gc5.3 tarball importgc5_3

63 files changed, 6038 insertions, 737 deletions

diff --git a/Makefile b/Makefile
index cfbfc452..4287a21d 100644
--- a/Makefile
+++ b/Makefile
@@ -7,9 +7,10 @@
 #      	 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.
-ABI_FLAG=
+ABI_FLAG= 
 CC=cc $(ABI_FLAG)
-CXX=CC $(ABI_FLAG)
+HOSTCC=$(CC)
+CXX=g++ $(ABI_FLAG)
 AS=as $(ABI_FLAG)
 #  The above doesn't work with gas, which doesn't run cpp.
 #  Define AS as `gcc -c -x assembler-with-cpp' instead.
@@ -35,6 +36,8 @@ CFLAGS= -O -DATOMIC_UNCOLLECTABLE -DNO_SIGNALS -DNO_EXECUTE_PERMISSION -DALL_INT
 # -D_SOLARIS_PTHREADS enables support for Solaris pthreads.
 #   Define SOLARIS_THREADS as well.
 # -DIRIX_THREADS enables support for Irix pthreads.  See README.irix.
+# -DHPUX_THREADS enables support for HP/UX 11 pthreads.
+#  Also requires -D_REENTRANT. See README.hp.
 # -DLINUX_THREADS enables support for Xavier Leroy's Linux threads.
 #   see README.linux.  -D_REENTRANT may also be required.
 # -DALL_INTERIOR_POINTERS allows all pointers to the interior
@@ -56,6 +59,8 @@ CFLAGS= -O -DATOMIC_UNCOLLECTABLE -DNO_SIGNALS -DNO_EXECUTE_PERMISSION -DALL_INT
 #   implementations, and it sometimes has a significant performance
 #   impact.  However, it is dangerous for many not-quite-ANSI C
 #   programs that call things like printf in asynchronous signal handlers.
+#   This is on by default.  Turning it off has not been extensively tested with
+#   compilers that reorder stores.  It should have been.
 # -DNO_EXECUTE_PERMISSION may cause some or all of the heap to not
 #   have execute permission, i.e. it may be impossible to execute
 #   code from the heap.  Currently this only affects the incremental
@@ -110,12 +115,6 @@ CFLAGS= -O -DATOMIC_UNCOLLECTABLE -DNO_SIGNALS -DNO_EXECUTE_PERMISSION -DALL_INT
 #   like a pointer, print both the address containing the value, and the
 #   value of the near-bogus-pointer.  Can be used to identifiy regions of
 #   memory that are likely to contribute misidentified pointers.
-# -DOLD_BLOCK_ALLOC Use the old, possibly faster, large block
-#   allocation strategy.  The new strategy tries harder to minimize
-#   fragmentation, sometimes at the expense of spending more time in the
-#   large block allocator and/or collecting more frequently.
-#   If you expect the allocator to promptly use an explicitly expanded
-#   heap, this is highly recommended.
 # -DKEEP_BACK_PTRS Add code to save back pointers in debugging headers
 #   for objects allocated with the debugging allocator.  If all objects
 #   through GC_MALLOC with GC_DEBUG defined, this allows the client
@@ -130,6 +129,24 @@ CFLAGS= -O -DATOMIC_UNCOLLECTABLE -DNO_SIGNALS -DNO_EXECUTE_PERMISSION -DALL_INT
 # -DCHECKSUMS reports on erroneously clear dirty bits, and unexpectedly
 #   altered stubborn objects, at substantial performance cost.
 #   Use only for debugging of the incremental collector.
+# -DGC_GCJ_SUPPORT includes support for gcj (and possibly other systems
+#   that include a pointer to a type descriptor in each allocated object).
+#   Building this way requires an ANSI C compiler.
+# -DUSE_I686_PREFETCH causes the collector to issue Pentium III style
+#   prefetch instructions.  No effect except on X86 Linux platforms.
+#   Assumes a very recent gcc-compatible compiler and assembler.
+#   (Gas prefetcht0 support was added around May 1999.)
+#   Empirically the code appears to still run correctly on Pentium II
+#   processors, though with no performance benefit.  May not run on other
+#   X86 processors?  In some cases this improves performance by
+#   15% or so.
+# -DUSE_3DNOW_PREFETCH causes the collector to issue AMD 3DNow style
+#   prefetch instructions.  Same restrictions as USE_I686_PREFETCH.
+#   UNTESTED!!
+# -DUSE_LD_WRAP in combination with the gld flags listed in README.linux
+#   causes the collector some system and pthread calls in a more transparent
+#   fashion than the usual macro-based approach.  Requires GNU ld, and
+#   currently probably works only with Linux.
 #
 
 
@@ -149,9 +166,9 @@ RANLIB= ranlib
 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 dbg_mlc.o malloc.o stubborn.o checksums.o solaris_threads.o irix_threads.o linux_threads.o typd_mlc.o ptr_chck.o mallocx.o solaris_pthreads.o
+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 dbg_mlc.o malloc.o stubborn.o checksums.o solaris_threads.o hpux_irix_threads.o linux_threads.o typd_mlc.o ptr_chck.o mallocx.o solaris_pthreads.o gcj_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 irix_threads.c linux_threads.c typd_mlc.c ptr_chck.c mallocx.c solaris_pthreads.c
+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 hpux_irix_threads.c linux_threads.c typd_mlc.c ptr_chck.c mallocx.c solaris_pthreads.c gcj_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/private/cord_pos.h cord/de_win.c cord/de_win.h cord/de_cmds.h cord/de_win.ICO cord/de_win.RC cord/SCOPTIONS.amiga cord/SMakefile.amiga
 
@@ -163,7 +180,8 @@ SRCS= $(CSRCS) mips_sgi_mach_dep.s rs6000_mach_dep.s alpha_mach_dep.s \
     threadlibs.c if_mach.c if_not_there.c gc_cpp.cc gc_cpp.h weakpointer.h \
     gcc_support.c mips_ultrix_mach_dep.s include/gc_alloc.h gc_alloc.h \
     include/new_gc_alloc.h include/javaxfc.h sparc_sunos4_mach_dep.s \
-    solaris_threads.h backptr.h $(CORD_SRCS)
+    solaris_threads.h backptr.h hpux_test_and_clear.s include/gc_gcj.h \
+    dbg_mlc.h $(CORD_SRCS)
 
 OTHER_FILES= Makefile PCR-Makefile OS2_MAKEFILE NT_MAKEFILE BCC_MAKEFILE \
            README test.c test_cpp.cc setjmp_t.c SMakefile.amiga \
@@ -180,7 +198,7 @@ OTHER_FILES= Makefile PCR-Makefile OS2_MAKEFILE NT_MAKEFILE BCC_MAKEFILE \
            add_gc_prefix.c README.solaris2 README.sgi README.hp README.uts \
 	   win32_threads.c NT_THREADS_MAKEFILE gc.mak README.dj Makefile.dj \
 	   README.alpha README.linux version.h Makefile.DLLs \
-	   WCC_MAKEFILE nursery.c nursery.h gc_copy_descr.h \
+	   WCC_MAKEFILE nursery.c include/gc_nursery.h include/gc_copy_descr.h \
 	   include/leak_detector.h
 
 CORD_INCLUDE_FILES= $(srcdir)/gc.h $(srcdir)/cord/cord.h $(srcdir)/cord/ec.h \
@@ -242,7 +260,7 @@ gc_cpp.o: $(srcdir)/gc_cpp.cc $(srcdir)/gc_cpp.h $(srcdir)/gc.h Makefile
 test_cpp: $(srcdir)/test_cpp.cc $(srcdir)/gc_cpp.h gc_cpp.o $(srcdir)/gc.h \
 base_lib $(UTILS)
 	rm -f test_cpp
-	./if_mach HP_PA "" $(CXX) $(CXXFLAGS) -o test_cpp $(srcdir)/test_cpp.cc gc_cpp.o gc.a -ldld
+	./if_mach HP_PA HPUX $(CXX) $(CXXFLAGS) -o test_cpp $(srcdir)/test_cpp.cc gc_cpp.o gc.a -ldld `./threadlibs`
 	./if_not_there test_cpp $(CXX) $(CXXFLAGS) -o test_cpp $(srcdir)/test_cpp.cc gc_cpp.o gc.a `./threadlibs`
 
 c++: gc_cpp.o $(srcdir)/gc_cpp.h test_cpp
@@ -276,7 +294,7 @@ libirixgc.so: $(OBJS) dyn_load.o
 
 # Linux shared library version of the collector
 liblinuxgc.so: $(OBJS) dyn_load.o
-	gcc -shared -o liblinuxgc.so $(OBJS) dyn_load.o -lo
+	gcc -shared -o liblinuxgc.so $(OBJS) dyn_load.o
 	ln liblinuxgc.so libgc.so
 
 # Alternative Linux rule.  This is preferable, but is likely to break the
@@ -298,13 +316,18 @@ mach_dep.o: $(srcdir)/mach_dep.c $(srcdir)/mips_sgi_mach_dep.s $(srcdir)/mips_ul
 	./if_mach MIPS RISCOS $(AS) -o mach_dep.o $(srcdir)/mips_ultrix_mach_dep.s
 	./if_mach MIPS ULTRIX $(AS) -o mach_dep.o $(srcdir)/mips_ultrix_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 ALPHA "" $(AS) -o mach_dep.o $(srcdir)/alpha_mach_dep.s
+#	alpha_mach_dep.s assumes that pointers are not saved in fp registers.
+#	Gcc on a 21264 can spill pointers to fp registers.  Oops.
 	./if_mach SPARC SUNOS5 $(AS) -o mach_dep.o $(srcdir)/sparc_mach_dep.s
 	./if_mach SPARC SUNOS4 $(AS) -o mach_dep.o $(srcdir)/sparc_sunos4_mach_dep.s
 	./if_mach SPARC OPENBSD $(AS) -o mach_dep.o $(srcdir)/sparc_sunos4_mach_dep.s
+	./if_mach HP_PA HPUX $(AS) -o hpux_test_and_clear.o $(srcdir)/hpux_test_and_clear.s
+	./if_mach HP_PA HPUX $(CC) -c -o md_tmp.o $(SPECIALCFLAGS) $(srcdir)/mach_dep.c
+	./if_mach HP_PA HPUX ld -r -o mach_dep.o md_tmp.o hpux_test_and_clear.o
 	./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 $(UTILS)
+mark_rts.o: $(srcdir)/mark_rts.c $(UTILS)
 	rm -f mark_rts.o
 	-./if_mach ALPHA OSF1 $(CC) -c $(CFLAGS) -Wo,-notail $(srcdir)/mark_rts.c
 	./if_not_there mark_rts.o $(CC) -c $(CFLAGS) $(srcdir)/mark_rts.c
@@ -329,27 +352,28 @@ cord/cordprnt.o: $(srcdir)/cord/cordprnt.c $(CORD_INCLUDE_FILES)
 cord/cordtest: $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a $(UTILS)
 	rm -f cord/cordtest
 	./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -lucb
-	./if_mach HP_PA "" $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -ldld
+	./if_mach HP_PA HPUX $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a -ldld `./threadlibs`
 	./if_not_there cord/cordtest $(CC) $(CFLAGS) -o cord/cordtest $(srcdir)/cord/cordtest.c $(CORD_OBJS) gc.a `./threadlibs`
 
 cord/de: $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(UTILS)
 	rm -f cord/de
 	./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -lucb `./threadlibs`
-	./if_mach HP_PA "" $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -ldld
+	./if_mach HP_PA HPUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) -ldld `./threadlibs`
 	./if_mach RS6000 "" $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
 	./if_mach I386 LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
-	./if_mach ALPHA LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
+	./if_mach ALPHA LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
+	./if_mach IA64 LINUX $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses `./threadlibs`
 	./if_mach M68K AMIGA $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a -lcurses
 	./if_not_there cord/de $(CC) $(CFLAGS) -o cord/de $(srcdir)/cord/de.c cord/cordbscs.o cord/cordxtra.o gc.a $(CURSES) `./threadlibs`
 
 if_mach: $(srcdir)/if_mach.c $(srcdir)/gcconfig.h
-	$(CC) $(CFLAGS) -o if_mach $(srcdir)/if_mach.c
+	$(HOSTCC) $(CFLAGS) -o if_mach $(srcdir)/if_mach.c
 
 threadlibs: $(srcdir)/threadlibs.c $(srcdir)/gcconfig.h Makefile
-	$(CC) $(CFLAGS) -o threadlibs $(srcdir)/threadlibs.c
+	$(HOSTCC) $(CFLAGS) -o threadlibs $(srcdir)/threadlibs.c
 
 if_not_there: $(srcdir)/if_not_there.c
-	$(CC) $(CFLAGS) -o if_not_there $(srcdir)/if_not_there.c
+	$(HOSTCC) $(CFLAGS) -o if_not_there $(srcdir)/if_not_there.c
 
 clean: 
 	rm -f gc.a *.o gctest gctest_dyn_link test_cpp \
@@ -357,16 +381,16 @@ clean:
 	      threadlibs $(CORD_OBJS) cord/cordtest cord/de
 	-rm -f *~
 
-gctest: test.o gc.a if_mach if_not_there
+gctest: test.o gc.a $(UTILS)
 	rm -f gctest
 	./if_mach SPARC DRSNX $(CC) $(CFLAGS) -o gctest  test.o gc.a -lucb
-	./if_mach HP_PA "" $(CC) $(CFLAGS) -o gctest  test.o gc.a -ldld
+	./if_mach HP_PA HPUX $(CC) $(CFLAGS) -o gctest  test.o gc.a -ldld `./threadlibs`
 	./if_not_there gctest $(CC) $(CFLAGS) -o gctest test.o gc.a `./threadlibs`
 
 # 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
+setjmp_test: $(srcdir)/setjmp_t.c $(srcdir)/gc.h $(UTILS)
 	$(CC) $(CFLAGS) -o setjmp_test $(srcdir)/setjmp_t.c
 
 test:  KandRtest cord/cordtest
@@ -382,7 +406,7 @@ add_gc_prefix: add_gc_prefix.c
 
 gc.tar: $(SRCS) $(OTHER_FILES) add_gc_prefix
 	./add_gc_prefix $(SRCS) $(OTHER_FILES) > /tmp/gc.tar-files
-	(cd $(srcdir)/.. ; tar cvfh - `cat /tmp/gc.tar-files`) > gc.tar
+	tar cvfh gc.tar `cat /tmp/gc.tar-files`
 
 pc_gc.tar: $(SRCS) $(OTHER_FILES)
 	tar cvfX pc_gc.tar pc_excludes $(SRCS) $(OTHER_FILES)
diff --git a/NT_MAKEFILE b/NT_MAKEFILE
index 52f6f4ae..a642262a 100644
--- a/NT_MAKEFILE
+++ b/NT_MAKEFILE
@@ -10,10 +10,10 @@ OBJS= alloc.obj reclaim.obj allchblk.obj misc.obj mach_dep.obj os_dep.obj mark_r
 all: gctest.exe cord\de.exe test_cpp.exe
 
 .c.obj:
-	$(cc) $(cdebug) $(cflags) $(cvars) -DSMALL_CONFIG -DSILENT -DALL_INTERIOR_POINTERS -D__STDC__ $*.c /Fo$*.obj
+	$(cc) $(cdebug) $(cflags) $(cvars) -DSILENT -DALL_INTERIOR_POINTERS -D__STDC__ $*.c /Fo$*.obj
 
 .cpp.obj:
-	$(cc) $(cdebug) $(cflags) $(cvars) -DSMALL_CONFIG -DSILENT -DALL_INTERIOR_POINTERS $*.CPP /Fo$*.obj
+	$(cc) $(cdebug) $(cflags) $(cvars) -DSILENT -DALL_INTERIOR_POINTERS $*.CPP /Fo$*.obj
 
 $(OBJS) test.obj: gc_priv.h gc_hdrs.h gc.h
 
diff --git a/NT_THREADS_MAKEFILE b/NT_THREADS_MAKEFILE
index 1f801cff..0fd22b70 120000..100644
--- a/NT_THREADS_MAKEFILE
+++ b/NT_THREADS_MAKEFILE
@@ -1 +1,2087 @@
-gc.mak
\ No newline at end of file
+# Microsoft Developer Studio Generated NMAKE File, Format Version 4.10
+# ** DO NOT EDIT **
+
+# TARGTYPE "Win32 (x86) Application" 0x0101
+# TARGTYPE "Win32 (x86) Dynamic-Link Library" 0x0102
+
+!IF "$(CFG)" == ""
+CFG=gctest - Win32 Release
+!MESSAGE No configuration specified.  Defaulting to cord - Win32 Debug.
+!ENDIF 
+
+!IF "$(CFG)" != "gc - Win32 Release" && "$(CFG)" != "gc - Win32 Debug" &&\
+ "$(CFG)" != "gctest - Win32 Release" && "$(CFG)" != "gctest - Win32 Debug" &&\
+ "$(CFG)" != "cord - Win32 Release" && "$(CFG)" != "cord - Win32 Debug"
+!MESSAGE Invalid configuration "$(CFG)" specified.
+!MESSAGE You can specify a configuration when running NMAKE on this makefile
+!MESSAGE by defining the macro CFG on the command line.  For example:
+!MESSAGE 
+!MESSAGE NMAKE /f "gc.mak" CFG="cord - Win32 Debug"
+!MESSAGE 
+!MESSAGE Possible choices for configuration are:
+!MESSAGE 
+!MESSAGE "gc - Win32 Release" (based on "Win32 (x86) Dynamic-Link Library")
+!MESSAGE "gc - Win32 Debug" (based on "Win32 (x86) Dynamic-Link Library")
+!MESSAGE "gctest - Win32 Release" (based on "Win32 (x86) Application")
+!MESSAGE "gctest - Win32 Debug" (based on "Win32 (x86) Application")
+!MESSAGE "cord - Win32 Release" (based on "Win32 (x86) Application")
+!MESSAGE "cord - Win32 Debug" (based on "Win32 (x86) Application")
+!MESSAGE 
+!ERROR An invalid configuration is specified.
+!ENDIF 
+
+!IF "$(OS)" == "Windows_NT"
+NULL=
+!ELSE 
+NULL=nul
+!ENDIF 
+################################################################################
+# Begin Project
+# PROP Target_Last_Scanned "gctest - Win32 Debug"
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+# PROP BASE Use_MFC 0
+# PROP BASE Use_Debug_Libraries 0
+# PROP BASE Output_Dir "Release"
+# PROP BASE Intermediate_Dir "Release"
+# PROP BASE Target_Dir ""
+# PROP Use_MFC 0
+# PROP Use_Debug_Libraries 0
+# PROP Output_Dir "Release"
+# PROP Intermediate_Dir "Release"
+# PROP Target_Dir ""
+OUTDIR=.\Release
+INTDIR=.\Release
+
+ALL : ".\Release\gc.dll" ".\Release\gc.bsc"
+
+CLEAN : 
+	-@erase ".\Release\allchblk.obj"
+	-@erase ".\Release\allchblk.sbr"
+	-@erase ".\Release\alloc.obj"
+	-@erase ".\Release\alloc.sbr"
+	-@erase ".\Release\blacklst.obj"
+	-@erase ".\Release\blacklst.sbr"
+	-@erase ".\Release\checksums.obj"
+	-@erase ".\Release\checksums.sbr"
+	-@erase ".\Release\dbg_mlc.obj"
+	-@erase ".\Release\dbg_mlc.sbr"
+	-@erase ".\Release\dyn_load.obj"
+	-@erase ".\Release\dyn_load.sbr"
+	-@erase ".\Release\finalize.obj"
+	-@erase ".\Release\finalize.sbr"
+	-@erase ".\Release\gc.bsc"
+	-@erase ".\Release\gc.dll"
+	-@erase ".\Release\gc.exp"
+	-@erase ".\Release\gc.lib"
+	-@erase ".\Release\headers.obj"
+	-@erase ".\Release\headers.sbr"
+	-@erase ".\Release\mach_dep.obj"
+	-@erase ".\Release\mach_dep.sbr"
+	-@erase ".\Release\malloc.obj"
+	-@erase ".\Release\malloc.sbr"
+	-@erase ".\Release\mallocx.obj"
+	-@erase ".\Release\mallocx.sbr"
+	-@erase ".\Release\mark.obj"
+	-@erase ".\Release\mark.sbr"
+	-@erase ".\Release\mark_rts.obj"
+	-@erase ".\Release\mark_rts.sbr"
+	-@erase ".\Release\misc.obj"
+	-@erase ".\Release\misc.sbr"
+	-@erase ".\Release\new_hblk.obj"
+	-@erase ".\Release\new_hblk.sbr"
+	-@erase ".\Release\obj_map.obj"
+	-@erase ".\Release\obj_map.sbr"
+	-@erase ".\Release\os_dep.obj"
+	-@erase ".\Release\os_dep.sbr"
+	-@erase ".\Release\ptr_chck.obj"
+	-@erase ".\Release\ptr_chck.sbr"
+	-@erase ".\Release\reclaim.obj"
+	-@erase ".\Release\reclaim.sbr"
+	-@erase ".\Release\stubborn.obj"
+	-@erase ".\Release\stubborn.sbr"
+	-@erase ".\Release\typd_mlc.obj"
+	-@erase ".\Release\typd_mlc.sbr"
+	-@erase ".\Release\win32_threads.obj"
+	-@erase ".\Release\win32_threads.sbr"
+
+"$(OUTDIR)" :
+    if not exist "$(OUTDIR)/$(NULL)" mkdir "$(OUTDIR)"
+
+CPP=cl.exe
+# ADD BASE CPP /nologo /MT /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /YX /c
+# ADD CPP /nologo /MD /W3 /GX /O2 /D "NDEBUG" /D "SILENT" /D "GC_BUILD" /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D "WIN32_THREADS" /FR /YX /c
+CPP_PROJ=/nologo /MD /W3 /GX /O2 /D "NDEBUG" /D "SILENT" /D "GC_BUILD" /D\
+ "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D\
+ "WIN32_THREADS" /FR"$(INTDIR)/" /Fp"$(INTDIR)/gc.pch" /YX /Fo"$(INTDIR)/" /c 
+CPP_OBJS=.\Release/
+CPP_SBRS=.\Release/
+
+.c{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.c{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+MTL=mktyplib.exe
+# ADD BASE MTL /nologo /D "NDEBUG" /win32
+# ADD MTL /nologo /D "NDEBUG" /win32
+MTL_PROJ=/nologo /D "NDEBUG" /win32 
+RSC=rc.exe
+# ADD BASE RSC /l 0x809 /d "NDEBUG"
+# ADD RSC /l 0x809 /d "NDEBUG"
+BSC32=bscmake.exe
+# ADD BASE BSC32 /nologo
+# ADD BSC32 /nologo
+BSC32_FLAGS=/nologo /o"$(OUTDIR)/gc.bsc" 
+BSC32_SBRS= \
+	".\Release\allchblk.sbr" \
+	".\Release\alloc.sbr" \
+	".\Release\blacklst.sbr" \
+	".\Release\checksums.sbr" \
+	".\Release\dbg_mlc.sbr" \
+	".\Release\dyn_load.sbr" \
+	".\Release\finalize.sbr" \
+	".\Release\headers.sbr" \
+	".\Release\mach_dep.sbr" \
+	".\Release\malloc.sbr" \
+	".\Release\mallocx.sbr" \
+	".\Release\mark.sbr" \
+	".\Release\mark_rts.sbr" \
+	".\Release\misc.sbr" \
+	".\Release\new_hblk.sbr" \
+	".\Release\obj_map.sbr" \
+	".\Release\os_dep.sbr" \
+	".\Release\ptr_chck.sbr" \
+	".\Release\reclaim.sbr" \
+	".\Release\stubborn.sbr" \
+	".\Release\typd_mlc.sbr" \
+	".\Release\win32_threads.sbr"
+
+".\Release\gc.bsc" : "$(OUTDIR)" $(BSC32_SBRS)
+    $(BSC32) @<<
+  $(BSC32_FLAGS) $(BSC32_SBRS)
+<<
+
+LINK32=link.exe
+# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /dll /machine:I386
+# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /dll /machine:I386
+LINK32_FLAGS=kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib\
+ advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib\
+ odbccp32.lib /nologo /subsystem:windows /dll /incremental:no\
+ /pdb:"$(OUTDIR)/gc.pdb" /machine:I386 /out:"$(OUTDIR)/gc.dll"\
+ /implib:"$(OUTDIR)/gc.lib" 
+LINK32_OBJS= \
+	".\Release\allchblk.obj" \
+	".\Release\alloc.obj" \
+	".\Release\blacklst.obj" \
+	".\Release\checksums.obj" \
+	".\Release\dbg_mlc.obj" \
+	".\Release\dyn_load.obj" \
+	".\Release\finalize.obj" \
+	".\Release\headers.obj" \
+	".\Release\mach_dep.obj" \
+	".\Release\malloc.obj" \
+	".\Release\mallocx.obj" \
+	".\Release\mark.obj" \
+	".\Release\mark_rts.obj" \
+	".\Release\misc.obj" \
+	".\Release\new_hblk.obj" \
+	".\Release\obj_map.obj" \
+	".\Release\os_dep.obj" \
+	".\Release\ptr_chck.obj" \
+	".\Release\reclaim.obj" \
+	".\Release\stubborn.obj" \
+	".\Release\typd_mlc.obj" \
+	".\Release\win32_threads.obj"
+
+".\Release\gc.dll" : "$(OUTDIR)" $(DEF_FILE) $(LINK32_OBJS)
+    $(LINK32) @<<
+  $(LINK32_FLAGS) $(LINK32_OBJS)
+<<
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+# PROP BASE Use_MFC 0
+# PROP BASE Use_Debug_Libraries 1
+# PROP BASE Output_Dir "Debug"
+# PROP BASE Intermediate_Dir "Debug"
+# PROP BASE Target_Dir ""
+# PROP Use_MFC 0
+# PROP Use_Debug_Libraries 1
+# PROP Output_Dir "Debug"
+# PROP Intermediate_Dir "Debug"
+# PROP Target_Dir ""
+OUTDIR=.\Debug
+INTDIR=.\Debug
+
+ALL : ".\Debug\gc.dll" ".\Debug\gc.bsc"
+
+CLEAN : 
+	-@erase ".\Debug\allchblk.obj"
+	-@erase ".\Debug\allchblk.sbr"
+	-@erase ".\Debug\alloc.obj"
+	-@erase ".\Debug\alloc.sbr"
+	-@erase ".\Debug\blacklst.obj"
+	-@erase ".\Debug\blacklst.sbr"
+	-@erase ".\Debug\checksums.obj"
+	-@erase ".\Debug\checksums.sbr"
+	-@erase ".\Debug\dbg_mlc.obj"
+	-@erase ".\Debug\dbg_mlc.sbr"
+	-@erase ".\Debug\dyn_load.obj"
+	-@erase ".\Debug\dyn_load.sbr"
+	-@erase ".\Debug\finalize.obj"
+	-@erase ".\Debug\finalize.sbr"
+	-@erase ".\Debug\gc.bsc"
+	-@erase ".\Debug\gc.dll"
+	-@erase ".\Debug\gc.exp"
+	-@erase ".\Debug\gc.lib"
+	-@erase ".\Debug\gc.map"
+	-@erase ".\Debug\gc.pdb"
+	-@erase ".\Debug\headers.obj"
+	-@erase ".\Debug\headers.sbr"
+	-@erase ".\Debug\mach_dep.obj"
+	-@erase ".\Debug\mach_dep.sbr"
+	-@erase ".\Debug\malloc.obj"
+	-@erase ".\Debug\malloc.sbr"
+	-@erase ".\Debug\mallocx.obj"
+	-@erase ".\Debug\mallocx.sbr"
+	-@erase ".\Debug\mark.obj"
+	-@erase ".\Debug\mark.sbr"
+	-@erase ".\Debug\mark_rts.obj"
+	-@erase ".\Debug\mark_rts.sbr"
+	-@erase ".\Debug\misc.obj"
+	-@erase ".\Debug\misc.sbr"
+	-@erase ".\Debug\new_hblk.obj"
+	-@erase ".\Debug\new_hblk.sbr"
+	-@erase ".\Debug\obj_map.obj"
+	-@erase ".\Debug\obj_map.sbr"
+	-@erase ".\Debug\os_dep.obj"
+	-@erase ".\Debug\os_dep.sbr"
+	-@erase ".\Debug\ptr_chck.obj"
+	-@erase ".\Debug\ptr_chck.sbr"
+	-@erase ".\Debug\reclaim.obj"
+	-@erase ".\Debug\reclaim.sbr"
+	-@erase ".\Debug\stubborn.obj"
+	-@erase ".\Debug\stubborn.sbr"
+	-@erase ".\Debug\typd_mlc.obj"
+	-@erase ".\Debug\typd_mlc.sbr"
+	-@erase ".\Debug\vc40.idb"
+	-@erase ".\Debug\vc40.pdb"
+	-@erase ".\Debug\win32_threads.obj"
+	-@erase ".\Debug\win32_threads.sbr"
+
+"$(OUTDIR)" :
+    if not exist "$(OUTDIR)/$(NULL)" mkdir "$(OUTDIR)"
+
+CPP=cl.exe
+# ADD BASE CPP /nologo /MTd /W3 /Gm /GX /Zi /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /YX /c
+# ADD CPP /nologo /MDd /W3 /Gm /GX /Zi /Od /D "_DEBUG" /D "SILENT" /D "GC_BUILD" /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D "WIN32_THREADS" /FR /YX /c
+CPP_PROJ=/nologo /MDd /W3 /Gm /GX /Zi /Od /D "_DEBUG" /D "SILENT" /D "GC_BUILD"\
+ /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D\
+ "WIN32_THREADS" /FR"$(INTDIR)/" /Fp"$(INTDIR)/gc.pch" /YX /Fo"$(INTDIR)/"\
+ /Fd"$(INTDIR)/" /c 
+CPP_OBJS=.\Debug/
+CPP_SBRS=.\Debug/
+
+.c{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.c{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+MTL=mktyplib.exe
+# ADD BASE MTL /nologo /D "_DEBUG" /win32
+# ADD MTL /nologo /D "_DEBUG" /win32
+MTL_PROJ=/nologo /D "_DEBUG" /win32 
+RSC=rc.exe
+# ADD BASE RSC /l 0x809 /d "_DEBUG"
+# ADD RSC /l 0x809 /d "_DEBUG"
+BSC32=bscmake.exe
+# ADD BASE BSC32 /nologo
+# ADD BSC32 /nologo
+BSC32_FLAGS=/nologo /o"$(OUTDIR)/gc.bsc" 
+BSC32_SBRS= \
+	".\Debug\allchblk.sbr" \
+	".\Debug\alloc.sbr" \
+	".\Debug\blacklst.sbr" \
+	".\Debug\checksums.sbr" \
+	".\Debug\dbg_mlc.sbr" \
+	".\Debug\dyn_load.sbr" \
+	".\Debug\finalize.sbr" \
+	".\Debug\headers.sbr" \
+	".\Debug\mach_dep.sbr" \
+	".\Debug\malloc.sbr" \
+	".\Debug\mallocx.sbr" \
+	".\Debug\mark.sbr" \
+	".\Debug\mark_rts.sbr" \
+	".\Debug\misc.sbr" \
+	".\Debug\new_hblk.sbr" \
+	".\Debug\obj_map.sbr" \
+	".\Debug\os_dep.sbr" \
+	".\Debug\ptr_chck.sbr" \
+	".\Debug\reclaim.sbr" \
+	".\Debug\stubborn.sbr" \
+	".\Debug\typd_mlc.sbr" \
+	".\Debug\win32_threads.sbr"
+
+".\Debug\gc.bsc" : "$(OUTDIR)" $(BSC32_SBRS)
+    $(BSC32) @<<
+  $(BSC32_FLAGS) $(BSC32_SBRS)
+<<
+
+LINK32=link.exe
+# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /dll /debug /machine:I386
+# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /dll /incremental:no /map /debug /machine:I386
+LINK32_FLAGS=kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib\
+ advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib\
+ odbccp32.lib /nologo /subsystem:windows /dll /incremental:no\
+ /pdb:"$(OUTDIR)/gc.pdb" /map:"$(INTDIR)/gc.map" /debug /machine:I386\
+ /out:"$(OUTDIR)/gc.dll" /implib:"$(OUTDIR)/gc.lib" 
+LINK32_OBJS= \
+	".\Debug\allchblk.obj" \
+	".\Debug\alloc.obj" \
+	".\Debug\blacklst.obj" \
+	".\Debug\checksums.obj" \
+	".\Debug\dbg_mlc.obj" \
+	".\Debug\dyn_load.obj" \
+	".\Debug\finalize.obj" \
+	".\Debug\headers.obj" \
+	".\Debug\mach_dep.obj" \
+	".\Debug\malloc.obj" \
+	".\Debug\mallocx.obj" \
+	".\Debug\mark.obj" \
+	".\Debug\mark_rts.obj" \
+	".\Debug\misc.obj" \
+	".\Debug\new_hblk.obj" \
+	".\Debug\obj_map.obj" \
+	".\Debug\os_dep.obj" \
+	".\Debug\ptr_chck.obj" \
+	".\Debug\reclaim.obj" \
+	".\Debug\stubborn.obj" \
+	".\Debug\typd_mlc.obj" \
+	".\Debug\win32_threads.obj"
+
+".\Debug\gc.dll" : "$(OUTDIR)" $(DEF_FILE) $(LINK32_OBJS)
+    $(LINK32) @<<
+  $(LINK32_FLAGS) $(LINK32_OBJS)
+<<
+
+!ELSEIF  "$(CFG)" == "gctest - Win32 Release"
+
+# PROP BASE Use_MFC 0
+# PROP BASE Use_Debug_Libraries 0
+# PROP BASE Output_Dir "gctest\Release"
+# PROP BASE Intermediate_Dir "gctest\Release"
+# PROP BASE Target_Dir "gctest"
+# PROP Use_MFC 0
+# PROP Use_Debug_Libraries 0
+# PROP Output_Dir "gctest\Release"
+# PROP Intermediate_Dir "gctest\Release"
+# PROP Target_Dir "gctest"
+OUTDIR=.\gctest\Release
+INTDIR=.\gctest\Release
+
+ALL : "gc - Win32 Release" ".\Release\gctest.exe"
+
+CLEAN : 
+	-@erase ".\gctest\Release\test.obj"
+	-@erase ".\Release\gctest.exe"
+
+"$(OUTDIR)" :
+    if not exist "$(OUTDIR)/$(NULL)" mkdir "$(OUTDIR)"
+
+CPP=cl.exe
+# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /YX /c
+# ADD CPP /nologo /MD /W3 /GX /O2 /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D "WIN32_THREADS" /YX /c
+CPP_PROJ=/nologo /MD /W3 /GX /O2 /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D\
+ "ALL_INTERIOR_POINTERS" /D "__STDC__" /D "WIN32_THREADS"\
+ /Fp"$(INTDIR)/gctest.pch" /YX /Fo"$(INTDIR)/" /c 
+CPP_OBJS=.\gctest\Release/
+CPP_SBRS=.\.
+
+.c{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.c{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+MTL=mktyplib.exe
+# ADD BASE MTL /nologo /D "NDEBUG" /win32
+# ADD MTL /nologo /D "NDEBUG" /win32
+MTL_PROJ=/nologo /D "NDEBUG" /win32 
+RSC=rc.exe
+# ADD BASE RSC /l 0x809 /d "NDEBUG"
+# ADD RSC /l 0x809 /d "NDEBUG"
+BSC32=bscmake.exe
+# ADD BASE BSC32 /nologo
+# ADD BSC32 /nologo
+BSC32_FLAGS=/nologo /o"$(OUTDIR)/gctest.bsc" 
+BSC32_SBRS= \
+	
+LINK32=link.exe
+# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386
+# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386 /out:"Release/gctest.exe"
+LINK32_FLAGS=kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib\
+ advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib\
+ odbccp32.lib /nologo /subsystem:windows /incremental:no\
+ /pdb:"$(OUTDIR)/gctest.pdb" /machine:I386 /out:"Release/gctest.exe" 
+LINK32_OBJS= \
+	".\gctest\Release\test.obj" \
+	".\Release\gc.lib"
+
+".\Release\gctest.exe" : "$(OUTDIR)" $(DEF_FILE) $(LINK32_OBJS)
+    $(LINK32) @<<
+  $(LINK32_FLAGS) $(LINK32_OBJS)
+<<
+
+!ELSEIF  "$(CFG)" == "gctest - Win32 Debug"
+
+# PROP BASE Use_MFC 0
+# PROP BASE Use_Debug_Libraries 1
+# PROP BASE Output_Dir "gctest\Debug"
+# PROP BASE Intermediate_Dir "gctest\Debug"
+# PROP BASE Target_Dir "gctest"
+# PROP Use_MFC 0
+# PROP Use_Debug_Libraries 1
+# PROP Output_Dir "gctest\Debug"
+# PROP Intermediate_Dir "gctest\Debug"
+# PROP Target_Dir "gctest"
+OUTDIR=.\gctest\Debug
+INTDIR=.\gctest\Debug
+
+ALL : "gc - Win32 Debug" ".\Debug\gctest.exe" ".\gctest\Debug\gctest.bsc"
+
+CLEAN : 
+	-@erase ".\Debug\gctest.exe"
+	-@erase ".\gctest\Debug\gctest.bsc"
+	-@erase ".\gctest\Debug\gctest.map"
+	-@erase ".\gctest\Debug\gctest.pdb"
+	-@erase ".\gctest\Debug\test.obj"
+	-@erase ".\gctest\Debug\test.sbr"
+	-@erase ".\gctest\Debug\vc40.idb"
+	-@erase ".\gctest\Debug\vc40.pdb"
+
+"$(OUTDIR)" :
+    if not exist "$(OUTDIR)/$(NULL)" mkdir "$(OUTDIR)"
+
+CPP=cl.exe
+# ADD BASE CPP /nologo /W3 /Gm /GX /Zi /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /YX /c
+# ADD CPP /nologo /MDd /W3 /Gm /GX /Zi /Od /D "_DEBUG" /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D "WIN32_THREADS" /FR /YX /c
+CPP_PROJ=/nologo /MDd /W3 /Gm /GX /Zi /Od /D "_DEBUG" /D "WIN32" /D "_WINDOWS"\
+ /D "ALL_INTERIOR_POINTERS" /D "__STDC__" /D "WIN32_THREADS" /FR"$(INTDIR)/"\
+ /Fp"$(INTDIR)/gctest.pch" /YX /Fo"$(INTDIR)/" /Fd"$(INTDIR)/" /c 
+CPP_OBJS=.\gctest\Debug/
+CPP_SBRS=.\gctest\Debug/
+
+.c{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.c{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+MTL=mktyplib.exe
+# ADD BASE MTL /nologo /D "_DEBUG" /win32
+# ADD MTL /nologo /D "_DEBUG" /win32
+MTL_PROJ=/nologo /D "_DEBUG" /win32 
+RSC=rc.exe
+# ADD BASE RSC /l 0x809 /d "_DEBUG"
+# ADD RSC /l 0x809 /d "_DEBUG"
+BSC32=bscmake.exe
+# ADD BASE BSC32 /nologo
+# ADD BSC32 /nologo
+BSC32_FLAGS=/nologo /o"$(OUTDIR)/gctest.bsc" 
+BSC32_SBRS= \
+	".\gctest\Debug\test.sbr"
+
+".\gctest\Debug\gctest.bsc" : "$(OUTDIR)" $(BSC32_SBRS)
+    $(BSC32) @<<
+  $(BSC32_FLAGS) $(BSC32_SBRS)
+<<
+
+LINK32=link.exe
+# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386
+# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /incremental:no /map /debug /machine:I386 /out:"Debug/gctest.exe"
+LINK32_FLAGS=kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib\
+ advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib\
+ odbccp32.lib /nologo /subsystem:windows /incremental:no\
+ /pdb:"$(OUTDIR)/gctest.pdb" /map:"$(INTDIR)/gctest.map" /debug /machine:I386\
+ /out:"Debug/gctest.exe" 
+LINK32_OBJS= \
+	".\Debug\gc.lib" \
+	".\gctest\Debug\test.obj"
+
+".\Debug\gctest.exe" : "$(OUTDIR)" $(DEF_FILE) $(LINK32_OBJS)
+    $(LINK32) @<<
+  $(LINK32_FLAGS) $(LINK32_OBJS)
+<<
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Release"
+
+# PROP BASE Use_MFC 0
+# PROP BASE Use_Debug_Libraries 0
+# PROP BASE Output_Dir "cord\Release"
+# PROP BASE Intermediate_Dir "cord\Release"
+# PROP BASE Target_Dir "cord"
+# PROP Use_MFC 0
+# PROP Use_Debug_Libraries 0
+# PROP Output_Dir "cord\Release"
+# PROP Intermediate_Dir "cord\Release"
+# PROP Target_Dir "cord"
+OUTDIR=.\cord\Release
+INTDIR=.\cord\Release
+
+ALL : "gc - Win32 Release" ".\Release\de.exe"
+
+CLEAN : 
+	-@erase ".\cord\Release\cordbscs.obj"
+	-@erase ".\cord\Release\cordxtra.obj"
+	-@erase ".\cord\Release\de.obj"
+	-@erase ".\cord\Release\de_win.obj"
+	-@erase ".\cord\Release\de_win.res"
+	-@erase ".\Release\de.exe"
+
+"$(OUTDIR)" :
+    if not exist "$(OUTDIR)/$(NULL)" mkdir "$(OUTDIR)"
+
+CPP=cl.exe
+# ADD BASE CPP /nologo /W3 /GX /O2 /D "WIN32" /D "NDEBUG" /D "_WINDOWS" /YX /c
+# ADD CPP /nologo /MD /W3 /GX /O2 /I "." /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /YX /c
+CPP_PROJ=/nologo /MD /W3 /GX /O2 /I "." /D "NDEBUG" /D "WIN32" /D "_WINDOWS" /D\
+ "ALL_INTERIOR_POINTERS" /Fp"$(INTDIR)/cord.pch" /YX /Fo"$(INTDIR)/" /c 
+CPP_OBJS=.\cord\Release/
+CPP_SBRS=.\.
+
+.c{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.c{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+MTL=mktyplib.exe
+# ADD BASE MTL /nologo /D "NDEBUG" /win32
+# ADD MTL /nologo /D "NDEBUG" /win32
+MTL_PROJ=/nologo /D "NDEBUG" /win32 
+RSC=rc.exe
+# ADD BASE RSC /l 0x809 /d "NDEBUG"
+# ADD RSC /l 0x809 /d "NDEBUG"
+RSC_PROJ=/l 0x809 /fo"$(INTDIR)/de_win.res" /d "NDEBUG" 
+BSC32=bscmake.exe
+# ADD BASE BSC32 /nologo
+# ADD BSC32 /nologo
+BSC32_FLAGS=/nologo /o"$(OUTDIR)/cord.bsc" 
+BSC32_SBRS= \
+	
+LINK32=link.exe
+# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386
+# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /machine:I386 /out:"Release/de.exe"
+LINK32_FLAGS=kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib\
+ advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib\
+ odbccp32.lib /nologo /subsystem:windows /incremental:no /pdb:"$(OUTDIR)/de.pdb"\
+ /machine:I386 /out:"Release/de.exe" 
+LINK32_OBJS= \
+	".\cord\Release\cordbscs.obj" \
+	".\cord\Release\cordxtra.obj" \
+	".\cord\Release\de.obj" \
+	".\cord\Release\de_win.obj" \
+	".\cord\Release\de_win.res" \
+	".\Release\gc.lib"
+
+".\Release\de.exe" : "$(OUTDIR)" $(DEF_FILE) $(LINK32_OBJS)
+    $(LINK32) @<<
+  $(LINK32_FLAGS) $(LINK32_OBJS)
+<<
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+# PROP BASE Use_MFC 0
+# PROP BASE Use_Debug_Libraries 1
+# PROP BASE Output_Dir "cord\Debug"
+# PROP BASE Intermediate_Dir "cord\Debug"
+# PROP BASE Target_Dir "cord"
+# PROP Use_MFC 0
+# PROP Use_Debug_Libraries 1
+# PROP Output_Dir "cord\Debug"
+# PROP Intermediate_Dir "cord\Debug"
+# PROP Target_Dir "cord"
+OUTDIR=.\cord\Debug
+INTDIR=.\cord\Debug
+
+ALL : "gc - Win32 Debug" ".\Debug\de.exe"
+
+CLEAN : 
+	-@erase ".\cord\Debug\cordbscs.obj"
+	-@erase ".\cord\Debug\cordxtra.obj"
+	-@erase ".\cord\Debug\de.obj"
+	-@erase ".\cord\Debug\de.pdb"
+	-@erase ".\cord\Debug\de_win.obj"
+	-@erase ".\cord\Debug\de_win.res"
+	-@erase ".\cord\Debug\vc40.idb"
+	-@erase ".\cord\Debug\vc40.pdb"
+	-@erase ".\Debug\de.exe"
+	-@erase ".\Debug\de.ilk"
+
+"$(OUTDIR)" :
+    if not exist "$(OUTDIR)/$(NULL)" mkdir "$(OUTDIR)"
+
+CPP=cl.exe
+# ADD BASE CPP /nologo /W3 /Gm /GX /Zi /Od /D "WIN32" /D "_DEBUG" /D "_WINDOWS" /YX /c
+# ADD CPP /nologo /MDd /W3 /Gm /GX /Zi /Od /I "." /D "_DEBUG" /D "WIN32" /D "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /YX /c
+CPP_PROJ=/nologo /MDd /W3 /Gm /GX /Zi /Od /I "." /D "_DEBUG" /D "WIN32" /D\
+ "_WINDOWS" /D "ALL_INTERIOR_POINTERS" /Fp"$(INTDIR)/cord.pch" /YX\
+ /Fo"$(INTDIR)/" /Fd"$(INTDIR)/" /c 
+CPP_OBJS=.\cord\Debug/
+CPP_SBRS=.\.
+
+.c{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_OBJS)}.obj:
+   $(CPP) $(CPP_PROJ) $<  
+
+.c{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cpp{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+.cxx{$(CPP_SBRS)}.sbr:
+   $(CPP) $(CPP_PROJ) $<  
+
+MTL=mktyplib.exe
+# ADD BASE MTL /nologo /D "_DEBUG" /win32
+# ADD MTL /nologo /D "_DEBUG" /win32
+MTL_PROJ=/nologo /D "_DEBUG" /win32 
+RSC=rc.exe
+# ADD BASE RSC /l 0x809 /d "_DEBUG"
+# ADD RSC /l 0x809 /d "_DEBUG"
+RSC_PROJ=/l 0x809 /fo"$(INTDIR)/de_win.res" /d "_DEBUG" 
+BSC32=bscmake.exe
+# ADD BASE BSC32 /nologo
+# ADD BSC32 /nologo
+BSC32_FLAGS=/nologo /o"$(OUTDIR)/cord.bsc" 
+BSC32_SBRS= \
+	
+LINK32=link.exe
+# ADD BASE LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386
+# ADD LINK32 kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib odbccp32.lib /nologo /subsystem:windows /debug /machine:I386 /out:"Debug/de.exe"
+LINK32_FLAGS=kernel32.lib user32.lib gdi32.lib winspool.lib comdlg32.lib\
+ advapi32.lib shell32.lib ole32.lib oleaut32.lib uuid.lib odbc32.lib\
+ odbccp32.lib /nologo /subsystem:windows /incremental:yes\
+ /pdb:"$(OUTDIR)/de.pdb" /debug /machine:I386 /out:"Debug/de.exe" 
+LINK32_OBJS= \
+	".\cord\Debug\cordbscs.obj" \
+	".\cord\Debug\cordxtra.obj" \
+	".\cord\Debug\de.obj" \
+	".\cord\Debug\de_win.obj" \
+	".\cord\Debug\de_win.res" \
+	".\Debug\gc.lib"
+
+".\Debug\de.exe" : "$(OUTDIR)" $(DEF_FILE) $(LINK32_OBJS)
+    $(LINK32) @<<
+  $(LINK32_FLAGS) $(LINK32_OBJS)
+<<
+
+!ENDIF 
+
+################################################################################
+# Begin Target
+
+# Name "gc - Win32 Release"
+# Name "gc - Win32 Debug"
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+!ENDIF 
+
+################################################################################
+# Begin Source File
+
+SOURCE=.\reclaim.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_RECLA=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_RECLA=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\reclaim.obj" : $(SOURCE) $(DEP_CPP_RECLA) "$(INTDIR)"
+
+".\Release\reclaim.sbr" : $(SOURCE) $(DEP_CPP_RECLA) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_RECLA=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_RECLA=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\reclaim.obj" : $(SOURCE) $(DEP_CPP_RECLA) "$(INTDIR)"
+
+".\Debug\reclaim.sbr" : $(SOURCE) $(DEP_CPP_RECLA) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\os_dep.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_OS_DE=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\STAT.H"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_OS_DE=\
+	".\il\PCR_IL.h"\
+	".\mm\PCR_MM.h"\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	".\vd\PCR_VD.h"\
+	
+
+".\Release\os_dep.obj" : $(SOURCE) $(DEP_CPP_OS_DE) "$(INTDIR)"
+
+".\Release\os_dep.sbr" : $(SOURCE) $(DEP_CPP_OS_DE) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_OS_DE=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\STAT.H"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_OS_DE=\
+	".\il\PCR_IL.h"\
+	".\mm\PCR_MM.h"\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	".\vd\PCR_VD.h"\
+	
+
+".\Debug\os_dep.obj" : $(SOURCE) $(DEP_CPP_OS_DE) "$(INTDIR)"
+
+".\Debug\os_dep.sbr" : $(SOURCE) $(DEP_CPP_OS_DE) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\misc.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_MISC_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MISC_=\
+	".\il\PCR_IL.h"\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\misc.obj" : $(SOURCE) $(DEP_CPP_MISC_) "$(INTDIR)"
+
+".\Release\misc.sbr" : $(SOURCE) $(DEP_CPP_MISC_) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_MISC_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MISC_=\
+	".\il\PCR_IL.h"\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\misc.obj" : $(SOURCE) $(DEP_CPP_MISC_) "$(INTDIR)"
+
+".\Debug\misc.sbr" : $(SOURCE) $(DEP_CPP_MISC_) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\mark_rts.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_MARK_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MARK_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\mark_rts.obj" : $(SOURCE) $(DEP_CPP_MARK_) "$(INTDIR)"
+
+".\Release\mark_rts.sbr" : $(SOURCE) $(DEP_CPP_MARK_) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_MARK_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MARK_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\mark_rts.obj" : $(SOURCE) $(DEP_CPP_MARK_) "$(INTDIR)"
+
+".\Debug\mark_rts.sbr" : $(SOURCE) $(DEP_CPP_MARK_) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\mach_dep.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_MACH_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MACH_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\mach_dep.obj" : $(SOURCE) $(DEP_CPP_MACH_) "$(INTDIR)"
+
+".\Release\mach_dep.sbr" : $(SOURCE) $(DEP_CPP_MACH_) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_MACH_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MACH_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\mach_dep.obj" : $(SOURCE) $(DEP_CPP_MACH_) "$(INTDIR)"
+
+".\Debug\mach_dep.sbr" : $(SOURCE) $(DEP_CPP_MACH_) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\headers.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_HEADE=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_HEADE=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\headers.obj" : $(SOURCE) $(DEP_CPP_HEADE) "$(INTDIR)"
+
+".\Release\headers.sbr" : $(SOURCE) $(DEP_CPP_HEADE) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_HEADE=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_HEADE=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\headers.obj" : $(SOURCE) $(DEP_CPP_HEADE) "$(INTDIR)"
+
+".\Debug\headers.sbr" : $(SOURCE) $(DEP_CPP_HEADE) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\alloc.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_ALLOC=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_ALLOC=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\alloc.obj" : $(SOURCE) $(DEP_CPP_ALLOC) "$(INTDIR)"
+
+".\Release\alloc.sbr" : $(SOURCE) $(DEP_CPP_ALLOC) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_ALLOC=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_ALLOC=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\alloc.obj" : $(SOURCE) $(DEP_CPP_ALLOC) "$(INTDIR)"
+
+".\Debug\alloc.sbr" : $(SOURCE) $(DEP_CPP_ALLOC) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\allchblk.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_ALLCH=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_ALLCH=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\allchblk.obj" : $(SOURCE) $(DEP_CPP_ALLCH) "$(INTDIR)"
+
+".\Release\allchblk.sbr" : $(SOURCE) $(DEP_CPP_ALLCH) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_ALLCH=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_ALLCH=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\allchblk.obj" : $(SOURCE) $(DEP_CPP_ALLCH) "$(INTDIR)"
+
+".\Debug\allchblk.sbr" : $(SOURCE) $(DEP_CPP_ALLCH) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\stubborn.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_STUBB=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_STUBB=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\stubborn.obj" : $(SOURCE) $(DEP_CPP_STUBB) "$(INTDIR)"
+
+".\Release\stubborn.sbr" : $(SOURCE) $(DEP_CPP_STUBB) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_STUBB=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_STUBB=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\stubborn.obj" : $(SOURCE) $(DEP_CPP_STUBB) "$(INTDIR)"
+
+".\Debug\stubborn.sbr" : $(SOURCE) $(DEP_CPP_STUBB) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\obj_map.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_OBJ_M=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_OBJ_M=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\obj_map.obj" : $(SOURCE) $(DEP_CPP_OBJ_M) "$(INTDIR)"
+
+".\Release\obj_map.sbr" : $(SOURCE) $(DEP_CPP_OBJ_M) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_OBJ_M=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_OBJ_M=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\obj_map.obj" : $(SOURCE) $(DEP_CPP_OBJ_M) "$(INTDIR)"
+
+".\Debug\obj_map.sbr" : $(SOURCE) $(DEP_CPP_OBJ_M) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\new_hblk.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_NEW_H=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_NEW_H=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\new_hblk.obj" : $(SOURCE) $(DEP_CPP_NEW_H) "$(INTDIR)"
+
+".\Release\new_hblk.sbr" : $(SOURCE) $(DEP_CPP_NEW_H) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_NEW_H=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_NEW_H=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\new_hblk.obj" : $(SOURCE) $(DEP_CPP_NEW_H) "$(INTDIR)"
+
+".\Debug\new_hblk.sbr" : $(SOURCE) $(DEP_CPP_NEW_H) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\mark.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_MARK_C=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MARK_C=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\mark.obj" : $(SOURCE) $(DEP_CPP_MARK_C) "$(INTDIR)"
+
+".\Release\mark.sbr" : $(SOURCE) $(DEP_CPP_MARK_C) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_MARK_C=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MARK_C=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\mark.obj" : $(SOURCE) $(DEP_CPP_MARK_C) "$(INTDIR)"
+
+".\Debug\mark.sbr" : $(SOURCE) $(DEP_CPP_MARK_C) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\malloc.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_MALLO=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MALLO=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\malloc.obj" : $(SOURCE) $(DEP_CPP_MALLO) "$(INTDIR)"
+
+".\Release\malloc.sbr" : $(SOURCE) $(DEP_CPP_MALLO) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_MALLO=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MALLO=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\malloc.obj" : $(SOURCE) $(DEP_CPP_MALLO) "$(INTDIR)"
+
+".\Debug\malloc.sbr" : $(SOURCE) $(DEP_CPP_MALLO) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\mallocx.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_MALLX=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MALLX=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\mallocx.obj" : $(SOURCE) $(DEP_CPP_MALLX) "$(INTDIR)"
+
+".\Release\mallocx.sbr" : $(SOURCE) $(DEP_CPP_MALLX) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_MALLX=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_MALLX=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\mallocx.obj" : $(SOURCE) $(DEP_CPP_MALLX) "$(INTDIR)"
+
+".\Debug\mallocx.sbr" : $(SOURCE) $(DEP_CPP_MALLX) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\finalize.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_FINAL=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_FINAL=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\finalize.obj" : $(SOURCE) $(DEP_CPP_FINAL) "$(INTDIR)"
+
+".\Release\finalize.sbr" : $(SOURCE) $(DEP_CPP_FINAL) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_FINAL=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_FINAL=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\finalize.obj" : $(SOURCE) $(DEP_CPP_FINAL) "$(INTDIR)"
+
+".\Debug\finalize.sbr" : $(SOURCE) $(DEP_CPP_FINAL) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\dbg_mlc.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_DBG_M=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_DBG_M=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\dbg_mlc.obj" : $(SOURCE) $(DEP_CPP_DBG_M) "$(INTDIR)"
+
+".\Release\dbg_mlc.sbr" : $(SOURCE) $(DEP_CPP_DBG_M) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_DBG_M=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_DBG_M=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\dbg_mlc.obj" : $(SOURCE) $(DEP_CPP_DBG_M) "$(INTDIR)"
+
+".\Debug\dbg_mlc.sbr" : $(SOURCE) $(DEP_CPP_DBG_M) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\blacklst.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_BLACK=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_BLACK=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\blacklst.obj" : $(SOURCE) $(DEP_CPP_BLACK) "$(INTDIR)"
+
+".\Release\blacklst.sbr" : $(SOURCE) $(DEP_CPP_BLACK) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_BLACK=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_BLACK=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\blacklst.obj" : $(SOURCE) $(DEP_CPP_BLACK) "$(INTDIR)"
+
+".\Debug\blacklst.sbr" : $(SOURCE) $(DEP_CPP_BLACK) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\typd_mlc.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_TYPD_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	".\gc_typed.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_TYPD_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\typd_mlc.obj" : $(SOURCE) $(DEP_CPP_TYPD_) "$(INTDIR)"
+
+".\Release\typd_mlc.sbr" : $(SOURCE) $(DEP_CPP_TYPD_) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_TYPD_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	".\gc_typed.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_TYPD_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\typd_mlc.obj" : $(SOURCE) $(DEP_CPP_TYPD_) "$(INTDIR)"
+
+".\Debug\typd_mlc.sbr" : $(SOURCE) $(DEP_CPP_TYPD_) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\ptr_chck.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_PTR_C=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_PTR_C=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\ptr_chck.obj" : $(SOURCE) $(DEP_CPP_PTR_C) "$(INTDIR)"
+
+".\Release\ptr_chck.sbr" : $(SOURCE) $(DEP_CPP_PTR_C) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_PTR_C=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_mark.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_PTR_C=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\ptr_chck.obj" : $(SOURCE) $(DEP_CPP_PTR_C) "$(INTDIR)"
+
+".\Debug\ptr_chck.sbr" : $(SOURCE) $(DEP_CPP_PTR_C) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\dyn_load.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_DYN_L=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\STAT.H"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_DYN_L=\
+	".\il\PCR_IL.h"\
+	".\mm\PCR_MM.h"\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\dyn_load.obj" : $(SOURCE) $(DEP_CPP_DYN_L) "$(INTDIR)"
+
+".\Release\dyn_load.sbr" : $(SOURCE) $(DEP_CPP_DYN_L) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_DYN_L=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\STAT.H"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_DYN_L=\
+	".\il\PCR_IL.h"\
+	".\mm\PCR_MM.h"\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\dyn_load.obj" : $(SOURCE) $(DEP_CPP_DYN_L) "$(INTDIR)"
+
+".\Debug\dyn_load.sbr" : $(SOURCE) $(DEP_CPP_DYN_L) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\win32_threads.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_WIN32=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_WIN32=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\win32_threads.obj" : $(SOURCE) $(DEP_CPP_WIN32) "$(INTDIR)"
+
+".\Release\win32_threads.sbr" : $(SOURCE) $(DEP_CPP_WIN32) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_WIN32=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_WIN32=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\win32_threads.obj" : $(SOURCE) $(DEP_CPP_WIN32) "$(INTDIR)"
+
+".\Debug\win32_threads.sbr" : $(SOURCE) $(DEP_CPP_WIN32) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\checksums.c
+
+!IF  "$(CFG)" == "gc - Win32 Release"
+
+DEP_CPP_CHECK=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_CHECK=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Release\checksums.obj" : $(SOURCE) $(DEP_CPP_CHECK) "$(INTDIR)"
+
+".\Release\checksums.sbr" : $(SOURCE) $(DEP_CPP_CHECK) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gc - Win32 Debug"
+
+DEP_CPP_CHECK=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_CHECK=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+".\Debug\checksums.obj" : $(SOURCE) $(DEP_CPP_CHECK) "$(INTDIR)"
+
+".\Debug\checksums.sbr" : $(SOURCE) $(DEP_CPP_CHECK) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+# End Target
+################################################################################
+# Begin Target
+
+# Name "gctest - Win32 Release"
+# Name "gctest - Win32 Debug"
+
+!IF  "$(CFG)" == "gctest - Win32 Release"
+
+!ELSEIF  "$(CFG)" == "gctest - Win32 Debug"
+
+!ENDIF 
+
+################################################################################
+# Begin Project Dependency
+
+# Project_Dep_Name "gc"
+
+!IF  "$(CFG)" == "gctest - Win32 Release"
+
+"gc - Win32 Release" : 
+   $(MAKE) /$(MAKEFLAGS) /F ".\gc.mak" CFG="gc - Win32 Release" 
+
+!ELSEIF  "$(CFG)" == "gctest - Win32 Debug"
+
+"gc - Win32 Debug" : 
+   $(MAKE) /$(MAKEFLAGS) /F ".\gc.mak" CFG="gc - Win32 Debug" 
+
+!ENDIF 
+
+# End Project Dependency
+################################################################################
+# Begin Source File
+
+SOURCE=.\test.c
+DEP_CPP_TEST_=\
+	".\gcconfig.h"\
+	".\gc.h"\
+	".\gc_hdrs.h"\
+	".\gc_priv.h"\
+	".\gc_typed.h"\
+	{$(INCLUDE)}"\sys\TYPES.H"\
+	
+NODEP_CPP_TEST_=\
+	".\th\PCR_Th.h"\
+	".\th\PCR_ThCrSec.h"\
+	".\th\PCR_ThCtl.h"\
+	
+
+!IF  "$(CFG)" == "gctest - Win32 Release"
+
+
+".\gctest\Release\test.obj" : $(SOURCE) $(DEP_CPP_TEST_) "$(INTDIR)"
+
+
+!ELSEIF  "$(CFG)" == "gctest - Win32 Debug"
+
+
+".\gctest\Debug\test.obj" : $(SOURCE) $(DEP_CPP_TEST_) "$(INTDIR)"
+
+".\gctest\Debug\test.sbr" : $(SOURCE) $(DEP_CPP_TEST_) "$(INTDIR)"
+
+
+!ENDIF 
+
+# End Source File
+# End Target
+################################################################################
+# Begin Target
+
+# Name "cord - Win32 Release"
+# Name "cord - Win32 Debug"
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+!ENDIF 
+
+################################################################################
+# Begin Project Dependency
+
+# Project_Dep_Name "gc"
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+"gc - Win32 Release" : 
+   $(MAKE) /$(MAKEFLAGS) /F ".\gc.mak" CFG="gc - Win32 Release" 
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+"gc - Win32 Debug" : 
+   $(MAKE) /$(MAKEFLAGS) /F ".\gc.mak" CFG="gc - Win32 Debug" 
+
+!ENDIF 
+
+# End Project Dependency
+################################################################################
+# Begin Source File
+
+SOURCE=.\cord\de_win.c
+DEP_CPP_DE_WI=\
+	".\cord\cord.h"\
+	".\cord\de_cmds.h"\
+	".\cord\de_win.h"\
+	".\cord\private\cord_pos.h"\
+	
+NODEP_CPP_DE_WI=\
+	".\cord\gc.h"\
+	
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+
+".\cord\Release\de_win.obj" : $(SOURCE) $(DEP_CPP_DE_WI) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+
+".\cord\Debug\de_win.obj" : $(SOURCE) $(DEP_CPP_DE_WI) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\cord\de.c
+DEP_CPP_DE_C2e=\
+	".\cord\cord.h"\
+	".\cord\de_cmds.h"\
+	".\cord\de_win.h"\
+	".\cord\private\cord_pos.h"\
+	
+NODEP_CPP_DE_C2e=\
+	".\cord\gc.h"\
+	
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+
+".\cord\Release\de.obj" : $(SOURCE) $(DEP_CPP_DE_C2e) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+
+".\cord\Debug\de.obj" : $(SOURCE) $(DEP_CPP_DE_C2e) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\cord\cordxtra.c
+DEP_CPP_CORDX=\
+	".\cord\cord.h"\
+	".\cord\ec.h"\
+	".\cord\private\cord_pos.h"\
+	
+NODEP_CPP_CORDX=\
+	".\cord\gc.h"\
+	
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+
+".\cord\Release\cordxtra.obj" : $(SOURCE) $(DEP_CPP_CORDX) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+
+".\cord\Debug\cordxtra.obj" : $(SOURCE) $(DEP_CPP_CORDX) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\cord\cordbscs.c
+DEP_CPP_CORDB=\
+	".\cord\cord.h"\
+	".\cord\private\cord_pos.h"\
+	
+NODEP_CPP_CORDB=\
+	".\cord\gc.h"\
+	
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+
+".\cord\Release\cordbscs.obj" : $(SOURCE) $(DEP_CPP_CORDB) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+
+".\cord\Debug\cordbscs.obj" : $(SOURCE) $(DEP_CPP_CORDB) "$(INTDIR)"
+   $(CPP) $(CPP_PROJ) $(SOURCE)
+
+
+!ENDIF 
+
+# End Source File
+################################################################################
+# Begin Source File
+
+SOURCE=.\cord\de_win.RC
+
+!IF  "$(CFG)" == "cord - Win32 Release"
+
+
+".\cord\Release\de_win.res" : $(SOURCE) "$(INTDIR)"
+   $(RSC) /l 0x809 /fo"$(INTDIR)/de_win.res" /i "cord" /d "NDEBUG" $(SOURCE)
+
+
+!ELSEIF  "$(CFG)" == "cord - Win32 Debug"
+
+
+".\cord\Debug\de_win.res" : $(SOURCE) "$(INTDIR)"
+   $(RSC) /l 0x809 /fo"$(INTDIR)/de_win.res" /i "cord" /d "_DEBUG" $(SOURCE)
+
+
+!ENDIF 
+
+# End Source File
+# End Target
+# End Project
+################################################################################
diff --git a/README b/README
index 80cb26ab..d2743163 100644
--- a/README
+++ b/README
@@ -1,6 +1,7 @@
 Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
 Copyright (c) 1991-1996 by Xerox Corporation.  All rights reserved.
 Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
 
 THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
 OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -11,7 +12,7 @@ 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 5.0alpha3 of a conservative garbage collector for C and C++.
+This is version 5.3 of a conservative garbage collector for C and C++.
 
 You might find a more recent version of this at
 
@@ -22,8 +23,8 @@ 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 (boehm@acm.org) at Xerox PARC
-and at SGI.
+Much of the code was rewritten by Hans-J. Boehm (boehm@acm.org) at Xerox PARC,
+SGI, and HP Labs.
 
 Some other contributors:  
 
@@ -475,9 +476,8 @@ intended to run with malloc/free (e.g. code with extreme real-time or
 portability constraints).  To do so define FIND_LEAK in Makefile
 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.  The collector will
-no longer reclaim inaccessible memory; in this form it is purely a
-debugging tool.
+object is found that has not been explicitly freed.  Such objects will
+also be automatically reclaimed.
   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
@@ -1497,7 +1497,153 @@ Since 5.0alpha2
  - Significantly changed the way leak detection is handled, as a consequence
    of the above.
 
+Since 5.0 alpha3
+ - Added protection fault handling patch for Linux/M68K from Fergus
+   Henderson and Roman Hodek.
+ - Removed the tests for SGI_SOURCE in new_gc_alloc.h.  This was causing that
+   interface to fail on nonSGI platforms.
+ - Changed the Linux stack finding code to use /proc, after changing it
+   to use HEURISTIC1.  (Thanks to David Mossberger for pointing out the
+   /proc hook.)
+ - Added HP/UX incremental GC support and HP/UX 11 thread support.
+   Thread support is currently still flakey.
+ - Added basic Linux/IA64 support.
+ - Integrated Anthony Green's PicoJava support.
+ - Integrated Scott Ananian's StrongARM/NetBSD support.
+ - Fixed some fairly serious performance bugs in the incremental
+   collector.  These have probably been there essentially forever.
+   (Mark bits were sometimes set before scanning dirty pages.
+   The reclaim phase unnecessarily dirtied full small object pages.)
+ - Changed the reclaim phase to ignore nearly full pages to avoid
+   touching them.
+ - Limited GC_black_list_spacing to roughly the heap growth increment.
+ - Changed full collection triggering heuristic to decrease full GC
+   frequency by default, but to explicitly trigger full GCs during
+   heap growth.  This doesn't always improve things, but on average it's
+   probably a win.
+ - GC_debug_free(0, ...) failed.  Thanks to Fergus Henderson for the
+   bug report and fix.
+
+Since 5.0 alpha4
+ - GC_malloc_explicitly_typed and friends sometimes failed to
+   initialize first word.
+ - Added allocation routines and support in the marker for mark descriptors
+   in a type structure referenced by the first word of an object.  This was
+   introduced to support gcj, but hopefully in a way that makes it
+   generically useful.
+ - Added GC_requested_heapsize, and inhibited collections in nonincremental
+   mode if the actual used heap size is less than what was explicitly
+   requested.
+ - The Solaris pthreads version of GC_pthread_create didn't handle a NULL
+   attribute pointer.  Solaris thread support used the wrong default thread
+   stack size.  (Thanks to Melissa O'Neill for the patch.)
+ - Changed PUSH_CONTENTS macro to no longer modify first parameter.
+   This usually doesn't matter, but it was certainly an accident waiting
+   to happen ...
+ - Added GC_register_finalizer_no_order and friends to gc.h.  They're
+   needed by Java implementations.
+ - Integrated a fix for a win32 deadlock resulting from clock() calling
+   malloc.  (Thanks to Chris Dodd.)
+ - Integrated Hiroshi Kawashima's port to Linux/MIPS.  This was designed
+   for a handheld platform, and may or may not be sufficient for other
+   machines.
+ - Fixed a va_arg problem with the %c specifier in cordprnt.c.  It appears
+   that this was always broken, but recent versions of gcc are the first to
+   report the (statically detectable) bug.
+ - Added an attempt at a more general solution to dlopen races/deadlocks.
+   GC_dlopen now temporarily disables collection.  Still not ideal, but ...
+ - Added -DUSE_I686_PREFETCH, -DUSE_3DNOW_PREFETCH, and support for IA64
+   prefetch instructions.  May improve performance measurably, but I'm not
+   sure the code will run correctly on processors that don't support the
+   instruction.  Won't build except with very recent gcc.
+ - Added caching for header lookups in the marker.  This seems to result
+   in a barely measurable performance gain.  Added support for interleaved
+   lookups of two pointers, but unconfigured that since the performance
+   gain is currently near zero, and it adds to code size.
+ - Changed Linux DATA_START definition to check both data_start and
+   __data_start, since nothing else seems to be portable.
+ - Added -DUSE_LD_WRAP to optionally take advantage of the GNU ld function
+   wrapping mechanism.  Probably currently useful only on Linux.
+ - Moved some variables for the scratch allocator into GC_arrays, on
+   Martin Hirzel's suggestion.
+ - Fixed a win32 threads bug that caused the collector to not look for
+   interior pointers from one of the thread stacks without
+   ALL_INTERIOR_POINTERS.  (Thanks to Jeff Sturm.)
+ - Added Mingw32 support.  (Thanks again to Jeff Sturm for the patch.)
+ - Changed the alpha port to use the generic register scanning code instead
+   of alpha_mach_dep.s.  Alpha_mach_dep.s doesn't look for pointers in fp
+   registers, but gcc sometimes spills pointers there.  (Thanks to Manuel
+   Serrano for helping me debug this by email.)  Changed the IA64 code to
+   do something similar for similar reasons.
+
+Since 5.0alpha6:
+ - -DREDIRECT_MALLOC was broken in alpha6. Fixed.
+ - Cleaned up gc_ccp.h slightly, thus also causing the HP C++ compiler to
+   accept it.
+ - Removed accidental reference to dbg_mlc.c, which caused dbg_mlc.o to be
+   linked into every executable.
+ - Added PREFETCH to bitmap marker.  Changed it to use the header cache.
+ - GC_push_marked sometimes pushed one object too many, resulting in a
+   segmentation fault in GC_mark_from_mark_stack.  This was probably an old
+   bug.  It finally showed up in gctest on win32.
+ - Gc_priv.h erroneously #defined GC_incremental to be TRUE instead of FALSE
+   when SMALL_CONFIG was defined.  This was no doubt a major performance bug for
+   the default win32 configuration.
+ - Removed -DSMALL_CONFIG from NT_MAKEFILE.  It seemed like an anchronism now
+   that the average PC has 64MB or so.
+ - Integrated Bryce McKinley's patches for linux threads and dynamic loading
+   from the libgcj tree.  Turned on dynamic loading support for Linux/PPC.
+ - Changed the stack finding code to use environ on HP/UX.  (Thanks
+   to Gustavo Rodriguez-Rivera for the suggestion.)  This should probably
+   be done on other platforms, too.  Since I can't test those, that'll
+   wait until after 5.0.
+
+Since 5.0alpha7:
+ - Fixed threadlibs.c for linux threads.  -DUSE_LD_WRAP was broken and
+   -ldl was omitted.  Fixed Linux stack finding code to handle
+   -DUSE_LD_WRAP correctly.
+ - Added MSWIN32 exception handler around marker, so that the collector
+   can recover from root segments that are unmapped during the collection.
+   This caused occasional failures under Windows 98, and may also be
+   an issue under Windows NT/2000.
+
+Since 5.0
+ - Fixed a gc.h header bug which showed up under Irix.  (Thanks to
+   Dan Sullivan.)
+ - Fixed a typo in GC_double_descr in typd_mlc.c not getting traced correctly.
+   This probably could result in objects described by array descriptors not
+   getting traced correctly.  (Thanks to Ben Hutchings for pointing this out.)
+ - The block nearly full tests in reclaim.c were not correct for 64 bit
+   environments.  This could result in unnecessary heap growth under unlikely
+   conditions.
+ - Removed use of CLEAR_DOUBLE from generic reclaim code, since odd sizes
+   could occur.
+
+Since 5.1
+ - dyn_load.c declared GC_scratch_last_end_ptr as an extern even if it
+   was defined as a macro.  This prevented the collector from building on
+   Irix.
+ - We quietly assumed that indirect mark descriptors were never 0.
+   Our own typed allocation interface violated that.  This could result
+   in segmentation faults in the marker with typed allocation.
+ - Fixed a _DUSE_MUNMAP bug in the heap block allocation code.
+   (Thanks to Ben Hutchings for the patch.)
+ - Taught the collector about VC++ handling array operator new.
+   (Thanks again to Ben Hutchings for the patch.)
+ - The two copies of gc_hdrs.h had diverged.  Made one a link to the other
+   again.
+
+Since 5.2
+ - Fixed _end declaration for OSF1.
+ - There were lots of spurious leak reports in leak detection mode, caused
+   by the fact that some pages were not being swept, and hence unmarked
+   objects weren't making it onto free lists.  (This bug dated back to 5.0.)
+ - Fixed a typo in the liblinuxgc.so Makefile rule.
+ - Added the GetExitCodeThread to Win32 GC_stop_world to (mostly) work
+   around a Windows 95 GetOpenFileName problem.  (Thanks to Jacob Navia.)
+
 To do:
+ - Integrate Linux/SPARC fixes.
  - Very large root set sizes (> 16 MB or so) could cause the collector
    to abort with an unexpected mark stack overflow.  (Thanks again to
    Peter Chubb.)  NOT YET FIXED.  Workaround is to increase the initial
@@ -1510,8 +1656,10 @@ To do:
    be possible to conditionally intercept mmap and use GC_exclude_static_roots.
    The real fix is to walk rld data structures, which looks possible.
  - Integrate MIT and DEC pthreads ports.
- - Deal with very uneven black-listing distributions.  If all the black listed
-   blocks reside in the newly allocated heap section, the heuristic for
-   temporarily ignoring black-listing fails, and the heap grows too much.
-   (This was observed in only one case, and could be worked around, but ...)
- - Some platform specific updates are waiting for 4.15alpha1. 
+ - Incremental collector should handle large objects better.  Currently,
+   it looks like the whole object is treated as dirty if any part of it
+   is.
+ - Cord/cordprnt.c doesn't build on a few platforms (notably PowerPC), since
+   we make some unwarranted assumptions about how varargs are handled.  This
+   currently makes the cord-aware versions of printf unusable on some platforms.
+   Fixing this is unfortunately not trivial.
diff --git a/README.QUICK b/README.QUICK
index 3273c8ba..ddebf82c 100644
--- a/README.QUICK
+++ b/README.QUICK
@@ -1,5 +1,7 @@
 Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
-Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
+Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+Copyright (c) 1999 by Hewlett-Packard. All rights reserved.
 
 THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
 OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
diff --git a/README.debugging b/README.debugging
index 80635c22..f4dd6567 100644
--- a/README.debugging
+++ b/README.debugging
@@ -40,7 +40,8 @@ void * big_realloc(void *p, size_t new_size)
 1) Consider using GC_malloc_atomic for objects containing nonpointers.  This is especially important for large arrays containg compressed data, pseudo-random numbers, and the like.  (This isn't all that likely to solve your problem, but it's a useful and easy optimization anyway, and this is a good time to try it.)   If you allocate large objects containg only one or two pointers at the beginning, either try the typed allocation primitives is gc.h, or separate out the pointerfree component.
 2) If you are using the collector in its default mode, with interior pointer recognition enabled, consider using GC_malloc_ignore_off_page to allocate large objects.  (See gc.h and above for details.  Large means > 100K in most environments.)
 3) GC_print_block_list() will print a list of all currently allocated heap blocks and what size objects they contain.  GC_print_hblkfreelist() will print a list of free heap blocks, and whether they are blacklisted.  GC_dump calls both of these, and also prints information about heap sections, and root segments.
-4) Write a tool that traces back references to the appropriate root.  Send me the code.  (I have code that does this for old PCR.)
+4) Build the collector with -DKEEP_BACK_PTRS, and use the backptr.h
+interface to determine why objects are being retained.
 
 
 ****If the collector appears to be losing objects:
@@ -54,5 +55,14 @@ void * big_realloc(void *p, size_t new_size)
 6) "print *GC_find_header(p)" in dbx or gdb will print the garbage collector block header information associated with the object p (e.g. object size, etc.)
 7) GC_is_marked(p) determines whether p is the base address of a marked object.  Note that objects allocated since the last collection should not be marked, and that unmarked objects are reclaimed incrementally.  It's usually most interesting to set a breakpoint in GC_finish_collection and then to determine how much of the damaged data structure is marked at that point.
 8) Look at the tracing facility in mark.c.  (Ignore this suggestion unless you are very familiar with collector internals.)
+9) [From Melissa O'Neill:]
+If you're using multiple threads, double check that all thread
+creation goes through the GC_ wrapper functions rather than
+calling the thread-creation functions themselves (e.g.,
+GC_pthread_create rather than pthread_create).  The gc.h header
+file includes suitable preprocessor definitions to accomplish
+this mapping transparently -- the question is: are you including
+it in all the modules that create threads?
+
 
 
diff --git a/README.hp b/README.hp
index b290590d..072ba538 100644
--- a/README.hp
+++ b/README.hp
@@ -8,4 +8,9 @@ The collector should compile with either plain cc or cc -Ae.  CC -Aa
 fails to define _HPUX_SOURCE and thus will not configure the collector
 correctly.
 
-There is currently no thread support.
+Incremental collection support was reccently added, and should now work.
+
+Thread support for HP/UX 11 Pthreads was also recently added.  It is still
+flakey in this release.  (It has only been tested on a uniprocessor.  Even
+there some fraction of thread creation calls fail with a not-yet-understood
+error return from sem_wait.)
diff --git a/README.linux b/README.linux
index b4f136af..e35e712e 100644
--- a/README.linux
+++ b/README.linux
@@ -31,16 +31,28 @@ To use threads, you need to abide by the following requirements:
 2) You must compile the collector with -DLINUX_THREADS and -D_REENTRANT
    specified in the Makefile.
 
-3) Every file that makes thread calls should define LINUX_THREADS and 
+3a) Every file that makes thread calls should define LINUX_THREADS and 
    _REENTRANT and then include gc.h.  Gc.h redefines some of the
    pthread primitives as macros which also provide the collector with
    information it requires.
 
-4) Currently dlopen() is probably not safe.  The collector must traverse
-   the list of libraries maintained by the runtime loader.  That can
-   probably be an inconsistent state when a thread calling the loader is
-   is stopped for GC.  (It's possible that this is fixable in the
-   same way it is handled for SOLARIS_THREADS, with GC_dlopen.)
+3b) A new alternative to (3a) is to build the collector with
+   -DUSE_LD_WRAP, and to link the final program with
+
+   (for ld) --wrap read --wrap dlopen --wrap pthread_create \
+	    --wrap pthread_join --wrap pthread_sigmask
+
+   (for gcc) -Wl,--wrap -Wl,read -Wl,--wrap -Wl,dlopen -Wl,--wrap \
+	     -Wl,pthread_create -Wl,--wrap -Wl,pthread_join -Wl,--wrap \
+	     -Wl,pthread_sigmask
+
+   In any case, _REENTRANT should be defined during compilation.
+
+4) Dlopen() disables collection during its execution.  (It can't run
+   concurrently with the collector, since the collector looks at its
+   data structures.  It can't acquire the allocator lock, since arbitrary
+   user startup code may run as part of dlopen().)  Under unusual
+   conditions, this may cause unexpected heap growth.
 
 5) The combination of LINUX_THREADS, REDIRECT_MALLOC, and incremental
    collection fails in seemingly random places.  This hasn't been tracked
@@ -48,3 +60,9 @@ To use threads, you need to abide by the following requirements:
    uses malloc, and thus can presumably get SIGSEGVs while inside the
    package.  There is no real guarantee that signals are handled properly
    at that point.
+
+6) Thread local storage may not be viewed as part of the root set by the
+   collector.  This probably depends on the linuxthreads version.  For the
+   time being, any collectable memory referenced by thread local storage should
+   also be referenced from elsewhere, or be allocated as uncollectable.
+   (This is really a bug that should be fixed somehow.)
diff --git a/README.solaris2 b/README.solaris2
index e5935131..cb15e30a 100644
--- a/README.solaris2
+++ b/README.solaris2
@@ -37,13 +37,10 @@ a thread stack.  If you know that you will only be running Solaris2.5
 or later, it should be possible to fix this by compiling the collector
 with -DSOLARIS23_MPROTECT_BUG_FIXED.
 
-Jeremy Fitzhardinge points out that there is a problem with the dlopen
-replacement, in that startup code in the library is run while the allocation
-lock is held.  This appears to be difficult to fix, since the collector does
-look at data structures maintained by dlopen, and hence some locking is needed
-around the dlopen call.  Defining USE_PROC_FOR_LIBRARIES will get address
-space layout information from /proc avoiding the dlopen lock.  But this has
-other disadvanatages, e.g. mmapped files may be scanned.
+Since 5.0 alpha5, dlopen disables collection temporarily,
+unless USE_PROC_FOR_LIBRARIES is defined.  In some unlikely cases, this
+can result in unpleasant heap growth.  But it seems better than the
+race/deadlock issues we had before.
 
 If solaris_threads are used on an X86 processor with malloc redirected to
 GC_malloc, it is necessary to call GC_thr_init explicitly before forking the
diff --git a/allchblk.c b/allchblk.c
index d8d0afdf..41624c75 100644
--- a/allchblk.c
+++ b/allchblk.c
@@ -2,6 +2,7 @@
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
  * Copyright (c) 1998-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -18,6 +19,7 @@
 #include <stdio.h>
 #include "gc_priv.h"
 
+GC_bool GC_use_entire_heap = 0;
 
 /*
  * Free heap blocks are kept on one of several free lists,
@@ -122,7 +124,7 @@ hdr * wanted;
 
 void GC_dump_regions()
 {
-    int i;
+    unsigned i;
     ptr_t start, end;
     ptr_t p;
     size_t bytes;
@@ -228,11 +230,15 @@ int n;
 	GC_ASSERT(HDR(GC_hblkfreelist[index]) == hhdr);
 	GC_hblkfreelist[index] = hhdr -> hb_next;
     } else {
-	PHDR(hhdr) -> hb_next = hhdr -> hb_next;
+	hdr *phdr;
+	GET_HDR(hhdr -> hb_prev, phdr);
+	phdr -> hb_next = hhdr -> hb_next;
     }
     if (0 != hhdr -> hb_next) {
+	hdr * nhdr;
 	GC_ASSERT(!IS_FORWARDING_ADDR_OR_NIL(NHDR(hhdr)));
-	NHDR(hhdr) -> hb_prev = hhdr -> hb_prev;
+	GET_HDR(hhdr -> hb_next, nhdr);
+	nhdr -> hb_prev = hhdr -> hb_prev;
     }
 }
 
@@ -243,13 +249,20 @@ struct hblk * GC_free_block_ending_at(h)
 struct hblk *h;
 {
     struct hblk * p = h - 1;
-    hdr * phdr = HDR(p);
+    hdr * phdr;
 
+    GET_HDR(p, phdr);
     while (0 != phdr && IS_FORWARDING_ADDR_OR_NIL(phdr)) {
 	p = FORWARDED_ADDR(p,phdr);
 	phdr = HDR(p);
     }
-    if (0 != phdr && HBLK_IS_FREE(phdr)) return p;
+    if (0 != phdr) {
+        if(HBLK_IS_FREE(phdr)) {
+	    return p;
+	} else {
+	    return 0;
+	}
+    }
     p = GC_prev_block(h - 1);
     if (0 != p) {
       phdr = HDR(p);
@@ -270,6 +283,7 @@ hdr * hhdr;
 {
     int index = GC_hblk_fl_from_blocks(divHBLKSZ(hhdr -> hb_sz));
     struct hblk *second = GC_hblkfreelist[index];
+    hdr * second_hdr;
 #   ifdef GC_ASSERTIONS
       struct hblk *next = (struct hblk *)((word)h + hhdr -> hb_sz);
       hdr * nexthdr = HDR(next);
@@ -282,7 +296,10 @@ hdr * hhdr;
     GC_hblkfreelist[index] = h;
     hhdr -> hb_next = second;
     hhdr -> hb_prev = 0;
-    if (0 != second) HDR(second) -> hb_prev = h;
+    if (0 != second) {
+      GET_HDR(second, second_hdr);
+      second_hdr -> hb_prev = h;
+    }
     GC_invalidate_map(hhdr);
 }
 
@@ -329,10 +346,10 @@ void GC_merge_unmapped(void)
     for (i = 0; i <= N_HBLK_FLS; ++i) {
       h = GC_hblkfreelist[i];
       while (h != 0) {
-	hhdr = HDR(h);
+	GET_HDR(h, hhdr);
 	size = hhdr->hb_sz;
 	next = (struct hblk *)((word)h + size);
-	nexthdr = HDR(next);
+	GET_HDR(next, nexthdr);
 	/* Coalesce with successor, if possible */
 	  if (0 != nexthdr && HBLK_IS_FREE(nexthdr)) {
 	    nextsize = nexthdr -> hb_sz;
@@ -397,8 +414,8 @@ int index;
     GC_remove_from_fl(hhdr, index);
     if (total_size == bytes) return h;
     rest = (struct hblk *)((word)h + bytes);
-    if (!GC_install_header(rest)) return(0);
-    rest_hdr = HDR(rest);
+    rest_hdr = GC_install_header(rest);
+    if (0 == rest_hdr) return(0);
     rest_hdr -> hb_sz = total_size - bytes;
     rest_hdr -> hb_flags = 0;
 #   ifdef GC_ASSERTIONS
@@ -505,16 +522,17 @@ int n;
 
     /* search for a big enough block in free list */
 	hbp = GC_hblkfreelist[n];
-	hhdr = HDR(hbp);
-	for(; 0 != hbp; hbp = hhdr -> hb_next, hhdr = HDR(hbp)) {
+	for(; 0 != hbp; hbp = hhdr -> hb_next) {
+	    GET_HDR(hbp, hhdr);
 	    size_avail = hhdr->hb_sz;
 	    if (size_avail < size_needed) continue;
-#	    ifdef PRESERVE_LAST
+	    if (!GC_use_entire_heap) {
 		if (size_avail != size_needed
+		    && USED_HEAP_SIZE >= GC_requested_heapsize
 		    && !GC_incremental && GC_should_collect()) {
 		    continue;
 		} 
-#	    endif
+	    }
 	    /* If the next heap block is obviously better, go on.	*/
 	    /* This prevents us from disassembling a single large block */
 	    /* to get tiny blocks.					*/
@@ -523,7 +541,7 @@ int n;
 	      
 	      thishbp = hhdr -> hb_next;
 	      if (thishbp != 0) {
-	        thishdr = HDR(thishbp);
+		GET_HDR(thishbp, thishdr);
 	        next_size = (signed_word)(thishdr -> hb_sz);
 	        if (next_size < size_avail
 	          && next_size >= size_needed
@@ -550,16 +568,16 @@ int n;
 	      size_avail -= (ptr_t)lasthbp - (ptr_t)hbp;
 	      thishbp = lasthbp;
 	      if (size_avail >= size_needed) {
-	        if (thishbp != hbp && GC_install_header(thishbp)) {
+	        if (thishbp != hbp &&
+		    0 != (thishdr = GC_install_header(thishbp))) {
 		  /* Make sure it's mapped before we mangle it. */
 #		    ifdef USE_MUNMAP
 		      if (!IS_MAPPED(hhdr)) {
-		        GC_remap((ptr_t)hbp, size_avail);
+		        GC_remap((ptr_t)hbp, hhdr -> hb_sz);
 		        hhdr -> hb_flags &= ~WAS_UNMAPPED;
 		      }
 #		    endif
 	          /* Split the block at thishbp */
-	              thishdr = HDR(thishbp);
 		      GC_split_block(hbp, hhdr, thishbp, thishdr, n);
 		  /* Advance to thishbp */
 		      hbp = thishbp;
@@ -597,8 +615,7 @@ int n;
 		      GC_large_free_bytes -= total_size;
 		      GC_remove_from_fl(hhdr, n);
 	              for (h = hbp; h < limit; h++) {
-	                if (h == hbp || GC_install_header(h)) {
-	                  hhdr = HDR(h);
+	                if (h == hbp || 0 != (hhdr = GC_install_header(h))) {
 	                  (void) setup_header(
 	                	  hhdr,
 	              		  BYTES_TO_WORDS(HBLKSIZE - HDR_BYTES),
@@ -621,7 +638,7 @@ int n;
 	    if( size_avail >= size_needed ) {
 #		ifdef USE_MUNMAP
 		  if (!IS_MAPPED(hhdr)) {
-		    GC_remap((ptr_t)hbp, size_avail);
+		    GC_remap((ptr_t)hbp, hhdr -> hb_sz);
 		    hhdr -> hb_flags &= ~WAS_UNMAPPED;
 		  }
 #	        endif
@@ -685,7 +702,7 @@ hdr *hhdr, *prevhdr, *nexthdr;
 signed_word size;
 
 
-    hhdr = HDR(hbp);
+    GET_HDR(hbp, hhdr);
     size = hhdr->hb_sz;
     size = HBLKSIZE * OBJ_SZ_TO_BLOCKS(size);
     GC_remove_counts(hbp, (word)size);
@@ -700,7 +717,7 @@ signed_word size;
     GC_ASSERT(IS_MAPPED(hhdr));
     GC_invalidate_map(hhdr);
     next = (struct hblk *)((word)hbp + size);
-    nexthdr = HDR(next);
+    GET_HDR(next, nexthdr);
     prev = GC_free_block_ending_at(hbp);
     /* Coalesce with successor, if possible */
       if(0 != nexthdr && HBLK_IS_FREE(nexthdr) && IS_MAPPED(nexthdr)) {
diff --git a/alloc.c b/alloc.c
index 1c57951f..e11ccdea 100644
--- a/alloc.c
+++ b/alloc.c
@@ -1,7 +1,8 @@
 /*
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
- * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1991-1996 by Xerox Corporation.  All rights reserved.
  * Copyright (c) 1998 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -62,8 +63,14 @@ word GC_gc_no = 0;
   int GC_incremental = 0;    /* By default, stop the world.	*/
 #endif
 
-int GC_full_freq = 4;	   /* Every 5th collection is a full	*/
-			   /* collection.			*/
+int GC_full_freq = 19;	   /* Every 20th collection is a full	*/
+			   /* collection, whether we need it 	*/
+			   /* or not.			        */
+
+GC_bool GC_need_full_gc = FALSE;
+			   /* Need full GC do to heap growth.	*/
+
+word GC_used_heap_size_after_full = 0;
 
 char * GC_copyright[] =
 {"Copyright 1988,1989 Hans-J. Boehm and Alan J. Demers ",
@@ -210,6 +217,7 @@ GC_bool GC_should_collect()
     return(GC_adj_words_allocd() >= min_words_allocd());
 }
 
+
 void GC_notify_full_gc()
 {
     if (GC_start_call_back != (void (*)())0) {
@@ -217,6 +225,8 @@ void GC_notify_full_gc()
     }
 }
 
+GC_bool GC_is_full_gc = FALSE;
+
 /* 
  * Initiate a garbage collection if appropriate.
  * Choose judiciously
@@ -226,7 +236,6 @@ void GC_notify_full_gc()
 void GC_maybe_gc()
 {
     static int n_partial_gcs = 0;
-    GC_bool is_full_gc = FALSE;
 
     if (GC_should_collect()) {
         if (!GC_incremental) {
@@ -234,7 +243,7 @@ void GC_maybe_gc()
             GC_gcollect_inner();
             n_partial_gcs = 0;
             return;
-        } else if (n_partial_gcs >= GC_full_freq) {
+        } else if (GC_need_full_gc || n_partial_gcs >= GC_full_freq) {
 #   	    ifdef PRINTSTATS
 	      GC_printf2(
 	        "***>Full mark for collection %lu after %ld allocd bytes\n",
@@ -246,7 +255,7 @@ void GC_maybe_gc()
 	    GC_clear_marks();
             n_partial_gcs = 0;
 	    GC_notify_full_gc();
- 	    is_full_gc = TRUE;
+ 	    GC_is_full_gc = TRUE;
         } else {
             n_partial_gcs++;
         }
@@ -260,7 +269,7 @@ void GC_maybe_gc()
 #           endif
             GC_finish_collection();
         } else {
-	    if (!is_full_gc) {
+	    if (!GC_is_full_gc) {
 		/* Count this as the first attempt */
 	        GC_n_attempts++;
 	    }
@@ -297,7 +306,9 @@ GC_stop_func stop_func;
     /* Make sure all blocks have been reclaimed, so sweep routines	*/
     /* don't see cleared mark bits.					*/
     /* If we're guaranteed to finish, then this is unnecessary.		*/
-	if (stop_func != GC_never_stop_func
+    /* In the find_leak case, we have to finish to guarantee that 	*/
+    /* previously unmarked objects are not reported as leaks.		*/
+	if ((GC_find_leak || stop_func != GC_never_stop_func)
 	    && !GC_reclaim_all(stop_func, FALSE)) {
 	    /* Aborted.  So far everything is still consistent.	*/
 	    return(FALSE);
@@ -307,6 +318,7 @@ GC_stop_func stop_func;
 #   ifdef SAVE_CALL_CHAIN
         GC_save_callers(GC_last_stack);
 #   endif
+    GC_is_full_gc = TRUE;
     if (!GC_stopped_mark(stop_func)) {
       if (!GC_incremental) {
     	/* We're partially done and have no way to complete or use 	*/
@@ -548,6 +560,14 @@ void GC_finish_collection()
 #   endif
     /* Reconstruct free lists to contain everything not marked */
         GC_start_reclaim(FALSE);
+        if (GC_is_full_gc)  {
+	    GC_used_heap_size_after_full = USED_HEAP_SIZE;
+	    GC_need_full_gc = FALSE;
+	} else {
+	    GC_need_full_gc =
+		 BYTES_TO_WORDS(USED_HEAP_SIZE - GC_used_heap_size_after_full)
+		 > min_words_allocd();
+	}
 
 #   ifdef PRINTSTATS
 	GC_printf2(
@@ -564,6 +584,7 @@ void GC_finish_collection()
 #   endif
 
       GC_n_attempts = 0;
+      GC_is_full_gc = FALSE;
     /* 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;
@@ -630,7 +651,8 @@ word bytes;
     if (GC_n_heap_sects >= MAX_HEAP_SECTS) {
     	ABORT("Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS");
     }
-    if (!GC_install_header(p)) {
+    phdr = GC_install_header(p);
+    if (0 == phdr) {
     	/* This is extremely unlikely. Can't add it.  This will		*/
     	/* almost certainly result in a	0 return from the allocator,	*/
     	/* which is entirely appropriate.				*/
@@ -640,7 +662,6 @@ word bytes;
     GC_heap_sects[GC_n_heap_sects].hs_bytes = bytes;
     GC_n_heap_sects++;
     words = BYTES_TO_WORDS(bytes - HDR_BYTES);
-    phdr = HDR(p);
     phdr -> hb_sz = words;
     phdr -> hb_map = (char *)1;   /* A value != GC_invalid_map	*/
     phdr -> hb_flags = 0;
@@ -789,6 +810,7 @@ word n;
     LOCK();
     if (!GC_is_initialized) GC_init_inner();
     result = (int)GC_expand_hp_inner(divHBLKSZ((word)bytes));
+    if (result) GC_requested_heapsize += bytes;
     UNLOCK();
     ENABLE_SIGNALS();
     return(result);
@@ -802,7 +824,8 @@ GC_bool GC_collect_or_expand(needed_blocks, ignore_off_page)
 word needed_blocks;
 GC_bool ignore_off_page;
 {
-    if (!GC_incremental && !GC_dont_gc && GC_should_collect()) {
+    if (!GC_incremental && !GC_dont_gc &&
+	(GC_dont_expand && GC_words_allocd > 0 || GC_should_collect())) {
       GC_notify_full_gc();
       GC_gcollect_inner();
     } else {
diff --git a/alpha_mach_dep.s b/alpha_mach_dep.s
index 087baeac..783ebf1f 100644
--- a/alpha_mach_dep.s
+++ b/alpha_mach_dep.s
@@ -1,5 +1,9 @@
  # $Id: alpha_mach_dep.s,v 1.2 1993/01/18 22:54:51 dosser Exp $
 
+# This is BROKEN on a 21264 running gcc, and probably in other cases.
+# The compiler may spill pointers to fp registers, and this code doesn't
+# scan those.
+
 # define call_push(x)    						\
 	lda   $16, 0(x);    	/* copy x to first argument register */	\
 	jsr   $26, GC_push_one; /* call GC_push_one, ret addr in $26 */	\
diff --git a/backptr.h b/backptr.h
index d34224e4..f5b7c5aa 100644
--- a/backptr.h
+++ b/backptr.h
@@ -29,9 +29,11 @@
 /*   source is heap object ==> *base_p != 0, *offset_p = offset       */
 /*   Returns 1 on success, 0 if source couldn't be determined.        */
 /* Dest can be any address within a heap object.                      */
-typedef enum {  GC_UNREFERENCED, /* No refence info available.		*/
+typedef enum {  GC_UNREFERENCED, /* No reference info available.	*/
 		GC_NO_SPACE,	/* Dest not allocated with debug alloc  */
 		GC_REFD_FROM_ROOT, /* Referenced directly by root *base_p */
+		GC_REFD_FROM_REG,  /* Referenced from a register, i.e.	*/
+				   /* a root without an address.	*/
 		GC_REFD_FROM_HEAP, /* Referenced from another heap obj. */
 		GC_FINALIZER_REFD /* Finalizable and hence accessible.  */
 } GC_ref_kind;
@@ -53,4 +55,9 @@ void * GC_generate_random_valid_address(void);
 /* source in dbg_mlc.c also serves as a sample client.	      */
 void GC_generate_random_backtrace(void);
 
+/* Print a backtrace from a specific address.  Used by the 	*/
+/* above.  The client should call GC_gcollect() immediately	*/
+/* before invocation.						*/
+void GC_print_backtrace(void *);
+
 
diff --git a/blacklst.c b/blacklst.c
index 0d623c0f..e5a3a26a 100644
--- a/blacklst.c
+++ b/blacklst.c
@@ -145,6 +145,13 @@ void GC_promote_black_lists()
     if (GC_black_list_spacing < 3 * HBLKSIZE) {
     	GC_black_list_spacing = 3 * HBLKSIZE;
     }
+    if (GC_black_list_spacing > MAXHINCR * HBLKSIZE) {
+	GC_black_list_spacing = MAXHINCR * HBLKSIZE;
+	/* Makes it easier to allocate really huge blocks, which otherwise */
+	/* may have problems with nonuniform blacklist distributions.	   */
+	/* This way we should always succeed immediately after growing the */ 
+	/* heap.							   */
+    }
 }
 
 void GC_unpromote_black_lists()
diff --git a/cord/cord.h b/cord/cord.h
index 584112fd..926089e8 100644
--- a/cord/cord.h
+++ b/cord/cord.h
@@ -41,7 +41,7 @@
  * This interface is fairly big, largely for performance reasons.
  * The most basic constants and functions:
  *
- * CORD - the type fo a cord;
+ * CORD - the type of a cord;
  * CORD_EMPTY - empty cord;
  * CORD_len(cord) - length of a cord;
  * CORD_cat(cord1,cord2) - concatenation of two cords;
diff --git a/cord/cordprnt.c b/cord/cordprnt.c
index 9c8cc873..8d57f046 100644
--- a/cord/cordprnt.c
+++ b/cord/cordprnt.c
@@ -233,7 +233,7 @@ int CORD_vsprintf(CORD * out, CORD format, va_list args)
 			if (width == NONE && prec == NONE) {
 			    register char c;
 
-			    c = va_arg(args, char);
+			    c = va_arg(args, int);
 			    CORD_ec_append(result, c);
 			    goto done;
 			}
diff --git a/cord/gc.h b/cord/gc.h
index 30614095..d7cd8996 100644
--- a/cord/gc.h
+++ b/cord/gc.h
@@ -1,7 +1,8 @@
 /* 
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
- * Copyright 1996 by Silicon Graphics.  All rights reserved.
+ * Copyright 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright 1999 by Hewlett-Packard Company.  All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -35,6 +36,14 @@
 #include "libgc_globals.h"
 #endif
 
+#if defined(__MINGW32__) && defined(WIN32_THREADS)
+# ifdef GC_BUILD
+#   define GC_API __declspec(dllexport)
+# else
+#   define GC_API __declspec(dllimport)
+# endif
+#endif
+
 #if defined(_MSC_VER) && defined(_DLL)
 # ifdef GC_BUILD
 #   define GC_API __declspec(dllexport)
@@ -58,9 +67,11 @@
 # if defined(__STDC__) || defined(__cplusplus)
 #   define GC_PROTO(args) args
     typedef void * GC_PTR;
+#   define GC_CONST const
 # else
 #   define GC_PROTO(args) ()
     typedef char * GC_PTR;
+#   define GC_CONST
 #  endif
 
 # ifdef __cplusplus
@@ -128,9 +139,26 @@ GC_API int GC_dont_expand;
 			/* Dont expand heap unless explicitly requested */
 			/* or forced to.				*/
 
+GC_API int GC_use_entire_heap;
+		/* Causes the nonincremental collector to use the	*/
+		/* entire heap before collecting.  This was the only 	*/
+		/* option for GC versions < 5.0.  This sometimes	*/
+		/* results in more large block fragmentation, since	*/
+		/* very larg blocks will tend to get broken up		*/
+		/* during each GC cycle.  It is likely to result in a	*/
+		/* larger working set, but lower collection		*/
+		/* frequencies, and hence fewer instructions executed	*/
+		/* in the collector.					*/
+
 GC_API int GC_full_freq;    /* Number of partial collections between	*/
 			    /* full collections.  Matters only if	*/
 			    /* GC_incremental is set.			*/
+			    /* Full collections are also triggered if	*/
+			    /* the collector detects a substantial	*/
+			    /* increase in the number of in-use heap	*/
+			    /* blocks.  Values in the tens are now	*/
+			    /* perfectly reasonable, unlike for		*/
+			    /* earlier GC versions.			*/
 			
 GC_API GC_word GC_non_gc_bytes;
 			/* Bytes not considered candidates for collection. */
@@ -297,6 +325,9 @@ GC_API int GC_try_to_collect GC_PROTO((GC_stop_func stop_func));
 /* Includes some pages that were allocated but never written.		*/
 GC_API size_t GC_get_heap_size GC_PROTO((void));
 
+/* Return a lower bound on the number of free bytes in the heap.	*/
+GC_API size_t GC_get_free_bytes GC_PROTO((void));
+
 /* Return the number of bytes allocated since the last collection.	*/
 GC_API size_t GC_get_bytes_since_gc GC_PROTO((void));
 
@@ -341,10 +372,10 @@ GC_API GC_PTR GC_malloc_atomic_ignore_off_page GC_PROTO((size_t lb));
 
 #ifdef GC_ADD_CALLER
 #  define GC_EXTRAS GC_RETURN_ADDR, __FILE__, __LINE__
-#  define GC_EXTRA_PARAMS GC_word ra, char * descr_string, int descr_int
+#  define GC_EXTRA_PARAMS GC_word ra, GC_CONST char * s, int i
 #else
 #  define GC_EXTRAS __FILE__, __LINE__
-#  define GC_EXTRA_PARAMS char * descr_string, int descr_int
+#  define GC_EXTRA_PARAMS GC_CONST char * s, int i
 #endif
 
 /* Debugging (annotated) allocation.  GC_gcollect will check 		*/
@@ -375,6 +406,8 @@ GC_API void GC_debug_end_stubborn_change GC_PROTO((GC_PTR));
 	GC_debug_register_finalizer(p, f, d, of, od)
 #   define GC_REGISTER_FINALIZER_IGNORE_SELF(p, f, d, of, od) \
 	GC_debug_register_finalizer_ignore_self(p, f, d, of, od)
+#   define GC_REGISTER_FINALIZER_NO_ORDER(p, f, d, of, od) \
+	GC_debug_register_finalizer_no_order(p, f, d, of, od)
 #   define GC_MALLOC_STUBBORN(sz) GC_debug_malloc_stubborn(sz, GC_EXTRAS);
 #   define GC_CHANGE_STUBBORN(p) GC_debug_change_stubborn(p)
 #   define GC_END_STUBBORN_CHANGE(p) GC_debug_end_stubborn_change(p)
@@ -391,6 +424,8 @@ GC_API void GC_debug_end_stubborn_change GC_PROTO((GC_PTR));
 	GC_register_finalizer(p, f, d, of, od)
 #   define GC_REGISTER_FINALIZER_IGNORE_SELF(p, f, d, of, od) \
 	GC_register_finalizer_ignore_self(p, f, d, of, od)
+#   define GC_REGISTER_FINALIZER_NO_ORDER(p, f, d, of, od) \
+	GC_register_finalizer_no_order(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)
@@ -469,6 +504,16 @@ GC_API void GC_debug_register_finalizer_ignore_self
 	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
 		  GC_finalization_proc *ofn, GC_PTR *ocd));
 
+/* Another version of the above.  It ignores all cycles.        */
+/* It should probably only be used by Java implementations.      */
+GC_API void GC_register_finalizer_no_order
+	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
+		  GC_finalization_proc *ofn, GC_PTR *ocd));
+GC_API void GC_debug_register_finalizer_no_order
+	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
+		  GC_finalization_proc *ofn, GC_PTR *ocd));
+
+
 /* 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	*/
@@ -525,6 +570,9 @@ GC_API int GC_unregister_disappearing_link GC_PROTO((GC_PTR * /* link */));
 GC_API GC_PTR GC_make_closure GC_PROTO((GC_finalization_proc fn, GC_PTR data));
 GC_API void GC_debug_invoke_finalizer GC_PROTO((GC_PTR obj, GC_PTR data));
 
+/* Returns !=0  if GC_invoke_finalizers has something to do. 		*/
+GC_API int GC_should_invoke_finalizers GC_PROTO((void));
+
 GC_API int GC_invoke_finalizers GC_PROTO((void));
 	/* Run finalizers for all objects that are ready to	*/
 	/* be finalized.  Return the number of finalizers	*/
@@ -688,7 +736,8 @@ GC_API void (*GC_is_visible_print_proc)
 # endif /* SOLARIS_THREADS */
 
 
-#if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#if !defined(USE_LD_WRAP) && \
+    (defined(IRIX_THREADS) || defined(LINUX_THREADS) || defined(HPUX_THREADS))
 /* We treat these similarly. */
 # include <pthread.h>
 # include <signal.h>
@@ -702,12 +751,13 @@ GC_API void (*GC_is_visible_print_proc)
 # define pthread_create GC_pthread_create
 # define pthread_sigmask GC_pthread_sigmask
 # define pthread_join GC_pthread_join
+# define dlopen GC_dlopen
 
-#endif /* IRIX_THREADS || LINUX_THREADS */
+#endif /* xxxxx_THREADS */
 
 # if defined(PCR) || defined(SOLARIS_THREADS) || defined(WIN32_THREADS) || \
 	defined(IRIX_THREADS) || defined(LINUX_THREADS) || \
-	defined(IRIX_JDK_THREADS)
+	defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
    	/* Any flavor of threads except SRC_M3.	*/
 /* This returns a list of objects, linked through their first		*/
 /* word.  Its use can greatly reduce lock contention problems, since	*/
diff --git a/dbg_mlc.c b/dbg_mlc.c
index 64832567..20768715 100644
--- a/dbg_mlc.c
+++ b/dbg_mlc.c
@@ -2,6 +2,7 @@
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
  * Copyright (c) 1997 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company.  All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -12,64 +13,14 @@
  * provided the above notices are retained, and a notice that the code was
  * modified is included with the above copyright notice.
  */
-# define I_HIDE_POINTERS
-# include "gc_priv.h"
-# ifdef KEEP_BACK_PTRS
-#   include "backptr.h"
-# endif
+
+#include "dbg_mlc.h"
 
 void GC_default_print_heap_obj_proc();
 GC_API void GC_register_finalizer_no_order
     	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
 		  GC_finalization_proc *ofn, GC_PTR *ocd));
 
-/* 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?	*/
-
-# 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.		*/
-
-
-/* Object header */
-typedef struct {
-#   ifdef KEEP_BACK_PTRS
-	ptr_t oh_back_ptr;
-#	define MARKED_FOR_FINALIZATION (ptr_t)(-1)
-	    /* Object was marked because it is finalizable. */
-#	ifdef ALIGN_DOUBLE
-	  word oh_dummy;
-#	endif
-#   endif
-    char * oh_string;		/* object descriptor string	*/
-    word oh_int;		/* object descriptor integers	*/
-#   ifdef NEED_CALLINFO
-      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)
-
-
-#ifdef SAVE_CALL_CHAIN
-#   define ADD_CALL_CHAIN(base, ra) GC_save_callers(((oh *)(base)) -> oh_ci)
-#   define PRINT_CALL_CHAIN(base) GC_print_callers(((oh *)(base)) -> oh_ci)
-#else
-# ifdef GC_ADD_CALLER
-#   define ADD_CALL_CHAIN(base, ra) ((oh *)(base)) -> oh_ci[0].ci_pc = (ra)
-#   define PRINT_CALL_CHAIN(base) GC_print_callers(((oh *)(base)) -> oh_ci)
-# else
-#   define ADD_CALL_CHAIN(base, ra)
-#   define PRINT_CALL_CHAIN(base)
-# endif
-#endif
 
 /* Check whether object with base pointer p has debugging info	*/ 
 /* p is assumed to point to a legitimate object in our part	*/
@@ -116,7 +67,7 @@ ptr_t p;
 
   /* Store information about the object referencing dest in *base_p	*/
   /* and *offset_p.							*/
-  /*   source is root ==> *base_p = 0, *offset_p = address		*/
+  /*   source is root ==> *base_p = address, *offset_p = 0		*/
   /*   source is heap object ==> *base_p != 0, *offset_p = offset 	*/
   /*   Returns 1 on success, 0 if source couldn't be determined.	*/
   /* Dest can be any address within a heap object.			*/
@@ -128,6 +79,7 @@ ptr_t p;
     if (!GC_has_debug_info((ptr_t) hdr)) return GC_NO_SPACE;
     bp = hdr -> oh_back_ptr;
     if (MARKED_FOR_FINALIZATION == bp) return GC_FINALIZER_REFD;
+    if (MARKED_FROM_REGISTER == bp) return GC_REFD_FROM_REG;
     if (0 == bp) return GC_UNREFERENCED;
     bp = REVEAL_POINTER(bp);
     bp_base = GC_base(bp);
@@ -177,18 +129,15 @@ ptr_t p;
     }
   }
 
-  /* Force a garbage collection and generate a backtrace from a	*/
-  /* random heap address.					*/
-  void GC_generate_random_backtrace(void)
+  /* Print back trace for p */
+  void GC_print_backtrace(void *p)
   {
-    void * current;
+    void *current = p;
     int i;
-    void * base;
-    size_t offset;
     GC_ref_kind source;
-    GC_gcollect();
-    current = GC_generate_random_valid_address();
-    GC_printf1("Chose address 0x%lx in object\n", (unsigned long)current);
+    size_t offset;
+    void *base;
+
     GC_print_heap_obj(GC_base(current));
     GC_err_printf0("\n");
     for (i = 0; ; ++i) {
@@ -207,6 +156,9 @@ ptr_t p;
 	case GC_REFD_FROM_ROOT:
 	  GC_err_printf1("root at 0x%lx\n", (unsigned long)base);
 	  goto out;
+	case GC_REFD_FROM_REG:
+	  GC_err_printf0("root in register\n");
+	  goto out;
 	case GC_FINALIZER_REFD:
 	  GC_err_printf0("list of finalizable objects\n");
 	  goto out;
@@ -221,6 +173,17 @@ ptr_t p;
     }
     out:;
   }
+
+  /* Force a garbage collection and generate a backtrace from a	*/
+  /* random heap address.					*/
+  void GC_generate_random_backtrace(void)
+  {
+    void * current;
+    GC_gcollect();
+    current = GC_generate_random_valid_address();
+    GC_printf1("Chose address 0x%lx in object\n", (unsigned long)current);
+    GC_print_backtrace(current);
+  }
     
 #endif /* KEEP_BACK_PTRS */
 
@@ -342,16 +305,8 @@ void GC_start_debugging()
     GC_register_displacement((word)sizeof(oh) + offset);
 }
 
-# ifdef GC_ADD_CALLER
-#   define EXTRA_ARGS word ra, char * s, int i
-#   define OPT_RA ra,
-# else
-#   define EXTRA_ARGS char * s, int i
-#   define OPT_RA
-# endif
-
 # ifdef __STDC__
-    GC_PTR GC_debug_malloc(size_t lb, EXTRA_ARGS)
+    GC_PTR GC_debug_malloc(size_t lb, GC_EXTRA_PARAMS)
 # else
     GC_PTR GC_debug_malloc(lb, s, i)
     size_t lb;
@@ -380,7 +335,7 @@ void GC_start_debugging()
 
 #ifdef STUBBORN_ALLOC
 # ifdef __STDC__
-    GC_PTR GC_debug_malloc_stubborn(size_t lb, EXTRA_ARGS)
+    GC_PTR GC_debug_malloc_stubborn(size_t lb, GC_EXTRA_PARAMS)
 # else
     GC_PTR GC_debug_malloc_stubborn(lb, s, i)
     size_t lb;
@@ -444,10 +399,34 @@ GC_PTR p;
     GC_end_stubborn_change(q);
 }
 
-#endif /* STUBBORN_ALLOC */
+#else /* !STUBBORN_ALLOC */
+
+# ifdef __STDC__
+    GC_PTR GC_debug_malloc_stubborn(size_t lb, GC_EXTRA_PARAMS)
+# else
+    GC_PTR GC_debug_malloc_stubborn(lb, s, i)
+    size_t lb;
+    char * s;
+    int i;
+# endif
+{
+    return GC_debug_malloc(lb, OPT_RA s, i);
+}
+
+void GC_debug_change_stubborn(p)
+GC_PTR p;
+{
+}
+
+void GC_debug_end_stubborn_change(p)
+GC_PTR p;
+{
+}
+
+#endif /* !STUBBORN_ALLOC */
 
 # ifdef __STDC__
-    GC_PTR GC_debug_malloc_atomic(size_t lb, EXTRA_ARGS)
+    GC_PTR GC_debug_malloc_atomic(size_t lb, GC_EXTRA_PARAMS)
 # else
     GC_PTR GC_debug_malloc_atomic(lb, s, i)
     size_t lb;
@@ -472,7 +451,7 @@ GC_PTR p;
 }
 
 # ifdef __STDC__
-    GC_PTR GC_debug_malloc_uncollectable(size_t lb, EXTRA_ARGS)
+    GC_PTR GC_debug_malloc_uncollectable(size_t lb, GC_EXTRA_PARAMS)
 # else
     GC_PTR GC_debug_malloc_uncollectable(lb, s, i)
     size_t lb;
@@ -498,7 +477,7 @@ GC_PTR p;
 
 #ifdef ATOMIC_UNCOLLECTABLE
 # ifdef __STDC__
-    GC_PTR GC_debug_malloc_atomic_uncollectable(size_t lb, EXTRA_ARGS)
+    GC_PTR GC_debug_malloc_atomic_uncollectable(size_t lb, GC_EXTRA_PARAMS)
 # else
     GC_PTR GC_debug_malloc_atomic_uncollectable(lb, s, i)
     size_t lb;
@@ -531,13 +510,15 @@ GC_PTR p;
     GC_PTR p;
 # endif
 {
-    register GC_PTR base = GC_base(p);
+    register GC_PTR base;
     register ptr_t clobbered;
     
+    if (0 == p) return;
+    base = GC_base(p);
     if (base == 0) {
         GC_err_printf1("Attempt to free invalid pointer %lx\n",
         	       (unsigned long)p);
-        if (p != 0) ABORT("free(invalid pointer)");
+        ABORT("free(invalid pointer)");
     }
     if ((ptr_t)p - (ptr_t)base != sizeof(oh)) {
         GC_err_printf1(
@@ -576,7 +557,7 @@ GC_PTR p;
 }
 
 # ifdef __STDC__
-    GC_PTR GC_debug_realloc(GC_PTR p, size_t lb, EXTRA_ARGS)
+    GC_PTR GC_debug_realloc(GC_PTR p, size_t lb, GC_EXTRA_PARAMS)
 # else
     GC_PTR GC_debug_realloc(p, lb, s, i)
     GC_PTR p;
diff --git a/dbg_mlc.h b/dbg_mlc.h
new file mode 100644
index 00000000..6abeda8a
--- /dev/null
+++ b/dbg_mlc.h
@@ -0,0 +1,102 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1997 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company.  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 mostly an internal header file.  Typical clients should
+ * not use it.  Clients that define their own object kinds with
+ * debugging allocators will probably want to include this, however.
+ * No attempt is made to keep the namespace clean.  This should not be
+ * included from header filrd that are frequently included by clients.
+ */
+
+#ifndef _DBG_MLC_H
+
+#define _DBG_MLC_H
+
+# define I_HIDE_POINTERS
+# include "gc_priv.h"
+# ifdef KEEP_BACK_PTRS
+#   include "backptr.h"
+# 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.		*/
+
+
+/* Object header */
+typedef struct {
+#   ifdef KEEP_BACK_PTRS
+	ptr_t oh_back_ptr;
+#	define MARKED_FOR_FINALIZATION (ptr_t)(-1)
+	    /* Object was marked because it is finalizable.	*/
+#	define MARKED_FROM_REGISTER (ptr_t)(-2)
+	    /* Object was marked from a rgister.  Hence the	*/
+	    /* source of the reference doesn't have an address.	*/
+#	ifdef ALIGN_DOUBLE
+	  word oh_dummy;
+#	endif
+#   endif
+    char * oh_string;		/* object descriptor string	*/
+    word oh_int;		/* object descriptor integers	*/
+#   ifdef NEED_CALLINFO
+      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))
+#define USR_PTR_FROM_BASE(p) ((ptr_t)(p) + sizeof(oh))
+
+/* There is no reason to ever add a byte at the end explicitly, since we */
+/* already add a guard word.						 */
+#undef ROUNDED_UP_WORDS
+#define ROUNDED_UP_WORDS(n) BYTES_TO_WORDS((n) + WORDS_TO_BYTES(1) - 1)
+
+#ifdef SAVE_CALL_CHAIN
+#   define ADD_CALL_CHAIN(base, ra) GC_save_callers(((oh *)(base)) -> oh_ci)
+#   define PRINT_CALL_CHAIN(base) GC_print_callers(((oh *)(base)) -> oh_ci)
+#else
+# ifdef GC_ADD_CALLER
+#   define ADD_CALL_CHAIN(base, ra) ((oh *)(base)) -> oh_ci[0].ci_pc = (ra)
+#   define PRINT_CALL_CHAIN(base) GC_print_callers(((oh *)(base)) -> oh_ci)
+# else
+#   define ADD_CALL_CHAIN(base, ra)
+#   define PRINT_CALL_CHAIN(base)
+# endif
+#endif
+
+# ifdef GC_ADD_CALLER
+#   define OPT_RA ra,
+# else
+#   define OPT_RA
+# 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.							*/
+GC_bool GC_has_debug_info(/* p */);
+
+/* 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 */);
+
+#endif /* _DBG_MLC_H */
diff --git a/dyn_load.c b/dyn_load.c
index d3df0a08..74aacd71 100644
--- a/dyn_load.c
+++ b/dyn_load.c
@@ -32,7 +32,9 @@
 #include "gc_priv.h"
 
 /* BTL: avoid circular redefinition of dlopen if SOLARIS_THREADS defined */
-# if defined(SOLARIS_THREADS) && defined(dlopen)
+# if (defined(LINUX_THREADS) || defined(SOLARIS_THREADS) \
+      || defined(HPUX_THREADS) || defined(IRIX_THREADS)) && defined(dlopen) \
+     && !defined(USE_LD_WRAP)
     /* To support threads in Solaris, gc.h interposes on dlopen by       */
     /* defining "dlopen" to be "GC_dlopen", which is implemented below.  */
     /* However, both GC_FirstDLOpenedLinkMap() and GC_dlopen() use the   */
@@ -47,7 +49,7 @@
 #if (defined(DYNAMIC_LOADING) || defined(MSWIN32)) && !defined(PCR)
 #if !defined(SUNOS4) && !defined(SUNOS5DL) && !defined(IRIX5) && \
     !defined(MSWIN32) && !(defined(ALPHA) && defined(OSF1)) && \
-    !defined(HP_PA) && !(defined(LINUX) && defined(__ELF__)) && \
+    !defined(HPUX) && !(defined(LINUX) && defined(__ELF__)) && \
     !defined(RS6000) && !defined(SCO_ELF)
  --> We only know how to find data segments of dynamic libraries for the
  --> above.  Additional SVR4 variants might not be too
@@ -159,32 +161,59 @@ static ptr_t GC_first_common()
 
 #endif  /* SUNOS4 ... */
 
-# if defined(SUNOS4) || defined(SUNOS5DL)
-/* 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
+# if defined(LINUX_THREADS) || defined(SOLARIS_THREADS) \
+     || defined(HPUX_THREADS) || defined(IRIX_THREADS)
+  /* Make sure we're not in the middle of a collection, and make	*/
+  /* sure we don't start any.	Returns previous value of GC_dont_gc.	*/
+  /* This is invoked prior to a dlopen call to avoid synchronization	*/
+  /* issues.  We can't just acquire the allocation lock, since startup 	*/
+  /* code in dlopen may try to allocate.				*/
+  /* This solution risks heap growth in the presence of many dlopen	*/
+  /* calls in either a multithreaded environment, or if the library	*/
+  /* initialization code allocates substantial amounts of GC'ed memory.	*/
+  /* But I don't know of a better solution.				*/
+  /* This can still deadlock if the client explicitly starts a GC 	*/
+  /* during the dlopen.  He shouldn't do that.				*/
+  static GC_bool disable_gc_for_dlopen()
+  {
+    GC_bool result;
+    LOCK();
+    result = GC_dont_gc;
+    while (GC_incremental && GC_collection_in_progress()) {
+	GC_collect_a_little_inner(1000);
+    }
+    GC_dont_gc = TRUE;
+    UNLOCK();
+    return(result);
+  }
 
-# 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>
+  /* Should probably happen for other operating	systems, too. */
+
+#include <dlfcn.h>
 
-void * GC_dlopen(const char *path, int mode)
+#ifdef USE_LD_WRAP
+  void * __wrap_dlopen(const char *path, int mode)
+#else
+  void * GC_dlopen(path, mode)
+  GC_CONST char * path;
+  int mode;
+#endif
 {
     void * result;
+    GC_bool dont_gc_save;
     
 #   ifndef USE_PROC_FOR_LIBRARIES
-      mutex_lock(&GC_allocate_ml);
+      dont_gc_save = disable_gc_for_dlopen();
+#   endif
+#   ifdef USE_LD_WRAP
+      result = __real_dlopen(path, mode);
+#   else
+      result = dlopen(path, mode);
 #   endif
-    result = dlopen(path, mode);
 #   ifndef USE_PROC_FOR_LIBRARIES
-      mutex_unlock(&GC_allocate_ml);
+      GC_dont_gc = dont_gc_save;
 #   endif
     return(result);
 }
@@ -195,6 +224,14 @@ void * GC_dlopen(const char *path, int mode)
 #   define dlopen GC_dlopen
 # endif
 
+# if defined(SUNOS4) || defined(SUNOS5DL)
+/* 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
+
 # ifndef USE_PROC_FOR_LIBRARIES
 void GC_register_dynamic_libraries()
 {
@@ -356,7 +393,9 @@ extern void * GC_roots_present();
 	/* The type is a lie, since the real type doesn't make sense here, */
 	/* and we only test for NULL.					   */
 
+#ifndef GC_scratch_last_end_ptr   /* Never an extern any more? */
 extern ptr_t GC_scratch_last_end_ptr; /* End of GC_scratch_alloc arena	*/
+#endif
 
 /* 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	*/
@@ -658,7 +697,7 @@ void GC_register_dynamic_libraries()
 }
 #endif
 
-#if defined(HP_PA)
+#if defined(HPUX)
 
 #include <errno.h>
 #include <dl.h>
@@ -681,6 +720,11 @@ void GC_register_dynamic_libraries()
 
       /* Check if this is the end of the list or if some error occured */
         if (status != 0) {
+#	 ifdef HPUX_THREADS
+	   /* I've seen errno values of 0.  The man page is not clear	*/
+	   /* as to whether errno should get set on a -1 return.	*/
+	   break;
+#	 else
           if (errno == EINVAL) {
               break; /* Moved past end of shared library list --> finished */
           } else {
@@ -691,6 +735,7 @@ void GC_register_dynamic_libraries()
 	      }
               ABORT("shl_get failed");
           }
+#	 endif
         }
 
 #     ifdef VERBOSE
@@ -713,7 +758,7 @@ void GC_register_dynamic_libraries()
         index++;
     }
 }
-#endif /* HP_PA */
+#endif /* HPUX */
 
 #ifdef RS6000
 #pragma alloca
diff --git a/finalize.c b/finalize.c
index 2ee927fe..e70a9ff1 100644
--- a/finalize.c
+++ b/finalize.c
@@ -619,7 +619,7 @@ void GC_finalize()
 void GC_enqueue_all_finalizers()
 {
     struct finalizable_object * curr_fo, * prev_fo, * next_fo;
-    ptr_t real_ptr, real_link;
+    ptr_t real_ptr;
     register int i;
     int fo_size;
     
@@ -694,6 +694,14 @@ GC_API void GC_finalize_all()
 }
 #endif
 
+/* Returns true if it is worth calling GC_invoke_finalizers. (Useful if	*/
+/* finalizers can only be called from some kind of `safe state' and	*/
+/* getting into that safe state is expensive.)				*/
+int GC_should_invoke_finalizers GC_PROTO((void))
+{
+    return GC_finalize_now != 0;
+}
+
 /* Invoke finalizers for all objects that are ready to be finalized.	*/
 /* Should be called without allocation lock.				*/
 int GC_invoke_finalizers()
diff --git a/gc.h b/gc.h
index 30614095..d7cd8996 100644
--- a/gc.h
+++ b/gc.h
@@ -1,7 +1,8 @@
 /* 
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
- * Copyright 1996 by Silicon Graphics.  All rights reserved.
+ * Copyright 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright 1999 by Hewlett-Packard Company.  All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -35,6 +36,14 @@
 #include "libgc_globals.h"
 #endif
 
+#if defined(__MINGW32__) && defined(WIN32_THREADS)
+# ifdef GC_BUILD
+#   define GC_API __declspec(dllexport)
+# else
+#   define GC_API __declspec(dllimport)
+# endif
+#endif
+
 #if defined(_MSC_VER) && defined(_DLL)
 # ifdef GC_BUILD
 #   define GC_API __declspec(dllexport)
@@ -58,9 +67,11 @@
 # if defined(__STDC__) || defined(__cplusplus)
 #   define GC_PROTO(args) args
     typedef void * GC_PTR;
+#   define GC_CONST const
 # else
 #   define GC_PROTO(args) ()
     typedef char * GC_PTR;
+#   define GC_CONST
 #  endif
 
 # ifdef __cplusplus
@@ -128,9 +139,26 @@ GC_API int GC_dont_expand;
 			/* Dont expand heap unless explicitly requested */
 			/* or forced to.				*/
 
+GC_API int GC_use_entire_heap;
+		/* Causes the nonincremental collector to use the	*/
+		/* entire heap before collecting.  This was the only 	*/
+		/* option for GC versions < 5.0.  This sometimes	*/
+		/* results in more large block fragmentation, since	*/
+		/* very larg blocks will tend to get broken up		*/
+		/* during each GC cycle.  It is likely to result in a	*/
+		/* larger working set, but lower collection		*/
+		/* frequencies, and hence fewer instructions executed	*/
+		/* in the collector.					*/
+
 GC_API int GC_full_freq;    /* Number of partial collections between	*/
 			    /* full collections.  Matters only if	*/
 			    /* GC_incremental is set.			*/
+			    /* Full collections are also triggered if	*/
+			    /* the collector detects a substantial	*/
+			    /* increase in the number of in-use heap	*/
+			    /* blocks.  Values in the tens are now	*/
+			    /* perfectly reasonable, unlike for		*/
+			    /* earlier GC versions.			*/
 			
 GC_API GC_word GC_non_gc_bytes;
 			/* Bytes not considered candidates for collection. */
@@ -297,6 +325,9 @@ GC_API int GC_try_to_collect GC_PROTO((GC_stop_func stop_func));
 /* Includes some pages that were allocated but never written.		*/
 GC_API size_t GC_get_heap_size GC_PROTO((void));
 
+/* Return a lower bound on the number of free bytes in the heap.	*/
+GC_API size_t GC_get_free_bytes GC_PROTO((void));
+
 /* Return the number of bytes allocated since the last collection.	*/
 GC_API size_t GC_get_bytes_since_gc GC_PROTO((void));
 
@@ -341,10 +372,10 @@ GC_API GC_PTR GC_malloc_atomic_ignore_off_page GC_PROTO((size_t lb));
 
 #ifdef GC_ADD_CALLER
 #  define GC_EXTRAS GC_RETURN_ADDR, __FILE__, __LINE__
-#  define GC_EXTRA_PARAMS GC_word ra, char * descr_string, int descr_int
+#  define GC_EXTRA_PARAMS GC_word ra, GC_CONST char * s, int i
 #else
 #  define GC_EXTRAS __FILE__, __LINE__
-#  define GC_EXTRA_PARAMS char * descr_string, int descr_int
+#  define GC_EXTRA_PARAMS GC_CONST char * s, int i
 #endif
 
 /* Debugging (annotated) allocation.  GC_gcollect will check 		*/
@@ -375,6 +406,8 @@ GC_API void GC_debug_end_stubborn_change GC_PROTO((GC_PTR));
 	GC_debug_register_finalizer(p, f, d, of, od)
 #   define GC_REGISTER_FINALIZER_IGNORE_SELF(p, f, d, of, od) \
 	GC_debug_register_finalizer_ignore_self(p, f, d, of, od)
+#   define GC_REGISTER_FINALIZER_NO_ORDER(p, f, d, of, od) \
+	GC_debug_register_finalizer_no_order(p, f, d, of, od)
 #   define GC_MALLOC_STUBBORN(sz) GC_debug_malloc_stubborn(sz, GC_EXTRAS);
 #   define GC_CHANGE_STUBBORN(p) GC_debug_change_stubborn(p)
 #   define GC_END_STUBBORN_CHANGE(p) GC_debug_end_stubborn_change(p)
@@ -391,6 +424,8 @@ GC_API void GC_debug_end_stubborn_change GC_PROTO((GC_PTR));
 	GC_register_finalizer(p, f, d, of, od)
 #   define GC_REGISTER_FINALIZER_IGNORE_SELF(p, f, d, of, od) \
 	GC_register_finalizer_ignore_self(p, f, d, of, od)
+#   define GC_REGISTER_FINALIZER_NO_ORDER(p, f, d, of, od) \
+	GC_register_finalizer_no_order(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)
@@ -469,6 +504,16 @@ GC_API void GC_debug_register_finalizer_ignore_self
 	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
 		  GC_finalization_proc *ofn, GC_PTR *ocd));
 
+/* Another version of the above.  It ignores all cycles.        */
+/* It should probably only be used by Java implementations.      */
+GC_API void GC_register_finalizer_no_order
+	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
+		  GC_finalization_proc *ofn, GC_PTR *ocd));
+GC_API void GC_debug_register_finalizer_no_order
+	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
+		  GC_finalization_proc *ofn, GC_PTR *ocd));
+
+
 /* 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	*/
@@ -525,6 +570,9 @@ GC_API int GC_unregister_disappearing_link GC_PROTO((GC_PTR * /* link */));
 GC_API GC_PTR GC_make_closure GC_PROTO((GC_finalization_proc fn, GC_PTR data));
 GC_API void GC_debug_invoke_finalizer GC_PROTO((GC_PTR obj, GC_PTR data));
 
+/* Returns !=0  if GC_invoke_finalizers has something to do. 		*/
+GC_API int GC_should_invoke_finalizers GC_PROTO((void));
+
 GC_API int GC_invoke_finalizers GC_PROTO((void));
 	/* Run finalizers for all objects that are ready to	*/
 	/* be finalized.  Return the number of finalizers	*/
@@ -688,7 +736,8 @@ GC_API void (*GC_is_visible_print_proc)
 # endif /* SOLARIS_THREADS */
 
 
-#if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#if !defined(USE_LD_WRAP) && \
+    (defined(IRIX_THREADS) || defined(LINUX_THREADS) || defined(HPUX_THREADS))
 /* We treat these similarly. */
 # include <pthread.h>
 # include <signal.h>
@@ -702,12 +751,13 @@ GC_API void (*GC_is_visible_print_proc)
 # define pthread_create GC_pthread_create
 # define pthread_sigmask GC_pthread_sigmask
 # define pthread_join GC_pthread_join
+# define dlopen GC_dlopen
 
-#endif /* IRIX_THREADS || LINUX_THREADS */
+#endif /* xxxxx_THREADS */
 
 # if defined(PCR) || defined(SOLARIS_THREADS) || defined(WIN32_THREADS) || \
 	defined(IRIX_THREADS) || defined(LINUX_THREADS) || \
-	defined(IRIX_JDK_THREADS)
+	defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
    	/* Any flavor of threads except SRC_M3.	*/
 /* This returns a list of objects, linked through their first		*/
 /* word.  Its use can greatly reduce lock contention problems, since	*/
diff --git a/gc_cpp.h b/gc_cpp.h
index ad7df5d7..3b86824c 100644
--- a/gc_cpp.h
+++ b/gc_cpp.h
@@ -16,12 +16,11 @@ the code was modified is included with the above copyright notice.
 C++ Interface to the Boehm Collector
 
     John R. Ellis and Jesse Hull 
-    Last modified on Mon Jul 24 15:43:42 PDT 1995 by ellis
 
 This interface provides access to the Boehm collector.  It provides
 basic facilities similar to those described in "Safe, Efficient
 Garbage Collection for C++", by John R. Elis and David L. Detlefs
-(ftp.parc.xerox.com:/pub/ellis/gc).
+(ftp://ftp.parc.xerox.com/pub/ellis/gc).
 
 All heap-allocated objects are either "collectable" or
 "uncollectable".  Programs must explicitly delete uncollectable
@@ -38,7 +37,7 @@ Objects derived from class "gc" are collectable.  For example:
     A* a = new A;       // a is collectable. 
 
 Collectable instances of non-class types can be allocated using the GC
-placement:
+(or UseGC) placement:
 
     typedef int A[ 10 ];
     A* a = new (GC) A;
@@ -124,6 +123,12 @@ invoked using the ANSI-conforming syntax t->~T().  If you're using
 cfront 3.0, you'll have to comment out the class gc_cleanup, which
 uses explicit invocation.
 
+5. GC name conflicts:
+
+Many other systems seem to use the identifier "GC" as an abbreviation
+for "Graphics Context".  Since version 5.0, GC placement has been replaced
+by UseGC.  GC is an alias for UseGC, unless GC_NAME_CONFLICT is defined.
+
 ****************************************************************************/
 
 #include "gc.h"
@@ -134,11 +139,15 @@ uses explicit invocation.
 
 #if ! defined( OPERATOR_NEW_ARRAY ) \
     && (__BORLANDC__ >= 0x450 || (__GNUC__ >= 2 && __GNUC_MINOR__ >= 6) \
-        || __WATCOMC__ >= 1050)
+        || __WATCOMC__ >= 1050 || _MSC_VER >= 1100)
 #   define OPERATOR_NEW_ARRAY
 #endif
 
-enum GCPlacement {GC, NoGC, PointerFreeGC};
+enum GCPlacement {UseGC,
+#ifndef GC_NAME_CONFLICT
+		  GC=UseGC,
+#endif
+                  NoGC, PointerFreeGC};
 
 class gc {public:
     inline void* operator new( size_t size );
@@ -191,6 +200,17 @@ inline void* operator new(
 
 #ifdef OPERATOR_NEW_ARRAY
 
+#ifdef _MSC_VER
+ /** This ensures that the system default operator new[] doesn't get
+  *  undefined, which is what seems to happen on VC++ 6 for some reason
+  *  if we define a multi-argument operator new[].
+  */
+ inline void *operator new[]( size_t size )
+ {
+    return ::operator new( size );
+ }
+#endif /* _MSC_VER */
+
 inline void* operator new[](
     size_t size, 
     GCPlacement gcp,
@@ -211,7 +231,7 @@ 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) 
+    if (gcp == UseGC) 
         return GC_MALLOC( size );
     else if (gcp == PointerFreeGC)
 	return GC_MALLOC_ATOMIC( size );
@@ -246,12 +266,12 @@ inline gc_cleanup::gc_cleanup() {
     GC_finalization_proc oldProc;
     void* oldData;
     void* base = GC_base( (void *) this );
-    if (0 == base) return;
-    GC_REGISTER_FINALIZER_IGNORE_SELF( 
-        base, cleanup, (void*) ((char*) this - (char*) base), 
+    if (0 != base)  {
+      GC_REGISTER_FINALIZER_IGNORE_SELF( 
+        base, (GC_finalization_proc)cleanup, (void*) ((char*) this - (char*) base), 
         &oldProc, &oldData );
-    if (0 != oldProc) {
-        GC_REGISTER_FINALIZER_IGNORE_SELF( base, oldProc, oldData, 0, 0 );}}
+      if (0 != oldProc) {
+        GC_REGISTER_FINALIZER_IGNORE_SELF( base, oldProc, oldData, 0, 0 );}}}
 
 inline void* operator new( 
     size_t size, 
@@ -261,7 +281,7 @@ inline void* operator new(
 {
     void* obj;
 
-    if (gcp == GC) {
+    if (gcp == UseGC) {
         obj = GC_MALLOC( size );
         if (cleanup != 0) 
             GC_REGISTER_FINALIZER_IGNORE_SELF( 
diff --git a/gc_hdrs.h b/gc_hdrs.h
index 60dc2ad3..6966a9a1 100644
--- a/gc_hdrs.h
+++ b/gc_hdrs.h
@@ -24,6 +24,17 @@ typedef struct hblkhdr hdr;
  * 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.
+ *
+ * This defines HDR, GET_HDR, and SET_HDR, the main macros used to
+ * retrieve and set object headers.  We also define some variants to
+ * retrieve 2 unrelated headers in interleaved fashion.  This
+ * slightly improves scheduling.
+ *
+ * Since 5.0 alpha 5, we can also take advantage of a header lookup
+ * cache.  This is a locally declared direct mapped cache, used inside
+ * the marker.  The HC_GET_HDR and HC_GET_HDR2 macros use and maintain this
+ * cache.  Assuming we get reasonable hit rates, this shaves a few
+ * memory references from each pointer validation.
  */
 
 # if CPP_WORDSZ > 32
@@ -45,6 +56,127 @@ typedef struct hblkhdr hdr;
 # define TOP_SZ (1 << LOG_TOP_SZ)
 # define BOTTOM_SZ (1 << LOG_BOTTOM_SZ)
 
+#ifndef SMALL_CONFIG
+# define USE_HDR_CACHE
+#endif
+
+/* #define COUNT_HDR_CACHE_HITS  */
+
+extern hdr * GC_invalid_header; /* header for an imaginary block 	*/
+				/* containing no objects.		*/
+
+
+/* Check whether p and corresponding hhdr point to long or invalid	*/
+/* object.  If so, advance them	to					*/
+/* beginning of	block, or set hhdr to GC_invalid_header.		*/
+#define ADVANCE(p, hhdr, source) \
+            if (IS_FORWARDING_ADDR_OR_NIL(hhdr)) { \
+              p = GC_FIND_START(p, hhdr, (word)source); \
+              if (p == 0) { \
+		hhdr = GC_invalid_header; \
+	      } else { \
+                hhdr = GC_find_header(p); \
+	      } \
+    	    }
+
+#ifdef USE_HDR_CACHE
+
+# ifdef COUNT_HDR_CACHE_HITS
+    extern word GC_hdr_cache_hits;
+    extern word GC_hdr_cache_misses;
+#   define HC_HIT() ++GC_hdr_cache_hits
+#   define HC_MISS() ++GC_hdr_cache_misses
+# else
+#   define HC_HIT()
+#   define HC_MISS()
+# endif
+
+  typedef struct hce {
+    word block_addr;  	/* right shifted by LOG_HBLKSIZE */
+    hdr * hce_hdr;
+  } hdr_cache_entry;
+
+# define HDR_CACHE_SIZE 8  /* power of 2 */
+
+# define DECLARE_HDR_CACHE \
+	hdr_cache_entry hdr_cache[HDR_CACHE_SIZE]
+
+# define INIT_HDR_CACHE BZERO(hdr_cache, sizeof(hdr_cache));
+
+# define HCE(h) hdr_cache + (((word)(h) >> LOG_HBLKSIZE) & (HDR_CACHE_SIZE-1))
+
+# define HCE_VALID_FOR(hce,h) ((hce) -> block_addr == \
+				((word)(h) >> LOG_HBLKSIZE))
+
+# define HCE_HDR(h) ((hce) -> hce_hdr)
+
+
+/* Analogous to GET_HDR, except that in the case of large objects, it	*/
+/* Returns the header for the object beginning, and updates p.		*/
+/* Returns &GC_bad_header instead of 0.  All of this saves a branch	*/
+/* in the fast path.							*/
+# define HC_GET_HDR(p, hhdr, source) \
+	{ \
+	  hdr_cache_entry * hce = HCE(p); \
+	  if (HCE_VALID_FOR(hce, p)) { \
+	    HC_HIT(); \
+	    hhdr = hce -> hce_hdr; \
+	  } else { \
+	    HC_MISS(); \
+	    GET_HDR(p, hhdr); \
+	    ADVANCE(p, hhdr, source); \
+	    hce -> block_addr = (word)(p) >> LOG_HBLKSIZE; \
+	    hce -> hce_hdr = hhdr; \
+	  } \
+	}
+
+# define HC_GET_HDR2(p1, hhdr1, source1, p2, hhdr2, source2) \
+	{ \
+	  hdr_cache_entry * hce1 = HCE(p1); \
+	  hdr_cache_entry * hce2 = HCE(p2); \
+	  if (HCE_VALID_FOR(hce1, p1)) { \
+	    HC_HIT(); \
+	    hhdr1 = hce1 -> hce_hdr; \
+	  } else { \
+	    HC_MISS(); \
+	    GET_HDR(p1, hhdr1); \
+	    ADVANCE(p1, hhdr1, source1); \
+	    hce1 -> block_addr = (word)(p1) >> LOG_HBLKSIZE; \
+	    hce1 -> hce_hdr = hhdr1; \
+	  } \
+	  if (HCE_VALID_FOR(hce2, p2)) { \
+	    HC_HIT(); \
+	    hhdr2 = hce2 -> hce_hdr; \
+	  } else { \
+	    HC_MISS(); \
+	    GET_HDR(p2, hhdr2); \
+	    ADVANCE(p2, hhdr2, source2); \
+	    hce2 -> block_addr = (word)(p2) >> LOG_HBLKSIZE; \
+	    hce2 -> hce_hdr = hhdr2; \
+	  } \
+	}
+
+#else /* !USE_HDR_CACHE */
+
+# define DECLARE_HDR_CACHE
+
+# define INIT_HDR_CACHE
+
+# define HC_GET_HDR(p, hhdr, source) \
+	{ \
+	  GET_HDR(p, hhdr); \
+	  ADVANCE(p, hhdr, source); \
+	}
+
+# define HC_GET_HDR2(p1, hhdr1, source1, p2, hhdr2, source2) \
+	{ \
+	  GET_HDR2(p1, hhdr1, p2, hhdr2); \
+	  ADVANCE(p1, hhdr1, source1); \
+	  ADVANCE(p2, hhdr2, source2); \
+	}
+
+#endif
+
 typedef struct bi {
     hdr * index[BOTTOM_SZ];
 	/*
@@ -97,6 +229,8 @@ typedef struct bi {
 #   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))
+#   define GET_HDR2(p1, hhdr1, p2, hhdr2) \
+	{ GET_HDR(p1, hhdr1); GET_HDR(p2, hhdr2); }
 # else /* hash */
 /*  Hash function for tree top level */
 #   define TL_HASH(hi) ((hi) & (TOP_SZ - 1))
@@ -123,6 +257,40 @@ typedef struct bi {
 #   define SET_HDR(p, hhdr) { register hdr ** _ha; GET_HDR_ADDR(p, _ha); \
 			      *_ha = (hhdr); }
 #   define HDR(p) GC_find_header((ptr_t)(p))
+    /* And some interleaved versions for two pointers at once.  	*/
+    /* This hopefully helps scheduling on processors like IA64.		*/
+#   define GET_BI2(p1, bottom_indx1, p2, bottom_indx2) \
+	{ \
+	    register word hi1 = \
+	        (word)(p1) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE); \
+	    register word hi2 = \
+	        (word)(p2) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE); \
+	    register bottom_index * _bi1 = GC_top_index[TL_HASH(hi1)]; \
+	    register bottom_index * _bi2 = GC_top_index[TL_HASH(hi2)]; \
+	    \
+	    while (_bi1 -> key != hi1 && _bi1 != GC_all_nils) \
+	    	_bi1 = _bi1 -> hash_link; \
+	    while (_bi2 -> key != hi2 && _bi2 != GC_all_nils) \
+	    	_bi2 = _bi2 -> hash_link; \
+	    (bottom_indx1) = _bi1; \
+	    (bottom_indx2) = _bi2; \
+	}
+#   define GET_HDR_ADDR2(p1, ha1, p2, ha2) \
+	{ \
+	    register bottom_index * bi1; \
+	    register bottom_index * bi2; \
+	    \
+	    GET_BI2(p1, bi1, p2, bi2);	\
+	    (ha1) = &(HDR_FROM_BI(bi1, p1)); \
+	    (ha2) = &(HDR_FROM_BI(bi2, p2)); \
+	}
+#   define GET_HDR2(p1, hhdr1, p2, hhdr2) \
+	{ register hdr ** _ha1;  \
+	  register hdr ** _ha2;  \
+	  GET_HDR_ADDR2(p1, _ha1, p2, _ha2); \
+	  (hhdr1) = *_ha1;  \
+	  (hhdr2) = *_ha2;  \
+	}
 # endif
 			    
 /* Is the result a forwarding address to someplace closer to the	*/
diff --git a/gc_mark.h b/gc_mark.h
index 4628323f..3a4908fb 100644
--- a/gc_mark.h
+++ b/gc_mark.h
@@ -20,6 +20,10 @@
 #ifndef GC_MARK_H
 # define GC_MARK_H
 
+# ifdef KEEP_BACK_PTRS
+#   include "dbg_mlc.h"
+# endif
+
 /* A client supplied mark procedure.  Returns new mark stack pointer.	*/
 /* Primary effect should be to push new entries on the mark stack.	*/
 /* Mark stack pointer values are passed and returned explicitly.	*/
@@ -41,8 +45,10 @@
 /* The real declarations of the following are in gc_priv.h, so that	*/
 /* we can avoid scanning the following table.				*/
 /*
-typedef struct ms_entry * (*mark_proc)(   word * addr, mark_stack_ptr,
-					  mark_stack_limit, env   );
+typedef struct ms_entry * (*mark_proc)(   word * addr,
+					  struct ms_entry *mark_stack_ptr,
+					  struct ms_entry *mark_stack_limit,
+					  word env   );
 					  
 # define LOG_MAX_MARK_PROCS 6
 # define MAX_MARK_PROCS (1 << LOG_MAX_MARK_PROCS)
@@ -51,6 +57,12 @@ extern mark_proc GC_mark_procs[MAX_MARK_PROCS];
 
 extern word GC_n_mark_procs;
 
+/* In a few cases it's necessary to assign statically known indices to	*/
+/* certain mark procs.  Thus we reserve a few for well known clients.	*/
+/* (This is necessary if mark descriptors are compiler generated.)	*/
+#define GC_RESERVED_MARK_PROCS 8
+#   define GCJ_RESERVED_MARK_PROC_INDEX 0
+
 /* 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.						*/
@@ -84,6 +96,13 @@ extern word GC_n_mark_procs;
 #define DS_PER_OBJECT 3	/* The real descriptor is at the		*/
 			/* byte displacement from the beginning of the	*/
 			/* object given by descr & ~DS_TAGS		*/
+			/* If the descriptor is negative, the real	*/
+			/* descriptor is at (*<object_start>) -		*/
+			/* (descr & ~DS_TAGS) - INDIR_PER_OBJ_BIAS	*/
+			/* The latter alternative can be used if each	*/
+			/* object contains a type descriptor in the	*/
+			/* first word.					*/
+#define INDIR_PER_OBJ_BIAS 0x10
 			
 typedef struct ms_entry {
     word * mse_start;   /* First word of object */
@@ -98,7 +117,7 @@ extern mse * GC_mark_stack_top;
 
 extern mse * GC_mark_stack;
 
-word GC_find_start();
+ptr_t GC_find_start();
 
 mse * GC_signal_mark_stack_overflow();
 
@@ -144,16 +163,60 @@ mse * GC_signal_mark_stack_overflow();
 # define PUSH_CONTENTS(current, mark_stack_top, mark_stack_limit, \
 		       source, exit_label) \
 { \
-    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, (word)source); \
-         if (current == 0) goto exit_label; \
-         hhdr = HDR(current); \
+    hdr * my_hhdr; \
+    ptr_t my_current = current; \
+ \
+    GET_HDR(my_current, my_hhdr); \
+    if (IS_FORWARDING_ADDR_OR_NIL(my_hhdr)) { \
+         my_current = GC_FIND_START(my_current, my_hhdr, (word)source); \
+         if (my_current == 0) goto exit_label; \
+         my_hhdr = GC_find_header(my_current); \
     } \
+    PUSH_CONTENTS_HDR(my_current, mark_stack_top, mark_stack_limit, \
+		  source, exit_label, my_hhdr);	\
+exit_label: ; \
+}
+
+/* As above, but use header cache for header lookup.	*/
+# define HC_PUSH_CONTENTS(current, mark_stack_top, mark_stack_limit, \
+		       source, exit_label) \
+{ \
+    hdr * my_hhdr; \
+    ptr_t my_current = current; \
+ \
+    HC_GET_HDR(my_current, my_hhdr, source); \
+    PUSH_CONTENTS_HDR(my_current, mark_stack_top, mark_stack_limit, \
+		  source, exit_label, my_hhdr);	\
+exit_label: ; \
+}
+
+/* As above, but deal with two pointers in interleaved fashion.	*/
+# define HC_PUSH_CONTENTS2(current1, current2, mark_stack_top, \
+			   mark_stack_limit, \
+		           source1, source2, exit_label1, exit_label2) \
+{ \
+    hdr * hhdr1; \
+    ptr_t my_current1 = current1; \
+    hdr * hhdr2; \
+    ptr_t my_current2 = current2; \
+ \
+    HC_GET_HDR2(my_current1, hhdr1, source1, my_current2, hhdr2, source2); \
+    PUSH_CONTENTS_HDR(my_current1, mark_stack_top, mark_stack_limit, \
+		  source1, exit_label1, hhdr1);	\
+exit_label1: ; \
+    if (0 != hhdr2) { \
+      PUSH_CONTENTS_HDR(my_current2, mark_stack_top, mark_stack_limit, \
+		  source2, exit_label2, hhdr2);	\
+    } \
+exit_label2: ; \
+}
+
+# define PUSH_CONTENTS_HDR(current, mark_stack_top, mark_stack_limit, \
+		           source, exit_label, hhdr) \
+{ \
+    int displ;  /* Displacement in block; first bytes, then words */ \
+    map_entry_type map_entry; \
+    \
     displ = HBLKDISPL(current); \
     map_entry = MAP_ENTRY((hhdr -> hb_map), displ); \
     if (map_entry == OBJ_INVALID) { \
@@ -177,10 +240,9 @@ mse * GC_signal_mark_stack_overflow();
     } \
     PUSH_OBJ(((word *)(HBLKPTR(current)) + displ), hhdr, \
     	     mark_stack_top, mark_stack_limit) \
-  exit_label: ; \
 }
 
-#ifdef PRINT_BLACK_LIST
+#if defined(PRINT_BLACK_LIST) || defined(KEEP_BACK_PTRS)
 #   define PUSH_ONE_CHECKED(p, ip, source) \
 	GC_push_one_checked(p, ip, (ptr_t)(source))
 #else
diff --git a/gc_priv.h b/gc_priv.h
index 5ce52a7a..ce685875 100644
--- a/gc_priv.h
+++ b/gc_priv.h
@@ -1,6 +1,9 @@
 /* 
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
+ *
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -41,7 +44,7 @@
 typedef GC_word word;
 typedef GC_signed_word signed_word;
 
-# ifndef CONFIG_H
+# ifndef GCCONFIG_H
 #   include "gcconfig.h"
 # endif
 
@@ -64,21 +67,22 @@ typedef char * ptr_t;	/* A generic pointer to which we can add	*/
 #       include <stddef.h>
 #   endif
 #   define VOLATILE volatile
-#   define CONST const
 #else
 #   ifdef MSWIN32
 #   	include <stdlib.h>
 #   endif
 #   define VOLATILE
-#   define CONST
 #endif
 
+#define CONST GC_CONST
+
 #if 0 /* was once defined for AMIGA */
 #   define GC_FAR __far
 #else
 #   define GC_FAR
 #endif
 
+
 /*********************************/
 /*                               */
 /* Definitions for conservative  */
@@ -170,15 +174,6 @@ typedef char * ptr_t;	/* A generic pointer to which we can add	*/
 		    /* May save significant amounts of space for obj_map  */
 		    /* entries.						  */
 
-#ifndef OLD_BLOCK_ALLOC
-   /* Macros controlling large block allocation strategy.	*/
-#  define EXACT_FIRST  	/* Make a complete pass through the large object */
-			/* free list before splitting a block		 */
-#  define PRESERVE_LAST /* Do not divide last allocated heap segment	 */
-			/* unless we would otherwise need to expand the	 */
-			/* heap.					 */
-#endif
-
 /* ALIGN_DOUBLE requires MERGE_SIZES at present. */
 # if defined(ALIGN_DOUBLE) && !defined(MERGE_SIZES)
 #   define MERGE_SIZES
@@ -278,6 +273,13 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #   define MS_TIME_DIFF(a,b) ((double) (a.tv_sec - b.tv_sec) * 1000.0 \
                                + (double) (a.tv_usec - b.tv_usec) / 1000.0)
 #else /* !BSD_TIME */
+# ifdef MSWIN32
+#   include <windows.h>
+#   include <winbase.h>
+#   define CLOCK_TYPE DWORD
+#   define GET_TIME(x) x = GetTickCount()
+#   define MS_TIME_DIFF(a,b) ((long)((a)-(b)))
+# else /* !MSWIN32, !BSD_TIME */
 #   include <time.h>
 #   if !defined(__STDC__) && defined(SPARC) && defined(SUNOS4)
       clock_t clock();	/* Not in time.h, where it belongs	*/
@@ -303,6 +305,7 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #   define GET_TIME(x) x = clock()
 #   define MS_TIME_DIFF(a,b) ((unsigned long) \
 		(1000.0*(double)((a)-(b))/(double)CLOCKS_PER_SEC))
+# endif /* !MSWIN32 */
 #endif /* !BSD_TIME */
 
 /* We use bzero and bcopy internally.  They may not be available.	*/
@@ -434,9 +437,12 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #    define LOCK() mutex_lock(&GC_allocate_ml);
 #    define UNLOCK() mutex_unlock(&GC_allocate_ml);
 #  endif
-#  ifdef LINUX_THREADS
+#  if defined(LINUX_THREADS) 
+#   if defined(I386)|| defined(POWERPC) || defined(ALPHA) || defined(IA64) \
+    || defined(M68K)
 #    include <pthread.h>
-#    ifdef __i386__
+#    define USE_SPIN_LOCK
+#    if defined(I386)
        inline static int GC_test_and_set(volatile unsigned int *addr) {
 	  int oldval;
 	  /* Note: the "xchg" instruction does not need a "lock" prefix */
@@ -445,15 +451,107 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 		: "0"(1), "m"(*(addr)));
 	  return oldval;
        }
-#    else
-       -- > Need implementation of GC_test_and_set()
 #    endif
-#    define GC_clear(addr) (*(addr) = 0)
+#    if defined(IA64)
+       inline static int GC_test_and_set(volatile unsigned int *addr) {
+	  int oldval;
+	  __asm__ __volatile__("xchg4 %0=%1,%2"
+		: "=r"(oldval), "=m"(*addr)
+		: "r"(1), "1"(*addr));
+	  return oldval;
+       }
+       inline static void GC_clear(volatile unsigned int *addr) {
+	 __asm__ __volatile__("st4.rel %0=r0" : "=m" (*addr));
+       }
+#      define GC_CLEAR_DEFINED
+#    endif
+#    ifdef M68K
+       /* Contributed by Tony Mantler.  I'm not sure how well it was	*/
+       /* tested.							*/
+       inline static int GC_test_and_set(volatile unsigned int *addr) {
+          char oldval; /* this must be no longer than 8 bits */
+
+          /* The return value is semi-phony. */
+          /* 'tas' sets bit 7 while the return */
+          /* value pretends bit 0 was set */
+          __asm__ __volatile__(
+                 "tas %1@; sne %0; negb %0"
+                 : "=d" (oldval)
+                 : "a" (addr));
+          return oldval;
+       }
+#    endif
+#    if defined(POWERPC)
+      inline static int GC_test_and_set(volatile unsigned int *addr) {
+        int oldval;
+        int temp = 1; // locked value
+
+        __asm__ __volatile__(
+               "1:\tlwarx %0,0,%3\n"   // load and reserve
+               "\tcmpwi %0, 0\n"       // if load is
+               "\tbne 2f\n"            //   non-zero, return already set
+               "\tstwcx. %2,0,%1\n"    // else store conditional
+               "\tbne- 1b\n"           // retry if lost reservation
+               "2:\t\n"                // oldval is zero if we set
+              : "=&r"(oldval), "=p"(addr)
+              : "r"(temp), "1"(addr)
+              : "memory");
+        return (int)oldval;
+      }
+      inline static void GC_clear(volatile unsigned int *addr) {
+	 __asm__ __volatile__("eieio");
+         *(addr) = 0;
+      }
+#     define GC_CLEAR_DEFINED
+#    endif
+#    ifdef ALPHA
+      inline static int GC_test_and_set(volatile unsigned int * addr)
+      {
+        unsigned long oldvalue;
+        unsigned long temp;
+
+        __asm__ __volatile__(
+                             "1:     ldl_l %0,%1\n"
+                             "       and %0,%3,%2\n"
+                             "       bne %2,2f\n"
+                             "       xor %0,%3,%0\n"
+                             "       stl_c %0,%1\n"
+                             "       beq %0,3f\n"
+                             "       mb\n"
+                             "2:\n"
+                             ".section .text2,\"ax\"\n"
+                             "3:     br 1b\n"
+                             ".previous"
+                             :"=&r" (temp), "=m" (*addr), "=&r" (oldvalue)
+                             :"Ir" (1), "m" (*addr));
+
+        return oldvalue;
+      }
+      /* Should probably also define GC_clear, since it needs	*/
+      /* a memory barrier ??					*/
+#    endif /* ALPHA */
+#    ifdef ARM32
+      inline static int GC_test_and_set(volatile unsigned int *addr) {
+        int oldval;
+        /* SWP on ARM is very similar to XCHG on x86.  Doesn't lock the
+         * bus because there are no SMP ARM machines.  If/when there are,
+         * this code will likely need to be updated. */
+        /* See linuxthreads/sysdeps/arm/pt-machine.h in glibc-2.1 */
+        __asm__ __volatile__("swp %0, %1, [%2]"
+      			     : "=r"(oldval)
+      			     : "r"(1), "r"(addr));
+        return oldval;
+      }
+#    endif
+#    ifndef GC_CLEAR_DEFINED
+       inline static void GC_clear(volatile unsigned int *addr) {
+	  /* Try to discourage gcc from moving anything past this. */
+	  __asm__ __volatile__(" ");
+          *(addr) = 0;
+       }
+#    endif
 
      extern volatile unsigned int GC_allocate_lock;
-	/* This is not a mutex because mutexes that obey the (optional)     */
-	/* POSIX scheduling rules are subject to convoys in high contention */
-	/* applications.  This is basically a spin lock.		    */
      extern pthread_t GC_lock_holder;
      extern void GC_lock(void);
 	/* Allocation lock holder.  Only set if acquired by client through */
@@ -462,31 +560,48 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #    define NO_THREAD (pthread_t)(-1)
 #    define UNSET_LOCK_HOLDER() GC_lock_holder = NO_THREAD
 #    define I_HOLD_LOCK() (pthread_equal(GC_lock_holder, pthread_self()))
-#    ifdef UNDEFINED
-#    	define LOCK() pthread_mutex_lock(&GC_allocate_ml)
-#    	define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
-#    else
-#	define LOCK() \
+#    define LOCK() \
 		{ if (GC_test_and_set(&GC_allocate_lock)) GC_lock(); }
-#	define UNLOCK() \
+#    define UNLOCK() \
 		GC_clear(&GC_allocate_lock)
-#    endif
-     extern GC_bool GC_collecting;
+     extern VOLATILE GC_bool GC_collecting;
 #    define ENTER_GC() \
 		{ \
 		    GC_collecting = 1; \
 		}
 #    define EXIT_GC() GC_collecting = 0;
+#   else /* LINUX_THREADS on hardware for which we don't know how	*/
+	 /* to do test and set.						*/
+#    include <pthread.h>
+     extern pthread_mutex_t GC_allocate_ml;
+#    define LOCK() pthread_mutex_lock(&GC_allocate_ml)
+#    define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
+#   endif
 #  endif /* LINUX_THREADS */
-#  if defined(IRIX_THREADS) || defined(IRIX_JDK_THREADS)
+#  if defined(HPUX_THREADS)
 #    include <pthread.h>
-#    include <mutex.h>
+     extern pthread_mutex_t GC_allocate_ml;
+#    define LOCK() pthread_mutex_lock(&GC_allocate_ml)
+#    define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
+#  endif
+#  if defined(IRIX_THREADS) || defined(IRIX_JDK_THREADS) 
+     /* This may also eventually be appropriate for HPUX_THREADS */
+#    include <pthread.h>
+#    ifndef HPUX_THREADS
+	/* This probably should never be included, but I can't test	*/
+	/* on Irix anymore.						*/
+#       include <mutex.h>
+#    endif
 
-#    if __mips < 3 || !(defined (_ABIN32) || defined(_ABI64)) \
+#    ifndef HPUX_THREADS
+#      if __mips < 3 || !(defined (_ABIN32) || defined(_ABI64)) \
 	|| !defined(_COMPILER_VERSION) || _COMPILER_VERSION < 700
 #        define GC_test_and_set(addr, v) test_and_set(addr,v)
-#    else
+#      else
 #	 define GC_test_and_set(addr, v) __test_and_set(addr,v)
+#      endif
+#    else
+       /* I couldn't find a way to do this inline on HP/UX	*/
 #    endif
      extern unsigned long GC_allocate_lock;
 	/* This is not a mutex because mutexes that obey the (optional) 	*/
@@ -500,15 +615,17 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #    define NO_THREAD (pthread_t)(-1)
 #    define UNSET_LOCK_HOLDER() GC_lock_holder = NO_THREAD
 #    define I_HOLD_LOCK() (pthread_equal(GC_lock_holder, pthread_self()))
-#    ifdef UNDEFINED
-#    	define LOCK() pthread_mutex_lock(&GC_allocate_ml)
-#    	define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
+#    ifdef HPUX_THREADS
+#      define LOCK() { if (!GC_test_and_clear(&GC_allocate_lock)) GC_lock(); }
+       /* The following is INCORRECT, since the memory model is too weak. */
+#      define UNLOCK() { GC_noop1(&GC_allocate_lock); \
+			*(volatile unsigned long *)(&GC_allocate_lock) = 1; }
 #    else
-#	define LOCK() { if (GC_test_and_set(&GC_allocate_lock, 1)) GC_lock(); }
-#       if __mips >= 3 && (defined (_ABIN32) || defined(_ABI64)) \
+#      define LOCK() { if (GC_test_and_set(&GC_allocate_lock, 1)) GC_lock(); }
+#      if __mips >= 3 && (defined (_ABIN32) || defined(_ABI64)) \
 	   && defined(_COMPILER_VERSION) && _COMPILER_VERSION >= 700
 #	    define UNLOCK() __lock_release(&GC_allocate_lock)
-#	else
+#      else
 	    /* The function call in the following should prevent the	*/
 	    /* compiler from moving assignments to below the UNLOCK.	*/
 	    /* This is probably not necessary for ucode or gcc 2.8.	*/
@@ -516,9 +633,9 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 	    /* versions.						*/
 #           define UNLOCK() { GC_noop1(&GC_allocate_lock); \
 			*(volatile unsigned long *)(&GC_allocate_lock) = 0; }
-#	endif
+#      endif
 #    endif
-     extern GC_bool GC_collecting;
+     extern VOLATILE GC_bool GC_collecting;
 #    define ENTER_GC() \
 		{ \
 		    GC_collecting = 1; \
@@ -607,7 +724,7 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 # else
 #   if defined(SOLARIS_THREADS) || defined(WIN32_THREADS) \
 	|| defined(IRIX_THREADS) || defined(LINUX_THREADS) \
-	|| defined(IRIX_JDK_THREADS)
+	|| defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
       void GC_stop_world();
       void GC_start_world();
 #     define STOP_WORLD() GC_stop_world()
@@ -857,6 +974,9 @@ struct hblkhdr {
 			    /* object starting at the ith word (header      */
 			    /* INCLUDED) in the heap block.                 */
 			    /* The lsb of word 0 is numbered 0.		    */
+			    /* Unused bits are invalid, and are 	    */
+			    /* occasionally set, e.g for uncollectable	    */
+			    /* objects.					    */
 };
 
 /*  heap block body */
@@ -891,8 +1011,10 @@ struct hblk {
 /* The type of mark procedures.  This really belongs in gc_mark.h.	*/
 /* But we put it here, so that we can avoid scanning the mark proc	*/
 /* table.								*/
-typedef struct ms_entry * (*mark_proc)(/* word * addr, mark_stack_ptr,
-					  mark_stack_limit, env */);
+typedef struct ms_entry * (*mark_proc)(/* word * addr,
+					  struct ms_entry *mark_stack_ptr,
+					  struct ms_entry *mark_stack_limit,
+					  word env */);
 # define LOG_MAX_MARK_PROCS 6
 # define MAX_MARK_PROCS (1 << LOG_MAX_MARK_PROCS)
 
@@ -969,6 +1091,7 @@ struct roots {
 struct _GC_arrays {
   word _heapsize;
   word _max_heapsize;
+  word _requested_heapsize;	/* Heap size due to explicit expansion */
   ptr_t _last_heap_addr;
   ptr_t _prev_heap_addr;
   word _large_free_bytes;
@@ -993,6 +1116,10 @@ struct _GC_arrays {
   word _mem_freed;
   	/* Number of explicitly deallocated words of memory	*/
   	/* since last collection.				*/
+  ptr_t _scratch_end_ptr;
+  ptr_t _scratch_last_end_ptr;
+	/* Used by headers.c, and can easily appear to point to	*/
+	/* heap.						*/
   mark_proc _mark_procs[MAX_MARK_PROCS];
   	/* Table of user-defined mark procedures.  There is	*/
 	/* a small number of these, which can be referenced	*/
@@ -1157,9 +1284,12 @@ GC_API GC_FAR struct _GC_arrays GC_arrays;
 # define GC_words_finalized GC_arrays._words_finalized
 # define GC_non_gc_bytes_at_gc GC_arrays._non_gc_bytes_at_gc
 # define GC_mem_freed GC_arrays._mem_freed
+# define GC_scratch_end_ptr GC_arrays._scratch_end_ptr
+# define GC_scratch_last_end_ptr GC_arrays._scratch_last_end_ptr
 # define GC_mark_procs GC_arrays._mark_procs
 # define GC_heapsize GC_arrays._heapsize
 # define GC_max_heapsize GC_arrays._max_heapsize
+# define GC_requested_heapsize GC_arrays._requested_heapsize
 # define GC_words_allocd_before_gc GC_arrays._words_allocd_before_gc
 # define GC_heap_sects GC_arrays._heap_sects
 # define GC_last_stack GC_arrays._last_stack
@@ -1194,6 +1324,8 @@ GC_API GC_FAR struct _GC_arrays GC_arrays;
 # define beginGC_arrays ((ptr_t)(&GC_arrays))
 # define endGC_arrays (((ptr_t)(&GC_arrays)) + (sizeof GC_arrays))
 
+#define USED_HEAP_SIZE (GC_heapsize - GC_large_free_bytes)
+
 /* Object kinds: */
 # define MAXOBJKINDS 16
 
@@ -1273,7 +1405,7 @@ extern GC_bool GC_objects_are_marked;	/* There are marked objects in  */
   extern GC_bool GC_incremental;
 			/* Using incremental/generational collection. */
 #else
-# define GC_incremental TRUE
+# define GC_incremental FALSE
 			/* Hopefully allow optimizer to remove some code. */
 #endif
 
@@ -1326,10 +1458,7 @@ extern ptr_t GC_greatest_plausible_heap_addr;
 ptr_t GC_approx_sp();
 
 GC_bool GC_should_collect();
-#ifdef PRESERVE_LAST
-    GC_bool GC_in_last_heap_sect(/* ptr_t */);
-	/* In last added heap section?  If so, avoid breaking up.	*/
-#endif
+
 void GC_apply_to_all_blocks(/*fn, client_data*/);
 			/* Invoke fn(hbp, client_data) for each 	*/
 			/* allocated heap block.			*/
@@ -1411,8 +1540,14 @@ extern void (*GC_start_call_back)(/* void */);
 			/* lock held.					*/
 			/* 0 by default.				*/
 void GC_push_regs();	/* Push register contents onto mark stack.	*/
+			/* If NURSERY is defined, the default push	*/
+			/* action can be overridden with GC_push_proc	*/
 void GC_remark();	/* Mark from all marked objects.  Used	*/
 		 	/* only if we had to drop something.	*/
+
+# ifdef NURSERY
+    extern void (*GC_push_proc)(ptr_t);
+# endif
 # if defined(MSWIN32)
   void __cdecl GC_push_one();
 # else
@@ -1600,9 +1735,10 @@ ptr_t GC_allocobj(/* sz_inn_words, kind */);
 				/* head.				*/
 
 void GC_init_headers();
-GC_bool GC_install_header(/*h*/);
+struct hblkhdr * GC_install_header(/*h*/);
 				/* Install a header for block h.	*/
-				/* Return FALSE on failure.		*/
+				/* Return 0 on failure, or the header	*/
+				/* otherwise.				*/
 GC_bool GC_install_counts(/*h, sz*/);
 				/* Set up forwarding counts for block	*/
 				/* h of size sz.			*/
diff --git a/gc_typed.h b/gc_typed.h
index e4a6b947..2e0598f2 100644
--- a/gc_typed.h
+++ b/gc_typed.h
@@ -61,6 +61,7 @@ GC_API GC_PTR GC_malloc_explicitly_typed
 			GC_PROTO((size_t size_in_bytes, GC_descr d));
 		/* Allocate an object whose layout is described by d.	*/
 		/* The resulting object MAY NOT BE PASSED TO REALLOC.	*/
+		/* The returned object is cleared.			*/
 
 GC_API GC_PTR GC_malloc_explicitly_typed_ignore_off_page
                         GC_PROTO((size_t size_in_bytes, GC_descr d));
@@ -75,6 +76,7 @@ GC_API GC_PTR GC_calloc_explicitly_typed
   	/* 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.				*/
+	/* Returned object is cleared.				*/
 
 #ifdef GC_DEBUG
 #   define GC_MALLOC_EXPLICTLY_TYPED(bytes, d) GC_MALLOC(bytes)
diff --git a/gcconfig.h b/gcconfig.h
index c9017d37..221983ca 100644
--- a/gcconfig.h
+++ b/gcconfig.h
@@ -13,9 +13,9 @@
  * modified is included with the above copyright notice.
  */
  
-#ifndef CONFIG_H
+#ifndef GCCONFIG_H
 
-# define CONFIG_H
+# define GCCONFIG_H
 
 /* Machine dependent parameters.  Some tuning parameters can be found	*/
 /* near the top of gc_private.h.					*/
@@ -53,6 +53,11 @@
 #    define NETBSD
 #    define mach_type_known
 # endif
+# if defined(__NetBSD__) && defined(arm32)
+#    define ARM32
+#    define NETBSD
+#    define mach_type_known
+# endif
 # if defined(vax)
 #    define VAX
 #    ifdef ultrix
@@ -64,15 +69,18 @@
 # endif
 # if defined(mips) || defined(__mips)
 #    define MIPS
-#    if defined(ultrix) || defined(__ultrix) || defined(__NetBSD__)
-#	define ULTRIX
-#    else
-#	if defined(_SYSTYPE_SVR4) || defined(SYSTYPE_SVR4) || defined(__SYSTYPE_SVR4__)
-#	  define IRIX5   /* or IRIX 6.X */
-#	else
-#	  define RISCOS  /* or IRIX 4.X */
-#	endif
-#    endif
+#    if !defined(LINUX)
+#      if defined(ultrix) || defined(__ultrix) || defined(__NetBSD__)
+#	 define ULTRIX
+#      else
+#	 if defined(_SYSTYPE_SVR4) || defined(SYSTYPE_SVR4) \
+	    || defined(__SYSTYPE_SVR4__)
+#	   define IRIX5   /* or IRIX 6.X */
+#	 else
+#	   define RISCOS  /* or IRIX 4.X */
+#	 endif
+#      endif
+#    endif /* !LINUX */
 #    define mach_type_known
 # endif
 # if defined(sequent) && defined(i386)
@@ -130,15 +138,22 @@
 #   define SYSV
 #   define mach_type_known
 # endif
-# if defined(_PA_RISC1_0) || defined(_PA_RISC1_1) \
+# if defined(_PA_RISC1_0) || defined(_PA_RISC1_1) || defined(_PA_RISC2_0) \
      || defined(hppa) || defined(__hppa__)
 #   define HP_PA
+#   ifndef LINUX
+#     define HPUX
+#   endif
 #   define mach_type_known
 # endif
 # if defined(LINUX) && (defined(i386) || defined(__i386__))
 #    define I386
 #    define mach_type_known
 # endif
+# if defined(LINUX) && (defined(__ia64__) || defined(__ia64))
+#    define IA64
+#    define mach_type_known
+# endif
 # if defined(LINUX) && defined(powerpc)
 #    define POWERPC
 #    define mach_type_known
@@ -147,10 +162,14 @@
 #    define M68K
 #    define mach_type_known
 # endif
-# if defined(LINUX) && defined(sparc)
+# if defined(LINUX) && (defined(sparc) || defined(__sparc__))
 #    define SPARC
 #    define mach_type_known
 # endif
+# if defined(LINUX) && defined(arm)
+#    define ARM32
+#    define mach_type_known
+# endif
 # if defined(__alpha) || defined(__alpha__)
 #   define ALPHA
 #   if !defined(LINUX)
@@ -243,6 +262,11 @@
 #   define CYGWIN32
 #   define mach_type_known
 # endif
+# if defined(__MINGW32__)
+#   define I386
+#   define MSWIN32
+#   define mach_type_known
+# endif
 # if defined(__BORLANDC__)
 #   define I386
 #   define MSWIN32
@@ -253,6 +277,10 @@
 #   define UTS4
 #   define mach_type_known
 # endif
+# if defined(__pj__)
+#   define PJ
+#   define mach_type_known
+# endif
 /* Ivan Demakov */
 # if defined(__WATCOMC__) && defined(__386__)
 #   define I386
@@ -307,6 +335,9 @@
 		    /* 		        (CX_UX and DGUX)		*/
 		    /* 		   S370	      ==> 370-like machine	*/
 		    /* 			running Amdahl UTS4		*/
+		    /* 		   ARM32      ==> Intel StrongARM	*/
+		    /* 		   IA64	      ==> Intel IA64		*/
+		    /*				  (e.g. Itanium)	*/
 
 
 /*
@@ -355,6 +386,11 @@
  *		to the nearest plausible page boundary, and use that instead
  *		of STACKBOTTOM.
  *
+ * Gustavo Rodriguez-Rivera points out that on most (all?) Unix machines,
+ * the value of environ is a pointer that can serve as STACKBOTTOM.
+ * I expect that HEURISTIC2 can be replaced by this approach, which
+ * interferes far less with debugging. 
+ *
  * 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:
@@ -392,6 +428,15 @@
  *
  * An architecture may define DYNAMIC_LOADING if dynamic_load.c
  * defined GC_register_dynamic_libraries() for the architecture.
+ *
+ * An architecture may define PREFETCH(x) to preload the cache with *x.
+ * This defaults to a no-op.
+ *
+ * PREFETCH_FOR_WRITE(x) is used if *x is about to be written.
+ *
+ * An architecture may also define CLEAR_DOUBLE(x) to be a fast way to
+ * clear the two words at GC_malloc-aligned address x.  By default,
+ * word stores of 0 are used instead.
  */
 
 
@@ -513,10 +558,11 @@
 				/* This was 2, but that didn't sound right. */
 #     define OS_TYPE "LINUX"
 #     define HEURISTIC1
+#     define DYNAMIC_LOADING
 #     undef STACK_GRAN
 #     define STACK_GRAN 0x10000000
 	/* Stack usually starts at 0x80000000 */
-#     define DATASTART GC_data_start
+#     define LINUX_DATA_START
       extern int _end;
 #     define DATAEND (&_end)
 #   endif
@@ -615,8 +661,8 @@
 #   ifdef LINUX
 #     define OS_TYPE "LINUX"
 #     ifdef __ELF__
-#         define DATASTART GC_data_start
-#         define DYNAMIC_LOADING
+#       define LINUX_DATA_START
+#       define DYNAMIC_LOADING
 #     else
           Linux Sparc non elf ?
 #     endif
@@ -684,13 +730,16 @@
 #   endif
 #   ifdef LINUX
 #	define OS_TYPE "LINUX"
-#       define HEURISTIC1
-#       undef STACK_GRAN
-#       define STACK_GRAN 0x10000000
-	/* STACKBOTTOM is usually 0xc0000000, but this changes with	*/
-	/* different kernel configurations.  In particular, systems	*/
-	/* with 2GB physical memory will usually move the user		*/
-	/* address space limit, and hence initial SP to 0x80000000.	*/
+#       define LINUX_STACKBOTTOM
+#	if 0
+#	  define HEURISTIC1
+#         undef STACK_GRAN
+#         define STACK_GRAN 0x10000000
+	  /* STACKBOTTOM is usually 0xc0000000, but this changes with	*/
+	  /* different kernel configurations.  In particular, systems	*/
+	  /* with 2GB physical memory will usually move the user	*/
+	  /* address space limit, and hence initial SP to 0x80000000.	*/
+#       endif
 #       if !defined(LINUX_THREADS) || !defined(REDIRECT_MALLOC)
 #	    define MPROTECT_VDB
 #	else
@@ -706,8 +755,7 @@
 #	     endif
 #	     include <features.h>
 #	     if defined(__GLIBC__) && __GLIBC__ >= 2
-		 extern int __data_start;
-#		 define DATASTART ((ptr_t)(&__data_start))
+#		 define LINUX_DATA_START
 #	     else
      	         extern char **__environ;
 #                define DATASTART ((ptr_t)(&__environ))
@@ -726,6 +774,26 @@
 	     extern int etext;
 #            define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
 #       endif
+#	ifdef USE_I686_PREFETCH
+#	  define PREFETCH(x) \
+	    __asm__ __volatile__ ("	prefetchnta	%0": : "m"(*(char *)(x)))
+	    /* Empirically prefetcht0 is much more effective at reducing	*/
+	    /* cache miss stalls for the targetted load instructions.  But it	*/
+	    /* seems to interfere enough with other cache traffic that the net	*/
+	    /* result is worse than prefetchnta.				*/
+#         if 0 
+	    /* Using prefetches for write seems to have a slight negative	*/
+	    /* impact on performance, at least for a PIII/500.			*/
+#	    define PREFETCH_FOR_WRITE(x) \
+	      __asm__ __volatile__ ("	prefetcht0	%0": : "m"(*(char *)(x)))
+#	  endif
+#	endif
+#	ifdef USE_3DNOW_PREFETCH
+#	  define PREFETCH(x) \
+	    __asm__ __volatile__ ("	prefetch	%0": : "m"(*(char *)(x)))
+#	  define PREFETCH_FOR_WRITE(x) 
+	    __asm__ __volatile__ ("	prefetchw	%0": : "m"(*(char *)(x)))
+#	endif
 #   endif
 #   ifdef CYGWIN32
 #       define OS_TYPE "CYGWIN32"
@@ -836,35 +904,48 @@
 
 # ifdef MIPS
 #   define MACH_TYPE "MIPS"
-#   ifndef IRIX5
-#     define DATASTART (ptr_t)0x10000000
-			      /* Could probably be slightly higher since */
-			      /* startup code allocates lots of stuff.   */
-#   else
-      extern int _fdata;
-#     define DATASTART ((ptr_t)(&_fdata))
-#     ifdef USE_MMAP
-#         define HEAP_START (ptr_t)0x30000000
-#     else
-#	  define HEAP_START DATASTART
-#     endif
-			      /* Lowest plausible heap address.		*/
-			      /* In the MMAP case, we map there.	*/
-			      /* In either case it is used to identify	*/
-			      /* heap sections so they're not 		*/
-			      /* considered as roots.			*/
-#   endif /* IRIX5 */
-#   define HEURISTIC2
 /* #   define STACKBOTTOM ((ptr_t)0x7fff8000)  sometimes also works.  */
+#   ifdef LINUX
+      /* This was developed for a linuxce style platform.  Probably	*/
+      /* needs to be tweaked for workstation class machines.		*/
+#     define OS_TYPE "LINUX"
+      extern int __data_start;
+#     define DATASTART ((ptr_t)(&__data_start))
+#     define ALIGNMENT 4
+#     define USE_GENERIC_PUSH_REGS 1
+#     define STACKBOTTOM 0x80000000
+	/* In many cases, this should probably use LINUX_STACKBOTTOM 	*/
+	/* instead. But some kernel versions seem to give the wrong	*/
+	/* value from /proc.						*/
+#   endif /* Linux */
 #   ifdef ULTRIX
+#	define HEURISTIC2
+#       define DATASTART (ptr_t)0x10000000
+			      /* Could probably be slightly higher since */
+			      /* startup code allocates lots of stuff.   */
 #	define OS_TYPE "ULTRIX"
 #       define ALIGNMENT 4
 #   endif
 #   ifdef RISCOS
+#	define HEURISTIC2
+#       define DATASTART (ptr_t)0x10000000
 #	define OS_TYPE "RISCOS"
 #   	define ALIGNMENT 4  /* Required by hardware */
 #   endif
 #   ifdef IRIX5
+#	define HEURISTIC2
+        extern int _fdata;
+#       define DATASTART ((ptr_t)(&_fdata))
+#       ifdef USE_MMAP
+#         define HEAP_START (ptr_t)0x30000000
+#       else
+#	  define HEAP_START DATASTART
+#       endif
+			      /* Lowest plausible heap address.		*/
+			      /* In the MMAP case, we map there.	*/
+			      /* In either case it is used to identify	*/
+			      /* heap sections so they're not 		*/
+			      /* considered as roots.			*/
 #	define OS_TYPE "IRIX5"
 #       define MPROTECT_VDB
 #       ifdef _MIPS_SZPTR
@@ -892,9 +973,17 @@
 # endif
 
 # ifdef HP_PA
+    /* OS is assumed to be HP/UX	*/
 #   define MACH_TYPE "HP_PA"
-#   define ALIGNMENT 4
-#   define ALIGN_DOUBLE
+#   define OS_TYPE "HPUX"
+#   ifdef __LP64__
+#     define CPP_WORDSZ 64
+#     define ALIGNMENT 8
+#   else
+#     define CPP_WORDSZ 32
+#     define ALIGNMENT 4
+#     define ALIGN_DOUBLE
+#   endif
     extern int __data_start;
 #   define DATASTART ((ptr_t)(&__data_start))
 #   if 0
@@ -907,22 +996,33 @@
 	/* this.)							*/
 #       define STACKBOTTOM ((ptr_t) 0x7b033000)  /* from /etc/conf/h/param.h */
 #   else
-#       define HEURISTIC2
+	/* Gustavo Rodriguez-Rivera suggested changing HEURISTIC2	*/
+	/* to this.  We'll probably do this on other platforms, too.	*/
+	/* For now I'll use it where I can test it.			*/
+	extern char ** environ;
+#       define STACKBOTTOM ((ptr_t)environ)
 #   endif
 #   define STACK_GROWS_UP
 #   define DYNAMIC_LOADING
+#   ifndef HPUX_THREADS
+#     define MPROTECT_VDB
+#   endif
 #   include <unistd.h>
 #   define GETPAGESIZE() sysconf(_SC_PAGE_SIZE)
-	/* They misspelled the Posix macro?	*/
 # endif
 
 # ifdef ALPHA
 #   define MACH_TYPE "ALPHA"
 #   define ALIGNMENT 8
+#   define USE_GENERIC_PUSH_REGS
+	/* Gcc and probably the DEC/Compaq compiler spill pointers to preserved	*/
+	/* fp registers in some cases when the target is a 21264.  The assembly	*/
+	/* code doesn't handle that yet, and version dependencies make that a	*/
+	/* bit tricky.  Do the easy thing for now.				*/
 #   ifdef OSF1
 #	define OS_TYPE "OSF1"
 #   	define DATASTART ((ptr_t) 0x140000000)
-	extern _end;
+	extern int _end;
 #   	define DATAEND ((ptr_t) &_end)
 #   	define HEURISTIC2
 	/* Normally HEURISTIC2 is too conervative, since		*/
@@ -939,13 +1039,9 @@
 #       define CPP_WORDSZ 64
 #       define STACKBOTTOM ((ptr_t) 0x120000000)
 #       ifdef __ELF__
-#   	  if 0
-	    /* __data_start apparently disappeared in some recent releases. */
-            extern int __data_start;
-#           define DATASTART &__data_start
-#	  endif
-#         define DATASTART GC_data_start
+#	  define LINUX_DATA_START
 #         define DYNAMIC_LOADING
+	  /* This doesn't work if the collector is in a dynamic library. */
 #       else
 #           define DATASTART ((ptr_t) 0x140000000)
 #       endif
@@ -957,6 +1053,48 @@
 #   endif
 # endif
 
+# ifdef IA64
+#   define MACH_TYPE "IA64"
+#   define ALIGN_DOUBLE
+	/* Requires 16 byte alignment for malloc */
+#   define ALIGNMENT 8
+#   define USE_GENERIC_PUSH_REGS
+	/* We need to get preserved registers in addition to register windows.	*/
+	/* That's easiest to do with setjmp.					*/
+#   ifdef HPUX
+	--> needs work
+#   endif
+#   ifdef LINUX
+#       define OS_TYPE "LINUX"
+#       define CPP_WORDSZ 64
+	/* This should really be done through /proc, but that	*/
+	/* requires we run on an IA64 kernel.			*/
+#       define STACKBOTTOM ((ptr_t) 0xa000000000000000l)
+	/* We also need the base address of the register stack	*/
+	/* backing store.  There is probably a better way to	*/
+	/* get that, too ...					*/
+#	define BACKING_STORE_BASE ((ptr_t) 0x9fffffff80000000l)
+#	if 1
+#	    define SEARCH_FOR_DATA_START
+#	    define DATASTART GC_data_start
+#	else
+	    extern int data_start;
+#	    define DATASTART ((ptr_t)(&data_start))
+#	endif
+#       define DYNAMIC_LOADING
+#	define MPROTECT_VDB
+		/* Requires Linux 2.3.47 or later.	*/
+	extern int _end;
+#	define DATAEND (&_end)
+#	define PREFETCH(x) \
+	  __asm__ ("	lfetch	[%0]": : "r"((void *)(x)))
+#	define PREFETCH_FOR_WRITE(x) \
+	  __asm__ ("	lfetch.excl	[%0]": : "r"((void *)(x)))
+#	define CLEAR_DOUBLE(x) \
+	  __asm__ ("	stf.spill	[%0]=f0": : "r"((void *)(x)))
+#   endif
+# endif
+
 # ifdef M88K
 #   define MACH_TYPE "M88K"
 #   define ALIGNMENT 4
@@ -987,6 +1125,69 @@
 #	define HEURISTIC2
 # endif
 
+# if defined(PJ)
+#   define ALIGNMENT 4
+    extern int _etext;
+#   define DATASTART ((ptr_t)(&_etext))
+#   define HEURISTIC1
+# endif
+
+# ifdef ARM32
+#   define CPP_WORDSZ 32
+#   define MACH_TYPE "ARM32"
+#   define ALIGNMENT 4
+#   ifdef NETBSD
+#       define OS_TYPE "NETBSD"
+#       define HEURISTIC2
+        extern char etext;
+#       define DATASTART ((ptr_t)(&etext))
+#       define USE_GENERIC_PUSH_REGS
+#   endif
+#   ifdef LINUX
+#       define OS_TYPE "LINUX"
+#       define HEURISTIC1
+#       undef STACK_GRAN
+#       define STACK_GRAN 0x10000000
+#       define USE_GENERIC_PUSH_REGS
+#       ifdef __ELF__
+#            define DYNAMIC_LOADING
+#	     include <features.h>
+#	     if defined(__GLIBC__) && __GLIBC__ >= 2
+#		 define LINUX_DATA_START
+#	     else
+     	         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.        */
+			      /* We could use _etext instead, but that  */
+			      /* would include .rodata, which may       */
+			      /* contain large read-only data tables    */
+			      /* that we'd rather not scan.		*/
+#	     endif
+	     extern int _end;
+#	     define DATAEND (&_end)
+#	else
+	     extern int etext;
+#            define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
+#       endif
+#   endif
+#endif
+
+#ifdef LINUX_DATA_START
+    /* Some Linux distributions arrange to define __data_start.  Some	*/
+    /* define data_start as a weak symbol.  The latter is technically	*/
+    /* broken, since the user program may define data_start, in which	*/
+    /* case we lose.  Nonetheless, we try both, prefering __data_start.	*/
+    /* We assume gcc.	*/
+#   pragma weak __data_start
+    extern int __data_start;
+#   pragma weak data_start
+    extern int data_start;
+#   define DATASTART ((ptr_t)(&__data_start != 0? &__data_start : &data_start))
+#endif
+
 # ifndef STACK_GROWS_UP
 #   define STACK_GROWS_DOWN
 # endif
@@ -1029,6 +1230,10 @@
 #   define SUNOS5SIGS
 # endif
 
+# if defined(HPUX)
+#   define SUNOS5SIGS
+# endif
+
 # if CPP_WORDSZ != 32 && CPP_WORDSZ != 64
    -> bad word size
 # endif
@@ -1063,6 +1268,26 @@
 #   define DEFAULT_VDB
 # endif
 
+# ifndef PREFETCH
+#   define PREFETCH(x)
+#   define NO_PREFETCH
+# endif
+
+# ifndef PREFETCH_FOR_WRITE
+#   define PREFETCH_FOR_WRITE(x)
+#   define NO_PREFETCH_FOR_WRITE
+# endif
+
+# ifndef CACHE_LINE_SIZE
+#   define CACHE_LINE_SIZE 32	/* Wild guess	*/
+# endif
+
+# ifndef CLEAR_DOUBLE
+#   define CLEAR_DOUBLE(x) \
+	((word*)x)[0] = 0; \
+	((word*)x)[1] = 0;
+# endif /* CLEAR_DOUBLE */
+
 # if defined(_SOLARIS_PTHREADS) && !defined(SOLARIS_THREADS)
 #   define SOLARIS_THREADS
 # endif
@@ -1078,10 +1303,13 @@
 # if defined(SOLARIS_THREADS) && !defined(SUNOS5)
 --> inconsistent configuration
 # endif
+# if defined(HPUX_THREADS) && !defined(HPUX)
+--> inconsistent configuration
+# endif
 # if defined(PCR) || defined(SRC_M3) || \
 	defined(SOLARIS_THREADS) || defined(WIN32_THREADS) || \
 	defined(IRIX_THREADS) || defined(LINUX_THREADS) || \
-	defined(IRIX_JDK_THREADS)
+	defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
 #   define THREADS
 # endif
 
@@ -1096,4 +1324,4 @@
 				/* include assembly code to do it well.	*/
 # endif
 
-# endif
+# endif /* GCCONFIG_H */
diff --git a/gcj_mlc.c b/gcj_mlc.c
new file mode 100644
index 00000000..eb67eac1
--- /dev/null
+++ b/gcj_mlc.c
@@ -0,0 +1,289 @@
+/*
+ * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company.  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, July 31, 1995 5:02 pm PDT */
+
+#ifdef GC_GCJ_SUPPORT
+
+/*
+ * This is an allocator interface tuned for gcj (the GNU/Cygnus static
+ * java compiler).
+ *
+ * Each allocated object has a pointer in its first word to a vtable,
+ * which for our purposes is simply a structure describing the type of
+ * the object.
+ * This descriptor structur contains a GC marking descriptor at offset
+ * MARK_DESCR_OFFSET.
+ *
+ * It is hoped that this interface may also be useful for other systems,
+ * possibly with some tuning of the constants.  But the immediate goal
+ * is to get better gcj performance.
+ *
+ * We assume:
+ *  1) We have an ANSI conforming C compiler.
+ *  2) Counting on explicit initialization of this interface is OK.
+ *  3) FASTLOCK is not a significant win.
+ */
+
+#include "gc_priv.h"
+#include "gc_mark.h"
+#include "include/gc_gcj.h"
+#include "dbg_mlc.h"
+
+GC_bool GC_gcj_malloc_initialized = FALSE;
+
+int GC_gcj_kind;	/* Object kind for objects with descriptors     */
+			/* in "vtable".					*/
+int GC_gcj_debug_kind;	/* The kind of objects that is always marked 	*/
+			/* with a mark proc call.			*/
+
+ptr_t * GC_gcjobjfreelist;
+ptr_t * GC_gcjdebugobjfreelist;
+
+void * GC_default_oom_action(void) { return 0; }
+
+void * (*GC_oom_action)(void) = GC_default_oom_action;
+
+/* Caller does not hold allocation lock. */
+void GC_init_gcj_malloc(int mp_index, void * /* really mark_proc */mp)
+{
+    register int i;
+    DCL_LOCK_STATE;
+
+    GC_init();	/* In case it's not already done.	*/
+    DISABLE_SIGNALS();
+    LOCK();
+    if (GC_gcj_malloc_initialized) {
+      UNLOCK();
+      ENABLE_SIGNALS();
+      return;
+    }
+    GC_gcj_malloc_initialized = TRUE;
+    GC_mark_procs[mp_index] = (mark_proc)mp;
+    if (mp_index >= GC_n_mark_procs) ABORT("GC_init_gcj_malloc: bad index");
+    /* Set up object kind gcj-style indirect descriptor. */
+      GC_gcjobjfreelist = (ptr_t *)
+          GC_generic_malloc_inner((MAXOBJSZ+1)*sizeof(ptr_t), PTRFREE);
+      if (GC_gcjobjfreelist == 0) ABORT("Couldn't allocate GC_gcjobjfreelist");
+      BZERO(GC_gcjobjfreelist, (MAXOBJSZ+1)*sizeof(ptr_t));
+      GC_gcj_kind = GC_n_kinds++;
+      GC_obj_kinds[GC_gcj_kind].ok_freelist = GC_gcjobjfreelist;
+      GC_obj_kinds[GC_gcj_kind].ok_reclaim_list = 0;
+      GC_obj_kinds[GC_gcj_kind].ok_descriptor =
+    	(((word)(-MARK_DESCR_OFFSET - INDIR_PER_OBJ_BIAS)) | DS_PER_OBJECT);
+      GC_obj_kinds[GC_gcj_kind].ok_relocate_descr = FALSE;
+      GC_obj_kinds[GC_gcj_kind].ok_init = TRUE;
+    /* Set up object kind for objects that require mark proc call.	*/
+      GC_gcjdebugobjfreelist = (ptr_t *)
+          GC_generic_malloc_inner((MAXOBJSZ+1)*sizeof(ptr_t), PTRFREE);
+      if (GC_gcjdebugobjfreelist == 0)
+	  ABORT("Couldn't allocate GC_gcjdebugobjfreelist");
+      BZERO(GC_gcjdebugobjfreelist, (MAXOBJSZ+1)*sizeof(ptr_t));
+      GC_gcj_debug_kind = GC_n_kinds++;
+      GC_obj_kinds[GC_gcj_debug_kind].ok_freelist = GC_gcjdebugobjfreelist;
+      GC_obj_kinds[GC_gcj_debug_kind].ok_reclaim_list = 0;
+      GC_obj_kinds[GC_gcj_debug_kind].ok_descriptor =
+    	MAKE_PROC(mp_index, 1 /* allocated with debug info */);
+      GC_obj_kinds[GC_gcj_debug_kind].ok_relocate_descr = FALSE;
+      GC_obj_kinds[GC_gcj_debug_kind].ok_init = TRUE;
+    UNLOCK();
+    ENABLE_SIGNALS();
+}
+
+ptr_t GC_clear_stack();
+
+#define GENERAL_MALLOC(lb,k) \
+    (GC_PTR)GC_clear_stack(GC_generic_malloc_inner((word)lb, k))
+    
+#define GENERAL_MALLOC_IOP(lb,k) \
+    (GC_PTR)GC_clear_stack(GC_generic_malloc_inner_ignore_off_page(lb, k))
+
+/* Allocate an object, clear it, and store the pointer to the	*/
+/* type structure (vtable in gcj).				*/
+/* This adds a byte at the end of the object if GC_malloc would.*/
+void * GC_gcj_malloc(size_t lb, void * ptr_to_struct_containing_descr)
+{
+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 = ALIGNED_WORDS(lb);
+#       endif
+	opp = &(GC_gcjobjfreelist[lw]);
+	LOCK();
+        if( (op = *opp) == 0 ) {
+            op = (ptr_t)GENERAL_MALLOC((word)lb, GC_gcj_kind);
+	    if (0 == op) {
+		UNLOCK();
+		return(GC_oom_action());
+	    }
+#	    ifdef MERGE_SIZES
+		lw = GC_size_map[lb];	/* May have been uninitialized.	*/
+#	    endif
+        } else {
+            *opp = obj_link(op);
+            GC_words_allocd += lw;
+            FASTUNLOCK();
+        }
+	*(void **)op = ptr_to_struct_containing_descr;
+	UNLOCK();
+    } else {
+	LOCK();
+	op = (ptr_t)GENERAL_MALLOC((word)lb, GC_gcj_kind);
+	if (0 == op) {
+	    UNLOCK();
+	    return(GC_oom_action());
+	}
+	*(void **)op = ptr_to_struct_containing_descr;
+	UNLOCK();
+    }
+    return((GC_PTR) op);
+}
+
+/* Similar to GC_gcj_malloc, but add debug info.  This is allocated	*/
+/* with GC_gcj_debug_kind.						*/
+GC_PTR GC_debug_gcj_malloc(size_t lb, void * ptr_to_struct_containing_descr,
+			   GC_EXTRA_PARAMS)
+{
+    GC_PTR result;
+
+    /* We clone the code from GC_debug_gcj_malloc, so that we 	*/
+    /* dont end up with extra frames on the stack, which could	*/
+    /* confuse the backtrace.					*/
+    LOCK();
+    result = GC_generic_malloc_inner(lb + DEBUG_BYTES, GC_gcj_debug_kind);
+    if (result == 0) {
+	UNLOCK();
+        GC_err_printf2("GC_debug_gcj_malloc(%ld, 0x%lx) returning NIL (",
+        	       (unsigned long) lb,
+		       (unsigned long) ptr_to_struct_containing_descr);
+        GC_err_puts(s);
+        GC_err_printf1(":%ld)\n", (unsigned long)i);
+        return(GC_oom_action());
+    }
+    *((void **)((ptr_t)result + sizeof(oh))) = ptr_to_struct_containing_descr;
+    UNLOCK();
+    if (!GC_debugging_started) {
+    	GC_start_debugging();
+    }
+    ADD_CALL_CHAIN(result, ra);
+    return (GC_store_debug_info(result, (word)lb, s, (word)i));
+}
+
+/* Similar to GC_gcj_malloc, but the size is in words, and we don't	*/
+/* adjust it.  The size is assumed to be such that it can be 	*/
+/* allocated as a small object.					*/
+void * GC_gcj_fast_malloc(size_t lw, void * ptr_to_struct_containing_descr)
+{
+ptr_t op;
+ptr_t * opp;
+DCL_LOCK_STATE;
+
+    opp = &(GC_gcjobjfreelist[lw]);
+    LOCK();
+    if( (op = *opp) == 0 ) {
+        op = (ptr_t)GC_clear_stack(
+		GC_generic_malloc_words_small_inner(lw, GC_gcj_kind));
+	if (0 == op) {
+	    UNLOCK();
+	    return(GC_oom_action());
+	}
+    } else {
+        *opp = obj_link(op);
+        GC_words_allocd += lw;
+    }
+    *(void **)op = ptr_to_struct_containing_descr;
+    UNLOCK();
+    return((GC_PTR) op);
+}
+
+/* And a debugging version of the above:	*/
+void * GC_debug_gcj_fast_malloc(size_t lw,
+				void * ptr_to_struct_containing_descr,
+				GC_EXTRA_PARAMS)
+{
+    GC_PTR result;
+    size_t lb = WORDS_TO_BYTES(lw);
+
+    /* We clone the code from GC_debug_gcj_malloc, so that we 	*/
+    /* dont end up with extra frames on the stack, which could	*/
+    /* confuse the backtrace.					*/
+    LOCK();
+    result = GC_generic_malloc_inner(lb + DEBUG_BYTES, GC_gcj_debug_kind);
+    if (result == 0) {
+	UNLOCK();
+        GC_err_printf2("GC_debug_gcj_fast_malloc(%ld, 0x%lx) returning NIL (",
+        	       (unsigned long) lw,
+		       (unsigned long) ptr_to_struct_containing_descr);
+        GC_err_puts(s);
+        GC_err_printf1(":%ld)\n", (unsigned long)i);
+        return(GC_oom_action());
+    }
+    *((void **)((ptr_t)result + sizeof(oh))) = ptr_to_struct_containing_descr;
+    UNLOCK();
+    if (!GC_debugging_started) {
+    	GC_start_debugging();
+    }
+    ADD_CALL_CHAIN(result, ra);
+    return (GC_store_debug_info(result, (word)lb, s, (word)i));
+}
+
+void * GC_gcj_malloc_ignore_off_page(size_t lb,
+				     void * ptr_to_struct_containing_descr) 
+{
+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 = ALIGNED_WORDS(lb);
+#       endif
+	opp = &(GC_gcjobjfreelist[lw]);
+	LOCK();
+        if( (op = *opp) == 0 ) {
+            op = (ptr_t)GENERAL_MALLOC_IOP(lb, GC_gcj_kind);
+#	    ifdef MERGE_SIZES
+		lw = GC_size_map[lb];	/* May have been uninitialized.	*/
+#	    endif
+        } else {
+            *opp = obj_link(op);
+            GC_words_allocd += lw;
+            FASTUNLOCK();
+        }
+	*(void **)op = ptr_to_struct_containing_descr;
+	UNLOCK();
+    } else {
+        op = (ptr_t)GENERAL_MALLOC_IOP(lb, GC_gcj_kind);
+        if (0 != op) {
+          *(void **)op = ptr_to_struct_containing_descr;
+	}
+        UNLOCK();
+    }
+    return((GC_PTR) op);
+}
+
+#else
+
+char GC_no_gcj_support;
+
+#endif  /* GC_GCJ_SUPPORT */
diff --git a/headers.c b/headers.c
index 9564a6a5..6e47bba8 100644
--- a/headers.c
+++ b/headers.c
@@ -50,10 +50,8 @@ ptr_t h;
  
 static ptr_t scratch_free_ptr = 0;
  
-ptr_t GC_scratch_end_ptr = 0;
-
-ptr_t GC_scratch_last_end_ptr = 0;
-		/* End point of last obtained scratch area */
+/* GC_scratch_last_end_ptr is end point of last obtained scratch area.  */
+/* GC_scratch_end_ptr is end point of current scratch area.		*/
  
 ptr_t GC_scratch_alloc(bytes)
 register word bytes;
@@ -128,6 +126,13 @@ hdr * hhdr;
     hhdr -> hb_next = (struct hblk *) hdr_free_list;
     hdr_free_list = hhdr;
 }
+
+hdr * GC_invalid_header;
+
+#ifdef USE_HDR_CACHE
+  word GC_hdr_cache_hits = 0;
+  word GC_hdr_cache_misses = 0;
+#endif
  
 void GC_init_headers()
 {
@@ -138,6 +143,8 @@ void GC_init_headers()
     for (i = 0; i < TOP_SZ; i++) {
         GC_top_index[i] = GC_all_nils;
     }
+    GC_invalid_header = alloc_hdr();
+    GC_invalidate_map(GC_invalid_header);
 }
 
 /* Make sure that there is a bottom level index block for address addr  */
@@ -191,10 +198,10 @@ word addr;
     return(TRUE);
 }
 
-/* Install a header for block h.  */
-/* The header is uninitialized.	  */
-/* Returns FALSE on failure.	  */
-GC_bool GC_install_header(h)
+/* Install a header for block h.	*/
+/* The header is uninitialized.	  	*/
+/* Returns the header or 0 on failure.	*/
+struct hblkhdr * GC_install_header(h)
 register struct hblk * h;
 {
     hdr * result;
@@ -205,7 +212,7 @@ register struct hblk * h;
 #   ifdef USE_MUNMAP
 	result -> hb_last_reclaimed = GC_gc_no;
 #   endif
-    return(result != 0);
+    return(result);
 }
 
 /* Set up forwarding counts for block h of size sz */
diff --git a/irix_threads.c b/hpux_irix_threads.c
index 5efca211..f6e74365 100644
--- a/irix_threads.c
+++ b/hpux_irix_threads.c
@@ -1,6 +1,7 @@
 /* 
- * Copyright (c) 1994 by Xerox Corporation.  All rights reserved.
- * Copyright (c) 1996 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -16,12 +17,19 @@
  * not guaranteed by the Pthread standard.  It may or may not be portable
  * to other implementations.
  *
+ * This now also includes an initial attempt at thread support for
+ * HP/UX 11.
+ *
  * Note that there is a lot of code duplication between linux_threads.c
- * and irix_threads.c; any changes made here may need to be reflected
+ * and hpux_irix_threads.c; any changes made here may need to be reflected
  * there too.
  */
 
-# if defined(IRIX_THREADS)
+# if defined(IRIX_THREADS) || defined(HPUX_THREADS)
+
+# if defined(HPUX_THREADS)
+#   include <sys/semaphore.h>
+# endif
 
 # include "gc_priv.h"
 # include <pthread.h>
@@ -169,8 +177,12 @@ ptr_t GC_stack_alloc(size_t * stack_size)
         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); */
-        result += GC_page_sz;
+#	ifdef STACK_GROWS_UP
+          /* mprotect(result + search_sz, GC_page_sz, PROT_NONE); */
+#	else
+          /* mprotect(result, GC_page_sz, PROT_NONE); */
+          result += GC_page_sz;
+#	endif
     }
     *stack_size = search_sz;
     return(result);
@@ -375,13 +387,14 @@ int GC_is_thread_stack(ptr_t addr)
 }
 # endif
 
-/* We hold allocation lock.  We assume the world is stopped.	*/
+/* We hold allocation lock.  Should do exactly the right thing if the	*/
+/* world is stopped.  Should not fail if it isn't.			*/
 void GC_push_all_stacks()
 {
     register int i;
     register GC_thread p;
     register ptr_t sp = GC_approx_sp();
-    register ptr_t lo, hi;
+    register ptr_t hot, cold;
     pthread_t me = pthread_self();
     
     if (!GC_thr_initialized) GC_thr_init();
@@ -390,17 +403,25 @@ void GC_push_all_stacks()
       for (p = GC_threads[i]; p != 0; p = p -> next) {
         if (p -> flags & FINISHED) continue;
         if (pthread_equal(p -> id, me)) {
-	    lo = GC_approx_sp();
+	    hot = GC_approx_sp();
 	} else {
-	    lo = p -> stack_ptr;
+	    hot = p -> stack_ptr;
 	}
         if (p -> stack_size != 0) {
-            hi = p -> stack + p -> stack_size;
+#	  ifdef STACK_GROWS_UP
+	    cold = p -> stack;
+#	  else
+            cold = p -> stack + p -> stack_size;
+#	  endif
         } else {
             /* The original stack. */
-            hi = GC_stackbottom;
+            cold = GC_stackbottom;
         }
-        GC_push_all_stack(lo, hi);
+#	ifdef STACK_GROWS_UP
+          GC_push_all_stack(cold, hot);
+#	else
+          GC_push_all_stack(hot, cold);
+#	endif
       }
     }
 }
@@ -531,6 +552,40 @@ void * GC_start_routine(void * arg)
     return(result);
 }
 
+# ifdef HPUX_THREADS
+  /* pthread_attr_t is not a structure, thus a simple structure copy	*/
+  /* won't work.							*/
+  static void copy_attr(pthread_attr_t * pa_ptr,
+			const pthread_attr_t  * source) {
+    int tmp;
+    size_t stmp;
+    void * vtmp;
+    struct sched_param sp_tmp;
+    pthread_spu_t ps_tmp;
+    (void) pthread_attr_init(pa_ptr);
+    (void) pthread_attr_getdetachstate(source, &tmp);
+    (void) pthread_attr_setdetachstate(pa_ptr, tmp);
+    (void) pthread_attr_getinheritsched(source, &tmp);
+    (void) pthread_attr_setinheritsched(pa_ptr, tmp);
+    (void) pthread_attr_getschedpolicy(source, &tmp);
+    (void) pthread_attr_setschedpolicy(pa_ptr, tmp);
+    (void) pthread_attr_getstacksize(source, &stmp);
+    (void) pthread_attr_setstacksize(pa_ptr, stmp);
+    (void) pthread_attr_getguardsize(source, &stmp);
+    (void) pthread_attr_setguardsize(pa_ptr, stmp);
+    (void) pthread_attr_getstackaddr(source, &vtmp);
+    (void) pthread_attr_setstackaddr(pa_ptr, vtmp);
+    (void) pthread_attr_getscope(source, &tmp);
+    (void) pthread_attr_setscope(pa_ptr, tmp);
+    (void) pthread_attr_getschedparam(source, &sp_tmp);
+    (void) pthread_attr_setschedparam(pa_ptr, &sp_tmp);
+    (void) pthread_attr_getprocessor_np(source, &ps_tmp, &tmp);
+    (void) pthread_attr_setprocessor_np(pa_ptr, ps_tmp, tmp);
+  }
+# else
+#   define copy_attr(pa_ptr, source) *(pa_ptr) = *(source)
+# endif
+
 int
 GC_pthread_create(pthread_t *new_thread,
 		  const pthread_attr_t *attr,
@@ -548,7 +603,9 @@ GC_pthread_create(pthread_t *new_thread,
 	/* library, which isn't visible to the collector.		 */
 
     if (0 == si) return(ENOMEM);
-    sem_init(&(si -> registered), 0, 0);
+    if (0 != sem_init(&(si -> registered), 0, 0)) {
+        ABORT("sem_init failed");
+    }
     si -> start_routine = start_routine;
     si -> arg = arg;
     LOCK();
@@ -557,7 +614,7 @@ GC_pthread_create(pthread_t *new_thread,
         stack = 0;
 	(void) pthread_attr_init(&new_attr);
     } else {
-        new_attr = *attr;
+	copy_attr(&new_attr, attr);
 	pthread_attr_getstackaddr(&new_attr, &stack);
     }
     pthread_attr_getstacksize(&new_attr, &stacksize);
@@ -586,24 +643,39 @@ GC_pthread_create(pthread_t *new_thread,
     /* This also ensures that we hold onto si until the child is done	*/
     /* with it.  Thus it doesn't matter whether it is otherwise		*/
     /* visible to the collector.					*/
-        if (0 != sem_wait(&(si -> registered))) ABORT("sem_wait failed");
+        while (0 != sem_wait(&(si -> registered))) {
+	  if (errno != EINTR) {
+	    GC_printf1("Sem_wait: errno = %ld\n", (unsigned long) errno);
+	    ABORT("sem_wait failed");
+	  }
+	}
         sem_destroy(&(si -> registered));
-    /* pthread_attr_destroy(&new_attr); */
+    pthread_attr_destroy(&new_attr);  /* Not a no-op under HPUX */
     return(result);
 }
 
+#ifndef HPUX_THREADS
+/* For now we use the pthreads locking primitives on HP/UX */
+
 GC_bool GC_collecting = 0; /* A hint that we're in the collector and       */
                         /* holding the allocation lock for an           */
                         /* extended period.                             */
 
 /* Reasonably fast spin locks.  Basically the same implementation */
-/* as STL alloc.h.  This isn't really the right way to do this.   */
-/* but until the POSIX scheduling mess gets straightened out ...  */
-
-unsigned long GC_allocate_lock = 0;
+/* as STL alloc.h.						  */
 
 #define SLEEP_THRESHOLD 3
 
+#ifdef HPUX
+   unsigned long GC_allocate_lock = 1;
+#  define GC_TRY_LOCK() GC_test_and_clear(&GC_allocate_lock)
+#  define GC_LOCK_TAKEN !GC_allocate_lock
+#else
+   unsigned long GC_allocate_lock = 0;
+#  define GC_TRY_LOCK() !GC_test_and_set(&GC_allocate_lock,1)
+#  define GC_LOCK_TAKEN GC_allocate_lock
+#endif
+
 void GC_lock()
 {
 #   define low_spin_max 30  /* spin cycles if we suspect uniprocessor */
@@ -616,7 +688,7 @@ void GC_lock()
 #   define PAUSE junk *= junk; junk *= junk; junk *= junk; junk *= junk
     int i;
 
-    if (!GC_test_and_set(&GC_allocate_lock, 1)) {
+    if (GC_TRY_LOCK()) {
         return;
     }
     junk = 0;
@@ -624,11 +696,11 @@ void GC_lock()
     my_last_spins = last_spins;
     for (i = 0; i < my_spin_max; i++) {
         if (GC_collecting) goto yield;
-        if (i < my_last_spins/2 || GC_allocate_lock) {
+        if (i < my_last_spins/2 || GC_LOCK_TAKEN) {
             PAUSE; 
             continue;
         }
-        if (!GC_test_and_set(&GC_allocate_lock, 1)) {
+        if (GC_TRY_LOCK()) {
 	    /*
              * got it!
              * Spinning worked.  Thus we're probably not being scheduled
@@ -644,7 +716,7 @@ void GC_lock()
     spin_max = low_spin_max;
 yield:
     for (i = 0;; ++i) {
-        if (!GC_test_and_set(&GC_allocate_lock, 1)) {
+        if (GC_TRY_LOCK()) {
             return;
         }
         if (i < SLEEP_THRESHOLD) {
@@ -662,7 +734,7 @@ yield:
     }
 }
 
-
+#endif /* !HPUX_THREADS */
 
 # else
 
diff --git a/hpux_test_and_clear.s b/hpux_test_and_clear.s
new file mode 100644
index 00000000..f09b2114
--- /dev/null
+++ b/hpux_test_and_clear.s
@@ -0,0 +1,21 @@
+	.SPACE $PRIVATE$
+	.SUBSPA $DATA$,QUAD=1,ALIGN=8,ACCESS=31
+	.SUBSPA $BSS$,QUAD=1,ALIGN=8,ACCESS=31,ZERO,SORT=82
+	.SPACE $TEXT$
+	.SUBSPA $LIT$,QUAD=0,ALIGN=8,ACCESS=44
+	.SUBSPA $CODE$,QUAD=0,ALIGN=8,ACCESS=44,CODE_ONLY
+	.IMPORT $global$,DATA
+	.IMPORT $$dyncall,MILLICODE
+	.SPACE $TEXT$
+	.SUBSPA $CODE$
+
+	.align 4
+	.EXPORT GC_test_and_clear,ENTRY,PRIV_LEV=3,ARGW0=GR,RTNVAL=GR
+GC_test_and_clear
+	.PROC
+	.CALLINFO FRAME=0,NO_CALLS
+	.ENTRY
+	ldcw,co (%r26),%r28
+	bv,n 0(%r2)
+	.EXIT
+	.PROCEND
diff --git a/include/backptr.h b/include/backptr.h
index d34224e4..f5b7c5aa 100644
--- a/include/backptr.h
+++ b/include/backptr.h
@@ -29,9 +29,11 @@
 /*   source is heap object ==> *base_p != 0, *offset_p = offset       */
 /*   Returns 1 on success, 0 if source couldn't be determined.        */
 /* Dest can be any address within a heap object.                      */
-typedef enum {  GC_UNREFERENCED, /* No refence info available.		*/
+typedef enum {  GC_UNREFERENCED, /* No reference info available.	*/
 		GC_NO_SPACE,	/* Dest not allocated with debug alloc  */
 		GC_REFD_FROM_ROOT, /* Referenced directly by root *base_p */
+		GC_REFD_FROM_REG,  /* Referenced from a register, i.e.	*/
+				   /* a root without an address.	*/
 		GC_REFD_FROM_HEAP, /* Referenced from another heap obj. */
 		GC_FINALIZER_REFD /* Finalizable and hence accessible.  */
 } GC_ref_kind;
@@ -53,4 +55,9 @@ void * GC_generate_random_valid_address(void);
 /* source in dbg_mlc.c also serves as a sample client.	      */
 void GC_generate_random_backtrace(void);
 
+/* Print a backtrace from a specific address.  Used by the 	*/
+/* above.  The client should call GC_gcollect() immediately	*/
+/* before invocation.						*/
+void GC_print_backtrace(void *);
+
 
diff --git a/include/cord.h b/include/cord.h
index 584112fd..926089e8 100644
--- a/include/cord.h
+++ b/include/cord.h
@@ -41,7 +41,7 @@
  * This interface is fairly big, largely for performance reasons.
  * The most basic constants and functions:
  *
- * CORD - the type fo a cord;
+ * CORD - the type of a cord;
  * CORD_EMPTY - empty cord;
  * CORD_len(cord) - length of a cord;
  * CORD_cat(cord1,cord2) - concatenation of two cords;
diff --git a/include/gc.h b/include/gc.h
index 30614095..d7cd8996 100644
--- a/include/gc.h
+++ b/include/gc.h
@@ -1,7 +1,8 @@
 /* 
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
- * Copyright 1996 by Silicon Graphics.  All rights reserved.
+ * Copyright 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright 1999 by Hewlett-Packard Company.  All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -35,6 +36,14 @@
 #include "libgc_globals.h"
 #endif
 
+#if defined(__MINGW32__) && defined(WIN32_THREADS)
+# ifdef GC_BUILD
+#   define GC_API __declspec(dllexport)
+# else
+#   define GC_API __declspec(dllimport)
+# endif
+#endif
+
 #if defined(_MSC_VER) && defined(_DLL)
 # ifdef GC_BUILD
 #   define GC_API __declspec(dllexport)
@@ -58,9 +67,11 @@
 # if defined(__STDC__) || defined(__cplusplus)
 #   define GC_PROTO(args) args
     typedef void * GC_PTR;
+#   define GC_CONST const
 # else
 #   define GC_PROTO(args) ()
     typedef char * GC_PTR;
+#   define GC_CONST
 #  endif
 
 # ifdef __cplusplus
@@ -128,9 +139,26 @@ GC_API int GC_dont_expand;
 			/* Dont expand heap unless explicitly requested */
 			/* or forced to.				*/
 
+GC_API int GC_use_entire_heap;
+		/* Causes the nonincremental collector to use the	*/
+		/* entire heap before collecting.  This was the only 	*/
+		/* option for GC versions < 5.0.  This sometimes	*/
+		/* results in more large block fragmentation, since	*/
+		/* very larg blocks will tend to get broken up		*/
+		/* during each GC cycle.  It is likely to result in a	*/
+		/* larger working set, but lower collection		*/
+		/* frequencies, and hence fewer instructions executed	*/
+		/* in the collector.					*/
+
 GC_API int GC_full_freq;    /* Number of partial collections between	*/
 			    /* full collections.  Matters only if	*/
 			    /* GC_incremental is set.			*/
+			    /* Full collections are also triggered if	*/
+			    /* the collector detects a substantial	*/
+			    /* increase in the number of in-use heap	*/
+			    /* blocks.  Values in the tens are now	*/
+			    /* perfectly reasonable, unlike for		*/
+			    /* earlier GC versions.			*/
 			
 GC_API GC_word GC_non_gc_bytes;
 			/* Bytes not considered candidates for collection. */
@@ -297,6 +325,9 @@ GC_API int GC_try_to_collect GC_PROTO((GC_stop_func stop_func));
 /* Includes some pages that were allocated but never written.		*/
 GC_API size_t GC_get_heap_size GC_PROTO((void));
 
+/* Return a lower bound on the number of free bytes in the heap.	*/
+GC_API size_t GC_get_free_bytes GC_PROTO((void));
+
 /* Return the number of bytes allocated since the last collection.	*/
 GC_API size_t GC_get_bytes_since_gc GC_PROTO((void));
 
@@ -341,10 +372,10 @@ GC_API GC_PTR GC_malloc_atomic_ignore_off_page GC_PROTO((size_t lb));
 
 #ifdef GC_ADD_CALLER
 #  define GC_EXTRAS GC_RETURN_ADDR, __FILE__, __LINE__
-#  define GC_EXTRA_PARAMS GC_word ra, char * descr_string, int descr_int
+#  define GC_EXTRA_PARAMS GC_word ra, GC_CONST char * s, int i
 #else
 #  define GC_EXTRAS __FILE__, __LINE__
-#  define GC_EXTRA_PARAMS char * descr_string, int descr_int
+#  define GC_EXTRA_PARAMS GC_CONST char * s, int i
 #endif
 
 /* Debugging (annotated) allocation.  GC_gcollect will check 		*/
@@ -375,6 +406,8 @@ GC_API void GC_debug_end_stubborn_change GC_PROTO((GC_PTR));
 	GC_debug_register_finalizer(p, f, d, of, od)
 #   define GC_REGISTER_FINALIZER_IGNORE_SELF(p, f, d, of, od) \
 	GC_debug_register_finalizer_ignore_self(p, f, d, of, od)
+#   define GC_REGISTER_FINALIZER_NO_ORDER(p, f, d, of, od) \
+	GC_debug_register_finalizer_no_order(p, f, d, of, od)
 #   define GC_MALLOC_STUBBORN(sz) GC_debug_malloc_stubborn(sz, GC_EXTRAS);
 #   define GC_CHANGE_STUBBORN(p) GC_debug_change_stubborn(p)
 #   define GC_END_STUBBORN_CHANGE(p) GC_debug_end_stubborn_change(p)
@@ -391,6 +424,8 @@ GC_API void GC_debug_end_stubborn_change GC_PROTO((GC_PTR));
 	GC_register_finalizer(p, f, d, of, od)
 #   define GC_REGISTER_FINALIZER_IGNORE_SELF(p, f, d, of, od) \
 	GC_register_finalizer_ignore_self(p, f, d, of, od)
+#   define GC_REGISTER_FINALIZER_NO_ORDER(p, f, d, of, od) \
+	GC_register_finalizer_no_order(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)
@@ -469,6 +504,16 @@ GC_API void GC_debug_register_finalizer_ignore_self
 	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
 		  GC_finalization_proc *ofn, GC_PTR *ocd));
 
+/* Another version of the above.  It ignores all cycles.        */
+/* It should probably only be used by Java implementations.      */
+GC_API void GC_register_finalizer_no_order
+	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
+		  GC_finalization_proc *ofn, GC_PTR *ocd));
+GC_API void GC_debug_register_finalizer_no_order
+	GC_PROTO((GC_PTR obj, GC_finalization_proc fn, GC_PTR cd,
+		  GC_finalization_proc *ofn, GC_PTR *ocd));
+
+
 /* 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	*/
@@ -525,6 +570,9 @@ GC_API int GC_unregister_disappearing_link GC_PROTO((GC_PTR * /* link */));
 GC_API GC_PTR GC_make_closure GC_PROTO((GC_finalization_proc fn, GC_PTR data));
 GC_API void GC_debug_invoke_finalizer GC_PROTO((GC_PTR obj, GC_PTR data));
 
+/* Returns !=0  if GC_invoke_finalizers has something to do. 		*/
+GC_API int GC_should_invoke_finalizers GC_PROTO((void));
+
 GC_API int GC_invoke_finalizers GC_PROTO((void));
 	/* Run finalizers for all objects that are ready to	*/
 	/* be finalized.  Return the number of finalizers	*/
@@ -688,7 +736,8 @@ GC_API void (*GC_is_visible_print_proc)
 # endif /* SOLARIS_THREADS */
 
 
-#if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#if !defined(USE_LD_WRAP) && \
+    (defined(IRIX_THREADS) || defined(LINUX_THREADS) || defined(HPUX_THREADS))
 /* We treat these similarly. */
 # include <pthread.h>
 # include <signal.h>
@@ -702,12 +751,13 @@ GC_API void (*GC_is_visible_print_proc)
 # define pthread_create GC_pthread_create
 # define pthread_sigmask GC_pthread_sigmask
 # define pthread_join GC_pthread_join
+# define dlopen GC_dlopen
 
-#endif /* IRIX_THREADS || LINUX_THREADS */
+#endif /* xxxxx_THREADS */
 
 # if defined(PCR) || defined(SOLARIS_THREADS) || defined(WIN32_THREADS) || \
 	defined(IRIX_THREADS) || defined(LINUX_THREADS) || \
-	defined(IRIX_JDK_THREADS)
+	defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
    	/* Any flavor of threads except SRC_M3.	*/
 /* This returns a list of objects, linked through their first		*/
 /* word.  Its use can greatly reduce lock contention problems, since	*/
diff --git a/gc_copy_descr.h b/include/gc_copy_descr.h
index 212c99e2..212c99e2 100644
--- a/gc_copy_descr.h
+++ b/include/gc_copy_descr.h
diff --git a/include/gc_cpp.h b/include/gc_cpp.h
index ad7df5d7..3b86824c 100644
--- a/include/gc_cpp.h
+++ b/include/gc_cpp.h
@@ -16,12 +16,11 @@ the code was modified is included with the above copyright notice.
 C++ Interface to the Boehm Collector
 
     John R. Ellis and Jesse Hull 
-    Last modified on Mon Jul 24 15:43:42 PDT 1995 by ellis
 
 This interface provides access to the Boehm collector.  It provides
 basic facilities similar to those described in "Safe, Efficient
 Garbage Collection for C++", by John R. Elis and David L. Detlefs
-(ftp.parc.xerox.com:/pub/ellis/gc).
+(ftp://ftp.parc.xerox.com/pub/ellis/gc).
 
 All heap-allocated objects are either "collectable" or
 "uncollectable".  Programs must explicitly delete uncollectable
@@ -38,7 +37,7 @@ Objects derived from class "gc" are collectable.  For example:
     A* a = new A;       // a is collectable. 
 
 Collectable instances of non-class types can be allocated using the GC
-placement:
+(or UseGC) placement:
 
     typedef int A[ 10 ];
     A* a = new (GC) A;
@@ -124,6 +123,12 @@ invoked using the ANSI-conforming syntax t->~T().  If you're using
 cfront 3.0, you'll have to comment out the class gc_cleanup, which
 uses explicit invocation.
 
+5. GC name conflicts:
+
+Many other systems seem to use the identifier "GC" as an abbreviation
+for "Graphics Context".  Since version 5.0, GC placement has been replaced
+by UseGC.  GC is an alias for UseGC, unless GC_NAME_CONFLICT is defined.
+
 ****************************************************************************/
 
 #include "gc.h"
@@ -134,11 +139,15 @@ uses explicit invocation.
 
 #if ! defined( OPERATOR_NEW_ARRAY ) \
     && (__BORLANDC__ >= 0x450 || (__GNUC__ >= 2 && __GNUC_MINOR__ >= 6) \
-        || __WATCOMC__ >= 1050)
+        || __WATCOMC__ >= 1050 || _MSC_VER >= 1100)
 #   define OPERATOR_NEW_ARRAY
 #endif
 
-enum GCPlacement {GC, NoGC, PointerFreeGC};
+enum GCPlacement {UseGC,
+#ifndef GC_NAME_CONFLICT
+		  GC=UseGC,
+#endif
+                  NoGC, PointerFreeGC};
 
 class gc {public:
     inline void* operator new( size_t size );
@@ -191,6 +200,17 @@ inline void* operator new(
 
 #ifdef OPERATOR_NEW_ARRAY
 
+#ifdef _MSC_VER
+ /** This ensures that the system default operator new[] doesn't get
+  *  undefined, which is what seems to happen on VC++ 6 for some reason
+  *  if we define a multi-argument operator new[].
+  */
+ inline void *operator new[]( size_t size )
+ {
+    return ::operator new( size );
+ }
+#endif /* _MSC_VER */
+
 inline void* operator new[](
     size_t size, 
     GCPlacement gcp,
@@ -211,7 +231,7 @@ 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) 
+    if (gcp == UseGC) 
         return GC_MALLOC( size );
     else if (gcp == PointerFreeGC)
 	return GC_MALLOC_ATOMIC( size );
@@ -246,12 +266,12 @@ inline gc_cleanup::gc_cleanup() {
     GC_finalization_proc oldProc;
     void* oldData;
     void* base = GC_base( (void *) this );
-    if (0 == base) return;
-    GC_REGISTER_FINALIZER_IGNORE_SELF( 
-        base, cleanup, (void*) ((char*) this - (char*) base), 
+    if (0 != base)  {
+      GC_REGISTER_FINALIZER_IGNORE_SELF( 
+        base, (GC_finalization_proc)cleanup, (void*) ((char*) this - (char*) base), 
         &oldProc, &oldData );
-    if (0 != oldProc) {
-        GC_REGISTER_FINALIZER_IGNORE_SELF( base, oldProc, oldData, 0, 0 );}}
+      if (0 != oldProc) {
+        GC_REGISTER_FINALIZER_IGNORE_SELF( base, oldProc, oldData, 0, 0 );}}}
 
 inline void* operator new( 
     size_t size, 
@@ -261,7 +281,7 @@ inline void* operator new(
 {
     void* obj;
 
-    if (gcp == GC) {
+    if (gcp == UseGC) {
         obj = GC_MALLOC( size );
         if (cleanup != 0) 
             GC_REGISTER_FINALIZER_IGNORE_SELF( 
diff --git a/include/gc_gcj.h b/include/gc_gcj.h
new file mode 100644
index 00000000..1dede515
--- /dev/null
+++ b/include/gc_gcj.h
@@ -0,0 +1,101 @@
+/* 
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
+ * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
+ * Copyright 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright 1999 by Hewlett-Packard Company.  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 assumes the collector has been compiled with GC_GCJ_SUPPORT */
+/* and that an ANSI C compiler is available.				 */
+
+/*
+ * We allocate objects whose first word contains a pointer to a struct
+ * describing the object type.  This struct contains a garbage collector mark
+ * descriptor at offset MARK_DESCR_OFFSET.  Alternatively, the objects
+ * may be marked by the mark procedure passed to GC_init_gcj_malloc.
+ */
+
+#ifndef MARK_DESCR_OFFSET
+#  define MARK_DESCR_OFFSET	sizeof(word)
+#endif
+	/* Gcj keeps GC descriptor as second word of vtable.	This	*/
+	/* probably needs to be adjusted for other clients.		*/
+	/* We currently assume that this offset is such that:		*/
+	/*	- all objects of this kind are large enough to have	*/
+	/*	  a value at that offset, and				*/
+	/* 	- it is not zero.					*/
+	/* These assumptions allow objects on the free list to be 	*/
+	/* marked normally.						*/
+
+#ifndef _GC_H
+#   include "gc.h"
+#endif
+
+/* The following allocators signal an out of memory condition with	*/
+/* return GC_oom_action();						*/
+/* The default GC_oom_action returns 0.					*/
+/* This functionality is currently restricted to the gcj allocators.	*/
+/* We may want to extend it to the others.				*/
+
+extern void * (*GC_oom_action)(void);
+
+/* The following function must be called before the gcj allocators	*/
+/* can be invoked.							*/
+/* mp_index and mp are the index and mark_proc (see gc_mark.h)		*/
+/* respectively for the allocated objects.  Mark_proc will be 		*/
+/* used to build the descriptor for objects allocated through the	*/
+/* debugging interface.  The mark_proc will be invoked on all such 	*/
+/* objects with an "environment" value of 1.  The client may chose	*/
+/* to use the same mark_proc for some of its generated mark descriptors.*/
+/* In that case, it should use a different "environment" value to	*/
+/* detect the presence or absence of the debug header.			*/
+
+/* the debugging interface.						*/
+/* Mp is really of type mark_proc, as defined in gc_mark.h.  We don't 	*/
+/* want to include that here for namespace pollution reasons.		*/
+extern void GC_init_gcj_malloc(int mp_index, void * /* really mark_proc */mp);
+
+/* Allocate an object, clear it, and store the pointer to the	*/
+/* type structure (vtable in gcj).				*/
+/* This adds a byte at the end of the object if GC_malloc would.*/
+extern void * GC_gcj_malloc(size_t lb, void * ptr_to_struct_containing_descr);
+/* The debug versions allocate such that the specified mark_proc	*/
+/* is always invoked.							*/
+extern void * GC_debug_gcj_malloc(size_t lb,
+				  void * ptr_to_struct_containing_descr,
+				  GC_EXTRA_PARAMS);
+
+/* Similar to the above, but the size is in words, and we don't	*/
+/* adjust it.  The size is assumed to be such that it can be 	*/
+/* allocated as a small object.					*/
+extern void * GC_gcj_fast_malloc(size_t lw,
+				 void * ptr_to_struct_containing_descr);
+extern void * GC_debug_gcj_fast_malloc(size_t lw,
+				 void * ptr_to_struct_containing_descr,
+				 GC_EXTRA_PARAMS);
+
+/* Similar to GC_gcj_malloc, but assumes that a pointer to near the	*/
+/* beginning of the resulting object is always maintained.		*/
+extern void * GC_gcj_malloc_ignore_off_page(size_t lb,
+				void * ptr_to_struct_containing_descr);
+
+# ifdef GC_DEBUG
+#   define GC_GCJ_MALLOC(s,d) GC_debug_gcj_malloc(s,d,GC_EXTRAS)
+#   define GC_GCJ_FAST_MALLOC(s,d) GC_debug_gcj_fast_malloc(s,d,GC_EXTRAS)
+#   define GC_GCJ_MALLOC_IGNORE_OFF_PAGE(s,d) GC_gcj_debug_malloc(s,d,GC_EXTRAS)
+# else
+#   define GC_GCJ_MALLOC(s,d) GC_gcj_malloc(s,d)
+#   define GC_GCJ_FAST_MALLOC(s,d) GC_gcj_fast_malloc(s,d)
+#   define GC_GCJ_MALLOC_IGNORE_OFF_PAGE(s,d) \
+	GC_gcj_debug_malloc_ignore_off_page(s,d)
+# endif
+
diff --git a/nursery.h b/include/gc_nursery.h
index d109ff09..d109ff09 100755
--- a/nursery.h
+++ b/include/gc_nursery.h
diff --git a/include/gc_typed.h b/include/gc_typed.h
index e4a6b947..2e0598f2 100644
--- a/include/gc_typed.h
+++ b/include/gc_typed.h
@@ -61,6 +61,7 @@ GC_API GC_PTR GC_malloc_explicitly_typed
 			GC_PROTO((size_t size_in_bytes, GC_descr d));
 		/* Allocate an object whose layout is described by d.	*/
 		/* The resulting object MAY NOT BE PASSED TO REALLOC.	*/
+		/* The returned object is cleared.			*/
 
 GC_API GC_PTR GC_malloc_explicitly_typed_ignore_off_page
                         GC_PROTO((size_t size_in_bytes, GC_descr d));
@@ -75,6 +76,7 @@ GC_API GC_PTR GC_calloc_explicitly_typed
   	/* 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.				*/
+	/* Returned object is cleared.				*/
 
 #ifdef GC_DEBUG
 #   define GC_MALLOC_EXPLICTLY_TYPED(bytes, d) GC_MALLOC(bytes)
diff --git a/include/new_gc_alloc.h b/include/new_gc_alloc.h
index 57713883..54b7bd44 100644
--- a/include/new_gc_alloc.h
+++ b/include/new_gc_alloc.h
@@ -318,12 +318,10 @@ class traceable_alloc_template {
 
 typedef traceable_alloc_template < 0 > traceable_alloc;
 
-#ifdef _SGI_SOURCE
-
 // We want to specialize simple_alloc so that it does the right thing
 // for all pointerfree types.  At the moment there is no portable way to
 // even approximate that.  The following approximation should work for
-// SGI compilers, and perhaps some others.
+// SGI compilers, and recent versions of g++.
 
 # define __GC_SPECIALIZE(T,alloc) \
 class simple_alloc<T, alloc> { \
@@ -451,6 +449,4 @@ __STL_END_NAMESPACE
 
 #endif /* __STL_USE_STD_ALLOCATORS */
 
-#endif /* _SGI_SOURCE */
-
 #endif /* GC_ALLOC_H */
diff --git a/include/private/gc_hdrs.h b/include/private/gc_hdrs.h
index 60dc2ad3..6966a9a1 100644
--- a/include/private/gc_hdrs.h
+++ b/include/private/gc_hdrs.h
@@ -24,6 +24,17 @@ typedef struct hblkhdr hdr;
  * 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.
+ *
+ * This defines HDR, GET_HDR, and SET_HDR, the main macros used to
+ * retrieve and set object headers.  We also define some variants to
+ * retrieve 2 unrelated headers in interleaved fashion.  This
+ * slightly improves scheduling.
+ *
+ * Since 5.0 alpha 5, we can also take advantage of a header lookup
+ * cache.  This is a locally declared direct mapped cache, used inside
+ * the marker.  The HC_GET_HDR and HC_GET_HDR2 macros use and maintain this
+ * cache.  Assuming we get reasonable hit rates, this shaves a few
+ * memory references from each pointer validation.
  */
 
 # if CPP_WORDSZ > 32
@@ -45,6 +56,127 @@ typedef struct hblkhdr hdr;
 # define TOP_SZ (1 << LOG_TOP_SZ)
 # define BOTTOM_SZ (1 << LOG_BOTTOM_SZ)
 
+#ifndef SMALL_CONFIG
+# define USE_HDR_CACHE
+#endif
+
+/* #define COUNT_HDR_CACHE_HITS  */
+
+extern hdr * GC_invalid_header; /* header for an imaginary block 	*/
+				/* containing no objects.		*/
+
+
+/* Check whether p and corresponding hhdr point to long or invalid	*/
+/* object.  If so, advance them	to					*/
+/* beginning of	block, or set hhdr to GC_invalid_header.		*/
+#define ADVANCE(p, hhdr, source) \
+            if (IS_FORWARDING_ADDR_OR_NIL(hhdr)) { \
+              p = GC_FIND_START(p, hhdr, (word)source); \
+              if (p == 0) { \
+		hhdr = GC_invalid_header; \
+	      } else { \
+                hhdr = GC_find_header(p); \
+	      } \
+    	    }
+
+#ifdef USE_HDR_CACHE
+
+# ifdef COUNT_HDR_CACHE_HITS
+    extern word GC_hdr_cache_hits;
+    extern word GC_hdr_cache_misses;
+#   define HC_HIT() ++GC_hdr_cache_hits
+#   define HC_MISS() ++GC_hdr_cache_misses
+# else
+#   define HC_HIT()
+#   define HC_MISS()
+# endif
+
+  typedef struct hce {
+    word block_addr;  	/* right shifted by LOG_HBLKSIZE */
+    hdr * hce_hdr;
+  } hdr_cache_entry;
+
+# define HDR_CACHE_SIZE 8  /* power of 2 */
+
+# define DECLARE_HDR_CACHE \
+	hdr_cache_entry hdr_cache[HDR_CACHE_SIZE]
+
+# define INIT_HDR_CACHE BZERO(hdr_cache, sizeof(hdr_cache));
+
+# define HCE(h) hdr_cache + (((word)(h) >> LOG_HBLKSIZE) & (HDR_CACHE_SIZE-1))
+
+# define HCE_VALID_FOR(hce,h) ((hce) -> block_addr == \
+				((word)(h) >> LOG_HBLKSIZE))
+
+# define HCE_HDR(h) ((hce) -> hce_hdr)
+
+
+/* Analogous to GET_HDR, except that in the case of large objects, it	*/
+/* Returns the header for the object beginning, and updates p.		*/
+/* Returns &GC_bad_header instead of 0.  All of this saves a branch	*/
+/* in the fast path.							*/
+# define HC_GET_HDR(p, hhdr, source) \
+	{ \
+	  hdr_cache_entry * hce = HCE(p); \
+	  if (HCE_VALID_FOR(hce, p)) { \
+	    HC_HIT(); \
+	    hhdr = hce -> hce_hdr; \
+	  } else { \
+	    HC_MISS(); \
+	    GET_HDR(p, hhdr); \
+	    ADVANCE(p, hhdr, source); \
+	    hce -> block_addr = (word)(p) >> LOG_HBLKSIZE; \
+	    hce -> hce_hdr = hhdr; \
+	  } \
+	}
+
+# define HC_GET_HDR2(p1, hhdr1, source1, p2, hhdr2, source2) \
+	{ \
+	  hdr_cache_entry * hce1 = HCE(p1); \
+	  hdr_cache_entry * hce2 = HCE(p2); \
+	  if (HCE_VALID_FOR(hce1, p1)) { \
+	    HC_HIT(); \
+	    hhdr1 = hce1 -> hce_hdr; \
+	  } else { \
+	    HC_MISS(); \
+	    GET_HDR(p1, hhdr1); \
+	    ADVANCE(p1, hhdr1, source1); \
+	    hce1 -> block_addr = (word)(p1) >> LOG_HBLKSIZE; \
+	    hce1 -> hce_hdr = hhdr1; \
+	  } \
+	  if (HCE_VALID_FOR(hce2, p2)) { \
+	    HC_HIT(); \
+	    hhdr2 = hce2 -> hce_hdr; \
+	  } else { \
+	    HC_MISS(); \
+	    GET_HDR(p2, hhdr2); \
+	    ADVANCE(p2, hhdr2, source2); \
+	    hce2 -> block_addr = (word)(p2) >> LOG_HBLKSIZE; \
+	    hce2 -> hce_hdr = hhdr2; \
+	  } \
+	}
+
+#else /* !USE_HDR_CACHE */
+
+# define DECLARE_HDR_CACHE
+
+# define INIT_HDR_CACHE
+
+# define HC_GET_HDR(p, hhdr, source) \
+	{ \
+	  GET_HDR(p, hhdr); \
+	  ADVANCE(p, hhdr, source); \
+	}
+
+# define HC_GET_HDR2(p1, hhdr1, source1, p2, hhdr2, source2) \
+	{ \
+	  GET_HDR2(p1, hhdr1, p2, hhdr2); \
+	  ADVANCE(p1, hhdr1, source1); \
+	  ADVANCE(p2, hhdr2, source2); \
+	}
+
+#endif
+
 typedef struct bi {
     hdr * index[BOTTOM_SZ];
 	/*
@@ -97,6 +229,8 @@ typedef struct bi {
 #   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))
+#   define GET_HDR2(p1, hhdr1, p2, hhdr2) \
+	{ GET_HDR(p1, hhdr1); GET_HDR(p2, hhdr2); }
 # else /* hash */
 /*  Hash function for tree top level */
 #   define TL_HASH(hi) ((hi) & (TOP_SZ - 1))
@@ -123,6 +257,40 @@ typedef struct bi {
 #   define SET_HDR(p, hhdr) { register hdr ** _ha; GET_HDR_ADDR(p, _ha); \
 			      *_ha = (hhdr); }
 #   define HDR(p) GC_find_header((ptr_t)(p))
+    /* And some interleaved versions for two pointers at once.  	*/
+    /* This hopefully helps scheduling on processors like IA64.		*/
+#   define GET_BI2(p1, bottom_indx1, p2, bottom_indx2) \
+	{ \
+	    register word hi1 = \
+	        (word)(p1) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE); \
+	    register word hi2 = \
+	        (word)(p2) >> (LOG_BOTTOM_SZ + LOG_HBLKSIZE); \
+	    register bottom_index * _bi1 = GC_top_index[TL_HASH(hi1)]; \
+	    register bottom_index * _bi2 = GC_top_index[TL_HASH(hi2)]; \
+	    \
+	    while (_bi1 -> key != hi1 && _bi1 != GC_all_nils) \
+	    	_bi1 = _bi1 -> hash_link; \
+	    while (_bi2 -> key != hi2 && _bi2 != GC_all_nils) \
+	    	_bi2 = _bi2 -> hash_link; \
+	    (bottom_indx1) = _bi1; \
+	    (bottom_indx2) = _bi2; \
+	}
+#   define GET_HDR_ADDR2(p1, ha1, p2, ha2) \
+	{ \
+	    register bottom_index * bi1; \
+	    register bottom_index * bi2; \
+	    \
+	    GET_BI2(p1, bi1, p2, bi2);	\
+	    (ha1) = &(HDR_FROM_BI(bi1, p1)); \
+	    (ha2) = &(HDR_FROM_BI(bi2, p2)); \
+	}
+#   define GET_HDR2(p1, hhdr1, p2, hhdr2) \
+	{ register hdr ** _ha1;  \
+	  register hdr ** _ha2;  \
+	  GET_HDR_ADDR2(p1, _ha1, p2, _ha2); \
+	  (hhdr1) = *_ha1;  \
+	  (hhdr2) = *_ha2;  \
+	}
 # endif
 			    
 /* Is the result a forwarding address to someplace closer to the	*/
diff --git a/include/private/gc_priv.h b/include/private/gc_priv.h
index 5ce52a7a..ce685875 100644
--- a/include/private/gc_priv.h
+++ b/include/private/gc_priv.h
@@ -1,6 +1,9 @@
 /* 
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
+ *
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -41,7 +44,7 @@
 typedef GC_word word;
 typedef GC_signed_word signed_word;
 
-# ifndef CONFIG_H
+# ifndef GCCONFIG_H
 #   include "gcconfig.h"
 # endif
 
@@ -64,21 +67,22 @@ typedef char * ptr_t;	/* A generic pointer to which we can add	*/
 #       include <stddef.h>
 #   endif
 #   define VOLATILE volatile
-#   define CONST const
 #else
 #   ifdef MSWIN32
 #   	include <stdlib.h>
 #   endif
 #   define VOLATILE
-#   define CONST
 #endif
 
+#define CONST GC_CONST
+
 #if 0 /* was once defined for AMIGA */
 #   define GC_FAR __far
 #else
 #   define GC_FAR
 #endif
 
+
 /*********************************/
 /*                               */
 /* Definitions for conservative  */
@@ -170,15 +174,6 @@ typedef char * ptr_t;	/* A generic pointer to which we can add	*/
 		    /* May save significant amounts of space for obj_map  */
 		    /* entries.						  */
 
-#ifndef OLD_BLOCK_ALLOC
-   /* Macros controlling large block allocation strategy.	*/
-#  define EXACT_FIRST  	/* Make a complete pass through the large object */
-			/* free list before splitting a block		 */
-#  define PRESERVE_LAST /* Do not divide last allocated heap segment	 */
-			/* unless we would otherwise need to expand the	 */
-			/* heap.					 */
-#endif
-
 /* ALIGN_DOUBLE requires MERGE_SIZES at present. */
 # if defined(ALIGN_DOUBLE) && !defined(MERGE_SIZES)
 #   define MERGE_SIZES
@@ -278,6 +273,13 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #   define MS_TIME_DIFF(a,b) ((double) (a.tv_sec - b.tv_sec) * 1000.0 \
                                + (double) (a.tv_usec - b.tv_usec) / 1000.0)
 #else /* !BSD_TIME */
+# ifdef MSWIN32
+#   include <windows.h>
+#   include <winbase.h>
+#   define CLOCK_TYPE DWORD
+#   define GET_TIME(x) x = GetTickCount()
+#   define MS_TIME_DIFF(a,b) ((long)((a)-(b)))
+# else /* !MSWIN32, !BSD_TIME */
 #   include <time.h>
 #   if !defined(__STDC__) && defined(SPARC) && defined(SUNOS4)
       clock_t clock();	/* Not in time.h, where it belongs	*/
@@ -303,6 +305,7 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #   define GET_TIME(x) x = clock()
 #   define MS_TIME_DIFF(a,b) ((unsigned long) \
 		(1000.0*(double)((a)-(b))/(double)CLOCKS_PER_SEC))
+# endif /* !MSWIN32 */
 #endif /* !BSD_TIME */
 
 /* We use bzero and bcopy internally.  They may not be available.	*/
@@ -434,9 +437,12 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #    define LOCK() mutex_lock(&GC_allocate_ml);
 #    define UNLOCK() mutex_unlock(&GC_allocate_ml);
 #  endif
-#  ifdef LINUX_THREADS
+#  if defined(LINUX_THREADS) 
+#   if defined(I386)|| defined(POWERPC) || defined(ALPHA) || defined(IA64) \
+    || defined(M68K)
 #    include <pthread.h>
-#    ifdef __i386__
+#    define USE_SPIN_LOCK
+#    if defined(I386)
        inline static int GC_test_and_set(volatile unsigned int *addr) {
 	  int oldval;
 	  /* Note: the "xchg" instruction does not need a "lock" prefix */
@@ -445,15 +451,107 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 		: "0"(1), "m"(*(addr)));
 	  return oldval;
        }
-#    else
-       -- > Need implementation of GC_test_and_set()
 #    endif
-#    define GC_clear(addr) (*(addr) = 0)
+#    if defined(IA64)
+       inline static int GC_test_and_set(volatile unsigned int *addr) {
+	  int oldval;
+	  __asm__ __volatile__("xchg4 %0=%1,%2"
+		: "=r"(oldval), "=m"(*addr)
+		: "r"(1), "1"(*addr));
+	  return oldval;
+       }
+       inline static void GC_clear(volatile unsigned int *addr) {
+	 __asm__ __volatile__("st4.rel %0=r0" : "=m" (*addr));
+       }
+#      define GC_CLEAR_DEFINED
+#    endif
+#    ifdef M68K
+       /* Contributed by Tony Mantler.  I'm not sure how well it was	*/
+       /* tested.							*/
+       inline static int GC_test_and_set(volatile unsigned int *addr) {
+          char oldval; /* this must be no longer than 8 bits */
+
+          /* The return value is semi-phony. */
+          /* 'tas' sets bit 7 while the return */
+          /* value pretends bit 0 was set */
+          __asm__ __volatile__(
+                 "tas %1@; sne %0; negb %0"
+                 : "=d" (oldval)
+                 : "a" (addr));
+          return oldval;
+       }
+#    endif
+#    if defined(POWERPC)
+      inline static int GC_test_and_set(volatile unsigned int *addr) {
+        int oldval;
+        int temp = 1; // locked value
+
+        __asm__ __volatile__(
+               "1:\tlwarx %0,0,%3\n"   // load and reserve
+               "\tcmpwi %0, 0\n"       // if load is
+               "\tbne 2f\n"            //   non-zero, return already set
+               "\tstwcx. %2,0,%1\n"    // else store conditional
+               "\tbne- 1b\n"           // retry if lost reservation
+               "2:\t\n"                // oldval is zero if we set
+              : "=&r"(oldval), "=p"(addr)
+              : "r"(temp), "1"(addr)
+              : "memory");
+        return (int)oldval;
+      }
+      inline static void GC_clear(volatile unsigned int *addr) {
+	 __asm__ __volatile__("eieio");
+         *(addr) = 0;
+      }
+#     define GC_CLEAR_DEFINED
+#    endif
+#    ifdef ALPHA
+      inline static int GC_test_and_set(volatile unsigned int * addr)
+      {
+        unsigned long oldvalue;
+        unsigned long temp;
+
+        __asm__ __volatile__(
+                             "1:     ldl_l %0,%1\n"
+                             "       and %0,%3,%2\n"
+                             "       bne %2,2f\n"
+                             "       xor %0,%3,%0\n"
+                             "       stl_c %0,%1\n"
+                             "       beq %0,3f\n"
+                             "       mb\n"
+                             "2:\n"
+                             ".section .text2,\"ax\"\n"
+                             "3:     br 1b\n"
+                             ".previous"
+                             :"=&r" (temp), "=m" (*addr), "=&r" (oldvalue)
+                             :"Ir" (1), "m" (*addr));
+
+        return oldvalue;
+      }
+      /* Should probably also define GC_clear, since it needs	*/
+      /* a memory barrier ??					*/
+#    endif /* ALPHA */
+#    ifdef ARM32
+      inline static int GC_test_and_set(volatile unsigned int *addr) {
+        int oldval;
+        /* SWP on ARM is very similar to XCHG on x86.  Doesn't lock the
+         * bus because there are no SMP ARM machines.  If/when there are,
+         * this code will likely need to be updated. */
+        /* See linuxthreads/sysdeps/arm/pt-machine.h in glibc-2.1 */
+        __asm__ __volatile__("swp %0, %1, [%2]"
+      			     : "=r"(oldval)
+      			     : "r"(1), "r"(addr));
+        return oldval;
+      }
+#    endif
+#    ifndef GC_CLEAR_DEFINED
+       inline static void GC_clear(volatile unsigned int *addr) {
+	  /* Try to discourage gcc from moving anything past this. */
+	  __asm__ __volatile__(" ");
+          *(addr) = 0;
+       }
+#    endif
 
      extern volatile unsigned int GC_allocate_lock;
-	/* This is not a mutex because mutexes that obey the (optional)     */
-	/* POSIX scheduling rules are subject to convoys in high contention */
-	/* applications.  This is basically a spin lock.		    */
      extern pthread_t GC_lock_holder;
      extern void GC_lock(void);
 	/* Allocation lock holder.  Only set if acquired by client through */
@@ -462,31 +560,48 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #    define NO_THREAD (pthread_t)(-1)
 #    define UNSET_LOCK_HOLDER() GC_lock_holder = NO_THREAD
 #    define I_HOLD_LOCK() (pthread_equal(GC_lock_holder, pthread_self()))
-#    ifdef UNDEFINED
-#    	define LOCK() pthread_mutex_lock(&GC_allocate_ml)
-#    	define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
-#    else
-#	define LOCK() \
+#    define LOCK() \
 		{ if (GC_test_and_set(&GC_allocate_lock)) GC_lock(); }
-#	define UNLOCK() \
+#    define UNLOCK() \
 		GC_clear(&GC_allocate_lock)
-#    endif
-     extern GC_bool GC_collecting;
+     extern VOLATILE GC_bool GC_collecting;
 #    define ENTER_GC() \
 		{ \
 		    GC_collecting = 1; \
 		}
 #    define EXIT_GC() GC_collecting = 0;
+#   else /* LINUX_THREADS on hardware for which we don't know how	*/
+	 /* to do test and set.						*/
+#    include <pthread.h>
+     extern pthread_mutex_t GC_allocate_ml;
+#    define LOCK() pthread_mutex_lock(&GC_allocate_ml)
+#    define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
+#   endif
 #  endif /* LINUX_THREADS */
-#  if defined(IRIX_THREADS) || defined(IRIX_JDK_THREADS)
+#  if defined(HPUX_THREADS)
 #    include <pthread.h>
-#    include <mutex.h>
+     extern pthread_mutex_t GC_allocate_ml;
+#    define LOCK() pthread_mutex_lock(&GC_allocate_ml)
+#    define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
+#  endif
+#  if defined(IRIX_THREADS) || defined(IRIX_JDK_THREADS) 
+     /* This may also eventually be appropriate for HPUX_THREADS */
+#    include <pthread.h>
+#    ifndef HPUX_THREADS
+	/* This probably should never be included, but I can't test	*/
+	/* on Irix anymore.						*/
+#       include <mutex.h>
+#    endif
 
-#    if __mips < 3 || !(defined (_ABIN32) || defined(_ABI64)) \
+#    ifndef HPUX_THREADS
+#      if __mips < 3 || !(defined (_ABIN32) || defined(_ABI64)) \
 	|| !defined(_COMPILER_VERSION) || _COMPILER_VERSION < 700
 #        define GC_test_and_set(addr, v) test_and_set(addr,v)
-#    else
+#      else
 #	 define GC_test_and_set(addr, v) __test_and_set(addr,v)
+#      endif
+#    else
+       /* I couldn't find a way to do this inline on HP/UX	*/
 #    endif
      extern unsigned long GC_allocate_lock;
 	/* This is not a mutex because mutexes that obey the (optional) 	*/
@@ -500,15 +615,17 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 #    define NO_THREAD (pthread_t)(-1)
 #    define UNSET_LOCK_HOLDER() GC_lock_holder = NO_THREAD
 #    define I_HOLD_LOCK() (pthread_equal(GC_lock_holder, pthread_self()))
-#    ifdef UNDEFINED
-#    	define LOCK() pthread_mutex_lock(&GC_allocate_ml)
-#    	define UNLOCK() pthread_mutex_unlock(&GC_allocate_ml)
+#    ifdef HPUX_THREADS
+#      define LOCK() { if (!GC_test_and_clear(&GC_allocate_lock)) GC_lock(); }
+       /* The following is INCORRECT, since the memory model is too weak. */
+#      define UNLOCK() { GC_noop1(&GC_allocate_lock); \
+			*(volatile unsigned long *)(&GC_allocate_lock) = 1; }
 #    else
-#	define LOCK() { if (GC_test_and_set(&GC_allocate_lock, 1)) GC_lock(); }
-#       if __mips >= 3 && (defined (_ABIN32) || defined(_ABI64)) \
+#      define LOCK() { if (GC_test_and_set(&GC_allocate_lock, 1)) GC_lock(); }
+#      if __mips >= 3 && (defined (_ABIN32) || defined(_ABI64)) \
 	   && defined(_COMPILER_VERSION) && _COMPILER_VERSION >= 700
 #	    define UNLOCK() __lock_release(&GC_allocate_lock)
-#	else
+#      else
 	    /* The function call in the following should prevent the	*/
 	    /* compiler from moving assignments to below the UNLOCK.	*/
 	    /* This is probably not necessary for ucode or gcc 2.8.	*/
@@ -516,9 +633,9 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 	    /* versions.						*/
 #           define UNLOCK() { GC_noop1(&GC_allocate_lock); \
 			*(volatile unsigned long *)(&GC_allocate_lock) = 0; }
-#	endif
+#      endif
 #    endif
-     extern GC_bool GC_collecting;
+     extern VOLATILE GC_bool GC_collecting;
 #    define ENTER_GC() \
 		{ \
 		    GC_collecting = 1; \
@@ -607,7 +724,7 @@ void GC_print_callers (/* struct callinfo info[NFRAMES] */);
 # else
 #   if defined(SOLARIS_THREADS) || defined(WIN32_THREADS) \
 	|| defined(IRIX_THREADS) || defined(LINUX_THREADS) \
-	|| defined(IRIX_JDK_THREADS)
+	|| defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
       void GC_stop_world();
       void GC_start_world();
 #     define STOP_WORLD() GC_stop_world()
@@ -857,6 +974,9 @@ struct hblkhdr {
 			    /* object starting at the ith word (header      */
 			    /* INCLUDED) in the heap block.                 */
 			    /* The lsb of word 0 is numbered 0.		    */
+			    /* Unused bits are invalid, and are 	    */
+			    /* occasionally set, e.g for uncollectable	    */
+			    /* objects.					    */
 };
 
 /*  heap block body */
@@ -891,8 +1011,10 @@ struct hblk {
 /* The type of mark procedures.  This really belongs in gc_mark.h.	*/
 /* But we put it here, so that we can avoid scanning the mark proc	*/
 /* table.								*/
-typedef struct ms_entry * (*mark_proc)(/* word * addr, mark_stack_ptr,
-					  mark_stack_limit, env */);
+typedef struct ms_entry * (*mark_proc)(/* word * addr,
+					  struct ms_entry *mark_stack_ptr,
+					  struct ms_entry *mark_stack_limit,
+					  word env */);
 # define LOG_MAX_MARK_PROCS 6
 # define MAX_MARK_PROCS (1 << LOG_MAX_MARK_PROCS)
 
@@ -969,6 +1091,7 @@ struct roots {
 struct _GC_arrays {
   word _heapsize;
   word _max_heapsize;
+  word _requested_heapsize;	/* Heap size due to explicit expansion */
   ptr_t _last_heap_addr;
   ptr_t _prev_heap_addr;
   word _large_free_bytes;
@@ -993,6 +1116,10 @@ struct _GC_arrays {
   word _mem_freed;
   	/* Number of explicitly deallocated words of memory	*/
   	/* since last collection.				*/
+  ptr_t _scratch_end_ptr;
+  ptr_t _scratch_last_end_ptr;
+	/* Used by headers.c, and can easily appear to point to	*/
+	/* heap.						*/
   mark_proc _mark_procs[MAX_MARK_PROCS];
   	/* Table of user-defined mark procedures.  There is	*/
 	/* a small number of these, which can be referenced	*/
@@ -1157,9 +1284,12 @@ GC_API GC_FAR struct _GC_arrays GC_arrays;
 # define GC_words_finalized GC_arrays._words_finalized
 # define GC_non_gc_bytes_at_gc GC_arrays._non_gc_bytes_at_gc
 # define GC_mem_freed GC_arrays._mem_freed
+# define GC_scratch_end_ptr GC_arrays._scratch_end_ptr
+# define GC_scratch_last_end_ptr GC_arrays._scratch_last_end_ptr
 # define GC_mark_procs GC_arrays._mark_procs
 # define GC_heapsize GC_arrays._heapsize
 # define GC_max_heapsize GC_arrays._max_heapsize
+# define GC_requested_heapsize GC_arrays._requested_heapsize
 # define GC_words_allocd_before_gc GC_arrays._words_allocd_before_gc
 # define GC_heap_sects GC_arrays._heap_sects
 # define GC_last_stack GC_arrays._last_stack
@@ -1194,6 +1324,8 @@ GC_API GC_FAR struct _GC_arrays GC_arrays;
 # define beginGC_arrays ((ptr_t)(&GC_arrays))
 # define endGC_arrays (((ptr_t)(&GC_arrays)) + (sizeof GC_arrays))
 
+#define USED_HEAP_SIZE (GC_heapsize - GC_large_free_bytes)
+
 /* Object kinds: */
 # define MAXOBJKINDS 16
 
@@ -1273,7 +1405,7 @@ extern GC_bool GC_objects_are_marked;	/* There are marked objects in  */
   extern GC_bool GC_incremental;
 			/* Using incremental/generational collection. */
 #else
-# define GC_incremental TRUE
+# define GC_incremental FALSE
 			/* Hopefully allow optimizer to remove some code. */
 #endif
 
@@ -1326,10 +1458,7 @@ extern ptr_t GC_greatest_plausible_heap_addr;
 ptr_t GC_approx_sp();
 
 GC_bool GC_should_collect();
-#ifdef PRESERVE_LAST
-    GC_bool GC_in_last_heap_sect(/* ptr_t */);
-	/* In last added heap section?  If so, avoid breaking up.	*/
-#endif
+
 void GC_apply_to_all_blocks(/*fn, client_data*/);
 			/* Invoke fn(hbp, client_data) for each 	*/
 			/* allocated heap block.			*/
@@ -1411,8 +1540,14 @@ extern void (*GC_start_call_back)(/* void */);
 			/* lock held.					*/
 			/* 0 by default.				*/
 void GC_push_regs();	/* Push register contents onto mark stack.	*/
+			/* If NURSERY is defined, the default push	*/
+			/* action can be overridden with GC_push_proc	*/
 void GC_remark();	/* Mark from all marked objects.  Used	*/
 		 	/* only if we had to drop something.	*/
+
+# ifdef NURSERY
+    extern void (*GC_push_proc)(ptr_t);
+# endif
 # if defined(MSWIN32)
   void __cdecl GC_push_one();
 # else
@@ -1600,9 +1735,10 @@ ptr_t GC_allocobj(/* sz_inn_words, kind */);
 				/* head.				*/
 
 void GC_init_headers();
-GC_bool GC_install_header(/*h*/);
+struct hblkhdr * GC_install_header(/*h*/);
 				/* Install a header for block h.	*/
-				/* Return FALSE on failure.		*/
+				/* Return 0 on failure, or the header	*/
+				/* otherwise.				*/
 GC_bool GC_install_counts(/*h, sz*/);
 				/* Set up forwarding counts for block	*/
 				/* h of size sz.			*/
diff --git a/include/private/gcconfig.h b/include/private/gcconfig.h
index c9017d37..221983ca 100644
--- a/include/private/gcconfig.h
+++ b/include/private/gcconfig.h
@@ -13,9 +13,9 @@
  * modified is included with the above copyright notice.
  */
  
-#ifndef CONFIG_H
+#ifndef GCCONFIG_H
 
-# define CONFIG_H
+# define GCCONFIG_H
 
 /* Machine dependent parameters.  Some tuning parameters can be found	*/
 /* near the top of gc_private.h.					*/
@@ -53,6 +53,11 @@
 #    define NETBSD
 #    define mach_type_known
 # endif
+# if defined(__NetBSD__) && defined(arm32)
+#    define ARM32
+#    define NETBSD
+#    define mach_type_known
+# endif
 # if defined(vax)
 #    define VAX
 #    ifdef ultrix
@@ -64,15 +69,18 @@
 # endif
 # if defined(mips) || defined(__mips)
 #    define MIPS
-#    if defined(ultrix) || defined(__ultrix) || defined(__NetBSD__)
-#	define ULTRIX
-#    else
-#	if defined(_SYSTYPE_SVR4) || defined(SYSTYPE_SVR4) || defined(__SYSTYPE_SVR4__)
-#	  define IRIX5   /* or IRIX 6.X */
-#	else
-#	  define RISCOS  /* or IRIX 4.X */
-#	endif
-#    endif
+#    if !defined(LINUX)
+#      if defined(ultrix) || defined(__ultrix) || defined(__NetBSD__)
+#	 define ULTRIX
+#      else
+#	 if defined(_SYSTYPE_SVR4) || defined(SYSTYPE_SVR4) \
+	    || defined(__SYSTYPE_SVR4__)
+#	   define IRIX5   /* or IRIX 6.X */
+#	 else
+#	   define RISCOS  /* or IRIX 4.X */
+#	 endif
+#      endif
+#    endif /* !LINUX */
 #    define mach_type_known
 # endif
 # if defined(sequent) && defined(i386)
@@ -130,15 +138,22 @@
 #   define SYSV
 #   define mach_type_known
 # endif
-# if defined(_PA_RISC1_0) || defined(_PA_RISC1_1) \
+# if defined(_PA_RISC1_0) || defined(_PA_RISC1_1) || defined(_PA_RISC2_0) \
      || defined(hppa) || defined(__hppa__)
 #   define HP_PA
+#   ifndef LINUX
+#     define HPUX
+#   endif
 #   define mach_type_known
 # endif
 # if defined(LINUX) && (defined(i386) || defined(__i386__))
 #    define I386
 #    define mach_type_known
 # endif
+# if defined(LINUX) && (defined(__ia64__) || defined(__ia64))
+#    define IA64
+#    define mach_type_known
+# endif
 # if defined(LINUX) && defined(powerpc)
 #    define POWERPC
 #    define mach_type_known
@@ -147,10 +162,14 @@
 #    define M68K
 #    define mach_type_known
 # endif
-# if defined(LINUX) && defined(sparc)
+# if defined(LINUX) && (defined(sparc) || defined(__sparc__))
 #    define SPARC
 #    define mach_type_known
 # endif
+# if defined(LINUX) && defined(arm)
+#    define ARM32
+#    define mach_type_known
+# endif
 # if defined(__alpha) || defined(__alpha__)
 #   define ALPHA
 #   if !defined(LINUX)
@@ -243,6 +262,11 @@
 #   define CYGWIN32
 #   define mach_type_known
 # endif
+# if defined(__MINGW32__)
+#   define I386
+#   define MSWIN32
+#   define mach_type_known
+# endif
 # if defined(__BORLANDC__)
 #   define I386
 #   define MSWIN32
@@ -253,6 +277,10 @@
 #   define UTS4
 #   define mach_type_known
 # endif
+# if defined(__pj__)
+#   define PJ
+#   define mach_type_known
+# endif
 /* Ivan Demakov */
 # if defined(__WATCOMC__) && defined(__386__)
 #   define I386
@@ -307,6 +335,9 @@
 		    /* 		        (CX_UX and DGUX)		*/
 		    /* 		   S370	      ==> 370-like machine	*/
 		    /* 			running Amdahl UTS4		*/
+		    /* 		   ARM32      ==> Intel StrongARM	*/
+		    /* 		   IA64	      ==> Intel IA64		*/
+		    /*				  (e.g. Itanium)	*/
 
 
 /*
@@ -355,6 +386,11 @@
  *		to the nearest plausible page boundary, and use that instead
  *		of STACKBOTTOM.
  *
+ * Gustavo Rodriguez-Rivera points out that on most (all?) Unix machines,
+ * the value of environ is a pointer that can serve as STACKBOTTOM.
+ * I expect that HEURISTIC2 can be replaced by this approach, which
+ * interferes far less with debugging. 
+ *
  * 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:
@@ -392,6 +428,15 @@
  *
  * An architecture may define DYNAMIC_LOADING if dynamic_load.c
  * defined GC_register_dynamic_libraries() for the architecture.
+ *
+ * An architecture may define PREFETCH(x) to preload the cache with *x.
+ * This defaults to a no-op.
+ *
+ * PREFETCH_FOR_WRITE(x) is used if *x is about to be written.
+ *
+ * An architecture may also define CLEAR_DOUBLE(x) to be a fast way to
+ * clear the two words at GC_malloc-aligned address x.  By default,
+ * word stores of 0 are used instead.
  */
 
 
@@ -513,10 +558,11 @@
 				/* This was 2, but that didn't sound right. */
 #     define OS_TYPE "LINUX"
 #     define HEURISTIC1
+#     define DYNAMIC_LOADING
 #     undef STACK_GRAN
 #     define STACK_GRAN 0x10000000
 	/* Stack usually starts at 0x80000000 */
-#     define DATASTART GC_data_start
+#     define LINUX_DATA_START
       extern int _end;
 #     define DATAEND (&_end)
 #   endif
@@ -615,8 +661,8 @@
 #   ifdef LINUX
 #     define OS_TYPE "LINUX"
 #     ifdef __ELF__
-#         define DATASTART GC_data_start
-#         define DYNAMIC_LOADING
+#       define LINUX_DATA_START
+#       define DYNAMIC_LOADING
 #     else
           Linux Sparc non elf ?
 #     endif
@@ -684,13 +730,16 @@
 #   endif
 #   ifdef LINUX
 #	define OS_TYPE "LINUX"
-#       define HEURISTIC1
-#       undef STACK_GRAN
-#       define STACK_GRAN 0x10000000
-	/* STACKBOTTOM is usually 0xc0000000, but this changes with	*/
-	/* different kernel configurations.  In particular, systems	*/
-	/* with 2GB physical memory will usually move the user		*/
-	/* address space limit, and hence initial SP to 0x80000000.	*/
+#       define LINUX_STACKBOTTOM
+#	if 0
+#	  define HEURISTIC1
+#         undef STACK_GRAN
+#         define STACK_GRAN 0x10000000
+	  /* STACKBOTTOM is usually 0xc0000000, but this changes with	*/
+	  /* different kernel configurations.  In particular, systems	*/
+	  /* with 2GB physical memory will usually move the user	*/
+	  /* address space limit, and hence initial SP to 0x80000000.	*/
+#       endif
 #       if !defined(LINUX_THREADS) || !defined(REDIRECT_MALLOC)
 #	    define MPROTECT_VDB
 #	else
@@ -706,8 +755,7 @@
 #	     endif
 #	     include <features.h>
 #	     if defined(__GLIBC__) && __GLIBC__ >= 2
-		 extern int __data_start;
-#		 define DATASTART ((ptr_t)(&__data_start))
+#		 define LINUX_DATA_START
 #	     else
      	         extern char **__environ;
 #                define DATASTART ((ptr_t)(&__environ))
@@ -726,6 +774,26 @@
 	     extern int etext;
 #            define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
 #       endif
+#	ifdef USE_I686_PREFETCH
+#	  define PREFETCH(x) \
+	    __asm__ __volatile__ ("	prefetchnta	%0": : "m"(*(char *)(x)))
+	    /* Empirically prefetcht0 is much more effective at reducing	*/
+	    /* cache miss stalls for the targetted load instructions.  But it	*/
+	    /* seems to interfere enough with other cache traffic that the net	*/
+	    /* result is worse than prefetchnta.				*/
+#         if 0 
+	    /* Using prefetches for write seems to have a slight negative	*/
+	    /* impact on performance, at least for a PIII/500.			*/
+#	    define PREFETCH_FOR_WRITE(x) \
+	      __asm__ __volatile__ ("	prefetcht0	%0": : "m"(*(char *)(x)))
+#	  endif
+#	endif
+#	ifdef USE_3DNOW_PREFETCH
+#	  define PREFETCH(x) \
+	    __asm__ __volatile__ ("	prefetch	%0": : "m"(*(char *)(x)))
+#	  define PREFETCH_FOR_WRITE(x) 
+	    __asm__ __volatile__ ("	prefetchw	%0": : "m"(*(char *)(x)))
+#	endif
 #   endif
 #   ifdef CYGWIN32
 #       define OS_TYPE "CYGWIN32"
@@ -836,35 +904,48 @@
 
 # ifdef MIPS
 #   define MACH_TYPE "MIPS"
-#   ifndef IRIX5
-#     define DATASTART (ptr_t)0x10000000
-			      /* Could probably be slightly higher since */
-			      /* startup code allocates lots of stuff.   */
-#   else
-      extern int _fdata;
-#     define DATASTART ((ptr_t)(&_fdata))
-#     ifdef USE_MMAP
-#         define HEAP_START (ptr_t)0x30000000
-#     else
-#	  define HEAP_START DATASTART
-#     endif
-			      /* Lowest plausible heap address.		*/
-			      /* In the MMAP case, we map there.	*/
-			      /* In either case it is used to identify	*/
-			      /* heap sections so they're not 		*/
-			      /* considered as roots.			*/
-#   endif /* IRIX5 */
-#   define HEURISTIC2
 /* #   define STACKBOTTOM ((ptr_t)0x7fff8000)  sometimes also works.  */
+#   ifdef LINUX
+      /* This was developed for a linuxce style platform.  Probably	*/
+      /* needs to be tweaked for workstation class machines.		*/
+#     define OS_TYPE "LINUX"
+      extern int __data_start;
+#     define DATASTART ((ptr_t)(&__data_start))
+#     define ALIGNMENT 4
+#     define USE_GENERIC_PUSH_REGS 1
+#     define STACKBOTTOM 0x80000000
+	/* In many cases, this should probably use LINUX_STACKBOTTOM 	*/
+	/* instead. But some kernel versions seem to give the wrong	*/
+	/* value from /proc.						*/
+#   endif /* Linux */
 #   ifdef ULTRIX
+#	define HEURISTIC2
+#       define DATASTART (ptr_t)0x10000000
+			      /* Could probably be slightly higher since */
+			      /* startup code allocates lots of stuff.   */
 #	define OS_TYPE "ULTRIX"
 #       define ALIGNMENT 4
 #   endif
 #   ifdef RISCOS
+#	define HEURISTIC2
+#       define DATASTART (ptr_t)0x10000000
 #	define OS_TYPE "RISCOS"
 #   	define ALIGNMENT 4  /* Required by hardware */
 #   endif
 #   ifdef IRIX5
+#	define HEURISTIC2
+        extern int _fdata;
+#       define DATASTART ((ptr_t)(&_fdata))
+#       ifdef USE_MMAP
+#         define HEAP_START (ptr_t)0x30000000
+#       else
+#	  define HEAP_START DATASTART
+#       endif
+			      /* Lowest plausible heap address.		*/
+			      /* In the MMAP case, we map there.	*/
+			      /* In either case it is used to identify	*/
+			      /* heap sections so they're not 		*/
+			      /* considered as roots.			*/
 #	define OS_TYPE "IRIX5"
 #       define MPROTECT_VDB
 #       ifdef _MIPS_SZPTR
@@ -892,9 +973,17 @@
 # endif
 
 # ifdef HP_PA
+    /* OS is assumed to be HP/UX	*/
 #   define MACH_TYPE "HP_PA"
-#   define ALIGNMENT 4
-#   define ALIGN_DOUBLE
+#   define OS_TYPE "HPUX"
+#   ifdef __LP64__
+#     define CPP_WORDSZ 64
+#     define ALIGNMENT 8
+#   else
+#     define CPP_WORDSZ 32
+#     define ALIGNMENT 4
+#     define ALIGN_DOUBLE
+#   endif
     extern int __data_start;
 #   define DATASTART ((ptr_t)(&__data_start))
 #   if 0
@@ -907,22 +996,33 @@
 	/* this.)							*/
 #       define STACKBOTTOM ((ptr_t) 0x7b033000)  /* from /etc/conf/h/param.h */
 #   else
-#       define HEURISTIC2
+	/* Gustavo Rodriguez-Rivera suggested changing HEURISTIC2	*/
+	/* to this.  We'll probably do this on other platforms, too.	*/
+	/* For now I'll use it where I can test it.			*/
+	extern char ** environ;
+#       define STACKBOTTOM ((ptr_t)environ)
 #   endif
 #   define STACK_GROWS_UP
 #   define DYNAMIC_LOADING
+#   ifndef HPUX_THREADS
+#     define MPROTECT_VDB
+#   endif
 #   include <unistd.h>
 #   define GETPAGESIZE() sysconf(_SC_PAGE_SIZE)
-	/* They misspelled the Posix macro?	*/
 # endif
 
 # ifdef ALPHA
 #   define MACH_TYPE "ALPHA"
 #   define ALIGNMENT 8
+#   define USE_GENERIC_PUSH_REGS
+	/* Gcc and probably the DEC/Compaq compiler spill pointers to preserved	*/
+	/* fp registers in some cases when the target is a 21264.  The assembly	*/
+	/* code doesn't handle that yet, and version dependencies make that a	*/
+	/* bit tricky.  Do the easy thing for now.				*/
 #   ifdef OSF1
 #	define OS_TYPE "OSF1"
 #   	define DATASTART ((ptr_t) 0x140000000)
-	extern _end;
+	extern int _end;
 #   	define DATAEND ((ptr_t) &_end)
 #   	define HEURISTIC2
 	/* Normally HEURISTIC2 is too conervative, since		*/
@@ -939,13 +1039,9 @@
 #       define CPP_WORDSZ 64
 #       define STACKBOTTOM ((ptr_t) 0x120000000)
 #       ifdef __ELF__
-#   	  if 0
-	    /* __data_start apparently disappeared in some recent releases. */
-            extern int __data_start;
-#           define DATASTART &__data_start
-#	  endif
-#         define DATASTART GC_data_start
+#	  define LINUX_DATA_START
 #         define DYNAMIC_LOADING
+	  /* This doesn't work if the collector is in a dynamic library. */
 #       else
 #           define DATASTART ((ptr_t) 0x140000000)
 #       endif
@@ -957,6 +1053,48 @@
 #   endif
 # endif
 
+# ifdef IA64
+#   define MACH_TYPE "IA64"
+#   define ALIGN_DOUBLE
+	/* Requires 16 byte alignment for malloc */
+#   define ALIGNMENT 8
+#   define USE_GENERIC_PUSH_REGS
+	/* We need to get preserved registers in addition to register windows.	*/
+	/* That's easiest to do with setjmp.					*/
+#   ifdef HPUX
+	--> needs work
+#   endif
+#   ifdef LINUX
+#       define OS_TYPE "LINUX"
+#       define CPP_WORDSZ 64
+	/* This should really be done through /proc, but that	*/
+	/* requires we run on an IA64 kernel.			*/
+#       define STACKBOTTOM ((ptr_t) 0xa000000000000000l)
+	/* We also need the base address of the register stack	*/
+	/* backing store.  There is probably a better way to	*/
+	/* get that, too ...					*/
+#	define BACKING_STORE_BASE ((ptr_t) 0x9fffffff80000000l)
+#	if 1
+#	    define SEARCH_FOR_DATA_START
+#	    define DATASTART GC_data_start
+#	else
+	    extern int data_start;
+#	    define DATASTART ((ptr_t)(&data_start))
+#	endif
+#       define DYNAMIC_LOADING
+#	define MPROTECT_VDB
+		/* Requires Linux 2.3.47 or later.	*/
+	extern int _end;
+#	define DATAEND (&_end)
+#	define PREFETCH(x) \
+	  __asm__ ("	lfetch	[%0]": : "r"((void *)(x)))
+#	define PREFETCH_FOR_WRITE(x) \
+	  __asm__ ("	lfetch.excl	[%0]": : "r"((void *)(x)))
+#	define CLEAR_DOUBLE(x) \
+	  __asm__ ("	stf.spill	[%0]=f0": : "r"((void *)(x)))
+#   endif
+# endif
+
 # ifdef M88K
 #   define MACH_TYPE "M88K"
 #   define ALIGNMENT 4
@@ -987,6 +1125,69 @@
 #	define HEURISTIC2
 # endif
 
+# if defined(PJ)
+#   define ALIGNMENT 4
+    extern int _etext;
+#   define DATASTART ((ptr_t)(&_etext))
+#   define HEURISTIC1
+# endif
+
+# ifdef ARM32
+#   define CPP_WORDSZ 32
+#   define MACH_TYPE "ARM32"
+#   define ALIGNMENT 4
+#   ifdef NETBSD
+#       define OS_TYPE "NETBSD"
+#       define HEURISTIC2
+        extern char etext;
+#       define DATASTART ((ptr_t)(&etext))
+#       define USE_GENERIC_PUSH_REGS
+#   endif
+#   ifdef LINUX
+#       define OS_TYPE "LINUX"
+#       define HEURISTIC1
+#       undef STACK_GRAN
+#       define STACK_GRAN 0x10000000
+#       define USE_GENERIC_PUSH_REGS
+#       ifdef __ELF__
+#            define DYNAMIC_LOADING
+#	     include <features.h>
+#	     if defined(__GLIBC__) && __GLIBC__ >= 2
+#		 define LINUX_DATA_START
+#	     else
+     	         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.        */
+			      /* We could use _etext instead, but that  */
+			      /* would include .rodata, which may       */
+			      /* contain large read-only data tables    */
+			      /* that we'd rather not scan.		*/
+#	     endif
+	     extern int _end;
+#	     define DATAEND (&_end)
+#	else
+	     extern int etext;
+#            define DATASTART ((ptr_t)((((word) (&etext)) + 0xfff) & ~0xfff))
+#       endif
+#   endif
+#endif
+
+#ifdef LINUX_DATA_START
+    /* Some Linux distributions arrange to define __data_start.  Some	*/
+    /* define data_start as a weak symbol.  The latter is technically	*/
+    /* broken, since the user program may define data_start, in which	*/
+    /* case we lose.  Nonetheless, we try both, prefering __data_start.	*/
+    /* We assume gcc.	*/
+#   pragma weak __data_start
+    extern int __data_start;
+#   pragma weak data_start
+    extern int data_start;
+#   define DATASTART ((ptr_t)(&__data_start != 0? &__data_start : &data_start))
+#endif
+
 # ifndef STACK_GROWS_UP
 #   define STACK_GROWS_DOWN
 # endif
@@ -1029,6 +1230,10 @@
 #   define SUNOS5SIGS
 # endif
 
+# if defined(HPUX)
+#   define SUNOS5SIGS
+# endif
+
 # if CPP_WORDSZ != 32 && CPP_WORDSZ != 64
    -> bad word size
 # endif
@@ -1063,6 +1268,26 @@
 #   define DEFAULT_VDB
 # endif
 
+# ifndef PREFETCH
+#   define PREFETCH(x)
+#   define NO_PREFETCH
+# endif
+
+# ifndef PREFETCH_FOR_WRITE
+#   define PREFETCH_FOR_WRITE(x)
+#   define NO_PREFETCH_FOR_WRITE
+# endif
+
+# ifndef CACHE_LINE_SIZE
+#   define CACHE_LINE_SIZE 32	/* Wild guess	*/
+# endif
+
+# ifndef CLEAR_DOUBLE
+#   define CLEAR_DOUBLE(x) \
+	((word*)x)[0] = 0; \
+	((word*)x)[1] = 0;
+# endif /* CLEAR_DOUBLE */
+
 # if defined(_SOLARIS_PTHREADS) && !defined(SOLARIS_THREADS)
 #   define SOLARIS_THREADS
 # endif
@@ -1078,10 +1303,13 @@
 # if defined(SOLARIS_THREADS) && !defined(SUNOS5)
 --> inconsistent configuration
 # endif
+# if defined(HPUX_THREADS) && !defined(HPUX)
+--> inconsistent configuration
+# endif
 # if defined(PCR) || defined(SRC_M3) || \
 	defined(SOLARIS_THREADS) || defined(WIN32_THREADS) || \
 	defined(IRIX_THREADS) || defined(LINUX_THREADS) || \
-	defined(IRIX_JDK_THREADS)
+	defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
 #   define THREADS
 # endif
 
@@ -1096,4 +1324,4 @@
 				/* include assembly code to do it well.	*/
 # endif
 
-# endif
+# endif /* GCCONFIG_H */
diff --git a/linux_threads.c b/linux_threads.c
index 8287dce6..4b808823 100644
--- a/linux_threads.c
+++ b/linux_threads.c
@@ -36,16 +36,26 @@
 # if defined(LINUX_THREADS)
 
 # include <pthread.h>
+# include <sched.h>
 # include <time.h>
 # include <errno.h>
 # include <unistd.h>
 # include <sys/mman.h>
 # include <sys/time.h>
 # include <semaphore.h>
+# include <signal.h>
+
+#ifdef USE_LD_WRAP
+#   define WRAP_FUNC(f) __wrap_##f
+#   define REAL_FUNC(f) __real_##f
+#else
+#   define WRAP_FUNC(f) GC_##f
+#   define REAL_FUNC(f) f
+#   undef pthread_create
+#   undef pthread_sigmask
+#   undef pthread_join
+#endif
 
-#undef pthread_create
-#undef pthread_sigmask
-#undef pthread_join
 
 void GC_thr_init();
 
@@ -86,8 +96,12 @@ typedef struct GC_Thread_Rep {
 #	define DETACHED 2	/* Thread is intended to be detached.	*/
 #	define MAIN_THREAD 4	/* True for the original thread only.	*/
 
-    ptr_t stack_end;
-    ptr_t stack_ptr;  		/* Valid only when stopped. */
+    ptr_t stack_end;		/* Cold end of the stack.		*/
+    ptr_t stack_ptr;  		/* Valid only when stopped.      	*/
+#   ifdef IA64
+	ptr_t backing_store_end;
+	ptr_t backing_store_ptr;
+#   endif
     int	signal;
     void * status;		/* The value returned from the thread.  */
     				/* Used only to avoid premature 	*/
@@ -138,6 +152,10 @@ static inline ptr_t GC_linux_thread_top_of_stack(void)
   return tos;
 }
 
+#ifdef IA64
+  extern word GC_save_regs_in_stack();
+#endif
+
 void GC_suspend_handler(int sig)
 {
     int dummy;
@@ -160,7 +178,9 @@ void GC_suspend_handler(int sig)
     /* to stop the world.  Thus concurrent modification of the	*/
     /* data structure is impossible.				*/
     me -> stack_ptr = (ptr_t)(&dummy);
-    me -> stack_end = GC_linux_thread_top_of_stack();
+#   ifdef IA64
+	me -> backing_store_ptr = (ptr_t)GC_save_regs_in_stack();
+#   endif
 
     /* Tell the thread that wants to stop the world that this   */
     /* thread has been stopped.  Note that sem_post() is  	*/
@@ -173,6 +193,11 @@ void GC_suspend_handler(int sig)
     /* is no race.						*/
     if (sigfillset(&mask) != 0) ABORT("sigfillset() failed");
     if (sigdelset(&mask, SIG_RESTART) != 0) ABORT("sigdelset() failed");
+#   ifdef NO_SIGNALS
+      if (sigdelset(&mask, SIGINT) != 0) ABORT("sigdelset() failed");
+      if (sigdelset(&mask, SIGQUIT) != 0) ABORT("sigdelset() failed");
+      if (sigdelset(&mask, SIGTERM) != 0) ABORT("sigdelset() failed");
+#   endif
     do {
 	    me->signal = 0;
 	    sigsuspend(&mask);             /* Wait for signal */
@@ -375,13 +400,21 @@ void GC_start_world()
     #endif
 }
 
-/* We hold allocation lock.  We assume the world is stopped.	*/
+# ifdef IA64
+#   define IF_IA64(x) x
+# else
+#   define IF_IA64(x)
+# endif
+/* We hold allocation lock.  Should do exactly the right thing if the	*/
+/* world is stopped.  Should not fail if it isn't.			*/
 void GC_push_all_stacks()
 {
-    register int i;
-    register GC_thread p;
-    register ptr_t sp = GC_approx_sp();
-    register ptr_t lo, hi;
+    int i;
+    GC_thread p;
+    ptr_t sp = GC_approx_sp();
+    ptr_t lo, hi;
+    /* On IA64, we also need to scan the register backing store. */
+    IF_IA64(ptr_t bs_lo; ptr_t bs_hi;)
     pthread_t me = pthread_self();
     
     if (!GC_thr_initialized) GC_thr_init();
@@ -393,25 +426,33 @@ void GC_push_all_stacks()
         if (p -> flags & FINISHED) continue;
         if (pthread_equal(p -> id, me)) {
 	    lo = GC_approx_sp();
+	    IF_IA64(bs_hi = (ptr_t)GC_save_regs_in_stack();)
 	} else {
 	    lo = p -> stack_ptr;
+	    IF_IA64(bs_hi = p -> backing_store_ptr;)
 	}
         if ((p -> flags & MAIN_THREAD) == 0) {
-	    if (pthread_equal(p -> id, me)) {
-		hi = GC_linux_thread_top_of_stack();
-	    } else {
-		hi = p -> stack_end;
-	    }
+	    hi = p -> stack_end;
+	    IF_IA64(bs_lo = p -> backing_store_end);
         } else {
             /* The original stack. */
             hi = GC_stackbottom;
+	    IF_IA64(bs_lo = BACKING_STORE_BASE;)
         }
         #if DEBUG_THREADS
             GC_printf3("Stack for thread 0x%lx = [%lx,%lx)\n",
     	        (unsigned long) p -> id,
 		(unsigned long) lo, (unsigned long) hi);
         #endif
+	if (0 == lo) ABORT("GC_push_all_stacks: sp not set!\n");
         GC_push_all_stack(lo, hi);
+#	ifdef IA64
+          if (pthread_equal(p -> id, me)) {
+	    GC_push_all_eager(bs_lo, bs_hi);
+	  } else {
+	    GC_push_all_stack(bs_lo, bs_hi);
+	  }
+#	endif
       }
     }
 }
@@ -420,6 +461,7 @@ void GC_push_all_stacks()
 /* We hold the allocation lock.	*/
 void GC_thr_init()
 {
+    int dummy;
     GC_thread t;
     struct sigaction act;
 
@@ -433,6 +475,14 @@ void GC_thr_init()
     if (sigfillset(&act.sa_mask) != 0) {
     	ABORT("sigfillset() failed");
     }
+#   ifdef NO_SIGNALS
+      if (sigdelset(&act.sa_mask, SIGINT) != 0
+	  || sigdelset(&act.sa_mask, SIGQUIT != 0)
+	  || sigdelset(&act.sa_mask, SIGTERM != 0)) {
+        ABORT("sigdelset() failed");
+      }
+#   endif
+
     /* SIG_RESTART is unmasked by the handler when necessary. 	*/
     act.sa_handler = GC_suspend_handler;
     if (sigaction(SIG_SUSPEND, &act, NULL) != 0) {
@@ -446,11 +496,11 @@ void GC_thr_init()
 
     /* Add the initial thread, so we can stop it.	*/
       t = GC_new_thread(pthread_self());
-      t -> stack_ptr = 0;
+      t -> stack_ptr = (ptr_t)(&dummy);
       t -> flags = DETACHED | MAIN_THREAD;
 }
 
-int GC_pthread_sigmask(int how, const sigset_t *set, sigset_t *oset)
+int WRAP_FUNC(pthread_sigmask)(int how, const sigset_t *set, sigset_t *oset)
 {
     sigset_t fudged_set;
     
@@ -459,7 +509,7 @@ int GC_pthread_sigmask(int how, const sigset_t *set, sigset_t *oset)
         sigdelset(&fudged_set, SIG_SUSPEND);
         set = &fudged_set;
     }
-    return(pthread_sigmask(how, set, oset));
+    return(REAL_FUNC(pthread_sigmask)(how, set, oset));
 }
 
 struct start_info {
@@ -483,10 +533,25 @@ void GC_thread_exit_proc(void *arg)
     } else {
 	me -> flags |= FINISHED;
     }
+    if (GC_incremental && GC_collection_in_progress()) {
+	int old_gc_no = GC_gc_no;
+
+	/* Make sure that no part of our stack is still on the mark stack, */
+	/* since it's about to be unmapped.				   */
+	while (GC_incremental && GC_collection_in_progress()
+	       && old_gc_no == GC_gc_no) {
+	    ENTER_GC();
+            GC_collect_a_little_inner(1);
+	    EXIT_GC();
+	    UNLOCK();
+	    sched_yield();
+	    LOCK();
+	}
+    }
     UNLOCK();
 }
 
-int GC_pthread_join(pthread_t thread, void **retval)
+int WRAP_FUNC(pthread_join)(pthread_t thread, void **retval)
 {
     int result;
     GC_thread thread_gc_id;
@@ -496,7 +561,7 @@ int GC_pthread_join(pthread_t thread, void **retval)
     /* This is guaranteed to be the intended one, since the thread id	*/
     /* cant have been recycled by pthreads.				*/
     UNLOCK();
-    result = pthread_join(thread, retval);
+    result = REAL_FUNC(pthread_join)(thread, retval);
     LOCK();
     /* Here the pthread thread id may have been recycled. */
     GC_delete_gc_thread(thread, thread_gc_id);
@@ -506,6 +571,7 @@ int GC_pthread_join(pthread_t thread, void **retval)
 
 void * GC_start_routine(void * arg)
 {
+    int dummy;
     struct start_info * si = arg;
     void * result;
     GC_thread me;
@@ -514,22 +580,45 @@ void * GC_start_routine(void * arg)
     void *start_arg;
 
     my_pthread = pthread_self();
+#   ifdef DEBUG_THREADS
+        GC_printf1("Starting thread 0x%lx\n", my_pthread);
+        GC_printf1("pid = %ld\n", (long) getpid());
+        GC_printf1("sp = 0x%lx\n", (long) &arg);
+#   endif
     LOCK();
     me = GC_new_thread(my_pthread);
     me -> flags = si -> flags;
     me -> stack_ptr = 0;
-    me -> stack_end = 0;
+    /* me -> stack_end = GC_linux_stack_base(); -- currently (11/99)	*/
+    /* doesn't work because the stack base in /proc/self/stat is the 	*/
+    /* one for the main thread.  There is a strong argument that that's	*/
+    /* a kernel bug, but a pervasive one.				*/
+#   ifdef STACK_GROWS_DOWN
+      me -> stack_end = (ptr_t)(((word)(&dummy) + (GC_page_size - 1))
+		                & ~(GC_page_size - 1));
+      me -> stack_ptr = me -> stack_end - 0x10;
+	/* Needs to be plausible, since an asynchronous stack mark	*/
+	/* should not crash.						*/
+#   else
+      me -> stack_end = (ptr_t)(((word)(&dummy) & ~(GC_page_size - 1));
+      me -> stack_ptr = me -> stack_end + 0x10;
+#   endif
+    /* This is dubious, since we may be more than a page into the stack, */
+    /* and hence skip some of it, though it's not clear that matters.	 */
+#   ifdef IA64
+      me -> backing_store_end = (ptr_t)
+			(GC_save_regs_in_stack() & ~(GC_page_size - 1));
+      /* This is also < 100% convincing.  We should also read this 	*/
+      /* from /proc, but the hook to do so isn't there yet.		*/
+#   endif /* IA64 */
     UNLOCK();
     start = si -> start_routine;
-    start_arg = si -> arg;
-    sem_post(&(si -> registered));
-    pthread_cleanup_push(GC_thread_exit_proc, si);
 #   ifdef DEBUG_THREADS
-        GC_printf1("Starting thread 0x%lx\n", pthread_self());
-        GC_printf1("pid = %ld\n", (long) getpid());
-        GC_printf1("sp = 0x%lx\n", (long) &arg);
 	GC_printf1("start_routine = 0x%lx\n", start);
 #   endif
+    start_arg = si -> arg;
+    sem_post(&(si -> registered));
+    pthread_cleanup_push(GC_thread_exit_proc, si);
     result = (*start)(start_arg);
 #if DEBUG_THREADS
         GC_printf1("Finishing thread 0x%x\n", pthread_self());
@@ -544,7 +633,7 @@ void * GC_start_routine(void * arg)
 }
 
 int
-GC_pthread_create(pthread_t *new_thread,
+WRAP_FUNC(pthread_create)(pthread_t *new_thread,
 		  const pthread_attr_t *attr,
                   void *(*start_routine)(void *), void *arg)
 {
@@ -576,7 +665,14 @@ GC_pthread_create(pthread_t *new_thread,
     if (PTHREAD_CREATE_DETACHED == detachstate) my_flags |= DETACHED;
     si -> flags = my_flags;
     UNLOCK();
-    result = pthread_create(new_thread, &new_attr, GC_start_routine, si);
+#   ifdef DEBUG_THREADS
+        GC_printf1("About to start new thread from thread 0x%X\n",
+		   pthread_self());
+#   endif
+    result = REAL_FUNC(pthread_create)(new_thread, &new_attr, GC_start_routine, si);
+#   ifdef DEBUG_THREADS
+        GC_printf1("Started thread 0x%X\n", *new_thread);
+#   endif
     /* Wait until child has been added to the thread table.		*/
     /* This also ensures that we hold onto si until the child is done	*/
     /* with it.  Thus it doesn't matter whether it is otherwise		*/
@@ -588,7 +684,9 @@ GC_pthread_create(pthread_t *new_thread,
     return(result);
 }
 
-GC_bool GC_collecting = 0;
+#if defined(USE_SPIN_LOCK)
+
+VOLATILE GC_bool GC_collecting = 0;
 			/* A hint that we're in the collector and       */
                         /* holding the allocation lock for an           */
                         /* extended period.                             */
@@ -661,5 +759,7 @@ yield:
     }
 }
 
+#endif /* known architecture */
+
 # endif /* LINUX_THREADS */
 
diff --git a/mach_dep.c b/mach_dep.c
index 53698604..12c3f076 100644
--- a/mach_dep.c
+++ b/mach_dep.c
@@ -62,6 +62,12 @@ asm static void PushMacRegisters()
 
 #endif /* __MWERKS__ */
 
+# if defined(SPARC) || defined(IA64)
+    /* Value returned from register flushing routine; either sp (SPARC) */
+    /* or ar.bsp (IA64)							*/
+    word GC_save_regs_ret_val;
+# 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       */
@@ -74,6 +80,24 @@ void GC_push_regs()
 #       ifdef RT
 	  register long TMP_SP; /* must be bound to r11 */
 #       endif
+
+#       if defined(MIPS) && defined(LINUX)
+	  /* I'm not sure whether this has actually been tested. */
+#         define call_push(x)     asm("move $4," x ";"); asm("jal GC_push_one")
+	  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");
+#         undef call_push
+#       endif	/* MIPS && LINUX */
+
 #       ifdef VAX
 	  /* VAX - generic code below does not work under 4.2 */
 	  /* r1 through r5 are caller save, and therefore     */
@@ -193,10 +217,11 @@ void GC_push_regs()
 #	endif	/* __MWERKS__ */
 #   endif	/* MACOS */
 
-#       if defined(I386) &&!defined(OS2) &&!defined(SVR4) &&!defined(MSWIN32) \
+#       if defined(I386) &&!defined(OS2) &&!defined(SVR4) \
+	&& (defined(__MINGW32__) || !defined(MSWIN32)) \
 	&& !defined(SCO) && !defined(SCO_ELF) \
  	&& !(defined(LINUX)       && defined(__ELF__)) \
-	&& !(defined(__FreeBSD__) && defined(__ELF__)) \
+	&& !(defined(FREEBSD) && defined(__ELF__)) \
 	&& !defined(DOS4GW)
 	/* I386 code, generic code does not appear to work */
 	/* It does appear to work under OS2, and asms dont */
@@ -211,20 +236,25 @@ void GC_push_regs()
 #       endif
 
 #	if ( defined(I386) && defined(LINUX) && defined(__ELF__) ) \
-	|| ( defined(I386) && defined(__FreeBSD__) && defined(__ELF__) )
+	|| ( defined(I386) && defined(FREEBSD) && defined(__ELF__) )
 
 	/* This is modified for Linux with ELF (Note: _ELF_ only) */
 	/* This section handles FreeBSD with ELF. */
-	  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 %ebp");  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");
+	/* Eax is caller-save and dead here.  Other caller-save 	*/
+	/* registers could also be skipped.  We assume there are no	*/
+	/* pointers in MMX registers, etc.				*/
+	/* We combine instructions in a single asm to prevent gcc from 	*/
+	/* inserting code in the middle.				*/
+	  asm("pushl %ecx; call GC_push_one; addl $4,%esp");
+	  asm("pushl %edx; call GC_push_one; addl $4,%esp");
+	  asm("pushl %ebp; call GC_push_one; addl $4,%esp");
+	  asm("pushl %esi; call GC_push_one; addl $4,%esp");
+	  asm("pushl %edi; call GC_push_one; addl $4,%esp");
+	  asm("pushl %ebx; call GC_push_one; addl $4,%esp");
 #	endif
 
-#       if defined(I386) && defined(MSWIN32) && !defined(USE_GENERIC)
+#       if defined(I386) && defined(MSWIN32) && !defined(__MINGW32__) \
+	   && !defined(USE_GENERIC)
 	/* I386 code, Microsoft variant		*/
 	  __asm  push eax
 	  __asm  call GC_push_one
@@ -268,12 +298,11 @@ void GC_push_regs()
 	  asm ("movd r7, tos"); asm ("bsr ?_GC_push_one"); asm ("adjspb $-4");
 #       endif
 
-#       ifdef SPARC
+#       if defined(SPARC)
 	  {
 	      word GC_save_regs_in_stack();
 	      
-	      /* generic code will not work */
-	      (void)GC_save_regs_in_stack();
+	      GC_save_regs_ret_val = GC_save_regs_in_stack();
 	  }
 #       endif
 
@@ -333,12 +362,22 @@ void GC_push_regs()
 #        endif /* !__GNUC__ */
 #       endif /* M68K/SYSV */
 
+#     if defined(PJ)
+	{
+	    register int * sp asm ("optop");
+	    extern int *__libc_stack_end;
+
+	    GC_push_all_stack (sp, __libc_stack_end);
+        }
+#     endif
 
       /* other machines... */
 #       if !(defined M68K) && !(defined VAX) && !(defined RT) 
 #	if !(defined SPARC) && !(defined I386) && !(defined NS32K)
-#	if !defined(POWERPC) && !defined(UTS4)
+#	if !defined(POWERPC) && !defined(UTS4) 
+#       if !defined(PJ) && !(defined(MIPS) && defined(LINUX))
 	    --> bad news <--
+#	endif
 #       endif
 #       endif
 #       endif
@@ -363,11 +402,24 @@ ptr_t cold_gc_frame;
 		for (; (char *)i < lim; i++) {
 		    *i = 0;
 		}
-#	    if defined(POWERPC) || defined(MSWIN32) || defined(UTS4)
+#	    if defined(POWERPC) || defined(MSWIN32) || defined(UTS4) || defined(LINUX)
 		(void) setjmp(regs);
 #	    else
 	        (void) _setjmp(regs);
 #	    endif
+#           if defined(SPARC) || defined(IA64)
+	      /* On a register window machine, we need to save register	*/
+	      /* contents on the stack for this to work.  The setjmp	*/
+	      /* is probably not needed on SPARC, since pointers are	*/
+	      /* only stored in windowed or scratch registers.  It is	*/
+	      /* needed on IA64, since some non-windowed registers are	*/
+	      /* preserved.						*/
+	      {
+	        word GC_save_regs_in_stack();
+	      
+	        GC_save_regs_ret_val = GC_save_regs_in_stack();
+	      }
+#           endif
 	    GC_push_current_stack(cold_gc_frame);
 	}
 }
@@ -398,6 +450,27 @@ ptr_t cold_gc_frame;
 #   endif
 # endif
 
+/* On IA64, we also need to flush register windows.  But they end	*/
+/* up on the other side of the stack segment.				*/
+/* Returns the backing store pointer for the register stack.		*/
+# ifdef IA64
+	asm("        .text");
+	asm("        .psr abi64");
+	asm("        .psr lsb");
+	asm("        .lsb");
+	asm("");
+	asm("        .text");
+	asm("        .align 16");
+	asm("        .global GC_save_regs_in_stack");
+	asm("        .proc GC_save_regs_in_stack");
+	asm("GC_save_regs_in_stack:");
+	asm("        .body");
+	asm("        flushrs");
+	asm("        ;;");
+	asm("        mov r8=ar.bsp");
+	asm("        br.ret.sptk.few rp");
+	asm("        .endp GC_save_regs_in_stack");
+# 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	*/
diff --git a/malloc.c b/malloc.c
index 66e62d29..a5a93ad8 100644
--- a/malloc.c
+++ b/malloc.c
@@ -81,6 +81,10 @@ register ptr_t *opp;
         /* but that's benign.				*/
         /* Volatile declarations may need to be added	*/
         /* to prevent the compiler from breaking things.*/
+	/* If we only execute the second of the 	*/
+	/* following assignments, we lose the free	*/
+	/* list, but that should still be OK, at least	*/
+	/* for garbage collected memory.		*/
         *opp = obj_link(op);
         obj_link(op) = 0;
     } else {
diff --git a/mallocx.c b/mallocx.c
index 8c07fa98..c8426652 100644
--- a/mallocx.c
+++ b/mallocx.c
@@ -134,22 +134,14 @@ void GC_incr_mem_freed(size_t n)
 
 /* Analogous to the above, but assumes a small object size, and 	*/
 /* bypasses MERGE_SIZES mechanism.  Used by gc_inline.h.		*/
-#ifdef __STDC__
-     ptr_t GC_generic_malloc_words_small(size_t lw, int k)
-#else 
-     ptr_t GC_generic_malloc_words_small(lw, k)
-     register word lw;
-     register int k;
-#endif
+ptr_t GC_generic_malloc_words_small_inner(lw, k)
+register word lw;
+register int k;
 {
 register ptr_t op;
 register ptr_t *opp;
 register struct obj_kind * kind = GC_obj_kinds + k;
-DCL_LOCK_STATE;
 
-    GC_INVOKE_FINALIZERS();
-    DISABLE_SIGNALS();
-    LOCK();
     opp = &(kind -> ok_freelist[lw]);
     if( (op = *opp) == 0 ) {
         if (!GC_is_initialized) {
@@ -167,6 +159,26 @@ DCL_LOCK_STATE;
     *opp = obj_link(op);
     obj_link(op) = 0;
     GC_words_allocd += lw;
+    return((ptr_t)op);
+}
+
+/* Analogous to the above, but assumes a small object size, and 	*/
+/* bypasses MERGE_SIZES mechanism.  Used by gc_inline.h.		*/
+#ifdef __STDC__
+     ptr_t GC_generic_malloc_words_small(size_t lw, int k)
+#else 
+     ptr_t GC_generic_malloc_words_small(lw, k)
+     register word lw;
+     register int k;
+#endif
+{
+register ptr_t op;
+DCL_LOCK_STATE;
+
+    GC_INVOKE_FINALIZERS();
+    DISABLE_SIGNALS();
+    LOCK();
+    op = GC_generic_malloc_words_small_inner(lw, k);
     UNLOCK();
     ENABLE_SIGNALS();
     return((ptr_t)op);
diff --git a/mark.c b/mark.c
index 34db472a..c54d894e 100644
--- a/mark.c
+++ b/mark.c
@@ -38,7 +38,7 @@ word x;
 
 /* mark_proc GC_mark_procs[MAX_MARK_PROCS] = {0} -- declared in gc_priv.h */
 
-word GC_n_mark_procs = 0;
+word GC_n_mark_procs = GC_RESERVED_MARK_PROCS;
 
 /* Initialize GC_obj_kinds properly and standard free lists properly.  	*/
 /* This must be done statically since they may be accessed before 	*/
@@ -87,6 +87,10 @@ struct obj_kind GC_obj_kinds[MAXOBJKINDS] = {
 #   define INITIAL_MARK_STACK_SIZE (1*HBLKSIZE)
 		/* INITIAL_MARK_STACK_SIZE * sizeof(mse) should be a 	*/
 		/* multiple of HBLKSIZE.				*/
+		/* The incremental collector actually likes a larger	*/
+		/* size, since it want to push all marked dirty objs	*/
+		/* before marking anything new.  Currently we let it	*/
+		/* grow dynamically.					*/
 # endif
 
 /*
@@ -248,13 +252,29 @@ static void alloc_mark_stack();
 GC_bool GC_mark_some(cold_gc_frame)
 ptr_t cold_gc_frame;
 {
+#ifdef MSWIN32
+  /* Windows 98 appears to asynchronously create and remove writable	*/
+  /* memory mappings, for reasons we haven't yet understood.  Since	*/
+  /* we look for writable regions to determine the root set, we may	*/
+  /* try to mark from an address range that disappeared since we 	*/
+  /* started the collection.  Thus we have to recover from faults here. */
+  /* This code does not appear to be necessary for Windows 95/NT/2000.	*/ 
+  /* Note that this code should never generate an incremental GC write	*/
+  /* fault.								*/
+  __try {
+#endif
     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_stack + GC_mark_stack_size
+		   - INITIAL_MARK_STACK_SIZE/2) {
+		/* Go ahead and mark, even though that might cause us to */
+		/* see more marked dirty objects later on.  Avoid this	 */
+		/* in the future.					 */
+		GC_mark_stack_too_small = TRUE;
     	        GC_mark_from_mark_stack();
     	        return(FALSE);
     	    } else {
@@ -333,6 +353,20 @@ ptr_t cold_gc_frame;
     	    ABORT("GC_mark_some: bad state");
     	    return(FALSE);
     }
+#ifdef MSWIN32
+  } __except (GetExceptionCode() == EXCEPTION_ACCESS_VIOLATION ?
+	    EXCEPTION_EXECUTE_HANDLER : EXCEPTION_CONTINUE_SEARCH) {
+#   ifdef PRINTSTATS
+	GC_printf0("Caught ACCESS_VIOLATION in marker. "
+		   "Memory mapping disappeared.\n");
+#   endif /* PRINTSTATS */
+    /* We have bad roots on the stack.  Discard mark stack.  	*/
+    /* Rescan from marked objects.  Redetermine roots.		*/
+    GC_invalidate_mark_state();	
+    scan_ptr = 0;
+    return FALSE;
+  }
+#endif /* MSWIN32 */
 }
 
 
@@ -356,20 +390,20 @@ GC_bool GC_mark_stack_empty()
 /* with IGNORE_OFF_PAGE set.						*/
 /*ARGSUSED*/
 # ifdef PRINT_BLACK_LIST
-  word GC_find_start(current, hhdr, source)
+  ptr_t GC_find_start(current, hhdr, source)
   word source;
 # else
-  word GC_find_start(current, hhdr)
+  ptr_t GC_find_start(current, hhdr)
 # define source 0
 # endif
-register word current;
+register ptr_t current;
 register hdr * hhdr;
 {
 #   ifdef ALL_INTERIOR_POINTERS
 	if (hhdr != 0) {
-	    register word orig = current;
+	    register ptr_t orig = current;
 	    
-	    current = (word)HBLKPTR(current) + HDR_BYTES;
+	    current = (ptr_t)HBLKPTR(current) + HDR_BYTES;
 	    do {
 	      current = current - HBLKSIZE*(word)hhdr;
 	      hhdr = HDR(current);
@@ -420,6 +454,12 @@ mse * msp;
  * 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.
+ * Note that this is the most performance critical routine in the
+ * collector.  Hence it contains all sorts of ugly hacks to speed
+ * things up.  In particular, we avoid procedure calls on the common
+ * path, we take advantage of peculiarities of the mark descriptor
+ * encoding, we optionally maintain a cache for the block address to
+ * header mapping, we prefetch when an object is "grayed", etc. 
  */
 void GC_mark_from_mark_stack()
 {
@@ -434,9 +474,12 @@ void GC_mark_from_mark_stack()
   register word descr;
   register ptr_t greatest_ha = GC_greatest_plausible_heap_addr;
   register ptr_t least_ha = GC_least_plausible_heap_addr;
+  DECLARE_HDR_CACHE;
+
 # define SPLIT_RANGE_WORDS 128  /* Must be power of 2.		*/
 
   GC_objects_are_marked = TRUE;
+  INIT_HDR_CACHE;
 # ifdef OS2 /* Use untweaked version to circumvent compiler problem */
   while (GC_mark_stack_top_reg >= GC_mark_stack_reg && credit >= 0) {
 # else
@@ -444,8 +487,13 @@ void GC_mark_from_mark_stack()
   	>= 0) {
 # endif
     current_p = GC_mark_stack_top_reg -> mse_start;
-  retry:
     descr = GC_mark_stack_top_reg -> mse_descr;
+  retry:
+    /* current_p and descr describe the current object.		*/
+    /* *GC_mark_stack_top_reg is vacant.			*/
+    /* The following is 0 only for small objects described by a simple	*/
+    /* length descriptor.  For many applications this is the common	*/
+    /* case, so we try to detect it quickly.				*/
     if (descr & ((~(WORDS_TO_BYTES(SPLIT_RANGE_WORDS) - 1)) | DS_TAGS)) {
       word tag = descr & DS_TAGS;
       
@@ -456,8 +504,8 @@ void GC_mark_from_mark_stack()
           /* 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);
+          GC_mark_stack_top_reg -> mse_descr =
+          		descr - WORDS_TO_BYTES(SPLIT_RANGE_WORDS-1);
           /* Make sure that pointers overlapping the two ranges are	*/
           /* considered. 						*/
           limit = (word *)((char *)limit + sizeof(word) - ALIGNMENT);
@@ -470,8 +518,9 @@ void GC_mark_from_mark_stack()
             if ((signed_word)descr < 0) {
               current = *current_p;
 	      if ((ptr_t)current >= least_ha && (ptr_t)current < greatest_ha) {
-                PUSH_CONTENTS(current, GC_mark_stack_top_reg, mark_stack_limit,
-			      current_p, exit1);
+		PREFETCH(current);
+                HC_PUSH_CONTENTS((ptr_t)current, GC_mark_stack_top_reg,
+			      mark_stack_limit, current_p, exit1);
 	      }
             }
 	    descr <<= 1;
@@ -487,24 +536,98 @@ void GC_mark_from_mark_stack()
               	    mark_stack_limit, ENV(descr));
           continue;
         case DS_PER_OBJECT:
-          GC_mark_stack_top_reg -> mse_descr =
-			*(word *)((ptr_t)current_p + descr - tag);
+	  if ((signed_word)descr >= 0) {
+	    /* Descriptor is in the object.	*/
+            descr = *(word *)((ptr_t)current_p + descr - DS_PER_OBJECT);
+	  } else {
+	    /* Descriptor is in type descriptor pointed to by first	*/
+	    /* word in object.						*/
+	    ptr_t type_descr = *(ptr_t *)current_p;
+	    /* type_descr is either a valid pointer to the descriptor	*/
+	    /* structure, or this object was on a free list.  If it 	*/
+	    /* it was anything but the last object on the free list,	*/
+	    /* we will misinterpret the next object on the free list as */
+	    /* the type descriptor, and get a 0 GC descriptor, which	*/
+	    /* is ideal.  Unfortunately, we need to check for the last	*/
+	    /* object case explicitly.					*/
+	    if (0 == type_descr) {
+		/* Rarely executed.	*/
+		GC_mark_stack_top_reg--;
+		continue;
+	    }
+            descr = *(word *)(type_descr
+			      - (descr - (DS_PER_OBJECT - INDIR_PER_OBJ_BIAS)));
+	  }
+	  if (0 == descr) {
+	    GC_mark_stack_top_reg--;
+	    continue;
+	  }
           goto retry;
       }
-    } else {
+    } else /* Small object with length descriptor */ {
       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;
-      if ((ptr_t)current >= least_ha && (ptr_t)current <  greatest_ha) {
-        PUSH_CONTENTS(current, GC_mark_stack_top_reg,
-		      mark_stack_limit, current_p, exit2);
+    {
+#     define PREF_DIST 4
+
+#     ifndef SMALL_CONFIG
+        word deferred;
+
+	/* Try to prefetch the next pointer to be examined asap.	*/
+	/* Empirically, this also seems to help slightly without	*/
+	/* prefetches, at least on linux/X86.  Presumably this loop 	*/
+	/* ends up with less register pressure, and gcc thus ends up 	*/
+	/* generating slightly better code.  Overall gcc code quality	*/
+	/* for this loop is still not great.				*/
+	for(;;) {
+	  PREFETCH((ptr_t)limit - PREF_DIST*CACHE_LINE_SIZE);
+	  deferred = *limit;
+	  limit = (word *)((char *)limit - ALIGNMENT);
+	  if ((ptr_t)deferred >= least_ha && (ptr_t)deferred <  greatest_ha) {
+	    PREFETCH(deferred);
+	    break;
+	  }
+	  if (current_p > limit) goto next_object;
+	  /* Unroll once, so we don't do too many of the prefetches 	*/
+	  /* based on limit.						*/
+	  deferred = *limit;
+	  limit = (word *)((char *)limit - ALIGNMENT);
+	  if ((ptr_t)deferred >= least_ha && (ptr_t)deferred <  greatest_ha) {
+	    PREFETCH(deferred);
+	    break;
+	  }
+	  if (current_p > limit) goto next_object;
+	}
+#     endif
+
+      while (current_p <= limit) {
+	/* Empirically, unrolling this loop doesn't help a lot.	*/
+	/* Since HC_PUSH_CONTENTS expands to a lot of code,	*/
+	/* we don't.						*/
+        current = *current_p;
+        PREFETCH((ptr_t)current_p + PREF_DIST*CACHE_LINE_SIZE);
+        if ((ptr_t)current >= least_ha && (ptr_t)current <  greatest_ha) {
+  	  /* Prefetch the contents of the object we just pushed.  It's	*/
+  	  /* likely we will need them soon.				*/
+  	  PREFETCH(current);
+          HC_PUSH_CONTENTS((ptr_t)current, GC_mark_stack_top_reg,
+  		           mark_stack_limit, current_p, exit2);
+        }
+        current_p = (word *)((char *)current_p + ALIGNMENT);
       }
-      current_p = (word *)((char *)current_p + ALIGNMENT);
+
+#     ifndef SMALL_CONFIG
+	/* We still need to mark the entry we previously prefetched.	*/
+	/* We alrady know that it passes the preliminary pointer	*/
+	/* validity test.						*/
+        HC_PUSH_CONTENTS((ptr_t)deferred, GC_mark_stack_top_reg,
+  		         mark_stack_limit, current_p, exit4);
+	next_object:;
+#     endif
     }
   }
   GC_mark_stack_top = GC_mark_stack_top_reg;
@@ -671,7 +794,13 @@ int all;
 # endif
 word p;
 {
-    GC_PUSH_ONE_STACK(p, 0);
+#   ifdef NURSERY
+      if (0 != GC_push_proc) {
+	GC_push_proc(p);
+	return;
+      }
+#   endif
+    GC_PUSH_ONE_STACK(p, MARKED_FROM_REGISTER);
 }
 
 # ifdef __STDC__
@@ -1014,6 +1143,7 @@ struct hblk *h;
 register hdr * hhdr;
 {
     register int sz = hhdr -> hb_sz;
+    register int descr = hhdr -> hb_descr;
     register word * p;
     register int word_no;
     register word * lim;
@@ -1021,19 +1151,14 @@ register hdr * hhdr;
     register mse * mark_stack_limit = &(GC_mark_stack[GC_mark_stack_size]);
     
     /* Some quick shortcuts: */
-	{ 
-	    struct obj_kind *ok = &(GC_obj_kinds[hhdr -> hb_obj_kind]);
-	    if ((0 | DS_LENGTH) == ok -> ok_descriptor
-		&& FALSE == ok -> ok_relocate_descr)
-		return;
-	}
+	if ((0 | DS_LENGTH) == descr) 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);
+        lim = (word *)h + HDR_WORDS;
     } else {
         lim = (word *)(h + 1) - sz;
     }
@@ -1056,10 +1181,6 @@ register hdr * hhdr;
       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 ignores user specified mark procs.  This currently	*/
-         /* doesn't matter, since marking from the whole object		*/
-         /* is always sufficient, and we will eventually use the user	*/
-         /* mark proc to avoid any bogus pointers.			*/
          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);
@@ -1115,7 +1236,7 @@ struct hblk *h;
 struct hblk * GC_push_next_marked_dirty(h)
 struct hblk *h;
 {
-    register hdr * hhdr = HDR(h);
+    register hdr * hhdr;
     
     if (!GC_dirty_maintained) { ABORT("dirty bits not set up"); }
     for (;;) {
diff --git a/mark_rts.c b/mark_rts.c
index 2f21ed32..5bafd07e 100644
--- a/mark_rts.c
+++ b/mark_rts.c
@@ -412,6 +412,8 @@ ptr_t cold_gc_frame;
 	if (0 == cold_gc_frame) return;
 #       ifdef STACK_GROWS_DOWN
     	  GC_push_all_eager(GC_approx_sp(), cold_gc_frame);
+	  /* For IA64, the register stack backing store is handled 	*/
+	  /* in the thread-specific code.				*/
 #       else
 	  GC_push_all_eager( cold_gc_frame, GC_approx_sp() );
 #       endif
@@ -419,6 +421,31 @@ ptr_t cold_gc_frame;
 #   	ifdef STACK_GROWS_DOWN
     	    GC_push_all_stack_partially_eager( GC_approx_sp(), GC_stackbottom,
 					       cold_gc_frame );
+#	    ifdef IA64
+	      /* We also need to push the register stack backing store. */
+	      /* This should really be done in the same way as the	*/
+	      /* regular stack.  For now we fudge it a bit.		*/
+	      /* Note that the backing store grows up, so we can't use	*/
+	      /* GC_push_all_stack_partially_eager.			*/
+	      {
+		extern word GC_save_regs_ret_val;
+			/* Previously set to backing store pointer.	*/
+		ptr_t bsp = (ptr_t) GC_save_regs_ret_val;
+	        ptr_t cold_gc_bs_pointer;
+#		ifdef ALL_INTERIOR_POINTERS
+	          cold_gc_bs_pointer = bsp - 2048;
+		  if (cold_gc_bs_pointer < BACKING_STORE_BASE) {
+		    cold_gc_bs_pointer = BACKING_STORE_BASE;
+		  }
+		  GC_push_all(BACKING_STORE_BASE, cold_gc_bs_pointer);
+#		else
+		  cold_gc_bs_pointer = BACKING_STORE_BASE;
+#		endif
+		GC_push_all_eager(cold_gc_bs_pointer, bsp);
+		/* All values should be sufficiently aligned that we	*/
+		/* dont have to worry about the boundary.		*/
+	      }
+#	    endif
 #       else
 	    GC_push_all_stack_partially_eager( GC_stackbottom, GC_approx_sp(),
 					       cold_gc_frame );
@@ -477,6 +504,9 @@ ptr_t cold_gc_frame;
 	/* In the USE_GENERIC_PUSH_REGS case, this is done inside	*/
 	/* GC_push_regs, so that we catch callee-save registers saved	*/
 	/* inside the GC_push_regs frame.				*/
+	/* In the case of linux threads on Ia64, the hot section of	*/
+	/* the main stack is marked here, but the register stack	*/
+	/* backing store is handled in the threads-specific code.	*/
 #   endif
     if (GC_push_other_roots != 0) (*GC_push_other_roots)();
     	/* In the threads case, this also pushes thread stacks.	*/
diff --git a/misc.c b/misc.c
index c3fce63d..780dc7b5 100644
--- a/misc.c
+++ b/misc.c
@@ -42,14 +42,16 @@
 #          ifdef WIN32_THREADS
 	      GC_API CRITICAL_SECTION GC_allocate_ml;
 #          else
-#             if defined(IRIX_THREADS) || defined(LINUX_THREADS) \
-		 || defined(IRIX_JDK_THREADS)
-#		ifdef UNDEFINED
-		    pthread_mutex_t GC_allocate_ml = PTHREAD_MUTEX_INITIALIZER;
-#		endif
+#             if defined(IRIX_THREADS) || defined(IRIX_JDK_THREADS) \
+		 || (defined(LINUX_THREADS) && defined(USE_SPIN_LOCK))
 	        pthread_t GC_lock_holder = NO_THREAD;
 #	      else
-	        --> declare allocator lock here
+#	        if defined(HPUX_THREADS) \
+		   || defined(LINUX_THREADS) && !defined(USE_SPIN_LOCK)
+		  pthread_mutex_t GC_allocate_ml = PTHREAD_MUTEX_INITIALIZER;
+#		else 
+	          --> declare allocator lock here
+#		endif
 #	      endif
 #	   endif
 #	endif
@@ -118,6 +120,15 @@ extern signed_word GC_mem_found;
 	for (i = 8*sizeof(word) + 1; i <= 16 * sizeof(word); i++) {
 	      GC_size_map[i] = (ROUNDED_UP_WORDS(i) + 1) & (~1);
 	}
+#	ifdef GC_GCJ_SUPPORT
+	   /* Make all sizes up to 32 words predictable, so that a 	*/
+	   /* compiler can statically perform the same computation,	*/
+	   /* or at least a computation that results in similar size	*/
+	   /* classes.							*/
+	   for (i = 16*sizeof(word) + 1; i <= 32 * sizeof(word); i++) {
+	      GC_size_map[i] = (ROUNDED_UP_WORDS(i) + 3) & (~3);
+	   }
+#	endif
 	/* We leave the rest of the array to be filled in on demand. */
     }
     
@@ -391,6 +402,11 @@ size_t GC_get_heap_size GC_PROTO(())
     return ((size_t) GC_heapsize);
 }
 
+size_t GC_get_free_bytes GC_PROTO(())
+{
+    return ((size_t) GC_large_free_bytes);
+}
+
 size_t GC_get_bytes_since_gc GC_PROTO(())
 {
     return ((size_t) WORDS_TO_BYTES(GC_words_allocd));
@@ -433,7 +449,8 @@ void GC_init_inner()
 #   ifdef MSWIN32
  	GC_init_win32();
 #   endif
-#   if defined(LINUX) && (defined(POWERPC) || defined(ALPHA) || defined(SPARC))
+#   if defined(SEARCH_FOR_DATA_START)
+	/* This doesn't really work if the collector is in a shared library. */
 	GC_init_linux_data_start();
 #   endif
 #   ifdef SOLARIS_THREADS
@@ -442,11 +459,12 @@ void GC_init_inner()
         GC_dirty_init();
 #   endif
 #   if defined(IRIX_THREADS) || defined(LINUX_THREADS) \
-       || defined(IRIX_JDK_THREADS)
+       || defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
         GC_thr_init();
 #   endif
 #   if !defined(THREADS) || defined(SOLARIS_THREADS) || defined(WIN32_THREADS) \
-       || defined(IRIX_THREADS) || defined(LINUX_THREADS)
+       || defined(IRIX_THREADS) || defined(LINUX_THREADS) \
+       || defined(HPUX_THREADS)
       if (GC_stackbottom == 0) {
 	GC_stackbottom = GC_get_stack_base();
       }
@@ -799,7 +817,8 @@ struct callinfo info[NFRAMES];
 
 #endif /* SAVE_CALL_CHAIN */
 
-# ifdef SRC_M3
+/* Needed by SRC_M3, gcj, and should perhaps be the official interface	*/
+/* to GC_dont_gc.							*/
 void GC_enable()
 {
     GC_dont_gc--;
@@ -809,7 +828,6 @@ void GC_disable()
 {
     GC_dont_gc++;
 }
-# endif
 
 #if !defined(NO_DEBUGGING)
 
diff --git a/new_hblk.c b/new_hblk.c
index 9f32ae0d..1e1273f8 100644
--- a/new_hblk.c
+++ b/new_hblk.c
@@ -103,10 +103,10 @@ ptr_t ofl;
     p[3] = 0;
     p += 4;
     for (; p < lim; p += 4) {
+	PREFETCH_FOR_WRITE(p+64);
         p[0] = (word)(p-4);
         p[1] = 0;
-        p[2] = 0;
-        p[3] = 0;
+	CLEAR_DOUBLE(p+2);
     };
     return((ptr_t)(p-4));
 }
@@ -141,6 +141,7 @@ ptr_t ofl;
     p[4] = (word)p;
     p += 8;
     for (; p < lim; p += 8) {
+	PREFETCH_FOR_WRITE(p+64);
         p[0] = (word)(p-4);
         p[4] = (word)p;
     };
@@ -179,6 +180,10 @@ int kind;
   /* Mark all objects if appropriate. */
       if (IS_UNCOLLECTABLE(kind)) GC_set_hdr_marks(HDR(h));
 
+  PREFETCH_FOR_WRITE((char *)h);
+  PREFETCH_FOR_WRITE((char *)h + 128);
+  PREFETCH_FOR_WRITE((char *)h + 256);
+  PREFETCH_FOR_WRITE((char *)h + 378);
   /* Handle small objects sizes more efficiently.  For larger objects 	*/
   /* the difference is less significant.				*/
 #  ifndef SMALL_CONFIG
diff --git a/nursery.c b/nursery.c
index 8bbb1015..ab83afba 100644
--- a/nursery.c
+++ b/nursery.c
@@ -11,12 +11,16 @@
  * modified is included with the above copyright notice.
  */
 
-
+#ifdef NURSERY
 ??? This implementation is incomplete.  If you are trying to
 ??? compile this you are doing something wrong.
 
 #include "nursery.h"
 
+#define SCAN_STATICS_FOR_NURSERY
+	/* If this is not defined, the collector will not see 	*/
+	/* references from static data areas to the nursery.	*/
+
 struct copy_obj {
     ptr_t forward;	/* Forwarding link for copied objects.	*/
     GC_copy_descriptor descr; /* Object descriptor	*/
@@ -68,6 +72,23 @@ unsigned short *GC_pinned;	/* Number of pinned objects in ith	*/
 
 GC_copy_alloc_state global_alloc_state = (ptr_t)(-1);	/* will overflow. */
 
+/* Array of known rescuing pointers from the heap to the nursery.	*/
+  ptr_t ** nursery_rescuers;
+  /* Pointer to one past the last slot in rescuer table	*/
+  ptr_t ** nursery_rescuers_end;
+  /* Maximum number of known rescuing pointers.			*/
+# define MAX_NURSERY_RESCUERS 32*1024
+  /* Add a rescuer to the list	*/
+# define ADD_NURSERY_RESCUER(p) \
+    if (nursery_rescuers_end >= nursery_rescuers + MAX_NURSERY_RESCUERS) { \
+      ABORT("Nursery recuers overflow"); /* Fix later !!! */ \
+    } else { \
+      *nursery_rescuers_end++ = p; \
+    }
+  /* Remove rescuer at the given position in the table	*/
+# define REMOVE_RESCUER(p) \
+    *p = *--nursery_rescuers_end
+
 /* Should be called with allocator lock held.	*/
 GC_nursery_init() {
     GC_nursery_start = GET_MEM(GC_nursery_size);
@@ -92,44 +113,158 @@ GC_nursery_init() {
     GC_nursery_blocks = GC_nursery_size/NURSERY_BLOCK_SIZE;
     GC_pinned = GC_scratch_alloc(GC_nursery_blocks * sizeof(unsigned short));
     BZERO(GC_pinned, GC_nursery_blocks);
+    nursery_rescuers = GET_MEM(MAX_NURSERY_RESCUERS * sizeof(ptr_t *));
+    nursery_rescuers_end = nursery_rescuers;
+    if (0 == GC_nursery_start || 0 == GC_nursery_map || 0 == nursery_rescuers)
+	ABORT("Insufficient memory for nursery");
+}
+
+#define PIN_OBJ(p) \
+    if (p >= GC_nursery_start && p < GC_nursery_end) { GC_pin_obj_checked(p); }
+
+/* Pin the object at p, if it's in the nursery.	*/
+void GC_pin_obj(ptr_t p) {
+    PIN_OBJ(p);
+}
+
+void (*GC_push_proc)(ptr_t) = 0;
+
+/* Pin the object at p, which is known to be in the nursery.	*/
+void GC_pin_obj_checked(ptr_t p) {
+    unsigned offset = p - GC_nursery_start;
+    unsigned word_offset = BYTES_TO_WORDS(offset);
+    unsigned blockno = (current - GC_nursery_start)/NURSERY_BLOCK_SIZE;
+    while (GC_nursery_map[word_offset] == NURSERY_MAP_NOT_START) {
+	--word_offset;    
+    }
+    if (GC_nursery_map[word_offset] != NURSERY_MAP_PINNED) {
+        GC_nursery_map[word_offset] = NURSERY_MAP_PINNED;
+        ++GC_pinned[blockno];
+        ??Push object at GC_nursery_start + WORDS_TO_BYTES(word_offset)
+        ??onto mark stack. 
+    }
+}
+
+void GC_scan_region_for_nursery(ptr_t low, ptr_t high) {
+#   if CPP_WORDSZ/8 != ALIGNMENT
+      --> fix this
+#   endif
+    word * l = (word *)((word)low + ALIGNMENT - 1 & ~(ALIGNMENT - 1));
+    word * h = (word *)((word)high & ~(ALIGNMENT - 1));
+    word * p;
+    for (p = l; p < h; ++p) {
+	PIN_OBJ(p);
+    }
 }
 
-/* Pin all nursery objects referenced from mark stack. */
-void GC_pin_mark_stack_objects(void) {
-    for each possible pointer current in a mark stack object
-	if (current >= GC_nursery_start && current < GC_nursery_end) {
-	    unsigned offset = current - GC_nursery_start;
-	    unsigned word_offset = BYTES_TO_WORDS(offset);
-	    unsigned blockno = (current - GC_nursery_start)/NURSERY_BLOCK_SIZE;
-	    while (GC_nursery_map[word_offset] == NURSERY_MAP_NOT_START) {
-		--word_offset;    
-	    }
-	    if (GC_nursery_map[word_offset] != NURSERY_MAP_PINNED) {
-	        GC_nursery_map[word_offset] = NURSERY_MAP_PINNED;
-	        ++GC_pinned[blockno];
-	        ??Push object at GC_nursery_start + WORDS_TO_BYTES(word_offset)
-	        ??onto stack. 
-	    }
+/* Invoke GC_scan_region_for_nursery on ranges that are not excluded. */
+void GC_scan_region_for_nursery_with_exclusions(ptr_t bottom, ptr_t top)
+{
+    struct exclusion * next;
+    ptr_t excl_start;
+
+    while (bottom < top) {
+        next = GC_next_exclusion(bottom);
+	if (0 == next || (excl_start = next -> e_start) >= top) {
+	    GC_scan_region_for_nursery(bottom, top);
+	    return;
 	}
+	if (excl_start > bottom)
+		GC_scan_region_for_nursery(bottom, excl_start);
+	bottom = next -> e_end;
     }
 }
 
+
+void GC_scan_stacks_for_nursery(void) {
+#   ifdef THREADS
+	--> fix this
+#   endif
+#   ifdef STACK_GROWS_DOWN
+      ptr_t stack_low = GC_approx_sp();
+      ptr_t stack_high = GC_stackbottom;
+#   else
+      ptr_t stack_low = GC_stackbottom;
+      ptr_t stack_high = GC_approx_sp();
+#   endif
+    GC_scan_region_for_nursery(stack_low, stack_high);
+#   ifdef IA64
+      GC_scan_region_for_nursery(BACKING_STORE_BASE,
+				 (ptr_t) GC_save_regs_ret_val);
+#   endif
+}
+
+void GC_scan_roots_for_nursery(void) {
+  /* Scan registers.	*/
+    /* Direct GC_push_one to call GC_pin_obj instead of marking	*/
+    /* and pushing objects.					*/
+    /* This is a bit ugly, but we don't have to touch the	*/
+    /* platform-dependent code.					*/
+     
+    void (*old_push_proc)(ptr_t) = GC_push_proc;
+    GC_push_proc = GC_pin_obj;
+    GC_push_regs();
+    GC_push_proc = old_push_proc;
+  GC_scan_stacks_for_nursery();
+# ifdef SCAN_STATICS_FOR_NURSERY
+#   if (defined(DYNAMIC_LOADING) || defined(MSWIN32) || defined(PCR)) \
+        && !defined(SRC_M3)
+      GC_remove_tmp_roots();
+      GC_register_dynamic_libraries();
+#   endif
+    /* Mark everything in static data areas                             */
+      for (i = 0; i < n_root_sets; i++) {
+        GC_scan_region_for_nursery_with_exclusions (
+			     GC_static_roots[i].r_start,
+			     GC_static_roots[i].r_end);
+     }
+# endif
+}
+
+/* Array of known rescuing pointers from the heap to the nursery.	*/
+ptr_t ** nursery_rescuers;
+
 /* Caller holds allocation lock.	*/
 void GC_collect_nursery(void) {
     int i;
     ptr_t scan_ptr = 0;
-    ?? old_mark_stack_top;
     STOP_WORLD;
     for (i = 0; i < GC_nursery_blocks; ++i) GC_pinned[i] = 0;
-    GC_push_all_roots();
-    old_mark_stack_top = GC_mark_stack_top();
-    GC_pin_mark_stack_objects();
+    GC_scan_roots_for_nursery();
+    /* All objects referenced by roots are now pinned. 		*/
+    /* Their contents are described by 			 	*/
+    /* mark stack entries.					*/
+
+    /* Pin blocks corresponding to valid allocation states.	*/
+    /* that probably happens automagically if the allocation	*/
+    /* states are kept where we can see them.			*/
+    /* It will take work if static roots are not scanned.	*/
+    /* We want to do this both for correctness and to avoid	*/
+    /* promoting very young objects.				*/
+
+    /* Somehow capture dirty bits.  Update rescuers array to	*/
+    /* reflect newly valid and invalid references from dirty 	*/
+    /* pages.  Other references should remain valid, since the	*/
+    /* referents should have been pinned.			*/
+
+    /* Traverse the old object heap.  Pin objects in the 	*/
+    /* nursery that are ambiguously referenced, copy those	*/
+    /* that are unambiguously referenced.			*/
+
+    /* Traverse objects in mark stack.				*/
+    /* If referenced object is in pinned block, add contents	*/
+    /* to mark stack.  If referenced object is forwarded,	*/
+    /* update pointer.  Otherwise reallocate the object	in the	*/
+    /* old heap, copy its contents, and then enqueue its 	*/
+    /* contents in the mark stack.				*/
     START_WORLD;
 }
 
 /* Initialize an allocation state so that it can be used for 	*/
 /* allocation.  This implicitly reserves a small section of the	*/
-/* nursery for use with his allocator.				*/
+/* nursery for use with this allocator.				*/
+/* Also called to replenish an allocator that has been 		*/
+/* exhausted.							*/
 void GC_init_copy_alloc_state(GC_copy_alloc_state *)
     unsigned next_block;
     ptr_t block_addr;
@@ -173,3 +308,5 @@ GC_PTR GC_copying_malloc2(GC_copy_descriptor *d, GC_copy_alloc_state *s) {
 
 GC_PTR GC_copying_malloc(GC_copy_descriptor *d) {
 }
+
+#endif /* NURSERY */
diff --git a/os_dep.c b/os_dep.c
index 81f74f3a..34b871d8 100644
--- a/os_dep.c
+++ b/os_dep.c
@@ -1,6 +1,8 @@
 /*
+ * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
  * Copyright (c) 1991-1995 by Xerox Corporation.  All rights reserved.
- * Copyright (c) 1996-1997 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company.  All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -31,7 +33,7 @@
       /* make sure the former gets defined to be the latter if appropriate. */
 #     include <features.h>
 #     if 2 <= __GLIBC__
-#       if 0 == __GLIBC_MINOR__
+#       if 2 == __GLIBC__ && 0 == __GLIBC_MINOR__
 	  /* glibc 2.1 no longer has sigcontext.h.  But signal.h	*/
 	  /* has the right declaration for glibc 2.1.			*/
 #         include <sigcontext.h>
@@ -54,17 +56,17 @@
 # include <signal.h>
 
 /* Blatantly OS dependent routines, except for those that are related 	*/
-/* dynamic loading.							*/
+/* to dynamic loading.							*/
 
 # if !defined(THREADS) && !defined(STACKBOTTOM) && defined(HEURISTIC2)
 #   define NEED_FIND_LIMIT
 # endif
 
-# if defined(IRIX_THREADS)
+# if defined(IRIX_THREADS) || defined(HPUX_THREADS)
 #   define NEED_FIND_LIMIT
 # endif
 
-# if (defined(SUNOS4) & defined(DYNAMIC_LOADING)) && !defined(PCR)
+# if (defined(SUNOS4) && defined(DYNAMIC_LOADING)) && !defined(PCR)
 #   define NEED_FIND_LIMIT
 # endif
 
@@ -72,7 +74,9 @@
 #   define NEED_FIND_LIMIT
 # endif
 
-# if defined(LINUX) && (defined(POWERPC) || defined(SPARC) || defined(ALPHA))
+# if defined(LINUX) && \
+     (defined(POWERPC) || defined(SPARC) || defined(ALPHA) || defined(IA64) \
+      || defined(MIPS))
 #   define NEED_FIND_LIMIT
 # endif
 
@@ -139,7 +143,8 @@
 # define OPT_PROT_EXEC 0
 #endif
 
-#if defined(LINUX) && (defined(POWERPC) || defined(SPARC) || defined(ALPHA))
+#if defined(SEARCH_FOR_DATA_START)
+  /* The following doesn't work if the GC is in a dynamic library.	*/
   /* The I386 case can be handled without a search.  The Alpha case	*/
   /* used to be handled differently as well, but the rules changed	*/
   /* for recent Linux versions.  This seems to be the easiest way to	*/
@@ -497,7 +502,7 @@ ptr_t GC_get_stack_base()
 
 #   if defined(SUNOS5SIGS) || defined(IRIX5) || defined(OSF1)
 	static struct sigaction old_segv_act;
-#	if defined(_sigargs) /* !Irix6.x */
+#	if defined(_sigargs) || defined(HPUX) /* !Irix6.x */
 	    static struct sigaction old_bus_act;
 #	endif
 #   else
@@ -525,10 +530,11 @@ ptr_t GC_get_stack_base()
 	        (void) sigaction(SIGSEGV, &act, 0);
 #	  else
 	        (void) sigaction(SIGSEGV, &act, &old_segv_act);
-#		ifdef _sigargs	/* Irix 5.x, not 6.x */
-		    /* Under 5.x, we may get SIGBUS.			*/
-		    /* Pthreads doesn't exist under 5.x, so we don't	*/
-		    /* have to worry in the threads case.		*/
+#		if defined(IRIX5) && defined(_sigargs) /* Irix 5.x, not 6.x */ \
+		   || defined(HPUX)
+		    /* Under Irix 5.x or HP/UX, we may get SIGBUS.	*/
+		    /* Pthreads doesn't exist under Irix 5.x, so we	*/
+		    /* don't have to worry in the threads case.		*/
 		    (void) sigaction(SIGBUS, &act, &old_bus_act);
 #		endif
 #	  endif	/* IRIX_THREADS */
@@ -544,7 +550,8 @@ ptr_t GC_get_stack_base()
     {
 #       if defined(SUNOS5SIGS) || defined(IRIX5) || defined(OSF1)
 	  (void) sigaction(SIGSEGV, &old_segv_act, 0);
-#	  ifdef _sigargs	/* Irix 5.x, not 6.x */
+#	  if defined(IRIX5) && defined(_sigargs) /* Irix 5.x, not 6.x */ \
+	     || defined(HPUX)
 	      (void) sigaction(SIGBUS, &old_bus_act, 0);
 #	  endif
 #       else
@@ -589,6 +596,55 @@ ptr_t GC_get_stack_base()
     }
 # endif
 
+#ifdef LINUX_STACKBOTTOM
+
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+
+# define STAT_SKIP 27   /* Number of fields preceding startstack	*/
+			/* field in /proc/self/stat			*/
+
+  ptr_t GC_linux_stack_base(void)
+  {
+    /* We read the stack base value from /proc/self/stat.  We do this	*/
+    /* using direct I/O system calls in order to avoid calling malloc   */
+    /* in case REDIRECT_MALLOC is defined.				*/ 
+#   define STAT_BUF_SIZE 4096
+#   ifdef USE_LD_WRAP
+#	define STAT_READ __real_read
+#   else
+#	define STAT_READ read
+#   endif    
+    char stat_buf[STAT_BUF_SIZE];
+    int f;
+    char c;
+    word result = 0;
+    size_t i, buf_offset = 0;
+
+    f = open("/proc/self/stat", O_RDONLY);
+    if (f < 0 || STAT_READ(f, stat_buf, STAT_BUF_SIZE) < 2 * STAT_SKIP) {
+	ABORT("Couldn't read /proc/self/stat");
+    }
+    c = stat_buf[buf_offset++];
+    /* Skip the required number of fields.  This number is hopefully	*/
+    /* constant across all Linux implementations.			*/
+      for (i = 0; i < STAT_SKIP; ++i) {
+	while (isspace(c)) c = stat_buf[buf_offset++];
+	while (!isspace(c)) c = stat_buf[buf_offset++];
+      }
+    while (isspace(c)) c = stat_buf[buf_offset++];
+    while (isdigit(c)) {
+      result *= 10;
+      result += c - '0';
+      c = stat_buf[buf_offset++];
+    }
+    close(f);
+    if (result < 0x10000000) ABORT("Absurd stack bottom value");
+    return (ptr_t)result;
+  }
+
+#endif /* LINUX_STACKBOTTOM */
 
 ptr_t GC_get_stack_base()
 {
@@ -610,6 +666,9 @@ ptr_t GC_get_stack_base()
 			      & ~STACKBOTTOM_ALIGNMENT_M1);
 #	   endif
 #	endif /* HEURISTIC1 */
+#	ifdef LINUX_STACKBOTTOM
+	   result = GC_linux_stack_base();
+#	endif
 #	ifdef HEURISTIC2
 #	    ifdef STACK_GROWS_DOWN
 		result = GC_find_limit((ptr_t)(&dummy), TRUE);
@@ -1443,7 +1502,7 @@ void GC_default_push_other_roots()
 
 # if defined(SOLARIS_THREADS) || defined(WIN32_THREADS) \
      || defined(IRIX_THREADS) || defined(LINUX_THREADS) \
-     || defined(IRIX_PCR_THREADS)
+     || defined(IRIX_JDK_THREADS) || defined(HPUX_THREADS)
 
 extern void GC_push_all_stacks();
 
@@ -1570,12 +1629,12 @@ struct hblk *h;
 #   include <sys/syscall.h>
 
 #   define PROTECT(addr, len) \
-    	  if (mprotect((caddr_t)(addr), (int)(len), \
+    	  if (mprotect((caddr_t)(addr), (size_t)(len), \
     	      	       PROT_READ | OPT_PROT_EXEC) < 0) { \
     	    ABORT("mprotect failed"); \
     	  }
 #   define UNPROTECT(addr, len) \
-    	  if (mprotect((caddr_t)(addr), (int)(len), \
+    	  if (mprotect((caddr_t)(addr), (size_t)(len), \
     	  	       PROT_WRITE | PROT_READ | OPT_PROT_EXEC ) < 0) { \
     	    ABORT("un-mprotect failed"); \
     	  }
@@ -1604,7 +1663,11 @@ struct hblk *h;
     typedef void (* SIG_PF)();
 #endif
 #if defined(SUNOS5SIGS) || defined(OSF1) || defined(LINUX)
+# ifdef __STDC__
     typedef void (* SIG_PF)(int);
+# else
+    typedef void (* SIG_PF)();
+# endif
 #endif
 #if defined(MSWIN32)
     typedef LPTOP_LEVEL_EXCEPTION_FILTER SIG_PF;
@@ -1616,17 +1679,34 @@ struct hblk *h;
     typedef void (* REAL_SIG_PF)(int, int, struct sigcontext *);
 #endif
 #if defined(SUNOS5SIGS)
-    typedef void (* REAL_SIG_PF)(int, struct siginfo *, void *);
+# ifdef HPUX
+#   define SIGINFO __siginfo
+# else
+#   define SIGINFO siginfo
+# endif
+# ifdef __STDC__
+    typedef void (* REAL_SIG_PF)(int, struct SIGINFO *, void *);
+# else
+    typedef void (* REAL_SIG_PF)();
+# endif
 #endif
 #if defined(LINUX)
 #   include <linux/version.h>
-#   if (LINUX_VERSION_CODE >= 0x20100) && !defined(M68K) || defined(ALPHA)
+#   if (LINUX_VERSION_CODE >= 0x20100) && !defined(M68K) || defined(ALPHA) || defined(IA64)
       typedef struct sigcontext s_c;
 #   else
       typedef struct sigcontext_struct s_c;
 #   endif
+#   if defined(ALPHA) || defined(M68K)
+      typedef void (* REAL_SIG_PF)(int, int, s_c *);
+#   else
+#     if defined(IA64)
+        typedef void (* REAL_SIG_PF)(int, siginfo_t *, s_c *);
+#     else
+        typedef void (* REAL_SIG_PF)(int, s_c);
+#     endif
+#   endif
 #   ifdef ALPHA
-    typedef void (* REAL_SIG_PF)(int, int, s_c *);
     /* Retrieve fault address from sigcontext structure by decoding	*/
     /* instruction.							*/
     char * get_fault_addr(s_c *sc) {
@@ -1638,8 +1718,6 @@ struct hblk *h;
 	faultaddr += (word) (((int)instr << 16) >> 16);
 	return (char *)faultaddr;
     }
-#   else /* !ALPHA */
-    typedef void (* REAL_SIG_PF)(int, s_c);
 #   endif /* !ALPHA */
 # endif
 
@@ -1675,21 +1753,41 @@ SIG_PF GC_old_segv_handler;	/* Also old MSWIN32 ACCESS_VIOLATION filter */
 #   endif
 # endif
 # if defined(LINUX)
-#   ifdef ALPHA
+#   if defined(ALPHA) || defined(M68K)
       void GC_write_fault_handler(int sig, int code, s_c * sc)
 #   else
-      void GC_write_fault_handler(int sig, s_c sc)
+#     if defined(IA64)
+        void GC_write_fault_handler(int sig, siginfo_t * si, s_c * scp)
+#     else
+        void GC_write_fault_handler(int sig, s_c sc)
+#     endif
 #   endif
 #   define SIG_OK (sig == SIGSEGV)
 #   define CODE_OK TRUE
-	/* Empirically c.trapno == 14, but is that useful?      */
-	/* We assume Intel architecture, so alignment		*/
-	/* faults are not possible.				*/
+	/* Empirically c.trapno == 14, on IA32, but is that useful?     */
+	/* Should probably consider alignment issues on other 		*/
+	/* architectures.						*/
 # endif
 # if defined(SUNOS5SIGS)
-    void GC_write_fault_handler(int sig, struct siginfo *scp, void * context)
-#   define SIG_OK (sig == SIGSEGV)
-#   define CODE_OK (scp -> si_code == SEGV_ACCERR)
+#  ifdef __STDC__
+    void GC_write_fault_handler(int sig, struct SIGINFO *scp, void * context)
+#  else
+    void GC_write_fault_handler(sig, scp, context)
+    int sig;
+    struct SIGINFO *scp;
+    void * context;
+#  endif
+#   ifdef HPUX
+#     define SIG_OK (sig == SIGSEGV || sig == SIGBUS)
+#     define CODE_OK (scp -> si_code == SEGV_ACCERR) \
+		     || (scp -> si_code == BUS_ADRERR) \
+		     || (scp -> si_code == BUS_UNKNOWN) \
+		     || (scp -> si_code == SEGV_UNKNOWN) \
+		     || (scp -> si_code == BUS_OBJERR)
+#   else
+#     define SIG_OK (sig == SIGSEGV)
+#     define CODE_OK (scp -> si_code == SEGV_ACCERR)
+#   endif
 # endif
 # if defined(MSWIN32)
     LONG WINAPI GC_write_fault_handler(struct _EXCEPTION_POINTERS *exc_info)
@@ -1741,7 +1839,18 @@ SIG_PF GC_old_segv_handler;	/* Also old MSWIN32 ACCESS_VIOLATION filter */
 #	  ifdef ALPHA
             char * addr = get_fault_addr(sc);
 #	  else
+#	    ifdef IA64
+	      char * addr = si -> si_addr;
+	      /* I believe this is claimed to work on all platforms for	*/
+	      /* Linux 2.3.47 and later.  Hopefully we don't have to	*/
+	      /* worry about earlier kernels on IA64.			*/
+#	    else
+#             if defined(POWERPC)
+                char * addr = (char *) (sc.regs->dar);
+#	      else
 		--> architecture not supported
+#	      endif
+#	    endif
 #	  endif
 #	endif
 #     endif
@@ -1794,10 +1903,14 @@ SIG_PF GC_old_segv_handler;	/* Also old MSWIN32 ACCESS_VIOLATION filter */
 		    return;
 #		endif
 #		if defined (LINUX)
-#		    ifdef ALPHA
+#		    if defined(ALPHA) || defined(M68K)
 		        (*(REAL_SIG_PF)old_handler) (sig, code, sc);
 #		    else 
+#		      if defined(IA64)
+		        (*(REAL_SIG_PF)old_handler) (sig, si, scp);
+#		      else
 		        (*(REAL_SIG_PF)old_handler) (sig, sc);
+#		      endif
 #		    endif
 		    return;
 #		endif
@@ -1909,7 +2022,7 @@ void GC_dirty_init()
       }
 #   endif
 #   if defined(SUNOS5SIGS) || defined(IRIX5)
-#     if defined(IRIX_THREADS) || defined(IRIX_PCR_THREADS)
+#     if defined(IRIX_THREADS) || defined(IRIX_JDK_THREADS)
       	sigaction(SIGSEGV, 0, &oldact);
       	sigaction(SIGSEGV, &act, 0);
 #     else
@@ -1935,6 +2048,15 @@ void GC_dirty_init()
 	  GC_err_printf0("Replaced other SIGSEGV handler\n");
 #       endif
       }
+#     ifdef HPUX
+      	  sigaction(SIGBUS, &act, &oldact);
+          GC_old_bus_handler = oldact.sa_handler;
+          if (GC_old_segv_handler != SIG_DFL) {
+#           ifdef PRINTSTATS
+	      GC_err_printf0("Replaced other SIGBUS handler\n");
+#           endif
+          }
+#     endif
 #    endif
 #   if defined(MSWIN32)
       GC_old_segv_handler = SetUnhandledExceptionFilter(GC_write_fault_handler);
@@ -2026,12 +2148,13 @@ word len;
     	      ((ptr_t)end_block - (ptr_t)start_block) + HBLKSIZE);
 }
 
-#ifndef MSWIN32
+#if !defined(MSWIN32) && !defined(LINUX_THREADS)
 /* Replacement for UNIX system call.	 */
 /* Other calls that write to the heap	 */
 /* should be handled similarly.		 */
 # if defined(__STDC__) && !defined(SUNOS4)
 #   include <unistd.h>
+#   include <sys/uio.h>
     ssize_t read(int fd, void *buf, size_t nbyte)
 # else
 #   ifndef LINT
@@ -2048,10 +2171,12 @@ word len;
     
     GC_begin_syscall();
     GC_unprotect_range(buf, (word)nbyte);
-#   ifdef IRIX5
+#   if defined(IRIX5) || defined(LINUX_THREADS)
 	/* Indirect system call may not always be easily available.	*/
 	/* We could call _read, but that would interfere with the	*/
 	/* libpthread interception of read.				*/
+	/* On Linux, we have to be careful with the linuxthreads	*/
+	/* read interception.						*/
 	{
 	    struct iovec iov;
 
@@ -2065,7 +2190,29 @@ word len;
     GC_end_syscall();
     return(result);
 }
-#endif /* !MSWIN32 */
+#endif /* !MSWIN32 && !LINUX */
+
+#ifdef USE_LD_WRAP
+    /* We use the GNU ld call wrapping facility.			*/
+    /* This requires that the linker be invoked with "--wrap read".	*/
+    /* This can be done by passing -Wl,"--wrap read" to gcc.		*/
+    /* I'm not sure that this actually wraps whatever version of read	*/
+    /* is called by stdio.  That code also mentions __read.		*/
+#   include <unistd.h>
+    ssize_t __wrap_read(int fd, void *buf, size_t nbyte)
+    {
+ 	int result;
+
+	GC_begin_syscall();
+    	GC_unprotect_range(buf, (word)nbyte);
+	result = __real_read(fd, buf, nbyte);
+	GC_end_syscall();
+	return(result);
+    }
+
+    /* We should probably also do this for __read, or whatever stdio	*/
+    /* actually calls.							*/
+#endif
 
 /*ARGSUSED*/
 GC_bool GC_page_was_ever_dirty(h)
diff --git a/reclaim.c b/reclaim.c
index 30859469..4ace1107 100644
--- a/reclaim.c
+++ b/reclaim.c
@@ -1,6 +1,8 @@
 /* 
  * Copyright 1988, 1989 Hans-J. Boehm, Alan J. Demers
- * Copyright (c) 1991-1994 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1991-1996 by Xerox Corporation.  All rights reserved.
+ * Copyright (c) 1996-1999 by Silicon Graphics.  All rights reserved.
+ * Copyright (c) 1999 by Hewlett-Packard Company. All rights reserved.
  *
  * THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED
  * OR IMPLIED.  ANY USE IS AT YOUR OWN RISK.
@@ -11,7 +13,6 @@
  * provided the above notices are retained, and a notice that the code was
  * modified is included with the above copyright notice.
  */
-/* Boehm, February 15, 1996 2:41 pm PST */
 
 #include <stdio.h>
 #include "gc_priv.h"
@@ -28,13 +29,8 @@ word sz;
     } 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));
-    }
+    GC_print_heap_obj(p);
+    GC_err_printf0("\n");
 }
 
 #   define FOUND_FREE(hblk, word_no) \
@@ -67,6 +63,139 @@ register hdr * hhdr;
     return(TRUE);
 }
 
+/* The following functions sometimes return a DONT_KNOW value. */
+#define DONT_KNOW  2
+
+#ifdef SMALL_CONFIG
+# define GC_block_nearly_full1(hhdr, pat1) DONT_KNOW
+# define GC_block_nearly_full3(hhdr, pat1, pat2) DONT_KNOW
+# define GC_block_nearly_full(hhdr) DONT_KNOW
+#else
+
+/*
+ * Test whether nearly all of the mark words consist of the same
+ * repeating pattern.
+ */
+#define FULL_THRESHOLD (MARK_BITS_SZ/16)
+
+GC_bool GC_block_nearly_full1(hhdr, pat1)
+hdr *hhdr;
+word pat1;
+{
+    unsigned i;
+    unsigned misses = 0;
+    GC_ASSERT((MARK_BITS_SZ & 1) == 0);
+    for (i = 0; i < MARK_BITS_SZ; ++i) {
+	if ((hhdr -> hb_marks[i] | ~pat1) != ONES) {
+	    if (++misses > FULL_THRESHOLD) return FALSE;
+	}
+    }
+    return TRUE;
+}
+
+/*
+ * Test whether the same repeating 3 word pattern occurs in nearly
+ * all the mark bit slots.
+ * This is used as a heuristic, so we're a bit sloppy and ignore
+ * the last one or two words.
+ */
+GC_bool GC_block_nearly_full3(hhdr, pat1, pat2, pat3)
+hdr *hhdr;
+word pat1, pat2, pat3;
+{
+    unsigned i;
+    unsigned misses = 0;
+
+    if (MARK_BITS_SZ < 4) {
+      return DONT_KNOW;
+    }
+    for (i = 0; i < MARK_BITS_SZ - 2; i += 3) {
+	if ((hhdr -> hb_marks[i] | ~pat1) != ONES) {
+	    if (++misses > FULL_THRESHOLD) return FALSE;
+	}
+	if ((hhdr -> hb_marks[i+1] | ~pat2) != ONES) {
+	    if (++misses > FULL_THRESHOLD) return FALSE;
+	}
+	if ((hhdr -> hb_marks[i+2] | ~pat3) != ONES) {
+	    if (++misses > FULL_THRESHOLD) return FALSE;
+	}
+    }
+    return TRUE;
+}
+
+/* Check whether a small object block is nearly full by looking at only */
+/* the mark bits.							*/
+/* We manually precomputed the mark bit patterns that need to be 	*/
+/* checked for, and we give up on the ones that are unlikely to occur,	*/
+/* or have period > 3.							*/
+/* This would be a lot easier with a mark bit per object instead of per	*/
+/* word, but that would rewuire computing object numbers in the mark	*/
+/* loop, which would require different data structures ...		*/
+GC_bool GC_block_nearly_full(hhdr)
+hdr *hhdr;
+{
+    int sz = hhdr -> hb_sz;
+
+#   if CPP_WORDSZ != 32 && CPP_WORDSZ != 64
+      return DONT_KNOW;	/* Shouldn't be used in any standard config.	*/
+#   endif
+    if (0 != HDR_WORDS) return DONT_KNOW;
+	/* Also shouldn't happen */
+#   if CPP_WORDSZ == 32
+      switch(sz) {
+        case 1:
+	  return GC_block_nearly_full1(hhdr, 0xffffffffl);
+	case 2:
+	  return GC_block_nearly_full1(hhdr, 0x55555555l);
+	case 4:
+	  return GC_block_nearly_full1(hhdr, 0x11111111l);
+	case 6:
+	  return GC_block_nearly_full3(hhdr, 0x41041041l,
+					      0x10410410l,
+					       0x04104104l);
+	case 8:
+	  return GC_block_nearly_full1(hhdr, 0x01010101l);
+	case 12:
+	  return GC_block_nearly_full3(hhdr, 0x01001001l,
+					      0x10010010l,
+					       0x00100100l);
+	case 16:
+	  return GC_block_nearly_full1(hhdr, 0x00010001l);
+	case 32:
+	  return GC_block_nearly_full1(hhdr, 0x00000001l);
+	default:
+	  return DONT_KNOW;
+      }
+#   endif
+#   if CPP_WORDSZ == 64
+      switch(sz) {
+        case 1:
+	  return GC_block_nearly_full1(hhdr, 0xffffffffffffffffl);
+	case 2:
+	  return GC_block_nearly_full1(hhdr, 0x5555555555555555l);
+	case 4:
+	  return GC_block_nearly_full1(hhdr, 0x1111111111111111l);
+	case 6:
+	  return GC_block_nearly_full3(hhdr, 0x1041041041041041l,
+					       0x4104104104104104l,
+					         0x0410410410410410l);
+	case 8:
+	  return GC_block_nearly_full1(hhdr, 0x0101010101010101l);
+	case 12:
+	  return GC_block_nearly_full3(hhdr, 0x1001001001001001l,
+					       0x0100100100100100l,
+					         0x0010010010010010l);
+	case 16:
+	  return GC_block_nearly_full1(hhdr, 0x0001000100010001l);
+	case 32:
+	  return GC_block_nearly_full1(hhdr, 0x0000000100000001l);
+	default:
+	  return DONT_KNOW;
+      }
+#   endif
+}
+#endif /* !SMALL_CONFIG */
+
 # ifdef GATHERSTATS
 #   define INCR_WORDS(sz) n_words_found += (sz)
 # else
@@ -187,8 +316,7 @@ register ptr_t list;
 	    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; \
+	    CLEAR_DOUBLE(p + start_displ + 2); \
 	    INCR_WORDS(4); \
 	}
     
@@ -448,10 +576,11 @@ register struct hblk *hbp;	/* ptr to current heap block		*/
 int report_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;
-    register int kind;
+    word sz;		/* size of objects in current block	*/
+    struct obj_kind * ok;
+    ptr_t * flh;
+    int kind;
+    GC_bool full;
     
     hhdr = HDR(hbp);
     sz = hhdr -> hb_sz;
@@ -459,7 +588,6 @@ int report_if_found;		/* Abort if a reclaimable object is found */
     kind = hhdr -> hb_obj_kind;
     ok = &GC_obj_kinds[kind];
     flh = &(ok -> ok_freelist[sz]);
-    GC_write_hint(hbp);
 
     if (report_if_found) {
 	GC_reclaim_check(hbp, hhdr, sz);
@@ -467,16 +595,41 @@ int report_if_found;		/* Abort if a reclaimable object is found */
       switch(sz) {
 #      ifndef SMALL_CONFIG
         case 1:
+#           if CPP_WORDSZ == 64
+	      full = GC_block_nearly_full1(hhdr, 0xffffffffffffffffl);
+#	    else
+	      full = GC_block_nearly_full1(hhdr, 0xffffffffl);
+#	    endif
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
+	    /* In the DONT_KNOW case, we let reclaim fault.	*/
             *flh = GC_reclaim1(hbp, hhdr, *flh);
             break;
         case 2:
+#           if CPP_WORDSZ == 64
+	      full = GC_block_nearly_full1(hhdr, 0x5555555555555555l);
+#	    else
+	      full = GC_block_nearly_full1(hhdr, 0x55555555l);
+#	    endif
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim_clear2(hbp, hhdr, *flh);
             break;
         case 4:
+#           if CPP_WORDSZ == 64
+	      full = GC_block_nearly_full1(hhdr, 0x1111111111111111l);
+#	    else
+	      full = GC_block_nearly_full1(hhdr, 0x11111111l);
+#	    endif
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim_clear4(hbp, hhdr, *flh);
             break;
 #      endif
         default:
+	    full = GC_block_nearly_full(hhdr);
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim_clear(hbp, hhdr, sz, *flh);
             break;
       }
@@ -484,20 +637,45 @@ int report_if_found;		/* Abort if a reclaimable object is found */
       switch(sz) {
 #      ifndef SMALL_CONFIG
         case 1:
+#           if CPP_WORDSZ == 64
+	      full = GC_block_nearly_full1(hhdr, 0xffffffffffffffffl);
+#	    else
+	      full = GC_block_nearly_full1(hhdr, 0xffffffffl);
+#	    endif
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim1(hbp, hhdr, *flh);
             break;
         case 2:
+#           if CPP_WORDSZ == 64
+	      full = GC_block_nearly_full1(hhdr, 0x5555555555555555l);
+#	    else
+	      full = GC_block_nearly_full1(hhdr, 0x55555555l);
+#	    endif
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim_uninit2(hbp, hhdr, *flh);
             break;
         case 4:
+#           if CPP_WORDSZ == 64
+	      full = GC_block_nearly_full1(hhdr, 0x1111111111111111l);
+#	    else
+	      full = GC_block_nearly_full1(hhdr, 0x11111111l);
+#	    endif
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim_uninit4(hbp, hhdr, *flh);
             break;
 #      endif
         default:
+	    full = GC_block_nearly_full(hhdr);
+	    if (TRUE == full) goto out;
+	    if (FALSE == full) GC_write_hint(hbp);
             *flh = GC_reclaim_uninit(hbp, hhdr, sz, *flh);
             break;
       }
     } 
+out:
     if (IS_UNCOLLECTABLE(kind)) GC_set_hdr_marks(hhdr);
 }
 
@@ -654,6 +832,12 @@ int report_if_found;		/* Abort if a GC_reclaimable object is found */
   /* 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)report_if_found);
+
+# ifdef EAGER_SWEEP
+    /* This is a very stupid thing to do.  We make it possible anyway,	*/
+    /* so that you can convince yourself that it really is very stupid.	*/
+    GC_reclaim_all((GC_stop_func)0, FALSE);
+# endif
     
 }
 
@@ -687,7 +871,7 @@ int kind;
  * Abort and return FALSE when/if (*stop_func)() returns TRUE.
  * If this returns TRUE, then it's safe to restart the world
  * with incorrectly cleared mark bits.
- * If ignore_old is TRUE, then reclain only blocks that have been 
+ * If ignore_old is TRUE, then reclaim only blocks that have been 
  * recently reclaimed, and discard the rest.
  * Stop_func may be 0.
  */
diff --git a/solaris_pthreads.c b/solaris_pthreads.c
index 58424577..467de484 100644
--- a/solaris_pthreads.c
+++ b/solaris_pthreads.c
@@ -77,14 +77,16 @@ GC_pthread_create(pthread_t *new_thread,
     pthread_attr_t  attr;
     word my_flags = 0;
     int  flag;
-    void * stack;
-    size_t stack_size;
+    void * stack = 0;
+    size_t stack_size = 0;
     int    n;
     struct sched_param schedparam;
    
-    (void)pthread_attr_getstacksize(attr_in, &stack_size);
-    (void)pthread_attr_getstackaddr(attr_in, &stack);
     (void)pthread_attr_init(&attr);
+    if (attr_in != 0) {
+	(void)pthread_attr_getstacksize(attr_in, &stack_size);
+	(void)pthread_attr_getstackaddr(attr_in, &stack);
+    }
 
     LOCK();
     if (!GC_thr_initialized) {
@@ -94,7 +96,11 @@ GC_pthread_create(pthread_t *new_thread,
 	    
     if (stack == 0) {
      	if (stack_size == 0)
-		stack_size = GC_min_stack_sz;
+		stack_size = 1048576;
+			  /* ^-- 1 MB (this was GC_min_stack_sz, but that
+			   * violates the pthread_create documentation which
+			   * says the default value if none is supplied is
+			   * 1MB) */
 	else
 		stack_size += thr_min_stack();
 
@@ -110,20 +116,22 @@ GC_pthread_create(pthread_t *new_thread,
     }
     (void)pthread_attr_setstacksize(&attr, stack_size);
     (void)pthread_attr_setstackaddr(&attr, stack);
-    (void)pthread_attr_getscope(attr_in, &n);
-    (void)pthread_attr_setscope(&attr, n);
-    (void)pthread_attr_getschedparam(attr_in, &schedparam);
-    (void)pthread_attr_setschedparam(&attr, &schedparam);
-    (void)pthread_attr_getschedpolicy(attr_in, &n);
-    (void)pthread_attr_setschedpolicy(&attr, n);
-    (void)pthread_attr_getinheritsched(attr_in, &n);
-    (void)pthread_attr_setinheritsched(&attr, n);
-
-    (void)pthread_attr_getdetachstate(attr_in, &flag);
-    if (flag == PTHREAD_CREATE_DETACHED) {
-	    my_flags |= DETACHED;
+    if (attr_in != 0) {
+	(void)pthread_attr_getscope(attr_in, &n);
+	(void)pthread_attr_setscope(&attr, n);
+	(void)pthread_attr_getschedparam(attr_in, &schedparam);
+	(void)pthread_attr_setschedparam(&attr, &schedparam);
+	(void)pthread_attr_getschedpolicy(attr_in, &n);
+	(void)pthread_attr_setschedpolicy(&attr, n);
+	(void)pthread_attr_getinheritsched(attr_in, &n);
+	(void)pthread_attr_setinheritsched(&attr, n);
+
+	(void)pthread_attr_getdetachstate(attr_in, &flag);
+	if (flag == PTHREAD_CREATE_DETACHED) {
+		my_flags |= DETACHED;
+	}
+	(void)pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_JOINABLE);
     }
-    (void)pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_JOINABLE);
     /*
      * thr_create can call malloc(), which if redirected will
      * attempt to acquire the allocation lock.
diff --git a/solaris_threads.c b/solaris_threads.c
index 65b2c651..c3b0b15b 100644
--- a/solaris_threads.c
+++ b/solaris_threads.c
@@ -661,7 +661,8 @@ void GC_my_stack_limits()
 }
 
 
-/* We hold allocation lock.  We assume the world is stopped.	*/
+/* We hold allocation lock.  Should do exactly the right thing if the	*/
+/* world is stopped.  Should not fail if it isn't.			*/
 void GC_push_all_stacks()
 {
     register int i;
@@ -900,7 +901,7 @@ GC_thr_create(void *stack_base, size_t stack_size,
     }
     GC_multithreaded++;
     if (stack == 0) {
-     	if (stack_size == 0) stack_size = GC_min_stack_sz;
+     	if (stack_size == 0) stack_size = 1024*1024;
      	stack = (void *)GC_stack_alloc(&stack_size);
      	if (stack == 0) {
 	    GC_multithreaded--;
diff --git a/test.c b/test.c
index b65632c9..96a54150 100644
--- a/test.c
+++ b/test.c
@@ -15,6 +15,8 @@
 /* An incomplete test for the garbage collector.  		*/
 /* Some more obscure entry points are not tested at all.	*/
 
+# undef GC_BUILD
+
 # if defined(mips) && defined(SYSTYPE_BSD43)
     /* MIPS RISCOS 4 */
 # else
@@ -45,7 +47,7 @@
 #   include <synch.h>
 # endif
 
-# if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+# if defined(IRIX_THREADS) || defined(LINUX_THREADS) || defined(HPUX_THREADS)
 #   include <pthread.h>
 # endif
 
@@ -53,9 +55,6 @@
 #   include <process.h>
     static CRITICAL_SECTION incr_cs;
 # endif
-# if defined(PCR) || defined(SOLARIS_THREADS) || defined(WIN32_THREADS)
-#   define THREADS
-# endif
 
 # ifdef AMIGA
    long __stack = 200000;
@@ -150,7 +149,6 @@ sexpr y;
     register sexpr r;
     
     r = (sexpr) GC_MALLOC_UNCOLLECTABLE(sizeof(struct SEXPR));
-assert(GC_is_marked(r));
     if (r == 0) {
         (void)GC_printf0("Out of memory\n");
         exit(1);
@@ -160,6 +158,76 @@ assert(GC_is_marked(r));
     return(r);
 }
 
+#ifdef GC_GCJ_SUPPORT
+
+#include "gc_mark.h"
+#include "dbg_mlc.h"
+#include "include/gc_gcj.h"
+
+/* The following struct emulates the vtable in gcj.	*/
+/* This assumes the default value of MARK_DESCR_OFFSET. */
+struct fake_vtable {
+  void * dummy;		/* class pointer in real gcj.	*/
+  size_t descr;
+};
+
+struct fake_vtable gcj_class_struct1 = { 0, sizeof(struct SEXPR)
+					    + sizeof(struct fake_vtable *) };
+			/* length based descriptor.	*/
+struct fake_vtable gcj_class_struct2 =
+				{ 0, (3l << (CPP_WORDSZ - 3)) | DS_BITMAP};
+			/* Bitmap based descriptor.	*/
+
+struct ms_entry * fake_gcj_mark_proc(word * addr,
+				     struct ms_entry *mark_stack_ptr,
+				     struct ms_entry *mark_stack_limit,
+				     word env   )
+{
+    sexpr x;
+    if (1 == env) {
+	/* Object allocated with debug allocator.	*/
+	addr = (word *)USR_PTR_FROM_BASE(addr);
+    }
+    x = (sexpr)(addr + 1); /* Skip the vtable pointer. */
+    /* We could just call PUSH_CONTENTS directly here.  But any real	*/
+    /* real client would try to filter out the obvious misses.		*/
+    if (0 != x -> sexpr_cdr) {
+	PUSH_CONTENTS((ptr_t)(x -> sexpr_cdr), mark_stack_ptr,
+			      mark_stack_limit, &(x -> sexpr_cdr), exit1);
+    }
+    if ((ptr_t)(x -> sexpr_car) > GC_least_plausible_heap_addr) {
+	PUSH_CONTENTS((ptr_t)(x -> sexpr_car), mark_stack_ptr,
+			      mark_stack_limit, &(x -> sexpr_car), exit2);
+    }
+    return(mark_stack_ptr);
+}
+
+sexpr gcj_cons(x, y)
+sexpr x;
+sexpr y;
+{
+    GC_word * r;
+    sexpr result;
+    static int count = 0;
+    
+    if (++count & 1) {
+        r = (GC_word *) GC_GCJ_FAST_MALLOC(3, &gcj_class_struct1);
+    } else {
+        r = (GC_word *) GC_GCJ_MALLOC(sizeof(struct SEXPR)
+				      + sizeof(struct fake_vtable*),
+				      &gcj_class_struct2);
+    }
+    if (r == 0) {
+        (void)GC_printf0("Out of memory\n");
+        exit(1);
+    }
+    result = (sexpr)(r + 1);
+    result -> sexpr_car = x;
+    result -> sexpr_cdr = y;
+    return(result);
+}
+#endif
+
 /* Return reverse(x) concatenated with y */
 sexpr reverse1(x, y)
 sexpr x, y;
@@ -187,6 +255,35 @@ int low, up;
     }
 }
 
+#ifdef GC_GCJ_SUPPORT
+/* Return reverse(x) concatenated with y */
+sexpr gcj_reverse1(x, y)
+sexpr x, y;
+{
+    if (is_nil(x)) {
+        return(y);
+    } else {
+        return( gcj_reverse1(cdr(x), gcj_cons(car(x), y)) );
+    }
+}
+
+sexpr gcj_reverse(x)
+sexpr x;
+{
+    return( gcj_reverse1(x, nil) );
+}
+
+sexpr gcj_ints(low, up)
+int low, up;
+{
+    if (low > up) {
+	return(nil);
+    } else {
+        return(gcj_cons(gcj_cons(INT_TO_SEXPR(low), nil), gcj_ints(low+1, up)));
+    }
+}
+#endif /* GC_GCJ_SUPPORT */
+
 /* To check uncollectable allocation we build lists with disguised cdr	*/
 /* pointers, and make sure they don't go away.				*/
 sexpr uncollectable_ints(low, up)
@@ -281,7 +378,7 @@ struct {
 }
 
 # if defined(IRIX_THREADS) || defined(LINUX_THREADS) \
-     || defined(SOLARIS_PTHREADS)
+     || defined(SOLARIS_PTHREADS) || defined(HPUX_THREADS)
     void fork_a_thread()
     {
       pthread_t t;
@@ -370,7 +467,12 @@ void reverse_test()
     g[799] = ints(1,18);
     h = (sexpr *)GC_MALLOC(1025 * sizeof(sexpr));
     h = (sexpr *)GC_REALLOC((GC_PTR)h, 2000 * sizeof(sexpr));
-    h[1999] = ints(1,19);
+#   ifdef GC_GCJ_SUPPORT
+      h[1999] = gcj_ints(1,200);
+      h[1999] = gcj_reverse(h[1999]);
+#   else
+      h[1999] = ints(1,200);
+#   endif
     /* Try to force some collections and reuse of small list elements */
       for (i = 0; i < 10; i++) {
         (void)ints(1, BIG);
@@ -415,7 +517,10 @@ void reverse_test()
     check_uncollectable_ints(d, 1, 100);
     check_ints(f[5], 1,17);
     check_ints(g[799], 1,18);
-    check_ints(h[1999], 1,19);
+#   ifdef GC_GCJ_SUPPORT
+      h[1999] = gcj_reverse(h[1999]);
+#   endif
+    check_ints(h[1999], 1,200);
 #   ifndef THREADS
 	a = 0;
 #   endif  
@@ -453,7 +558,7 @@ VOLATILE int dropped_something = 0;
     static mutex_t incr_lock;
     mutex_lock(&incr_lock);
 # endif
-# if  defined(IRIX_THREADS) || defined(LINUX_THREADS)
+# if  defined(IRIX_THREADS) || defined(LINUX_THREADS) || defined(HPUX_THREADS)
     static pthread_mutex_t incr_lock = PTHREAD_MUTEX_INITIALIZER;
     pthread_mutex_lock(&incr_lock);
 # endif
@@ -471,7 +576,7 @@ VOLATILE int dropped_something = 0;
 # ifdef SOLARIS_THREADS
     mutex_unlock(&incr_lock);
 # endif
-# if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+# if defined(IRIX_THREADS) || defined(LINUX_THREADS) || defined(HPUX_THREADS)
     pthread_mutex_unlock(&incr_lock);
 # endif
 # ifdef WIN32_THREADS
@@ -532,7 +637,8 @@ int n;
 	    static mutex_t incr_lock;
 	    mutex_lock(&incr_lock);
 #	  endif
-#         if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#         if defined(IRIX_THREADS) || defined(LINUX_THREADS) \
+	     || defined(HPUX_THREADS)
             static pthread_mutex_t incr_lock = PTHREAD_MUTEX_INITIALIZER;
             pthread_mutex_lock(&incr_lock);
 #         endif
@@ -548,7 +654,8 @@ int n;
 #	  ifdef SOLARIS_THREADS
 	    mutex_unlock(&incr_lock);
 #	  endif
-#	  if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#	  if defined(IRIX_THREADS) || defined(LINUX_THREADS) \
+	     || defined(HPUX_THREADS)
 	    pthread_mutex_unlock(&incr_lock);
 #	  endif
 #         ifdef WIN32_THREADS
@@ -644,7 +751,7 @@ void * alloc8bytes()
 #else
 
 # if defined(_SOLARIS_PTHREADS) || defined(IRIX_THREADS) \
-     || defined(LINUX_THREADS)
+     || defined(LINUX_THREADS) || defined(HPUX_THREADS)
 pthread_key_t fl_key;
 
 void * alloc8bytes()
@@ -760,6 +867,10 @@ void typed_test()
     old = 0;
     for (i = 0; i < 4000; i++) {
         new = (GC_word *) GC_malloc_explicitly_typed(4 * sizeof(GC_word), d1);
+        if (0 != new[0] || 0 != new[1]) {
+	    GC_printf0("Bad initialization by GC_malloc_explicitly_typed\n");
+	    FAIL;
+	}
         new[0] = 17;
         new[1] = (GC_word)old;
         old = new;
@@ -783,6 +894,10 @@ void typed_test()
           new = (GC_word *) GC_calloc_explicitly_typed(1001,
           					       3 * sizeof(GC_word),
         					       d2);
+          if (0 != new[0] || 0 != new[1]) {
+	    GC_printf0("Bad initialization by GC_malloc_explicitly_typed\n");
+	    FAIL;
+	  }
         }
         new[0] = 17;
         new[1] = (GC_word)old;
@@ -857,6 +972,7 @@ void run_one_test()
     	(void)GC_printf0("GC_malloc_uncollectable(0) failed\n");
 	    FAIL;
     }
+    GC_FREE(0);
     GC_is_valid_displacement_print_proc = fail_proc1;
     GC_is_visible_print_proc = fail_proc1;
     x = GC_malloc(16);
@@ -879,7 +995,7 @@ void run_one_test()
 	FAIL;
     }
     if (!TEST_FAIL_COUNT(1)) {
-#	if!(defined(RS6000) || defined(POWERPC))
+#	if!(defined(RS6000) || defined(POWERPC) || defined(IA64))
 	  /* ON RS6000s function pointers point to a descriptor in the	*/
 	  /* data segment, so there should have been no failures.	*/
     	  (void)GC_printf0("GC_is_visible produced wrong failure indication\n");
@@ -906,6 +1022,10 @@ void run_one_test()
     /* Test floating point alignment */
 	*(double *)GC_MALLOC(sizeof(double)) = 1.0;
 	*(double *)GC_MALLOC(sizeof(double)) = 1.0;
+#   ifdef GC_GCJ_SUPPORT
+      GC_REGISTER_DISPLACEMENT(sizeof(struct fake_vtable *));
+      GC_init_gcj_malloc(0, (void *)fake_gcj_mark_proc);
+#   endif
     /* Repeated list reversal test. */
 	reverse_test();
 #   ifdef PRINTSTATS
@@ -1030,8 +1150,9 @@ void SetMinimumStack(long minSize)
 
 
 #if !defined(PCR) && !defined(SOLARIS_THREADS) && !defined(WIN32_THREADS) \
-  && !defined(IRIX_THREADS) && !defined(LINUX_THREADS) || defined(LINT)
-#ifdef MSWIN32
+  && !defined(IRIX_THREADS) && !defined(LINUX_THREADS) \
+  && !defined(HPUX_THREADS) || defined(LINT)
+#if defined(MSWIN32) && !defined(__MINGW32__)
   int APIENTRY WinMain(HINSTANCE instance, HINSTANCE prev, LPSTR cmd, int n)
 #else
   int main()
@@ -1113,19 +1234,24 @@ int APIENTRY WinMain(HINSTANCE instance, HINSTANCE prev, LPSTR cmd, int n)
 # endif
   InitializeCriticalSection(&incr_cs);
   (void) GC_set_warn_proc(warn_proc);
-  for (i = 0; i < NTEST; i++) {
+# if NTEST > 0
+   for (i = 0; i < NTEST; i++) {
     h[i] = (HANDLE)_beginthreadex(NULL, 0, thr_run_one_test, 0, 0, &thread_id);
     if (h[i] == (HANDLE)-1) {
       (void)GC_printf1("Thread creation failed %lu\n", (unsigned long)GetLastError());
       FAIL;
     }
-  }
+   }
+# endif /* NTEST > 0 */
   run_one_test();
-  for (i = 0; i < NTEST; i++)
+# if NTEST > 0
+   for (i = 0; i < NTEST; i++) {
     if (WaitForSingleObject(h[i], INFINITE) != WAIT_OBJECT_0) {
       (void)GC_printf1("Thread wait failed %lu\n", (unsigned long)GetLastError());
       FAIL;
     }
+   }
+# endif /* NTEST > 0 */
   check_heap_stats();
   (void)fflush(stdout);
   return(0);
@@ -1161,7 +1287,8 @@ test()
 }
 #endif
 
-#if defined(SOLARIS_THREADS) || defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#if defined(SOLARIS_THREADS) || defined(IRIX_THREADS) \
+ || defined(HPUX_THREADS) || defined(LINUX_THREADS)
 void * thr_run_one_test(void * arg)
 {
     run_one_test();
@@ -1222,7 +1349,7 @@ main()
 	*((volatile char *)&code - 1024*1024) = 0;      /* Require 1 Mb */
 #   endif /* IRIX_THREADS */
     pthread_attr_init(&attr);
-#   ifdef IRIX_THREADS
+#   if defined(IRIX_THREADS) || defined(HPUX_THREADS)
     	pthread_attr_setstacksize(&attr, 1000000);
 #   endif
     n_tests = 0;
@@ -1232,7 +1359,7 @@ main()
 	(void) GC_printf0("Emulating dirty bits with mprotect/signals\n");
 #   endif
     (void) GC_set_warn_proc(warn_proc);
-    if (pthread_key_create(&fl_key, 0) != 0) {
+    if ((code = pthread_key_create(&fl_key, 0)) != 0) {
         (void)GC_printf1("Key creation failed %lu\n", (unsigned long)code);
     	FAIL;
     }
@@ -1260,4 +1387,4 @@ main()
     return(0);
 }
 #endif /* pthreads */
-#endif /* SOLARIS_THREADS || IRIX_THREADS || LINUX_THREADS */
+#endif /* SOLARIS_THREADS || IRIX_THREADS || LINUX_THREADS || HPUX_THREADS */
diff --git a/test_cpp.cc b/test_cpp.cc
index 3160b098..0d450776 100644
--- a/test_cpp.cc
+++ b/test_cpp.cc
@@ -37,6 +37,12 @@ extern "C" {
 #ifdef MSWIN32
 #   include <windows.h>
 #endif
+#ifdef GC_NAME_CONFLICT
+#   define USE_GC UseGC
+    struct foo * GC;
+#else
+#   define USE_GC GC
+#endif
 
 
 #define my_assert( e ) \
@@ -214,7 +220,7 @@ int APIENTRY WinMain(
         for (i = 0; i < 1000; i++) {
             C* c = new C( 2 );
             C c1( 2 );           /* stack allocation should work too */
-            D* d = ::new (GC, D::CleanUp, (void*) i) D( i );
+            D* d = ::new (USE_GC, D::CleanUp, (void*) i) D( i );
             F* f = new F;
             if (0 == i % 10) delete c;}
 
@@ -222,9 +228,9 @@ int APIENTRY WinMain(
             drop the references to them immediately, forcing many
             collections. */
         for (i = 0; i < 1000000; i++) {
-            A* a = new (GC) A( i );
+            A* a = new (USE_GC) A( i );
             B* b = new B( i );
-            b = new (GC) B( i );
+            b = new (USE_GC) B( i );
             if (0 == i % 10) {
                 B::Deleting( 1 );
                 delete b;
diff --git a/threadlibs.c b/threadlibs.c
index 4a0a6cfc..7379454c 100644
--- a/threadlibs.c
+++ b/threadlibs.c
@@ -3,9 +3,21 @@
 
 int main()
 {
-#   if defined(IRIX_THREADS) || defined(LINUX_THREADS)
+#   if defined(LINUX_THREADS)
+#     ifdef USE_LD_WRAP
+	printf("-Wl,--wrap -Wl,read -Wl,--wrap -Wl,dlopen "
+	       "-Wl,--wrap -Wl,pthread_create -Wl,--wrap -Wl,pthread_join "
+	       "-Wl,--wrap -Wl,pthread_sigmask -lpthread -ldl\n");
+#     else
+	printf("-lpthread -ldl\n");
+#     endif
+#   endif
+#   if defined(IRIX_THREADS)
 	printf("-lpthread\n");
 #   endif
+#   if defined(HPUX_THREADS)
+	printf("-lpthread -lrt\n");
+#   endif
 #   ifdef SOLARIS_THREADS
         printf("-lthread -ldl\n");
 #   endif
diff --git a/typd_mlc.c b/typd_mlc.c
index 74f455d9..a198f6c1 100644
--- a/typd_mlc.c
+++ b/typd_mlc.c
@@ -175,7 +175,7 @@ GC_descr GC_double_descr(descriptor, nwords)
 register GC_descr descriptor;
 register word nwords;
 {
-    if (descriptor && DS_TAGS == DS_LENGTH) {
+    if (descriptor & DS_TAGS == DS_LENGTH) {
         descriptor = GC_bm_table[BYTES_TO_WORDS((word)descriptor)];
     };
     descriptor |= (descriptor & ~DS_TAGS) >> nwords;
@@ -430,7 +430,7 @@ word env;
     	if (bm & 1) {
     	    current = *current_p;
     	    if ((ptr_t)current >= least_ha && (ptr_t)current <= greatest_ha) {
-    	        PUSH_CONTENTS(current, mark_stack_ptr,
+    	        PUSH_CONTENTS((ptr_t)current, mark_stack_ptr,
 			      mark_stack_limit, current_p, exit1);
     	    }
     	}
@@ -665,6 +665,7 @@ DCL_LOCK_STATE;
 #	    endif
         } else {
             *opp = obj_link(op);
+	    obj_link(op) = 0;
             GC_words_allocd += lw;
             FASTUNLOCK();
         }
@@ -708,6 +709,7 @@ DCL_LOCK_STATE;
 #	    endif
         } else {
             *opp = obj_link(op);
+	    obj_link(op) = 0;
             GC_words_allocd += lw;
             FASTUNLOCK();
         }
@@ -717,7 +719,7 @@ DCL_LOCK_STATE;
        lw = BYTES_TO_WORDS(GC_size(op));
    }
    if (op != NULL)
-   ((word *)op)[lw - 1] = d;
+       ((word *)op)[lw - 1] = d;
    return((GC_PTR) op);
 }
 
@@ -772,6 +774,7 @@ DCL_LOCK_STATE;
 #	    endif
         } else {
             *opp = obj_link(op);
+	    obj_link(op) = 0;
             GC_words_allocd += lw;
             FASTUNLOCK();
         }
diff --git a/version.h b/version.h
index 97ac5f5e..5319f56d 100644
--- a/version.h
+++ b/version.h
@@ -1,6 +1,6 @@
 #define GC_VERSION_MAJOR 5
-#define GC_VERSION_MINOR 0
-#define GC_ALPHA_VERSION 3
+#define GC_VERSION_MINOR 3
+#define GC_ALPHA_VERSION GC_NOT_ALPHA
 
 #   define GC_NOT_ALPHA 0xff
 
diff --git a/win32_threads.c b/win32_threads.c
index f6f74bd1..7c8de4d2 100755
--- a/win32_threads.c
+++ b/win32_threads.c
@@ -2,8 +2,10 @@
 
 #include "gc_priv.h"
 
+#if 0
 #define STRICT
 #include <windows.h>
+#endif
 
 #define MAX_THREADS 64
 
@@ -31,6 +33,20 @@ void GC_stop_world()
   for (i = 0; i < MAX_THREADS; i++)
     if (thread_table[i].stack != 0
 	&& thread_table[i].id != thread_id) {
+      /* Apparently the Windows 95 GetOpenFileName call creates	*/
+      /* a thread that does not properly get cleaned up, and		*/
+      /* SuspendThread on its descriptor may provoke a crash.		*/
+      /* This reduces the probability of that event, though it still	*/
+      /* appears there's a race here.					*/
+      DWORD exitCode; 
+      if (GetExitCodeThread(thread_table[i].handle,&exitCode) &&
+            exitCode != STILL_ACTIVE) {
+            thread_table[i].stack = 0;
+          thread_table[i].in_use = FALSE;
+          CloseHandle(thread_table[i].handle);
+          BZERO(&thread_table[i].context, sizeof(CONTEXT));
+          continue;
+      }
       if (SuspendThread(thread_table[i].handle) == (DWORD)-1)
 	ABORT("SuspendThread failed");
       thread_table[i].suspended = TRUE;
@@ -61,7 +77,7 @@ ptr_t GC_current_stackbottom()
   ABORT("no thread table entry for current thread");
 }
 
-ptr_t GC_get_lo_stack_addr(ptr_t s)
+static ptr_t GC_get_lo_stack_addr(ptr_t s)
 {
     ptr_t bottom;
     MEMORY_BASIC_INFORMATION info;
@@ -81,7 +97,7 @@ void GC_push_all_stacks()
     if (thread_table[i].stack) {
       ptr_t bottom = GC_get_lo_stack_addr(thread_table[i].stack);
       if (thread_table[i].id == thread_id)
-	GC_push_all(&i, thread_table[i].stack);
+	GC_push_all_stack(&i, thread_table[i].stack);
       else {
 	thread_table[i].context.ContextFlags
 			= (CONTEXT_INTEGER|CONTEXT_CONTROL);