_ALLOC
+    	GC_read_changed();
+#   endif
+#   ifdef CHECKSUMS
+	{
+	    extern void GC_check_dirty();
+	    
+	    GC_check_dirty();
+	}
+#   endif
+#   ifdef GATHERSTATS
+	GC_n_rescuing_pages = 0;
+#   endif
+}
+
+/* Initiate partial marking.	*/
+/*ARGSUSED*/
+void GC_initiate_partial()
+{
+    if (GC_dirty_maintained) GC_read_dirty();
+#   ifdef STUBBORN_ALLOC
+    	GC_read_changed();
+#   endif
+#   ifdef CHECKSUMS
+	{
+	    extern void GC_check_dirty();
+	    
+	    if (GC_dirty_maintained) GC_check_dirty();
+	}
+#   endif
+#   ifdef GATHERSTATS
+	GC_n_rescuing_pages = 0;
+#   endif
+    if (GC_mark_state == MS_NONE) {
+        GC_mark_state = MS_PUSH_RESCUERS;
+    } else if (GC_mark_state != MS_INVALID) {
+    	ABORT("unexpected state");
+    } /* else this is really a full collection, and mark	*/
+      /* bits are invalid.					*/
+    scan_ptr = 0;
+}
+
+
+static void alloc_mark_stack();
+
+/* Perform a small amount of marking.			*/
+/* We try to touch roughly a page of memory.		*/
+/* Return TRUE if we just finished a mark phase.	*/
+bool GC_mark_some()
+{
+    switch(GC_mark_state) {
+    	case MS_NONE:
+    	    return(FALSE);
+    	    
+    	case MS_PUSH_RESCUERS:
+    	    if (GC_mark_stack_top
+    	        >= GC_mark_stack + INITIAL_MARK_STACK_SIZE/4) {
+    	        GC_mark_from_mark_stack();
+    	        return(FALSE);
+    	    } else {
+    	        scan_ptr = GC_push_next_marked_dirty(scan_ptr);
+    	        if (scan_ptr == 0) {
+#		    ifdef PRINTSTATS
+			GC_printf1("Marked from %lu dirty pages\n",
+				   (unsigned long)GC_n_rescuing_pages);
+#		    endif
+    	    	    GC_push_roots(FALSE);
+    	    	    GC_objects_are_marked = TRUE;
+    	    	    if (GC_mark_state != MS_INVALID) {
+    	    	        GC_mark_state = MS_ROOTS_PUSHED;
+    	    	    }
+    	    	}
+    	    }
+    	    return(FALSE);
+    	
+    	case MS_PUSH_UNCOLLECTABLE:
+    	    if (GC_mark_stack_top
+    	        >= GC_mark_stack + INITIAL_MARK_STACK_SIZE/4) {
+    	        GC_mark_from_mark_stack();
+    	        return(FALSE);
+    	    } else {
+    	        scan_ptr = GC_push_next_marked_uncollectable(scan_ptr);
+    	        if (scan_ptr == 0) {
+    	    	    GC_push_roots(TRUE);
+    	    	    GC_objects_are_marked = TRUE;
+    	    	    if (GC_mark_state != MS_INVALID) {
+    	    	        GC_mark_state = MS_ROOTS_PUSHED;
+    	    	    }
+    	    	}
+    	    }
+    	    return(FALSE);
+    	
+    	case MS_ROOTS_PUSHED:
+    	    if (GC_mark_stack_top >= GC_mark_stack) {
+    	        GC_mark_from_mark_stack();
+    	        return(FALSE);
+    	    } else {
+    	        GC_mark_state = MS_NONE;
+    	        if (GC_mark_stack_too_small) {
+    	            alloc_mark_stack(2*GC_mark_stack_size);
+    	        }
+    	        return(TRUE);
+    	    }
+    	    
+    	case MS_INVALID:
+    	case MS_PARTIALLY_INVALID:
+	    if (!GC_objects_are_marked) {
+		GC_mark_state = MS_PUSH_UNCOLLECTABLE;
+		return(FALSE);
+	    }
+    	    if (GC_mark_stack_top >= GC_mark_stack) {
+    	        GC_mark_from_mark_stack();
+    	        return(FALSE);
+    	    }
+    	    if (scan_ptr == 0
+    	        && (GC_mark_state == MS_INVALID || GC_mark_stack_too_small)) {
+    	        alloc_mark_stack(2*GC_mark_stack_size);
+		GC_mark_state = MS_PARTIALLY_INVALID;
+    	    }
+    	    scan_ptr = GC_push_next_marked(scan_ptr);
+    	    if (scan_ptr == 0 && GC_mark_state == MS_PARTIALLY_INVALID) {
+    	    	GC_push_roots(TRUE);
+    	    	GC_objects_are_marked = TRUE;
+    	    	if (GC_mark_state != MS_INVALID) {
+    	    	    GC_mark_state = MS_ROOTS_PUSHED;
+    	    	}
+    	    }
+    	    return(FALSE);
+    	default:
+    	    ABORT("GC_mark_some: bad state");
+    	    return(FALSE);
+    }
+}
+
+
+bool GC_mark_stack_empty()
+{
+    return(GC_mark_stack_top < GC_mark_stack);
+}	
+
+#ifdef PROF_MARKER
+    word GC_prof_array[10];
+#   define PROF(n) GC_prof_array[n]++
+#else
+#   define PROF(n)
+#endif
+
+/* Given a pointer to someplace other than a small object page or the	*/
+/* first page of a large object, return a pointer either to the		*/
+/* start of the large object or NIL.					*/
+/* In the latter case black list the address current.			*/
+/* Returns NIL without black listing if current points to a block	*/
+/* with IGNORE_OFF_PAGE set.						*/
+/*ARGSUSED*/
+word GC_find_start(current, hhdr)
+register word current;
+register hdr * hhdr;
+{
+#   ifdef ALL_INTERIOR_POINTERS
+	if (hhdr != 0) {
+	    register word orig = current;
+	    
+	    current = (word)HBLKPTR(current) + HDR_BYTES;
+	    do {
+	      current = current - HBLKSIZE*(int)hhdr;
+	      hhdr = HDR(current);
+	    } while(IS_FORWARDING_ADDR_OR_NIL(hhdr));
+	    /* current points to the start of the large object */
+	    if (hhdr -> hb_flags & IGNORE_OFF_PAGE) return(0);
+	    if ((word *)orig - (word *)current
+	         >= (ptrdiff_t)(hhdr->hb_sz)) {
+	        /* Pointer past the end of the block */
+	        GC_ADD_TO_BLACK_LIST_NORMAL(orig);
+	        return(0);
+	    }
+	    return(current);
+	} else {
+	    GC_ADD_TO_BLACK_LIST_NORMAL(current);
+	    return(0);
+        }
+#   else
+        GC_ADD_TO_BLACK_LIST_NORMAL(current);
+        return(0);
+#   endif
+}
+
+mse * GC_signal_mark_stack_overflow(msp)
+mse * msp;
+{
+    GC_mark_state = MS_INVALID;
+#   ifdef PRINTSTATS
+	GC_printf1("Mark stack overflow; current size = %lu entries\n",
+	    	    GC_mark_stack_size);
+#    endif
+     return(msp-INITIAL_MARK_STACK_SIZE/8);
+}
+
+
+/*
+ * Mark objects pointed to by the regions described by
+ * mark stack entries between GC_mark_stack and GC_mark_stack_top,
+ * inclusive.  Assumes the upper limit of a mark stack entry
+ * is never 0.  A mark stack entry never has size 0.
+ * We try to traverse on the order of a hblk of memory before we return.
+ * Caller is responsible for calling this until the mark stack is empty.
+ */
+void GC_mark_from_mark_stack()
+{
+  mse * GC_mark_stack_reg = GC_mark_stack;
+  mse * GC_mark_stack_top_reg = GC_mark_stack_top;
+  mse * mark_stack_limit = &(GC_mark_stack[GC_mark_stack_size]);
+  int credit = HBLKSIZE;	/* Remaining credit for marking work	*/
+  register word * current_p;	/* Pointer to current candidate ptr.	*/
+  register word current;	/* Candidate pointer.			*/
+  register word * limit;	/* (Incl) limit of current candidate 	*/
+  				/* range				*/
+  register word descr;
+  register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
+  register ptr_t least_ha = GC_least_plausible_heap_addr;
+# define SPLIT_RANGE_WORDS 128  /* Must be power of 2.		*/
+
+  GC_objects_are_marked = TRUE;
+# ifdef OS2 /* Use untweaked version to circumvent compiler problem */
+  while (GC_mark_stack_top_reg >= GC_mark_stack_reg && credit >= 0) {
+# else
+  while ((((ptr_t)GC_mark_stack_top_reg - (ptr_t)GC_mark_stack_reg) | credit)
+  	>= 0) {
+# endif
+    current_p = GC_mark_stack_top_reg -> mse_start;
+    descr = GC_mark_stack_top_reg -> mse_descr;
+  retry:  
+    if (descr & ((~(WORDS_TO_BYTES(SPLIT_RANGE_WORDS) - 1)) | DS_TAGS)) {
+      word tag = descr & DS_TAGS;
+      
+      switch(tag) {
+        case DS_LENGTH:
+          /* Large length.					        */
+          /* Process part of the range to avoid pushing too much on the	*/
+          /* stack.							*/
+          GC_mark_stack_top_reg -> mse_start =
+         	limit = current_p + SPLIT_RANGE_WORDS-1;
+          GC_mark_stack_top_reg -> mse_descr -=
+          		WORDS_TO_BYTES(SPLIT_RANGE_WORDS-1);
+          /* Make sure that pointers overlapping the two ranges are	*/
+          /* considered. 						*/
+          limit += sizeof(word) - ALIGNMENT;
+          break;
+        case DS_BITMAP:
+          GC_mark_stack_top_reg--;
+          descr &= ~DS_TAGS;
+          credit -= WORDS_TO_BYTES(WORDSZ/2); /* guess */
+          while (descr != 0) {
+            if ((signed_word)descr < 0) {
+              current = *current_p++;
+              descr <<= 1;
+              if ((ptr_t)current < least_ha) continue;
+              if ((ptr_t)current >= greatest_ha) continue;
+              PUSH_CONTENTS(current, GC_mark_stack_top_reg, mark_stack_limit);
+            } else {
+              descr <<= 1;
+              current_p++;
+            }
+          }
+          continue;
+        case DS_PROC:
+          GC_mark_stack_top_reg--;
+          credit -= PROC_BYTES;
+          GC_mark_stack_top_reg =
+              (*PROC(descr))
+              	    (current_p, GC_mark_stack_top_reg,
+              	    mark_stack_limit, ENV(descr));
+          continue;
+        case DS_PER_OBJECT:
+          descr = *(word *)((ptr_t)current_p + descr - tag);
+          goto retry;
+      }
+    } else {
+      GC_mark_stack_top_reg--;
+      limit = (word *)(((ptr_t)current_p) + (word)descr);
+    }
+    /* The simple case in which we're scanning a range.	*/
+    credit -= (ptr_t)limit - (ptr_t)current_p;
+    limit -= 1;
+    while (current_p <= limit) {
+      current = *current_p;
+      current_p = (word *)((char *)current_p + ALIGNMENT);
+      if ((ptr_t)current < least_ha) continue;
+      if ((ptr_t)current >= greatest_ha) continue;
+      PUSH_CONTENTS(current, GC_mark_stack_top_reg, mark_stack_limit);
+    }
+  }
+  GC_mark_stack_top = GC_mark_stack_top_reg;
+}
+
+/* Allocate or reallocate space for mark stack of size s words  */
+/* May silently fail.						*/
+static void alloc_mark_stack(n)
+word n;
+{
+    mse * new_stack = (mse *)GC_scratch_alloc(n * sizeof(struct ms_entry));
+    
+    GC_mark_stack_too_small = FALSE;
+    if (GC_mark_stack_size != 0) {
+        if (new_stack != 0) {
+          word displ = HBLKDISPL(GC_mark_stack);
+          word size = GC_mark_stack_size * sizeof(struct ms_entry);
+          
+          /* Recycle old space */
+            if (displ == 0) {
+              GC_add_to_heap((struct hblk *)GC_mark_stack, size);
+	    } else {
+	      GC_add_to_heap((struct hblk *)
+	      			((word)GC_mark_stack - displ + HBLKSIZE),
+	      		     size - HBLKSIZE);
+	    }
+          GC_mark_stack = new_stack;
+          GC_mark_stack_size = n;
+#	  ifdef PRINTSTATS
+	      GC_printf1("Grew mark stack to %lu frames\n",
+		    	 (unsigned long) GC_mark_stack_size);
+#	  endif
+        } else {
+#	  ifdef PRINTSTATS
+	      GC_printf1("Failed to grow mark stack to %lu frames\n",
+		    	 (unsigned long) n);
+#	  endif
+        }
+    } else {
+        if (new_stack == 0) {
+            GC_err_printf0("No space for mark stack\n");
+            EXIT();
+        }
+        GC_mark_stack = new_stack;
+        GC_mark_stack_size = n;
+    }
+    GC_mark_stack_top = GC_mark_stack-1;
+}
+
+void GC_mark_init()
+{
+    alloc_mark_stack(INITIAL_MARK_STACK_SIZE);
+}
+
+/*
+ * Push all locations between b and t onto the mark stack.
+ * b is the first location to be checked. t is one past the last
+ * location to be checked.
+ * Should only be used if there is no possibility of mark stack
+ * overflow.
+ */
+void GC_push_all(bottom, top)
+ptr_t bottom;
+ptr_t top;
+{
+    register word length;
+    
+    bottom = (ptr_t)(((word) bottom + ALIGNMENT-1) & ~(ALIGNMENT-1));
+    top = (ptr_t)(((word) top) & ~(ALIGNMENT-1));
+    if (top == 0 || bottom == top) return;
+    GC_mark_stack_top++;
+    if (GC_mark_stack_top >= GC_mark_stack + GC_mark_stack_size) {
+	ABORT("unexpected mark stack overflow");
+    }
+    length = top - bottom;
+#   if DS_TAGS > ALIGNMENT - 1
+	length += DS_TAGS;
+	length &= ~DS_TAGS;
+#   endif
+    GC_mark_stack_top -> mse_start = (word *)bottom;
+    GC_mark_stack_top -> mse_descr = length;
+}
+
+/*
+ * Analogous to the above, but push only those pages that may have been
+ * dirtied.  A block h is assumed dirty if dirty_fn(h) != 0.
+ * We use push_fn to actually push the block.
+ * Will not overflow mark stack if push_fn pushes a small fixed number
+ * of entries.  (This is invoked only if push_fn pushes a single entry,
+ * or if it marks each object before pushing it, thus ensuring progress
+ * in the event of a stack overflow.)
+ */
+void GC_push_dirty(bottom, top, dirty_fn, push_fn)
+ptr_t bottom;
+ptr_t top;
+int (*dirty_fn)(/* struct hblk * h */);
+void (*push_fn)(/* ptr_t bottom, ptr_t top */);
+{
+    register struct hblk * h;
+
+    bottom = (ptr_t)(((long) bottom + ALIGNMENT-1) & ~(ALIGNMENT-1));
+    top = (ptr_t)(((long) top) & ~(ALIGNMENT-1));
+
+    if (top == 0 || bottom == top) return;
+    h = HBLKPTR(bottom + HBLKSIZE);
+    if (top <= (ptr_t) h) {
+  	if ((*dirty_fn)(h-1)) {
+	    (*push_fn)(bottom, top);
+	}
+	return;
+    }
+    if ((*dirty_fn)(h-1)) {
+        (*push_fn)(bottom, (ptr_t)h);
+    }
+    while ((ptr_t)(h+1) <= top) {
+	if ((*dirty_fn)(h)) {
+	    if ((word)(GC_mark_stack_top - GC_mark_stack)
+		> 3 * GC_mark_stack_size / 4) {
+	 	/* Danger of mark stack overflow */
+		(*push_fn)((ptr_t)h, top);
+		return;
+	    } else {
+		(*push_fn)((ptr_t)h, (ptr_t)(h+1));
+	    }
+	}
+	h++;
+    }
+    if ((ptr_t)h != top) {
+	if ((*dirty_fn)(h)) {
+            (*push_fn)((ptr_t)h, top);
+        }
+    }
+    if (GC_mark_stack_top >= GC_mark_stack + GC_mark_stack_size) {
+        ABORT("unexpected mark stack overflow");
+    }
+}
+
+# ifndef SMALL_CONFIG
+void GC_push_conditional(bottom, top, all)
+ptr_t bottom;
+ptr_t top;
+{
+    if (all) {
+      if (GC_dirty_maintained) {
+#	ifdef PROC_VDB
+	    /* Pages that were never dirtied cannot contain pointers	*/
+	    GC_push_dirty(bottom, top, GC_page_was_ever_dirty, GC_push_all);
+#	else
+	    GC_push_all(bottom, top);
+#	endif
+      } else {
+      	GC_push_all(bottom, top);
+      }
+    } else {
+	GC_push_dirty(bottom, top, GC_page_was_dirty, GC_push_all);
+    }
+}
+#endif
+
+/*
+ * Push a single value onto mark stack. Mark from the object pointed to by p.
+ * GC_push_one is normally called by GC_push_regs, and thus must be defined.
+ * P is considered valid even if it is an interior pointer.
+ * Previously marked objects are not pushed.  Hence we make progress even
+ * if the mark stack overflows.
+ */
+# define GC_PUSH_ONE_STACK(p) \
+    if ((ptr_t)(p) >= GC_least_plausible_heap_addr 	\
+	 && (ptr_t)(p) < GC_greatest_plausible_heap_addr) {	\
+	 GC_push_one_checked(p,TRUE);	\
+    }
+
+/*
+ * As above, but interior pointer recognition as for
+ * normal for heap pointers.
+ */
+# ifdef ALL_INTERIOR_POINTERS
+#   define AIP TRUE
+# else
+#   define AIP FALSE
+# endif
+# define GC_PUSH_ONE_HEAP(p) \
+    if ((ptr_t)(p) >= GC_least_plausible_heap_addr 	\
+	 && (ptr_t)(p) < GC_greatest_plausible_heap_addr) {	\
+	 GC_push_one_checked(p,AIP);	\
+    }
+
+# ifdef MSWIN32
+  void __cdecl GC_push_one(p)
+# else
+  void GC_push_one(p)
+# endif
+word p;
+{
+    GC_PUSH_ONE_STACK(p);
+}
+
+# ifdef __STDC__
+#   define BASE(p) (word)GC_base((void *)(p))
+# else
+#   define BASE(p) (word)GC_base((char *)(p))
+# endif
+
+/* As above, but argument passed preliminary test. */
+void GC_push_one_checked(p, interior_ptrs)
+register word p;
+register bool interior_ptrs;
+{
+    register word r;
+    register hdr * hhdr; 
+    register int displ;
+  
+    GET_HDR(p, hhdr);
+    if (IS_FORWARDING_ADDR_OR_NIL(hhdr)) {
+        if (hhdr != 0 && interior_ptrs) {
+          r = BASE(p);
+	  hhdr = HDR(r);
+	  displ = BYTES_TO_WORDS(HBLKDISPL(r));
+	} else {
+	  hhdr = 0;
+	}
+    } else {
+        register map_entry_type map_entry;
+        
+        displ = HBLKDISPL(p);
+        map_entry = MAP_ENTRY((hhdr -> hb_map), displ);
+        if (map_entry == OBJ_INVALID) {
+          if (interior_ptrs) {
+            r = BASE(p);
+	    displ = BYTES_TO_WORDS(HBLKDISPL(r));
+	    if (r == 0) hhdr = 0;
+          } else {
+            hhdr = 0;
+          }
+        } else {
+          displ = BYTES_TO_WORDS(displ);
+          displ -= map_entry;
+          r = (word)((word *)(HBLKPTR(p)) + displ);
+        }
+    }
+    /* If hhdr != 0 then r == GC_base(p), only we did it faster. */
+    /* displ is the word index within the block.		 */
+    if (hhdr == 0) {
+    	if (interior_ptrs) {
+	    GC_add_to_black_list_stack(p);
+	} else {
+	    GC_ADD_TO_BLACK_LIST_NORMAL(p);
+	}
+    } else {
+	if (!mark_bit_from_hdr(hhdr, displ)) {
+	    set_mark_bit_from_hdr(hhdr, displ);
+	    PUSH_OBJ((word *)r, hhdr, GC_mark_stack_top,
+	             &(GC_mark_stack[GC_mark_stack_size]));
+	}
+    }
+}
+
+/*
+ * A version of GC_push_all that treats all interior pointers as valid
+ */
+void GC_push_all_stack(bottom, top)
+ptr_t bottom;
+ptr_t top;
+{
+# ifdef ALL_INTERIOR_POINTERS
+    GC_push_all(bottom, top);
+# else
+    word * b = (word *)(((long) bottom + ALIGNMENT-1) & ~(ALIGNMENT-1));
+    word * t = (word *)(((long) top) & ~(ALIGNMENT-1));
+    register word *p;
+    register word q;
+    register word *lim;
+    register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
+    register ptr_t least_ha = GC_least_plausible_heap_addr;
+#   define GC_greatest_plausible_heap_addr greatest_ha
+#   define GC_least_plausible_heap_addr least_ha
+
+    if (top == 0) return;
+    /* check all pointers in range and put in push if they appear */
+    /* to be valid.						  */
+      lim = t - 1 /* longword */;
+      for (p = b; p <= lim; p = (word *)(((char *)p) + ALIGNMENT)) {
+	q = *p;
+	GC_PUSH_ONE_STACK(q);
+      }
+#   undef GC_greatest_plausible_heap_addr
+#   undef GC_least_plausible_heap_addr
+# endif
+}
+
+#ifndef SMALL_CONFIG
+/* Push all objects reachable from marked objects in the given block */
+/* of size 1 objects.						     */
+void GC_push_marked1(h, hhdr)
+struct hblk *h;
+register hdr * hhdr;
+{
+    word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p;
+    word *plim;
+    register int i;
+    register word q;
+    register word mark_word;
+    register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
+    register ptr_t least_ha = GC_least_plausible_heap_addr;
+#   define GC_greatest_plausible_heap_addr greatest_ha
+#   define GC_least_plausible_heap_addr least_ha
+    
+    p = (word *)(h->hb_body);
+    plim = (word *)(((word)h) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    i = 0;
+	    while(mark_word != 0) {
+	      if (mark_word & 1) {
+	          q = p[i];
+	          GC_PUSH_ONE_HEAP(q);
+	      }
+	      i++;
+	      mark_word >>= 1;
+	    }
+	    p += WORDSZ;
+	}
+#   undef GC_greatest_plausible_heap_addr
+#   undef GC_least_plausible_heap_addr        
+}
+
+
+/* Push all objects reachable from marked objects in the given block */
+/* of size 2 objects.						     */
+void GC_push_marked2(h, hhdr)
+struct hblk *h;
+register hdr * hhdr;
+{
+    word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p;
+    word *plim;
+    register int i;
+    register word q;
+    register word mark_word;
+    register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
+    register ptr_t least_ha = GC_least_plausible_heap_addr;
+#   define GC_greatest_plausible_heap_addr greatest_ha
+#   define GC_least_plausible_heap_addr least_ha
+    
+    p = (word *)(h->hb_body);
+    plim = (word *)(((word)h) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    i = 0;
+	    while(mark_word != 0) {
+	      if (mark_word & 1) {
+	          q = p[i];
+	          GC_PUSH_ONE_HEAP(q);
+	          q = p[i+1];
+	          GC_PUSH_ONE_HEAP(q);
+	      }
+	      i += 2;
+	      mark_word >>= 2;
+	    }
+	    p += WORDSZ;
+	}
+#   undef GC_greatest_plausible_heap_addr
+#   undef GC_least_plausible_heap_addr        
+}
+
+/* Push all objects reachable from marked objects in the given block */
+/* of size 4 objects.						     */
+/* There is a risk of mark stack overflow here.  But we handle that. */
+/* And only unmarked objects get pushed, so it's not very likely.    */
+void GC_push_marked4(h, hhdr)
+struct hblk *h;
+register hdr * hhdr;
+{
+    word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p;
+    word *plim;
+    register int i;
+    register word q;
+    register word mark_word;
+    register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
+    register ptr_t least_ha = GC_least_plausible_heap_addr;
+#   define GC_greatest_plausible_heap_addr greatest_ha
+#   define GC_least_plausible_heap_addr least_ha
+    
+    p = (word *)(h->hb_body);
+    plim = (word *)(((word)h) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    i = 0;
+	    while(mark_word != 0) {
+	      if (mark_word & 1) {
+	          q = p[i];
+	          GC_PUSH_ONE_HEAP(q);
+	          q = p[i+1];
+	          GC_PUSH_ONE_HEAP(q);
+	          q = p[i+2];
+	          GC_PUSH_ONE_HEAP(q);
+	          q = p[i+3];
+	          GC_PUSH_ONE_HEAP(q);
+	      }
+	      i += 4;
+	      mark_word >>= 4;
+	    }
+	    p += WORDSZ;
+	}
+#   undef GC_greatest_plausible_heap_addr
+#   undef GC_least_plausible_heap_addr        
+}
+
+#endif /* SMALL_CONFIG */
+
+/* Push all objects reachable from marked objects in the given block */
+void GC_push_marked(h, hhdr)
+struct hblk *h;
+register hdr * hhdr;
+{
+    register int sz = hhdr -> hb_sz;
+    register word * p;
+    register int word_no;
+    register word * lim;
+    register mse * GC_mark_stack_top_reg;
+    register mse * mark_stack_limit = &(GC_mark_stack[GC_mark_stack_size]);
+    
+    /* Some quick shortcuts: */
+        if (hhdr -> hb_obj_kind == PTRFREE) return;
+        if (GC_block_empty(hhdr)/* nothing marked */) return;
+#   ifdef GATHERSTATS
+        GC_n_rescuing_pages++;
+#   endif
+    GC_objects_are_marked = TRUE;
+    if (sz > MAXOBJSZ) {
+        lim = (word *)(h + 1);
+    } else {
+        lim = (word *)(h + 1) - sz;
+    }
+    
+    switch(sz) {
+#   ifndef SMALL_CONFIG    
+     case 1:
+       GC_push_marked1(h, hhdr);
+       break;
+     case 2:
+       GC_push_marked2(h, hhdr);
+       break;
+     case 4:
+       GC_push_marked4(h, hhdr);
+       break;
+#   endif       
+     default:
+      GC_mark_stack_top_reg = GC_mark_stack_top;
+      for (p = (word *)h + HDR_WORDS, word_no = HDR_WORDS; p <= lim;
+         p += sz, word_no += sz) {
+         /* This needs manual optimization: */
+         if (mark_bit_from_hdr(hhdr, word_no)) {
+           /* Mark from fields inside the object */
+             PUSH_OBJ((word *)p, hhdr, GC_mark_stack_top_reg, mark_stack_limit);
+#	     ifdef GATHERSTATS
+		/* Subtract this object from total, since it was	*/
+		/* added in twice.					*/
+		GC_composite_in_use -= sz;
+#	     endif
+         }
+      }
+      GC_mark_stack_top = GC_mark_stack_top_reg;
+    }
+}
+
+#ifndef SMALL_CONFIG
+/* Test whether any page in the given block is dirty	*/
+bool GC_block_was_dirty(h, hhdr)
+struct hblk *h;
+register hdr * hhdr;
+{
+    register int sz = hhdr -> hb_sz;
+    
+    if (sz < MAXOBJSZ) {
+         return(GC_page_was_dirty(h));
+    } else {
+    	 register ptr_t p = (ptr_t)h;
+         sz += HDR_WORDS;
+         sz = WORDS_TO_BYTES(sz);
+         while (p < (ptr_t)h + sz) {
+             if (GC_page_was_dirty((struct hblk *)p)) return(TRUE);
+             p += HBLKSIZE;
+         }
+         return(FALSE);
+    }
+}
+#endif /* SMALL_CONFIG */
+
+/* Similar to GC_push_next_marked, but return address of next block	*/
+struct hblk * GC_push_next_marked(h)
+struct hblk *h;
+{
+    register hdr * hhdr;
+    
+    h = GC_next_block(h);
+    if (h == 0) return(0);
+    hhdr = HDR(h);
+    GC_push_marked(h, hhdr);
+    return(h + OBJ_SZ_TO_BLOCKS(hhdr -> hb_sz));
+}
+
+#ifndef SMALL_CONFIG
+/* Identical to above, but mark only from dirty pages	*/
+struct hblk * GC_push_next_marked_dirty(h)
+struct hblk *h;
+{
+    register hdr * hhdr = HDR(h);
+    
+    if (!GC_dirty_maintained) { ABORT("dirty bits not set up"); }
+    for (;;) {
+        h = GC_next_block(h);
+        if (h == 0) return(0);
+        hhdr = HDR(h);
+#	ifdef STUBBORN_ALLOC
+          if (hhdr -> hb_obj_kind == STUBBORN) {
+            if (GC_page_was_changed(h) && GC_block_was_dirty(h, hhdr)) {
+                break;
+            }
+          } else {
+            if (GC_block_was_dirty(h, hhdr)) break;
+          }
+#	else
+	  if (GC_block_was_dirty(h, hhdr)) break;
+#	endif
+        h += OBJ_SZ_TO_BLOCKS(hhdr -> hb_sz);
+    }
+    GC_push_marked(h, hhdr);
+    return(h + OBJ_SZ_TO_BLOCKS(hhdr -> hb_sz));
+}
+#endif
+
+/* Similar to above, but for uncollectable pages.  Needed since we	*/
+/* do not clear marks for such pages, even for full collections.	*/
+struct hblk * GC_push_next_marked_uncollectable(h)
+struct hblk *h;
+{
+    register hdr * hhdr = HDR(h);
+    
+    for (;;) {
+        h = GC_next_block(h);
+        if (h == 0) return(0);
+        hhdr = HDR(h);
+	if (hhdr -> hb_obj_kind == UNCOLLECTABLE) break;
+        h += OBJ_SZ_TO_BLOCKS(hhdr -> hb_sz);
+    }
+    GC_push_marked(h, hhdr);
+    return(h + OBJ_SZ_TO_BLOCKS(hhdr -> hb_sz));
+}
+
+
diff --git a/mark_rts.c b/mark_rts.c
new file mode 100644
index 00000000..376746f1
--- /dev/null
+++ b/mark_rts.c
@@ -0,0 +1,280 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 1:58 pm PDT */
+# include <stdio.h>
+# include "gc_priv.h"
+
+# ifdef PCR
+#   define MAX_ROOT_SETS 1024
+# else
+#   ifdef MSWIN32
+#	define MAX_ROOT_SETS 512
+	    /* Under NT, we add only written pages, which can result 	*/
+	    /* in many small root sets.					*/
+#   else
+#       define MAX_ROOT_SETS 64
+#   endif
+# endif
+
+/* Data structure for list of root sets.				*/
+/* We keep a hash table, so that we can filter out duplicate additions.	*/
+/* Under Win32, we need to do a better job of filtering overlaps, so	*/
+/* we resort to sequential search, and pay the price.			*/
+struct roots {
+	ptr_t r_start;
+	ptr_t r_end;
+#	ifndef MSWIN32
+	  struct roots * r_next;
+#	endif
+};
+
+static struct roots static_roots[MAX_ROOT_SETS];
+
+static int n_root_sets = 0;
+
+	/* static_roots[0..n_root_sets) contains the valid root sets. */
+
+#ifndef MSWIN32
+#   define LOG_RT_SIZE 6
+#   define RT_SIZE (1 << LOG_RT_SIZE)  /* Power of 2, may be != MAX_ROOT_SETS */
+
+    static struct roots * root_index[RT_SIZE];
+	/* Hash table header.  Used only to check whether a range is 	*/
+	/* already present.						*/
+
+static int rt_hash(addr)
+char * addr;
+{
+    word result = (word) addr;
+#   if CPP_WORDSZ > 8*LOG_RT_SIZE
+	result ^= result >> 8*LOG_RT_SIZE;
+#   endif
+#   if CPP_WORDSZ > 4*LOG_RT_SIZE
+	result ^= result >> 4*LOG_RT_SIZE;
+#   endif
+    result ^= result >> 2*LOG_RT_SIZE;
+    result ^= result >> LOG_RT_SIZE;
+    result &= (RT_SIZE-1);
+    return(result);
+}
+
+/* Is a range starting at b already in the table? If so return a	*/
+/* pointer to it, else NIL.						*/
+struct roots * GC_roots_present(b)
+char *b;
+{
+    register int h = rt_hash(b);
+    register struct roots *p = root_index[h];
+    
+    while (p != 0) {
+        if (p -> r_start == (ptr_t)b) return(p);
+        p = p -> r_next;
+    }
+    return(FALSE);
+}
+
+/* Add the given root structure to the index. */
+static void add_roots_to_index(p)
+struct roots *p;
+{
+    register int h = rt_hash(p -> r_start);
+    
+    p -> r_next = root_index[h];
+    root_index[h] = p;
+}
+
+# else /* MSWIN32 */
+
+#   define add_roots_to_index(p)
+
+# endif
+
+
+
+
+word GC_root_size = 0;
+
+void GC_add_roots(b, e)
+char * b; char * e;
+{
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    GC_add_roots_inner(b, e);
+    UNLOCK();
+    ENABLE_SIGNALS();
+}
+
+
+/* Add [b,e) to the root set.  Adding the same interval a second time	*/
+/* is a moderately fast noop, and hence benign.  We do not handle	*/
+/* different but overlapping intervals efficiently.  (We do handle	*/
+/* them correctly.)							*/
+void GC_add_roots_inner(b, e)
+char * b; char * e;
+{
+    struct roots * old;
+    
+    /* We exclude GC data structures from root sets.  It's usually safe	*/
+    /* to mark from those, but it is a waste of time.			*/
+    if ( (ptr_t)b < endGC_arrays && (ptr_t)e > beginGC_arrays) {
+        if ((ptr_t)e <= endGC_arrays) {
+            if ((ptr_t)b >= beginGC_arrays) return;
+            e = (char *)beginGC_arrays;
+        } else if ((ptr_t)b >= beginGC_arrays) {
+            b = (char *)endGC_arrays;
+        } else {
+            GC_add_roots_inner(b, (char *)beginGC_arrays);
+            GC_add_roots_inner((char *)endGC_arrays, e);
+            return;
+        }
+    }
+#   ifdef MSWIN32
+      /* Spend the time to ensure that there are no overlapping	*/
+      /* or adjacent intervals.					*/
+      /* This could be done faster with e.g. a			*/
+      /* balanced tree.  But the execution time here is		*/
+      /* virtually guaranteed to be dominated by the time it	*/
+      /* takes to scan the roots.				*/
+      {
+        register int i;
+        
+        for (i = 0; i < n_root_sets; i++) {
+            old = static_roots + i;
+            if ((ptr_t)b <= old -> r_end && (ptr_t)e >= old -> r_start) {
+                if ((ptr_t)b < old -> r_start) {
+                    old -> r_start = (ptr_t)b;
+                }
+                if ((ptr_t)e > old -> r_end) {
+                    old -> r_end = (ptr_t)e;
+                }
+                break;
+            }
+        }
+        if (i < n_root_sets) {
+          /* merge other overlapping intervals */
+            struct roots *other;
+            
+            for (i++; i < n_root_sets; i++) {
+              other = static_roots + i;
+              b = (char *)(other -> r_start);
+              e = (char *)(other -> r_end);
+              if ((ptr_t)b <= old -> r_end && (ptr_t)e >= old -> r_start) {
+                if ((ptr_t)b < old -> r_start) {
+                    old -> r_start = (ptr_t)b;
+                }
+                if ((ptr_t)e > old -> r_end) {
+                    old -> r_end = (ptr_t)e;
+                }
+                /* Delete this entry. */
+                  other -> r_start = static_roots[n_root_sets-1].r_start;
+                  other -> r_end = static_roots[n_root_sets-1].r_end;
+                  n_root_sets--;
+              }
+            }
+          return;
+        }
+      }
+#   else
+      old = GC_roots_present(b);
+      if (old != 0) {
+        if ((ptr_t)e <= old -> r_end) /* already there */ return;
+        /* else extend */
+        GC_root_size += (ptr_t)e - old -> r_end;
+        old -> r_end = (ptr_t)e;
+        return;
+      }
+#   endif
+    if (n_root_sets == MAX_ROOT_SETS) {
+        ABORT("Too many root sets\n");
+    }
+    static_roots[n_root_sets].r_start = (ptr_t)b;
+    static_roots[n_root_sets].r_end = (ptr_t)e;
+#   ifndef MSWIN32
+      static_roots[n_root_sets].r_next = 0;
+#   endif
+    add_roots_to_index(static_roots + n_root_sets);
+    GC_root_size += (ptr_t)e - (ptr_t)b;
+    n_root_sets++;
+}
+
+void GC_clear_roots()
+{
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    n_root_sets = 0;
+    GC_root_size = 0;
+    UNLOCK();
+    ENABLE_SIGNALS();
+}
+
+# ifndef THREADS
+ptr_t GC_approx_sp()
+{
+    word dummy;
+    
+    return((ptr_t)(&dummy));
+}
+# endif
+
+/*
+ * Call the mark routines (GC_tl_push for a single pointer, GC_push_conditional
+ * on groups of pointers) on every top level accessible pointer.
+ * If all is FALSE, arrange to push only possibly altered values.
+ */
+
+void GC_push_roots(all)
+bool all;
+{
+    register int i;
+
+    /*
+     * push registers - i.e., call GC_push_one(r) for each
+     * register contents r.
+     */
+        GC_push_regs(); /* usually defined in machine_dep.c */
+        
+    /*
+     * Next push static data.  This must happen early on, since it's
+     * not robust against mark stack overflow.
+     */
+     /* Reregister dynamic libraries, in case one got added.	*/
+#      if (defined(DYNAMIC_LOADING) || defined(MSWIN32) || defined(PCR)) \
+           && !defined(SRC_M3)
+         GC_register_dynamic_libraries();
+#      endif
+     /* Mark everything in static data areas                             */
+       for (i = 0; i < n_root_sets; i++) {
+         GC_push_conditional(static_roots[i].r_start,
+			     static_roots[i].r_end, all);
+       }
+
+    /*
+     * Now traverse stacks.
+     */
+#   ifndef THREADS
+        /* Mark everything on the stack.           */
+#   	  ifdef STACK_GROWS_DOWN
+	    GC_push_all_stack( GC_approx_sp(), GC_stackbottom );
+#	  else
+	    GC_push_all_stack( GC_stackbottom, GC_approx_sp() );
+#	  endif
+#   endif
+    if (GC_push_other_roots != 0) (*GC_push_other_roots)();
+    	/* In the threads case, this also pushes thread stacks.	*/
+}
+
diff --git a/mips_mach_dep.s b/mips_mach_dep.s
new file mode 100644
index 00000000..178224e3
--- /dev/null
+++ b/mips_mach_dep.s
@@ -0,0 +1,26 @@
+# define call_push(x)     move    $4,x;    jal     GC_push_one
+
+    .text
+ # Mark from machine registers that are saved by C compiler
+    .globl  GC_push_regs
+    .ent    GC_push_regs
+GC_push_regs:
+    subu    $sp,8       ## Need to save only return address
+    sw      $31,4($sp)
+    .mask   0x80000000,-4
+    .frame  $sp,8,$31
+    call_push($2)
+    call_push($3)
+    call_push($16)
+    call_push($17)
+    call_push($18)
+    call_push($19)
+    call_push($20)
+    call_push($21)
+    call_push($22)
+    call_push($23)
+    call_push($30)
+    lw      $31,4($sp)
+    addu    $sp,8
+    j       $31
+    .end    GC_push_regs
diff --git a/misc.c b/misc.c
new file mode 100644
index 00000000..f4b5d9ca
--- /dev/null
+++ b/misc.c
@@ -0,0 +1,610 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:04 pm PDT */
+
+#define DEBUG       /* Some run-time consistency checks */
+#undef DEBUG
+#define VERBOSE
+#undef VERBOSE
+
+#include <stdio.h>
+#include <signal.h>
+#define I_HIDE_POINTERS	/* To make GC_call_with_alloc_lock visible */
+#include "gc_priv.h"
+
+# ifdef THREADS
+#   ifdef PCR
+#     include "il/PCR_IL.h"
+      PCR_Th_ML GC_allocate_ml;
+#   else
+#     ifdef SRC_M3
+	/* Critical section counter is defined in the M3 runtime 	*/
+	/* That's all we use.						*/
+#     else
+#	ifdef SOLARIS_THREADS
+	  mutex_t GC_allocate_ml;	/* Implicitly initialized.	*/
+#	else
+	  --> declare allocator lock here
+#	endif
+#     endif
+#   endif
+# endif
+
+GC_FAR struct _GC_arrays GC_arrays = { 0 };
+
+
+bool GC_debugging_started = FALSE;
+	/* defined here so we don't have to load debug_malloc.o */
+
+void (*GC_check_heap)() = (void (*)())0;
+
+ptr_t GC_stackbottom = 0;
+
+bool GC_dont_gc = 0;
+
+bool GC_quiet = 0;
+
+extern signed_word GC_mem_found;
+
+# ifdef MERGE_SIZES
+    /* Set things up so that GC_size_map[i] >= words(i),		*/
+    /* but not too much bigger						*/
+    /* and so that size_map contains relatively few distinct entries 	*/
+    /* This is stolen from Russ Atkinson's Cedar quantization		*/
+    /* alogrithm (but we precompute it).				*/
+
+
+    void GC_init_size_map()
+    {
+	register unsigned i;
+
+	/* Map size 0 to 1.  This avoids problems at lower levels. */
+	  GC_size_map[0] = 1;
+	/* One word objects don't have to be 2 word aligned.	   */
+	  for (i = 1; i < sizeof(word); i++) {
+	      GC_size_map[i] = 1;
+	  }
+	  GC_size_map[sizeof(word)] = ROUNDED_UP_WORDS(sizeof(word));
+	for (i = sizeof(word) + 1; i <= 8 * sizeof(word); i++) {
+#           ifdef ALIGN_DOUBLE
+	      GC_size_map[i] = (ROUNDED_UP_WORDS(i) + 1) & (~1);
+#           else
+	      GC_size_map[i] = ROUNDED_UP_WORDS(i);
+#           endif
+	}
+	for (i = 8*sizeof(word) + 1; i <= 16 * sizeof(word); i++) {
+	      GC_size_map[i] = (ROUNDED_UP_WORDS(i) + 1) & (~1);
+	}
+	/* We leave the rest of the array to be filled in on demand. */
+    }
+    
+    /* Fill in additional entries in GC_size_map, including the ith one */
+    /* We assume the ith entry is currently 0.				*/
+    /* Note that a filled in section of the array ending at n always    */
+    /* has length at least n/4.						*/
+    void GC_extend_size_map(i)
+    word i;
+    {
+        word orig_word_sz = ROUNDED_UP_WORDS(i);
+        word word_sz = orig_word_sz;
+    	register word byte_sz = WORDS_TO_BYTES(word_sz);
+    				/* The size we try to preserve.		*/
+    				/* Close to to i, unless this would	*/
+    				/* introduce too many distinct sizes.	*/
+    	word smaller_than_i = byte_sz - (byte_sz >> 3);
+    	word much_smaller_than_i = byte_sz - (byte_sz >> 2);
+    	register word low_limit;	/* The lowest indexed entry we 	*/
+    					/* initialize.			*/
+    	register word j;
+    	
+    	if (GC_size_map[smaller_than_i] == 0) {
+    	    low_limit = much_smaller_than_i;
+    	    while (GC_size_map[low_limit] != 0) low_limit++;
+    	} else {
+    	    low_limit = smaller_than_i + 1;
+    	    while (GC_size_map[low_limit] != 0) low_limit++;
+    	    word_sz = ROUNDED_UP_WORDS(low_limit);
+    	    word_sz += word_sz >> 3;
+    	    if (word_sz < orig_word_sz) word_sz = orig_word_sz;
+    	}
+#	ifdef ALIGN_DOUBLE
+	    word_sz += 1;
+	    word_sz &= ~1;
+#	endif
+	if (word_sz > MAXOBJSZ) {
+	    word_sz = MAXOBJSZ;
+	}
+    	byte_sz = WORDS_TO_BYTES(word_sz);
+#	ifdef ADD_BYTE_AT_END
+	    /* We need one extra byte; don't fill in GC_size_map[byte_sz] */
+	    byte_sz--;
+#	endif
+
+    	for (j = low_limit; j <= byte_sz; j++) GC_size_map[j] = word_sz;  
+    }
+# endif
+
+
+/*
+ * The following is a gross hack to deal with a problem that can occur
+ * on machines that are sloppy about stack frame sizes, notably SPARC.
+ * Bogus pointers may be written to the stack and not cleared for
+ * a LONG time, because they always fall into holes in stack frames
+ * that are not written.  We partially address this by clearing
+ * sections of the stack whenever we get control.
+ */
+word GC_stack_last_cleared = 0;	/* GC_no when we last did this */
+# define CLEAR_SIZE 213
+# define DEGRADE_RATE 50
+
+word GC_min_sp;		/* Coolest stack pointer value from which we've */
+			/* already cleared the stack.			*/
+			
+# ifdef STACK_GROWS_DOWN
+#   define COOLER_THAN >
+#   define HOTTER_THAN <
+#   define MAKE_COOLER(x,y) if ((word)(x)+(y) > (word)(x)) {(x) += (y);} \
+			    else {(x) = (word)ONES;}
+#   define MAKE_HOTTER(x,y) (x) -= (y)
+# else
+#   define COOLER_THAN <
+#   define HOTTER_THAN >
+#   define MAKE_COOLER(x,y) if ((word)(x)-(y) < (word)(x)) {(x) -= (y);} else {(x) = 0;}
+#   define MAKE_HOTTER(x,y) (x) += (y)
+# endif
+
+word GC_high_water;
+			/* "hottest" stack pointer value we have seen	*/
+			/* recently.  Degrades over time.		*/
+
+word GC_stack_upper_bound()
+{
+    word dummy;
+    
+    return((word)(&dummy));
+}
+
+word GC_words_allocd_at_reset;
+
+#if defined(ASM_CLEAR_CODE) && !defined(THREADS)
+  extern ptr_t GC_clear_stack_inner();
+#endif  
+
+#if !defined(ASM_CLEAR_CODE) && !defined(THREADS)
+/* Clear the stack up to about limit.  Return arg. */
+/*ARGSUSED*/
+ptr_t GC_clear_stack_inner(arg, limit)
+ptr_t arg;
+word limit;
+{
+    word dummy[CLEAR_SIZE];
+    
+    BZERO(dummy, CLEAR_SIZE*sizeof(word));
+    if ((word)(dummy) COOLER_THAN limit) {
+        (void) GC_clear_stack_inner(arg, limit);
+    }
+    /* Make sure the recursive call is not a tail call, and the bzero	*/
+    /* call is not recognized as dead code.				*/
+    GC_noop(dummy);
+    return(arg);
+}
+#endif
+
+
+/* Clear some of the inaccessible part of the stack.  Returns its	*/
+/* argument, so it can be used in a tail call position, hence clearing  */
+/* another frame.							*/
+ptr_t GC_clear_stack(arg)
+ptr_t arg;
+{
+    register word sp = GC_stack_upper_bound();
+    register word limit;
+#   ifdef THREADS
+        word dummy[CLEAR_SIZE];;
+#   endif
+    
+#   define SLOP 400
+	/* Extra bytes we clear every time.  This clears our own	*/
+	/* activation record, and should cause more frequent		*/
+	/* clearing near the cold end of the stack, a good thing.	*/
+#   define GC_SLOP 4000
+	/* We make GC_high_water this much hotter than we really saw   	*/
+	/* saw it, to cover for GC noise etc. above our current frame.	*/
+#   define CLEAR_THRESHOLD 100000
+	/* We restart the clearing process after this many bytes of	*/
+	/* allocation.  Otherwise very heavily recursive programs	*/
+	/* with sparse stacks may result in heaps that grow almost	*/
+	/* without bounds.  As the heap gets larger, collection 	*/
+	/* frequency decreases, thus clearing frequency would decrease, */
+	/* thus more junk remains accessible, thus the heap gets	*/
+	/* larger ...							*/
+# ifdef THREADS
+    BZERO(dummy, CLEAR_SIZE*sizeof(word));
+# else
+    if (GC_gc_no > GC_stack_last_cleared) {
+        /* Start things over, so we clear the entire stack again */
+        if (GC_stack_last_cleared == 0) GC_high_water = (word) GC_stackbottom;
+        GC_min_sp = GC_high_water;
+        GC_stack_last_cleared = GC_gc_no;
+        GC_words_allocd_at_reset = GC_words_allocd;
+    }
+    /* Adjust GC_high_water */
+        MAKE_COOLER(GC_high_water, WORDS_TO_BYTES(DEGRADE_RATE) + GC_SLOP);
+        if (sp HOTTER_THAN GC_high_water) {
+            GC_high_water = sp;
+        }
+        MAKE_HOTTER(GC_high_water, GC_SLOP);
+    limit = GC_min_sp;
+    MAKE_HOTTER(limit, SLOP);
+    if (sp COOLER_THAN limit) {
+        limit &= ~0xf;	/* Make it sufficiently aligned for assembly	*/
+        		/* implementations of GC_clear_stack_inner.	*/
+        GC_min_sp = sp;
+        return(GC_clear_stack_inner(arg, limit));
+    } else if (WORDS_TO_BYTES(GC_words_allocd - GC_words_allocd_at_reset)
+    	       > CLEAR_THRESHOLD) {
+    	/* Restart clearing process, but limit how much clearing we do. */
+    	GC_min_sp = sp;
+    	MAKE_HOTTER(GC_min_sp, CLEAR_THRESHOLD/4);
+    	if (GC_min_sp HOTTER_THAN GC_high_water) GC_min_sp = GC_high_water;
+    	GC_words_allocd_at_reset = GC_words_allocd;
+    }  
+# endif
+  return(arg);
+}
+
+
+/* Return a pointer to the base address of p, given a pointer to a	*/
+/* an address within an object.  Return 0 o.w.				*/
+# ifdef __STDC__
+    extern_ptr_t GC_base(extern_ptr_t p)
+# else
+    extern_ptr_t GC_base(p)
+    extern_ptr_t p;
+# endif
+{
+    register word r;
+    register struct hblk *h;
+    register hdr *candidate_hdr;
+    
+    r = (word)p;
+    h = HBLKPTR(r);
+    candidate_hdr = HDR(r);
+    if (candidate_hdr == 0) return(0);
+    /* If it's a pointer to the middle of a large object, move it	*/
+    /* to the beginning.						*/
+	while (IS_FORWARDING_ADDR_OR_NIL(candidate_hdr)) {
+	   h = h - (int)candidate_hdr;
+	   r = (word)h + HDR_BYTES;
+	   candidate_hdr = HDR(h);
+	}
+    if (candidate_hdr -> hb_map == GC_invalid_map) return(0);
+    /* Make sure r points to the beginning of the object */
+	r &= ~(WORDS_TO_BYTES(1) - 1);
+        {
+	    register int offset =
+	        	(word *)r - (word *)(HBLKPTR(r)) - HDR_WORDS;
+	    register signed_word sz = candidate_hdr -> hb_sz;
+	    register int correction;
+	        
+	    correction = offset % sz;
+	    r -= (WORDS_TO_BYTES(correction));
+	    if (((word *)r + sz) > (word *)(h + 1)
+	        && sz <= BYTES_TO_WORDS(HBLKSIZE) - HDR_WORDS) {
+	        return(0);
+	    }
+	}
+    return((extern_ptr_t)r);
+}
+
+/* Return the size of an object, given a pointer to its base.		*/
+/* (For small obects this also happens to work from interior pointers,	*/
+/* but that shouldn't be relied upon.)					*/
+# ifdef __STDC__
+    size_t GC_size(extern_ptr_t p)
+# else
+    size_t GC_size(p)
+    extern_ptr_t p;
+# endif
+{
+    register int sz;
+    register hdr * hhdr = HDR(p);
+    
+    sz = WORDS_TO_BYTES(hhdr -> hb_sz);
+    if (sz < 0) {
+        return(-sz);
+    } else {
+        return(sz);
+    }
+}
+
+size_t GC_get_heap_size()
+{
+    return ((size_t) GC_heapsize);
+}
+
+bool GC_is_initialized = FALSE;
+
+void GC_init()
+{
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    GC_init_inner();
+    UNLOCK();
+    ENABLE_SIGNALS();
+
+}
+
+#ifdef MSWIN32
+    extern void GC_init_win32();
+#endif
+
+void GC_init_inner()
+{
+    word dummy;
+    
+    if (GC_is_initialized) return;
+    GC_is_initialized = TRUE;
+#   ifdef MSWIN32
+ 	GC_init_win32();
+#   endif
+#   ifdef SOLARIS_THREADS
+	/* We need dirty bits in order to find live stack sections.	*/
+        GC_dirty_init();
+#   endif
+#   if !defined(THREADS) || defined(SOLARIS_THREADS)
+      if (GC_stackbottom == 0) {
+	GC_stackbottom = GC_get_stack_base();
+      }
+#   endif
+    if  (sizeof (ptr_t) != sizeof(word)) {
+        ABORT("sizeof (ptr_t) != sizeof(word)\n");
+    }
+    if  (sizeof (signed_word) != sizeof(word)) {
+        ABORT("sizeof (signed_word) != sizeof(word)\n");
+    }
+    if  (sizeof (struct hblk) != HBLKSIZE) {
+        ABORT("sizeof (struct hblk) != HBLKSIZE\n");
+    }
+#   ifndef THREADS
+#     if defined(STACK_GROWS_UP) && defined(STACK_GROWS_DOWN)
+  	ABORT(
+  	  "Only one of STACK_GROWS_UP and STACK_GROWS_DOWN should be defd\n");
+#     endif
+#     if !defined(STACK_GROWS_UP) && !defined(STACK_GROWS_DOWN)
+  	ABORT(
+  	  "One of STACK_GROWS_UP and STACK_GROWS_DOWN should be defd\n");
+#     endif
+#     ifdef STACK_GROWS_DOWN
+        if ((word)(&dummy) > (word)GC_stackbottom) {
+          GC_err_printf0(
+          	"STACK_GROWS_DOWN is defd, but stack appears to grow up\n");
+          GC_err_printf2("sp = 0x%lx, GC_stackbottom = 0x%lx\n",
+          	   	 (unsigned long) (&dummy),
+          	   	 (unsigned long) GC_stackbottom);
+          ABORT("stack direction 3\n");
+        }
+#     else
+        if ((word)(&dummy) < (word)GC_stackbottom) {
+          GC_err_printf0(
+          	"STACK_GROWS_UP is defd, but stack appears to grow down\n");
+          GC_err_printf2("sp = 0x%lx, GC_stackbottom = 0x%lx\n",
+          	       	 (unsigned long) (&dummy),
+          	     	 (unsigned long) GC_stackbottom);
+          ABORT("stack direction 4");
+        }
+#     endif
+#   endif
+#   if !defined(_AUX_SOURCE) || defined(__GNUC__)
+      if ((word)(-1) < (word)0) {
+    	GC_err_printf0("The type word should be an unsigned integer type\n");
+    	GC_err_printf0("It appears to be signed\n");
+    	ABORT("word");
+      }
+#   endif
+    if ((signed_word)(-1) >= (signed_word)0) {
+    	GC_err_printf0(
+    		"The type signed_word should be a signed integer type\n");
+    	GC_err_printf0("It appears to be unsigned\n");
+    	ABORT("signed_word");
+    }
+    
+    GC_init_headers();
+    /* Add initial guess of root sets */
+      GC_register_data_segments();
+    GC_bl_init();
+    GC_mark_init();
+    if (!GC_expand_hp_inner((word)MINHINCR)) {
+        GC_err_printf0("Can't start up: not enough memory\n");
+        EXIT();
+    }
+    /* Preallocate large object map.  It's otherwise inconvenient to 	*/
+    /* deal with failure.						*/
+      if (!GC_add_map_entry((word)0)) {
+        GC_err_printf0("Can't start up: not enough memory\n");
+        EXIT();
+      }
+    GC_register_displacement_inner(0L);
+#   ifdef MERGE_SIZES
+      GC_init_size_map();
+#   endif
+#   ifdef PCR
+      PCR_IL_Lock(PCR_Bool_false, PCR_allSigsBlocked, PCR_waitForever);
+      PCR_IL_Unlock();
+      GC_pcr_install();
+#   endif
+    /* Get black list set up */
+      GC_gcollect_inner();
+#   ifdef STUBBORN_ALLOC
+    	GC_stubborn_init();
+#   endif
+    /* Convince lint that some things are used */
+#   ifdef LINT
+      {
+          extern char * GC_copyright[];
+          extern GC_read();
+          
+          GC_noop(GC_copyright, GC_find_header, GC_print_block_list,
+                  GC_push_one, GC_call_with_alloc_lock, GC_read,
+                  GC_print_hblkfreelist, GC_dont_expand);
+      }
+#   endif
+}
+
+void GC_enable_incremental()
+{
+    DCL_LOCK_STATE;
+    
+# ifndef FIND_LEAK
+    DISABLE_SIGNALS();
+    LOCK();
+    if (GC_incremental) goto out;
+#   ifndef SOLARIS_THREADS
+        GC_dirty_init();
+#   endif
+    if (!GC_is_initialized) {
+        GC_init_inner();
+    }
+    if (GC_dont_gc) {
+        /* Can't easily do it. */
+        UNLOCK();
+    	ENABLE_SIGNALS();
+    	return;
+    }
+    if (GC_words_allocd > 0) {
+    	/* There may be unmarked reachable objects	*/
+    	GC_gcollect_inner();
+    }   /* else we're OK in assuming everything's	*/
+    	/* clean since nothing can point to an	  	*/
+    	/* unmarked object.			  	*/
+    GC_read_dirty();
+    GC_incremental = TRUE;
+out:
+    UNLOCK();
+    ENABLE_SIGNALS();
+# endif
+}
+
+#if defined(OS2) || defined(MSWIN32)
+    FILE * GC_stdout = NULL;
+    FILE * GC_stderr = NULL;
+#endif
+
+#ifdef MSWIN32
+  void GC_set_files()
+  {
+    if (GC_stdout == NULL) {
+	GC_stdout = fopen("gc.log", "wt");
+    }
+    if (GC_stderr == NULL) {
+	GC_stderr = GC_stdout;
+    }
+  }
+#endif
+
+#ifdef OS2
+  void GC_set_files()
+  {
+      if (GC_stdout == NULL) {
+	GC_stdout = stdout;
+    }
+    if (GC_stderr == NULL) {
+	GC_stderr = stderr;
+    }
+  }
+#endif
+
+/* A version of printf that is unlikely to call malloc, and is thus safer */
+/* to call from the collector in case malloc has been bound to GC_malloc. */
+/* Assumes that no more than 1023 characters are written at once.	  */
+/* Assumes that all arguments have been converted to something of the	  */
+/* same size as long, and that the format conversions expect something	  */
+/* of that size.							  */
+void GC_printf(format, a, b, c, d, e, f)
+char * format;
+long a, b, c, d, e, f;
+{
+    char buf[1025];
+    
+    if (GC_quiet) return;
+    buf[1024] = 0x15;
+    (void) sprintf(buf, format, a, b, c, d, e, f);
+    if (buf[1024] != 0x15) ABORT("GC_printf clobbered stack");
+#   if defined(OS2) || defined(MSWIN32)
+      GC_set_files();
+      /* We hope this doesn't allocate */
+      if (fwrite(buf, 1, strlen(buf), GC_stdout) != strlen(buf))
+          ABORT("write to stdout failed");
+      fflush(GC_stdout);
+#   else
+      if (write(1, buf, strlen(buf)) < 0) ABORT("write to stdout failed");
+#   endif
+}
+
+void GC_err_printf(format, a, b, c, d, e, f)
+char * format;
+long a, b, c, d, e, f;
+{
+    char buf[1025];
+    
+    buf[1024] = 0x15;
+    (void) sprintf(buf, format, a, b, c, d, e, f);
+    if (buf[1024] != 0x15) ABORT("GC_err_printf clobbered stack");
+#   if defined(OS2) || defined(MSWIN32)
+      GC_set_files();
+      /* We hope this doesn't allocate */
+      if (fwrite(buf, 1, strlen(buf), GC_stderr) != strlen(buf))
+          ABORT("write to stderr failed");
+      fflush(GC_stderr);
+#   else
+      if (write(2, buf, strlen(buf)) < 0) ABORT("write to stderr failed");
+#   endif
+}
+
+void GC_err_puts(s)
+char *s;
+{
+#   if defined(OS2) || defined(MSWIN32)
+      GC_set_files();
+      /* We hope this doesn't allocate */
+      if (fwrite(s, 1, strlen(s), GC_stderr) != strlen(s))
+          ABORT("write to stderr failed");
+      fflush(GC_stderr);
+#   else
+      if (write(2, s, strlen(s)) < 0) ABORT("write to stderr failed");
+#   endif
+}
+
+#ifndef PCR
+void GC_abort(msg)
+char * msg;
+{
+    GC_err_printf1("%s\n", msg);
+    (void) abort();
+}
+#endif
+
+# ifdef SRC_M3
+void GC_enable()
+{
+    GC_dont_gc--;
+}
+
+void GC_disable()
+{
+    GC_dont_gc++;
+}
+# endif
diff --git a/new_hblk.c b/new_hblk.c
new file mode 100644
index 00000000..436c38fc
--- /dev/null
+++ b/new_hblk.c
@@ -0,0 +1,239 @@
+/*
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ * This file contains the functions:
+ *	ptr_t GC_build_flXXX(h, old_fl)
+ *	void GC_new_hblk(n)
+ */
+/* Boehm, May 19, 1994 2:09 pm PDT */
+
+
+# include <stdio.h>
+# include "gc_priv.h"
+
+#ifndef SMALL_CONFIG
+/*
+ * Build a free list for size 1 objects inside hblk h.  Set the last link to
+ * be ofl.  Return a pointer tpo the first free list entry.
+ */
+ptr_t GC_build_fl1(h, ofl)
+struct hblk *h;
+ptr_t ofl;
+{
+    register word * p = (word *)h;
+    register word * lim = (word *)(h + 1);
+    
+    p[0] = (word)ofl;
+    p[1] = (word)(p);
+    p[2] = (word)(p+1);
+    p[3] = (word)(p+2);
+    p += 4;
+    for (; p < lim; p += 4) {
+        p[0] = (word)(p-1);
+        p[1] = (word)(p);
+        p[2] = (word)(p+1);
+        p[3] = (word)(p+2);
+    };
+    return((ptr_t)(p-1));
+}
+
+/* The same for size 2 cleared objects */
+ptr_t GC_build_fl_clear2(h, ofl)
+struct hblk *h;
+ptr_t ofl;
+{
+    register word * p = (word *)h;
+    register word * lim = (word *)(h + 1);
+    
+    p[0] = (word)ofl;
+    p[1] = 0;
+    p[2] = (word)p;
+    p[3] = 0;
+    p += 4;
+    for (; p < lim; p += 4) {
+        p[0] = (word)(p-2);
+        p[1] = 0;
+        p[2] = (word)p;
+        p[3] = 0;
+    };
+    return((ptr_t)(p-2));
+}
+
+/* The same for size 3 cleared objects */
+ptr_t GC_build_fl_clear3(h, ofl)
+struct hblk *h;
+ptr_t ofl;
+{
+    register word * p = (word *)h;
+    register word * lim = (word *)(h + 1) - 2;
+    
+    p[0] = (word)ofl;
+    p[1] = 0;
+    p[2] = 0;
+    p += 3;
+    for (; p < lim; p += 3) {
+        p[0] = (word)(p-3);
+        p[1] = 0;
+        p[2] = 0;
+    };
+    return((ptr_t)(p-3));
+}
+
+/* The same for size 4 cleared objects */
+ptr_t GC_build_fl_clear4(h, ofl)
+struct hblk *h;
+ptr_t ofl;
+{
+    register word * p = (word *)h;
+    register word * lim = (word *)(h + 1);
+    
+    p[0] = (word)ofl;
+    p[1] = 0;
+    p[2] = 0;
+    p[3] = 0;
+    p += 4;
+    for (; p < lim; p += 4) {
+        p[0] = (word)(p-4);
+        p[1] = 0;
+        p[2] = 0;
+        p[3] = 0;
+    };
+    return((ptr_t)(p-4));
+}
+
+/* The same for size 2 uncleared objects */
+ptr_t GC_build_fl2(h, ofl)
+struct hblk *h;
+ptr_t ofl;
+{
+    register word * p = (word *)h;
+    register word * lim = (word *)(h + 1);
+    
+    p[0] = (word)ofl;
+    p[2] = (word)p;
+    p += 4;
+    for (; p < lim; p += 4) {
+        p[0] = (word)(p-2);
+        p[2] = (word)p;
+    };
+    return((ptr_t)(p-2));
+}
+
+/* The same for size 4 uncleared objects */
+ptr_t GC_build_fl4(h, ofl)
+struct hblk *h;
+ptr_t ofl;
+{
+    register word * p = (word *)h;
+    register word * lim = (word *)(h + 1);
+    
+    p[0] = (word)ofl;
+    p[4] = (word)p;
+    p += 8;
+    for (; p < lim; p += 8) {
+        p[0] = (word)(p-4);
+        p[4] = (word)p;
+    };
+    return((ptr_t)(p-4));
+}
+
+#endif /* !SMALL_CONFIG */
+
+/*
+ * Allocate a new heapblock for small objects of size n.
+ * Add all of the heapblock's objects to the free list for objects
+ * of that size.  Will fail to do anything if we are out of memory.
+ */
+void GC_new_hblk(sz, kind)
+register word sz;
+int kind;
+{
+    register word *p,
+		  *prev;
+    word *last_object;		/* points to last object in new hblk	*/
+    register struct hblk *h;	/* the new heap block			*/
+    register bool clear = GC_obj_kinds[kind].ok_init;
+
+#   ifdef PRINTSTATS
+	if ((sizeof (struct hblk)) > HBLKSIZE) {
+	    ABORT("HBLK SZ inconsistency");
+        }
+#   endif
+
+  /* Allocate a new heap block */
+    h = GC_allochblk(sz, kind, 0);
+    if (h == 0) return;
+
+  /* Handle small objects sizes more efficiently.  For larger objects 	*/
+  /* the difference is less significant.				*/
+#  ifndef SMALL_CONFIG
+    switch (sz) {
+        case 1: GC_obj_kinds[kind].ok_freelist[1] =
+        	  GC_build_fl1(h, GC_obj_kinds[kind].ok_freelist[1]);
+        	return;
+        case 2: if (clear) {
+         	    GC_obj_kinds[kind].ok_freelist[2] =
+        	      GC_build_fl_clear2(h, GC_obj_kinds[kind].ok_freelist[2]);
+        	} else {
+        	    GC_obj_kinds[kind].ok_freelist[2] =
+        	      GC_build_fl2(h, GC_obj_kinds[kind].ok_freelist[2]);
+        	}
+        	return;
+        case 3: if (clear) {
+         	    GC_obj_kinds[kind].ok_freelist[3] =
+        	      GC_build_fl_clear3(h, GC_obj_kinds[kind].ok_freelist[3]);
+        	    return;
+        	} else {
+        	    /* It's messy to do better than the default here. */
+        	    break;
+        	}
+        case 4: if (clear) {
+        	    GC_obj_kinds[kind].ok_freelist[4] =
+        	      GC_build_fl_clear4(h, GC_obj_kinds[kind].ok_freelist[4]);
+        	} else {
+        	    GC_obj_kinds[kind].ok_freelist[4] =
+        	      GC_build_fl4(h, GC_obj_kinds[kind].ok_freelist[4]);
+        	}
+        	return;
+        default:
+        	break;
+    }
+#  endif /* !SMALL_CONFIG */
+    
+  /* Clear the page if necessary. */
+    if (clear) BZERO(h, HBLKSIZE);
+    
+  /* Add objects to free list */
+    p = &(h -> hb_body[sz]);	/* second object in *h	*/
+    prev = &(h -> hb_body[0]);       	/* One object behind p	*/
+    last_object = (word *)((char *)h + HBLKSIZE);
+    last_object -= sz;
+			    /* Last place for last object to start */
+
+  /* make a list of all objects in *h with head as last object */
+    while (p <= last_object) {
+      /* current object's link points to last object */
+        obj_link(p) = (ptr_t)prev;
+	prev = p;
+	p += sz;
+    }
+    p -= sz;			/* p now points to last object */
+
+  /*
+   * put p (which is now head of list of objects in *h) as first
+   * pointer in the appropriate free list for this size.
+   */
+      obj_link(h -> hb_body) = GC_obj_kinds[kind].ok_freelist[sz];
+      GC_obj_kinds[kind].ok_freelist[sz] = ((ptr_t)p);
+}
+
diff --git a/obj_map.c b/obj_map.c
new file mode 100644
index 00000000..e728c37c
--- /dev/null
+++ b/obj_map.c
@@ -0,0 +1,137 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991, 1992 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 1:59 pm PDT */
+  
+/* Routines for maintaining maps describing heap block
+ * layouts for various object sizes.  Allows fast pointer validity checks
+ * and fast location of object start locations on machines (such as SPARC)
+ * with slow division.
+ */
+ 
+# include "gc_priv.h"
+
+char * GC_invalid_map = 0;
+
+/* Invalidate the object map associated with a block.	Free blocks	*/
+/* are identified by invalid maps.					*/
+void GC_invalidate_map(hhdr)
+hdr *hhdr;
+{
+    register int displ;
+    
+    if (GC_invalid_map == 0) {
+        GC_invalid_map = GC_scratch_alloc(MAP_SIZE);
+        if (GC_invalid_map == 0) {
+            GC_err_printf0(
+            	"Cant initialize GC_invalid_map: insufficient memory\n");
+            EXIT();
+        }
+        for (displ = 0; displ < HBLKSIZE; displ++) {
+            MAP_ENTRY(GC_invalid_map, displ) = OBJ_INVALID;
+        }
+    }
+    hhdr -> hb_map = GC_invalid_map;
+}
+
+/* Consider pointers that are offset bytes displaced from the beginning */
+/* of an object to be valid.                                            */
+void GC_register_displacement(offset) 
+word offset;
+{
+# ifndef ALL_INTERIOR_POINTERS
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    GC_register_displacement_inner(offset);
+    UNLOCK();
+    ENABLE_SIGNALS();
+# endif
+}
+
+void GC_register_displacement_inner(offset) 
+word offset;
+{
+# ifndef ALL_INTERIOR_POINTERS
+    register unsigned i;
+    
+    if (offset > MAX_OFFSET) {
+        ABORT("Bad argument to GC_register_displacement");
+    }
+    if (!GC_valid_offsets[offset]) {
+      GC_valid_offsets[offset] = TRUE;
+      GC_modws_valid_offsets[offset % sizeof(word)] = TRUE;
+      for (i = 0; i <= MAXOBJSZ; i++) {
+          if (GC_obj_map[i] != 0) {
+             if (i == 0) {
+               GC_obj_map[i][offset + HDR_BYTES] = (char)BYTES_TO_WORDS(offset);
+             } else {
+               register unsigned j;
+               register unsigned lb = WORDS_TO_BYTES(i);
+               
+               if (offset < lb) {
+                 for (j = offset + HDR_BYTES; j < HBLKSIZE; j += lb) {
+                   GC_obj_map[i][j] = (char)BYTES_TO_WORDS(offset);
+                 }
+               }
+             }
+          }
+      }
+    }
+# endif
+}
+
+
+/* Add a heap block map for objects of size sz to obj_map.	*/
+/* Return FALSE on failure.					*/
+bool GC_add_map_entry(sz)
+word sz;
+{
+    register unsigned obj_start;
+    register unsigned displ;
+    register char * new_map;
+    
+    if (sz > MAXOBJSZ) sz = 0;
+    if (GC_obj_map[sz] != 0) {
+        return(TRUE);
+    }
+    new_map = GC_scratch_alloc(MAP_SIZE);
+    if (new_map == 0) return(FALSE);
+#   ifdef PRINTSTATS
+        GC_printf1("Adding block map for size %lu\n", (unsigned long)sz);
+#   endif
+    for (displ = 0; displ < HBLKSIZE; displ++) {
+        MAP_ENTRY(new_map,displ) = OBJ_INVALID;
+    }
+    if (sz == 0) {
+        for(displ = 0; displ <= MAX_OFFSET; displ++) {
+            if (OFFSET_VALID(displ)) {
+                MAP_ENTRY(new_map,displ+HDR_BYTES) = BYTES_TO_WORDS(displ);
+            }
+        }
+    } else {
+        for (obj_start = HDR_BYTES;
+             obj_start + WORDS_TO_BYTES(sz) <= HBLKSIZE;
+             obj_start += WORDS_TO_BYTES(sz)) {
+             for (displ = 0; displ < WORDS_TO_BYTES(sz); displ++) {
+                 if (OFFSET_VALID(displ)) {
+                     MAP_ENTRY(new_map, obj_start + displ) =
+                     				BYTES_TO_WORDS(displ);
+                 }
+             }
+        }
+    }
+    GC_obj_map[sz] = new_map;
+    return(TRUE);
+}
diff --git a/os_dep.c b/os_dep.c
new file mode 100644
index 00000000..89932bd8
--- /dev/null
+++ b/os_dep.c
@@ -0,0 +1,1645 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:10 pm PDT */
+# if !defined(OS2) && !defined(PCR) && !defined(AMIGA)
+#   include <sys/types.h>
+# endif
+# include "gc_priv.h"
+# include <stdio.h>
+# include <signal.h>
+
+/* Blatantly OS dependent routines, except for those that are related 	*/
+/* dynamic loading.							*/
+
+#ifdef FREEBSD
+#  include <machine/trap.h>
+#endif
+
+#ifdef AMIGA
+# include <proto/exec.h>
+# include <proto/dos.h>
+# include <dos/dosextens.h>
+# include <workbench/startup.h>
+#endif
+
+#ifdef MSWIN32
+# define WIN32_LEAN_AND_MEAN
+# define NOSERVICE
+# include <windows.h>
+#endif
+
+#ifdef IRIX5
+# include <sys/uio.h>
+#endif
+
+#ifdef PCR
+# include "il/PCR_IL.h"
+# include "th/PCR_ThCtl.h"
+# include "mm/PCR_MM.h"
+#endif
+
+# ifdef OS2
+
+# include <stddef.h>
+
+# ifndef __IBMC__ /* e.g. EMX */
+
+struct exe_hdr {
+    unsigned short      magic_number;
+    unsigned short      padding[29];
+    long                new_exe_offset;
+};
+
+#define E_MAGIC(x)      (x).magic_number
+#define EMAGIC          0x5A4D  
+#define E_LFANEW(x)     (x).new_exe_offset
+
+struct e32_exe {
+    unsigned char       magic_number[2]; 
+    unsigned char       byte_order; 
+    unsigned char       word_order; 
+    unsigned long       exe_format_level;
+    unsigned short      cpu;       
+    unsigned short      os;
+    unsigned long       padding1[13];
+    unsigned long       object_table_offset;
+    unsigned long       object_count;    
+    unsigned long       padding2[31];
+};
+
+#define E32_MAGIC1(x)   (x).magic_number[0]
+#define E32MAGIC1       'L'
+#define E32_MAGIC2(x)   (x).magic_number[1]
+#define E32MAGIC2       'X'
+#define E32_BORDER(x)   (x).byte_order
+#define E32LEBO         0
+#define E32_WORDER(x)   (x).word_order
+#define E32LEWO         0
+#define E32_CPU(x)      (x).cpu
+#define E32CPU286       1
+#define E32_OBJTAB(x)   (x).object_table_offset
+#define E32_OBJCNT(x)   (x).object_count
+
+struct o32_obj {
+    unsigned long       size;  
+    unsigned long       base;
+    unsigned long       flags;  
+    unsigned long       pagemap;
+    unsigned long       mapsize; 
+    unsigned long       reserved;
+};
+
+#define O32_FLAGS(x)    (x).flags
+#define OBJREAD         0x0001L
+#define OBJWRITE        0x0002L
+#define OBJINVALID      0x0080L
+#define O32_SIZE(x)     (x).size
+#define O32_BASE(x)     (x).base
+
+# else  /* IBM's compiler */
+
+# define INCL_DOSEXCEPTIONS
+# define INCL_DOSPROCESS
+# define INCL_DOSERRORS
+# define INCL_DOSMODULEMGR
+# define INCL_DOSMEMMGR
+# include <os2.h>
+
+/* A kludge to get around what appears to be a header file bug */
+# ifndef WORD
+#   define WORD unsigned short
+# endif
+# ifndef DWORD
+#   define DWORD unsigned long
+# endif
+
+# define EXE386 1
+# include <newexe.h>
+# include <exe386.h>
+
+# endif  /* __IBMC__ */
+
+/* Disable and enable signals during nontrivial allocations	*/
+
+void GC_disable_signals(void)
+{
+    ULONG nest;
+    
+    DosEnterMustComplete(&nest);
+    if (nest != 1) ABORT("nested GC_disable_signals");
+}
+
+void GC_enable_signals(void)
+{
+    ULONG nest;
+    
+    DosExitMustComplete(&nest);
+    if (nest != 0) ABORT("GC_enable_signals");
+}
+
+
+# else
+
+#  if !defined(PCR) && !defined(AMIGA) && !defined(MSWIN32)
+
+#   ifdef sigmask
+	/* Use the traditional BSD interface */
+#	define SIGSET_T int
+#	define SIG_DEL(set, signal) (set) &= ~(sigmask(signal))
+#	define SIG_FILL(set)  (set) = 0x7fffffff
+    	  /* Setting the leading bit appears to provoke a bug in some	*/
+    	  /* longjmp implementations.  Most systems appear not to have	*/
+    	  /* a signal 32.						*/
+#	define SIGSETMASK(old, new) (old) = sigsetmask(new)
+#   else
+	/* Use POSIX/SYSV interface	*/
+#	define SIGSET_T sigset_t
+#	define SIG_DEL(set, signal) sigdelset(&(set), (signal))
+#	define SIG_FILL(set) sigfillset(&set)
+#	define SIGSETMASK(old, new) sigprocmask(SIG_SETMASK, &(new), &(old))
+#   endif
+
+static bool mask_initialized = FALSE;
+
+static SIGSET_T new_mask;
+
+static SIGSET_T old_mask;
+
+static SIGSET_T dummy;
+
+#if defined(PRINTSTATS) && !defined(THREADS)
+# define CHECK_SIGNALS
+  int GC_sig_disabled = 0;
+#endif
+
+void GC_disable_signals()
+{
+    if (!mask_initialized) {
+    	SIG_FILL(new_mask);
+
+	SIG_DEL(new_mask, SIGSEGV);
+	SIG_DEL(new_mask, SIGILL);
+	SIG_DEL(new_mask, SIGQUIT);
+#	ifdef SIGBUS
+	    SIG_DEL(new_mask, SIGBUS);
+#	endif
+#	ifdef SIGIOT
+	    SIG_DEL(new_mask, SIGIOT);
+#	endif
+#	ifdef SIGEMT
+	    SIG_DEL(new_mask, SIGEMT);
+#	endif
+#	ifdef SIGTRAP
+	    SIG_DEL(new_mask, SIGTRAP);
+#	endif 
+	mask_initialized = TRUE;
+    }
+#   ifdef CHECK_SIGNALS
+	if (GC_sig_disabled != 0) ABORT("Nested disables");
+	GC_sig_disabled++;
+#   endif
+    SIGSETMASK(old_mask,new_mask);
+}
+
+void GC_enable_signals()
+{
+#   ifdef CHECK_SIGNALS
+	if (GC_sig_disabled != 1) ABORT("Unmatched enable");
+	GC_sig_disabled--;
+#   endif
+    SIGSETMASK(dummy,old_mask);
+}
+
+#  endif  /* !PCR */
+
+# endif /*!OS/2 */
+
+/*
+ * Find the base of the stack.
+ * Used only in single-threaded environment.
+ * With threads, GC_mark_roots needs to know how to do this.
+ * Called with allocator lock held.
+ */
+# ifdef MSWIN32
+
+/* Get the page size.	*/
+word GC_page_size = 0;
+
+word GC_get_page_size()
+{
+    SYSTEM_INFO sysinfo;
+    
+    if (GC_page_size == 0) {
+        GetSystemInfo(&sysinfo);
+        GC_page_size = sysinfo.dwPageSize;
+    }
+    return(GC_page_size);
+}
+
+# define is_writable(prot) ((prot) == PAGE_READWRITE \
+			    || (prot) == PAGE_WRITECOPY \
+			    || (prot) == PAGE_EXECUTE_READWRITE \
+			    || (prot) == PAGE_EXECUTE_WRITECOPY)
+/* Return the number of bytes that are writable starting at p.	*/
+/* The pointer p is assumed to be page aligned.			*/
+/* If base is not 0, *base becomes the beginning of the 	*/
+/* allocation region containing p.				*/
+word GC_get_writable_length(ptr_t p, ptr_t *base)
+{
+    MEMORY_BASIC_INFORMATION buf;
+    word result;
+    word protect;
+    
+    result = VirtualQuery(p, &buf, sizeof(buf));
+    if (result != sizeof(buf)) ABORT("Weird VirtualQuery result");
+    if (base != 0) *base = (ptr_t)(buf.AllocationBase);
+    protect = (buf.Protect & ~(PAGE_GUARD | PAGE_NOCACHE));
+    if (!is_writable(protect)) {
+        return(0);
+    }
+    if (buf.State != MEM_COMMIT) return(0);
+    return(buf.RegionSize);
+}
+
+ptr_t GC_get_stack_base()
+{
+    int dummy;
+    ptr_t sp = (ptr_t)(&dummy);
+    ptr_t trunc_sp = (ptr_t)((word)sp & ~(GC_get_page_size() - 1));
+    word size = GC_get_writable_length(trunc_sp, 0);
+   
+    return(trunc_sp + size);
+}
+
+
+# else
+
+# ifdef OS2
+
+ptr_t GC_get_stack_base()
+{
+    PTIB ptib;
+    PPIB ppib;
+    
+    if (DosGetInfoBlocks(&ptib, &ppib) != NO_ERROR) {
+    	GC_err_printf0("DosGetInfoBlocks failed\n");
+    	ABORT("DosGetInfoBlocks failed\n");
+    }
+    return((ptr_t)(ptib -> tib_pstacklimit));
+}
+
+# else
+
+# ifdef AMIGA
+
+ptr_t GC_get_stack_base()
+{
+    extern struct WBStartup *_WBenchMsg;
+    extern long __base;
+    extern long __stack;
+    struct Task *task;
+    struct Process *proc;
+    struct CommandLineInterface *cli;
+    long size;
+
+    if ((task = FindTask(0)) == 0) {
+	GC_err_puts("Cannot find own task structure\n");
+	ABORT("task missing");
+    }
+    proc = (struct Process *)task;
+    cli = BADDR(proc->pr_CLI);
+
+    if (_WBenchMsg != 0 || cli == 0) {
+	size = (char *)task->tc_SPUpper - (char *)task->tc_SPLower;
+    } else {
+	size = cli->cli_DefaultStack * 4;
+    }
+    return (ptr_t)(__base + GC_max(size, __stack));
+}
+
+# else
+
+# if !defined(THREADS) && !defined(STACKBOTTOM) && defined(HEURISTIC2)
+#   define NEED_FIND_LIMIT
+# endif
+
+# if defined(SUNOS4) & defined(DYNAMIC_LOADING)
+#   define NEED_FIND_LIMIT
+# endif
+
+# ifdef NEED_FIND_LIMIT
+  /* Some tools to implement HEURISTIC2	*/
+#   define MIN_PAGE_SIZE 256	/* Smallest conceivable page size, bytes */
+#   include <setjmp.h>
+    /* static */ jmp_buf GC_jmp_buf;
+    
+    /*ARGSUSED*/
+    void GC_fault_handler(sig)
+    int sig;
+    {
+        longjmp(GC_jmp_buf, 1);
+    }
+
+#   ifdef __STDC__
+	typedef void (*handler)(int);
+#   else
+	typedef void (*handler)();
+#   endif
+
+    /* Return the first nonaddressible location > p (up) or 	*/
+    /* the smallest location q s.t. [q,p] is addressible (!up).	*/
+    ptr_t GC_find_limit(p, up)
+    ptr_t p;
+    bool up;
+    {
+        static VOLATILE ptr_t result;
+    		/* Needs to be static, since otherwise it may not be	*/
+    		/* preserved across the longjmp.  Can safely be 	*/
+    		/* static since it's only called once, with the		*/
+    		/* allocation lock held.				*/
+
+        static handler old_segv_handler, old_bus_handler;
+      		/* See above for static declaration.			*/
+
+    	old_segv_handler = signal(SIGSEGV, GC_fault_handler);
+#	ifdef SIGBUS
+	   old_bus_handler = signal(SIGBUS, GC_fault_handler);
+#	endif
+	if (setjmp(GC_jmp_buf) == 0) {
+	    result = (ptr_t)(((word)(p))
+			      & ~(MIN_PAGE_SIZE-1));
+	    for (;;) {
+ 	        if (up) {
+		    result += MIN_PAGE_SIZE;
+ 	        } else {
+		    result -= MIN_PAGE_SIZE;
+ 	        }
+		GC_noop(*result);
+	    }
+	}
+	(void) signal(SIGSEGV, old_segv_handler);
+#	ifdef SIGBUS
+	    (void) signal(SIGBUS, old_bus_handler);
+#	endif
+ 	if (!up) {
+	    result += MIN_PAGE_SIZE;
+ 	}
+	return(result);
+    }
+# endif
+
+
+ptr_t GC_get_stack_base()
+{
+    word dummy;
+    ptr_t result;
+
+#   define STACKBOTTOM_ALIGNMENT_M1 0xffffff
+
+#   ifdef STACKBOTTOM
+	return(STACKBOTTOM);
+#   else
+#	ifdef HEURISTIC1
+#	   ifdef STACK_GROWS_DOWN
+	     result = (ptr_t)((((word)(&dummy))
+	     		       + STACKBOTTOM_ALIGNMENT_M1)
+			      & ~STACKBOTTOM_ALIGNMENT_M1);
+#	   else
+	     result = (ptr_t)(((word)(&dummy))
+			      & ~STACKBOTTOM_ALIGNMENT_M1);
+#	   endif
+#	endif /* HEURISTIC1 */
+#	ifdef HEURISTIC2
+#	    ifdef STACK_GROWS_DOWN
+		result = GC_find_limit((ptr_t)(&dummy), TRUE);
+#	    else
+		result = GC_find_limit((ptr_t)(&dummy), FALSE);
+#	    endif
+#	endif /* HEURISTIC2 */
+    	return(result);
+#   endif /* STACKBOTTOM */
+}
+
+# endif /* ! AMIGA */
+# endif /* ! OS2 */
+# endif /* ! MSWIN32 */
+
+/*
+ * Register static data segment(s) as roots.
+ * If more data segments are added later then they need to be registered
+ * add that point (as we do with SunOS dynamic loading),
+ * or GC_mark_roots needs to check for them (as we do with PCR).
+ * Called with allocator lock held.
+ */
+
+# ifdef OS2
+
+void GC_register_data_segments()
+{
+    PTIB ptib;
+    PPIB ppib;
+    HMODULE module_handle;
+#   define PBUFSIZ 512
+    UCHAR path[PBUFSIZ];
+    FILE * myexefile;
+    struct exe_hdr hdrdos;	/* MSDOS header.	*/
+    struct e32_exe hdr386;	/* Real header for my executable */
+    struct o32_obj seg;	/* Currrent segment */
+    int nsegs;
+    
+    
+    if (DosGetInfoBlocks(&ptib, &ppib) != NO_ERROR) {
+    	GC_err_printf0("DosGetInfoBlocks failed\n");
+    	ABORT("DosGetInfoBlocks failed\n");
+    }
+    module_handle = ppib -> pib_hmte;
+    if (DosQueryModuleName(module_handle, PBUFSIZ, path) != NO_ERROR) {
+    	GC_err_printf0("DosQueryModuleName failed\n");
+    	ABORT("DosGetInfoBlocks failed\n");
+    }
+    myexefile = fopen(path, "rb");
+    if (myexefile == 0) {
+        GC_err_puts("Couldn't open executable ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Failed to open executable\n");
+    }
+    if (fread((char *)(&hdrdos), 1, sizeof hdrdos, myexefile) < sizeof hdrdos) {
+        GC_err_puts("Couldn't read MSDOS header from ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Couldn't read MSDOS header");
+    }
+    if (E_MAGIC(hdrdos) != EMAGIC) {
+        GC_err_puts("Executable has wrong DOS magic number: ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Bad DOS magic number");
+    }
+    if (fseek(myexefile, E_LFANEW(hdrdos), SEEK_SET) != 0) {
+        GC_err_puts("Seek to new header failed in ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Bad DOS magic number");
+    }
+    if (fread((char *)(&hdr386), 1, sizeof hdr386, myexefile) < sizeof hdr386) {
+        GC_err_puts("Couldn't read MSDOS header from ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Couldn't read OS/2 header");
+    }
+    if (E32_MAGIC1(hdr386) != E32MAGIC1 || E32_MAGIC2(hdr386) != E32MAGIC2) {
+        GC_err_puts("Executable has wrong OS/2 magic number:");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Bad OS/2 magic number");
+    }
+    if ( E32_BORDER(hdr386) != E32LEBO || E32_WORDER(hdr386) != E32LEWO) {
+        GC_err_puts("Executable %s has wrong byte order: ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Bad byte order");
+    }
+    if ( E32_CPU(hdr386) == E32CPU286) {
+        GC_err_puts("GC can't handle 80286 executables: ");
+        GC_err_puts(path); GC_err_puts("\n");
+        EXIT();
+    }
+    if (fseek(myexefile, E_LFANEW(hdrdos) + E32_OBJTAB(hdr386),
+    	      SEEK_SET) != 0) {
+        GC_err_puts("Seek to object table failed: ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Seek to object table failed");
+    }
+    for (nsegs = E32_OBJCNT(hdr386); nsegs > 0; nsegs--) {
+      int flags;
+      if (fread((char *)(&seg), 1, sizeof seg, myexefile) < sizeof seg) {
+        GC_err_puts("Couldn't read obj table entry from ");
+        GC_err_puts(path); GC_err_puts("\n");
+        ABORT("Couldn't read obj table entry");
+      }
+      flags = O32_FLAGS(seg);
+      if (!(flags & OBJWRITE)) continue;
+      if (!(flags & OBJREAD)) continue;
+      if (flags & OBJINVALID) {
+          GC_err_printf0("Object with invalid pages?\n");
+          continue;
+      } 
+      GC_add_roots_inner(O32_BASE(seg), O32_BASE(seg)+O32_SIZE(seg));
+    }
+}
+
+# else
+
+# ifdef MSWIN32
+  /* Unfortunately, we have to handle win32s very differently from NT, 	*/
+  /* Since VirtualQuery has very different semantics.  In particular,	*/
+  /* under win32s a VirtualQuery call on an unmapped page returns an	*/
+  /* invalid result.  Under GC_register_data_segments is a noop and	*/
+  /* all real work is done by GC_register_dynamic_libraries.  Under	*/
+  /* win32s, we cannot find the data segments associated with dll's.	*/
+  /* We rgister the main data segment here.				*/
+  bool GC_win32s = FALSE;	/* We're running under win32s.	*/
+  
+  void GC_init_win32()
+  {
+      if (GetVersion() & 0x80000000) GC_win32s = TRUE;
+  }
+  
+  /* Return the smallest address a such that VirtualQuery		*/
+  /* returns correct results for all addresses between a and start.	*/
+  /* Assumes VirtualQuery returns correct information for start.	*/
+  ptr_t GC_least_described_address(ptr_t start)
+  {  
+    MEMORY_BASIC_INFORMATION buf;
+    SYSTEM_INFO sysinfo;
+    DWORD result;
+    LPVOID limit;
+    ptr_t p;
+    LPVOID q;
+    
+    GetSystemInfo(&sysinfo);
+    limit = sysinfo.lpMinimumApplicationAddress;
+    p = (ptr_t)((word)start & ~(GC_get_page_size() - 1));
+    for (;;) {
+    	q = (LPVOID)(p - GC_get_page_size());
+    	if ((ptr_t)q > (ptr_t)p /* underflow */ || q < limit) break;
+    	result = VirtualQuery(q, &buf, sizeof(buf));
+    	if (result != sizeof(buf)) break;
+    	p = (ptr_t)(buf.AllocationBase);
+    }
+    return(p);
+  }
+  
+  /* Is p the start of either the malloc heap, or of one of our */
+  /* heap sections?						*/
+  bool GC_is_heap_base (ptr_t p)
+  {
+     static ptr_t malloc_heap_pointer = 0;
+     register unsigned i;
+     register DWORD result;
+     
+     if (malloc_heap_pointer = 0) {
+        MEMORY_BASIC_INFORMATION buf;
+        result = VirtualQuery(malloc(1), &buf, sizeof(buf));
+        if (result != sizeof(buf)) {
+            ABORT("Weird VirtualQuery result");
+        }
+        malloc_heap_pointer = (ptr_t)(buf.AllocationBase);
+     }
+     if (p == malloc_heap_pointer) return(TRUE);
+     for (i = 0; i < GC_n_heap_bases; i++) {
+         if (GC_heap_bases[i] == p) return(TRUE);
+     }
+     return(FALSE);
+  }
+  
+  void GC_register_root_section(ptr_t static_root)
+  {
+      MEMORY_BASIC_INFORMATION buf;
+      SYSTEM_INFO sysinfo;
+      DWORD result;
+      DWORD protect;
+      LPVOID p;
+      char * base;
+      char * limit, * new_limit;
+    
+      if (!GC_win32s) return;
+      p = base = limit = GC_least_described_address(static_root);
+      GetSystemInfo(&sysinfo);
+      while (p < sysinfo.lpMaximumApplicationAddress) {
+        result = VirtualQuery(p, &buf, sizeof(buf));
+        if (result != sizeof(buf) || GC_is_heap_base(buf.AllocationBase)) break;
+        new_limit = (char *)p + buf.RegionSize;
+        protect = buf.Protect;
+        if (buf.State == MEM_COMMIT
+            && is_writable(protect)) {
+            if ((char *)p == limit) {
+                limit = new_limit;
+            } else {
+                if (base != limit) GC_add_roots_inner(base, limit);
+                base = p;
+                limit = new_limit;
+            }
+        }
+        if (p > (LPVOID)new_limit /* overflow */) break;
+        p = (LPVOID)new_limit;
+      }
+      if (base != limit) GC_add_roots_inner(base, limit);
+  }
+  
+  void GC_register_data_segments()
+  {
+      static char dummy;
+      
+      GC_register_root_section((ptr_t)(&dummy));
+  }
+# else
+# ifdef AMIGA
+
+  void GC_register_data_segments()
+  {
+    extern struct WBStartup *_WBenchMsg;
+    struct Process	*proc;
+    struct CommandLineInterface *cli;
+    BPTR myseglist;
+    ULONG *data;
+
+    if ( _WBenchMsg != 0 ) {
+	if ((myseglist = _WBenchMsg->sm_Segment) == 0) {
+	    GC_err_puts("No seglist from workbench\n");
+	    return;
+	}
+    } else {
+	if ((proc = (struct Process *)FindTask(0)) == 0) {
+	    GC_err_puts("Cannot find process structure\n");
+	    return;
+	}
+	if ((cli = BADDR(proc->pr_CLI)) == 0) {
+	    GC_err_puts("No CLI\n");
+	    return;
+	}
+	if ((myseglist = cli->cli_Module) == 0) {
+	    GC_err_puts("No seglist from CLI\n");
+	    return;
+	}
+    }
+
+    for (data = (ULONG *)BADDR(myseglist); data != 0;
+         data = (ULONG *)BADDR(data[0])) {
+	GC_add_roots_inner((char *)&data[1], ((char *)&data[1]) + data[-1]);
+    }
+  }
+
+
+# else
+
+void GC_register_data_segments()
+{
+#   ifndef NEXT
+        extern int end;
+#   endif
+ 
+#   if !defined(PCR) && !defined(SRC_M3) && !defined(NEXT)
+      GC_add_roots_inner(DATASTART, (char *)(&end));
+#   endif
+#   if !defined(PCR) && defined(NEXT)
+      GC_add_roots_inner(DATASTART, (char *) get_end());
+#   endif
+    /* Dynamic libraries are added at every collection, since they may  */
+    /* change.								*/
+}
+
+# endif  /* ! AMIGA */
+# endif  /* ! MSWIN32 */
+# endif  /* ! OS2 */
+
+/*
+ * Auxiliary routines for obtaining memory from OS.
+ */
+ 
+# if !defined(OS2) && !defined(PCR) && !defined(AMIGA) && !defined(MSWIN32)
+
+extern caddr_t sbrk();
+# ifdef __STDC__
+#   define SBRK_ARG_T size_t
+# else
+#   define SBRK_ARG_T int
+# endif
+
+# ifdef RS6000
+/* The compiler seems to generate speculative reads one past the end of	*/
+/* an allocated object.  Hence we need to make sure that the page 	*/
+/* following the last heap page is also mapped.				*/
+ptr_t GC_unix_get_mem(bytes)
+word bytes;
+{
+    caddr_t cur_brk = sbrk(0);
+    caddr_t result;
+    SBRK_ARG_T lsbs = (word)cur_brk & (HBLKSIZE-1);
+    static caddr_t my_brk_val = 0;
+    
+    if (lsbs != 0) {
+        if(sbrk(HBLKSIZE - lsbs) == (caddr_t)(-1)) return(0);
+    }
+    if (cur_brk == my_brk_val) {
+    	/* Use the extra block we allocated last time. */
+        result = (ptr_t)sbrk((SBRK_ARG_T)bytes);
+        if (result == (caddr_t)(-1)) return(0);
+        result -= HBLKSIZE;
+    } else {
+        result = (ptr_t)sbrk(HBLKSIZE + (SBRK_ARG_T)bytes);
+        if (result == (caddr_t)(-1)) return(0);
+    }
+    my_brk_val = result + bytes + HBLKSIZE;	/* Always HBLKSIZE aligned */
+    return((ptr_t)result);
+}
+
+#else
+ptr_t GC_unix_get_mem(bytes)
+word bytes;
+{
+    caddr_t cur_brk = sbrk(0);
+    caddr_t result;
+    SBRK_ARG_T lsbs = (word)cur_brk & (HBLKSIZE-1);
+    
+    if (lsbs != 0) {
+        if(sbrk(HBLKSIZE - lsbs) == (caddr_t)(-1)) return(0);
+    }
+    result = sbrk((SBRK_ARG_T)bytes);
+    if (result == (caddr_t)(-1)) return(0);
+    return((ptr_t)result);
+}
+#endif
+
+# endif
+
+# ifdef __OS2__
+
+void * os2_alloc(size_t bytes)
+{
+    void * result;
+
+    if (DosAllocMem(&result, bytes, PAG_EXECUTE | PAG_READ |
+    				    PAG_WRITE | PAG_COMMIT)
+		    != NO_ERROR) {
+	return(0);
+    }
+    if (result == 0) return(os2_alloc(bytes));
+    return(result);
+}
+
+# endif /* OS2 */
+
+
+# ifdef MSWIN32
+word GC_n_heap_bases = 0;
+
+ptr_t GC_win32_get_mem(bytes)
+word bytes;
+{
+    ptr_t result;
+    
+    if (GC_win32s) {
+    	/* VirtualAlloc doesn't like PAGE_EXECUTE_READWRITE.	*/
+    	/* There are also unconfirmed rumors of other		*/
+    	/* problems, so we dodge the issue.			*/
+        result = (ptr_t) GlobalAlloc(0, bytes + HBLKSIZE);
+        result = (ptr_t)(((word)result + HBLKSIZE) & ~(HBLKSIZE-1));
+    } else {
+        result = (ptr_t) VirtualAlloc(NULL, bytes,
+    				      MEM_COMMIT | MEM_RESERVE,
+    				      PAGE_EXECUTE_READWRITE);
+    }
+    if (HBLKDISPL(result) != 0) ABORT("Bad VirtualAlloc result");
+    	/* If I read the documentation correctly, this can	*/
+    	/* only happen if HBLKSIZE > 64k or not a power of 2.	*/
+    if (GC_n_heap_bases >= MAX_HEAP_SECTS) ABORT("Too many heap sections");
+    GC_heap_bases[GC_n_heap_bases++] = result;
+    return(result);			  
+}
+
+# endif
+
+/* Routine for pushing any additional roots.  In THREADS 	*/
+/* environment, this is also responsible for marking from 	*/
+/* thread stacks.  In the SRC_M3 case, it also handles		*/
+/* global variables.						*/
+#ifndef THREADS
+void (*GC_push_other_roots)() = 0;
+#else /* THREADS */
+
+# ifdef PCR
+PCR_ERes GC_push_thread_stack(PCR_Th_T *t, PCR_Any dummy)
+{
+    struct PCR_ThCtl_TInfoRep info;
+    PCR_ERes result;
+    
+    info.ti_stkLow = info.ti_stkHi = 0;
+    result = PCR_ThCtl_GetInfo(t, &info);
+    GC_push_all_stack((ptr_t)(info.ti_stkLow), (ptr_t)(info.ti_stkHi));
+    return(result);
+}
+
+/* Push the contents of an old object. We treat this as stack	*/
+/* data only becasue that makes it robust against mark stack	*/
+/* overflow.							*/
+PCR_ERes GC_push_old_obj(void *p, size_t size, PCR_Any data)
+{
+    GC_push_all_stack((ptr_t)p, (ptr_t)p + size);
+    return(PCR_ERes_okay);
+}
+
+
+void GC_default_push_other_roots()
+{
+    /* Traverse data allocated by previous memory managers.		*/
+	{
+	  extern struct PCR_MM_ProcsRep * GC_old_allocator;
+	  
+	  if ((*(GC_old_allocator->mmp_enumerate))(PCR_Bool_false,
+	  					   GC_push_old_obj, 0)
+	      != PCR_ERes_okay) {
+	      ABORT("Old object enumeration failed");
+	  }
+	}
+    /* Traverse all thread stacks. */
+	if (PCR_ERes_IsErr(
+                PCR_ThCtl_ApplyToAllOtherThreads(GC_push_thread_stack,0))
+              || PCR_ERes_IsErr(GC_push_thread_stack(PCR_Th_CurrThread(), 0))) {
+              ABORT("Thread stack marking failed\n");
+	}
+}
+
+# endif /* PCR */
+
+# ifdef SRC_M3
+
+# ifdef ALL_INTERIOR_POINTERS
+    --> misconfigured
+# endif
+
+
+extern void ThreadF__ProcessStacks();
+
+void GC_push_thread_stack(start, stop)
+word start, stop;
+{
+   GC_push_all_stack((ptr_t)start, (ptr_t)stop + sizeof(word));
+}
+
+/* Push routine with M3 specific calling convention. */
+GC_m3_push_root(dummy1, p, dummy2, dummy3)
+word *p;
+ptr_t dummy1, dummy2;
+int dummy3;
+{
+    word q = *p;
+    
+    if ((ptr_t)(q) >= GC_least_plausible_heap_addr
+	 && (ptr_t)(q) < GC_greatest_plausible_heap_addr) {
+	 GC_push_one_checked(q,FALSE);
+    }
+}
+
+/* M3 set equivalent to RTHeap.TracedRefTypes */
+typedef struct { int elts[1]; }  RefTypeSet;
+RefTypeSet GC_TracedRefTypes = {{0x1}};
+
+/* From finalize.c */
+extern void GC_push_finalizer_structures();
+
+/* From stubborn.c: */
+# ifdef STUBBORN_ALLOC
+    extern extern_ptr_t * GC_changing_list_start;
+# endif
+
+
+void GC_default_push_other_roots()
+{
+    /* Use the M3 provided routine for finding static roots.	*/
+    /* This is a bit dubious, since it presumes no C roots.	*/
+    /* We handle the collector roots explicitly.		*/
+       {
+# 	 ifdef STUBBORN_ALLOC
+           GC_push_one(GC_changing_list_start);
+#	 endif
+      	 GC_push_finalizer_structures();
+      	 RTMain__GlobalMapProc(GC_m3_push_root, 0, GC_TracedRefTypes);
+       }
+	if (GC_words_allocd > 0) {
+	    ThreadF__ProcessStacks(GC_push_thread_stack);
+	}
+	/* Otherwise this isn't absolutely necessary, and we have	*/
+	/* startup ordering problems.					*/
+}
+
+# endif /* SRC_M3 */
+
+# ifdef SOLARIS_THREADS
+
+void GC_default_push_other_roots()
+{
+    GC_push_all_stacks();
+}
+
+# endif /* SOLARIS_THREADS */
+
+void (*GC_push_other_roots)() = GC_default_push_other_roots;
+
+#endif
+
+/*
+ * Routines for accessing dirty  bits on virtual pages.
+ * We plan to eventaually implement four strategies for doing so:
+ * DEFAULT_VDB:	A simple dummy implementation that treats every page
+ *		as possibly dirty.  This makes incremental collection
+ *		useless, but the implementation is still correct.
+ * PCR_VDB:	Use PPCRs virtual dirty bit facility.
+ * PROC_VDB:	Use the /proc facility for reading dirty bits.  Only
+ *		works under some SVR4 variants.  Even then, it may be
+ *		too slow to be entirely satisfactory.  Requires reading
+ *		dirty bits for entire address space.  Implementations tend
+ *		to assume that the client is a (slow) debugger.
+ * MPROTECT_VDB:Protect pages and then catch the faults to keep track of
+ *		dirtied pages.  The implementation (and implementability)
+ *		is highly system dependent.  This usually fails when system
+ *		calls write to a protected page.  We prevent the read system
+ *		call from doing so.  It is the clients responsibility to
+ *		make sure that other system calls are similarly protected
+ *		or write only to the stack.
+ */
+ 
+bool GC_dirty_maintained;
+
+# ifdef DEFAULT_VDB
+
+/* All of the following assume the allocation lock is held, and	*/
+/* signals are disabled.					*/
+
+/* The client asserts that unallocated pages in the heap are never	*/
+/* written.								*/
+
+/* Initialize virtual dirty bit implementation.			*/
+void GC_dirty_init()
+{
+}
+
+/* Retrieve system dirty bits for heap to a local buffer.	*/
+/* Restore the systems notion of which pages are dirty.		*/
+void GC_read_dirty()
+{}
+
+/* Is the HBLKSIZE sized page at h marked dirty in the local buffer?	*/
+/* If the actual page size is different, this returns TRUE if any	*/
+/* of the pages overlapping h are dirty.  This routine may err on the	*/
+/* side of labelling pages as dirty (and this implementation does).	*/
+/*ARGSUSED*/
+bool GC_page_was_dirty(h)
+struct hblk *h;
+{
+    return(TRUE);
+}
+
+/*
+ * The following two routines are typically less crucial.  They matter
+ * most with large dynamic libraries, or if we can't accurately identify
+ * stacks, e.g. under Solaris 2.X.  Otherwise the following default
+ * versions are adequate.
+ */
+ 
+/* Could any valid GC heap pointer ever have been written to this page?	*/
+/*ARGSUSED*/
+bool GC_page_was_ever_dirty(h)
+struct hblk *h;
+{
+    return(TRUE);
+}
+
+/* Reset the n pages starting at h to "was never dirty" status.	*/
+void GC_is_fresh(h, n)
+struct hblk *h;
+word n;
+{
+}
+
+/* A call hints that h is about to be written.	*/
+/* May speed up some dirty bit implementations.	*/
+/*ARGSUSED*/
+void GC_write_hint(h)
+struct hblk *h;
+{
+}
+
+# endif /* DEFAULT_VDB */
+
+
+# ifdef MPROTECT_VDB
+
+/*
+ * See DEFAULT_VDB for interface descriptions.
+ */
+
+/*
+ * This implementation maintains dirty bits itself by catching write
+ * faults and keeping track of them.  We assume nobody else catches
+ * SIGBUS or SIGSEGV.  We assume no write faults occur in system calls
+ * except as a result of a read system call.  This means clients must
+ * either ensure that system calls do not touch the heap, or must
+ * provide their own wrappers analogous to the one for read.
+ * We assume the page size is a multiple of HBLKSIZE.
+ * This implementation is currently SunOS 4.X and IRIX 5.X specific, though we
+ * tried to use portable code where easily possible.  It is known
+ * not to work under a number of other systems.
+ */
+
+# include <sys/mman.h>
+# include <signal.h>
+# include <sys/syscall.h>
+
+VOLATILE page_hash_table GC_dirty_pages;
+				/* Pages dirtied since last GC_read_dirty. */
+
+word GC_page_size;
+
+bool GC_just_outside_heap(addr)
+word addr;
+{
+    register int i;
+    register word start;
+    register word end;
+    word mask = GC_page_size-1;
+    
+    for (i = 0; i < GC_n_heap_sects; i++) {
+    	start = (word) GC_heap_sects[i].hs_start;
+    	end = start + (word)GC_heap_sects[i].hs_bytes;
+        if (addr < start && addr >= (start & ~mask)
+            || addr >= end && addr < ((end + mask) & ~mask)) {
+            return(TRUE);
+        }
+    }
+    return(FALSE);
+}
+
+#if defined(SUNOS4) || defined(FREEBSD)
+    typedef void (* SIG_PF)();
+#endif
+
+#if defined(ALPHA) /* OSF1 */
+    typedef void (* SIG_PF)(int);
+#endif
+#if defined(IRIX5) || defined(ALPHA) /* OSF1 */
+    typedef void (* REAL_SIG_PF)(int, int, struct sigcontext *);
+#endif
+
+SIG_PF GC_old_bus_handler;
+SIG_PF GC_old_segv_handler;
+
+/*ARGSUSED*/
+# if defined (SUNOS4) || defined(FREEBSD)
+    void GC_write_fault_handler(sig, code, scp, addr)
+    int sig, code;
+    struct sigcontext *scp;
+    char * addr;
+#   ifdef SUNOS4
+#     define SIG_OK (sig == SIGSEGV || sig == SIGBUS)
+#     define CODE_OK (FC_CODE(code) == FC_PROT \
+              	    || (FC_CODE(code) == FC_OBJERR \
+              	       && FC_ERRNO(code) == FC_PROT))
+#   endif
+#   ifdef FREEBSD
+#     define SIG_OK (sig == SIGBUS)
+#     define CODE_OK (code == BUS_PAGE_FAULT)
+#   endif
+# endif
+# if defined(IRIX5) || defined(ALPHA) /* OSF1 */
+#   include <errno.h>
+    void GC_write_fault_handler(int sig, int code, struct sigcontext *scp)
+#   define SIG_OK (sig == SIGSEGV)
+#   ifdef ALPHA
+#     define CODE_OK (code == 2 /* experimentally determined */)
+#   endif
+#   ifdef IRIX5
+#     define CODE_OK (code == EACCES)
+#   endif
+# endif
+{
+    register int i;
+#   ifdef IRIX5
+	char * addr = (char *) (scp -> sc_badvaddr);
+#   endif
+#   ifdef ALPHA
+	char * addr = (char *) (scp -> sc_traparg_a0);
+#   endif
+    
+    if (SIG_OK && CODE_OK) {
+        register struct hblk * h =
+        		(struct hblk *)((word)addr & ~(GC_page_size-1));
+        
+        if (HDR(addr) == 0 && !GC_just_outside_heap((word)addr)) {
+            SIG_PF old_handler;
+            
+            if (sig == SIGSEGV) {
+            	old_handler = GC_old_segv_handler;
+            } else {
+                old_handler = GC_old_bus_handler;
+            }
+            if (old_handler == SIG_DFL) {
+                ABORT("Unexpected bus error or segmentation fault");
+            } else {
+#		if defined (SUNOS4) || defined(FREEBSD)
+		  (*old_handler) (sig, code, scp, addr);
+#		else
+		  (*(REAL_SIG_PF)old_handler) (sig, code, scp);
+#		endif
+		return;
+            }
+        }
+        for (i = 0; i < divHBLKSZ(GC_page_size); i++) {
+            register int index = PHT_HASH(h+i);
+            
+            set_pht_entry_from_index(GC_dirty_pages, index);
+        }
+        if (mprotect((caddr_t)h, (int)GC_page_size,
+            PROT_WRITE | PROT_READ | PROT_EXEC) < 0) {
+    	    ABORT("mprotect failed in handler");
+    	}
+#	if defined(IRIX5) || defined(ALPHA)
+	    /* IRIX resets the signal handler each time. */
+	    signal(SIGSEGV, (SIG_PF) GC_write_fault_handler);
+#	endif
+    	/* The write may not take place before dirty bits are read.	*/
+    	/* But then we'll fault again ...				*/
+    	return;
+    }
+
+    ABORT("Unexpected bus error or segmentation fault");
+}
+
+void GC_write_hint(h)
+struct hblk *h;
+{
+    register struct hblk * h_trunc =
+        		(struct hblk *)((word)h & ~(GC_page_size-1));
+    register int i;
+    register bool found_clean = FALSE;
+    
+    for (i = 0; i < divHBLKSZ(GC_page_size); i++) {
+        register int index = PHT_HASH(h_trunc+i);
+            
+        if (!get_pht_entry_from_index(GC_dirty_pages, index)) {
+            found_clean = TRUE;
+            set_pht_entry_from_index(GC_dirty_pages, index);
+        }
+    }
+    if (found_clean) {
+   	if (mprotect((caddr_t)h_trunc, (int)GC_page_size,
+            PROT_WRITE | PROT_READ | PROT_EXEC) < 0) {
+    	    ABORT("mprotect failed in GC_write_hint");
+    	}
+    }
+}
+				 
+void GC_dirty_init()
+{
+    GC_dirty_maintained = TRUE;
+    GC_page_size = getpagesize();
+    if (GC_page_size % HBLKSIZE != 0) {
+        GC_err_printf0("Page size not multiple of HBLKSIZE\n");
+        ABORT("Page size not multiple of HBLKSIZE");
+    }
+#   if defined(SUNOS4) || defined(FREEBSD)
+      GC_old_bus_handler = signal(SIGBUS, GC_write_fault_handler);
+      if (GC_old_bus_handler == SIG_IGN) {
+        GC_err_printf0("Previously ignored bus error!?");
+        GC_old_bus_handler == SIG_DFL;
+      }
+      if (GC_old_bus_handler != SIG_DFL) {
+#	ifdef PRINTSTATS
+          GC_err_printf0("Replaced other SIGBUS handler\n");
+#	endif
+      }
+#   endif
+#   if defined(IRIX5) || defined(ALPHA) || defined(SUNOS4)
+      GC_old_segv_handler = signal(SIGSEGV, (SIG_PF)GC_write_fault_handler);
+      if (GC_old_segv_handler == SIG_IGN) {
+        GC_err_printf0("Previously ignored segmentation violation!?");
+        GC_old_segv_handler == SIG_DFL;
+      }
+      if (GC_old_segv_handler != SIG_DFL) {
+#	ifdef PRINTSTATS
+          GC_err_printf0("Replaced other SIGSEGV handler\n");
+#	endif
+      }
+#   endif
+}
+
+
+
+void GC_protect_heap()
+{
+    word ps = GC_page_size;
+    word pmask = (ps-1);
+    ptr_t start;
+    word offset;
+    word len;
+    int i;
+    
+    for (i = 0; i < GC_n_heap_sects; i++) {
+        offset = (word)(GC_heap_sects[i].hs_start) & pmask;
+        start = GC_heap_sects[i].hs_start - offset;
+        len = GC_heap_sects[i].hs_bytes + offset;
+        len += ps-1; len &= ~pmask;
+    	if (mprotect((caddr_t)start, (int)len, PROT_READ | PROT_EXEC) < 0) {
+    	    ABORT("mprotect failed");
+    	}
+    }
+}
+
+# ifdef THREADS
+--> The following is broken.  We can lose dirty bits.  We would need
+--> the signal handler to cooperate, as in PCR.
+# endif
+
+void GC_read_dirty()
+{
+    BCOPY(GC_dirty_pages, GC_grungy_pages,
+          (sizeof GC_dirty_pages));
+    BZERO(GC_dirty_pages, (sizeof GC_dirty_pages));
+    GC_protect_heap();
+}
+
+bool GC_page_was_dirty(h)
+struct hblk * h;
+{
+    register word index = PHT_HASH(h);
+    
+    return(HDR(h) == 0 || get_pht_entry_from_index(GC_grungy_pages, index));
+}
+
+/*
+ * If this code needed to be thread-safe, the following would need to
+ * acquire and release the allocation lock.  This is tricky, since e.g.
+ * the cord package issues a read while it already holds the allocation lock.
+ */
+ 
+# ifdef THREADS
+	--> fix this
+# endif
+void GC_begin_syscall()
+{
+}
+
+void GC_end_syscall()
+{
+}
+
+void GC_unprotect_range(addr, len)
+ptr_t addr;
+word len;
+{
+    struct hblk * start_block;
+    struct hblk * end_block;
+    register struct hblk *h;
+    ptr_t obj_start;
+    
+    if (!GC_incremental) return;
+    obj_start = GC_base(addr);
+    if (obj_start == 0) return;
+    if (GC_base(addr + len - 1) != obj_start) {
+        ABORT("GC_unprotect_range(range bigger than object)");
+    }
+    start_block = (struct hblk *)((word)addr & ~(GC_page_size - 1));
+    end_block = (struct hblk *)((word)(addr + len - 1) & ~(GC_page_size - 1));
+    end_block += GC_page_size/HBLKSIZE - 1;
+    for (h = start_block; h <= end_block; h++) {
+        register word index = PHT_HASH(h);
+        
+        set_pht_entry_from_index(GC_dirty_pages, index);
+    }
+    if (mprotect((caddr_t)start_block,
+    	         (int)((ptr_t)end_block - (ptr_t)start_block)
+    	         + HBLKSIZE,
+    	         PROT_WRITE | PROT_READ | PROT_EXEC) < 0) {
+    	ABORT("mprotect failed in GC_unprotect_range");
+    }
+}
+
+/* Replacement for UNIX system call.	 */
+/* Other calls that write to the heap	 */
+/* should be handled similarly.		 */
+# ifndef LINT
+  int read(fd, buf, nbyte)
+# else
+  int GC_read(fd, buf, nbyte)
+# endif
+int fd;
+char *buf;
+int nbyte;
+{
+    int result;
+    
+    GC_begin_syscall();
+    GC_unprotect_range(buf, (word)nbyte);
+#   ifdef IRIX5
+	/* Indirect system call exists, but is undocumented, and	*/
+	/* always seems to return EINVAL.  There seems to be no		*/
+	/* general way to wrap system calls, since the system call	*/
+	/* convention appears to require an immediate argument for	*/
+	/* the system call number, and building the required code	*/
+	/* in the data segment also seems dangerous.  We can fake it	*/
+	/* for read; anything else is up to the client.			*/
+	{
+	    struct iovec iov;
+
+	    iov.iov_base = buf;
+	    iov.iov_len = nbyte;
+	    result = readv(fd, &iov, 1);
+	}
+#   else
+    	result = syscall(SYS_read, fd, buf, nbyte);
+#   endif
+    GC_end_syscall();
+    return(result);
+}
+
+/*ARGSUSED*/
+bool GC_page_was_ever_dirty(h)
+struct hblk *h;
+{
+    return(TRUE);
+}
+
+/* Reset the n pages starting at h to "was never dirty" status.	*/
+/*ARGSUSED*/
+void GC_is_fresh(h, n)
+struct hblk *h;
+word n;
+{
+}
+
+# endif /* MPROTECT_VDB */
+
+# ifdef PROC_VDB
+
+/*
+ * See DEFAULT_VDB for interface descriptions.
+ */
+ 
+/*
+ * This implementaion assumes a Solaris 2.X like /proc pseudo-file-system
+ * from which we can read page modified bits.  This facility is far from
+ * optimal (e.g. we would like to get the info for only some of the
+ * address space), but it avoids intercepting system calls.
+ */
+
+#include <sys/types.h>
+#include <sys/signal.h>
+#include <sys/fault.h>
+#include <sys/syscall.h>
+#include <sys/procfs.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+
+#define BUFSZ 20000
+char *GC_proc_buf;
+
+page_hash_table GC_written_pages = { 0 };	/* Pages ever dirtied	*/
+
+#ifdef SOLARIS_THREADS
+/* We don't have exact sp values for threads.  So we count on	*/
+/* occasionally declaring stack pages to be fresh.  Thus we 	*/
+/* need a real implementation of GC_is_fresh.  We can't clear	*/
+/* entries in GC_written_pages, since that would declare all	*/
+/* pages with the given hash address to be fresh.		*/
+#   define MAX_FRESH_PAGES 8*1024	/* Must be power of 2 */
+    struct hblk ** GC_fresh_pages;	/* A direct mapped cache.	*/
+    					/* Collisions are dropped.	*/
+
+#   define FRESH_PAGE_SLOT(h) (divHBLKSZ((word)(h)) & (MAX_FRESH_PAGES-1))
+#   define ADD_FRESH_PAGE(h) \
+	GC_fresh_pages[FRESH_PAGE_SLOT(h)] = (h)
+#   define PAGE_IS_FRESH(h) \
+	(GC_fresh_pages[FRESH_PAGE_SLOT(h)] == (h) && (h) != 0)
+#endif
+
+/* Add all pages in pht2 to pht1 */
+void GC_or_pages(pht1, pht2)
+page_hash_table pht1, pht2;
+{
+    register int i;
+    
+    for (i = 0; i < PHT_SIZE; i++) pht1[i] |= pht2[i];
+}
+
+int GC_proc_fd;
+
+void GC_dirty_init()
+{
+    int fd;
+    char buf[30];
+
+    GC_dirty_maintained = TRUE;
+    if (GC_words_allocd != 0 || GC_words_allocd_before_gc != 0) {
+    	register int i;
+    
+        for (i = 0; i < PHT_SIZE; i++) GC_written_pages[i] = (word)(-1);
+#       ifdef PRINTSTATS
+	    GC_printf1("Allocated words:%lu:all pages may have been written\n",
+	    	       (unsigned long)
+	    	      		(GC_words_allocd + GC_words_allocd_before_gc));
+#	endif       
+    }
+    sprintf(buf, "/proc/%d", getpid());
+    fd = open(buf, O_RDONLY);
+    if (fd < 0) {
+    	ABORT("/proc open failed");
+    }
+    GC_proc_fd = ioctl(fd, PIOCOPENPD, 0);
+    if (GC_proc_fd < 0) {
+    	ABORT("/proc ioctl failed");
+    }
+    GC_proc_buf = GC_scratch_alloc(BUFSZ);
+#   ifdef SOLARIS_THREADS
+	GC_fresh_pages = (struct hblk **)
+	  GC_scratch_alloc(MAX_FRESH_PAGES * sizeof (struct hblk *));
+	if (GC_fresh_pages == 0) {
+	    GC_err_printf0("No space for fresh pages\n");
+	    EXIT();
+	}
+	BZERO(GC_fresh_pages, MAX_FRESH_PAGES * sizeof (struct hblk *));
+#   endif
+}
+
+/* Ignore write hints. They don't help us here.	*/
+/*ARGSUSED*/
+void GC_write_hint(h)
+struct hblk *h;
+{
+}
+
+void GC_read_dirty()
+{
+    unsigned long ps, np;
+    int nmaps;
+    ptr_t vaddr;
+    struct prasmap * map;
+    char * bufp;
+    ptr_t current_addr, limit;
+    int i;
+
+    BZERO(GC_grungy_pages, (sizeof GC_grungy_pages));
+    
+    bufp = GC_proc_buf;
+    if (read(GC_proc_fd, bufp, BUFSZ) <= 0) {
+        ABORT("/proc read failed: BUFSZ too small?\n");
+    }
+    /* Copy dirty bits into GC_grungy_pages */
+    	nmaps = ((struct prpageheader *)bufp) -> pr_nmap;
+	/* printf( "nmaps = %d, PG_REFERENCED = %d, PG_MODIFIED = %d\n",
+		     nmaps, PG_REFERENCED, PG_MODIFIED); */
+	bufp = bufp + sizeof(struct prpageheader);
+	for (i = 0; i < nmaps; i++) {
+	    map = (struct prasmap *)bufp;
+	    vaddr = (ptr_t)(map -> pr_vaddr);
+	    ps = map -> pr_pagesize;
+	    np = map -> pr_npage;
+	    /* printf("vaddr = 0x%X, ps = 0x%X, np = 0x%X\n", vaddr, ps, np); */
+	    limit = vaddr + ps * np;
+	    bufp += sizeof (struct prasmap);
+	    for (current_addr = vaddr;
+	         current_addr < limit; current_addr += ps){
+	        if ((*bufp++) & PG_MODIFIED) {
+	            register struct hblk * h = (struct hblk *) current_addr;
+	            
+	            while ((ptr_t)h < current_addr + ps) {
+	                register word index = PHT_HASH(h);
+	                
+	                set_pht_entry_from_index(GC_grungy_pages, index);
+#			ifdef SOLARIS_THREADS
+			  {
+			    register int slot = FRESH_PAGE_SLOT(h);
+			    
+			    if (GC_fresh_pages[slot] == h) {
+			        GC_fresh_pages[slot] = 0;
+			    }
+			  }
+#			endif
+	                h++;
+	            }
+	        }
+	    }
+	    bufp += sizeof(long) - 1;
+	    bufp = (char *)((unsigned long)bufp & ~(sizeof(long)-1));
+	}
+    /* Update GC_written_pages. */
+        GC_or_pages(GC_written_pages, GC_grungy_pages);
+#   ifdef SOLARIS_THREADS
+      /* Make sure that old stacks are considered completely clean	*/
+      /* unless written again.						*/
+	GC_old_stacks_are_fresh();
+#   endif
+}
+
+bool GC_page_was_dirty(h)
+struct hblk *h;
+{
+    register word index = PHT_HASH(h);
+    register bool result;
+    
+    result = get_pht_entry_from_index(GC_grungy_pages, index);
+#   ifdef SOLARIS_THREADS
+	if (result && PAGE_IS_FRESH(h)) result = FALSE;
+	/* This happens only if page was declared fresh since	*/
+	/* the read_dirty call, e.g. because it's in an unused  */
+	/* thread stack.  It's OK to treat it as clean, in	*/
+	/* that case.  And it's consistent with 		*/
+	/* GC_page_was_ever_dirty.				*/
+#   endif
+    return(result);
+}
+
+bool GC_page_was_ever_dirty(h)
+struct hblk *h;
+{
+    register word index = PHT_HASH(h);
+    register bool result;
+    
+    result = get_pht_entry_from_index(GC_written_pages, index);
+#   ifdef SOLARIS_THREADS
+	if (result && PAGE_IS_FRESH(h)) result = FALSE;
+#   endif
+    return(result);
+}
+
+void GC_is_fresh(h, n)
+struct hblk *h;
+word n;
+{
+
+    register word index;
+    
+#   ifdef SOLARIS_THREADS
+      register word i;
+      
+      if (GC_fresh_pages != 0) {
+        for (i = 0; i < n; i++) {
+          PAGE_IS_FRESH(h + n);
+        }
+      }
+#   endif
+}
+
+# endif /* PROC_VDB */
+
+
+# ifdef PCR_VDB
+
+# include "vd/PCR_VD.h"
+
+# define NPAGES (32*1024)	/* 128 MB */
+
+PCR_VD_DB  GC_grungy_bits[NPAGES];
+
+ptr_t GC_vd_base;	/* Address corresponding to GC_grungy_bits[0]	*/
+			/* HBLKSIZE aligned.				*/
+
+void GC_dirty_init()
+{
+    GC_dirty_maintained = TRUE;
+    /* For the time being, we assume the heap generally grows up */
+    GC_vd_base = GC_heap_sects[0].hs_start;
+    if (GC_vd_base == 0) {
+   	ABORT("Bad initial heap segment");
+    }
+    if (PCR_VD_Start(HBLKSIZE, GC_vd_base, NPAGES*HBLKSIZE)
+	!= PCR_ERes_okay) {
+	ABORT("dirty bit initialization failed");
+    }
+}
+
+void GC_read_dirty()
+{
+    /* lazily enable dirty bits on newly added heap sects */
+    {
+        static int onhs = 0;
+        int nhs = GC_n_heap_sects;
+        for( ; onhs < nhs; onhs++ ) {
+            PCR_VD_WriteProtectEnable(
+                    GC_heap_sects[onhs].hs_start,
+                    GC_heap_sects[onhs].hs_bytes );
+        }
+    }
+
+
+    if (PCR_VD_Clear(GC_vd_base, NPAGES*HBLKSIZE, GC_grungy_bits)
+        != PCR_ERes_okay) {
+	ABORT("dirty bit read failed");
+    }
+}
+
+bool GC_page_was_dirty(h)
+struct hblk *h;
+{
+    if((ptr_t)h < GC_vd_base || (ptr_t)h >= GC_vd_base + NPAGES*HBLKSIZE) {
+	return(TRUE);
+    }
+    return(GC_grungy_bits[h - (struct hblk *)GC_vd_base] & PCR_VD_DB_dirtyBit);
+}
+
+/*ARGSUSED*/
+void GC_write_hint(h)
+struct hblk *h;
+{
+    PCR_VD_WriteProtectDisable(h, HBLKSIZE);
+    PCR_VD_WriteProtectEnable(h, HBLKSIZE);
+}
+
+# endif /* PCR_VDB */
+
+
+
+
diff --git a/pc_excludes b/pc_excludes
new file mode 100644
index 00000000..6f1465fa
--- /dev/null
+++ b/pc_excludes
@@ -0,0 +1,16 @@
+solaris_threads.c
+pcr_interface.c
+real_malloc.c
+mips_mach_dep.s
+rs6000_mach_dep.s
+alpha_mach_dep.s
+sparc_mach_dep.s
+PCR-Makefile
+setjmp_t.c
+SMakefile.amiga
+SCoptions.amiga
+README.amiga
+callprocs
+gc.man
+pc_excludes
+barrett_diagram
diff --git a/pcr_interface.c b/pcr_interface.c
new file mode 100644
index 00000000..0985c8f8
--- /dev/null
+++ b/pcr_interface.c
@@ -0,0 +1,114 @@
+/* 
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 1:59 pm PDT */
+# include "gc_priv.h"
+
+# ifdef PCR
+/*
+ * Note that POSIX PCR requires an ANSI C compiler.  Hence we are allowed
+ * to make the same assumption here.
+ * We wrap all of the allocator functions to avoid questions of
+ * compatibility between the prototyped and nonprototyped versions of the f
+ */
+# include "mm/PCR_MM.h"
+
+# define MY_MAGIC 17L
+
+void * GC_AllocProc(size_t size, PCR_Bool ptrFree, PCR_Bool clear )
+{
+    if (ptrFree) {
+        void * result = (void *)GC_malloc_atomic(size);
+        if (clear && result != 0) BZERO(result, size);
+        return(result);
+    } else {
+        return((void *)GC_malloc(size));
+    }
+}
+
+# define GC_ReallocProc GC_realloc
+
+# define GC_FreeProc GC_free
+
+typedef struct {
+  PCR_ERes (*ed_proc)(void *p, size_t size, PCR_Any data);
+  bool ed_pointerfree;
+  PCR_ERes ed_fail_code;
+  PCR_Any ed_client_data;
+} enumerate_data;
+
+void GC_enumerate_block(h, ed)
+register struct hblk *h;
+enumerate_data * ed;
+{
+    register hdr * hhdr;
+    register int sz;
+    word *p;
+    word * lim;
+    
+    hhdr = HDR(h);
+    sz = hhdr -> hb_sz;
+    if (sz >= 0 && ed -> ed_pointerfree
+    	|| sz <= 0 && !(ed -> ed_pointerfree)) return;
+    if (sz < 0) sz = -sz;
+    lim = (word *)(h+1) - sz;
+    p = (word *)h;
+    do {
+        if (PCR_ERes_IsErr(ed -> ed_fail_code)) return;
+        ed -> ed_fail_code =
+            (*(ed -> ed_proc))(p, WORDS_TO_BYTES(sz), ed -> ed_client_data);
+        p+= sz;
+    } while (p <= lim);
+}
+
+struct PCR_MM_ProcsRep * GC_old_allocator = 0;
+
+PCR_ERes GC_EnumerateProc(
+    PCR_Bool ptrFree,
+    PCR_ERes (*proc)(void *p, size_t size, PCR_Any data),
+    PCR_Any data
+)
+{
+    enumerate_data ed;
+    
+    ed.ed_proc = proc;
+    ed.ed_pointerfree = ptrFree;
+    ed.ed_fail_code = PCR_ERes_okay;
+    ed.ed_client_data = data;
+    GC_apply_to_all_blocks(GC_enumerate_block, &ed);
+    if (ed.ed_fail_code != PCR_ERes_okay) {
+        return(ed.ed_fail_code);
+    } else {
+    	/* Also enumerate objects allocated by my predecessors */
+    	return((*(GC_old_allocator->mmp_enumerate))(ptrFree, proc, data));
+    }
+}
+
+void GC_DummyFreeProc(void *p) {};
+
+void GC_DummyShutdownProc(void) {};
+
+struct PCR_MM_ProcsRep GC_Rep = {
+	MY_MAGIC,
+	GC_AllocProc,
+	GC_ReallocProc,
+	GC_DummyFreeProc,  	/* mmp_free */
+	GC_FreeProc,  		/* mmp_unsafeFree */
+	GC_EnumerateProc,
+	GC_DummyShutdownProc	/* mmp_shutdown */
+};
+
+void GC_pcr_install()
+{
+    PCR_MM_Install(&GC_Rep, &GC_old_allocator);
+}
+# endif
diff --git a/real_malloc.c b/real_malloc.c
new file mode 100644
index 00000000..dece9fdc
--- /dev/null
+++ b/real_malloc.c
@@ -0,0 +1,36 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:04 pm PDT */
+
+
+# ifdef PCR
+/*
+ * This definition should go in its own file that includes no other
+ * header files.  Otherwise, we risk not getting the underlying system
+ * malloc.
+ */
+# define PCR_NO_RENAME
+# include <stdlib.h>
+
+# ifdef __STDC__
+    char * real_malloc(size_t size)
+# else 
+    char * real_malloc()
+    int size;
+# endif
+{
+    return((char *)malloc(size));
+}
+#endif /* PCR */
+
diff --git a/reclaim.c b/reclaim.c
new file mode 100644
index 00000000..004cbf1d
--- /dev/null
+++ b/reclaim.c
@@ -0,0 +1,705 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:00 pm PDT */
+
+#include <stdio.h>
+#include "gc_priv.h"
+
+signed_word GC_mem_found = 0;
+			/* Number of longwords of memory GC_reclaimed     */
+
+# ifdef FIND_LEAK
+static report_leak(p, sz)
+ptr_t p;
+word sz;
+{
+    if (HDR(p) -> hb_obj_kind == PTRFREE) {
+        GC_err_printf0("Leaked atomic object at ");
+    } else {
+        GC_err_printf0("Leaked composite object at ");
+    }
+    if (GC_debugging_started && GC_has_debug_info(p)) {
+        GC_print_obj(p);
+    } else {
+        GC_err_printf2("0x%lx (appr. size = %ld)\n",
+       		      (unsigned long)p,
+       		      (unsigned long)WORDS_TO_BYTES(sz));
+    }
+}
+
+#   define FOUND_FREE(hblk, word_no) \
+      if (abort_if_found) { \
+         report_leak((long)hblk + WORDS_TO_BYTES(word_no), \
+         	     HDR(hblk) -> hb_sz); \
+      }
+# else
+#   define FOUND_FREE(hblk, word_no)
+# endif
+
+/*
+ * reclaim phase
+ *
+ */
+
+
+/*
+ * Test whether a block is completely empty, i.e. contains no marked
+ * objects.  This does not require the block to be in physical
+ * memory.
+ */
+ 
+bool GC_block_empty(hhdr)
+register hdr * hhdr;
+{
+    register word *p = (word *)(&(hhdr -> hb_marks[0]));
+    register word * plim =
+	    		(word *)(&(hhdr -> hb_marks[MARK_BITS_SZ]));
+    while (p < plim) {
+	if (*p++) return(FALSE);
+    }
+    return(TRUE);
+}
+
+# ifdef GATHERSTATS
+#   define INCR_WORDS(sz) n_words_found += (sz)
+# else
+#   define INCR_WORDS(sz)
+# endif
+/*
+ * Restore unmarked small objects in h of size sz to the object
+ * free list.  Returns the new list.
+ * Clears unmarked objects.
+ */
+/*ARGSUSED*/
+ptr_t GC_reclaim_clear(hbp, hhdr, sz, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+register hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+register word sz;
+{
+    register int word_no;
+    register word *p, *q, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif        
+    
+    p = (word *)(hbp->hb_body);
+    word_no = HDR_WORDS;
+    plim = (word *)((((word)hbp) + HBLKSIZE)
+		   - WORDS_TO_BYTES(sz));
+
+    /* go through all words in block */
+	while( p <= plim )  {
+	    if( mark_bit_from_hdr(hhdr, word_no) ) {
+		p += sz;
+	    } else {
+		FOUND_FREE(hbp, word_no);
+		INCR_WORDS(sz);
+		/* object is available - put on list */
+		    obj_link(p) = list;
+		    list = ((ptr_t)p);
+		/* Clear object, advance p to next object in the process */
+		    q = p + sz;
+                    p++; /* Skip link field */
+                    while (p < q) {
+			*p++ = 0;
+		    }
+	    }
+	    word_no += sz;
+	}
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+}
+
+#ifndef SMALL_CONFIG
+
+/*
+ * A special case for 2 word composite objects (e.g. cons cells):
+ */
+/*ARGSUSED*/
+ptr_t GC_reclaim_clear2(hbp, hhdr, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+{
+    register word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif
+    register word mark_word;
+    register int i;
+#   define DO_OBJ(start_displ) \
+	if (!(mark_word & ((word)1 << start_displ))) { \
+	    FOUND_FREE(hbp, p - (word *)hbp + start_displ); \
+	    p[start_displ] = (word)list; \
+	    list = (ptr_t)(p+start_displ); \
+	    p[start_displ+1] = 0; \
+	    INCR_WORDS(2); \
+	}
+    
+    p = (word *)(hbp->hb_body);
+    plim = (word *)(((word)hbp) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    for (i = 0; i < WORDSZ; i += 8) {
+		DO_OBJ(0);
+		DO_OBJ(2);
+		DO_OBJ(4);
+		DO_OBJ(6);
+		p += 8;
+		mark_word >>= 8;
+	    }
+	}	        
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+#   undef DO_OBJ
+}
+
+/*
+ * Another special case for 4 word composite objects:
+ */
+/*ARGSUSED*/
+ptr_t GC_reclaim_clear4(hbp, hhdr, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+{
+    register word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif
+    register word mark_word;
+#   define DO_OBJ(start_displ) \
+	if (!(mark_word & ((word)1 << start_displ))) { \
+	    FOUND_FREE(hbp, p - (word *)hbp + start_displ); \
+	    p[start_displ] = (word)list; \
+	    list = (ptr_t)(p+start_displ); \
+	    p[start_displ+1] = 0; \
+	    p[start_displ+2] = 0; \
+	    p[start_displ+3] = 0; \
+	    INCR_WORDS(4); \
+	}
+    
+    p = (word *)(hbp->hb_body);
+    plim = (word *)(((word)hbp) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    DO_OBJ(0);
+	    DO_OBJ(4);
+	    DO_OBJ(8);
+	    DO_OBJ(12);
+	    DO_OBJ(16);
+	    DO_OBJ(20);
+	    DO_OBJ(24);
+	    DO_OBJ(28);
+#	    if CPP_WORDSZ == 64
+	      DO_OBJ(32);
+	      DO_OBJ(36);
+	      DO_OBJ(40);
+	      DO_OBJ(44);
+	      DO_OBJ(48);
+	      DO_OBJ(52);
+	      DO_OBJ(56);
+	      DO_OBJ(60);
+#	    endif
+	    p += WORDSZ;
+	}	        
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+#   undef DO_OBJ
+}
+
+#endif /* !SMALL_CONFIG */
+
+/* The same thing, but don't clear objects: */
+/*ARGSUSED*/
+ptr_t GC_reclaim_uninit(hbp, hhdr, sz, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+register hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+register word sz;
+{
+    register int word_no;
+    register word *p, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif
+    
+    p = (word *)(hbp->hb_body);
+    word_no = HDR_WORDS;
+    plim = (word *)((((word)hbp) + HBLKSIZE)
+		   - WORDS_TO_BYTES(sz));
+
+    /* go through all words in block */
+	while( p <= plim )  {
+	    if( !mark_bit_from_hdr(hhdr, word_no) ) {
+		FOUND_FREE(hbp, word_no);
+		INCR_WORDS(sz);
+		/* object is available - put on list */
+		    obj_link(p) = list;
+		    list = ((ptr_t)p);
+	    }
+	    p += sz;
+	    word_no += sz;
+	}
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+}
+
+#ifndef SMALL_CONFIG
+/*
+ * Another special case for 2 word atomic objects:
+ */
+/*ARGSUSED*/
+ptr_t GC_reclaim_uninit2(hbp, hhdr, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+{
+    register word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif
+    register word mark_word;
+    register int i;
+#   define DO_OBJ(start_displ) \
+	if (!(mark_word & ((word)1 << start_displ))) { \
+	    FOUND_FREE(hbp, p - (word *)hbp + start_displ); \
+	    p[start_displ] = (word)list; \
+	    list = (ptr_t)(p+start_displ); \
+	    INCR_WORDS(2); \
+	}
+    
+    p = (word *)(hbp->hb_body);
+    plim = (word *)(((word)hbp) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    for (i = 0; i < WORDSZ; i += 8) {
+		DO_OBJ(0);
+		DO_OBJ(2);
+		DO_OBJ(4);
+		DO_OBJ(6);
+		p += 8;
+		mark_word >>= 8;
+	    }
+	}	        
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+#   undef DO_OBJ
+}
+
+/*
+ * Another special case for 4 word atomic objects:
+ */
+/*ARGSUSED*/
+ptr_t GC_reclaim_uninit4(hbp, hhdr, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+{
+    register word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif
+    register word mark_word;
+#   define DO_OBJ(start_displ) \
+	if (!(mark_word & ((word)1 << start_displ))) { \
+	    FOUND_FREE(hbp, p - (word *)hbp + start_displ); \
+	    p[start_displ] = (word)list; \
+	    list = (ptr_t)(p+start_displ); \
+	    INCR_WORDS(4); \
+	}
+    
+    p = (word *)(hbp->hb_body);
+    plim = (word *)(((word)hbp) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    DO_OBJ(0);
+	    DO_OBJ(4);
+	    DO_OBJ(8);
+	    DO_OBJ(12);
+	    DO_OBJ(16);
+	    DO_OBJ(20);
+	    DO_OBJ(24);
+	    DO_OBJ(28);
+#	    if CPP_WORDSZ == 64
+	      DO_OBJ(32);
+	      DO_OBJ(36);
+	      DO_OBJ(40);
+	      DO_OBJ(44);
+	      DO_OBJ(48);
+	      DO_OBJ(52);
+	      DO_OBJ(56);
+	      DO_OBJ(60);
+#	    endif
+	    p += WORDSZ;
+	}	        
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+#   undef DO_OBJ
+}
+
+/* Finally the one word case, which never requires any clearing: */
+/*ARGSUSED*/
+ptr_t GC_reclaim1(hbp, hhdr, list, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+hdr * hhdr;
+bool abort_if_found;		/* Abort if a reclaimable object is found */
+register ptr_t list;
+{
+    register word * mark_word_addr = &(hhdr->hb_marks[divWORDSZ(HDR_WORDS)]);
+    register word *p, *plim;
+#   ifdef GATHERSTATS
+        register int n_words_found = 0;
+#   endif
+    register word mark_word;
+    register int i;
+#   define DO_OBJ(start_displ) \
+	if (!(mark_word & ((word)1 << start_displ))) { \
+	    FOUND_FREE(hbp, p - (word *)hbp + start_displ); \
+	    p[start_displ] = (word)list; \
+	    list = (ptr_t)(p+start_displ); \
+	    INCR_WORDS(1); \
+	}
+    
+    p = (word *)(hbp->hb_body);
+    plim = (word *)(((word)hbp) + HBLKSIZE);
+
+    /* go through all words in block */
+	while( p < plim )  {
+	    mark_word = *mark_word_addr++;
+	    for (i = 0; i < WORDSZ; i += 4) {
+		DO_OBJ(0);
+		DO_OBJ(1);
+		DO_OBJ(2);
+		DO_OBJ(3);
+		p += 4;
+		mark_word >>= 4;
+	    }
+	}	        
+#   ifdef GATHERSTATS
+	GC_mem_found += n_words_found;
+#   endif
+    return(list);
+#   undef DO_OBJ
+}
+
+#endif /* !SMALL_CONFIG */
+
+/*
+ * Restore unmarked small objects in the block pointed to by hbp
+ * to the appropriate object free list.
+ * If entirely empty blocks are to be completely deallocated, then
+ * caller should perform that check.
+ */
+void GC_reclaim_small_nonempty_block(hbp, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+int abort_if_found;		/* Abort if a reclaimable object is found */
+{
+    hdr * hhdr;
+    register word sz;		/* size of objects in current block	*/
+    register struct obj_kind * ok;
+    register ptr_t * flh;
+    
+    hhdr = HDR(hbp);
+    sz = hhdr -> hb_sz;
+    hhdr -> hb_last_reclaimed = (unsigned short) GC_gc_no;
+    ok = &GC_obj_kinds[hhdr -> hb_obj_kind];
+    flh = &(ok -> ok_freelist[sz]);
+    GC_write_hint(hbp);
+
+    if (ok -> ok_init) {
+      switch(sz) {
+#      ifndef SMALL_CONFIG
+        case 1:
+            *flh = GC_reclaim1(hbp, hhdr, *flh, abort_if_found);
+            break;
+        case 2:
+            *flh = GC_reclaim_clear2(hbp, hhdr, *flh, abort_if_found);
+            break;
+        case 4:
+            *flh = GC_reclaim_clear4(hbp, hhdr, *flh, abort_if_found);
+            break;
+#      endif
+        default:
+            *flh = GC_reclaim_clear(hbp, hhdr, sz, *flh, abort_if_found);
+            break;
+      }
+    } else {
+      switch(sz) {
+#      ifndef SMALL_CONFIG
+        case 1:
+            *flh = GC_reclaim1(hbp, hhdr, *flh, abort_if_found);
+            break;
+        case 2:
+            *flh = GC_reclaim_uninit2(hbp, hhdr, *flh, abort_if_found);
+            break;
+        case 4:
+            *flh = GC_reclaim_uninit4(hbp, hhdr, *flh, abort_if_found);
+            break;
+#      endif
+        default:
+            *flh = GC_reclaim_uninit(hbp, hhdr, sz, *flh, abort_if_found);
+            break;
+      }
+    } 
+}
+
+/*
+ * Restore an unmarked large object or an entirely empty blocks of small objects
+ * to the heap block free list.
+ * Otherwise enqueue the block for later processing
+ * by GC_reclaim_small_nonempty_block.
+ * If abort_if_found is TRUE, then process any block immediately.
+ */
+void GC_reclaim_block(hbp, abort_if_found)
+register struct hblk *hbp;	/* ptr to current heap block		*/
+word abort_if_found;		/* Abort if a reclaimable object is found */
+{
+    register hdr * hhdr;
+    register word sz;		/* size of objects in current block	*/
+    register struct obj_kind * ok;
+    struct hblk ** rlh;
+
+    hhdr = HDR(hbp);
+    sz = hhdr -> hb_sz;
+    ok = &GC_obj_kinds[hhdr -> hb_obj_kind];
+
+    if( sz > MAXOBJSZ ) {  /* 1 big object */
+        if( !mark_bit_from_hdr(hhdr, HDR_WORDS) ) {
+	    FOUND_FREE(hbp, HDR_WORDS);
+#	    ifdef GATHERSTATS
+	        GC_mem_found += sz;
+#	    endif
+	    GC_freehblk(hbp);
+	}
+    } else {
+        bool empty = GC_block_empty(hhdr);
+        if (abort_if_found) {
+    	  GC_reclaim_small_nonempty_block(hbp, (int)abort_if_found);
+        } else if (empty) {
+#	  ifdef GATHERSTATS
+            GC_mem_found += BYTES_TO_WORDS(HBLKSIZE);
+#	  endif
+          GC_freehblk(hbp);
+        } else {
+          /* group of smaller objects, enqueue the real work */
+          rlh = &(ok -> ok_reclaim_list[sz]);
+          hhdr -> hb_next = *rlh;
+          *rlh = hbp;
+        }
+    }
+}
+
+/* Routines to gather and print heap block info 	*/
+/* intended for debugging.  Otherwise should be called	*/
+/* with lock.						*/
+static number_of_blocks;
+static total_bytes;
+
+/* Number of set bits in a word.  Not performance critical.	*/
+static int set_bits(n)
+word n;
+{
+    register word m = n;
+    register int result = 0;
+    
+    while (m > 0) {
+    	if (m & 1) result++;
+    	m >>= 1;
+    }
+    return(result);
+}
+
+/* Return the number of set mark bits in the given header	*/
+int GC_n_set_marks(hhdr)
+hdr * hhdr;
+{
+    register int result = 0;
+    register int i;
+    
+    for (i = 0; i < MARK_BITS_SZ; i++) {
+        result += set_bits(hhdr -> hb_marks[i]);
+    }
+    return(result);
+}
+
+/*ARGSUSED*/
+void GC_print_block_descr(h, dummy)
+struct hblk *h;
+word dummy;
+{
+    register hdr * hhdr = HDR(h);
+    register bytes = WORDS_TO_BYTES(hhdr -> hb_sz);
+    
+    GC_printf3("(%lu:%lu,%lu)", (unsigned long)(hhdr -> hb_obj_kind),
+    			        (unsigned long)bytes,
+    			        (unsigned long)(GC_n_set_marks(hhdr)));
+    bytes += HDR_BYTES + HBLKSIZE-1;
+    bytes &= ~(HBLKSIZE-1);
+    total_bytes += bytes;
+    number_of_blocks++;
+}
+
+void GC_print_block_list()
+{
+    GC_printf0("(kind(0=ptrfree,1=normal,2=unc.,3=stubborn):size_in_bytes, #_marks_set)\n");
+    number_of_blocks = 0;
+    total_bytes = 0;
+    GC_apply_to_all_blocks(GC_print_block_descr, (word)0);
+    GC_printf2("\nblocks = %lu, bytes = %lu\n",
+    	       (unsigned long)number_of_blocks,
+    	       (unsigned long)total_bytes);
+}
+
+/*
+ * Do the same thing on the entire heap, after first clearing small object
+ * free lists (if we are not just looking for leaks).
+ */
+void GC_start_reclaim(abort_if_found)
+int abort_if_found;		/* Abort if a GC_reclaimable object is found */
+{
+    int kind;
+    
+    /* Clear reclaim- and free-lists */
+      for (kind = 0; kind < GC_n_kinds; kind++) {
+        register ptr_t *fop;
+        register ptr_t *lim;
+        register struct hblk ** hbpp;
+        register struct hblk ** hlim;
+          
+        if (!abort_if_found) {
+            lim = &(GC_obj_kinds[kind].ok_freelist[MAXOBJSZ+1]);
+	    for( fop = GC_obj_kinds[kind].ok_freelist; fop < lim; fop++ ) {
+	      *fop = 0;
+	    }
+	} /* otherwise free list objects are marked, 	*/
+	  /* and its safe to leave them			*/
+	hlim = &(GC_obj_kinds[kind].ok_reclaim_list[MAXOBJSZ+1]);
+	for( hbpp = GC_obj_kinds[kind].ok_reclaim_list;
+	    hbpp < hlim; hbpp++ ) {
+	    *hbpp = 0;
+	}
+      }
+    
+#   ifdef PRINTBLOCKS
+        GC_printf0("GC_reclaim: current block sizes:\n");
+        GC_print_block_list();
+#   endif
+
+  /* Go through all heap blocks (in hblklist) and reclaim unmarked objects */
+  /* or enqueue the block for later processing.				   */
+    GC_apply_to_all_blocks(GC_reclaim_block, (word)abort_if_found);
+    
+}
+
+/*
+ * Sweep blocks of the indicated object size and kind until either the
+ * appropriate free list is nonempty, or there are no more blocks to
+ * sweep.
+ */
+void GC_continue_reclaim(sz, kind)
+word sz;	/* words */
+int kind;
+{
+    register hdr * hhdr;
+    register struct hblk * hbp;
+    register struct obj_kind * ok = &(GC_obj_kinds[kind]);
+    struct hblk ** rlh = &(ok -> ok_reclaim_list[sz]);
+    ptr_t *flh = &(ok -> ok_freelist[sz]);
+    
+    
+    while ((hbp = *rlh) != 0) {
+        hhdr = HDR(hbp);
+        *rlh = hhdr -> hb_next;
+        GC_reclaim_small_nonempty_block(hbp, FALSE);
+        if (*flh != 0) break;
+    }
+}
+
+/*
+ * Reclaim all blocks that have been recently reclaimed.
+ * Clear lists of blocks waiting to be reclaimed.
+ * Must be done before clearing mark bits with the world running,
+ * since otherwise a subsequent reclamation of block would see
+ * the wrong mark bits.
+ * SHOULD PROBABLY BE INCREMENTAL
+ */
+void GC_reclaim_or_delete_all()
+{
+    register word sz;
+    register int kind;
+    register hdr * hhdr;
+    register struct hblk * hbp;
+    register struct obj_kind * ok;
+    struct hblk ** rlh;
+#   ifdef PRINTTIMES
+	CLOCK_TYPE start_time;
+	CLOCK_TYPE done_time;
+	
+	GET_TIME(start_time);
+#   endif
+    
+    for (kind = 0; kind < GC_n_kinds; kind++) {
+    	ok = &(GC_obj_kinds[kind]);
+    	for (sz = 1; sz <= MAXOBJSZ; sz++) {
+    	    rlh = &(ok -> ok_reclaim_list[sz]);
+    	    while ((hbp = *rlh) != 0) {
+        	hhdr = HDR(hbp);
+        	*rlh = hhdr -> hb_next;
+        	if (hhdr -> hb_last_reclaimed == GC_gc_no - 1) {
+        	    /* It's likely we'll need it this time, too	*/
+        	    /* It's been touched recently, so this	*/
+        	    /* shouldn't trigger paging.		*/
+        	    GC_reclaim_small_nonempty_block(hbp, FALSE);
+        	}
+            }
+        }
+    }
+#   ifdef PRINTTIMES
+	GET_TIME(done_time);
+	GC_printf1("Disposing of reclaim lists took %lu msecs\n",
+	           MS_TIME_DIFF(done_time,start_time));
+#   endif
+}
diff --git a/rs6000_mach_dep.s b/rs6000_mach_dep.s
new file mode 100644
index 00000000..e0dbe808
--- /dev/null
+++ b/rs6000_mach_dep.s
@@ -0,0 +1,105 @@
+    .csect
+    .set   r0,0
+    .set   r1,1
+    .set   r2,2
+    .set   r3,3
+    .set   r4,4
+    .set   r5,5
+    .set   r6,6
+    .set   r7,7
+    .set   r8,8
+    .set   r9,9
+    .set   r10,10
+    .set   r11,11
+    .set   r12,12
+    .set   r13,13
+    .set   r14,14
+    .set   r15,15
+    .set   r16,16
+    .set   r17,17
+    .set   r18,18
+    .set   r19,19
+    .set   r20,20
+    .set   r21,21
+    .set   r22,22
+    .set   r23,23
+    .set   r24,24
+    .set   r25,25
+    .set   r26,26
+    .set   r27,27
+    .set   r28,28
+    .set   r29,29
+    .set   r30,30
+    .set   r31,31
+
+ # Mark from machine registers that are saved by C compiler
+    .globl  .GC_push_regs
+.GC_push_regs:
+    .extern .GC_push_one
+    stu	    r1,-64(r1)  # reserve stack frame
+    mflr    r0		# save link register
+    st      r0,0x48(r1)
+    oril    r3,r2,0x0   # mark from r2
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r13,0x0   # mark from r13-r31
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r14,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r15,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r16,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r17,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r18,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r19,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r20,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r21,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r22,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r23,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r24,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r25,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r26,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r27,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r28,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r29,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r30,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    oril    r3,r31,0x0
+    bl 	    .GC_push_one
+    cror    15,15,15
+    l       r0,0x48(r1)
+    mtlr    r0
+    ai      r1,r1,64
+    br
diff --git a/setjmp_t.c b/setjmp_t.c
new file mode 100644
index 00000000..14dcd30c
--- /dev/null
+++ b/setjmp_t.c
@@ -0,0 +1,151 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:01 pm PDT */
+ 
+/* Check whether setjmp actually saves registers in jmp_buf. */
+/* If it doesn't, the generic mark_regs code won't work.     */
+/* Compilers vary as to whether they will put x in a 	     */
+/* (callee-save) register without -O.  The code is	     */
+/* contrived such that any decent compiler should put x in   */
+/* a callee-save register with -O.  Thus it is is 	     */
+/* recommended that this be run optimized.  (If the machine  */
+/* has no callee-save registers, then the generic code is    */
+/* safe, but this will not be noticed by this piece of       */
+/* code.)						     */
+#include <stdio.h>
+#include <setjmp.h>
+#include "config.h"
+
+#ifdef __hpux
+/* X/OPEN PG3 defines "void* sbrk();" and this clashes with the definition */
+/* in gc_private.h, so we set the clock backwards with _CLASSIC_XOPEN_TYPES. */
+/* This is for HP-UX 8.0.
+/* sbrk() is not used in this file, of course.  W. Underwood, 15 Jun 1992 */
+#define _CLASSIC_XOPEN_TYPES
+#include <unistd.h>
+int
+getpagesize()
+{
+    return sysconf(_SC_PAGE_SIZE);
+}
+#endif
+
+#if defined(SUNOS5)
+#define _CLASSIC_XOPEN_TYPES
+#include <unistd.h>
+int
+getpagesize()
+{
+    return sysconf(_SC_PAGESIZE);
+}
+#endif
+
+#ifdef _AUX_SOURCE
+#include <sys/mmu.h>
+int
+getpagesize()
+{
+   return PAGESIZE;
+}
+#endif
+
+#ifdef AMIGA
+int
+getpagesize()
+{
+    return(4096);
+}
+#endif
+
+#ifdef __OS2__
+#define INCL_DOSFILEMGR
+#define INCL_DOSMISC
+#define INCL_DOSERRORS
+#include <os2.h>
+
+int
+getpagesize()
+{
+    ULONG result[1];
+    
+    if (DosQuerySysInfo(QSV_PAGE_SIZE, QSV_PAGE_SIZE,
+    		        (void *)result, sizeof(ULONG)) != NO_ERROR) {
+    	fprintf(stderr, "DosQuerySysInfo failed\n");
+    	result[0] = 4096;
+    }
+    return((int)(result[0]));
+}
+#endif
+
+struct {char a_a; char * a_b;} a;
+
+int * nested_sp()
+{
+    int dummy;
+    
+    return(&dummy);
+}
+
+main()
+{
+	int dummy;
+	long ps = getpagesize();
+	jmp_buf b;
+	register int x = strlen("a");  /* 1, slightly disguised */
+	static int y = 0;
+
+	if (nested_sp() < &dummy) {
+	  printf("Stack appears to grow down, which is the default.\n");
+	  printf("A good guess for STACKBOTTOM on this machine is 0x%X.\n",
+	         ((long)(&dummy) + ps) & ~(ps-1));
+	} else {
+	  printf("Stack appears to grow up.\n");
+	  printf("Define STACK_GROWS_UP in gc_private.h\n");
+	  printf("A good guess for STACKBOTTOM on this machine is 0x%X.\n",
+	         ((long)(&dummy) + ps) & ~(ps-1));
+	}
+	printf("Note that this may vary between machines of ostensibly\n");
+	printf("the same architecture (e.g. Sun 3/50s and 3/80s).\n");
+	printf("A good guess for ALIGNMENT on this machine is %d.\n",
+	       (unsigned long)(&(a.a_b))-(unsigned long)(&a));
+	
+	/* Encourage the compiler to keep x in a callee-save register */
+	x = 2*x-1;
+	printf("");
+	x = 2*x-1;
+	setjmp(b);
+	if (y == 1) {
+	    if (x == 2) {
+		printf("Generic mark_regs code probably wont work\n");
+#		if defined(SPARC) || defined(RS6000) || defined(VAX) || defined(MIPS) || defined(M68K) || defined(I386) || defined(NS32K) || defined(RT)
+		    printf("Assembly code supplied\n");
+#		else
+		    printf("Need assembly code\n");
+#		endif
+	    } else if (x == 1) {
+		printf("Generic mark_regs code may work\n");
+	    } else {
+		printf("Very strange setjmp implementation\n");
+	    }
+	}
+	y++;
+	x = 2;
+	if (y == 1) longjmp(b,1);
+	return(0);
+}
+
+int g(x)
+int x;
+{
+	return(x);
+}
diff --git a/solaris_threads.c b/solaris_threads.c
new file mode 100644
index 00000000..94f461e8
--- /dev/null
+++ b/solaris_threads.c
@@ -0,0 +1,516 @@
+/* 
+ * Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/*
+ * Support code for Solaris threads.  Provides functionality we wish Sun
+ * had provided.  Relies on some information we probably shouldn't rely on.
+ */
+/* Boehm, May 19, 1994 2:05 pm PDT */
+
+# if defined(SOLARIS_THREADS)
+
+# include "gc_priv.h"
+# include <thread.h>
+# include <synch.h>
+# include <sys/types.h>
+# include <sys/mman.h>
+# include <sys/time.h>
+# include <sys/resource.h>
+# define _CLASSIC_XOPEN_TYPES
+# include <unistd.h>
+
+#undef thr_join
+#undef thr_create
+#undef thr_suspend
+#undef thr_continue
+
+mutex_t GC_thr_lock;		/* Acquired before allocation lock	*/
+cond_t GC_prom_join_cv;		/* Broadcast whenany thread terminates	*/
+cond_t GC_create_cv;		/* Signalled when a new undetached	*/
+				/* thread starts.			*/
+
+bool GC_thr_initialized = FALSE;
+
+size_t GC_min_stack_sz;
+
+size_t GC_page_sz;
+
+# define N_FREE_LISTS 25
+ptr_t GC_stack_free_lists[N_FREE_LISTS] = { 0 };
+		/* GC_stack_free_lists[i] is free list for stacks of 	*/
+		/* size GC_min_stack_sz*2**i.				*/
+		/* Free lists are linked through first word.		*/
+
+/* Return a stack of size at least *stack_size.  *stack_size is	*/
+/* replaced by the actual stack size.				*/
+/* Caller holds GC_thr_lock.					*/
+ptr_t GC_stack_alloc(size_t * stack_size)
+{
+    register size_t requested_sz = *stack_size;
+    register size_t search_sz = GC_min_stack_sz;
+    register int index = 0;	/* = log2(search_sz/GC_min_stack_sz) */
+    register ptr_t result;
+    
+    while (search_sz < requested_sz) {
+        search_sz *= 2;
+        index++;
+    }
+    if ((result = GC_stack_free_lists[index]) == 0
+        && (result = GC_stack_free_lists[index+1]) != 0) {
+        /* Try next size up. */
+        search_sz *= 2; index++;
+    }
+    if (result != 0) {
+        GC_stack_free_lists[index] = *(ptr_t *)result;
+    } else {
+        result = (ptr_t) GC_scratch_alloc(search_sz + 2*GC_page_sz);
+        result = (ptr_t)(((word)result + GC_page_sz) & ~(GC_page_sz - 1));
+        /* Protect hottest page to detect overflow. */
+        mprotect(result, GC_page_sz, PROT_NONE);
+        GC_is_fresh((struct hblk *)result, divHBLKSZ(search_sz));
+        result += GC_page_sz;
+    }
+    *stack_size = search_sz;
+    return(result);
+}
+
+/* Caller holds GC_thr_lock.					*/
+void GC_stack_free(ptr_t stack, size_t size)
+{
+    register int index = 0;
+    register size_t search_sz = GC_min_stack_sz;
+    
+    while (search_sz < size) {
+        search_sz *= 2;
+        index++;
+    }
+    if (search_sz != size) ABORT("Bad stack size");
+    *(ptr_t *)stack = GC_stack_free_lists[index];
+    GC_stack_free_lists[index] = stack;
+}
+
+void GC_my_stack_limits();
+
+/* Notify virtual dirty bit implementation that known empty parts of	*/
+/* stacks do not contain useful data.					*/ 
+void GC_old_stacks_are_fresh()
+{
+    register int i;
+    register ptr_t p;
+    register size_t sz;
+    register struct hblk * h;
+    int dummy;
+    
+    if (!GC_thr_initialized) GC_thr_init();
+    for (i = 0, sz= GC_min_stack_sz; i < N_FREE_LISTS;
+         i++, sz *= 2) {
+         for (p = GC_stack_free_lists[i]; p != 0; p = *(ptr_t *)p) {
+             h = (struct hblk *)(((word)p + HBLKSIZE-1) & ~(HBLKSIZE-1));
+             if ((ptr_t)h == p) {
+                 GC_is_fresh((struct hblk *)p, divHBLKSZ(sz));
+             } else {
+                 GC_is_fresh((struct hblk *)p, divHBLKSZ(sz) - 1);
+                 BZERO(p, (ptr_t)h - p);
+             }
+         }
+    }
+    GC_my_stack_limits();
+}
+
+/* The set of all known threads.  We intercept thread creation and 	*/
+/* joins.  We never actually create detached threads.  We allocate all 	*/
+/* new thread stacks ourselves.  These allow us to maintain this	*/
+/* data structure.							*/
+/* Protected by GC_thr_lock.						*/
+/* Some of this should be declared vaolatile, but that's incosnsistent	*/
+/* with some library routine declarations.  In particular, the 		*/
+/* definition of cond_t doesn't mention volatile!			*/
+typedef struct GC_Thread_Rep {
+    struct GC_Thread_Rep * next;
+    thread_t id;
+    word flags;
+#	define FINISHED 1   	/* Thread has exited.	*/
+#	define DETACHED 2	/* Thread is intended to be detached.	*/
+#	define CLIENT_OWNS_STACK	4
+				/* Stack was supplied by client.	*/
+#	define SUSPENDED 8	/* Currently suspended.	*/	
+    ptr_t stack;
+    size_t stack_size;
+    cond_t join_cv;
+    void * status;
+} * GC_thread;
+
+# define THREAD_TABLE_SZ 128	/* Must be power of 2	*/
+volatile GC_thread GC_threads[THREAD_TABLE_SZ];
+
+/* Add a thread to GC_threads.  We assume it wasn't already there.	*/
+/* Caller holds GC_thr_lock if there is > 1 thread.			*/
+/* Initial caller may hold allocation lock.				*/
+GC_thread GC_new_thread(thread_t id)
+{
+    int hv = ((word)id) % THREAD_TABLE_SZ;
+    GC_thread result;
+    static struct GC_Thread_Rep first_thread;
+    static bool first_thread_used = FALSE;
+    
+    if (!first_thread_used) {
+    	result = &first_thread;
+    	first_thread_used = TRUE;
+    	/* Dont acquire allocation lock, since we may already hold it. */
+    } else {
+        result = GC_NEW(struct GC_Thread_Rep);
+    }
+    if (result == 0) return(0);
+    result -> id = id;
+    result -> next = GC_threads[hv];
+    GC_threads[hv] = result;
+    /* result -> finished = 0; */
+    (void) cond_init(&(result->join_cv), USYNC_THREAD, 0);
+    return(result);
+}
+
+/* Delete a thread from GC_threads.  We assume it is there.	*/
+/* (The code intentionally traps if it wasn't.)			*/
+/* Caller holds GC_thr_lock.					*/
+void GC_delete_thread(thread_t id)
+{
+    int hv = ((word)id) % THREAD_TABLE_SZ;
+    register GC_thread p = GC_threads[hv];
+    register GC_thread prev = 0;
+    
+    while (p -> id != id) {
+        prev = p;
+        p = p -> next;
+    }
+    if (prev == 0) {
+        GC_threads[hv] = p -> next;
+    } else {
+        prev -> next = p -> next;
+    }
+}
+
+/* Return the GC_thread correpsonding to a given thread_t.	*/
+/* Returns 0 if it's not there.					*/
+/* Caller holds GC_thr_lock.					*/
+GC_thread GC_lookup_thread(thread_t id)
+{
+    int hv = ((word)id) % THREAD_TABLE_SZ;
+    register GC_thread p = GC_threads[hv];
+    
+    while (p != 0 && p -> id != id) p = p -> next;
+    return(p);
+}
+
+/* Notify dirty bit implementation of unused parts of my stack. */
+void GC_my_stack_limits()
+{
+    int dummy;
+    register ptr_t hottest = (ptr_t)((word)(&dummy) & ~(HBLKSIZE-1));
+    register GC_thread me = GC_lookup_thread(thr_self());
+    register size_t stack_size = me -> stack_size;
+    register ptr_t stack;
+    
+    if (stack_size == 0) {
+      /* original thread */
+        struct rlimit rl;
+         
+        if (getrlimit(RLIMIT_STACK, &rl) != 0) ABORT("getrlimit failed");
+        /* Empirically, what should be the stack page with lowest	*/
+        /* address is actually inaccessible.				*/
+        stack_size = ((word)rl.rlim_cur & ~(HBLKSIZE-1)) - GC_page_sz;
+        stack = GC_stackbottom - stack_size + GC_page_sz;
+    } else {
+        stack = me -> stack;
+    }
+    if (stack > hottest || stack + stack_size < hottest) {
+    	ABORT("sp out of bounds");
+    }
+    GC_is_fresh((struct hblk *)stack, divHBLKSZ(hottest - stack));
+}
+
+
+/* Caller holds allocation lock.	*/
+void GC_stop_world()
+{
+    thread_t my_thread = thr_self();
+    register int i;
+    register GC_thread p;
+    
+    for (i = 0; i < THREAD_TABLE_SZ; i++) {
+      for (p = GC_threads[i]; p != 0; p = p -> next) {
+        if (p -> id != my_thread && !(p -> flags & SUSPENDED)) {
+            if (thr_suspend(p -> id) < 0) ABORT("thr_suspend failed");
+        }
+      }
+    }
+}
+
+/* Caller holds allocation lock.	*/
+void GC_start_world()
+{
+    thread_t my_thread = thr_self();
+    register int i;
+    register GC_thread p;
+    
+    for (i = 0; i < THREAD_TABLE_SZ; i++) {
+      for (p = GC_threads[i]; p != 0; p = p -> next) {
+        if (p -> id != my_thread && !(p -> flags & SUSPENDED)) {
+            if (thr_continue(p -> id) < 0) ABORT("thr_continue failed");
+        }
+      }
+    }
+}
+
+
+void GC_push_all_stacks()
+{
+    /* We assume the world is stopped. */
+    register int i;
+    register GC_thread p;
+    word dummy;
+    register ptr_t sp = (ptr_t) (&dummy);
+    register ptr_t bottom, top;
+    struct rlimit rl;
+    
+#   define PUSH(bottom,top) \
+      if (GC_dirty_maintained) { \
+	GC_push_dirty((bottom), (top), GC_page_was_ever_dirty, \
+		      GC_push_all_stack); \
+      } else { \
+        GC_push_all((bottom), (top)); \
+      }
+    if (!GC_thr_initialized) GC_thr_init();
+    for (i = 0; i < THREAD_TABLE_SZ; i++) {
+      for (p = GC_threads[i]; p != 0; p = p -> next) {
+        if (p -> stack_size != 0) {
+            bottom = p -> stack;
+            top = p -> stack + p -> stack_size;
+        } else {
+            /* The original stack. */
+            if (getrlimit(RLIMIT_STACK, &rl) != 0) ABORT("getrlimit failed");
+            bottom = GC_stackbottom - rl.rlim_cur + GC_page_sz;
+            top = GC_stackbottom;
+        }
+        if ((word)sp > (word)bottom && (word)sp < (word)top) bottom = sp;
+        PUSH(bottom, top);
+      }
+    }
+}
+
+/* The only thread that ever really performs a thr_join.	*/
+void * GC_thr_daemon(void * dummy)
+{
+    void *status;
+    thread_t departed;
+    register GC_thread t;
+    register int i;
+    register int result;
+    
+    for(;;) {
+      start:
+        result = thr_join((thread_t)0, &departed, &status);
+    	mutex_lock(&GC_thr_lock);
+    	if (result != 0) {
+    	    /* No more threads; wait for create. */
+    	    for (i = 0; i < THREAD_TABLE_SZ; i++) {
+    	        for (t = GC_threads[i]; t != 0; t = t -> next) {
+                    if (!(t -> flags & (DETACHED | FINISHED))) {
+                      mutex_unlock(&GC_thr_lock);
+                      goto start; /* Thread started just before we */
+                      		  /* acquired the lock.		   */
+                    }
+                }
+            }
+            cond_wait(&GC_create_cv, &GC_thr_lock);
+            mutex_unlock(&GC_thr_lock);
+            goto start;
+    	}
+    	t = GC_lookup_thread(departed);
+    	if (!(t -> flags & CLIENT_OWNS_STACK)) {
+    	    GC_stack_free(t -> stack, t -> stack_size);
+    	}
+    	if (t -> flags & DETACHED) {
+    	    GC_delete_thread(departed);
+    	} else {
+    	    t -> status = status;
+    	    t -> flags |= FINISHED;
+    	    cond_signal(&(t -> join_cv));
+    	    cond_broadcast(&GC_prom_join_cv);
+    	}
+    	mutex_unlock(&GC_thr_lock);
+    }
+}
+
+GC_thr_init()
+{
+    GC_thread t;
+    /* This gets called from the first thread creation, so	*/
+    /* mutual exclusion is not an issue.			*/
+    GC_thr_initialized = TRUE;
+    GC_min_stack_sz = ((thr_min_stack() + HBLKSIZE-1) & ~(HBLKSIZE - 1));
+    GC_page_sz = sysconf(_SC_PAGESIZE);
+    mutex_init(&GC_thr_lock, USYNC_THREAD, 0);
+    cond_init(&GC_prom_join_cv, USYNC_THREAD, 0);
+    cond_init(&GC_create_cv, USYNC_THREAD, 0);
+    /* Add the initial thread, so we can stop it.	*/
+      t = GC_new_thread(thr_self());
+      t -> stack_size = 0;
+      t -> flags = DETACHED;
+    if (thr_create(0 /* stack */, 0 /* stack_size */, GC_thr_daemon,
+    		   0 /* arg */, THR_DETACHED | THR_DAEMON,
+    		   0 /* thread_id */) != 0) {
+    	ABORT("Cant fork daemon");
+    }
+    
+}
+
+/* We acquire the allocation lock to prevent races with 	*/
+/* stopping/starting world.					*/
+int GC_thr_suspend(thread_t target_thread)
+{
+    GC_thread t;
+    int result;
+    
+    mutex_lock(&GC_thr_lock);
+    LOCK();
+    result = thr_suspend(target_thread);
+    if (result == 0) {
+    	t = GC_lookup_thread(target_thread);
+    	if (t == 0) ABORT("thread unknown to GC");
+        t -> flags |= SUSPENDED;
+    }
+    UNLOCK();
+    mutex_unlock(&GC_thr_lock);
+    return(result);
+}
+
+int GC_thr_continue(thread_t target_thread)
+{
+    GC_thread t;
+    int result;
+    
+    mutex_lock(&GC_thr_lock);
+    LOCK();
+    result = thr_continue(target_thread);
+    if (result == 0) {
+    	t = GC_lookup_thread(target_thread);
+    	if (t == 0) ABORT("thread unknown to GC");
+        t -> flags &= ~SUSPENDED;
+    }
+    UNLOCK();
+    mutex_unlock(&GC_thr_lock);
+    return(result);
+}
+
+int GC_thr_join(thread_t wait_for, thread_t *departed, void **status)
+{
+    register GC_thread t;
+    int result = 0;
+    
+    mutex_lock(&GC_thr_lock);
+    if (wait_for == 0) {
+        register int i;
+        register bool thread_exists;
+    
+    	for (;;) {
+    	  thread_exists = FALSE;
+    	  for (i = 0; i < THREAD_TABLE_SZ; i++) {
+    	    for (t = GC_threads[i]; t != 0; t = t -> next) {
+              if (!(t -> flags & DETACHED)) {
+                if (t -> flags & FINISHED) {
+                  goto found;
+                }
+                thread_exists = TRUE;
+              }
+            }
+          }
+          if (!thread_exists) {
+              result = ESRCH;
+    	      goto out;
+          }
+          cond_wait(&GC_prom_join_cv, &GC_thr_lock);
+        }
+    } else {
+        t = GC_lookup_thread(wait_for);
+    	if (t == 0 || t -> flags & DETACHED) {
+    	    result = ESRCH;
+    	    goto out;
+    	}
+    	if (wait_for == thr_self()) {
+    	    result = EDEADLK;
+    	    goto out;
+    	}
+    	while (!(t -> flags & FINISHED)) {
+            cond_wait(&(t -> join_cv), &GC_thr_lock);
+    	}
+    	
+    }
+  found:
+    if (status) *status = t -> status;
+    if (departed) *departed = t -> id;
+    cond_destroy(&(t -> join_cv));
+    GC_delete_thread(t -> id);
+  out:
+    mutex_unlock(&GC_thr_lock);
+    return(result);
+}
+
+
+int
+GC_thr_create(void *stack_base, size_t stack_size,
+              void *(*start_routine)(void *), void *arg, long flags,
+              thread_t *new_thread)
+{
+    int result;
+    GC_thread t;
+    thread_t my_new_thread;
+    word my_flags = 0;
+    void * stack = stack_base;
+   
+    if (!GC_thr_initialized) GC_thr_init();
+    mutex_lock(&GC_thr_lock);
+    if (stack == 0) {
+     	if (stack_size == 0) stack_size = GC_min_stack_sz;
+     	stack = (void *)GC_stack_alloc(&stack_size);
+     	if (stack == 0) {
+     	    mutex_unlock(&GC_thr_lock);
+     	    return(ENOMEM);
+     	}
+    } else {
+    	my_flags |= CLIENT_OWNS_STACK;
+    }
+    if (flags & THR_DETACHED) my_flags |= DETACHED;
+    if (flags & THR_SUSPENDED) my_flags |= SUSPENDED;
+    result = thr_create(stack, stack_size, start_routine,
+   		        arg, flags & ~THR_DETACHED, &my_new_thread);
+    if (result == 0) {
+        t = GC_new_thread(my_new_thread);
+        t -> flags = my_flags;
+        if (!(my_flags & DETACHED)) cond_init(&(t -> join_cv), USYNC_THREAD, 0);
+        t -> stack = stack;
+        t -> stack_size = stack_size;
+        if (new_thread != 0) *new_thread = my_new_thread;
+        cond_signal(&GC_create_cv);
+    } else if (!(my_flags & CLIENT_OWNS_STACK)) {
+      	GC_stack_free(stack, stack_size);
+    }        
+    mutex_unlock(&GC_thr_lock);  
+    return(result);
+}
+
+# else
+
+#ifndef LINT
+  int GC_no_sunOS_threads;
+#endif
+
+# endif /* SOLARIS_THREADS */
diff --git a/sparc_mach_dep.s b/sparc_mach_dep.s
new file mode 100644
index 00000000..a6a0a241
--- /dev/null
+++ b/sparc_mach_dep.s
@@ -0,0 +1,38 @@
+!	SPARCompiler 3.0 and later apparently no loner handles
+!	asm outside functions.  So we need a separate .s file
+!	This is only set up for SunOS 5, not SunOS 4.
+!	Assumes this is called before the stack contents are
+!	examined.
+
+	.seg 	"text"
+	.globl	GC_save_regs_in_stack
+	.globl 	GC_push_regs
+GC_save_regs_in_stack:
+GC_push_regs:
+	ta	0x3   ! ST_FLUSH_WINDOWS
+	mov	%sp,%o0
+	retl
+	nop
+	
+	.globl	GC_clear_stack_inner
+GC_clear_stack_inner:
+	mov	%sp,%o2		! Save sp
+	add	%sp,-8,%o3	! p = sp-8
+	clr	%g1		! [g0,g1] = 0
+	add	%o1,-0x60,%sp	! Move sp out of the way,
+				! so that traps still work.
+				! Includes some extra words
+				! so we can be sloppy below.
+loop:
+	std	%g0,[%o3]	! *(long long *)p = 0
+	cmp	%o3,%o1
+	bgu	loop		! if (p > limit) goto loop
+	  add	%o3,-8,%o3	! p -= 8 (delay slot)
+	retl
+	  mov	%o2,%sp		! Restore sp., delay slot
+	
+		
+		
+		
+		
+	
\ No newline at end of file
diff --git a/stubborn.c b/stubborn.c
new file mode 100644
index 00000000..e674977b
--- /dev/null
+++ b/stubborn.c
@@ -0,0 +1,315 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ */
+/* Boehm, May 19, 1994 2:11 pm PDT */
+
+
+#include "gc_priv.h"
+
+# ifdef STUBBORN_ALLOC
+/* Stubborn object (hard to change, nearly immutable) allocation. */
+
+extern ptr_t GC_clear_stack();	/* in misc.c, behaves like identity */
+
+#define GENERAL_MALLOC(lb,k) \
+    (extern_ptr_t)GC_clear_stack(GC_generic_malloc((word)lb, k))
+
+/* Data structure representing immutable objects that 	*/
+/* are still being initialized.				*/
+/* This is a bit baroque in order to avoid acquiring	*/
+/* the lock twice for a typical allocation.		*/
+
+extern_ptr_t * GC_changing_list_start;
+
+# ifdef THREADS
+  VOLATILE extern_ptr_t * VOLATILE GC_changing_list_current;
+# else
+  extern_ptr_t * GC_changing_list_current;
+# endif
+	/* Points at last added element.  Also (ab)used for		*/
+	/* synchronization.  Updates and reads are assumed atomic.	*/
+
+extern_ptr_t * GC_changing_list_limit;
+	/* Points at the last word of the buffer, which is always 0	*/
+	/* All entries in (GC_changing_list_current,			*/
+	/* GC_changing_list_limit] are 0				*/
+
+
+void GC_stubborn_init()
+{
+#   define INIT_SIZE 10
+
+    GC_changing_list_start = (extern_ptr_t *)
+    			GC_generic_malloc_inner(
+    				(word)(INIT_SIZE * sizeof(extern_ptr_t)),
+    				PTRFREE);
+    BZERO(GC_changing_list_start,
+    	  INIT_SIZE * sizeof(extern_ptr_t));
+    if (GC_changing_list_start == 0) {
+        GC_err_printf0("Insufficient space to start up\n");
+        ABORT("GC_stubborn_init: put of space");
+    }
+    GC_changing_list_current = GC_changing_list_start;
+    GC_changing_list_limit = GC_changing_list_start + INIT_SIZE - 1;
+    * GC_changing_list_limit = 0;
+}
+
+/* Compact and possibly grow GC_uninit_list.  The old copy is		*/
+/* left alone.	Lock must be held.					*/
+/* When called GC_changing_list_current == GC_changing_list_limit	*/
+/* which is one past the current element.				*/
+/* When we finish GC_changing_list_current again points one past last	*/
+/* element.								*/
+/* Invariant while this is running: GC_changing_list_current    	*/
+/* points at a word containing 0.					*/
+/* Returns FALSE on failure.						*/
+bool GC_compact_changing_list()
+{
+    register extern_ptr_t *p, *q;
+    register word count = 0;
+    word old_size = GC_changing_list_limit-GC_changing_list_start+1;
+    register word new_size = old_size;
+    extern_ptr_t * new_list;
+    
+    for (p = GC_changing_list_start; p < GC_changing_list_limit; p++) {
+        if (*p != 0) count++;
+    }
+    if (2 * count > old_size) new_size = 2 * count;
+    new_list = (extern_ptr_t *)
+    		GC_generic_malloc_inner(
+    				new_size * sizeof(extern_ptr_t), PTRFREE);
+    		/* PTRFREE is a lie.  But we don't want the collector to  */
+    		/* consider these.  We do want the list itself to be  	  */
+    		/* collectable.						  */
+    if (new_list == 0) return(FALSE);
+    BZERO(new_list, new_size * sizeof(extern_ptr_t));
+    q = new_list;
+    for (p = GC_changing_list_start; p < GC_changing_list_limit; p++) {
+        if (*p != 0) *q++ = *p;
+    }
+    GC_changing_list_start = new_list;
+    GC_changing_list_limit = new_list + new_size - 1;
+    GC_changing_list_current = q;
+    return(TRUE);
+}
+
+/* Add p to changing list.  Clear p on failure.	*/
+# define ADD_CHANGING(p) \
+	{	\
+	    register struct hblk * h = HBLKPTR(p);	\
+	    register word index = PHT_HASH(h);	\
+	    \
+	    set_pht_entry_from_index(GC_changed_pages, index);	\
+	}	\
+	if (*GC_changing_list_current != 0 \
+	    && ++GC_changing_list_current == GC_changing_list_limit) { \
+	    if (!GC_compact_changing_list()) (p) = 0; \
+	} \
+	*GC_changing_list_current = p;
+
+void GC_change_stubborn(p)
+extern_ptr_t p;
+{
+    DCL_LOCK_STATE;
+    
+    DISABLE_SIGNALS();
+    LOCK();
+    ADD_CHANGING(p);
+    UNLOCK();
+    ENABLE_SIGNALS();
+}
+
+void GC_end_stubborn_change(p)
+extern_ptr_t p;
+{
+#   ifdef THREADS
+      register VOLATILE extern_ptr_t * my_current = GC_changing_list_current;
+#   else
+      register extern_ptr_t * my_current = GC_changing_list_current;
+#   endif
+    register bool tried_quick;
+    DCL_LOCK_STATE;
+    
+    if (*my_current == p) {
+        /* Hopefully the normal case.					*/
+        /* Compaction could not have been running when we started.	*/
+        *my_current = 0;
+#	ifdef THREADS
+          if (my_current == GC_changing_list_current) {
+            /* Compaction can't have run in the interim. 	*/
+            /* We got away with the quick and dirty approach.   */
+            return;
+          }
+          tried_quick = TRUE;
+#	else
+	  return;
+#	endif
+    } else {
+        tried_quick = FALSE;
+    }
+    DISABLE_SIGNALS();
+    LOCK();
+    my_current = GC_changing_list_current;
+    for (; my_current >= GC_changing_list_start; my_current--) {
+        if (*my_current == p) {
+            *my_current = 0;
+            UNLOCK();
+            ENABLE_SIGNALS();
+            return;
+        }
+    }
+    if (!tried_quick) {
+        GC_err_printf1("Bad arg to GC_end_stubborn_change: 0x%lx\n",
+        	       (unsigned long)p);
+        ABORT("Bad arg to GC_end_stubborn_change");
+    }
+    UNLOCK();
+    ENABLE_SIGNALS();
+}
+
+/* Allocate lb bytes of composite (pointerful) data	*/
+/* No pointer fields may be changed after a call to	*/
+/* GC_end_stubborn_change(p) where p is the value	*/
+/* returned by GC_malloc_stubborn.			*/
+# ifdef __STDC__
+    extern_ptr_t GC_malloc_stubborn(size_t lb)
+# else
+    extern_ptr_t GC_malloc_stubborn(lb)
+    size_t lb;
+# endif
+{
+register ptr_t op;
+register ptr_t *opp;
+register word lw;
+ptr_t result;
+DCL_LOCK_STATE;
+
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+#       endif
+	opp = &(GC_sobjfreelist[lw]);
+	FASTLOCK();
+        if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {
+            FASTUNLOCK();
+            result = GC_generic_malloc((word)lb, STUBBORN);
+            goto record;
+        }
+        *opp = obj_link(op);
+        obj_link(op) = 0;
+        GC_words_allocd += lw;
+        result = (extern_ptr_t) op;
+        ADD_CHANGING(result);
+        FASTUNLOCK();
+        return((extern_ptr_t)result);
+   } else {
+       result = (extern_ptr_t)
+          	GC_generic_malloc((word)lb, STUBBORN);
+   }
+record:
+   DISABLE_SIGNALS();
+   LOCK();
+   ADD_CHANGING(result);
+   UNLOCK();
+   ENABLE_SIGNALS();
+   return((extern_ptr_t)GC_clear_stack(result));
+}
+
+
+/* Functions analogous to GC_read_dirty and GC_page_was_dirty.	*/
+/* Report pages on which stubborn objects were changed.		*/
+void GC_read_changed()
+{
+    register extern_ptr_t * p = GC_changing_list_start;
+    register extern_ptr_t q;
+    register struct hblk * h;
+    register word index;
+    
+    if (p == 0) /* initializing */ return;
+    BCOPY(GC_changed_pages, GC_prev_changed_pages,
+          (sizeof GC_changed_pages));
+    BZERO(GC_changed_pages, (sizeof GC_changed_pages));
+    for (; p <= GC_changing_list_current; p++) {
+        if ((q = *p) != 0) {
+            h = HBLKPTR(q);
+            index = PHT_HASH(h);
+            set_pht_entry_from_index(GC_changed_pages, index);
+        }
+    }
+}
+
+bool GC_page_was_changed(h)
+struct hblk * h;
+{
+    register word index = PHT_HASH(h);
+    
+    return(get_pht_entry_from_index(GC_prev_changed_pages, index));
+}
+
+/* Remove unreachable entries from changed list. Should only be	*/
+/* called with mark bits consistent and lock held.		*/
+void GC_clean_changing_list()
+{
+    register extern_ptr_t * p = GC_changing_list_start;
+    register extern_ptr_t q;
+    register ptr_t r;
+    register unsigned long count = 0;
+    register unsigned long dropped_count = 0;
+    
+    if (p == 0) /* initializing */ return;
+    for (; p <= GC_changing_list_current; p++) {
+        if ((q = *p) != 0) {
+            count++;
+            r = (ptr_t)GC_base(q);
+            if (r == 0 || !GC_is_marked(r)) {
+                *p = 0;
+                dropped_count++;
+	    }
+        }
+    }
+#   ifdef PRINTSTATS
+      if (count > 0) {
+        GC_printf2("%lu entries in changing list: reclaimed %lu\n",
+                  (unsigned long)count, (unsigned long)dropped_count);
+      }
+#   endif
+}
+
+#else /* !STUBBORN_ALLOC */
+
+# ifdef __STDC__
+    extern_ptr_t GC_malloc_stubborn(size_t lb)
+# else
+    extern_ptr_t GC_malloc_stubborn(lb)
+    size_t lb;
+# endif
+{
+    return(GC_malloc(lb));
+}
+
+/*ARGSUSED*/
+void GC_end_stubborn_change(p)
+extern_ptr_t p;
+{
+}
+
+/*ARGSUSED*/
+void GC_change_stubborn(p)
+extern_ptr_t p;
+{
+}
+
+
+#endif
diff --git a/test.c b/test.c
new file mode 100644
index 00000000..070d892e
--- /dev/null
+++ b/test.c
@@ -0,0 +1,764 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to copy this garbage collector for any purpose,
+ * provided the above notices are retained on all copies.
+ */
+/* Boehm, May 6, 1994 3:32 pm PDT */
+/* An incomplete test for the garbage collector.  		*/
+/* Some more obscure entry points are not tested at all.	*/
+
+# include <stdlib.h>
+# include <stdio.h>
+# include "gc.h"
+# include "gc_typed.h"
+# include "gc_priv.h"	/* For output and some statistics	*/
+# include "config.h"
+
+# ifdef MSWIN32
+#   include <windows.h>
+# endif
+
+# ifdef PCR
+#   include "th/PCR_ThCrSec.h"
+#   include "th/PCR_Th.h"
+# endif
+
+# ifdef SOLARIS_THREADS
+#   include <thread.h>
+#   include <synch.h>
+# endif
+
+# if defined(PCR) || defined(SOLARIS_THREADS)
+#   define THREADS
+# endif
+
+# ifdef AMIGA
+   long __stack = 200000;
+# endif
+
+# define FAIL (void)abort()
+
+/* AT_END may be defined to excercise the interior pointer test	*/
+/* if the collector is configured with ALL_INTERIOR_POINTERS.   */
+/* As it stands, this test should succeed with either		*/
+/* configuration.  In the FIND_LEAK configuration, it should	*/
+/* find lots of leaks, since we free almost nothing.		*/
+
+struct SEXPR {
+    struct SEXPR * sexpr_car;
+    struct SEXPR * sexpr_cdr;
+};
+
+# ifdef __STDC__
+    typedef void * void_star;
+# else
+    typedef char * void_star;
+# endif
+
+typedef struct SEXPR * sexpr;
+
+extern sexpr cons();
+
+# define nil ((sexpr) 0)
+# define car(x) ((x) -> sexpr_car)
+# define cdr(x) ((x) -> sexpr_cdr)
+# define is_nil(x) ((x) == nil)
+
+
+int extra_count = 0;        /* Amount of space wasted in cons node */
+
+/* Silly implementation of Lisp cons. Intentionally wastes lots of space */
+/* to test collector.                                                    */
+sexpr cons (x, y)
+sexpr x;
+sexpr y;
+{
+    register sexpr r;
+    register int *p;
+    register my_extra = extra_count;
+    
+    r = (sexpr) GC_MALLOC_STUBBORN(sizeof(struct SEXPR) + my_extra);
+    if (r == 0) {
+        (void)GC_printf0("Out of memory\n");
+        exit(1);
+    }
+    for (p = (int *)r;
+         ((char *)p) < ((char *)r) + my_extra + sizeof(struct SEXPR); p++) {
+	if (*p) {
+	    (void)GC_printf1("Found nonzero at 0x%lx - allocator is broken\n",
+	    		     (unsigned long)p);
+	    FAIL;
+        }
+        *p = 13;
+    }
+#   ifdef AT_END
+	r = (sexpr)((char *)r + (my_extra & ~7));
+#   endif
+    r -> sexpr_car = x;
+    r -> sexpr_cdr = y;
+    my_extra++;
+    if ( my_extra >= 5000 ) {
+        extra_count = 0;
+    } else {
+        extra_count = my_extra;
+    }
+    GC_END_STUBBORN_CHANGE((char *)r);
+    return(r);
+}
+
+sexpr small_cons (x, y)
+sexpr x;
+sexpr y;
+{
+    register sexpr r;
+    
+    r = (sexpr) GC_MALLOC(sizeof(struct SEXPR));
+    if (r == 0) {
+        (void)GC_printf0("Out of memory\n");
+        exit(1);
+    }
+    r -> sexpr_car = x;
+    r -> sexpr_cdr = y;
+    return(r);
+}
+
+sexpr small_cons_uncollectable (x, y)
+sexpr x;
+sexpr y;
+{
+    register sexpr r;
+    
+    r = (sexpr) GC_MALLOC_UNCOLLECTABLE(sizeof(struct SEXPR));
+    if (r == 0) {
+        (void)GC_printf0("Out of memory\n");
+        exit(1);
+    }
+    r -> sexpr_car = x;
+    r -> sexpr_cdr = (sexpr) (~(unsigned long)y);
+    return(r);
+}
+
+/* Return reverse(x) concatenated with y */
+sexpr reverse1(x, y)
+sexpr x, y;
+{
+    if (is_nil(x)) {
+        return(y);
+    } else {
+        return( reverse1(cdr(x), cons(car(x), y)) );
+    }
+}
+
+sexpr reverse(x)
+sexpr x;
+{
+    return( reverse1(x, nil) );
+}
+
+sexpr ints(low, up)
+int low, up;
+{
+    if (low > up) {
+	return(nil);
+    } else {
+        return(small_cons(small_cons((sexpr)low, (sexpr)0), ints(low+1, up)));
+    }
+}
+
+/* Too check uncollectable allocation we build lists with disguised cdr	*/
+/* pointers, and make sure they don't go away.				*/
+sexpr uncollectable_ints(low, up)
+int low, up;
+{
+    if (low > up) {
+	return(nil);
+    } else {
+        return(small_cons_uncollectable(small_cons((sexpr)low, (sexpr)0),
+               uncollectable_ints(low+1, up)));
+    }
+}
+
+void check_ints(list, low, up)
+sexpr list;
+int low, up;
+{
+    if ((int)(car(car(list))) != low) {
+        (void)GC_printf0(
+           "List reversal produced incorrect list - collector is broken\n");
+        exit(1);
+    }
+    if (low == up) {
+        if (cdr(list) != nil) {
+           (void)GC_printf0("List too long - collector is broken\n");
+           exit(1);
+        }
+    } else {
+        check_ints(cdr(list), low+1, up);
+    }
+}
+
+# define UNCOLLECTABLE_CDR(x) (sexpr)(~(unsigned long)(cdr(x)))
+
+void check_uncollectable_ints(list, low, up)
+sexpr list;
+int low, up;
+{
+    if ((int)(car(car(list))) != low) {
+        (void)GC_printf0(
+           "Uncollectable list corrupted - collector is broken\n");
+        exit(1);
+    }
+    if (low == up) {
+        if (UNCOLLECTABLE_CDR(list) != nil) {
+           (void)GC_printf0("Uncollectable ist too long - collector is broken\n");
+           exit(1);
+        }
+    } else {
+        check_uncollectable_ints(UNCOLLECTABLE_CDR(list), low+1, up);
+    }
+}
+
+/* Not used, but useful for debugging: */
+void print_int_list(x)
+sexpr x;
+{
+    if (is_nil(x)) {
+        (void)GC_printf0("NIL\n");
+    } else {
+        (void)GC_printf1("(%ld)", (long)(car(car(x))));
+        if (!is_nil(cdr(x))) {
+            (void)GC_printf0(", ");
+            (void)print_int_list(cdr(x));
+        } else {
+            (void)GC_printf0("\n");
+        }
+    }
+}
+
+/* Try to force a to be strangely aligned */
+struct {
+  char dummy;
+  sexpr aa;
+} A;
+#define a A.aa
+
+/*
+ * Repeatedly reverse lists built out of very different sized cons cells.
+ * Check that we didn't lose anything.
+ */
+void reverse_test()
+{
+    int i;
+    sexpr b;
+    sexpr c;
+    sexpr d;
+    sexpr e;
+#   if defined(MSWIN32)
+      /* Win32S only allows 128K stacks */
+#     define BIG 1000
+#   else
+#     define BIG 4500
+#   endif
+
+    a = ints(1, 49);
+    b = ints(1, 50);
+    c = ints(1, BIG);
+    d = uncollectable_ints(1, 100);
+    e = uncollectable_ints(1, 1);
+    /* Superficially test interior pointer recognition on stack */
+    c = (sexpr)((char *)c + sizeof(char *));
+    d = (sexpr)((char *)d + sizeof(char *));
+#   ifdef __STDC__
+        GC_FREE((void *)e);
+#   else
+        GC_FREE((char *)e);
+#   endif
+    for (i = 0; i < 50; i++) {
+        b = reverse(reverse(b));
+    }
+    check_ints(b,1,50);
+    for (i = 0; i < 60; i++) {
+    	/* This maintains the invariant that a always points to a list of */
+    	/* 49 integers.  Thus this is thread safe without locks.	  */
+        a = reverse(reverse(a));
+#	if !defined(AT_END) && !defined(THREADS)
+	  /* This is not thread safe, since realloc explicitly deallocates */
+          if (i & 1) {
+            a = (sexpr)GC_REALLOC((void_star)a, 500);
+          } else {
+            a = (sexpr)GC_REALLOC((void_star)a, 8200);
+          }
+#	endif
+    }
+    check_ints(a,1,49);
+    check_ints(b,1,50);
+    c = (sexpr)((char *)c - sizeof(char *));
+    d = (sexpr)((char *)d - sizeof(char *));
+    check_ints(c,1,BIG);
+    check_uncollectable_ints(d, 1, 100);
+    a = b = c = 0;
+}
+
+/*
+ * The rest of this builds balanced binary trees, checks that they don't
+ * disappear, and tests finalization.
+ */
+typedef struct treenode {
+    int level;
+    struct treenode * lchild;
+    struct treenode * rchild;
+} tn;
+
+int finalizable_count = 0;
+int finalized_count = 0;
+int dropped_something = 0;
+
+# ifdef __STDC__
+  void finalizer(void * obj, void * client_data)
+# else
+  void finalizer(obj, client_data)
+  char * obj;
+  char * client_data;
+# endif
+{
+  tn * t = (tn *)obj;
+
+# ifdef PCR
+     PCR_ThCrSec_EnterSys();
+# endif
+# ifdef SOLARIS_THREADS
+    static mutex_t incr_lock;
+    mutex_lock(&incr_lock);
+# endif
+  if ((int)client_data != t -> level) {
+     (void)GC_printf0("Wrong finalization data - collector is broken\n");
+     FAIL;
+  }
+  finalized_count++;
+# ifdef PCR
+    PCR_ThCrSec_ExitSys();
+# endif
+# ifdef SOLARIS_THREADS
+    mutex_unlock(&incr_lock);
+# endif
+}
+
+size_t counter = 0;
+
+# define MAX_FINALIZED 8000
+GC_FAR GC_word live_indicators[MAX_FINALIZED] = {0};
+int live_indicators_count = 0;
+
+tn * mktree(n)
+int n;
+{
+    tn * result = (tn *)GC_MALLOC(sizeof(tn));
+    
+    if (n == 0) return(0);
+    if (result == 0) {
+        (void)GC_printf0("Out of memory\n");
+        exit(1);
+    }
+    result -> level = n;
+    result -> lchild = mktree(n-1);
+    result -> rchild = mktree(n-1);
+    if (counter++ % 17 == 0 && n >= 2) {
+        tn * tmp = result -> lchild -> rchild;
+        
+        result -> lchild -> rchild = result -> rchild -> lchild;
+        result -> rchild -> lchild = tmp;
+    }
+    if (counter++ % 119 == 0) {
+        int my_index;
+        
+        {
+#	  ifdef PCR
+ 	    PCR_ThCrSec_EnterSys();
+#	  endif
+#	  ifdef SOLARIS_THREADS
+	    static mutex_t incr_lock;
+	    mutex_lock(&incr_lock);
+#	  endif
+		/* Losing a count here causes erroneous report of failure. */
+          finalizable_count++;
+          my_index = live_indicators_count++;
+#	  ifdef PCR
+ 	    PCR_ThCrSec_ExitSys();
+#	  endif
+#	  ifdef SOLARIS_THREADS
+	    mutex_unlock(&incr_lock);
+#	  endif
+	}
+
+        GC_REGISTER_FINALIZER((void_star)result, finalizer, (void_star)n,
+        		      (GC_finalization_proc *)0, (void_star *)0);
+        live_indicators[my_index] = 13;
+        if (GC_general_register_disappearing_link(
+         	(void_star *)(&(live_indicators[my_index])),
+         	(void_star)result) != 0) {
+         	GC_printf0("GC_general_register_disappearing_link failed\n");
+         	FAIL;
+        }
+        if (GC_unregister_disappearing_link(
+         	(void_star *)
+         	   (&(live_indicators[my_index]))) == 0) {
+         	GC_printf0("GC_unregister_disappearing_link failed\n");
+         	FAIL;
+        }
+        if (GC_general_register_disappearing_link(
+         	(void_star *)(&(live_indicators[my_index])),
+         	(void_star)result) != 0) {
+         	GC_printf0("GC_general_register_disappearing_link failed 2\n");
+         	FAIL;
+        }
+    }
+    return(result);
+}
+
+void chktree(t,n)
+tn *t;
+int n;
+{
+    if (n == 0 && t != 0) {
+        (void)GC_printf0("Clobbered a leaf - collector is broken\n");
+        FAIL;
+    }
+    if (n == 0) return;
+    if (t -> level != n) {
+        (void)GC_printf1("Lost a node at level %lu - collector is broken\n",
+        		 (unsigned long)n);
+        FAIL;
+    }
+    if (counter++ % 373 == 0) (void) GC_MALLOC(counter%5001);
+    chktree(t -> lchild, n-1);
+    if (counter++ % 73 == 0) (void) GC_MALLOC(counter%373);
+    chktree(t -> rchild, n-1);
+}
+
+# ifdef SOLARIS_THREADS
+thread_key_t fl_key;
+
+void * alloc8bytes()
+{
+    void ** my_free_list_ptr;
+    void * my_free_list;
+    
+    if (thr_getspecific(fl_key, (void **)(&my_free_list_ptr)) != 0) {
+    	(void)GC_printf0("thr_getspecific failed\n");
+    	FAIL;
+    }
+    if (my_free_list_ptr == 0) {
+        my_free_list_ptr = GC_NEW_UNCOLLECTABLE(void *);
+        if (thr_setspecific(fl_key, my_free_list_ptr) != 0) {
+    	    (void)GC_printf0("thr_setspecific failed\n");
+    	    FAIL;
+        }
+    }
+    my_free_list = *my_free_list_ptr;
+    if (my_free_list == 0) {
+        my_free_list = GC_malloc_many(8);
+        if (my_free_list == 0) {
+            (void)GC_printf0("alloc8bytes out of memory\n");
+    	    FAIL;
+        }
+    }
+    *my_free_list_ptr = GC_NEXT(my_free_list);
+    GC_NEXT(my_free_list) = 0;
+    return(my_free_list);
+}
+
+#else
+# define alloc8bytes() GC_MALLOC_ATOMIC(8)
+#endif
+
+void alloc_small(n)
+int n;
+{
+    register int i;
+    
+    for (i = 0; i < n; i += 8) {
+        if (alloc8bytes() == 0) {
+            (void)GC_printf0("Out of memory\n");
+            FAIL;
+        }
+    }
+}
+
+void tree_test()
+{
+    tn * root;
+    register int i;
+    
+    root = mktree(16);
+    alloc_small(5000000);
+    chktree(root, 16);
+    if (finalized_count && ! dropped_something) {
+        (void)GC_printf0("Premature finalization - collector is broken\n");
+        FAIL;
+    }
+    dropped_something = 1;
+    root = mktree(16);
+    chktree(root, 16);
+    for (i = 16; i >= 0; i--) {
+        root = mktree(i);
+        chktree(root, i);
+    }
+    alloc_small(5000000);
+}
+
+unsigned n_tests = 0;
+
+/* A very simple test of explicitly typed allocation	*/
+void typed_test()
+{
+    GC_word * old, * new;
+    GC_word bm3 = 0x3;
+    GC_word bm2 = 0x2;
+    GC_word bm_large = 0xf7ff7fff;
+    GC_descr d1 = GC_make_descriptor(&bm3, 2);
+    GC_descr d2 = GC_make_descriptor(&bm2, 2);
+#   ifndef LINT
+      GC_descr dummy = GC_make_descriptor(&bm_large, 32);
+#   endif
+    GC_descr d3 = GC_make_descriptor(&bm_large, 32);
+    register int i;
+    
+    old = 0;
+    for (i = 0; i < 4000; i++) {
+        new = (GC_word *) GC_malloc_explicitly_typed(4 * sizeof(GC_word), d1);
+        new[0] = 17;
+        new[1] = (GC_word)old;
+        old = new;
+        new = (GC_word *) GC_malloc_explicitly_typed(4 * sizeof(GC_word), d2);
+        new[0] = 17;
+        new[1] = (GC_word)old;
+        old = new;
+        new = (GC_word *) GC_malloc_explicitly_typed(33 * sizeof(GC_word), d3);
+        new[0] = 17;
+        new[1] = (GC_word)old;
+        old = new;
+        new = (GC_word *) GC_calloc_explicitly_typed(4, 2 * sizeof(GC_word),
+        					     d1);
+        new[0] = 17;
+        new[1] = (GC_word)old;
+        old = new;
+        if (i & 0xff) {
+          new = (GC_word *) GC_calloc_explicitly_typed(7, 3 * sizeof(GC_word),
+        					     d2);
+        } else {
+          new = (GC_word *) GC_calloc_explicitly_typed(1001,
+          					       3 * sizeof(GC_word),
+        					       d2);
+        }
+        new[0] = 17;
+        new[1] = (GC_word)old;
+        old = new;
+    }
+    for (i = 0; i < 20000; i++) {
+        if (new[0] != 17) {
+            (void)GC_printf1("typed alloc failed at %lu\n",
+            		     (unsigned long)i);
+            FAIL;
+        }
+        new[0] = 0;
+        old = new;
+        new = (GC_word *)(old[1]);
+    }
+}
+
+void run_one_test()
+{
+    DCL_LOCK_STATE;
+    
+#   ifndef GC_DEBUG
+	if (GC_size(GC_MALLOC(7)) != 8
+	    || GC_size(GC_MALLOC(15)) != 16) {
+	    (void)GC_printf0("GC_size produced unexpected results\n");
+	    FAIL;
+	}
+#   endif
+    reverse_test();
+#   ifdef PRINTSTATS
+	GC_printf0("-------------Finished reverse_test\n");
+#   endif
+    typed_test();
+#   ifdef PRINTSTATS
+	GC_printf0("-------------Finished typed_test\n");
+#   endif
+    tree_test();
+    LOCK();
+    n_tests++;
+    UNLOCK();
+    
+}
+
+void check_heap_stats()
+{
+    unsigned long max_heap_sz;
+    register int i;
+    int still_live;
+    
+    if (sizeof(char *) > 4) {
+        max_heap_sz = 13000000;
+    } else {
+    	max_heap_sz = 10000000;
+    }
+#   ifdef GC_DEBUG
+	max_heap_sz *= 2;
+#       ifdef SPARC
+	    max_heap_sz *= 2;
+#       endif
+#   endif
+    /* Garbage collect repeatedly so that all inaccessible objects	*/
+    /* can be finalized.						*/
+      for (i = 0; i < 16; i++) {
+        GC_gcollect();
+      }
+    (void)GC_printf1("Completed %lu tests\n", (unsigned long)n_tests);
+    (void)GC_printf2("Finalized %lu/%lu objects - ",
+    		     (unsigned long)finalized_count,
+    		     (unsigned long)finalizable_count);
+    if (finalized_count > finalizable_count
+        || finalized_count < finalizable_count/2) {
+        (void)GC_printf0("finalization is probably broken\n");
+        FAIL;
+    } else {
+        (void)GC_printf0("finalization is probably ok\n");
+    }
+    still_live = 0;
+    for (i = 0; i < MAX_FINALIZED; i++) {
+    	if (live_indicators[i] != 0) {
+    	    still_live++;
+    	}
+    }
+    if (still_live != finalizable_count - finalized_count) {
+        (void)GC_printf1
+            ("%lu disappearing links remain - disappearing links are broken\n",
+             (unsigned long) still_live);
+        FAIL;
+    }
+    (void)GC_printf1("Total number of bytes allocated is %lu\n",
+    		(unsigned long)
+    	           WORDS_TO_BYTES(GC_words_allocd + GC_words_allocd_before_gc));
+    (void)GC_printf1("Final heap size is %lu bytes\n",
+    		     (unsigned long)GC_get_heap_size());
+    if (WORDS_TO_BYTES(GC_words_allocd + GC_words_allocd_before_gc)
+        < 33500000*n_tests) {
+        (void)GC_printf0("Incorrect execution - missed some allocations\n");
+        FAIL;
+    }
+    if (GC_get_heap_size() > max_heap_sz*n_tests) {
+        (void)GC_printf0("Unexpected heap growth - collector may be broken\n");
+        FAIL;
+    }
+    (void)GC_printf0("Collector appears to work\n");
+}
+
+#if !defined(PCR) && !defined(SOLARIS_THREADS) || defined(LINT)
+#ifdef MSWIN32
+  int APIENTRY WinMain(HINSTANCE instance, HINSTANCE prev, LPSTR cmd, int n)
+#else
+  int main()
+#endif
+{
+    n_tests = 0;
+#   if defined(MPROTECT_VDB) || defined(PROC_VDB)
+      GC_enable_incremental();
+      (void) GC_printf0("Switched to incremental mode\n");
+#     if defined(MPROTECT_VDB)
+	(void)GC_printf0("Emulating dirty bits with mprotect/signals\n");
+#     else
+	(void)GC_printf0("Reading dirty bits from /proc\n");
+#      endif
+#   endif
+    run_one_test();
+    check_heap_stats();
+    (void)fflush(stdout);
+#   ifdef LINT
+	/* Entry points we should be testing, but aren't.		   */
+	/* Some can be tested by defining GC_DEBUG at the top of this file */
+	/* This is a bit SunOS4 specific.				   */			
+	GC_noop(GC_expand_hp, GC_add_roots, GC_clear_roots,
+	        GC_register_disappearing_link,
+	        GC_print_obj, GC_debug_change_stubborn,
+	        GC_debug_end_stubborn_change, GC_debug_malloc_uncollectable,
+	        GC_debug_free, GC_debug_realloc, GC_generic_malloc_words_small,
+	        GC_init, GC_make_closure, GC_debug_invoke_finalizer,
+	        GC_page_was_ever_dirty, GC_is_fresh,
+		GC_malloc_ignore_off_page);
+#   endif
+    return(0);
+}
+# endif
+
+#ifdef PCR
+test()
+{
+    PCR_Th_T * th1;
+    PCR_Th_T * th2;
+    int code;
+
+    n_tests = 0;
+    GC_enable_incremental();
+    th1 = PCR_Th_Fork(run_one_test, 0);
+    th2 = PCR_Th_Fork(run_one_test, 0);
+    run_one_test();
+    if (PCR_Th_T_Join(th1, &code, NIL, PCR_allSigsBlocked, PCR_waitForever)
+        != PCR_ERes_okay || code != 0) {
+        (void)GC_printf0("Thread 1 failed\n");
+    }
+    if (PCR_Th_T_Join(th2, &code, NIL, PCR_allSigsBlocked, PCR_waitForever)
+        != PCR_ERes_okay || code != 0) {
+        (void)GC_printf0("Thread 2 failed\n");
+    }
+    check_heap_stats();
+    (void)fflush(stdout);
+    return(0);
+}
+#endif
+
+#ifdef SOLARIS_THREADS
+void * thr_run_one_test(void * arg)
+{
+    run_one_test();
+    return(0);
+}
+main()
+{
+    thread_t th1;
+    thread_t th2;
+    int code;
+
+    n_tests = 0;
+    GC_enable_incremental();
+    if (thr_keycreate(&fl_key, GC_free) != 0) {
+        (void)GC_printf1("Key creation failed %lu\n", (unsigned long)code);
+    	FAIL;
+    }
+    if ((code = thr_create(0, 1024*1024, thr_run_one_test, 0, 0, &th1)) != 0) {
+    	(void)GC_printf1("Thread 1 creation failed %lu\n", (unsigned long)code);
+    	FAIL;
+    }
+    if ((code = thr_create(0, 1024*1024, thr_run_one_test, 0, THR_NEW_LWP, &th2)) != 0) {
+    	(void)GC_printf1("Thread 2 creation failed %lu\n", (unsigned long)code);
+    	FAIL;
+    }
+    run_one_test();
+    if ((code = thr_join(th1, 0, 0)) != 0) {
+        (void)GC_printf1("Thread 1 failed %lu\n", (unsigned long)code);
+        FAIL;
+    }
+    if (thr_join(th2, 0, 0) != 0) {
+        (void)GC_printf1("Thread 2 failed %lu\n", (unsigned long)code);
+        FAIL;
+    }
+    check_heap_stats();
+    (void)fflush(stdout);
+    return(0);
+}
+#endif
diff --git a/typd_mlc.c b/typd_mlc.c
new file mode 100644
index 00000000..b04cbbeb
--- /dev/null
+++ b/typd_mlc.c
@@ -0,0 +1,777 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ *
+ * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
+ * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
+ *
+ * Permission is hereby granted to use or copy this program
+ * for any purpose,  provided the above notices are retained on all copies.
+ * Permission to modify the code and to distribute modified code is granted,
+ * provided the above notices are retained, and a notice that the code was
+ * modified is included with the above copyright notice.
+ *
+ */
+/* Boehm, May 19, 1994 2:06 pm PDT */
+
+
+/*
+ * Some simple primitives for allocation with explicit type information.
+ * Simple objects are allocated such that they contain a GC_descr at the
+ * end (in the last allocated word).  This descriptor may be a procedure
+ * which then examines an extended descriptor passed as its environment.
+ *
+ * Arrays are treated as simple objects if they have sufficiently simple
+ * structure.  Otherwise they are allocated from an array kind that supplies
+ * a special mark procedure.  These arrays contain a pointer to a
+ * complex_descriptor as their last word.
+ * This is done because the environment field is too small, and the collector
+ * must trace the complex_descriptor.
+ *
+ * Note that descriptors inside objects may appear cleared, if we encounter a
+ * false refrence to an object on a free list.  In the GC_descr case, this
+ * is OK, since a 0 descriptor corresponds to examining no fields.
+ * In the complex_descriptor case, we explicitly check for that case.
+ *
+ * MAJOR PARTS OF THIS CODE HAVE NOT BEEN TESTED AT ALL and are not testable,
+ * since they are not accessible through the current interface.
+ */
+
+#include "gc_priv.h"
+#include "gc_mark.h"
+#include "gc_typed.h"
+
+# ifdef ADD_BYTE_AT_END
+#   define EXTRA_BYTES (sizeof(word) - 1)
+# else
+#   define EXTRA_BYTES (sizeof(word))
+# endif
+
+bool GC_explicit_typing_initialized = FALSE;
+
+int GC_explicit_kind;	/* Object kind for objects with indirect	*/
+			/* (possibly extended) descriptors.		*/
+
+int GC_array_kind;	/* Object kind for objects with complex		*/
+			/* descriptors and GC_array_mark_proc.		*/
+
+/* Extended descriptors.  GC_typed_mark_proc understands these.	*/
+/* These are used for simple objects that are larger than what	*/
+/* can be described by a BITMAP_BITS sized bitmap.		*/
+typedef struct {
+	word ed_bitmap;	/* lsb corresponds to first word.	*/
+	bool ed_continued;	/* next entry is continuation.	*/
+} ext_descr;
+
+/* Array descriptors.  GC_array_mark_proc understands these.	*/
+/* We may eventually need to add provisions for headers and	*/
+/* trailers.  Hence we provide for tree structured descriptors, */
+/* though we don't really use them currently.			*/
+typedef union ComplexDescriptor {
+    struct LeafDescriptor {	/* Describes simple array	*/
+        word ld_tag;
+#	define LEAF_TAG 1
+	word ld_size;		/* bytes per element	*/
+				/* multiple of ALIGNMENT	*/
+	word ld_nelements;	/* Number of elements.	*/
+	GC_descr ld_descriptor; /* A simple length, bitmap,	*/
+				/* or procedure descriptor.	*/
+    } ld;
+    struct ComplexArrayDescriptor {
+        word ad_tag;
+#	define ARRAY_TAG 2
+	word ad_nelements;
+	union ComplexDescriptor * ad_element_descr;
+    } ad;
+    struct SequenceDescriptor {
+        word sd_tag;
+#	define SEQUENCE_TAG 3
+	union ComplexDescriptor * sd_first;
+	union ComplexDescriptor * sd_second;
+    } sd;
+} complex_descriptor;
+#define TAG ld.ld_tag
+
+ext_descr * GC_ext_descriptors;	/* Points to array of extended 	*/
+				/* descriptors.			*/
+
+word GC_ed_size = 0;	/* Current size of above arrays.	*/
+# define ED_INITIAL_SIZE 100;
+
+word GC_avail_descr = 0;	/* Next available slot.		*/
+
+int GC_typed_mark_proc_index;	/* Indices of my mark		*/
+int GC_array_mark_proc_index;	/* procedures.			*/
+
+/* Add a multiword bitmap to GC_ext_descriptors arrays.  Return	*/
+/* starting index.						*/
+/* Returns -1 on failure.					*/
+/* Caller does not hold allocation lock.			*/
+signed_word GC_add_ext_descriptor(bm, nbits)
+GC_bitmap bm;
+word nbits;
+{
+    register size_t nwords = divWORDSZ(nbits + WORDSZ-1);
+    register signed_word result;
+    register word i;
+    register word last_part;
+    register int extra_bits;
+    DCL_LOCK_STATE;
+
+    DISABLE_SIGNALS();
+    LOCK();
+    while (GC_avail_descr + nwords >= GC_ed_size) {
+    	ext_descr * new;
+    	size_t new_size;
+    	word ed_size = GC_ed_size;
+    	
+    	UNLOCK();
+        ENABLE_SIGNALS();
+    	if (ed_size == 0) {
+    	    new_size = ED_INITIAL_SIZE;
+    	} else {
+    	    new_size = 2 * ed_size;
+    	    if (new_size > MAX_ENV) return(-1);
+    	} 
+    	new = (ext_descr *) GC_malloc_atomic(new_size * sizeof(ext_descr));
+    	if (new == 0) return(-1);
+    	DISABLE_SIGNALS();
+        LOCK();
+        if (ed_size == GC_ed_size) {
+            if (GC_avail_descr != 0) {
+    	        BCOPY(GC_ext_descriptors, new,
+    	              GC_avail_descr * sizeof(ext_descr));
+    	    }
+    	    GC_ed_size = new_size;
+    	    GC_ext_descriptors = new;
+    	}  /* else another thread already resized it in the meantime */
+    }
+    result = GC_avail_descr;
+    for (i = 0; i < nwords-1; i++) {
+        GC_ext_descriptors[result + i].ed_bitmap = bm[i];
+        GC_ext_descriptors[result + i].ed_continued = TRUE;
+    }
+    last_part = bm[i];
+    /* Clear irrelevant bits. */
+    extra_bits = nwords * WORDSZ - nbits;
+    last_part <<= extra_bits;
+    last_part >>= extra_bits;
+    GC_ext_descriptors[result + i].ed_bitmap = last_part;
+    GC_ext_descriptors[result + i].ed_continued = FALSE;
+    GC_avail_descr += nwords;
+    UNLOCK();
+    ENABLE_SIGNALS();
+    return(result);
+}
+
+/* Table of bitmap descriptors for n word long all pointer objects.	*/
+GC_descr GC_bm_table[WORDSZ/2];
+	
+/* Return a descriptor for the concatenation of 2 nwords long objects,	*/
+/* each of which is described by descriptor.				*/
+/* The result is known to be short enough to fit into a bitmap		*/
+/* descriptor.								*/
+/* Descriptor is a DS_LENGTH or DS_BITMAP descriptor.			*/
+GC_descr GC_double_descr(descriptor, nwords)
+register GC_descr descriptor;
+register word nwords;
+{
+    if (descriptor && DS_TAGS == DS_LENGTH) {
+        descriptor = GC_bm_table[BYTES_TO_WORDS((word)descriptor)];
+    };
+    descriptor |= (descriptor & ~DS_TAGS) >> nwords;
+    return(descriptor);
+}
+
+complex_descriptor * GC_make_sequence_descriptor();
+
+/* Build a descriptor for an array with nelements elements,	*/
+/* each of which can be described by a simple descriptor.	*/
+/* We try to optimize some common cases.			*/
+/* If the result is COMPLEX, then a complex_descr* is returned  */
+/* in *complex_d.							*/
+/* If the result is LEAF, then we built a LeafDescriptor in	*/
+/* the structure pointed to by leaf.				*/
+/* The tag in the leaf structure is not set.			*/
+/* If the result is SIMPLE, then a GC_descr			*/
+/* is returned in *simple_d.					*/
+/* If the result is NO_MEM, then				*/
+/* we failed to allocate the descriptor.			*/
+/* The implementation knows that DS_LENGTH is 0.		*/
+/* *leaf, *complex_d, and *simple_d may be used as temporaries	*/
+/* during the construction.					*/
+# define COMPLEX 2
+# define LEAF 1
+# define SIMPLE 0
+# define NO_MEM (-1)
+int GC_make_array_descriptor(nelements, size, descriptor,
+			     simple_d, complex_d, leaf)
+word size;
+word nelements;
+GC_descr descriptor;
+GC_descr *simple_d;
+complex_descriptor **complex_d;
+struct LeafDescriptor * leaf;
+{
+#   define OPT_THRESHOLD 50
+	/* For larger arrays, we try to combine descriptors of adjacent	*/
+	/* descriptors to speed up marking, and to reduce the amount	*/
+	/* of space needed on the mark stack.				*/
+    if ((descriptor & DS_TAGS) == DS_LENGTH) {
+      if ((word)descriptor == size) {
+    	*simple_d = nelements * descriptor;
+    	return(SIMPLE);
+      } else if ((word)descriptor == 0) {
+        *simple_d = (GC_descr)0;
+        return(SIMPLE);
+      }
+    }
+    if (nelements <= OPT_THRESHOLD) {
+      if (nelements <= 1) {
+        if (nelements == 1) {
+            *simple_d = descriptor;
+            return(SIMPLE);
+        } else {
+            *simple_d = (GC_descr)0;
+            return(SIMPLE);
+        }
+      }
+    } else if (size <= BITMAP_BITS/2
+    	       && (descriptor & DS_TAGS) != DS_PROC
+    	       && (size & (sizeof(word)-1)) == 0) {
+      int result =      
+          GC_make_array_descriptor(nelements/2, 2*size,
+      				   GC_double_descr(descriptor,
+      				   		   BYTES_TO_WORDS(size)),
+      				   simple_d, complex_d, leaf);
+      if ((nelements & 1) == 0) {
+          return(result);
+      } else {
+          struct LeafDescriptor * one_element =
+              (struct LeafDescriptor *)
+        	GC_malloc_atomic(sizeof(struct LeafDescriptor));
+          
+          if (result == NO_MEM || one_element == 0) return(NO_MEM);
+          one_element -> ld_tag = LEAF_TAG;
+          one_element -> ld_size = size;
+          one_element -> ld_nelements = 1;
+          one_element -> ld_descriptor = descriptor;
+          switch(result) {
+            case SIMPLE:
+            {
+              struct LeafDescriptor * beginning =
+                (struct LeafDescriptor *)
+        	  GC_malloc_atomic(sizeof(struct LeafDescriptor));
+              if (beginning == 0) return(NO_MEM);
+              beginning -> ld_tag = LEAF_TAG;
+              beginning -> ld_size = size;
+              beginning -> ld_nelements = 1;
+              beginning -> ld_descriptor = *simple_d;
+              *complex_d = GC_make_sequence_descriptor(
+              			(complex_descriptor *)beginning,
+              			(complex_descriptor *)one_element);
+              break;
+            }
+            case LEAF:
+            {
+              struct LeafDescriptor * beginning =
+                (struct LeafDescriptor *)
+        	  GC_malloc_atomic(sizeof(struct LeafDescriptor));
+              if (beginning == 0) return(NO_MEM);
+              beginning -> ld_tag = LEAF_TAG;
+              beginning -> ld_size = leaf -> ld_size;
+              beginning -> ld_nelements = leaf -> ld_nelements;
+              beginning -> ld_descriptor = leaf -> ld_descriptor;
+              *complex_d = GC_make_sequence_descriptor(
+              			(complex_descriptor *)beginning,
+              			(complex_descriptor *)one_element);
+              break;
+            }
+            case COMPLEX:
+              *complex_d = GC_make_sequence_descriptor(
+              			*complex_d,
+              			(complex_descriptor *)one_element);
+              break;
+          }
+          return(COMPLEX);
+      }
+    }
+    {
+        leaf -> ld_size = size;
+        leaf -> ld_nelements = nelements;
+        leaf -> ld_descriptor = descriptor;
+        return(LEAF);
+    }
+}
+
+complex_descriptor * GC_make_sequence_descriptor(first, second)
+complex_descriptor * first;
+complex_descriptor * second;
+{
+    struct SequenceDescriptor * result =
+        (struct SequenceDescriptor *)
+        	GC_malloc(sizeof(struct SequenceDescriptor));
+    /* Can't result in overly conservative marking, since tags are	*/
+    /* very small integers. Probably faster than maintaining type	*/
+    /* info.								*/    
+    if (result != 0) {
+    	result -> sd_tag = SEQUENCE_TAG;
+        result -> sd_first = first;
+        result -> sd_second = second;
+    }
+    return((complex_descriptor *)result);
+}
+
+#ifdef UNDEFINED
+complex_descriptor * GC_make_complex_array_descriptor(nelements, descr)
+word nelements;
+complex_descriptor * descr;
+{
+    struct ComplexArrayDescriptor * result =
+        (struct ComplexArrayDescriptor *)
+        	GC_malloc(sizeof(struct ComplexArrayDescriptor));
+    
+    if (result != 0) {
+    	result -> ad_tag = ARRAY_TAG;
+        result -> ad_nelements = nelements;
+        result -> ad_element_descr = descr;
+    }
+    return((complex_descriptor *)result);
+}
+#endif
+
+ptr_t * GC_eobjfreelist;
+
+ptr_t * GC_arobjfreelist;
+
+struct hblk ** GC_ereclaim_list;
+
+struct hblk ** GC_arreclaim_list;
+
+mse * GC_typed_mark_proc();
+
+mse * GC_array_mark_proc();
+
+GC_descr GC_generic_array_descr;
+
+/* Caller does not hold allocation lock. */
+void GC_init_explicit_typing()
+{
+    register int i;
+    DCL_LOCK_STATE;
+
+    
+#   ifdef PRINTSTATS
+     	if (sizeof(struct LeafDescriptor) % sizeof(word) != 0)
+     	    ABORT("Bad leaf descriptor size");
+#   endif
+    DISABLE_SIGNALS();
+    LOCK();
+    if (GC_explicit_typing_initialized) {
+      UNLOCK();
+      ENABLE_SIGNALS();
+      return;
+    }
+    GC_explicit_typing_initialized = TRUE;
+    /* Set up object kind with simple indirect descriptor. */
+      GC_eobjfreelist = (ptr_t *)
+          GC_generic_malloc_inner((MAXOBJSZ+1)*sizeof(ptr_t), PTRFREE);
+      if (GC_eobjfreelist == 0) ABORT("Couldn't allocate GC_eobjfreelist");
+      BZERO(GC_eobjfreelist, (MAXOBJSZ+1)*sizeof(ptr_t));
+      GC_ereclaim_list = (struct hblk **)
+    	GC_generic_malloc_inner((MAXOBJSZ+1)*sizeof(struct hblk *), PTRFREE);
+      if (GC_ereclaim_list == 0)
+      				ABORT("Couldn't allocate GC_ereclaim_list");
+      BZERO(GC_ereclaim_list, (MAXOBJSZ+1)*sizeof(struct hblk *));
+      GC_explicit_kind = GC_n_kinds++;
+      GC_obj_kinds[GC_explicit_kind].ok_freelist = GC_eobjfreelist;
+      GC_obj_kinds[GC_explicit_kind].ok_reclaim_list = GC_ereclaim_list;
+      GC_obj_kinds[GC_explicit_kind].ok_descriptor =
+    		(((word)WORDS_TO_BYTES(-1)) | DS_PER_OBJECT);
+      GC_obj_kinds[GC_explicit_kind].ok_relocate_descr = TRUE;
+      GC_obj_kinds[GC_explicit_kind].ok_init = TRUE;
+    		/* Descriptors are in the last word of the object. */
+      GC_typed_mark_proc_index = GC_n_mark_procs;
+      GC_mark_procs[GC_typed_mark_proc_index] = GC_typed_mark_proc;
+      GC_n_mark_procs++;
+        /* Moving this up breaks DEC AXP compiler.      */
+    /* Set up object kind with array descriptor. */
+      GC_arobjfreelist = (ptr_t *)
+          GC_generic_malloc_inner((MAXOBJSZ+1)*sizeof(ptr_t), PTRFREE);
+      if (GC_arobjfreelist == 0) ABORT("Couldn't allocate GC_arobjfreelist");
+      BZERO(GC_arobjfreelist, (MAXOBJSZ+1)*sizeof(ptr_t));
+      GC_arreclaim_list = (struct hblk **)
+    	GC_generic_malloc_inner((MAXOBJSZ+1)*sizeof(struct hblk *), PTRFREE);
+      if (GC_arreclaim_list == 0) ABORT("Couldn't allocate GC_arreclaim_list");
+      BZERO(GC_arreclaim_list, (MAXOBJSZ+1)*sizeof(struct hblk *));
+      if (GC_arreclaim_list == 0) ABORT("Couldn't allocate GC_arreclaim_list");
+      if (GC_n_mark_procs >= MAX_MARK_PROCS)
+      		ABORT("No slot for array mark proc");
+      GC_array_mark_proc_index = GC_n_mark_procs++;
+      if (GC_n_kinds >= MAXOBJKINDS)
+      		ABORT("No kind available for array objects");
+      GC_array_kind = GC_n_kinds++;
+      GC_obj_kinds[GC_array_kind].ok_freelist = GC_arobjfreelist;
+      GC_obj_kinds[GC_array_kind].ok_reclaim_list = GC_arreclaim_list;
+      GC_obj_kinds[GC_array_kind].ok_descriptor =
+    		MAKE_PROC(GC_array_mark_proc_index, 0);;
+      GC_obj_kinds[GC_array_kind].ok_relocate_descr = FALSE;
+      GC_obj_kinds[GC_array_kind].ok_init = TRUE;
+    		/* Descriptors are in the last word of the object. */
+            GC_mark_procs[GC_array_mark_proc_index] = GC_array_mark_proc;
+      for (i = 0; i < WORDSZ/2; i++) {
+          GC_descr d = (((word)(-1)) >> (WORDSZ - i)) << (WORDSZ - i);
+          d |= DS_BITMAP;
+          GC_bm_table[i] = d;
+      }
+      GC_generic_array_descr = MAKE_PROC(GC_array_mark_proc_index, 0); 
+    UNLOCK();
+    ENABLE_SIGNALS();
+}
+
+mse * GC_typed_mark_proc(addr, mark_stack_ptr, mark_stack_limit, env)
+register word * addr;
+register mse * mark_stack_ptr;
+mse * mark_stack_limit;
+word env;
+{
+    register word bm = GC_ext_descriptors[env].ed_bitmap;
+    register word * current_p = addr;
+    register word current;
+    register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
+    register ptr_t least_ha = GC_least_plausible_heap_addr;
+    
+    for (; bm != 0; bm >>= 1, current_p++) {
+    	if (bm & 1) {
+    	    current = *current_p;
+    	    if ((ptr_t)current >= least_ha && (ptr_t)current <= greatest_ha) {
+    	        PUSH_CONTENTS(current, mark_stack_ptr, mark_stack_limit);
+    	    }
+    	}
+    }
+    if (GC_ext_descriptors[env].ed_continued) {
+        /* Push an entry with the rest of the descriptor back onto the	*/
+        /* stack.  Thus we never do too much work at once.  Note that	*/
+        /* we also can't overflow the mark stack unless we actually 	*/
+        /* mark something.						*/
+        mark_stack_ptr++;
+        if (mark_stack_ptr >= mark_stack_limit) {
+            mark_stack_ptr = GC_signal_mark_stack_overflow(mark_stack_ptr);
+        }
+        mark_stack_ptr -> mse_start = addr + WORDSZ;
+        mark_stack_ptr -> mse_descr =
+        	MAKE_PROC(GC_typed_mark_proc_index, env+1);
+    }
+    return(mark_stack_ptr);
+}
+
+/* Return the size of the object described by d.  It would be faster to	*/
+/* store this directly, or to compute it as part of			*/
+/* GC_push_complex_descriptor, but hopefully it doesn't matter.		*/
+word GC_descr_obj_size(d)
+register complex_descriptor *d;
+{
+    switch(d -> TAG) {
+      case LEAF_TAG:
+      	return(d -> ld.ld_nelements * d -> ld.ld_size);
+      case ARRAY_TAG:
+        return(d -> ad.ad_nelements
+               * GC_descr_obj_size(d -> ad.ad_element_descr));
+      case SEQUENCE_TAG:
+        return(GC_descr_obj_size(d -> sd.sd_first)
+               + GC_descr_obj_size(d -> sd.sd_second));
+      default:
+        ABORT("Bad complex descriptor");
+        /*NOTREACHED*/
+    }
+}
+
+/* Push descriptors for the object at addr with complex descriptor d	*/
+/* onto the mark stack.  Return 0 if the mark stack overflowed.  	*/
+mse * GC_push_complex_descriptor(addr, d, msp, msl)
+word * addr;
+register complex_descriptor *d;
+register mse * msp;
+mse * msl;
+{
+    register ptr_t current = (ptr_t) addr;
+    register word nelements;
+    register word sz;
+    register word i;
+    
+    switch(d -> TAG) {
+      case LEAF_TAG:
+        {
+          register GC_descr descr = d -> ld.ld_descriptor;
+          
+          nelements = d -> ld.ld_nelements;
+          if (msl - msp <= (ptrdiff_t)nelements) return(0);
+          sz = d -> ld.ld_size;
+          for (i = 0; i < nelements; i++) {
+              msp++;
+              msp -> mse_start = (word *)current;
+              msp -> mse_descr = descr;
+              current += sz;
+          }
+          return(msp);
+        }
+      case ARRAY_TAG:
+        {
+          register complex_descriptor *descr = d -> ad.ad_element_descr;
+          
+          nelements = d -> ad.ad_nelements;
+          sz = GC_descr_obj_size(descr);
+          for (i = 0; i < nelements; i++) {
+              msp = GC_push_complex_descriptor((word *)current, descr,
+              					msp, msl);
+              if (msp == 0) return(0);
+              current += sz;
+          }
+          return(msp);
+        }
+      case SEQUENCE_TAG:
+        {
+          sz = GC_descr_obj_size(d -> sd.sd_first);
+          msp = GC_push_complex_descriptor((word *)current, d -> sd.sd_first,
+          				   msp, msl);
+          if (msp == 0) return(0);
+          current += sz;
+          msp = GC_push_complex_descriptor((word *)current, d -> sd.sd_second,
+          				   msp, msl);
+          return(msp);
+        }
+      default:
+        ABORT("Bad complex descriptor");
+        /*NOTREACHED*/
+    }
+}
+
+/*ARGSUSED*/
+mse * GC_array_mark_proc(addr, mark_stack_ptr, mark_stack_limit, env)
+register word * addr;
+register mse * mark_stack_ptr;
+mse * mark_stack_limit;
+word env;
+{
+    register hdr * hhdr = HDR(addr);
+    register word sz = hhdr -> hb_sz;
+    register complex_descriptor * descr = (complex_descriptor *)(addr[sz-1]);
+    mse * orig_mark_stack_ptr = mark_stack_ptr;
+    mse * new_mark_stack_ptr;
+    
+    if (descr == 0) {
+    	/* Found a reference to a free list entry.  Ignore it. */
+    	return(orig_mark_stack_ptr);
+    }
+    /* In use counts were already updated when array descriptor was	*/
+    /* pushed.  Here we only replace it by subobject descriptors, so 	*/
+    /* no update is necessary.						*/
+    new_mark_stack_ptr = GC_push_complex_descriptor(addr, descr,
+    						    mark_stack_ptr,
+    						    mark_stack_limit-1);
+    if (new_mark_stack_ptr == 0) {
+    	/* Doesn't fit.  Conservatively push the whole array as a unit	*/
+    	/* and request a mark stack expansion.				*/
+    	/* This cannot cause a mark stack overflow, since it replaces	*/
+    	/* the original array entry.					*/
+    	GC_mark_stack_too_small = TRUE;
+    	new_mark_stack_ptr = orig_mark_stack_ptr + 1;
+    	new_mark_stack_ptr -> mse_start = addr;
+    	new_mark_stack_ptr -> mse_descr = WORDS_TO_BYTES(sz) | DS_LENGTH;
+    } else {
+        /* Push descriptor itself */
+        new_mark_stack_ptr++;
+        new_mark_stack_ptr -> mse_start = addr + sz - 1;
+        new_mark_stack_ptr -> mse_descr = sizeof(word) | DS_LENGTH;
+    }
+    return(new_mark_stack_ptr);
+}
+
+#if defined(__STDC__) || defined(__cplusplus)
+  GC_descr GC_make_descriptor(GC_bitmap bm, size_t len)
+#else
+  GC_descr GC_make_descriptor(bm, len)
+  GC_bitmap bm;
+  size_t len;
+#endif
+{
+    register signed_word last_set_bit = len - 1;
+    register word result;
+    register int i;
+#   define HIGH_BIT (((word)1) << (WORDSZ - 1))
+    
+    if (!GC_explicit_typing_initialized) GC_init_explicit_typing();
+    while (last_set_bit >= 0 && !GC_get_bit(bm, last_set_bit)) last_set_bit --;
+    if (last_set_bit < 0) return(0 /* no pointers */);
+#   if ALIGNMENT == CPP_WORDSZ/8
+    {
+      register bool all_bits_set = TRUE;
+      for (i = 0; i < last_set_bit; i++) {
+    	if (!GC_get_bit(bm, i)) {
+    	    all_bits_set = FALSE;
+    	    break;
+    	}
+      }
+      if (all_bits_set) {
+    	/* An initial section contains all pointers.  Use length descriptor. */
+        return(WORDS_TO_BYTES(last_set_bit+1) | DS_LENGTH);
+      }
+    }
+#   endif
+    if (last_set_bit < BITMAP_BITS) {
+    	/* Hopefully the common case.			*/
+    	/* Build bitmap descriptor (with bits reversed)	*/
+    	result = HIGH_BIT;
+    	for (i = last_set_bit - 1; i >= 0; i--) {
+    	    result >>= 1;
+    	    if (GC_get_bit(bm, i)) result |= HIGH_BIT;
+    	}
+    	result |= DS_BITMAP;
+    	return(result);
+    } else {
+    	signed_word index;
+    	
+    	index = GC_add_ext_descriptor(bm, (word)last_set_bit+1);
+    	if (index == -1) return(WORDS_TO_BYTES(last_set_bit+1) | DS_LENGTH);
+    				/* Out of memory: use conservative	*/
+    				/* approximation.			*/
+    	result = MAKE_PROC(GC_typed_mark_proc_index, (word)index);
+    	return(result);
+    }
+}
+
+ptr_t GC_clear_stack();
+
+#define GENERAL_MALLOC(lb,k) \
+    (extern_ptr_t)GC_clear_stack(GC_generic_malloc((word)lb, k))
+    
+#if defined(__STDC__) || defined(__cplusplus)
+  extern void * GC_malloc_explicitly_typed(size_t lb, GC_descr d)
+#else
+  extern char * GC_malloc_explicitly_typed(lb, d)
+  size_t lb;
+  GC_descr d;
+#endif
+{
+register ptr_t op;
+register ptr_t * opp;
+register word lw;
+DCL_LOCK_STATE;
+
+    lb += EXTRA_BYTES;
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+#       endif
+	opp = &(GC_eobjfreelist[lw]);
+	FASTLOCK();
+        if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {
+            FASTUNLOCK();
+            op = (ptr_t)GENERAL_MALLOC((word)lb, GC_explicit_kind);
+#	    ifdef MERGE_SIZES
+		lw = GC_size_map[lb];	/* May have been uninitialized.	*/            
+#	    endif
+        } else {
+            *opp = obj_link(op);
+            GC_words_allocd += lw;
+            FASTUNLOCK();
+        }
+   } else {
+       op = (ptr_t)GENERAL_MALLOC((word)lb, GC_explicit_kind);
+       lw = BYTES_TO_WORDS(GC_size(op));
+   }
+   ((word *)op)[lw - 1] = d;
+   return((extern_ptr_t) op);
+}
+
+#if defined(__STDC__) || defined(__cplusplus)
+  void * GC_calloc_explicitly_typed(size_t n,
+  				    size_t lb,
+  				    GC_descr d)
+#else
+  char * GC_calloc_explicitly_typed(n, lb, d)
+  size_t n;
+  size_t lb;
+  GC_descr d;
+#endif
+{
+register ptr_t op;
+register ptr_t * opp;
+register word lw;
+GC_descr simple_descr;
+complex_descriptor *complex_descr;
+register int descr_type;
+struct LeafDescriptor leaf;
+DCL_LOCK_STATE;
+
+    descr_type = GC_make_array_descriptor((word)n, (word)lb, d,
+    					  &simple_descr, &complex_descr, &leaf);
+    switch(descr_type) {
+    	case NO_MEM: return(0);
+    	case SIMPLE: return(GC_malloc_explicitly_typed(n*lb, simple_descr));
+    	case LEAF:
+    	    lb *= n;
+    	    lb += sizeof(struct LeafDescriptor) + EXTRA_BYTES;
+    	    break;
+    	case COMPLEX:
+    	    lb *= n;
+    	    lb += EXTRA_BYTES;
+    	    break;
+    }
+    if( SMALL_OBJ(lb) ) {
+#       ifdef MERGE_SIZES
+	  lw = GC_size_map[lb];
+#	else
+	  lw = ROUNDED_UP_WORDS(lb);
+#       endif
+	opp = &(GC_arobjfreelist[lw]);
+	FASTLOCK();
+        if( !FASTLOCK_SUCCEEDED() || (op = *opp) == 0 ) {
+            FASTUNLOCK();
+            op = (ptr_t)GENERAL_MALLOC((word)lb, GC_array_kind);
+#	    ifdef MERGE_SIZES
+		lw = GC_size_map[lb];	/* May have been uninitialized.	*/            
+#	    endif
+        } else {
+            *opp = obj_link(op);
+            GC_words_allocd += lw;
+            FASTUNLOCK();
+        }
+   } else {
+       op = (ptr_t)GENERAL_MALLOC((word)lb, GC_array_kind);
+       lw = BYTES_TO_WORDS(GC_size(op));
+   }
+   if (descr_type == LEAF) {
+       /* Set up the descriptor inside the object itself. */
+       VOLATILE struct LeafDescriptor * lp =
+           (struct LeafDescriptor *)
+               ((word *)op
+                + lw - (BYTES_TO_WORDS(sizeof(struct LeafDescriptor)) + 1));
+                
+       lp -> ld_tag = LEAF_TAG;
+       lp -> ld_size = leaf.ld_size;
+       lp -> ld_nelements = leaf.ld_nelements;
+       lp -> ld_descriptor = leaf.ld_descriptor;
+       ((VOLATILE word *)op)[lw - 1] = (word)lp;
+   } else {
+       extern unsigned GC_finalization_failures;
+       unsigned ff = GC_finalization_failures;
+       
+       ((word *)op)[lw - 1] = (word)complex_descr;
+       /* Make sure the descriptor is cleared once there is any danger	*/
+       /* it may have been collected.					*/
+       (void)
+         GC_general_register_disappearing_link((extern_ptr_t *)
+         					  ((word *)op+lw-1),
+       					          (extern_ptr_t) op);
+       if (ff != GC_finalization_failures) {
+           /* We may have failed to register op due to lack of memory.	*/
+           /* We were out of memory very recently, so we can safely 	*/
+           /* punt.							*/
+           ((word *)op)[lw - 1] = 0;
+           return(0);
+       }			          
+   }
+   return((extern_ptr_t) op);
+}