svn merge -r514:HEAD ^/branches/released .

25 files changed, 949 insertions, 1775 deletions

diff --git a/openssl/crypto/bn/Makefile b/openssl/crypto/bn/Makefile
index f5e8f65a4..aabc4f56b 100644
--- a/openssl/crypto/bn/Makefile
+++ b/openssl/crypto/bn/Makefile
@@ -12,8 +12,6 @@ MAKEFILE=	Makefile
 AR=		ar r
 
 BN_ASM=		bn_asm.o
-# or use
-#BN_ASM=	bn86-elf.o
 
 CFLAGS= $(INCLUDES) $(CFLAG)
 ASFLAGS= $(INCLUDES) $(ASFLAG)
@@ -28,13 +26,13 @@ LIBSRC=	bn_add.c bn_div.c bn_exp.c bn_lib.c bn_ctx.c bn_mul.c bn_mod.c \
 	bn_print.c bn_rand.c bn_shift.c bn_word.c bn_blind.c \
 	bn_kron.c bn_sqrt.c bn_gcd.c bn_prime.c bn_err.c bn_sqr.c bn_asm.c \
 	bn_recp.c bn_mont.c bn_mpi.c bn_exp2.c bn_gf2m.c bn_nist.c \
-	bn_depr.c bn_x931p.c bn_const.c bn_opt.c
+	bn_depr.c bn_const.c
 
 LIBOBJ=	bn_add.o bn_div.o bn_exp.o bn_lib.o bn_ctx.o bn_mul.o bn_mod.o \
 	bn_print.o bn_rand.o bn_shift.o bn_word.o bn_blind.o \
 	bn_kron.o bn_sqrt.o bn_gcd.o bn_prime.o bn_err.o bn_sqr.o $(BN_ASM) \
 	bn_recp.o bn_mont.o bn_mpi.o bn_exp2.o bn_gf2m.o bn_nist.o \
-	bn_depr.o bn_x931p.o bn_const.o bn_opt.o
+	bn_depr.o bn_const.o
 
 SRC= $(LIBSRC)
 
@@ -58,36 +56,25 @@ bnbug: bnbug.c ../../libcrypto.a top
 	cc -g -I../../include bnbug.c -o bnbug ../../libcrypto.a
 
 lib:	$(LIBOBJ)
-	$(ARX) $(LIB) $(LIBOBJ)
+	$(AR) $(LIB) $(LIBOBJ)
 	$(RANLIB) $(LIB) || echo Never mind.
 	@touch lib
 
-# ELF
-bn86-elf.s:	asm/bn-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) bn-586.pl elf $(CFLAGS) > ../$@)
-co86-elf.s:	asm/co-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) co-586.pl elf $(CFLAGS) > ../$@)
-mo86-elf.s:	asm/mo-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) mo-586.pl elf $(CFLAGS) > ../$@)
-# COFF
-bn86-cof.s: asm/bn-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) bn-586.pl coff $(CFLAGS) > ../$@)
-co86-cof.s: asm/co-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) co-586.pl coff $(CFLAGS) > ../$@)
-mo86-cof.s: asm/mo-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) mo-586.pl coff $(CFLAGS) > ../$@)
-# a.out
-bn86-out.s: asm/bn-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) bn-586.pl a.out $(CFLAGS) > ../$@)
-co86-out.s: asm/co-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) co-586.pl a.out $(CFLAGS) > ../$@)
-mo86-out.s: asm/mo-586.pl ../perlasm/x86asm.pl
-	(cd asm; $(PERL) mo-586.pl a.out $(CFLAGS) > ../$@)
+bn-586.s:	asm/bn-586.pl ../perlasm/x86asm.pl
+	$(PERL) asm/bn-586.pl $(PERLASM_SCHEME) $(CFLAGS) $(PROCESSOR) > $@
+co-586.s:	asm/co-586.pl ../perlasm/x86asm.pl
+	$(PERL) asm/co-586.pl $(PERLASM_SCHEME) $(CFLAGS) $(PROCESSOR) > $@
+x86-mont.s:	asm/x86-mont.pl ../perlasm/x86asm.pl
+	$(PERL) asm/x86-mont.pl $(PERLASM_SCHEME) $(CFLAGS) $(PROCESSOR) > $@
 
 sparcv8.o:	asm/sparcv8.S
 	$(CC) $(CFLAGS) -c asm/sparcv8.S
-sparcv8plus.o:	asm/sparcv8plus.S
-	$(CC) $(CFLAGS) -c asm/sparcv8plus.S
+bn-sparcv9.o:	asm/sparcv8plus.S
+	$(CC) $(CFLAGS) -c -o $@ asm/sparcv8plus.S
+sparcv9a-mont.s:	asm/sparcv9a-mont.pl
+	$(PERL) asm/sparcv9a-mont.pl $(CFLAGS) > $@
+sparcv9-mont.s:		asm/sparcv9-mont.pl
+	$(PERL) asm/sparcv9-mont.pl $(CFLAGS) > $@
 
 bn-mips3.o:	asm/mips3.s
 	@if [ "$(CC)" = "gcc" ]; then \
@@ -95,10 +82,13 @@ bn-mips3.o:	asm/mips3.s
 		as -$$ABI -O -o $@ asm/mips3.s; \
 	else	$(CC) -c $(CFLAGS) -o $@ asm/mips3.s; fi
 
+bn-s390x.o:	asm/s390x.S
+	$(CC) $(CFLAGS) -c -o $@ asm/s390x.S
+
 x86_64-gcc.o:	asm/x86_64-gcc.c
 	$(CC) $(CFLAGS) -c -o $@ asm/x86_64-gcc.c
 x86_64-mont.s:	asm/x86_64-mont.pl
-	$(PERL) asm/x86_64-mont.pl $@
+	$(PERL) asm/x86_64-mont.pl $(PERLASM_SCHEME) > $@
 
 bn-ia64.s:	asm/ia64.S
 	$(CC) $(CFLAGS) -E asm/ia64.S > $@
@@ -111,12 +101,14 @@ pa-risc2.o: asm/pa-risc2.s
 	/usr/ccs/bin/as -o pa-risc2.o asm/pa-risc2.s
 
 # ppc - AIX, Linux, MacOS X...
-linux_ppc32.s: asm/ppc.pl;	$(PERL) $< $@
-linux_ppc64.s: asm/ppc.pl;	$(PERL) $< $@
-aix_ppc32.s: asm/ppc.pl;	$(PERL) asm/ppc.pl $@
-aix_ppc64.s: asm/ppc.pl;	$(PERL) asm/ppc.pl $@
-osx_ppc32.s: asm/ppc.pl;	$(PERL) $< $@
-osx_ppc64.s: asm/ppc.pl;	$(PERL) $< $@
+bn-ppc.s:	asm/ppc.pl;	$(PERL) asm/ppc.pl $(PERLASM_SCHEME) $@
+ppc-mont.s:	asm/ppc-mont.pl;$(PERL) asm/ppc-mont.pl $(PERLASM_SCHEME) $@
+
+alpha-mont.s:	asm/alpha-mont.pl
+	$(PERL) $< | $(CC) -E - | tee $@ > /dev/null
+
+# GNU make "catch all"
+%-mont.s:	asm/%-mont.pl;	$(PERL) $< $(CFLAGS) > $@
 
 files:
 	$(PERL) $(TOP)/util/files.pl Makefile >> $(TOP)/MINFO
@@ -184,8 +176,11 @@ bn_blind.o: ../../include/openssl/lhash.h ../../include/openssl/opensslconf.h
 bn_blind.o: ../../include/openssl/opensslv.h ../../include/openssl/ossl_typ.h
 bn_blind.o: ../../include/openssl/safestack.h ../../include/openssl/stack.h
 bn_blind.o: ../../include/openssl/symhacks.h ../cryptlib.h bn_blind.c bn_lcl.h
-bn_const.o: ../../include/openssl/e_os2.h ../../include/openssl/opensslconf.h
-bn_const.o: ../../include/openssl/ossl_typ.h bn.h bn_const.c
+bn_const.o: ../../include/openssl/crypto.h ../../include/openssl/e_os2.h
+bn_const.o: ../../include/openssl/opensslconf.h
+bn_const.o: ../../include/openssl/opensslv.h ../../include/openssl/ossl_typ.h
+bn_const.o: ../../include/openssl/safestack.h ../../include/openssl/stack.h
+bn_const.o: ../../include/openssl/symhacks.h bn.h bn_const.c
 bn_ctx.o: ../../e_os.h ../../include/openssl/bio.h ../../include/openssl/bn.h
 bn_ctx.o: ../../include/openssl/buffer.h ../../include/openssl/crypto.h
 bn_ctx.o: ../../include/openssl/e_os2.h ../../include/openssl/err.h
@@ -292,13 +287,6 @@ bn_nist.o: ../../include/openssl/lhash.h ../../include/openssl/opensslconf.h
 bn_nist.o: ../../include/openssl/opensslv.h ../../include/openssl/ossl_typ.h
 bn_nist.o: ../../include/openssl/safestack.h ../../include/openssl/stack.h
 bn_nist.o: ../../include/openssl/symhacks.h ../cryptlib.h bn_lcl.h bn_nist.c
-bn_opt.o: ../../e_os.h ../../include/openssl/bio.h ../../include/openssl/bn.h
-bn_opt.o: ../../include/openssl/buffer.h ../../include/openssl/crypto.h
-bn_opt.o: ../../include/openssl/e_os2.h ../../include/openssl/err.h
-bn_opt.o: ../../include/openssl/lhash.h ../../include/openssl/opensslconf.h
-bn_opt.o: ../../include/openssl/opensslv.h ../../include/openssl/ossl_typ.h
-bn_opt.o: ../../include/openssl/safestack.h ../../include/openssl/stack.h
-bn_opt.o: ../../include/openssl/symhacks.h ../cryptlib.h bn_lcl.h bn_opt.c
 bn_prime.o: ../../e_os.h ../../include/openssl/bio.h ../../include/openssl/bn.h
 bn_prime.o: ../../include/openssl/buffer.h ../../include/openssl/crypto.h
 bn_prime.o: ../../include/openssl/e_os2.h ../../include/openssl/err.h
@@ -357,6 +345,3 @@ bn_word.o: ../../include/openssl/lhash.h ../../include/openssl/opensslconf.h
 bn_word.o: ../../include/openssl/opensslv.h ../../include/openssl/ossl_typ.h
 bn_word.o: ../../include/openssl/safestack.h ../../include/openssl/stack.h
 bn_word.o: ../../include/openssl/symhacks.h ../cryptlib.h bn_lcl.h bn_word.c
-bn_x931p.o: ../../include/openssl/bn.h ../../include/openssl/e_os2.h
-bn_x931p.o: ../../include/openssl/opensslconf.h
-bn_x931p.o: ../../include/openssl/ossl_typ.h bn_x931p.c
diff --git a/openssl/crypto/bn/asm/armv4-mont.pl b/openssl/crypto/bn/asm/armv4-mont.pl
index 05d5dc1a4..14e0d2d1d 100644
--- a/openssl/crypto/bn/asm/armv4-mont.pl
+++ b/openssl/crypto/bn/asm/armv4-mont.pl
@@ -193,6 +193,7 @@ bn_mul_mont:
 	bx	lr			@ interoperable with Thumb ISA:-)
 .size	bn_mul_mont,.-bn_mul_mont
 .asciz	"Montgomery multiplication for ARMv4, CRYPTOGAMS by <appro\@openssl.org>"
+.align	2
 ___
 
 $code =~ s/\bbx\s+lr\b/.word\t0xe12fff1e/gm;	# make it possible to compile with -march=armv4
diff --git a/openssl/crypto/bn/asm/bn-586.pl b/openssl/crypto/bn/asm/bn-586.pl
index 26c2685a7..332ef3e91 100644
--- a/openssl/crypto/bn/asm/bn-586.pl
+++ b/openssl/crypto/bn/asm/bn-586.pl
@@ -1,6 +1,7 @@
 #!/usr/local/bin/perl
 
-push(@INC,"perlasm","../../perlasm");
+$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
+push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
 &asm_init($ARGV[0],$0);
@@ -24,38 +25,25 @@ sub bn_mul_add_words
 	{
 	local($name)=@_;
 
-	&function_begin($name,$sse2?"EXTRN\t_OPENSSL_ia32cap_P:DWORD":"");
+	&function_begin_B($name,$sse2?"EXTRN\t_OPENSSL_ia32cap_P:DWORD":"");
 
-	&comment("");
-	$Low="eax";
-	$High="edx";
-	$a="ebx";
-	$w="ebp";
-	$r="edi";
-	$c="esi";
-
-	&xor($c,$c);		# clear carry
-	&mov($r,&wparam(0));	#
-
-	&mov("ecx",&wparam(2));	#
-	&mov($a,&wparam(1));	#
-
-	&and("ecx",0xfffffff8);	# num / 8
-	&mov($w,&wparam(3));	#
-
-	&push("ecx");		# Up the stack for a tmp variable
-
-	&jz(&label("maw_finish"));
+	$r="eax";
+	$a="edx";
+	$c="ecx";
 
 	if ($sse2) {
 		&picmeup("eax","OPENSSL_ia32cap_P");
 		&bt(&DWP(0,"eax"),26);
-		&jnc(&label("maw_loop"));
+		&jnc(&label("maw_non_sse2"));
 
-		&movd("mm0",$w);		# mm0 = w
+		&mov($r,&wparam(0));
+		&mov($a,&wparam(1));
+		&mov($c,&wparam(2));
+		&movd("mm0",&wparam(3));	# mm0 = w
 		&pxor("mm1","mm1");		# mm1 = carry_in
-
-		&set_label("maw_sse2_loop",0);
+		&jmp(&label("maw_sse2_entry"));
+		
+	&set_label("maw_sse2_unrolled",16);
 		&movd("mm3",&DWP(0,$r,"",0));	# mm3 = r[0]
 		&paddq("mm1","mm3");		# mm1 = carry_in + r[0]
 		&movd("mm2",&DWP(0,$a,"",0));	# mm2 = a[0]
@@ -112,42 +100,82 @@ sub bn_mul_add_words
 		&psrlq("mm1",32);		# mm1 = carry6
 		&paddq("mm1","mm3");		# mm1 = carry6 + r[7] + w*a[7]
 		&movd(&DWP(28,$r,"",0),"mm1");
-		&add($r,32);
+		&lea($r,&DWP(32,$r));
 		&psrlq("mm1",32);		# mm1 = carry_out
 
-		&sub("ecx",8);
+		&sub($c,8);
+		&jz(&label("maw_sse2_exit"));
+	&set_label("maw_sse2_entry");
+		&test($c,0xfffffff8);
+		&jnz(&label("maw_sse2_unrolled"));
+
+	&set_label("maw_sse2_loop",4);
+		&movd("mm2",&DWP(0,$a));	# mm2 = a[i]
+		&movd("mm3",&DWP(0,$r));	# mm3 = r[i]
+		&pmuludq("mm2","mm0");		# a[i] *= w
+		&lea($a,&DWP(4,$a));
+		&paddq("mm1","mm3");		# carry += r[i]
+		&paddq("mm1","mm2");		# carry += a[i]*w
+		&movd(&DWP(0,$r),"mm1");	# r[i] = carry_low
+		&sub($c,1);
+		&psrlq("mm1",32);		# carry = carry_high
+		&lea($r,&DWP(4,$r));
 		&jnz(&label("maw_sse2_loop"));
-
-		&movd($c,"mm1");		# c = carry_out
+	&set_label("maw_sse2_exit");
+		&movd("eax","mm1");		# c = carry_out
 		&emms();
+		&ret();
 
-		&jmp(&label("maw_finish"));
+	&set_label("maw_non_sse2",16);
 	}
 
-	&set_label("maw_loop",0);
+	# function_begin prologue
+	&push("ebp");
+	&push("ebx");
+	&push("esi");
+	&push("edi");
+
+	&comment("");
+	$Low="eax";
+	$High="edx";
+	$a="ebx";
+	$w="ebp";
+	$r="edi";
+	$c="esi";
+
+	&xor($c,$c);		# clear carry
+	&mov($r,&wparam(0));	#
+
+	&mov("ecx",&wparam(2));	#
+	&mov($a,&wparam(1));	#
+
+	&and("ecx",0xfffffff8);	# num / 8
+	&mov($w,&wparam(3));	#
 
-	&mov(&swtmp(0),"ecx");	#
+	&push("ecx");		# Up the stack for a tmp variable
+
+	&jz(&label("maw_finish"));
+
+	&set_label("maw_loop",16);
 
 	for ($i=0; $i<32; $i+=4)
 		{
 		&comment("Round $i");
 
-		 &mov("eax",&DWP($i,$a,"",0)); 	# *a
+		 &mov("eax",&DWP($i,$a)); 	# *a
 		&mul($w);			# *a * w
-		&add("eax",$c);		# L(t)+= *r
-		 &mov($c,&DWP($i,$r,"",0));	# L(t)+= *r
+		&add("eax",$c);			# L(t)+= c
 		&adc("edx",0);			# H(t)+=carry
-		 &add("eax",$c);		# L(t)+=c
+		 &add("eax",&DWP($i,$r));	# L(t)+= *r
 		&adc("edx",0);			# H(t)+=carry
-		 &mov(&DWP($i,$r,"",0),"eax");	# *r= L(t);
+		 &mov(&DWP($i,$r),"eax");	# *r= L(t);
 		&mov($c,"edx");			# c=  H(t);
 		}
 
 	&comment("");
-	&mov("ecx",&swtmp(0));	#
-	&add($a,32);
-	&add($r,32);
 	&sub("ecx",8);
+	&lea($a,&DWP(32,$a));
+	&lea($r,&DWP(32,$r));
 	&jnz(&label("maw_loop"));
 
 	&set_label("maw_finish",0);
@@ -160,16 +188,15 @@ sub bn_mul_add_words
 	for ($i=0; $i<7; $i++)
 		{
 		&comment("Tail Round $i");
-		 &mov("eax",&DWP($i*4,$a,"",0));# *a
+		 &mov("eax",&DWP($i*4,$a));	# *a
 		&mul($w);			# *a * w
 		&add("eax",$c);			# L(t)+=c
-		 &mov($c,&DWP($i*4,$r,"",0));	# L(t)+= *r
 		&adc("edx",0);			# H(t)+=carry
-		 &add("eax",$c);
+		 &add("eax",&DWP($i*4,$r));	# L(t)+= *r
 		&adc("edx",0);			# H(t)+=carry
 		 &dec("ecx") if ($i != 7-1);
-		&mov(&DWP($i*4,$r,"",0),"eax");	# *r= L(t);
-		 &mov($c,"edx");			# c=  H(t);
+		&mov(&DWP($i*4,$r),"eax");	# *r= L(t);
+		 &mov($c,"edx");		# c=  H(t);
 		&jz(&label("maw_end")) if ($i != 7-1);
 		}
 	&set_label("maw_end",0);
@@ -184,7 +211,45 @@ sub bn_mul_words
 	{
 	local($name)=@_;
 
-	&function_begin($name,"");
+	&function_begin_B($name,$sse2?"EXTRN\t_OPENSSL_ia32cap_P:DWORD":"");
+
+	$r="eax";
+	$a="edx";
+	$c="ecx";
+
+	if ($sse2) {
+		&picmeup("eax","OPENSSL_ia32cap_P");
+		&bt(&DWP(0,"eax"),26);
+		&jnc(&label("mw_non_sse2"));
+
+		&mov($r,&wparam(0));
+		&mov($a,&wparam(1));
+		&mov($c,&wparam(2));
+		&movd("mm0",&wparam(3));	# mm0 = w
+		&pxor("mm1","mm1");		# mm1 = carry = 0
+
+	&set_label("mw_sse2_loop",16);
+		&movd("mm2",&DWP(0,$a));	# mm2 = a[i]
+		&pmuludq("mm2","mm0");		# a[i] *= w
+		&lea($a,&DWP(4,$a));
+		&paddq("mm1","mm2");		# carry += a[i]*w
+		&movd(&DWP(0,$r),"mm1");	# r[i] = carry_low
+		&sub($c,1);
+		&psrlq("mm1",32);		# carry = carry_high
+		&lea($r,&DWP(4,$r));
+		&jnz(&label("mw_sse2_loop"));
+
+		&movd("eax","mm1");		# return carry
+		&emms();
+		&ret();
+	&set_label("mw_non_sse2",16);
+	}
+
+	# function_begin prologue
+	&push("ebp");
+	&push("ebx");
+	&push("esi");
+	&push("edi");
 
 	&comment("");
 	$Low="eax";
@@ -257,7 +322,40 @@ sub bn_sqr_words
 	{
 	local($name)=@_;
 
-	&function_begin($name,"");
+	&function_begin_B($name,$sse2?"EXTRN\t_OPENSSL_ia32cap_P:DWORD":"");
+
+	$r="eax";
+	$a="edx";
+	$c="ecx";
+
+	if ($sse2) {
+		&picmeup("eax","OPENSSL_ia32cap_P");
+		&bt(&DWP(0,"eax"),26);
+		&jnc(&label("sqr_non_sse2"));
+
+		&mov($r,&wparam(0));
+		&mov($a,&wparam(1));
+		&mov($c,&wparam(2));
+
+	&set_label("sqr_sse2_loop",16);
+		&movd("mm0",&DWP(0,$a));	# mm0 = a[i]
+		&pmuludq("mm0","mm0");		# a[i] *= a[i]
+		&lea($a,&DWP(4,$a));		# a++
+		&movq(&QWP(0,$r),"mm0");	# r[i] = a[i]*a[i]
+		&sub($c,1);
+		&lea($r,&DWP(8,$r));		# r += 2
+		&jnz(&label("sqr_sse2_loop"));
+
+		&emms();
+		&ret();
+	&set_label("sqr_non_sse2",16);
+	}
+
+	# function_begin prologue
+	&push("ebp");
+	&push("ebx");
+	&push("esi");
+	&push("edi");
 
 	&comment("");
 	$r="esi";
@@ -313,12 +411,13 @@ sub bn_div_words
 	{
 	local($name)=@_;
 
-	&function_begin($name,"");
+	&function_begin_B($name,"");
 	&mov("edx",&wparam(0));	#
 	&mov("eax",&wparam(1));	#
-	&mov("ebx",&wparam(2));	#
-	&div("ebx");
-	&function_end($name);
+	&mov("ecx",&wparam(2));	#
+	&div("ecx");
+	&ret();
+	&function_end_B($name);
 	}
 
 sub bn_add_words
diff --git a/openssl/crypto/bn/asm/co-586.pl b/openssl/crypto/bn/asm/co-586.pl
index 5d962cb95..57101a6bd 100644
--- a/openssl/crypto/bn/asm/co-586.pl
+++ b/openssl/crypto/bn/asm/co-586.pl
@@ -1,6 +1,7 @@
 #!/usr/local/bin/perl
 
-push(@INC,"perlasm","../../perlasm");
+$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
+push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
 &asm_init($ARGV[0],$0);
diff --git a/openssl/crypto/bn/asm/mo-586.pl b/openssl/crypto/bn/asm/mo-586.pl
deleted file mode 100644
index 098229309..000000000
--- a/openssl/crypto/bn/asm/mo-586.pl
+++ /dev/null
@@ -1,603 +0,0 @@
-#!/usr/bin/env perl
-
-# This is crypto/bn/asm/x86-mont.pl (with asciz from crypto/perlasm/x86asm.pl)
-# from OpenSSL 0.9.9-dev 
-
-sub ::asciz
-{ my @str=unpack("C*",shift);
-    push @str,0;
-    while ($#str>15) {
-	&data_byte(@str[0..15]);
-	foreach (0..15) { shift @str; }
-    }
-    &data_byte(@str) if (@str);
-}
-
-# ====================================================================
-# Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
-# project. The module is, however, dual licensed under OpenSSL and
-# CRYPTOGAMS licenses depending on where you obtain it. For further
-# details see http://www.openssl.org/~appro/cryptogams/.
-# ====================================================================
-
-# October 2005
-#
-# This is a "teaser" code, as it can be improved in several ways...
-# First of all non-SSE2 path should be implemented (yes, for now it
-# performs Montgomery multiplication/convolution only on SSE2-capable
-# CPUs such as P4, others fall down to original code). Then inner loop
-# can be unrolled and modulo-scheduled to improve ILP and possibly
-# moved to 128-bit XMM register bank (though it would require input
-# rearrangement and/or increase bus bandwidth utilization). Dedicated
-# squaring procedure should give further performance improvement...
-# Yet, for being draft, the code improves rsa512 *sign* benchmark by
-# 110%(!), rsa1024 one - by 70% and rsa4096 - by 20%:-)
-
-# December 2006
-#
-# Modulo-scheduling SSE2 loops results in further 15-20% improvement.
-# Integer-only code [being equipped with dedicated squaring procedure]
-# gives ~40% on rsa512 sign benchmark...
-
-push(@INC,"perlasm","../../perlasm");
-require "x86asm.pl";
-
-&asm_init($ARGV[0],$0);
-
-$sse2=0;
-for (@ARGV) { $sse2=1 if (/-DOPENSSL_IA32_SSE2/); }
-
-&external_label("OPENSSL_ia32cap_P") if ($sse2);
-
-&function_begin("bn_mul_mont");
-
-$i="edx";
-$j="ecx";
-$ap="esi";	$tp="esi";		# overlapping variables!!!
-$rp="edi";	$bp="edi";		# overlapping variables!!!
-$np="ebp";
-$num="ebx";
-
-$_num=&DWP(4*0,"esp");			# stack top layout
-$_rp=&DWP(4*1,"esp");
-$_ap=&DWP(4*2,"esp");
-$_bp=&DWP(4*3,"esp");
-$_np=&DWP(4*4,"esp");
-$_n0=&DWP(4*5,"esp");	$_n0q=&QWP(4*5,"esp");
-$_sp=&DWP(4*6,"esp");
-$_bpend=&DWP(4*7,"esp");
-$frame=32;				# size of above frame rounded up to 16n
-
-	&xor	("eax","eax");
-	&mov	("edi",&wparam(5));	# int num
-	&cmp	("edi",4);
-	&jl	(&label("just_leave"));
-
-	&lea	("esi",&wparam(0));	# put aside pointer to argument block
-	&lea	("edx",&wparam(1));	# load ap
-	&mov	("ebp","esp");		# saved stack pointer!
-	&add	("edi",2);		# extra two words on top of tp
-	&neg	("edi");
-	&lea	("esp",&DWP(-$frame,"esp","edi",4));	# alloca($frame+4*(num+2))
-	&neg	("edi");
-
-	# minimize cache contention by arraning 2K window between stack
-	# pointer and ap argument [np is also position sensitive vector,
-	# but it's assumed to be near ap, as it's allocated at ~same
-	# time].
-	&mov	("eax","esp");
-	&sub	("eax","edx");
-	&and	("eax",2047);
-	&sub	("esp","eax");		# this aligns sp and ap modulo 2048
-
-	&xor	("edx","esp");
-	&and	("edx",2048);
-	&xor	("edx",2048);
-	&sub	("esp","edx");		# this splits them apart modulo 4096
-
-	&and	("esp",-64);		# align to cache line
-
-	################################# load argument block...
-	&mov	("eax",&DWP(0*4,"esi"));# BN_ULONG *rp
-	&mov	("ebx",&DWP(1*4,"esi"));# const BN_ULONG *ap
-	&mov	("ecx",&DWP(2*4,"esi"));# const BN_ULONG *bp
-	&mov	("edx",&DWP(3*4,"esi"));# const BN_ULONG *np
-	&mov	("esi",&DWP(4*4,"esi"));# const BN_ULONG *n0
-	#&mov	("edi",&DWP(5*4,"esi"));# int num
-
-	&mov	("esi",&DWP(0,"esi"));	# pull n0[0]
-	&mov	($_rp,"eax");		# ... save a copy of argument block
-	&mov	($_ap,"ebx");
-	&mov	($_bp,"ecx");
-	&mov	($_np,"edx");
-	&mov	($_n0,"esi");
-	&lea	($num,&DWP(-3,"edi"));	# num=num-1 to assist modulo-scheduling
-	#&mov	($_num,$num);		# redundant as $num is not reused
-	&mov	($_sp,"ebp");		# saved stack pointer!
-
-if($sse2) {
-$acc0="mm0";	# mmx register bank layout
-$acc1="mm1";
-$car0="mm2";
-$car1="mm3";
-$mul0="mm4";
-$mul1="mm5";
-$temp="mm6";
-$mask="mm7";
-
-	&picmeup("eax","OPENSSL_ia32cap_P");
-	&bt	(&DWP(0,"eax"),26);
-	&jnc	(&label("non_sse2"));
-
-	&mov	("eax",-1);
-	&movd	($mask,"eax");		# mask 32 lower bits
-
-	&mov	($ap,$_ap);		# load input pointers
-	&mov	($bp,$_bp);
-	&mov	($np,$_np);
-
-	&xor	($i,$i);		# i=0
-	&xor	($j,$j);		# j=0
-
-	&movd	($mul0,&DWP(0,$bp));		# bp[0]
-	&movd	($mul1,&DWP(0,$ap));		# ap[0]
-	&movd	($car1,&DWP(0,$np));		# np[0]
-
-	&pmuludq($mul1,$mul0);			# ap[0]*bp[0]
-	&movq	($car0,$mul1);
-	&movq	($acc0,$mul1);			# I wish movd worked for
-	&pand	($acc0,$mask);			# inter-register transfers
-
-	&pmuludq($mul1,$_n0q);			# *=n0
-
-	&pmuludq($car1,$mul1);			# "t[0]"*np[0]*n0
-	&paddq	($car1,$acc0);
-
-	&movd	($acc1,&DWP(4,$np));		# np[1]
-	&movd	($acc0,&DWP(4,$ap));		# ap[1]
-
-	&psrlq	($car0,32);
-	&psrlq	($car1,32);
-
-	&inc	($j);				# j++
-&set_label("1st",16);
-	&pmuludq($acc0,$mul0);			# ap[j]*bp[0]
-	&pmuludq($acc1,$mul1);			# np[j]*m1
-	&paddq	($car0,$acc0);			# +=c0
-	&paddq	($car1,$acc1);			# +=c1
-
-	&movq	($acc0,$car0);
-	&pand	($acc0,$mask);
-	&movd	($acc1,&DWP(4,$np,$j,4));	# np[j+1]
-	&paddq	($car1,$acc0);			# +=ap[j]*bp[0];
-	&movd	($acc0,&DWP(4,$ap,$j,4));	# ap[j+1]
-	&psrlq	($car0,32);
-	&movd	(&DWP($frame-4,"esp",$j,4),$car1);	# tp[j-1]=
-	&psrlq	($car1,32);
-
-	&lea	($j,&DWP(1,$j));
-	&cmp	($j,$num);
-	&jl	(&label("1st"));
-
-	&pmuludq($acc0,$mul0);			# ap[num-1]*bp[0]
-	&pmuludq($acc1,$mul1);			# np[num-1]*m1
-	&paddq	($car0,$acc0);			# +=c0
-	&paddq	($car1,$acc1);			# +=c1
-
-	&movq	($acc0,$car0);
-	&pand	($acc0,$mask);
-	&paddq	($car1,$acc0);			# +=ap[num-1]*bp[0];
-	&movd	(&DWP($frame-4,"esp",$j,4),$car1);	# tp[num-2]=
-
-	&psrlq	($car0,32);
-	&psrlq	($car1,32);
-
-	&paddq	($car1,$car0);
-	&movq	(&QWP($frame,"esp",$num,4),$car1);	# tp[num].tp[num-1]
-
-	&inc	($i);				# i++
-&set_label("outer");
-	&xor	($j,$j);			# j=0
-
-	&movd	($mul0,&DWP(0,$bp,$i,4));	# bp[i]
-	&movd	($mul1,&DWP(0,$ap));		# ap[0]
-	&movd	($temp,&DWP($frame,"esp"));	# tp[0]
-	&movd	($car1,&DWP(0,$np));		# np[0]
-	&pmuludq($mul1,$mul0);			# ap[0]*bp[i]
-
-	&paddq	($mul1,$temp);			# +=tp[0]
-	&movq	($acc0,$mul1);
-	&movq	($car0,$mul1);
-	&pand	($acc0,$mask);
-
-	&pmuludq($mul1,$_n0q);			# *=n0
-
-	&pmuludq($car1,$mul1);
-	&paddq	($car1,$acc0);
-
-	&movd	($temp,&DWP($frame+4,"esp"));	# tp[1]
-	&movd	($acc1,&DWP(4,$np));		# np[1]
-	&movd	($acc0,&DWP(4,$ap));		# ap[1]
-
-	&psrlq	($car0,32);
-	&psrlq	($car1,32);
-	&paddq	($car0,$temp);			# +=tp[1]
-
-	&inc	($j);				# j++
-	&dec	($num);
-&set_label("inner");
-	&pmuludq($acc0,$mul0);			# ap[j]*bp[i]
-	&pmuludq($acc1,$mul1);			# np[j]*m1
-	&paddq	($car0,$acc0);			# +=c0
-	&paddq	($car1,$acc1);			# +=c1
-
-	&movq	($acc0,$car0);
-	&movd	($temp,&DWP($frame+4,"esp",$j,4));# tp[j+1]
-	&pand	($acc0,$mask);
-	&movd	($acc1,&DWP(4,$np,$j,4));	# np[j+1]
-	&paddq	($car1,$acc0);			# +=ap[j]*bp[i]+tp[j]
-	&movd	($acc0,&DWP(4,$ap,$j,4));	# ap[j+1]
-	&psrlq	($car0,32);
-	&movd	(&DWP($frame-4,"esp",$j,4),$car1);# tp[j-1]=
-	&psrlq	($car1,32);
-	&paddq	($car0,$temp);			# +=tp[j+1]
-
-	&dec	($num);
-	&lea	($j,&DWP(1,$j));		# j++
-	&jnz	(&label("inner"));
-
-	&mov	($num,$j);
-	&pmuludq($acc0,$mul0);			# ap[num-1]*bp[i]
-	&pmuludq($acc1,$mul1);			# np[num-1]*m1
-	&paddq	($car0,$acc0);			# +=c0
-	&paddq	($car1,$acc1);			# +=c1
-
-	&movq	($acc0,$car0);
-	&pand	($acc0,$mask);
-	&paddq	($car1,$acc0);			# +=ap[num-1]*bp[i]+tp[num-1]
-	&movd	(&DWP($frame-4,"esp",$j,4),$car1);	# tp[num-2]=
-	&psrlq	($car0,32);
-	&psrlq	($car1,32);
-
-	&movd	($temp,&DWP($frame+4,"esp",$num,4));	# += tp[num]
-	&paddq	($car1,$car0);
-	&paddq	($car1,$temp);
-	&movq	(&QWP($frame,"esp",$num,4),$car1);	# tp[num].tp[num-1]
-
-	&lea	($i,&DWP(1,$i));		# i++
-	&cmp	($i,$num);
-	&jle	(&label("outer"));
-
-	&emms	();				# done with mmx bank
-	&jmp	(&label("common_tail"));
-
-&set_label("non_sse2",16);
-}
-
-if (0) {
-	&mov	("esp",$_sp);
-	&xor	("eax","eax");	# signal "not fast enough [yet]"
-	&jmp	(&label("just_leave"));
-	# While the below code provides competitive performance for
-	# all key lengthes on modern Intel cores, it's still more
-	# than 10% slower for 4096-bit key elsewhere:-( "Competitive"
-	# means compared to the original integer-only assembler.
-	# 512-bit RSA sign is better by ~40%, but that's about all
-	# one can say about all CPUs...
-} else {
-$inp="esi";	# integer path uses these registers differently
-$word="edi";
-$carry="ebp";
-
-	&mov	($inp,$_ap);
-	&lea	($carry,&DWP(1,$num));
-	&mov	($word,$_bp);
-	&xor	($j,$j);				# j=0
-	&mov	("edx",$inp);
-	&and	($carry,1);				# see if num is even
-	&sub	("edx",$word);				# see if ap==bp
-	&lea	("eax",&DWP(4,$word,$num,4));		# &bp[num]
-	&or	($carry,"edx");
-	&mov	($word,&DWP(0,$word));			# bp[0]
-	&jz	(&label("bn_sqr_mont"));
-	&mov	($_bpend,"eax");
-	&mov	("eax",&DWP(0,$inp));
-	&xor	("edx","edx");
-
-&set_label("mull",16);
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[j]*bp[0]
-	&add	($carry,"eax");
-	&lea	($j,&DWP(1,$j));
-	&adc	("edx",0);
-	&mov	("eax",&DWP(0,$inp,$j,4));		# ap[j+1]
-	&cmp	($j,$num);
-	&mov	(&DWP($frame-4,"esp",$j,4),$carry);	# tp[j]=
-	&jl	(&label("mull"));
-
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[num-1]*bp[0]
-	 &mov	($word,$_n0);
-	&add	("eax",$carry);
-	 &mov	($inp,$_np);
-	&adc	("edx",0);
-	 &imul	($word,&DWP($frame,"esp"));		# n0*tp[0]
-
-	&mov	(&DWP($frame,"esp",$num,4),"eax");	# tp[num-1]=
-	&xor	($j,$j);
-	&mov	(&DWP($frame+4,"esp",$num,4),"edx");	# tp[num]=
-	&mov	(&DWP($frame+8,"esp",$num,4),$j);	# tp[num+1]=
-
-	&mov	("eax",&DWP(0,$inp));			# np[0]
-	&mul	($word);				# np[0]*m
-	&add	("eax",&DWP($frame,"esp"));		# +=tp[0]
-	&mov	("eax",&DWP(4,$inp));			# np[1]
-	&adc	("edx",0);
-	&inc	($j);
-
-	&jmp	(&label("2ndmadd"));
-
-&set_label("1stmadd",16);
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[j]*bp[i]
-	&add	($carry,&DWP($frame,"esp",$j,4));	# +=tp[j]
-	&lea	($j,&DWP(1,$j));
-	&adc	("edx",0);
-	&add	($carry,"eax");
-	&mov	("eax",&DWP(0,$inp,$j,4));		# ap[j+1]
-	&adc	("edx",0);
-	&cmp	($j,$num);
-	&mov	(&DWP($frame-4,"esp",$j,4),$carry);	# tp[j]=
-	&jl	(&label("1stmadd"));
-
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[num-1]*bp[i]
-	&add	("eax",&DWP($frame,"esp",$num,4));	# +=tp[num-1]
-	 &mov	($word,$_n0);
-	&adc	("edx",0);
-	 &mov	($inp,$_np);
-	&add	($carry,"eax");
-	&adc	("edx",0);
-	 &imul	($word,&DWP($frame,"esp"));		# n0*tp[0]
-
-	&xor	($j,$j);
-	&add	("edx",&DWP($frame+4,"esp",$num,4));	# carry+=tp[num]
-	&mov	(&DWP($frame,"esp",$num,4),$carry);	# tp[num-1]=
-	&adc	($j,0);
-	 &mov	("eax",&DWP(0,$inp));			# np[0]
-	&mov	(&DWP($frame+4,"esp",$num,4),"edx");	# tp[num]=
-	&mov	(&DWP($frame+8,"esp",$num,4),$j);	# tp[num+1]=
-
-	&mul	($word);				# np[0]*m
-	&add	("eax",&DWP($frame,"esp"));		# +=tp[0]
-	&mov	("eax",&DWP(4,$inp));			# np[1]
-	&adc	("edx",0);
-	&mov	($j,1);
-
-&set_label("2ndmadd",16);
-	&mov	($carry,"edx");
-	&mul	($word);				# np[j]*m
-	&add	($carry,&DWP($frame,"esp",$j,4));	# +=tp[j]
-	&lea	($j,&DWP(1,$j));
-	&adc	("edx",0);
-	&add	($carry,"eax");
-	&mov	("eax",&DWP(0,$inp,$j,4));		# np[j+1]
-	&adc	("edx",0);
-	&cmp	($j,$num);
-	&mov	(&DWP($frame-8,"esp",$j,4),$carry);	# tp[j-1]=
-	&jl	(&label("2ndmadd"));
-
-	&mov	($carry,"edx");
-	&mul	($word);				# np[j]*m
-	&add	($carry,&DWP($frame,"esp",$num,4));	# +=tp[num-1]
-	&adc	("edx",0);
-	&add	($carry,"eax");
-	&adc	("edx",0);
-	&mov	(&DWP($frame-4,"esp",$num,4),$carry);	# tp[num-2]=
-
-	&xor	("eax","eax");
-	 &mov	($j,$_bp);				# &bp[i]
-	&add	("edx",&DWP($frame+4,"esp",$num,4));	# carry+=tp[num]
-	&adc	("eax",&DWP($frame+8,"esp",$num,4));	# +=tp[num+1]
-	 &lea	($j,&DWP(4,$j));
-	&mov	(&DWP($frame,"esp",$num,4),"edx");	# tp[num-1]=
-	 &cmp	($j,$_bpend);
-	&mov	(&DWP($frame+4,"esp",$num,4),"eax");	# tp[num]=
-	&je	(&label("common_tail"));
-
-	&mov	($word,&DWP(0,$j));			# bp[i+1]
-	&mov	($inp,$_ap);
-	&mov	($_bp,$j);				# &bp[++i]
-	&xor	($j,$j);
-	&xor	("edx","edx");
-	&mov	("eax",&DWP(0,$inp));
-	&jmp	(&label("1stmadd"));
-
-&set_label("bn_sqr_mont",16);
-$sbit=$num;
-	&mov	($_num,$num);
-	&mov	($_bp,$j);				# i=0
-
-	&mov	("eax",$word);				# ap[0]
-	&mul	($word);				# ap[0]*ap[0]
-	&mov	(&DWP($frame,"esp"),"eax");		# tp[0]=
-	&mov	($sbit,"edx");
-	&shr	("edx",1);
-	&and	($sbit,1);
-	&inc	($j);
-&set_label("sqr",16);
-	&mov	("eax",&DWP(0,$inp,$j,4));		# ap[j]
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[j]*ap[0]
-	&add	("eax",$carry);
-	&lea	($j,&DWP(1,$j));
-	&adc	("edx",0);
-	&lea	($carry,&DWP(0,$sbit,"eax",2));
-	&shr	("eax",31);
-	&cmp	($j,$_num);
-	&mov	($sbit,"eax");
-	&mov	(&DWP($frame-4,"esp",$j,4),$carry);	# tp[j]=
-	&jl	(&label("sqr"));
-
-	&mov	("eax",&DWP(0,$inp,$j,4));		# ap[num-1]
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[num-1]*ap[0]
-	&add	("eax",$carry);
-	 &mov	($word,$_n0);
-	&adc	("edx",0);
-	 &mov	($inp,$_np);
-	&lea	($carry,&DWP(0,$sbit,"eax",2));
-	 &imul	($word,&DWP($frame,"esp"));		# n0*tp[0]
-	&shr	("eax",31);
-	&mov	(&DWP($frame,"esp",$j,4),$carry);	# tp[num-1]=
-
-	&lea	($carry,&DWP(0,"eax","edx",2));
-	 &mov	("eax",&DWP(0,$inp));			# np[0]
-	&shr	("edx",31);
-	&mov	(&DWP($frame+4,"esp",$j,4),$carry);	# tp[num]=
-	&mov	(&DWP($frame+8,"esp",$j,4),"edx");	# tp[num+1]=
-
-	&mul	($word);				# np[0]*m
-	&add	("eax",&DWP($frame,"esp"));		# +=tp[0]
-	&mov	($num,$j);
-	&adc	("edx",0);
-	&mov	("eax",&DWP(4,$inp));			# np[1]
-	&mov	($j,1);
-
-&set_label("3rdmadd",16);
-	&mov	($carry,"edx");
-	&mul	($word);				# np[j]*m
-	&add	($carry,&DWP($frame,"esp",$j,4));	# +=tp[j]
-	&adc	("edx",0);
-	&add	($carry,"eax");
-	&mov	("eax",&DWP(4,$inp,$j,4));		# np[j+1]
-	&adc	("edx",0);
-	&mov	(&DWP($frame-4,"esp",$j,4),$carry);	# tp[j-1]=
-
-	&mov	($carry,"edx");
-	&mul	($word);				# np[j+1]*m
-	&add	($carry,&DWP($frame+4,"esp",$j,4));	# +=tp[j+1]
-	&lea	($j,&DWP(2,$j));
-	&adc	("edx",0);
-	&add	($carry,"eax");
-	&mov	("eax",&DWP(0,$inp,$j,4));		# np[j+2]
-	&adc	("edx",0);
-	&cmp	($j,$num);
-	&mov	(&DWP($frame-8,"esp",$j,4),$carry);	# tp[j]=
-	&jl	(&label("3rdmadd"));
-
-	&mov	($carry,"edx");
-	&mul	($word);				# np[j]*m
-	&add	($carry,&DWP($frame,"esp",$num,4));	# +=tp[num-1]
-	&adc	("edx",0);
-	&add	($carry,"eax");
-	&adc	("edx",0);
-	&mov	(&DWP($frame-4,"esp",$num,4),$carry);	# tp[num-2]=
-
-	&mov	($j,$_bp);				# i
-	&xor	("eax","eax");
-	&mov	($inp,$_ap);
-	&add	("edx",&DWP($frame+4,"esp",$num,4));	# carry+=tp[num]
-	&adc	("eax",&DWP($frame+8,"esp",$num,4));	# +=tp[num+1]
-	&mov	(&DWP($frame,"esp",$num,4),"edx");	# tp[num-1]=
-	&cmp	($j,$num);
-	&mov	(&DWP($frame+4,"esp",$num,4),"eax");	# tp[num]=
-	&je	(&label("common_tail"));
-
-	&mov	($word,&DWP(4,$inp,$j,4));		# ap[i]
-	&lea	($j,&DWP(1,$j));
-	&mov	("eax",$word);
-	&mov	($_bp,$j);				# ++i
-	&mul	($word);				# ap[i]*ap[i]
-	&add	("eax",&DWP($frame,"esp",$j,4));	# +=tp[i]
-	&adc	("edx",0);
-	&mov	(&DWP($frame,"esp",$j,4),"eax");	# tp[i]=
-	&xor	($carry,$carry);
-	&cmp	($j,$num);
-	&lea	($j,&DWP(1,$j));
-	&je	(&label("sqrlast"));
-
-	&mov	($sbit,"edx");				# zaps $num
-	&shr	("edx",1);
-	&and	($sbit,1);
-&set_label("sqradd",16);
-	&mov	("eax",&DWP(0,$inp,$j,4));		# ap[j]
-	&mov	($carry,"edx");
-	&mul	($word);				# ap[j]*ap[i]
-	&add	("eax",$carry);
-	&lea	($carry,&DWP(0,"eax","eax"));
-	&adc	("edx",0);
-	&shr	("eax",31);
-	&add	($carry,&DWP($frame,"esp",$j,4));	# +=tp[j]
-	&lea	($j,&DWP(1,$j));
-	&adc	("eax",0);
-	&add	($carry,$sbit);
-	&adc	("eax",0);
-	&cmp	($j,$_num);
-	&mov	(&DWP($frame-4,"esp",$j,4),$carry);	# tp[j]=
-	&mov	($sbit,"eax");
-	&jle	(&label("sqradd"));
-
-	&mov	($carry,"edx");
-	&lea	("edx",&DWP(0,$sbit,"edx",2));
-	&shr	($carry,31);
-&set_label("sqrlast");
-	&mov	($word,$_n0);
-	&mov	($inp,$_np);
-	&imul	($word,&DWP($frame,"esp"));		# n0*tp[0]
-
-	&add	("edx",&DWP($frame,"esp",$j,4));	# +=tp[num]
-	&mov	("eax",&DWP(0,$inp));			# np[0]
-	&adc	($carry,0);
-	&mov	(&DWP($frame,"esp",$j,4),"edx");	# tp[num]=
-	&mov	(&DWP($frame+4,"esp",$j,4),$carry);	# tp[num+1]=
-
-	&mul	($word);				# np[0]*m
-	&add	("eax",&DWP($frame,"esp"));		# +=tp[0]
-	&lea	($num,&DWP(-1,$j));
-	&adc	("edx",0);
-	&mov	($j,1);
-	&mov	("eax",&DWP(4,$inp));			# np[1]
-
-	&jmp	(&label("3rdmadd"));
-}
-
-&set_label("common_tail",16);
-	&mov	($np,$_np);			# load modulus pointer
-	&mov	($rp,$_rp);			# load result pointer
-	&lea	($tp,&DWP($frame,"esp"));	# [$ap and $bp are zapped]
-
-	&mov	("eax",&DWP(0,$tp));		# tp[0]
-	&mov	($j,$num);			# j=num-1
-	&xor	($i,$i);			# i=0 and clear CF!
-
-&set_label("sub",16);
-	&sbb	("eax",&DWP(0,$np,$i,4));
-	&mov	(&DWP(0,$rp,$i,4),"eax");	# rp[i]=tp[i]-np[i]
-	&dec	($j);				# doesn't affect CF!
-	&mov	("eax",&DWP(4,$tp,$i,4));	# tp[i+1]
-	&lea	($i,&DWP(1,$i));		# i++
-	&jge	(&label("sub"));
-
-	&sbb	("eax",0);			# handle upmost overflow bit
-	&and	($tp,"eax");
-	&not	("eax");
-	&mov	($np,$rp);
-	&and	($np,"eax");
-	&or	($tp,$np);			# tp=carry?tp:rp
-
-&set_label("copy",16);				# copy or in-place refresh
-	&mov	("eax",&DWP(0,$tp,$num,4));
-	&mov	(&DWP(0,$rp,$num,4),"eax");	# rp[i]=tp[i]
-	&mov	(&DWP($frame,"esp",$num,4),$j);	# zap temporary vector
-	&dec	($num);
-	&jge	(&label("copy"));
-
-	&mov	("esp",$_sp);		# pull saved stack pointer
-	&mov	("eax",1);
-&set_label("just_leave");
-&function_end("bn_mul_mont");
-
-&asciz("Montgomery Multiplication for x86, CRYPTOGAMS by <appro\@openssl.org>");
-
-&asm_finish();
diff --git a/openssl/crypto/bn/asm/ppc.pl b/openssl/crypto/bn/asm/ppc.pl
index 08e005347..37c65d351 100644
--- a/openssl/crypto/bn/asm/ppc.pl
+++ b/openssl/crypto/bn/asm/ppc.pl
@@ -100,9 +100,9 @@
 #	me a note at schari@us.ibm.com
 #
 
-$opf = shift;
+$flavour = shift;
 
-if ($opf =~ /32\.s/) {
+if ($flavour =~ /32/) {
 	$BITS=	32;
 	$BNSZ=	$BITS/8;
 	$ISA=	"\"ppc\"";
@@ -125,7 +125,7 @@ if ($opf =~ /32\.s/) {
 	$INSR=	"insrwi";	# insert right
 	$ROTL=	"rotlwi";	# rotate left by immediate
 	$TR=	"tw";		# conditional trap
-} elsif ($opf =~ /64\.s/) {
+} elsif ($flavour =~ /64/) {
 	$BITS=	64;
 	$BNSZ=	$BITS/8;
 	$ISA=	"\"ppc64\"";
@@ -149,93 +149,16 @@ if ($opf =~ /32\.s/) {
 	$INSR=	"insrdi";	# insert right 
 	$ROTL=	"rotldi";	# rotate left by immediate
 	$TR=	"td";		# conditional trap
-} else { die "nonsense $opf"; }
+} else { die "nonsense $flavour"; }
 
-( defined shift || open STDOUT,">$opf" ) || die "can't open $opf: $!";
+$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
+( $xlate="${dir}ppc-xlate.pl" and -f $xlate ) or
+( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or
+die "can't locate ppc-xlate.pl";
 
-# function entry points from the AIX code
-#
-# There are other, more elegant, ways to handle this. We (IBM) chose
-# this approach as it plays well with scripts we run to 'namespace'
-# OpenSSL .i.e. we add a prefix to all the public symbols so we can
-# co-exist in the same process with other implementations of OpenSSL.
-# 'cleverer' ways of doing these substitutions tend to hide data we
-# need to be obvious.
-#
-my @items = ("bn_sqr_comba4",
-	     "bn_sqr_comba8",
-	     "bn_mul_comba4",
-	     "bn_mul_comba8",
-	     "bn_sub_words",
-	     "bn_add_words",
-	     "bn_div_words",
-	     "bn_sqr_words",
-	     "bn_mul_words",
-	     "bn_mul_add_words");
+open STDOUT,"| $^X $xlate $flavour ".shift || die "can't call $xlate: $!";
 
-if    ($opf =~ /linux/)	{  do_linux();	}
-elsif ($opf =~ /aix/)	{  do_aix();	}
-elsif ($opf =~ /osx/)	{  do_osx();	}
-else			{  do_bsd();	}
-
-sub do_linux {
-    $d=&data();
-
-    if ($BITS==64) {
-      foreach $t (@items) {
-        $d =~ s/\.$t:/\
-\t.section\t".opd","aw"\
-\t.align\t3\
-\t.globl\t$t\
-$t:\
-\t.quad\t.$t,.TOC.\@tocbase,0\
-\t.size\t$t,24\
-\t.previous\n\
-\t.type\t.$t,\@function\
-\t.globl\t.$t\
-.$t:/g;
-      }
-    }
-    else {
-      foreach $t (@items) {
-        $d=~s/\.$t/$t/g;
-      }
-    }
-    # hide internal labels to avoid pollution of name table...
-    $d=~s/Lppcasm_/.Lppcasm_/gm;
-    print $d;
-}
-
-sub do_aix {
-    # AIX assembler is smart enough to please the linker without
-    # making us do something special...
-    print &data();
-}
-
-# MacOSX 32 bit
-sub do_osx {
-    $d=&data();
-    # Change the bn symbol prefix from '.' to '_'
-    foreach $t (@items) {
-      $d=~s/\.$t/_$t/g;
-    }
-    # Change .machine to something OS X asm will accept
-    $d=~s/\.machine.*/.text/g;
-    $d=~s/\#/;/g; # change comment from '#' to ';'
-    print $d;
-}
-
-# BSD (Untested)
-sub do_bsd {
-    $d=&data();
-    foreach $t (@items) {
-      $d=~s/\.$t/_$t/g;
-    }
-    print $d;
-}
-
-sub data {
-	local($data)=<<EOF;
+$data=<<EOF;
 #--------------------------------------------------------------------
 #
 #
@@ -297,33 +220,20 @@ sub data {
 #
 #	Defines to be used in the assembly code.
 #	
-.set r0,0	# we use it as storage for value of 0
-.set SP,1	# preserved
-.set RTOC,2	# preserved 
-.set r3,3	# 1st argument/return value
-.set r4,4	# 2nd argument/volatile register
-.set r5,5	# 3rd argument/volatile register
-.set r6,6	# ...
-.set r7,7
-.set r8,8
-.set r9,9
-.set r10,10
-.set r11,11
-.set r12,12
-.set r13,13	# not used, nor any other "below" it...
-
-.set BO_IF_NOT,4
-.set BO_IF,12
-.set BO_dCTR_NZERO,16
-.set BO_dCTR_ZERO,18
-.set BO_ALWAYS,20
-.set CR0_LT,0;
-.set CR0_GT,1;
-.set CR0_EQ,2
-.set CR1_FX,4;
-.set CR1_FEX,5;
-.set CR1_VX,6
-.set LR,8
+#.set r0,0	# we use it as storage for value of 0
+#.set SP,1	# preserved
+#.set RTOC,2	# preserved 
+#.set r3,3	# 1st argument/return value
+#.set r4,4	# 2nd argument/volatile register
+#.set r5,5	# 3rd argument/volatile register
+#.set r6,6	# ...
+#.set r7,7
+#.set r8,8
+#.set r9,9
+#.set r10,10
+#.set r11,11
+#.set r12,12
+#.set r13,13	# not used, nor any other "below" it...
 
 #	Declare function names to be global
 #	NOTE:	For gcc these names MUST be changed to remove
@@ -344,7 +254,7 @@ sub data {
 	
 # .text section
 	
-	.machine	$ISA
+	.machine	"any"
 
 #
 #	NOTE:	The following label name should be changed to
@@ -478,7 +388,7 @@ sub data {
 
 	$ST		r9,`6*$BNSZ`(r3)	#r[6]=c1
 	$ST		r10,`7*$BNSZ`(r3)	#r[7]=c2
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 #
@@ -903,7 +813,7 @@ sub data {
 	$ST		r9, `15*$BNSZ`(r3)	#r[15]=c1;
 
 
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 
 	.long	0x00000000
 
@@ -1055,7 +965,7 @@ sub data {
 
 	$ST	r10,`6*$BNSZ`(r3)	#r[6]=c1
 	$ST	r11,`7*$BNSZ`(r3)	#r[7]=c2
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 #
@@ -1591,7 +1501,7 @@ sub data {
 	adde	r10,r10,r9
 	$ST	r12,`14*$BNSZ`(r3)	#r[14]=c3;
 	$ST	r10,`15*$BNSZ`(r3)	#r[15]=c1;
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 #
@@ -1623,7 +1533,7 @@ sub data {
 	subfc.	r7,r0,r6        # If r6 is 0 then result is 0.
 				# if r6 > 0 then result !=0
 				# In either case carry bit is set.
-	bc	BO_IF,CR0_EQ,Lppcasm_sub_adios
+	beq	Lppcasm_sub_adios
 	addi	r4,r4,-$BNSZ
 	addi	r3,r3,-$BNSZ
 	addi	r5,r5,-$BNSZ
@@ -1635,11 +1545,11 @@ Lppcasm_sub_mainloop:
 				# if carry = 1 this is r7-r8. Else it
 				# is r7-r8 -1 as we need.
 	$STU	r6,$BNSZ(r3)
-	bc	BO_dCTR_NZERO,CR0_EQ,Lppcasm_sub_mainloop
+	bdnz-	Lppcasm_sub_mainloop
 Lppcasm_sub_adios:	
 	subfze	r3,r0		# if carry bit is set then r3 = 0 else -1
 	andi.	r3,r3,1         # keep only last bit.
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 
@@ -1670,7 +1580,7 @@ Lppcasm_sub_adios:
 #	check for r6 = 0. Is this needed?
 #
 	addic.	r6,r6,0		#test r6 and clear carry bit.
-	bc	BO_IF,CR0_EQ,Lppcasm_add_adios
+	beq	Lppcasm_add_adios
 	addi	r4,r4,-$BNSZ
 	addi	r3,r3,-$BNSZ
 	addi	r5,r5,-$BNSZ
@@ -1680,10 +1590,10 @@ Lppcasm_add_mainloop:
 	$LDU	r8,$BNSZ(r5)
 	adde	r8,r7,r8
 	$STU	r8,$BNSZ(r3)
-	bc	BO_dCTR_NZERO,CR0_EQ,Lppcasm_add_mainloop
+	bdnz-	Lppcasm_add_mainloop
 Lppcasm_add_adios:	
 	addze	r3,r0			#return carry bit.
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 #
@@ -1707,24 +1617,24 @@ Lppcasm_add_adios:
 #	r5 = d
 	
 	$UCMPI	0,r5,0			# compare r5 and 0
-	bc	BO_IF_NOT,CR0_EQ,Lppcasm_div1	# proceed if d!=0
+	bne	Lppcasm_div1		# proceed if d!=0
 	li	r3,-1			# d=0 return -1
-	bclr	BO_ALWAYS,CR0_LT	
+	blr
 Lppcasm_div1:
 	xor	r0,r0,r0		#r0=0
 	li	r8,$BITS
 	$CNTLZ.	r7,r5			#r7 = num leading 0s in d.
-	bc	BO_IF,CR0_EQ,Lppcasm_div2	#proceed if no leading zeros
+	beq	Lppcasm_div2		#proceed if no leading zeros
 	subf	r8,r7,r8		#r8 = BN_num_bits_word(d)
 	$SHR.	r9,r3,r8		#are there any bits above r8'th?
 	$TR	16,r9,r0		#if there're, signal to dump core...
 Lppcasm_div2:
 	$UCMP	0,r3,r5			#h>=d?
-	bc	BO_IF,CR0_LT,Lppcasm_div3	#goto Lppcasm_div3 if not
+	blt	Lppcasm_div3		#goto Lppcasm_div3 if not
 	subf	r3,r5,r3		#h-=d ; 
 Lppcasm_div3:				#r7 = BN_BITS2-i. so r7=i
 	cmpi	0,0,r7,0		# is (i == 0)?
-	bc	BO_IF,CR0_EQ,Lppcasm_div4
+	beq	Lppcasm_div4
 	$SHL	r3,r3,r7		# h = (h<< i)
 	$SHR	r8,r4,r8		# r8 = (l >> BN_BITS2 -i)
 	$SHL	r5,r5,r7		# d<<=i
@@ -1741,7 +1651,7 @@ Lppcasm_divouterloop:
 	$SHRI	r11,r4,`$BITS/2`	#r11= (l&BN_MASK2h)>>BN_BITS4
 					# compute here for innerloop.
 	$UCMP	0,r8,r9			# is (h>>BN_BITS4)==dh
-	bc	BO_IF_NOT,CR0_EQ,Lppcasm_div5	# goto Lppcasm_div5 if not
+	bne	Lppcasm_div5		# goto Lppcasm_div5 if not
 
 	li	r8,-1
 	$CLRU	r8,r8,`$BITS/2`		#q = BN_MASK2l 
@@ -1762,9 +1672,9 @@ Lppcasm_divinnerloop:
 					# the following 2 instructions do that
 	$SHLI	r7,r10,`$BITS/2`	# r7 = (t<<BN_BITS4)
 	or	r7,r7,r11		# r7|=((l&BN_MASK2h)>>BN_BITS4)
-	$UCMP	1,r6,r7			# compare (tl <= r7)
-	bc	BO_IF_NOT,CR0_EQ,Lppcasm_divinnerexit
-	bc	BO_IF_NOT,CR1_FEX,Lppcasm_divinnerexit
+	$UCMP	cr1,r6,r7		# compare (tl <= r7)
+	bne	Lppcasm_divinnerexit
+	ble	cr1,Lppcasm_divinnerexit
 	addi	r8,r8,-1		#q--
 	subf	r12,r9,r12		#th -=dh
 	$CLRU	r10,r5,`$BITS/2`	#r10=dl. t is no longer needed in loop.
@@ -1773,14 +1683,14 @@ Lppcasm_divinnerloop:
 Lppcasm_divinnerexit:
 	$SHRI	r10,r6,`$BITS/2`	#t=(tl>>BN_BITS4)
 	$SHLI	r11,r6,`$BITS/2`	#tl=(tl<<BN_BITS4)&BN_MASK2h;
-	$UCMP	1,r4,r11		# compare l and tl
+	$UCMP	cr1,r4,r11		# compare l and tl
 	add	r12,r12,r10		# th+=t
-	bc	BO_IF_NOT,CR1_FX,Lppcasm_div7  # if (l>=tl) goto Lppcasm_div7
+	bge	cr1,Lppcasm_div7	# if (l>=tl) goto Lppcasm_div7
 	addi	r12,r12,1		# th++
 Lppcasm_div7:
 	subf	r11,r11,r4		#r11=l-tl
-	$UCMP	1,r3,r12		#compare h and th
-	bc	BO_IF_NOT,CR1_FX,Lppcasm_div8	#if (h>=th) goto Lppcasm_div8
+	$UCMP	cr1,r3,r12		#compare h and th
+	bge	cr1,Lppcasm_div8	#if (h>=th) goto Lppcasm_div8
 	addi	r8,r8,-1		# q--
 	add	r3,r5,r3		# h+=d
 Lppcasm_div8:
@@ -1791,12 +1701,12 @@ Lppcasm_div8:
 					# the following 2 instructions will do this.
 	$INSR	r11,r12,`$BITS/2`,`$BITS/2`	# r11 is the value we want rotated $BITS/2.
 	$ROTL	r3,r11,`$BITS/2`	# rotate by $BITS/2 and store in r3
-	bc	BO_dCTR_ZERO,CR0_EQ,Lppcasm_div9#if (count==0) break ;
+	bdz	Lppcasm_div9		#if (count==0) break ;
 	$SHLI	r0,r8,`$BITS/2`		#ret =q<<BN_BITS4
 	b	Lppcasm_divouterloop
 Lppcasm_div9:
 	or	r3,r8,r0
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 #
@@ -1822,7 +1732,7 @@ Lppcasm_div9:
 #	No unrolling done here. Not performance critical.
 
 	addic.	r5,r5,0			#test r5.
-	bc	BO_IF,CR0_EQ,Lppcasm_sqr_adios
+	beq	Lppcasm_sqr_adios
 	addi	r4,r4,-$BNSZ
 	addi	r3,r3,-$BNSZ
 	mtctr	r5
@@ -1833,9 +1743,9 @@ Lppcasm_sqr_mainloop:
 	$UMULH  r8,r6,r6
 	$STU	r7,$BNSZ(r3)
 	$STU	r8,$BNSZ(r3)
-	bc	BO_dCTR_NZERO,CR0_EQ,Lppcasm_sqr_mainloop
+	bdnz-	Lppcasm_sqr_mainloop
 Lppcasm_sqr_adios:	
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 
@@ -1858,7 +1768,7 @@ Lppcasm_sqr_adios:
 	xor	r0,r0,r0
 	xor	r12,r12,r12		# used for carry
 	rlwinm.	r7,r5,30,2,31		# num >> 2
-	bc	BO_IF,CR0_EQ,Lppcasm_mw_REM
+	beq	Lppcasm_mw_REM
 	mtctr	r7
 Lppcasm_mw_LOOP:	
 					#mul(rp[0],ap[0],w,c1);
@@ -1896,11 +1806,11 @@ Lppcasm_mw_LOOP:
 	
 	addi	r3,r3,`4*$BNSZ`
 	addi	r4,r4,`4*$BNSZ`
-	bc	BO_dCTR_NZERO,CR0_EQ,Lppcasm_mw_LOOP
+	bdnz-	Lppcasm_mw_LOOP
 
 Lppcasm_mw_REM:
 	andi.	r5,r5,0x3
-	bc	BO_IF,CR0_EQ,Lppcasm_mw_OVER
+	beq	Lppcasm_mw_OVER
 					#mul(rp[0],ap[0],w,c1);
 	$LD	r8,`0*$BNSZ`(r4)
 	$UMULL	r9,r6,r8
@@ -1912,7 +1822,7 @@ Lppcasm_mw_REM:
 	
 	addi	r5,r5,-1
 	cmpli	0,0,r5,0
-	bc	BO_IF,CR0_EQ,Lppcasm_mw_OVER
+	beq	Lppcasm_mw_OVER
 
 	
 					#mul(rp[1],ap[1],w,c1);
@@ -1926,7 +1836,7 @@ Lppcasm_mw_REM:
 	
 	addi	r5,r5,-1
 	cmpli	0,0,r5,0
-	bc	BO_IF,CR0_EQ,Lppcasm_mw_OVER
+	beq	Lppcasm_mw_OVER
 	
 					#mul_add(rp[2],ap[2],w,c1);
 	$LD	r8,`2*$BNSZ`(r4)
@@ -1939,7 +1849,7 @@ Lppcasm_mw_REM:
 		
 Lppcasm_mw_OVER:	
 	addi	r3,r12,0
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 
 #
@@ -1964,7 +1874,7 @@ Lppcasm_mw_OVER:
 	xor	r0,r0,r0		#r0 = 0
 	xor	r12,r12,r12  		#r12 = 0 . used for carry		
 	rlwinm.	r7,r5,30,2,31		# num >> 2
-	bc	BO_IF,CR0_EQ,Lppcasm_maw_leftover	# if (num < 4) go LPPCASM_maw_leftover
+	beq	Lppcasm_maw_leftover	# if (num < 4) go LPPCASM_maw_leftover
 	mtctr	r7
 Lppcasm_maw_mainloop:	
 					#mul_add(rp[0],ap[0],w,c1);
@@ -2017,11 +1927,11 @@ Lppcasm_maw_mainloop:
 	$ST	r11,`3*$BNSZ`(r3)
 	addi	r3,r3,`4*$BNSZ`
 	addi	r4,r4,`4*$BNSZ`
-	bc	BO_dCTR_NZERO,CR0_EQ,Lppcasm_maw_mainloop
+	bdnz-	Lppcasm_maw_mainloop
 	
 Lppcasm_maw_leftover:
 	andi.	r5,r5,0x3
-	bc	BO_IF,CR0_EQ,Lppcasm_maw_adios
+	beq	Lppcasm_maw_adios
 	addi	r3,r3,-$BNSZ
 	addi	r4,r4,-$BNSZ
 					#mul_add(rp[0],ap[0],w,c1);
@@ -2036,7 +1946,7 @@ Lppcasm_maw_leftover:
 	addze	r12,r10
 	$ST	r9,0(r3)
 	
-	bc	BO_dCTR_ZERO,CR0_EQ,Lppcasm_maw_adios
+	bdz	Lppcasm_maw_adios
 					#mul_add(rp[1],ap[1],w,c1);
 	$LDU	r8,$BNSZ(r4)	
 	$UMULL	r9,r6,r8
@@ -2048,7 +1958,7 @@ Lppcasm_maw_leftover:
 	addze	r12,r10
 	$ST	r9,0(r3)
 	
-	bc	BO_dCTR_ZERO,CR0_EQ,Lppcasm_maw_adios
+	bdz	Lppcasm_maw_adios
 					#mul_add(rp[2],ap[2],w,c1);
 	$LDU	r8,$BNSZ(r4)
 	$UMULL	r9,r6,r8
@@ -2062,17 +1972,10 @@ Lppcasm_maw_leftover:
 		
 Lppcasm_maw_adios:	
 	addi	r3,r12,0
-	bclr	BO_ALWAYS,CR0_LT
+	blr
 	.long	0x00000000
 	.align	4
 EOF
-	$data =~ s/\`([^\`]*)\`/eval $1/gem;
-
-	# if some assembler chokes on some simplified mnemonic,
-	# this is the spot to fix it up, e.g.:
-	# GNU as doesn't seem to accept cmplw, 32-bit unsigned compare
-	$data =~ s/^(\s*)cmplw(\s+)([^,]+),(.*)/$1cmpl$2$3,0,$4/gm;
-	# assembler X doesn't accept li, load immediate value
-	#$data =~ s/^(\s*)li(\s+)([^,]+),(.*)/$1addi$2$3,0,$4/gm;
-	return($data);
-}
+$data =~ s/\`([^\`]*)\`/eval $1/gem;
+print $data;
+close STDOUT;
diff --git a/openssl/crypto/bn/asm/sparcv8plus.S b/openssl/crypto/bn/asm/sparcv8plus.S
index 8c56e2e7e..63de1860f 100644
--- a/openssl/crypto/bn/asm/sparcv8plus.S
+++ b/openssl/crypto/bn/asm/sparcv8plus.S
@@ -144,6 +144,19 @@
  *	    }
  */
 
+#if defined(__SUNPRO_C) && defined(__sparcv9)
+  /* They've said -xarch=v9 at command line */
+  .register	%g2,#scratch
+  .register	%g3,#scratch
+# define	FRAME_SIZE	-192
+#elif defined(__GNUC__) && defined(__arch64__)
+  /* They've said -m64 at command line */
+  .register	%g2,#scratch
+  .register	%g3,#scratch
+# define	FRAME_SIZE	-192
+#else 
+# define	FRAME_SIZE	-96
+#endif 
 /*
  * GNU assembler can't stand stuw:-(
  */
@@ -619,8 +632,6 @@ bn_sub_words:
  *							Andy.
  */
 
-#define FRAME_SIZE	-96
-
 /*
  * Here is register usage map for *all* routines below.
  */
diff --git a/openssl/crypto/bn/asm/x86_64-gcc.c b/openssl/crypto/bn/asm/x86_64-gcc.c
index f13f52dd8..acb0b4011 100644
--- a/openssl/crypto/bn/asm/x86_64-gcc.c
+++ b/openssl/crypto/bn/asm/x86_64-gcc.c
@@ -1,4 +1,5 @@
-#ifdef __SUNPRO_C
+#include "../bn_lcl.h"
+#if !(defined(__GNUC__) && __GNUC__>=2)
 # include "../bn_asm.c"	/* kind of dirty hack for Sun Studio */
 #else
 /*
@@ -54,7 +55,15 @@
  *    machine.
  */
 
+#ifdef _WIN64
+#define BN_ULONG unsigned long long
+#else
 #define BN_ULONG unsigned long
+#endif
+
+#undef mul
+#undef mul_add
+#undef sqr
 
 /*
  * "m"(a), "+m"(r)	is the way to favor DirectPath µ-code;
@@ -97,7 +106,7 @@
 		: "a"(a)		\
 		: "cc");
 
-BN_ULONG bn_mul_add_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w)
+BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 	{
 	BN_ULONG c1=0;
 
@@ -121,7 +130,7 @@ BN_ULONG bn_mul_add_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w)
 	return(c1);
 	} 
 
-BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w)
+BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 	{
 	BN_ULONG c1=0;
 
@@ -144,7 +153,7 @@ BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w)
 	return(c1);
 	} 
 
-void bn_sqr_words(BN_ULONG *r, BN_ULONG *a, int n)
+void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
         {
 	if (n <= 0) return;
 
@@ -175,14 +184,14 @@ BN_ULONG bn_div_words(BN_ULONG h, BN_ULONG l, BN_ULONG d)
 	return ret;
 }
 
-BN_ULONG bn_add_words (BN_ULONG *rp, BN_ULONG *ap, BN_ULONG *bp,int n)
+BN_ULONG bn_add_words (BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp,int n)
 { BN_ULONG ret=0,i=0;
 
 	if (n <= 0) return 0;
 
 	asm (
 	"	subq	%2,%2		\n"
-	".align 16			\n"
+	".p2align 4			\n"
 	"1:	movq	(%4,%2,8),%0	\n"
 	"	adcq	(%5,%2,8),%0	\n"
 	"	movq	%0,(%3,%2,8)	\n"
@@ -198,14 +207,14 @@ BN_ULONG bn_add_words (BN_ULONG *rp, BN_ULONG *ap, BN_ULONG *bp,int n)
 }
 
 #ifndef SIMICS
-BN_ULONG bn_sub_words (BN_ULONG *rp, BN_ULONG *ap, BN_ULONG *bp,int n)
+BN_ULONG bn_sub_words (BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp,int n)
 { BN_ULONG ret=0,i=0;
 
 	if (n <= 0) return 0;
 
 	asm (
 	"	subq	%2,%2		\n"
-	".align 16			\n"
+	".p2align 4			\n"
 	"1:	movq	(%4,%2,8),%0	\n"
 	"	sbbq	(%5,%2,8),%0	\n"
 	"	movq	%0,(%3,%2,8)	\n"
@@ -485,7 +494,7 @@ void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
 	r[7]=c2;
 	}
 
-void bn_sqr_comba8(BN_ULONG *r, BN_ULONG *a)
+void bn_sqr_comba8(BN_ULONG *r, const BN_ULONG *a)
 	{
 	BN_ULONG t1,t2;
 	BN_ULONG c1,c2,c3;
@@ -561,7 +570,7 @@ void bn_sqr_comba8(BN_ULONG *r, BN_ULONG *a)
 	r[15]=c1;
 	}
 
-void bn_sqr_comba4(BN_ULONG *r, BN_ULONG *a)
+void bn_sqr_comba4(BN_ULONG *r, const BN_ULONG *a)
 	{
 	BN_ULONG t1,t2;
 	BN_ULONG c1,c2,c3;
diff --git a/openssl/crypto/bn/asm/x86_64-mont.pl b/openssl/crypto/bn/asm/x86_64-mont.pl
index c43b69592..3b7a6f243 100644
--- a/openssl/crypto/bn/asm/x86_64-mont.pl
+++ b/openssl/crypto/bn/asm/x86_64-mont.pl
@@ -15,14 +15,18 @@
 # respectful 50%. It remains to be seen if loop unrolling and
 # dedicated squaring routine can provide further improvement...
 
-$output=shift;
+$flavour = shift;
+$output  = shift;
+if ($flavour =~ /\./) { $output = $flavour; undef $flavour; }
+
+$win64=0; $win64=1 if ($flavour =~ /[nm]asm|mingw64/ || $output =~ /\.asm$/);
 
 $0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
 ( $xlate="${dir}x86_64-xlate.pl" and -f $xlate ) or
 ( $xlate="${dir}../../perlasm/x86_64-xlate.pl" and -f $xlate) or
 die "can't locate x86_64-xlate.pl";
 
-open STDOUT,"| $^X $xlate $output";
+open STDOUT,"| $^X $xlate $flavour $output";
 
 # int bn_mul_mont(
 $rp="%rdi";	# BN_ULONG *rp,
@@ -55,13 +59,14 @@ bn_mul_mont:
 	push	%r15
 
 	mov	${num}d,${num}d
-	lea	2($num),%rax
-	mov	%rsp,%rbp
-	neg	%rax
-	lea	(%rsp,%rax,8),%rsp	# tp=alloca(8*(num+2))
+	lea	2($num),%r10
+	mov	%rsp,%r11
+	neg	%r10
+	lea	(%rsp,%r10,8),%rsp	# tp=alloca(8*(num+2))
 	and	\$-1024,%rsp		# minimize TLB usage
 
-	mov	%rbp,8(%rsp,$num,8)	# tp[num+1]=%rsp
+	mov	%r11,8(%rsp,$num,8)	# tp[num+1]=%rsp
+.Lprologue:
 	mov	%rdx,$bp		# $bp reassigned, remember?
 
 	mov	($n0),$n0		# pull n0[0] value
@@ -197,18 +202,129 @@ bn_mul_mont:
 	dec	$j
 	jge	.Lcopy
 
-	mov	8(%rsp,$num,8),%rsp	# restore %rsp
+	mov	8(%rsp,$num,8),%rsi	# restore %rsp
 	mov	\$1,%rax
+	mov	(%rsi),%r15
+	mov	8(%rsi),%r14
+	mov	16(%rsi),%r13
+	mov	24(%rsi),%r12
+	mov	32(%rsi),%rbp
+	mov	40(%rsi),%rbx
+	lea	48(%rsi),%rsp
+.Lepilogue:
+	ret
+.size	bn_mul_mont,.-bn_mul_mont
+.asciz	"Montgomery Multiplication for x86_64, CRYPTOGAMS by <appro\@openssl.org>"
+.align	16
+___
+
+# EXCEPTION_DISPOSITION handler (EXCEPTION_RECORD *rec,ULONG64 frame,
+#		CONTEXT *context,DISPATCHER_CONTEXT *disp)
+if ($win64) {
+$rec="%rcx";
+$frame="%rdx";
+$context="%r8";
+$disp="%r9";
+
+$code.=<<___;
+.extern	__imp_RtlVirtualUnwind
+.type	se_handler,\@abi-omnipotent
+.align	16
+se_handler:
+	push	%rsi
+	push	%rdi
+	push	%rbx
+	push	%rbp
+	push	%r12
+	push	%r13
+	push	%r14
+	push	%r15
+	pushfq
+	sub	\$64,%rsp
+
+	mov	120($context),%rax	# pull context->Rax
+	mov	248($context),%rbx	# pull context->Rip
+
+	lea	.Lprologue(%rip),%r10
+	cmp	%r10,%rbx		# context->Rip<.Lprologue
+	jb	.Lin_prologue
+
+	mov	152($context),%rax	# pull context->Rsp
+
+	lea	.Lepilogue(%rip),%r10
+	cmp	%r10,%rbx		# context->Rip>=.Lepilogue
+	jae	.Lin_prologue
+
+	mov	192($context),%r10	# pull $num
+	mov	8(%rax,%r10,8),%rax	# pull saved stack pointer
+	lea	48(%rax),%rax
+
+	mov	-8(%rax),%rbx
+	mov	-16(%rax),%rbp
+	mov	-24(%rax),%r12
+	mov	-32(%rax),%r13
+	mov	-40(%rax),%r14
+	mov	-48(%rax),%r15
+	mov	%rbx,144($context)	# restore context->Rbx
+	mov	%rbp,160($context)	# restore context->Rbp
+	mov	%r12,216($context)	# restore context->R12
+	mov	%r13,224($context)	# restore context->R13
+	mov	%r14,232($context)	# restore context->R14
+	mov	%r15,240($context)	# restore context->R15
+
+.Lin_prologue:
+	mov	8(%rax),%rdi
+	mov	16(%rax),%rsi
+	mov	%rax,152($context)	# restore context->Rsp
+	mov	%rsi,168($context)	# restore context->Rsi
+	mov	%rdi,176($context)	# restore context->Rdi
+
+	mov	40($disp),%rdi		# disp->ContextRecord
+	mov	$context,%rsi		# context
+	mov	\$154,%ecx		# sizeof(CONTEXT)
+	.long	0xa548f3fc		# cld; rep movsq
+
+	mov	$disp,%rsi
+	xor	%rcx,%rcx		# arg1, UNW_FLAG_NHANDLER
+	mov	8(%rsi),%rdx		# arg2, disp->ImageBase
+	mov	0(%rsi),%r8		# arg3, disp->ControlPc
+	mov	16(%rsi),%r9		# arg4, disp->FunctionEntry
+	mov	40(%rsi),%r10		# disp->ContextRecord
+	lea	56(%rsi),%r11		# &disp->HandlerData
+	lea	24(%rsi),%r12		# &disp->EstablisherFrame
+	mov	%r10,32(%rsp)		# arg5
+	mov	%r11,40(%rsp)		# arg6
+	mov	%r12,48(%rsp)		# arg7
+	mov	%rcx,56(%rsp)		# arg8, (NULL)
+	call	*__imp_RtlVirtualUnwind(%rip)
+
+	mov	\$1,%eax		# ExceptionContinueSearch
+	add	\$64,%rsp
+	popfq
 	pop	%r15
 	pop	%r14
 	pop	%r13
 	pop	%r12
 	pop	%rbp
 	pop	%rbx
+	pop	%rdi
+	pop	%rsi
 	ret
-.size	bn_mul_mont,.-bn_mul_mont
-.asciz	"Montgomery Multiplication for x86_64, CRYPTOGAMS by <appro\@openssl.org>"
+.size	se_handler,.-se_handler
+
+.section	.pdata
+.align	4
+	.rva	.LSEH_begin_bn_mul_mont
+	.rva	.LSEH_end_bn_mul_mont
+	.rva	.LSEH_info_bn_mul_mont
+
+.section	.xdata
+.align	8
+.LSEH_info_bn_mul_mont:
+	.byte	9,0,0,0
+	.rva	se_handler
 ___
+}
 
 print $code;
 close STDOUT;
diff --git a/openssl/crypto/bn/bn.h b/openssl/crypto/bn/bn.h
index f1719a587..e484b7fc1 100644
--- a/openssl/crypto/bn/bn.h
+++ b/openssl/crypto/bn/bn.h
@@ -56,6 +56,59 @@
  * [including the GNU Public Licence.]
  */
 /* ====================================================================
+ * Copyright (c) 1998-2006 The OpenSSL Project.  All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer. 
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *
+ * 3. All advertising materials mentioning features or use of this
+ *    software must display the following acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit. (http://www.openssl.org/)"
+ *
+ * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
+ *    endorse or promote products derived from this software without
+ *    prior written permission. For written permission, please contact
+ *    openssl-core@openssl.org.
+ *
+ * 5. Products derived from this software may not be called "OpenSSL"
+ *    nor may "OpenSSL" appear in their names without prior written
+ *    permission of the OpenSSL Project.
+ *
+ * 6. Redistributions of any form whatsoever must retain the following
+ *    acknowledgment:
+ *    "This product includes software developed by the OpenSSL Project
+ *    for use in the OpenSSL Toolkit (http://www.openssl.org/)"
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
+ * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+ * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
+ * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+ * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ * ====================================================================
+ *
+ * This product includes cryptographic software written by Eric Young
+ * (eay@cryptsoft.com).  This product includes software written by Tim
+ * Hudson (tjh@cryptsoft.com).
+ *
+ */
+/* ====================================================================
  * Copyright 2002 Sun Microsystems, Inc. ALL RIGHTS RESERVED.
  *
  * Portions of the attached software ("Contribution") are developed by 
@@ -77,6 +130,7 @@
 #include <stdio.h> /* FILE */
 #endif
 #include <openssl/ossl_typ.h>
+#include <openssl/crypto.h>
 
 #ifdef  __cplusplus
 extern "C" {
@@ -94,9 +148,11 @@ extern "C" {
 /* #define BN_DEBUG */
 /* #define BN_DEBUG_RAND */
 
+#ifndef OPENSSL_SMALL_FOOTPRINT
 #define BN_MUL_COMBA
 #define BN_SQR_COMBA
 #define BN_RECURSION
+#endif
 
 /* This next option uses the C libraries (2 word)/(1 word) function.
  * If it is not defined, I use my C version (which is slower).
@@ -137,6 +193,8 @@ extern "C" {
 #define BN_DEC_FMT1	"%lu"
 #define BN_DEC_FMT2	"%019lu"
 #define BN_DEC_NUM	19
+#define BN_HEX_FMT1	"%lX"
+#define BN_HEX_FMT2	"%016lX"
 #endif
 
 /* This is where the long long data type is 64 bits, but long is 32.
@@ -162,83 +220,37 @@ extern "C" {
 #define BN_DEC_FMT1	"%llu"
 #define BN_DEC_FMT2	"%019llu"
 #define BN_DEC_NUM	19
+#define BN_HEX_FMT1	"%llX"
+#define BN_HEX_FMT2	"%016llX"
 #endif
 
 #ifdef THIRTY_TWO_BIT
 #ifdef BN_LLONG
-# if defined(OPENSSL_SYS_WIN32) && !defined(__GNUC__)
+# if defined(_WIN32) && !defined(__GNUC__)
 #  define BN_ULLONG	unsigned __int64
+#  define BN_MASK	(0xffffffffffffffffI64)
 # else
 #  define BN_ULLONG	unsigned long long
+#  define BN_MASK	(0xffffffffffffffffLL)
 # endif
 #endif
-#define BN_ULONG	unsigned long
-#define BN_LONG		long
+#define BN_ULONG	unsigned int
+#define BN_LONG		int
 #define BN_BITS		64
 #define BN_BYTES	4
 #define BN_BITS2	32
 #define BN_BITS4	16
-#ifdef OPENSSL_SYS_WIN32
-/* VC++ doesn't like the LL suffix */
-#define BN_MASK		(0xffffffffffffffffL)
-#else
-#define BN_MASK		(0xffffffffffffffffLL)
-#endif
 #define BN_MASK2	(0xffffffffL)
 #define BN_MASK2l	(0xffff)
 #define BN_MASK2h1	(0xffff8000L)
 #define BN_MASK2h	(0xffff0000L)
 #define BN_TBIT		(0x80000000L)
 #define BN_DEC_CONV	(1000000000L)
-#define BN_DEC_FMT1	"%lu"
-#define BN_DEC_FMT2	"%09lu"
-#define BN_DEC_NUM	9
-#endif
-
-#ifdef SIXTEEN_BIT
-#ifndef BN_DIV2W
-#define BN_DIV2W
-#endif
-#define BN_ULLONG	unsigned long
-#define BN_ULONG	unsigned short
-#define BN_LONG		short
-#define BN_BITS		32
-#define BN_BYTES	2
-#define BN_BITS2	16
-#define BN_BITS4	8
-#define BN_MASK		(0xffffffff)
-#define BN_MASK2	(0xffff)
-#define BN_MASK2l	(0xff)
-#define BN_MASK2h1	(0xff80)
-#define BN_MASK2h	(0xff00)
-#define BN_TBIT		(0x8000)
-#define BN_DEC_CONV	(100000)
 #define BN_DEC_FMT1	"%u"
-#define BN_DEC_FMT2	"%05u"
-#define BN_DEC_NUM	5
-#endif
-
-#ifdef EIGHT_BIT
-#ifndef BN_DIV2W
-#define BN_DIV2W
-#endif
-#define BN_ULLONG	unsigned short
-#define BN_ULONG	unsigned char
-#define BN_LONG		char
-#define BN_BITS		16
-#define BN_BYTES	1
-#define BN_BITS2	8
-#define BN_BITS4	4
-#define BN_MASK		(0xffff)
-#define BN_MASK2	(0xff)
-#define BN_MASK2l	(0xf)
-#define BN_MASK2h1	(0xf8)
-#define BN_MASK2h	(0xf0)
-#define BN_TBIT		(0x80)
-#define BN_DEC_CONV	(100)
-#define BN_DEC_FMT1	"%u"
-#define BN_DEC_FMT2	"%02u"
-#define BN_DEC_NUM	2
+#define BN_DEC_FMT2	"%09u"
+#define BN_DEC_NUM	9
+#define BN_HEX_FMT1	"%X"
+#define BN_HEX_FMT2	"%08X"
 #endif
 
 #define BN_DEFAULT_BITS	1280
@@ -303,12 +315,8 @@ struct bn_mont_ctx_st
 	BIGNUM N;      /* The modulus */
 	BIGNUM Ni;     /* R*(1/R mod N) - N*Ni = 1
 	                * (Ni is only stored for bignum algorithm) */
-#if 0
-	/* OpenSSL 0.9.9 preview: */
-	BN_ULONG n0[2];/* least significant word(s) of Ni */
-#else
-	BN_ULONG n0;   /* least significant word of Ni */
-#endif
+	BN_ULONG n0[2];/* least significant word(s) of Ni;
+	                  (type changed with 0.9.9, was "BN_ULONG n0;" before) */
 	int flags;
 	};
 
@@ -504,6 +512,7 @@ char *	BN_bn2hex(const BIGNUM *a);
 char *	BN_bn2dec(const BIGNUM *a);
 int 	BN_hex2bn(BIGNUM **a, const char *str);
 int 	BN_dec2bn(BIGNUM **a, const char *str);
+int	BN_asc2bn(BIGNUM **a, const char *str);
 int	BN_gcd(BIGNUM *r,const BIGNUM *a,const BIGNUM *b,BN_CTX *ctx);
 int	BN_kronecker(const BIGNUM *a,const BIGNUM *b,BN_CTX *ctx); /* returns -2 for error */
 BIGNUM *BN_mod_inverse(BIGNUM *ret,
@@ -531,17 +540,6 @@ int	BN_is_prime_ex(const BIGNUM *p,int nchecks, BN_CTX *ctx, BN_GENCB *cb);
 int	BN_is_prime_fasttest_ex(const BIGNUM *p,int nchecks, BN_CTX *ctx,
 		int do_trial_division, BN_GENCB *cb);
 
-int BN_X931_generate_Xpq(BIGNUM *Xp, BIGNUM *Xq, int nbits, BN_CTX *ctx);
-
-int BN_X931_derive_prime_ex(BIGNUM *p, BIGNUM *p1, BIGNUM *p2,
-			const BIGNUM *Xp, const BIGNUM *Xp1, const BIGNUM *Xp2,
-			const BIGNUM *e, BN_CTX *ctx, BN_GENCB *cb);
-int BN_X931_generate_prime_ex(BIGNUM *p, BIGNUM *p1, BIGNUM *p2,
-			BIGNUM *Xp1, BIGNUM *Xp2,
-			const BIGNUM *Xp,
-			const BIGNUM *e, BN_CTX *ctx,
-			BN_GENCB *cb);
-
 BN_MONT_CTX *BN_MONT_CTX_new(void );
 void BN_MONT_CTX_init(BN_MONT_CTX *ctx);
 int BN_mod_mul_montgomery(BIGNUM *r,const BIGNUM *a,const BIGNUM *b,
@@ -560,19 +558,22 @@ BN_MONT_CTX *BN_MONT_CTX_set_locked(BN_MONT_CTX **pmont, int lock,
 #define	BN_BLINDING_NO_UPDATE	0x00000001
 #define	BN_BLINDING_NO_RECREATE	0x00000002
 
-BN_BLINDING *BN_BLINDING_new(const BIGNUM *A, const BIGNUM *Ai, /* const */ BIGNUM *mod);
+BN_BLINDING *BN_BLINDING_new(const BIGNUM *A, const BIGNUM *Ai, BIGNUM *mod);
 void BN_BLINDING_free(BN_BLINDING *b);
 int BN_BLINDING_update(BN_BLINDING *b,BN_CTX *ctx);
 int BN_BLINDING_convert(BIGNUM *n, BN_BLINDING *b, BN_CTX *ctx);
 int BN_BLINDING_invert(BIGNUM *n, BN_BLINDING *b, BN_CTX *ctx);
 int BN_BLINDING_convert_ex(BIGNUM *n, BIGNUM *r, BN_BLINDING *b, BN_CTX *);
 int BN_BLINDING_invert_ex(BIGNUM *n, const BIGNUM *r, BN_BLINDING *b, BN_CTX *);
+#ifndef OPENSSL_NO_DEPRECATED
 unsigned long BN_BLINDING_get_thread_id(const BN_BLINDING *);
 void BN_BLINDING_set_thread_id(BN_BLINDING *, unsigned long);
+#endif
+CRYPTO_THREADID *BN_BLINDING_thread_id(BN_BLINDING *);
 unsigned long BN_BLINDING_get_flags(const BN_BLINDING *);
 void BN_BLINDING_set_flags(BN_BLINDING *, unsigned long);
 BN_BLINDING *BN_BLINDING_create_param(BN_BLINDING *b,
-	const BIGNUM *e, /* const */ BIGNUM *m, BN_CTX *ctx,
+	const BIGNUM *e, BIGNUM *m, BN_CTX *ctx,
 	int (*bn_mod_exp)(BIGNUM *r, const BIGNUM *a, const BIGNUM *p,
 			  const BIGNUM *m, BN_CTX *ctx, BN_MONT_CTX *m_ctx),
 	BN_MONT_CTX *m_ctx);
@@ -625,24 +626,24 @@ int	BN_GF2m_mod_solve_quad(BIGNUM *r, const BIGNUM *a, const BIGNUM *p,
  *     t^p[0] + t^p[1] + ... + t^p[k]
  * where m = p[0] > p[1] > ... > p[k] = 0.
  */
-int	BN_GF2m_mod_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[]);
+int	BN_GF2m_mod_arr(BIGNUM *r, const BIGNUM *a, const int p[]);
 	/* r = a mod p */
 int	BN_GF2m_mod_mul_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b,
-	const unsigned int p[], BN_CTX *ctx); /* r = (a * b) mod p */
-int	BN_GF2m_mod_sqr_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[],
+	const int p[], BN_CTX *ctx); /* r = (a * b) mod p */
+int	BN_GF2m_mod_sqr_arr(BIGNUM *r, const BIGNUM *a, const int p[],
 	BN_CTX *ctx); /* r = (a * a) mod p */
-int	BN_GF2m_mod_inv_arr(BIGNUM *r, const BIGNUM *b, const unsigned int p[],
+int	BN_GF2m_mod_inv_arr(BIGNUM *r, const BIGNUM *b, const int p[],
 	BN_CTX *ctx); /* r = (1 / b) mod p */
 int	BN_GF2m_mod_div_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b,
-	const unsigned int p[], BN_CTX *ctx); /* r = (a / b) mod p */
+	const int p[], BN_CTX *ctx); /* r = (a / b) mod p */
 int	BN_GF2m_mod_exp_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b,
-	const unsigned int p[], BN_CTX *ctx); /* r = (a ^ b) mod p */
+	const int p[], BN_CTX *ctx); /* r = (a ^ b) mod p */
 int	BN_GF2m_mod_sqrt_arr(BIGNUM *r, const BIGNUM *a,
-	const unsigned int p[], BN_CTX *ctx); /* r = sqrt(a) mod p */
+	const int p[], BN_CTX *ctx); /* r = sqrt(a) mod p */
 int	BN_GF2m_mod_solve_quad_arr(BIGNUM *r, const BIGNUM *a,
-	const unsigned int p[], BN_CTX *ctx); /* r^2 + r = a mod p */
-int	BN_GF2m_poly2arr(const BIGNUM *a, unsigned int p[], int max);
-int	BN_GF2m_arr2poly(const unsigned int p[], BIGNUM *a);
+	const int p[], BN_CTX *ctx); /* r^2 + r = a mod p */
+int	BN_GF2m_poly2arr(const BIGNUM *a, int p[], int max);
+int	BN_GF2m_arr2poly(const int p[], BIGNUM *a);
 
 /* faster mod functions for the 'NIST primes' 
  * 0 <= a < p^2 */
@@ -751,10 +752,12 @@ int RAND_pseudo_bytes(unsigned char *buf,int num);
 #define bn_correct_top(a) \
         { \
         BN_ULONG *ftl; \
-	if ((a)->top > 0) \
+	int tmp_top = (a)->top; \
+	if (tmp_top > 0) \
 		{ \
-		for (ftl= &((a)->d[(a)->top-1]); (a)->top > 0; (a)->top--) \
-		if (*(ftl--)) break; \
+		for (ftl= &((a)->d[tmp_top-1]); tmp_top > 0; tmp_top--) \
+			if (*(ftl--)) break; \
+		(a)->top = tmp_top; \
 		} \
 	bn_pollute(a); \
 	}
diff --git a/openssl/crypto/bn/bn_asm.c b/openssl/crypto/bn/bn_asm.c
index 99bc2de49..c43c91cc0 100644
--- a/openssl/crypto/bn/bn_asm.c
+++ b/openssl/crypto/bn/bn_asm.c
@@ -75,6 +75,7 @@ BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 	assert(num >= 0);
 	if (num <= 0) return(c1);
 
+#ifndef OPENSSL_SMALL_FOOTPRINT
 	while (num&~3)
 		{
 		mul_add(rp[0],ap[0],w,c1);
@@ -83,11 +84,11 @@ BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 		mul_add(rp[3],ap[3],w,c1);
 		ap+=4; rp+=4; num-=4;
 		}
-	if (num)
+#endif
+	while (num)
 		{
-		mul_add(rp[0],ap[0],w,c1); if (--num==0) return c1;
-		mul_add(rp[1],ap[1],w,c1); if (--num==0) return c1;
-		mul_add(rp[2],ap[2],w,c1); return c1;
+		mul_add(rp[0],ap[0],w,c1);
+		ap++; rp++; num--;
 		}
 	
 	return(c1);
@@ -100,6 +101,7 @@ BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 	assert(num >= 0);
 	if (num <= 0) return(c1);
 
+#ifndef OPENSSL_SMALL_FOOTPRINT
 	while (num&~3)
 		{
 		mul(rp[0],ap[0],w,c1);
@@ -108,11 +110,11 @@ BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 		mul(rp[3],ap[3],w,c1);
 		ap+=4; rp+=4; num-=4;
 		}
-	if (num)
+#endif
+	while (num)
 		{
-		mul(rp[0],ap[0],w,c1); if (--num == 0) return c1;
-		mul(rp[1],ap[1],w,c1); if (--num == 0) return c1;
-		mul(rp[2],ap[2],w,c1);
+		mul(rp[0],ap[0],w,c1);
+		ap++; rp++; num--;
 		}
 	return(c1);
 	} 
@@ -121,6 +123,8 @@ void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
         {
 	assert(n >= 0);
 	if (n <= 0) return;
+
+#ifndef OPENSSL_SMALL_FOOTPRINT
 	while (n&~3)
 		{
 		sqr(r[0],r[1],a[0]);
@@ -129,11 +133,11 @@ void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
 		sqr(r[6],r[7],a[3]);
 		a+=4; r+=8; n-=4;
 		}
-	if (n)
+#endif
+	while (n)
 		{
-		sqr(r[0],r[1],a[0]); if (--n == 0) return;
-		sqr(r[2],r[3],a[1]); if (--n == 0) return;
-		sqr(r[4],r[5],a[2]);
+		sqr(r[0],r[1],a[0]);
+		a++; r+=2; n--;
 		}
 	}
 
@@ -150,18 +154,20 @@ BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 	bl=LBITS(w);
 	bh=HBITS(w);
 
-	for (;;)
+#ifndef OPENSSL_SMALL_FOOTPRINT
+	while (num&~3)
 		{
 		mul_add(rp[0],ap[0],bl,bh,c);
-		if (--num == 0) break;
 		mul_add(rp[1],ap[1],bl,bh,c);
-		if (--num == 0) break;
 		mul_add(rp[2],ap[2],bl,bh,c);
-		if (--num == 0) break;
 		mul_add(rp[3],ap[3],bl,bh,c);
-		if (--num == 0) break;
-		ap+=4;
-		rp+=4;
+		ap+=4; rp+=4; num-=4;
+		}
+#endif
+	while (num)
+		{
+		mul_add(rp[0],ap[0],bl,bh,c);
+		ap++; rp++; num--;
 		}
 	return(c);
 	} 
@@ -177,18 +183,20 @@ BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
 	bl=LBITS(w);
 	bh=HBITS(w);
 
-	for (;;)
+#ifndef OPENSSL_SMALL_FOOTPRINT
+	while (num&~3)
 		{
 		mul(rp[0],ap[0],bl,bh,carry);
-		if (--num == 0) break;
 		mul(rp[1],ap[1],bl,bh,carry);
-		if (--num == 0) break;
 		mul(rp[2],ap[2],bl,bh,carry);
-		if (--num == 0) break;
 		mul(rp[3],ap[3],bl,bh,carry);
-		if (--num == 0) break;
-		ap+=4;
-		rp+=4;
+		ap+=4; rp+=4; num-=4;
+		}
+#endif
+	while (num)
+		{
+		mul(rp[0],ap[0],bl,bh,carry);
+		ap++; rp++; num--;
 		}
 	return(carry);
 	} 
@@ -197,22 +205,21 @@ void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
         {
 	assert(n >= 0);
 	if (n <= 0) return;
-	for (;;)
+
+#ifndef OPENSSL_SMALL_FOOTPRINT
+	while (n&~3)
 		{
 		sqr64(r[0],r[1],a[0]);
-		if (--n == 0) break;
-
 		sqr64(r[2],r[3],a[1]);
-		if (--n == 0) break;
-
 		sqr64(r[4],r[5],a[2]);
-		if (--n == 0) break;
-
 		sqr64(r[6],r[7],a[3]);
-		if (--n == 0) break;
-
-		a+=4;
-		r+=8;
+		a+=4; r+=8; n-=4;
+		}
+#endif
+	while (n)
+		{
+		sqr64(r[0],r[1],a[0]);
+		a++; r+=2; n--;
 		}
 	}
 
@@ -303,31 +310,30 @@ BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
 	assert(n >= 0);
 	if (n <= 0) return((BN_ULONG)0);
 
-	for (;;)
+#ifndef OPENSSL_SMALL_FOOTPRINT
+	while (n&~3)
 		{
 		ll+=(BN_ULLONG)a[0]+b[0];
 		r[0]=(BN_ULONG)ll&BN_MASK2;
 		ll>>=BN_BITS2;
-		if (--n <= 0) break;
-
 		ll+=(BN_ULLONG)a[1]+b[1];
 		r[1]=(BN_ULONG)ll&BN_MASK2;
 		ll>>=BN_BITS2;
-		if (--n <= 0) break;
-
 		ll+=(BN_ULLONG)a[2]+b[2];
 		r[2]=(BN_ULONG)ll&BN_MASK2;
 		ll>>=BN_BITS2;
-		if (--n <= 0) break;
-
 		ll+=(BN_ULLONG)a[3]+b[3];
 		r[3]=(BN_ULONG)ll&BN_MASK2;
 		ll>>=BN_BITS2;
-		if (--n <= 0) break;
-
-		a+=4;
-		b+=4;
-		r+=4;
+		a+=4; b+=4; r+=4; n-=4;
+		}
+#endif
+	while (n)
+		{
+		ll+=(BN_ULLONG)a[0]+b[0];
+		r[0]=(BN_ULONG)ll&BN_MASK2;
+		ll>>=BN_BITS2;
+		a++; b++; r++; n--;
 		}
 	return((BN_ULONG)ll);
 	}
@@ -340,7 +346,8 @@ BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
 	if (n <= 0) return((BN_ULONG)0);
 
 	c=0;
-	for (;;)
+#ifndef OPENSSL_SMALL_FOOTPRINT
+	while (n&~3)
 		{
 		t=a[0];
 		t=(t+c)&BN_MASK2;
@@ -348,35 +355,36 @@ BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
 		l=(t+b[0])&BN_MASK2;
 		c+=(l < t);
 		r[0]=l;
-		if (--n <= 0) break;
-
 		t=a[1];
 		t=(t+c)&BN_MASK2;
 		c=(t < c);
 		l=(t+b[1])&BN_MASK2;
 		c+=(l < t);
 		r[1]=l;
-		if (--n <= 0) break;
-
 		t=a[2];
 		t=(t+c)&BN_MASK2;
 		c=(t < c);
 		l=(t+b[2])&BN_MASK2;
 		c+=(l < t);
 		r[2]=l;
-		if (--n <= 0) break;
-
 		t=a[3];
 		t=(t+c)&BN_MASK2;
 		c=(t < c);
 		l=(t+b[3])&BN_MASK2;
 		c+=(l < t);
 		r[3]=l;
-		if (--n <= 0) break;
-
-		a+=4;
-		b+=4;
-		r+=4;
+		a+=4; b+=4; r+=4; n-=4;
+		}
+#endif
+	while(n)
+		{
+		t=a[0];
+		t=(t+c)&BN_MASK2;
+		c=(t < c);
+		l=(t+b[0])&BN_MASK2;
+		c+=(l < t);
+		r[0]=l;
+		a++; b++; r++; n--;
 		}
 	return((BN_ULONG)c);
 	}
@@ -390,36 +398,35 @@ BN_ULONG bn_sub_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b, int n)
 	assert(n >= 0);
 	if (n <= 0) return((BN_ULONG)0);
 
-	for (;;)
+#ifndef OPENSSL_SMALL_FOOTPRINT
+	while (n&~3)
 		{
 		t1=a[0]; t2=b[0];
 		r[0]=(t1-t2-c)&BN_MASK2;
 		if (t1 != t2) c=(t1 < t2);
-		if (--n <= 0) break;
-
 		t1=a[1]; t2=b[1];
 		r[1]=(t1-t2-c)&BN_MASK2;
 		if (t1 != t2) c=(t1 < t2);
-		if (--n <= 0) break;
-
 		t1=a[2]; t2=b[2];
 		r[2]=(t1-t2-c)&BN_MASK2;
 		if (t1 != t2) c=(t1 < t2);
-		if (--n <= 0) break;
-
 		t1=a[3]; t2=b[3];
 		r[3]=(t1-t2-c)&BN_MASK2;
 		if (t1 != t2) c=(t1 < t2);
-		if (--n <= 0) break;
-
-		a+=4;
-		b+=4;
-		r+=4;
+		a+=4; b+=4; r+=4; n-=4;
+		}
+#endif
+	while (n)
+		{
+		t1=a[0]; t2=b[0];
+		r[0]=(t1-t2-c)&BN_MASK2;
+		if (t1 != t2) c=(t1 < t2);
+		a++; b++; r++; n--;
 		}
 	return(c);
 	}
 
-#ifdef BN_MUL_COMBA
+#if defined(BN_MUL_COMBA) && !defined(OPENSSL_SMALL_FOOTPRINT)
 
 #undef bn_mul_comba8
 #undef bn_mul_comba4
@@ -820,18 +827,134 @@ void bn_sqr_comba4(BN_ULONG *r, const BN_ULONG *a)
 	r[6]=c1;
 	r[7]=c2;
 	}
+
+#ifdef OPENSSL_NO_ASM
+#ifdef OPENSSL_BN_ASM_MONT
+#include <alloca.h>
+/*
+ * This is essentially reference implementation, which may or may not
+ * result in performance improvement. E.g. on IA-32 this routine was
+ * observed to give 40% faster rsa1024 private key operations and 10%
+ * faster rsa4096 ones, while on AMD64 it improves rsa1024 sign only
+ * by 10% and *worsens* rsa4096 sign by 15%. Once again, it's a
+ * reference implementation, one to be used as starting point for
+ * platform-specific assembler. Mentioned numbers apply to compiler
+ * generated code compiled with and without -DOPENSSL_BN_ASM_MONT and
+ * can vary not only from platform to platform, but even for compiler
+ * versions. Assembler vs. assembler improvement coefficients can
+ * [and are known to] differ and are to be documented elsewhere.
+ */
+int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, const BN_ULONG *np,const BN_ULONG *n0p, int num)
+	{
+	BN_ULONG c0,c1,ml,*tp,n0;
+#ifdef mul64
+	BN_ULONG mh;
+#endif
+	volatile BN_ULONG *vp;
+	int i=0,j;
+
+#if 0	/* template for platform-specific implementation */
+	if (ap==bp)	return bn_sqr_mont(rp,ap,np,n0p,num);
+#endif
+	vp = tp = alloca((num+2)*sizeof(BN_ULONG));
+
+	n0 = *n0p;
+
+	c0 = 0;
+	ml = bp[0];
+#ifdef mul64
+	mh = HBITS(ml);
+	ml = LBITS(ml);
+	for (j=0;j<num;++j)
+		mul(tp[j],ap[j],ml,mh,c0);
+#else
+	for (j=0;j<num;++j)
+		mul(tp[j],ap[j],ml,c0);
+#endif
+
+	tp[num]   = c0;
+	tp[num+1] = 0;
+	goto enter;
+
+	for(i=0;i<num;i++)
+		{
+		c0 = 0;
+		ml = bp[i];
+#ifdef mul64
+		mh = HBITS(ml);
+		ml = LBITS(ml);
+		for (j=0;j<num;++j)
+			mul_add(tp[j],ap[j],ml,mh,c0);
+#else
+		for (j=0;j<num;++j)
+			mul_add(tp[j],ap[j],ml,c0);
+#endif
+		c1 = (tp[num] + c0)&BN_MASK2;
+		tp[num]   = c1;
+		tp[num+1] = (c1<c0?1:0);
+	enter:
+		c1  = tp[0];
+		ml = (c1*n0)&BN_MASK2;
+		c0 = 0;
+#ifdef mul64
+		mh = HBITS(ml);
+		ml = LBITS(ml);
+		mul_add(c1,np[0],ml,mh,c0);
+#else
+		mul_add(c1,ml,np[0],c0);
+#endif
+		for(j=1;j<num;j++)
+			{
+			c1 = tp[j];
+#ifdef mul64
+			mul_add(c1,np[j],ml,mh,c0);
+#else
+			mul_add(c1,ml,np[j],c0);
+#endif
+			tp[j-1] = c1&BN_MASK2;
+			}
+		c1        = (tp[num] + c0)&BN_MASK2;
+		tp[num-1] = c1;
+		tp[num]   = tp[num+1] + (c1<c0?1:0);
+		}
+
+	if (tp[num]!=0 || tp[num-1]>=np[num-1])
+		{
+		c0 = bn_sub_words(rp,tp,np,num);
+		if (tp[num]!=0 || c0==0)
+			{
+			for(i=0;i<num+2;i++)	vp[i] = 0;
+			return 1;
+			}
+		}
+	for(i=0;i<num;i++)	rp[i] = tp[i],	vp[i] = 0;
+	vp[num]   = 0;
+	vp[num+1] = 0;
+	return 1;
+	}
+#else
+/*
+ * Return value of 0 indicates that multiplication/convolution was not
+ * performed to signal the caller to fall down to alternative/original
+ * code-path.
+ */
+int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, const BN_ULONG *np,const BN_ULONG *n0, int num)
+{	return 0;	}
+#endif /* OPENSSL_BN_ASM_MONT */
+#endif
+
 #else /* !BN_MUL_COMBA */
 
 /* hmm... is it faster just to do a multiply? */
 #undef bn_sqr_comba4
-void bn_sqr_comba4(BN_ULONG *r, BN_ULONG *a)
+void bn_sqr_comba4(BN_ULONG *r, const BN_ULONG *a)
 	{
 	BN_ULONG t[8];
 	bn_sqr_normal(r,a,4,t);
 	}
 
 #undef bn_sqr_comba8
-void bn_sqr_comba8(BN_ULONG *r, BN_ULONG *a)
+void bn_sqr_comba8(BN_ULONG *r, const BN_ULONG *a)
 	{
 	BN_ULONG t[16];
 	bn_sqr_normal(r,a,8,t);
@@ -857,4 +980,51 @@ void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
 	r[15]=bn_mul_add_words(&(r[7]),a,8,b[7]);
 	}
 
+#ifdef OPENSSL_NO_ASM
+#ifdef OPENSSL_BN_ASM_MONT
+#include <alloca.h>
+int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, const BN_ULONG *np,const BN_ULONG *n0p, int num)
+	{
+	BN_ULONG c0,c1,*tp,n0=*n0p;
+	volatile BN_ULONG *vp;
+	int i=0,j;
+
+	vp = tp = alloca((num+2)*sizeof(BN_ULONG));
+
+	for(i=0;i<=num;i++)	tp[i]=0;
+
+	for(i=0;i<num;i++)
+		{
+		c0         = bn_mul_add_words(tp,ap,num,bp[i]);
+		c1         = (tp[num] + c0)&BN_MASK2;
+		tp[num]    = c1;
+		tp[num+1]  = (c1<c0?1:0);
+
+		c0         = bn_mul_add_words(tp,np,num,tp[0]*n0);
+		c1         = (tp[num] + c0)&BN_MASK2;
+		tp[num]    = c1;
+		tp[num+1] += (c1<c0?1:0);
+		for(j=0;j<=num;j++)	tp[j]=tp[j+1];
+		}
+
+	if (tp[num]!=0 || tp[num-1]>=np[num-1])
+		{
+		c0 = bn_sub_words(rp,tp,np,num);
+		if (tp[num]!=0 || c0==0)
+			{
+			for(i=0;i<num+2;i++)	vp[i] = 0;
+			return 1;
+			}
+		}
+	for(i=0;i<num;i++)	rp[i] = tp[i],	vp[i] = 0;
+	vp[num]   = 0;
+	vp[num+1] = 0;
+	return 1;
+	}
+#else
+int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, const BN_ULONG *np,const BN_ULONG *n0, int num)
+{	return 0;	}
+#endif /* OPENSSL_BN_ASM_MONT */
+#endif
+
 #endif /* !BN_MUL_COMBA */
diff --git a/openssl/crypto/bn/bn_blind.c b/openssl/crypto/bn/bn_blind.c
index c11fb4ccc..e060592fd 100644
--- a/openssl/crypto/bn/bn_blind.c
+++ b/openssl/crypto/bn/bn_blind.c
@@ -1,6 +1,6 @@
 /* crypto/bn/bn_blind.c */
 /* ====================================================================
- * Copyright (c) 1998-2005 The OpenSSL Project.  All rights reserved.
+ * Copyright (c) 1998-2006 The OpenSSL Project.  All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
@@ -121,8 +121,11 @@ struct bn_blinding_st
 	BIGNUM *Ai;
 	BIGNUM *e;
 	BIGNUM *mod; /* just a reference */
+#ifndef OPENSSL_NO_DEPRECATED
 	unsigned long thread_id; /* added in OpenSSL 0.9.6j and 0.9.7b;
 				  * used only by crypto/rsa/rsa_eay.c, rsa_lib.c */
+#endif
+	CRYPTO_THREADID tid;
 	unsigned int  counter;
 	unsigned long flags;
 	BN_MONT_CTX *m_ctx;
@@ -131,7 +134,7 @@ struct bn_blinding_st
 			  BN_MONT_CTX *m_ctx);
 	};
 
-BN_BLINDING *BN_BLINDING_new(const BIGNUM *A, const BIGNUM *Ai, /* const */ BIGNUM *mod)
+BN_BLINDING *BN_BLINDING_new(const BIGNUM *A, const BIGNUM *Ai, BIGNUM *mod)
 	{
 	BN_BLINDING *ret=NULL;
 
@@ -158,6 +161,7 @@ BN_BLINDING *BN_BLINDING_new(const BIGNUM *A, const BIGNUM *Ai, /* const */ BIGN
 		BN_set_flags(ret->mod, BN_FLG_CONSTTIME);
 
 	ret->counter = BN_BLINDING_COUNTER;
+	CRYPTO_THREADID_current(&ret->tid);
 	return(ret);
 err:
 	if (ret != NULL) BN_BLINDING_free(ret);
@@ -263,6 +267,7 @@ int BN_BLINDING_invert_ex(BIGNUM *n, const BIGNUM *r, BN_BLINDING *b, BN_CTX *ct
 	return(ret);
 	}
 
+#ifndef OPENSSL_NO_DEPRECATED
 unsigned long BN_BLINDING_get_thread_id(const BN_BLINDING *b)
 	{
 	return b->thread_id;
@@ -272,6 +277,12 @@ void BN_BLINDING_set_thread_id(BN_BLINDING *b, unsigned long n)
 	{
 	b->thread_id = n;
 	}
+#endif
+
+CRYPTO_THREADID *BN_BLINDING_thread_id(BN_BLINDING *b)
+	{
+	return &b->tid;
+	}
 
 unsigned long BN_BLINDING_get_flags(const BN_BLINDING *b)
 	{
@@ -284,7 +295,7 @@ void BN_BLINDING_set_flags(BN_BLINDING *b, unsigned long flags)
 	}
 
 BN_BLINDING *BN_BLINDING_create_param(BN_BLINDING *b,
-	const BIGNUM *e, /* const */ BIGNUM *m, BN_CTX *ctx,
+	const BIGNUM *e, BIGNUM *m, BN_CTX *ctx,
 	int (*bn_mod_exp)(BIGNUM *r, const BIGNUM *a, const BIGNUM *p,
 			  const BIGNUM *m, BN_CTX *ctx, BN_MONT_CTX *m_ctx),
 	BN_MONT_CTX *m_ctx)
diff --git a/openssl/crypto/bn/bn_ctx.c b/openssl/crypto/bn/bn_ctx.c
index b3452f1a9..3f2256f67 100644
--- a/openssl/crypto/bn/bn_ctx.c
+++ b/openssl/crypto/bn/bn_ctx.c
@@ -161,7 +161,7 @@ static void ctxdbg(BN_CTX *ctx)
 	fprintf(stderr,"(%08x): ", (unsigned int)ctx);
 	while(bnidx < ctx->used)
 		{
-		fprintf(stderr,"%02x ", item->vals[bnidx++ % BN_CTX_POOL_SIZE].dmax);
+		fprintf(stderr,"%03x ", item->vals[bnidx++ % BN_CTX_POOL_SIZE].dmax);
 		if(!(bnidx % BN_CTX_POOL_SIZE))
 			item = item->next;
 		}
@@ -171,8 +171,8 @@ static void ctxdbg(BN_CTX *ctx)
 	while(fpidx < stack->depth)
 		{
 		while(bnidx++ < stack->indexes[fpidx])
-			fprintf(stderr,"   ");
-		fprintf(stderr,"^^ ");
+			fprintf(stderr,"    ");
+		fprintf(stderr,"^^^ ");
 		bnidx++;
 		fpidx++;
 		}
diff --git a/openssl/crypto/bn/bn_div.c b/openssl/crypto/bn/bn_div.c
index 1e8e57626..802a43d64 100644
--- a/openssl/crypto/bn/bn_div.c
+++ b/openssl/crypto/bn/bn_div.c
@@ -102,7 +102,7 @@ int BN_div(BIGNUM *dv, BIGNUM *rem, const BIGNUM *m, const BIGNUM *d,
 	/* The next 2 are needed so we can do a dv->d[0]|=1 later
 	 * since BN_lshift1 will only work once there is a value :-) */
 	BN_zero(dv);
-	bn_wexpand(dv,1);
+	if(bn_wexpand(dv,1) == NULL) goto end;
 	dv->top=1;
 
 	if (!BN_lshift(D,D,nm-nd)) goto end;
@@ -229,7 +229,8 @@ int BN_div(BIGNUM *dv, BIGNUM *rm, const BIGNUM *num, const BIGNUM *divisor,
 	if (dv == NULL)
 		res=BN_CTX_get(ctx);
 	else	res=dv;
-	if (sdiv == NULL || res == NULL) goto err;
+	if (sdiv == NULL || res == NULL || tmp == NULL || snum == NULL)
+		goto err;
 
 	/* First we normalise the numbers */
 	norm_shift=BN_BITS2-((BN_num_bits(divisor))%BN_BITS2);
@@ -336,7 +337,7 @@ X) -> 0x%08X\n",
 				t2 -= d1;
 				}
 #else /* !BN_LLONG */
-			BN_ULONG t2l,t2h,ql,qh;
+			BN_ULONG t2l,t2h;
 
 			q=bn_div_words(n0,n1,d0);
 #ifdef BN_DEBUG_LEVITTE
@@ -354,9 +355,12 @@ X) -> 0x%08X\n",
 			t2l = d1 * q;
 			t2h = BN_UMULT_HIGH(d1,q);
 #else
+			{
+			BN_ULONG ql, qh;
 			t2l=LBITS(d1); t2h=HBITS(d1);
 			ql =LBITS(q);  qh =HBITS(q);
 			mul64(t2l,t2h,ql,qh); /* t2=(BN_ULLONG)d1*q; */
+			}
 #endif
 
 			for (;;)
@@ -560,7 +564,7 @@ X) -> 0x%08X\n",
 				t2 -= d1;
 				}
 #else /* !BN_LLONG */
-			BN_ULONG t2l,t2h,ql,qh;
+			BN_ULONG t2l,t2h;
 
 			q=bn_div_words(n0,n1,d0);
 #ifdef BN_DEBUG_LEVITTE
@@ -578,9 +582,12 @@ X) -> 0x%08X\n",
 			t2l = d1 * q;
 			t2h = BN_UMULT_HIGH(d1,q);
 #else
+			{
+			BN_ULONG ql, qh;
 			t2l=LBITS(d1); t2h=HBITS(d1);
 			ql =LBITS(q);  qh =HBITS(q);
 			mul64(t2l,t2h,ql,qh); /* t2=(BN_ULLONG)d1*q; */
+			}
 #endif
 
 			for (;;)
diff --git a/openssl/crypto/bn/bn_exp.c b/openssl/crypto/bn/bn_exp.c
index 70a33f0d9..d9b6c737f 100644
--- a/openssl/crypto/bn/bn_exp.c
+++ b/openssl/crypto/bn/bn_exp.c
@@ -134,7 +134,8 @@ int BN_exp(BIGNUM *r, const BIGNUM *a, const BIGNUM *p, BN_CTX *ctx)
 		rr = BN_CTX_get(ctx);
 	else
 		rr = r;
-	if ((v = BN_CTX_get(ctx)) == NULL) goto err;
+	v = BN_CTX_get(ctx);
+	if (rr == NULL || v == NULL) goto err;
 
 	if (BN_copy(v,a) == NULL) goto err;
 	bits=BN_num_bits(p);
diff --git a/openssl/crypto/bn/bn_gf2m.c b/openssl/crypto/bn/bn_gf2m.c
index 306f029f2..527b0fa15 100644
--- a/openssl/crypto/bn/bn_gf2m.c
+++ b/openssl/crypto/bn/bn_gf2m.c
@@ -121,74 +121,12 @@ static const BN_ULONG SQR_tb[16] =
     SQR_tb[(w) >> 12 & 0xF] << 24 | SQR_tb[(w) >>  8 & 0xF] << 16 | \
     SQR_tb[(w) >>  4 & 0xF] <<  8 | SQR_tb[(w)       & 0xF]
 #endif
-#ifdef SIXTEEN_BIT
-#define SQR1(w) \
-    SQR_tb[(w) >> 12 & 0xF] <<  8 | SQR_tb[(w) >>  8 & 0xF]
-#define SQR0(w) \
-    SQR_tb[(w) >>  4 & 0xF] <<  8 | SQR_tb[(w)       & 0xF]
-#endif
-#ifdef EIGHT_BIT
-#define SQR1(w) \
-    SQR_tb[(w) >>  4 & 0xF]
-#define SQR0(w) \
-    SQR_tb[(w)       & 15]
-#endif
 
 /* Product of two polynomials a, b each with degree < BN_BITS2 - 1,
  * result is a polynomial r with degree < 2 * BN_BITS - 1
  * The caller MUST ensure that the variables have the right amount
  * of space allocated.
  */
-#ifdef EIGHT_BIT
-static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)
-	{
-	register BN_ULONG h, l, s;
-	BN_ULONG tab[4], top1b = a >> 7;
-	register BN_ULONG a1, a2;
-
-	a1 = a & (0x7F); a2 = a1 << 1;
-
-	tab[0] = 0; tab[1] = a1; tab[2] = a2; tab[3] = a1^a2;
-
-	s = tab[b      & 0x3]; l  = s;
-	s = tab[b >> 2 & 0x3]; l ^= s << 2; h  = s >> 6;
-	s = tab[b >> 4 & 0x3]; l ^= s << 4; h ^= s >> 4;
-	s = tab[b >> 6      ]; l ^= s << 6; h ^= s >> 2;
-	
-	/* compensate for the top bit of a */
-
-	if (top1b & 01) { l ^= b << 7; h ^= b >> 1; } 
-
-	*r1 = h; *r0 = l;
-	} 
-#endif
-#ifdef SIXTEEN_BIT
-static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)
-	{
-	register BN_ULONG h, l, s;
-	BN_ULONG tab[4], top1b = a >> 15; 
-	register BN_ULONG a1, a2;
-
-	a1 = a & (0x7FFF); a2 = a1 << 1;
-
-	tab[0] = 0; tab[1] = a1; tab[2] = a2; tab[3] = a1^a2;
-
-	s = tab[b      & 0x3]; l  = s;
-	s = tab[b >> 2 & 0x3]; l ^= s <<  2; h  = s >> 14;
-	s = tab[b >> 4 & 0x3]; l ^= s <<  4; h ^= s >> 12;
-	s = tab[b >> 6 & 0x3]; l ^= s <<  6; h ^= s >> 10;
-	s = tab[b >> 8 & 0x3]; l ^= s <<  8; h ^= s >>  8;
-	s = tab[b >>10 & 0x3]; l ^= s << 10; h ^= s >>  6;
-	s = tab[b >>12 & 0x3]; l ^= s << 12; h ^= s >>  4;
-	s = tab[b >>14      ]; l ^= s << 14; h ^= s >>  2;
-
-	/* compensate for the top bit of a */
-
-	if (top1b & 01) { l ^= b << 15; h ^= b >> 1; } 
-
-	*r1 = h; *r0 = l;
-	} 
-#endif
 #ifdef THIRTY_TWO_BIT
 static void bn_GF2m_mul_1x1(BN_ULONG *r1, BN_ULONG *r0, const BN_ULONG a, const BN_ULONG b)
 	{
@@ -294,7 +232,8 @@ int	BN_GF2m_add(BIGNUM *r, const BIGNUM *a, const BIGNUM *b)
 	if (a->top < b->top) { at = b; bt = a; }
 	else { at = a; bt = b; }
 
-	bn_wexpand(r, at->top);
+	if(bn_wexpand(r, at->top) == NULL)
+		return 0;
 
 	for (i = 0; i < bt->top; i++)
 		{
@@ -320,7 +259,7 @@ int	BN_GF2m_add(BIGNUM *r, const BIGNUM *a, const BIGNUM *b)
 
 
 /* Performs modular reduction of a and store result in r.  r could be a. */
-int BN_GF2m_mod_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[])
+int BN_GF2m_mod_arr(BIGNUM *r, const BIGNUM *a, const int p[])
 	{
 	int j, k;
 	int n, dN, d0, d1;
@@ -421,11 +360,11 @@ int BN_GF2m_mod_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[])
 int	BN_GF2m_mod(BIGNUM *r, const BIGNUM *a, const BIGNUM *p)
 	{
 	int ret = 0;
-	const int max = BN_num_bits(p);
-	unsigned int *arr=NULL;
+	const int max = BN_num_bits(p) + 1;
+	int *arr=NULL;
 	bn_check_top(a);
 	bn_check_top(p);
-	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;
+	if ((arr = (int *)OPENSSL_malloc(sizeof(int) * max)) == NULL) goto err;
 	ret = BN_GF2m_poly2arr(p, arr, max);
 	if (!ret || ret > max)
 		{
@@ -443,7 +382,7 @@ err:
 /* Compute the product of two polynomials a and b, reduce modulo p, and store
  * the result in r.  r could be a or b; a could be b.
  */
-int	BN_GF2m_mod_mul_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const unsigned int p[], BN_CTX *ctx)
+int	BN_GF2m_mod_mul_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const int p[], BN_CTX *ctx)
 	{
 	int zlen, i, j, k, ret = 0;
 	BIGNUM *s;
@@ -499,12 +438,12 @@ err:
 int	BN_GF2m_mod_mul(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const BIGNUM *p, BN_CTX *ctx)
 	{
 	int ret = 0;
-	const int max = BN_num_bits(p);
-	unsigned int *arr=NULL;
+	const int max = BN_num_bits(p) + 1;
+	int *arr=NULL;
 	bn_check_top(a);
 	bn_check_top(b);
 	bn_check_top(p);
-	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;
+	if ((arr = (int *)OPENSSL_malloc(sizeof(int) * max)) == NULL) goto err;
 	ret = BN_GF2m_poly2arr(p, arr, max);
 	if (!ret || ret > max)
 		{
@@ -520,7 +459,7 @@ err:
 
 
 /* Square a, reduce the result mod p, and store it in a.  r could be a. */
-int	BN_GF2m_mod_sqr_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[], BN_CTX *ctx)
+int	BN_GF2m_mod_sqr_arr(BIGNUM *r, const BIGNUM *a, const int p[], BN_CTX *ctx)
 	{
 	int i, ret = 0;
 	BIGNUM *s;
@@ -555,12 +494,12 @@ err:
 int	BN_GF2m_mod_sqr(BIGNUM *r, const BIGNUM *a, const BIGNUM *p, BN_CTX *ctx)
 	{
 	int ret = 0;
-	const int max = BN_num_bits(p);
-	unsigned int *arr=NULL;
+	const int max = BN_num_bits(p) + 1;
+	int *arr=NULL;
 
 	bn_check_top(a);
 	bn_check_top(p);
-	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;
+	if ((arr = (int *)OPENSSL_malloc(sizeof(int) * max)) == NULL) goto err;
 	ret = BN_GF2m_poly2arr(p, arr, max);
 	if (!ret || ret > max)
 		{
@@ -642,7 +581,7 @@ err:
  * function is only provided for convenience; for best performance, use the 
  * BN_GF2m_mod_inv function.
  */
-int BN_GF2m_mod_inv_arr(BIGNUM *r, const BIGNUM *xx, const unsigned int p[], BN_CTX *ctx)
+int BN_GF2m_mod_inv_arr(BIGNUM *r, const BIGNUM *xx, const int p[], BN_CTX *ctx)
 	{
 	BIGNUM *field;
 	int ret = 0;
@@ -768,7 +707,7 @@ err:
  * function is only provided for convenience; for best performance, use the 
  * BN_GF2m_mod_div function.
  */
-int BN_GF2m_mod_div_arr(BIGNUM *r, const BIGNUM *yy, const BIGNUM *xx, const unsigned int p[], BN_CTX *ctx)
+int BN_GF2m_mod_div_arr(BIGNUM *r, const BIGNUM *yy, const BIGNUM *xx, const int p[], BN_CTX *ctx)
 	{
 	BIGNUM *field;
 	int ret = 0;
@@ -793,7 +732,7 @@ err:
  * the result in r.  r could be a.
  * Uses simple square-and-multiply algorithm A.5.1 from IEEE P1363.
  */
-int	BN_GF2m_mod_exp_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const unsigned int p[], BN_CTX *ctx)
+int	BN_GF2m_mod_exp_arr(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const int p[], BN_CTX *ctx)
 	{
 	int ret = 0, i, n;
 	BIGNUM *u;
@@ -839,12 +778,12 @@ err:
 int BN_GF2m_mod_exp(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, const BIGNUM *p, BN_CTX *ctx)
 	{
 	int ret = 0;
-	const int max = BN_num_bits(p);
-	unsigned int *arr=NULL;
+	const int max = BN_num_bits(p) + 1;
+	int *arr=NULL;
 	bn_check_top(a);
 	bn_check_top(b);
 	bn_check_top(p);
-	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;
+	if ((arr = (int *)OPENSSL_malloc(sizeof(int) * max)) == NULL) goto err;
 	ret = BN_GF2m_poly2arr(p, arr, max);
 	if (!ret || ret > max)
 		{
@@ -862,7 +801,7 @@ err:
  * the result in r.  r could be a.
  * Uses exponentiation as in algorithm A.4.1 from IEEE P1363.
  */
-int	BN_GF2m_mod_sqrt_arr(BIGNUM *r, const BIGNUM *a, const unsigned int p[], BN_CTX *ctx)
+int	BN_GF2m_mod_sqrt_arr(BIGNUM *r, const BIGNUM *a, const int p[], BN_CTX *ctx)
 	{
 	int ret = 0;
 	BIGNUM *u;
@@ -898,11 +837,11 @@ err:
 int BN_GF2m_mod_sqrt(BIGNUM *r, const BIGNUM *a, const BIGNUM *p, BN_CTX *ctx)
 	{
 	int ret = 0;
-	const int max = BN_num_bits(p);
-	unsigned int *arr=NULL;
+	const int max = BN_num_bits(p) + 1;
+	int *arr=NULL;
 	bn_check_top(a);
 	bn_check_top(p);
-	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) * max)) == NULL) goto err;
+	if ((arr = (int *)OPENSSL_malloc(sizeof(int) * max)) == NULL) goto err;
 	ret = BN_GF2m_poly2arr(p, arr, max);
 	if (!ret || ret > max)
 		{
@@ -919,10 +858,9 @@ err:
 /* Find r such that r^2 + r = a mod p.  r could be a. If no r exists returns 0.
  * Uses algorithms A.4.7 and A.4.6 from IEEE P1363.
  */
-int BN_GF2m_mod_solve_quad_arr(BIGNUM *r, const BIGNUM *a_, const unsigned int p[], BN_CTX *ctx)
+int BN_GF2m_mod_solve_quad_arr(BIGNUM *r, const BIGNUM *a_, const int p[], BN_CTX *ctx)
 	{
-	int ret = 0, count = 0;
-	unsigned int j;
+	int ret = 0, count = 0, j;
 	BIGNUM *a, *z, *rho, *w, *w2, *tmp;
 
 	bn_check_top(a_);
@@ -1017,11 +955,11 @@ err:
 int BN_GF2m_mod_solve_quad(BIGNUM *r, const BIGNUM *a, const BIGNUM *p, BN_CTX *ctx)
 	{
 	int ret = 0;
-	const int max = BN_num_bits(p);
-	unsigned int *arr=NULL;
+	const int max = BN_num_bits(p) + 1;
+	int *arr=NULL;
 	bn_check_top(a);
 	bn_check_top(p);
-	if ((arr = (unsigned int *)OPENSSL_malloc(sizeof(unsigned int) *
+	if ((arr = (int *)OPENSSL_malloc(sizeof(int) *
 						max)) == NULL) goto err;
 	ret = BN_GF2m_poly2arr(p, arr, max);
 	if (!ret || ret > max)
@@ -1037,20 +975,17 @@ err:
 	}
 
 /* Convert the bit-string representation of a polynomial
- * ( \sum_{i=0}^n a_i * x^i , where a_0 is *not* zero) into an array
- * of integers corresponding to the bits with non-zero coefficient.
+ * ( \sum_{i=0}^n a_i * x^i) into an array of integers corresponding 
+ * to the bits with non-zero coefficient.  Array is terminated with -1.
  * Up to max elements of the array will be filled.  Return value is total
- * number of coefficients that would be extracted if array was large enough.
+ * number of array elements that would be filled if array was large enough.
  */
-int BN_GF2m_poly2arr(const BIGNUM *a, unsigned int p[], int max)
+int BN_GF2m_poly2arr(const BIGNUM *a, int p[], int max)
 	{
 	int i, j, k = 0;
 	BN_ULONG mask;
 
-	if (BN_is_zero(a) || !BN_is_bit_set(a, 0))
-		/* a_0 == 0 => return error (the unsigned int array
-		 * must be terminated by 0)
-		 */
+	if (BN_is_zero(a))
 		return 0;
 
 	for (i = a->top - 1; i >= 0; i--)
@@ -1070,24 +1005,28 @@ int BN_GF2m_poly2arr(const BIGNUM *a, unsigned int p[], int max)
 			}
 		}
 
+	if (k < max) {
+		p[k] = -1;
+		k++;
+	}
+
 	return k;
 	}
 
 /* Convert the coefficient array representation of a polynomial to a 
- * bit-string.  The array must be terminated by 0.
+ * bit-string.  The array must be terminated by -1.
  */
-int BN_GF2m_arr2poly(const unsigned int p[], BIGNUM *a)
+int BN_GF2m_arr2poly(const int p[], BIGNUM *a)
 	{
 	int i;
 
 	bn_check_top(a);
 	BN_zero(a);
-	for (i = 0; p[i] != 0; i++)
+	for (i = 0; p[i] != -1; i++)
 		{
 		if (BN_set_bit(a, p[i]) == 0)
 			return 0;
 		}
-	BN_set_bit(a, 0);
 	bn_check_top(a);
 
 	return 1;
diff --git a/openssl/crypto/bn/bn_lcl.h b/openssl/crypto/bn/bn_lcl.h
index 27ac4397a..8e5e98e3f 100644
--- a/openssl/crypto/bn/bn_lcl.h
+++ b/openssl/crypto/bn/bn_lcl.h
@@ -255,7 +255,8 @@ extern "C" {
 	     : "r"(a), "r"(b));		\
 	ret;			})
 #  endif	/* compiler */
-# elif defined(__x86_64) && defined(SIXTY_FOUR_BIT_LONG)
+# elif (defined(__x86_64) || defined(__x86_64__)) && \
+       (defined(SIXTY_FOUR_BIT_LONG) || defined(SIXTY_FOUR_BIT))
 #  if defined(__GNUC__)
 #   define BN_UMULT_HIGH(a,b)	({	\
 	register BN_ULONG ret,discard;	\
diff --git a/openssl/crypto/bn/bn_lib.c b/openssl/crypto/bn/bn_lib.c
index 32a8fbaf5..5470fbe6e 100644
--- a/openssl/crypto/bn/bn_lib.c
+++ b/openssl/crypto/bn/bn_lib.c
@@ -133,15 +133,34 @@ int BN_get_params(int which)
 
 const BIGNUM *BN_value_one(void)
 	{
-	static BN_ULONG data_one=1L;
-	static BIGNUM const_one={&data_one,1,1,0,BN_FLG_STATIC_DATA};
+	static const BN_ULONG data_one=1L;
+	static const BIGNUM const_one={(BN_ULONG *)&data_one,1,1,0,BN_FLG_STATIC_DATA};
 
 	return(&const_one);
 	}
 
+char *BN_options(void)
+	{
+	static int init=0;
+	static char data[16];
+
+	if (!init)
+		{
+		init++;
+#ifdef BN_LLONG
+		BIO_snprintf(data,sizeof data,"bn(%d,%d)",
+			     (int)sizeof(BN_ULLONG)*8,(int)sizeof(BN_ULONG)*8);
+#else
+		BIO_snprintf(data,sizeof data,"bn(%d,%d)",
+			     (int)sizeof(BN_ULONG)*8,(int)sizeof(BN_ULONG)*8);
+#endif
+		}
+	return(data);
+	}
+
 int BN_num_bits_word(BN_ULONG l)
 	{
-	static const char bits[256]={
+	static const unsigned char bits[256]={
 		0,1,2,2,3,3,3,3,4,4,4,4,4,4,4,4,
 		5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,5,
 		6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,6,
@@ -216,7 +235,7 @@ int BN_num_bits_word(BN_ULONG l)
 		else
 #endif
 			{
-#if defined(SIXTEEN_BIT) || defined(THIRTY_TWO_BIT) || defined(SIXTY_FOUR_BIT) || defined(SIXTY_FOUR_BIT_LONG)
+#if defined(THIRTY_TWO_BIT) || defined(SIXTY_FOUR_BIT) || defined(SIXTY_FOUR_BIT_LONG)
 			if (l & 0xff00L)
 				return(bits[(int)(l>>8)]+8);
 			else	
@@ -744,7 +763,7 @@ int BN_is_bit_set(const BIGNUM *a, int n)
 	i=n/BN_BITS2;
 	j=n%BN_BITS2;
 	if (a->top <= i) return 0;
-	return(((a->d[i])>>j)&((BN_ULONG)1));
+	return (int)(((a->d[i])>>j)&((BN_ULONG)1));
 	}
 
 int BN_mask_bits(BIGNUM *a, int n)
diff --git a/openssl/crypto/bn/bn_mont.c b/openssl/crypto/bn/bn_mont.c
index 4799b152d..7224637ab 100644
--- a/openssl/crypto/bn/bn_mont.c
+++ b/openssl/crypto/bn/bn_mont.c
@@ -122,26 +122,10 @@
 
 #define MONT_WORD /* use the faster word-based algorithm */
 
-#if defined(MONT_WORD) && defined(OPENSSL_BN_ASM_MONT) && (BN_BITS2<=32)
-/* This condition means we have a specific non-default build:
- * In the 0.9.8 branch, OPENSSL_BN_ASM_MONT is normally not set for any
- * BN_BITS2<=32 platform; an explicit "enable-montasm" is required.
- * I.e., if we are here, the user intentionally deviates from the
- * normal stable build to get better Montgomery performance from
- * the 0.9.9-dev backport.
- *
- * In this case only, we also enable BN_from_montgomery_word()
- * (another non-stable feature from 0.9.9-dev).
- */
-#define MONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD
-#endif
-
-#ifdef MONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD
+#ifdef MONT_WORD
 static int BN_from_montgomery_word(BIGNUM *ret, BIGNUM *r, BN_MONT_CTX *mont);
 #endif
 
-
-
 int BN_mod_mul_montgomery(BIGNUM *r, const BIGNUM *a, const BIGNUM *b,
 			  BN_MONT_CTX *mont, BN_CTX *ctx)
 	{
@@ -153,11 +137,7 @@ int BN_mod_mul_montgomery(BIGNUM *r, const BIGNUM *a, const BIGNUM *b,
 	if (num>1 && a->top==num && b->top==num)
 		{
 		if (bn_wexpand(r,num) == NULL) return(0);
-#if 0 /* for OpenSSL 0.9.9 mont->n0 */
 		if (bn_mul_mont(r->d,a->d,b->d,mont->N.d,mont->n0,num))
-#else
-		if (bn_mul_mont(r->d,a->d,b->d,mont->N.d,&mont->n0,num))
-#endif
 			{
 			r->neg = a->neg^b->neg;
 			r->top = num;
@@ -181,7 +161,7 @@ int BN_mod_mul_montgomery(BIGNUM *r, const BIGNUM *a, const BIGNUM *b,
 		if (!BN_mul(tmp,a,b,ctx)) goto err;
 		}
 	/* reduce from aRR to aR */
-#ifdef MONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD
+#ifdef MONT_WORD
 	if (!BN_from_montgomery_word(r,tmp,mont)) goto err;
 #else
 	if (!BN_from_montgomery(r,tmp,mont,ctx)) goto err;
@@ -193,7 +173,7 @@ err:
 	return(ret);
 	}
 
-#ifdef MONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD
+#ifdef MONT_WORD
 static int BN_from_montgomery_word(BIGNUM *ret, BIGNUM *r, BN_MONT_CTX *mont)
 	{
 	BIGNUM *n;
@@ -217,15 +197,15 @@ static int BN_from_montgomery_word(BIGNUM *ret, BIGNUM *r, BN_MONT_CTX *mont)
 	nrp= &(r->d[nl]);
 
 	/* clear the top words of T */
+#if 1
 	for (i=r->top; i<max; i++) /* memset? XXX */
 		r->d[i]=0;
+#else
+	memset(&(r->d[r->top]),0,(max-r->top)*sizeof(BN_ULONG)); 
+#endif
 
 	r->top=max;
-#if 0 /* for OpenSSL 0.9.9 mont->n0 */
 	n0=mont->n0[0];
-#else
-	n0=mont->n0;
-#endif
 
 #ifdef BN_COUNT
 	fprintf(stderr,"word BN_from_montgomery_word %d * %d\n",nl,nl);
@@ -270,6 +250,8 @@ static int BN_from_montgomery_word(BIGNUM *ret, BIGNUM *r, BN_MONT_CTX *mont)
 		}
 	al=r->top-ri;
 
+#define BRANCH_FREE 1
+#if BRANCH_FREE
 	if (bn_wexpand(ret,ri) == NULL) return(0);
 	x=0-(((al-ri)>>(sizeof(al)*8-1))&1);
 	ret->top=x=(ri&~x)|(al&x);	/* min(ri,al) */
@@ -317,164 +299,8 @@ static int BN_from_montgomery_word(BIGNUM *ret, BIGNUM *r, BN_MONT_CTX *mont)
 		rp[i]=nrp[i], ap[i]=0;
 	bn_correct_top(r);
 	bn_correct_top(ret);
-	bn_check_top(ret);
-
-	return(1);
-	}
-
-int BN_from_montgomery(BIGNUM *ret, const BIGNUM *a, BN_MONT_CTX *mont,
-	     BN_CTX *ctx)
-	{
-	int retn=0;
-	BIGNUM *t;
-
-	BN_CTX_start(ctx);
-	if ((t = BN_CTX_get(ctx)) && BN_copy(t,a))
-		retn = BN_from_montgomery_word(ret,t,mont);
-	BN_CTX_end(ctx);
-	return retn;
-	}
-
-#else /* !MONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD */
-
-int BN_from_montgomery(BIGNUM *ret, const BIGNUM *a, BN_MONT_CTX *mont,
-	     BN_CTX *ctx)
-	{
-	int retn=0;
-
-#ifdef MONT_WORD
-	BIGNUM *n,*r;
-	BN_ULONG *ap,*np,*rp,n0,v,*nrp;
-	int al,nl,max,i,x,ri;
-
-	BN_CTX_start(ctx);
-	if ((r = BN_CTX_get(ctx)) == NULL) goto err;
-
-	if (!BN_copy(r,a)) goto err;
-	n= &(mont->N);
-
-	ap=a->d;
-	/* mont->ri is the size of mont->N in bits (rounded up
-	   to the word size) */
-	al=ri=mont->ri/BN_BITS2;
-	
-	nl=n->top;
-	if ((al == 0) || (nl == 0)) { r->top=0; return(1); }
-
-	max=(nl+al+1); /* allow for overflow (no?) XXX */
-	if (bn_wexpand(r,max) == NULL) goto err;
-
-	r->neg=a->neg^n->neg;
-	np=n->d;
-	rp=r->d;
-	nrp= &(r->d[nl]);
-
-	/* clear the top words of T */
-#if 1
-	for (i=r->top; i<max; i++) /* memset? XXX */
-		r->d[i]=0;
 #else
-	memset(&(r->d[r->top]),0,(max-r->top)*sizeof(BN_ULONG)); 
-#endif
-
-	r->top=max;
-	n0=mont->n0;
-
-#ifdef BN_COUNT
-	fprintf(stderr,"word BN_from_montgomery %d * %d\n",nl,nl);
-#endif
-	for (i=0; i<nl; i++)
-		{
-#ifdef __TANDEM
-                {
-                   long long t1;
-                   long long t2;
-                   long long t3;
-                   t1 = rp[0] * (n0 & 0177777);
-                   t2 = 037777600000l;
-                   t2 = n0 & t2;
-                   t3 = rp[0] & 0177777;
-                   t2 = (t3 * t2) & BN_MASK2;
-                   t1 = t1 + t2;
-                   v=bn_mul_add_words(rp,np,nl,(BN_ULONG) t1);
-                }
-#else
-		v=bn_mul_add_words(rp,np,nl,(rp[0]*n0)&BN_MASK2);
-#endif
-		nrp++;
-		rp++;
-		if (((nrp[-1]+=v)&BN_MASK2) >= v)
-			continue;
-		else
-			{
-			if (((++nrp[0])&BN_MASK2) != 0) continue;
-			if (((++nrp[1])&BN_MASK2) != 0) continue;
-			for (x=2; (((++nrp[x])&BN_MASK2) == 0); x++) ;
-			}
-		}
-	bn_correct_top(r);
-	
-	/* mont->ri will be a multiple of the word size and below code
-	 * is kind of BN_rshift(ret,r,mont->ri) equivalent */
-	if (r->top <= ri)
-		{
-		ret->top=0;
-		retn=1;
-		goto err;
-		}
-	al=r->top-ri;
-
-# define BRANCH_FREE 1
-# if BRANCH_FREE
-	if (bn_wexpand(ret,ri) == NULL) goto err;
-	x=0-(((al-ri)>>(sizeof(al)*8-1))&1);
-	ret->top=x=(ri&~x)|(al&x);	/* min(ri,al) */
-	ret->neg=r->neg;
-
-	rp=ret->d;
-	ap=&(r->d[ri]);
-
-	{
-	size_t m1,m2;
-
-	v=bn_sub_words(rp,ap,np,ri);
-	/* this ----------------^^ works even in al<ri case
-	 * thanks to zealous zeroing of top of the vector in the
-	 * beginning. */
-
-	/* if (al==ri && !v) || al>ri) nrp=rp; else nrp=ap; */
-	/* in other words if subtraction result is real, then
-	 * trick unconditional memcpy below to perform in-place
-	 * "refresh" instead of actual copy. */
-	m1=0-(size_t)(((al-ri)>>(sizeof(al)*8-1))&1);	/* al<ri */
-	m2=0-(size_t)(((ri-al)>>(sizeof(al)*8-1))&1);	/* al>ri */
-	m1|=m2;			/* (al!=ri) */
-	m1|=(0-(size_t)v);	/* (al!=ri || v) */
-	m1&=~m2;		/* (al!=ri || v) && !al>ri */
-	nrp=(BN_ULONG *)(((size_t)rp&~m1)|((size_t)ap&m1));
-	}
-
-	/* 'i<ri' is chosen to eliminate dependency on input data, even
-	 * though it results in redundant copy in al<ri case. */
-	for (i=0,ri-=4; i<ri; i+=4)
-		{
-		BN_ULONG t1,t2,t3,t4;
-		
-		t1=nrp[i+0];
-		t2=nrp[i+1];
-		t3=nrp[i+2];	ap[i+0]=0;
-		t4=nrp[i+3];	ap[i+1]=0;
-		rp[i+0]=t1;	ap[i+2]=0;
-		rp[i+1]=t2;	ap[i+3]=0;
-		rp[i+2]=t3;
-		rp[i+3]=t4;
-		}
-	for (ri+=4; i<ri; i++)
-		rp[i]=nrp[i], ap[i]=0;
-	bn_correct_top(r);
-	bn_correct_top(ret);
-# else
-	if (bn_wexpand(ret,al) == NULL) goto err;
+	if (bn_wexpand(ret,al) == NULL) return(0);
 	ret->top=al;
 	ret->neg=r->neg;
 
@@ -497,8 +323,30 @@ int BN_from_montgomery(BIGNUM *ret, const BIGNUM *a, BN_MONT_CTX *mont,
 	al+=4;
 	for (; i<al; i++)
 		rp[i]=ap[i];
-# endif
-#else /* !MONT_WORD */ 
+
+	if (BN_ucmp(ret, &(mont->N)) >= 0)
+		{
+		if (!BN_usub(ret,ret,&(mont->N))) return(0);
+		}
+#endif
+	bn_check_top(ret);
+
+	return(1);
+	}
+#endif	/* MONT_WORD */
+
+int BN_from_montgomery(BIGNUM *ret, const BIGNUM *a, BN_MONT_CTX *mont,
+	     BN_CTX *ctx)
+	{
+	int retn=0;
+#ifdef MONT_WORD
+	BIGNUM *t;
+
+	BN_CTX_start(ctx);
+	if ((t = BN_CTX_get(ctx)) && BN_copy(t,a))
+		retn = BN_from_montgomery_word(ret,t,mont);
+	BN_CTX_end(ctx);
+#else /* !MONT_WORD */
 	BIGNUM *t1,*t2;
 
 	BN_CTX_start(ctx);
@@ -515,21 +363,18 @@ int BN_from_montgomery(BIGNUM *ret, const BIGNUM *a, BN_MONT_CTX *mont,
 	if (!BN_mul(t1,t2,&mont->N,ctx)) goto err;
 	if (!BN_add(t2,a,t1)) goto err;
 	if (!BN_rshift(ret,t2,mont->ri)) goto err;
-#endif /* MONT_WORD */
 
-#if !defined(BRANCH_FREE) || BRANCH_FREE==0
 	if (BN_ucmp(ret, &(mont->N)) >= 0)
 		{
 		if (!BN_usub(ret,ret,&(mont->N))) goto err;
 		}
-#endif
 	retn=1;
 	bn_check_top(ret);
  err:
 	BN_CTX_end(ctx);
+#endif /* MONT_WORD */
 	return(retn);
 	}
-#endif /* MONT_FROM_WORD___NON_DEFAULT_0_9_8_BUILD */
 
 BN_MONT_CTX *BN_MONT_CTX_new(void)
 	{
@@ -549,11 +394,7 @@ void BN_MONT_CTX_init(BN_MONT_CTX *ctx)
 	BN_init(&(ctx->RR));
 	BN_init(&(ctx->N));
 	BN_init(&(ctx->Ni));
-#if 0 /* for OpenSSL 0.9.9 mont->n0 */
 	ctx->n0[0] = ctx->n0[1] = 0;
-#else
-	ctx->n0 = 0;
-#endif
 	ctx->flags=0;
 	}
 
@@ -585,26 +426,22 @@ int BN_MONT_CTX_set(BN_MONT_CTX *mont, const BIGNUM *mod, BN_CTX *ctx)
 		BIGNUM tmod;
 		BN_ULONG buf[2];
 
-		mont->ri=(BN_num_bits(mod)+(BN_BITS2-1))/BN_BITS2*BN_BITS2;
-		BN_zero(R);
-#if 0 /* for OpenSSL 0.9.9 mont->n0, would be "#if defined(OPENSSL_BN_ASM_MONT) && (BN_BITS2<=32)",
-         only certain BN_BITS2<=32 platforms actually need this */
-		if (!(BN_set_bit(R,2*BN_BITS2))) goto err;	/* R */
-#else
-		if (!(BN_set_bit(R,BN_BITS2))) goto err;	/* R */
-#endif
-
-		buf[0]=mod->d[0]; /* tmod = N mod word size */
-		buf[1]=0;
-
 		BN_init(&tmod);
 		tmod.d=buf;
-		tmod.top = buf[0] != 0 ? 1 : 0;
 		tmod.dmax=2;
 		tmod.neg=0;
 
-#if 0 /* for OpenSSL 0.9.9 mont->n0, would be "#if defined(OPENSSL_BN_ASM_MONT) && (BN_BITS2<=32)";
-         only certain BN_BITS2<=32 platforms actually need this */
+		mont->ri=(BN_num_bits(mod)+(BN_BITS2-1))/BN_BITS2*BN_BITS2;
+
+#if defined(OPENSSL_BN_ASM_MONT) && (BN_BITS2<=32)
+		/* Only certain BN_BITS2<=32 platforms actually make use of
+		 * n0[1], and we could use the #else case (with a shorter R
+		 * value) for the others.  However, currently only the assembler
+		 * files do know which is which. */
+
+		BN_zero(R);
+		if (!(BN_set_bit(R,2*BN_BITS2))) goto err;
+
 								tmod.top=0;
 		if ((buf[0] = mod->d[0]))			tmod.top=1;
 		if ((buf[1] = mod->top>1 ? mod->d[1] : 0))	tmod.top=2;
@@ -632,6 +469,12 @@ int BN_MONT_CTX_set(BN_MONT_CTX *mont, const BIGNUM *mod, BN_CTX *ctx)
 		mont->n0[0] = (Ri->top > 0) ? Ri->d[0] : 0;
 		mont->n0[1] = (Ri->top > 1) ? Ri->d[1] : 0;
 #else
+		BN_zero(R);
+		if (!(BN_set_bit(R,BN_BITS2))) goto err;	/* R */
+
+		buf[0]=mod->d[0]; /* tmod = N mod word size */
+		buf[1]=0;
+		tmod.top = buf[0] != 0 ? 1 : 0;
 							/* Ri = R^-1 mod N*/
 		if ((BN_mod_inverse(Ri,R,&tmod,ctx)) == NULL)
 			goto err;
@@ -647,12 +490,8 @@ int BN_MONT_CTX_set(BN_MONT_CTX *mont, const BIGNUM *mod, BN_CTX *ctx)
 		if (!BN_div(Ri,NULL,Ri,&tmod,ctx)) goto err;
 		/* Ni = (R*Ri-1)/N,
 		 * keep only least significant word: */
-# if 0 /* for OpenSSL 0.9.9 mont->n0 */
 		mont->n0[0] = (Ri->top > 0) ? Ri->d[0] : 0;
 		mont->n0[1] = 0;
-# else
-		mont->n0 = (Ri->top > 0) ? Ri->d[0] : 0;
-# endif
 #endif
 		}
 #else /* !MONT_WORD */
@@ -689,12 +528,8 @@ BN_MONT_CTX *BN_MONT_CTX_copy(BN_MONT_CTX *to, BN_MONT_CTX *from)
 	if (!BN_copy(&(to->N),&(from->N))) return NULL;
 	if (!BN_copy(&(to->Ni),&(from->Ni))) return NULL;
 	to->ri=from->ri;
-#if 0 /* for OpenSSL 0.9.9 mont->n0 */
 	to->n0[0]=from->n0[0];
 	to->n0[1]=from->n0[1];
-#else
-	to->n0=from->n0;
-#endif
 	return(to);
 	}
 
diff --git a/openssl/crypto/bn/bn_mul.c b/openssl/crypto/bn/bn_mul.c
index b848c8cc6..a0e9ec3b4 100644
--- a/openssl/crypto/bn/bn_mul.c
+++ b/openssl/crypto/bn/bn_mul.c
@@ -1028,17 +1028,19 @@ int BN_mul(BIGNUM *r, const BIGNUM *a, const BIGNUM *b, BN_CTX *ctx)
 			assert(j <= al || j <= bl);
 			k = j+j;
 			t = BN_CTX_get(ctx);
+			if (t == NULL)
+				goto err;
 			if (al > j || bl > j)
 				{
-				bn_wexpand(t,k*4);
-				bn_wexpand(rr,k*4);
+				if (bn_wexpand(t,k*4) == NULL) goto err;
+				if (bn_wexpand(rr,k*4) == NULL) goto err;
 				bn_mul_part_recursive(rr->d,a->d,b->d,
 					j,al-j,bl-j,t->d);
 				}
 			else	/* al <= j || bl <= j */
 				{
-				bn_wexpand(t,k*2);
-				bn_wexpand(rr,k*2);
+				if (bn_wexpand(t,k*2) == NULL) goto err;
+				if (bn_wexpand(rr,k*2) == NULL) goto err;
 				bn_mul_recursive(rr->d,a->d,b->d,
 					j,al-j,bl-j,t->d);
 				}
diff --git a/openssl/crypto/bn/bn_opt.c b/openssl/crypto/bn/bn_opt.c
deleted file mode 100644
index 21cbb38f6..000000000
--- a/openssl/crypto/bn/bn_opt.c
+++ /dev/null
@@ -1,87 +0,0 @@
-/* crypto/bn/bn_opt.c */
-/* Copyright (C) 1995-1998 Eric Young (eay@cryptsoft.com)
- * All rights reserved.
- *
- * This package is an SSL implementation written
- * by Eric Young (eay@cryptsoft.com).
- * The implementation was written so as to conform with Netscapes SSL.
- * 
- * This library is free for commercial and non-commercial use as long as
- * the following conditions are aheared to.  The following conditions
- * apply to all code found in this distribution, be it the RC4, RSA,
- * lhash, DES, etc., code; not just the SSL code.  The SSL documentation
- * included with this distribution is covered by the same copyright terms
- * except that the holder is Tim Hudson (tjh@cryptsoft.com).
- * 
- * Copyright remains Eric Young's, and as such any Copyright notices in
- * the code are not to be removed.
- * If this package is used in a product, Eric Young should be given attribution
- * as the author of the parts of the library used.
- * This can be in the form of a textual message at program startup or
- * in documentation (online or textual) provided with the package.
- * 
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- * 1. Redistributions of source code must retain the copyright
- *    notice, this list of conditions and the following disclaimer.
- * 2. Redistributions in binary form must reproduce the above copyright
- *    notice, this list of conditions and the following disclaimer in the
- *    documentation and/or other materials provided with the distribution.
- * 3. All advertising materials mentioning features or use of this software
- *    must display the following acknowledgement:
- *    "This product includes cryptographic software written by
- *     Eric Young (eay@cryptsoft.com)"
- *    The word 'cryptographic' can be left out if the rouines from the library
- *    being used are not cryptographic related :-).
- * 4. If you include any Windows specific code (or a derivative thereof) from 
- *    the apps directory (application code) you must include an acknowledgement:
- *    "This product includes software written by Tim Hudson (tjh@cryptsoft.com)"
- * 
- * THIS SOFTWARE IS PROVIDED BY ERIC YOUNG ``AS IS'' AND
- * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
- * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
- * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
- * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
- * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
- * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
- * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
- * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
- * SUCH DAMAGE.
- * 
- * The licence and distribution terms for any publically available version or
- * derivative of this code cannot be changed.  i.e. this code cannot simply be
- * copied and put under another distribution licence
- * [including the GNU Public Licence.]
- */
-
-#ifndef BN_DEBUG
-# undef NDEBUG /* avoid conflicting definitions */
-# define NDEBUG
-#endif
-
-#include <assert.h>
-#include <limits.h>
-#include <stdio.h>
-#include "cryptlib.h"
-#include "bn_lcl.h"
-
-char *BN_options(void)
-	{
-	static int init=0;
-	static char data[16];
-
-	if (!init)
-		{
-		init++;
-#ifdef BN_LLONG
-		BIO_snprintf(data,sizeof data,"bn(%d,%d)",
-			     (int)sizeof(BN_ULLONG)*8,(int)sizeof(BN_ULONG)*8);
-#else
-		BIO_snprintf(data,sizeof data,"bn(%d,%d)",
-			     (int)sizeof(BN_ULONG)*8,(int)sizeof(BN_ULONG)*8);
-#endif
-		}
-	return(data);
-	}
diff --git a/openssl/crypto/bn/bn_print.c b/openssl/crypto/bn/bn_print.c
index 810dde34e..bebb466d0 100644
--- a/openssl/crypto/bn/bn_print.c
+++ b/openssl/crypto/bn/bn_print.c
@@ -294,6 +294,27 @@ err:
 	return(0);
 	}
 
+int BN_asc2bn(BIGNUM **bn, const char *a)
+	{
+	const char *p = a;
+	if (*p == '-')
+		p++;
+
+	if (p[0] == '0' && (p[1] == 'X' || p[1] == 'x'))
+		{		
+		if (!BN_hex2bn(bn, p + 2))
+			return 0;
+		}
+	else
+		{
+		if (!BN_dec2bn(bn, p))
+			return 0;
+		}
+	if (*a == '-')
+		(*bn)->neg = 1;
+	return 1;
+	}
+
 #ifndef OPENSSL_NO_BIO
 #ifndef OPENSSL_NO_FP_API
 int BN_print_fp(FILE *fp, const BIGNUM *a)
diff --git a/openssl/crypto/bn/bn_x931p.c b/openssl/crypto/bn/bn_x931p.c
deleted file mode 100644
index 04c5c874e..000000000
--- a/openssl/crypto/bn/bn_x931p.c
+++ /dev/null
@@ -1,272 +0,0 @@
-/* bn_x931p.c */
-/* Written by Dr Stephen N Henson (steve@openssl.org) for the OpenSSL
- * project 2005.
- */
-/* ====================================================================
- * Copyright (c) 2005 The OpenSSL Project.  All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- *
- * 1. Redistributions of source code must retain the above copyright
- *    notice, this list of conditions and the following disclaimer. 
- *
- * 2. Redistributions in binary form must reproduce the above copyright
- *    notice, this list of conditions and the following disclaimer in
- *    the documentation and/or other materials provided with the
- *    distribution.
- *
- * 3. All advertising materials mentioning features or use of this
- *    software must display the following acknowledgment:
- *    "This product includes software developed by the OpenSSL Project
- *    for use in the OpenSSL Toolkit. (http://www.OpenSSL.org/)"
- *
- * 4. The names "OpenSSL Toolkit" and "OpenSSL Project" must not be used to
- *    endorse or promote products derived from this software without
- *    prior written permission. For written permission, please contact
- *    licensing@OpenSSL.org.
- *
- * 5. Products derived from this software may not be called "OpenSSL"
- *    nor may "OpenSSL" appear in their names without prior written
- *    permission of the OpenSSL Project.
- *
- * 6. Redistributions of any form whatsoever must retain the following
- *    acknowledgment:
- *    "This product includes software developed by the OpenSSL Project
- *    for use in the OpenSSL Toolkit (http://www.OpenSSL.org/)"
- *
- * THIS SOFTWARE IS PROVIDED BY THE OpenSSL PROJECT ``AS IS'' AND ANY
- * EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
- * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
- * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE OpenSSL PROJECT OR
- * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
- * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
- * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
- * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
- * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
- * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
- * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
- * OF THE POSSIBILITY OF SUCH DAMAGE.
- * ====================================================================
- *
- * This product includes cryptographic software written by Eric Young
- * (eay@cryptsoft.com).  This product includes software written by Tim
- * Hudson (tjh@cryptsoft.com).
- *
- */
-
-#include <stdio.h>
-#include <openssl/bn.h>
-
-/* X9.31 routines for prime derivation */
-
-/* X9.31 prime derivation. This is used to generate the primes pi
- * (p1, p2, q1, q2) from a parameter Xpi by checking successive odd
- * integers.
- */
-
-static int bn_x931_derive_pi(BIGNUM *pi, const BIGNUM *Xpi, BN_CTX *ctx,
-			BN_GENCB *cb)
-	{
-	int i = 0;
-	if (!BN_copy(pi, Xpi))
-		return 0;
-	if (!BN_is_odd(pi) && !BN_add_word(pi, 1))
-		return 0;
-	for(;;)
-		{
-		i++;
-		BN_GENCB_call(cb, 0, i);
-		/* NB 27 MR is specificed in X9.31 */
-		if (BN_is_prime_fasttest_ex(pi, 27, ctx, 1, cb))
-			break;
-		if (!BN_add_word(pi, 2))
-			return 0;
-		}
-	BN_GENCB_call(cb, 2, i);
-	return 1;
-	}
-
-/* This is the main X9.31 prime derivation function. From parameters
- * Xp1, Xp2 and Xp derive the prime p. If the parameters p1 or p2 are
- * not NULL they will be returned too: this is needed for testing.
- */
-
-int BN_X931_derive_prime_ex(BIGNUM *p, BIGNUM *p1, BIGNUM *p2,
-			const BIGNUM *Xp, const BIGNUM *Xp1, const BIGNUM *Xp2,
-			const BIGNUM *e, BN_CTX *ctx, BN_GENCB *cb)
-	{
-	int ret = 0;
-
-	BIGNUM *t, *p1p2, *pm1;
-
-	/* Only even e supported */
-	if (!BN_is_odd(e))
-		return 0;
-
-	BN_CTX_start(ctx);
-	if (!p1)
-		p1 = BN_CTX_get(ctx);
-
-	if (!p2)
-		p2 = BN_CTX_get(ctx);
-
-	t = BN_CTX_get(ctx);
-
-	p1p2 = BN_CTX_get(ctx);
-
-	pm1 = BN_CTX_get(ctx);
-
-	if (!bn_x931_derive_pi(p1, Xp1, ctx, cb))
-		goto err;
-
-	if (!bn_x931_derive_pi(p2, Xp2, ctx, cb))
-		goto err;
-
-	if (!BN_mul(p1p2, p1, p2, ctx))
-		goto err;
-
-	/* First set p to value of Rp */
-
-	if (!BN_mod_inverse(p, p2, p1, ctx))
-		goto err;
-
-	if (!BN_mul(p, p, p2, ctx))
-		goto err;
-
-	if (!BN_mod_inverse(t, p1, p2, ctx))
-		goto err;
-
-	if (!BN_mul(t, t, p1, ctx))
-		goto err;
-
-	if (!BN_sub(p, p, t))
-		goto err;
-
-	if (p->neg && !BN_add(p, p, p1p2))
-		goto err;
-
-	/* p now equals Rp */
-
-	if (!BN_mod_sub(p, p, Xp, p1p2, ctx))
-		goto err;
-
-	if (!BN_add(p, p, Xp))
-		goto err;
-
-	/* p now equals Yp0 */
-
-	for (;;)
-		{
-		int i = 1;
-		BN_GENCB_call(cb, 0, i++);
-		if (!BN_copy(pm1, p))
-			goto err;
-		if (!BN_sub_word(pm1, 1))
-			goto err;
-		if (!BN_gcd(t, pm1, e, ctx))
-			goto err;
-		if (BN_is_one(t)
-		/* X9.31 specifies 8 MR and 1 Lucas test or any prime test
-		 * offering similar or better guarantees 50 MR is considerably 
-		 * better.
-		 */
-			&& BN_is_prime_fasttest_ex(p, 50, ctx, 1, cb))
-			break;
-		if (!BN_add(p, p, p1p2))
-			goto err;
-		}
-
-	BN_GENCB_call(cb, 3, 0);
-
-	ret = 1;
-
-	err:
-
-	BN_CTX_end(ctx);
-
-	return ret;
-	}
-
-/* Generate pair of paramters Xp, Xq for X9.31 prime generation.
- * Note: nbits paramter is sum of number of bits in both.
- */
-
-int BN_X931_generate_Xpq(BIGNUM *Xp, BIGNUM *Xq, int nbits, BN_CTX *ctx)
-	{
-	BIGNUM *t;
-	int i;
-	/* Number of bits for each prime is of the form
-	 * 512+128s for s = 0, 1, ...
-	 */
-	if ((nbits < 1024) || (nbits & 0xff))
-		return 0;
-	nbits >>= 1;
-	/* The random value Xp must be between sqrt(2) * 2^(nbits-1) and
-	 * 2^nbits - 1. By setting the top two bits we ensure that the lower
-	 * bound is exceeded.
-	 */
-	if (!BN_rand(Xp, nbits, 1, 0))
-		return 0;
-
-	BN_CTX_start(ctx);
-	t = BN_CTX_get(ctx);
-
-	for (i = 0; i < 1000; i++)
-		{
-		if (!BN_rand(Xq, nbits, 1, 0))
-			return 0;
-		/* Check that |Xp - Xq| > 2^(nbits - 100) */
-		BN_sub(t, Xp, Xq);
-		if (BN_num_bits(t) > (nbits - 100))
-			break;
-		}
-
-	BN_CTX_end(ctx);
-
-	if (i < 1000)
-		return 1;
-
-	return 0;
-
-	}
-
-/* Generate primes using X9.31 algorithm. Of the values p, p1, p2, Xp1
- * and Xp2 only 'p' needs to be non-NULL. If any of the others are not NULL
- * the relevant parameter will be stored in it.
- *
- * Due to the fact that |Xp - Xq| > 2^(nbits - 100) must be satisfied Xp and Xq
- * are generated using the previous function and supplied as input.
- */
-
-int BN_X931_generate_prime_ex(BIGNUM *p, BIGNUM *p1, BIGNUM *p2,
-			BIGNUM *Xp1, BIGNUM *Xp2,
-			const BIGNUM *Xp,
-			const BIGNUM *e, BN_CTX *ctx,
-			BN_GENCB *cb)
-	{
-	int ret = 0;
-
-	BN_CTX_start(ctx);
-	if (!Xp1)
-		Xp1 = BN_CTX_get(ctx);
-	if (!Xp2)
-		Xp2 = BN_CTX_get(ctx);
-
-	if (!BN_rand(Xp1, 101, 0, 0))
-		goto error;
-	if (!BN_rand(Xp2, 101, 0, 0))
-		goto error;
-	if (!BN_X931_derive_prime_ex(p, p1, p2, Xp, Xp1, Xp2, e, ctx, cb))
-		goto error;
-
-	ret = 1;
-
-	error:
-	BN_CTX_end(ctx);
-
-	return ret;
-
-	}
-
diff --git a/openssl/crypto/bn/bntest.c b/openssl/crypto/bn/bntest.c
index cf190380f..0cd99c5b4 100644
--- a/openssl/crypto/bn/bntest.c
+++ b/openssl/crypto/bn/bntest.c
@@ -486,7 +486,7 @@ static void print_word(BIO *bp,BN_ULONG w)
 		return;
 		}
 #endif
-	BIO_printf(bp,"%lX",w);
+	BIO_printf(bp,BN_HEX_FMT1,w);
 	}
 
 int test_div_word(BIO *bp)
@@ -732,6 +732,8 @@ int test_mont(BIO *bp, BN_CTX *ctx)
 	BN_init(&n);
 
 	mont=BN_MONT_CTX_new();
+	if (mont == NULL)
+		return 0;
 
 	BN_bntest_rand(&a,100,0,0); /**/
 	BN_bntest_rand(&b,100,0,0); /**/
@@ -1027,7 +1029,7 @@ int test_exp(BIO *bp, BN_CTX *ctx)
 		BN_bntest_rand(a,20+i*5,0,0); /**/
 		BN_bntest_rand(b,2+i,0,0); /**/
 
-		if (!BN_exp(d,a,b,ctx))
+		if (BN_exp(d,a,b,ctx) <= 0)
 			return(0);
 
 		if (bp != NULL)
@@ -1116,8 +1118,8 @@ int test_gf2m_mod(BIO *bp)
 	{
 	BIGNUM *a,*b[2],*c,*d,*e;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1174,8 +1176,8 @@ int test_gf2m_mod_mul(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d,*e,*f,*g,*h;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1245,8 +1247,8 @@ int test_gf2m_mod_sqr(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1304,8 +1306,8 @@ int test_gf2m_mod_inv(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1359,8 +1361,8 @@ int test_gf2m_mod_div(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d,*e,*f;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1422,8 +1424,8 @@ int test_gf2m_mod_exp(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d,*e,*f;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1493,8 +1495,8 @@ int test_gf2m_mod_sqrt(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d,*e,*f;
 	int i, j, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
@@ -1552,8 +1554,8 @@ int test_gf2m_mod_solve_quad(BIO *bp,BN_CTX *ctx)
 	{
 	BIGNUM *a,*b[2],*c,*d,*e;
 	int i, j, s = 0, t, ret = 0;
-	unsigned int p0[] = {163,7,6,3,0};
-	unsigned int p1[] = {193,15,0};
+	int p0[] = {163,7,6,3,0,-1};
+	int p1[] = {193,15,0,-1};
 
 	a=BN_new();
 	b[0]=BN_new();
diff --git a/openssl/crypto/bn/exptest.c b/openssl/crypto/bn/exptest.c
index f598a07cf..074a8e882 100644
--- a/openssl/crypto/bn/exptest.c
+++ b/openssl/crypto/bn/exptest.c
@@ -163,7 +163,7 @@ int main(int argc, char *argv[])
 		  	{
 			if (BN_cmp(r_simple,r_mont) != 0)
 				printf("\nsimple and mont results differ\n");
-			if (BN_cmp(r_simple,r_mont) != 0)
+			if (BN_cmp(r_simple,r_mont_const) != 0)
 				printf("\nsimple and mont const time results differ\n");
 			if (BN_cmp(r_simple,r_recp) != 0)
 				printf("\nsimple and recp results differ\n");
@@ -187,7 +187,7 @@ int main(int argc, char *argv[])
 	BN_free(b);
 	BN_free(m);
 	BN_CTX_free(ctx);
-	ERR_remove_state(0);
+	ERR_remove_thread_state(NULL);
 	CRYPTO_mem_leaks(out);
 	BIO_free(out);
 	printf(" done\n");