From: Richard Levitte <levitte@openssl.org>
Date: Sun, 17 Sep 2000 20:37:33 +0000 (+0000)
Subject: The rest of the merge in of the main trunk, a few conflicts resolved.
X-Git-Tag: OpenSSL-engine-0_9_6-beta2
X-Git-Url: https://git.librecmc.org/?a=commitdiff_plain;h=ce6990739eaa9f8a4a232cdb2fa008ce08480fec;p=oweals%2Fopenssl.git

The rest of the merge in of the main trunk, a few conflicts resolved.
---

diff --git a/Configure b/Configure
index afe66c440e..0c92260a06 100755
--- a/Configure
+++ b/Configure
@@ -221,12 +221,21 @@ my %table=(
 #   crypto/sha/sha_lcl.h.
 #					<appro@fy.chalmers.se>
 #
-"hpux-parisc-cc","cc:-Ae +O3 +ESlit -z -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY:::-ldld:BN_LLONG DES_PTR DES_UNROLL DES_RISC1::::::::::dl",
+#!#"hpux-parisc-cc","cc:-Ae +O3 +ESlit -z -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY:::-ldld:BN_LLONG DES_PTR DES_UNROLL DES_RISC1::::::::::dl",
 # Since there is mention of this in shlib/hpux10-cc.sh
 "hpux-parisc-cc-o4","cc:-Ae +O4 +ESlit -z -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY:::-ldld:BN_LLONG DES_PTR DES_UNROLL DES_RISC1::::::::::dl",
 "hpux-parisc-gcc","gcc:-O3 -DB_ENDIAN -DBN_DIV2W:::-ldld:BN_LLONG DES_PTR DES_UNROLL DES_RISC1::::::::::dl",
 "hpux64-parisc-cc","cc:-Ae +DD64 +O3 +ESlit -z -DB_ENDIAN -DMD32_XARRAY::-D_REENTRANT:-ldld:SIXTY_FOUR_BIT_LONG MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT::::::::::dl",
 
+# More attempts at unified 10.X and 11.X targets for HP C compiler.
+#
+# Chris Ruemmler <ruemmler@cup.hp.com>
+# Kevin Steves <ks@hp.se>
+"hpux-parisc-cc","cc:+O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY::-D_REENTRANT:-ldl:MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT::::::::::dl",
+"hpux-parisc2-cc","cc:+DA2.0 +DS2.0 +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DMD32_XARRAY::-D_REENTRANT:-ldl:SIXTY_FOUR_BIT MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT:asm/pa-risc2.o:::::::::dl",
+"hpux64-parisc2-cc","cc:+DD64 +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DMD32_XARRAY::-D_REENTRANT:-ldl:SIXTY_FOUR_BIT_LONG MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT:asm/pa-risc2W.o:::::::::dl",
+"hpux-parisc1_1-cc","cc:+DA1.1 +DS1.1 +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DMD32_XARRAY::-D_REENTRANT:-ldl:MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT::::::::::dl",
+
 # HPUX 9.X config.
 # Don't use the bundled cc.  It is broken.  Use HP ANSI C if possible, or
 # egcs.  gcc 2.8.1 is also broken.
diff --git a/README b/README
index 739dd0d62f..46252db971 100644
--- a/README
+++ b/README
@@ -1,5 +1,5 @@
 
- OpenSSL 0.9.6-beta1 [engine] 11 Sep 2000
+ OpenSSL 0.9.6-beta2 [engine] 17 Sep 2000
 
  Copyright (c) 1998-2000 The OpenSSL Project
  Copyright (c) 1995-1998 Eric A. Young, Tim J. Hudson
diff --git a/TABLE b/TABLE
index eb561a69b0..d2d06f3b12 100644
--- a/TABLE
+++ b/TABLE
@@ -1162,11 +1162,11 @@ $shared_cflag =
 
 *** hpux-parisc-cc
 $cc           = cc
-$cflags       = -Ae +O3 +ESlit -z -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY
+$cflags       = +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DBN_DIV2W -DMD32_XARRAY
 $unistd       = 
-$thread_cflag = 
-$lflags       = -ldld
-$bn_ops       = BN_LLONG DES_PTR DES_UNROLL DES_RISC1
+$thread_cflag = -D_REENTRANT
+$lflags       = -ldl
+$bn_ops       = MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT
 $bn_obj       = 
 $des_obj      = 
 $bf_obj       = 
@@ -1220,6 +1220,46 @@ $dso_scheme   = dl
 $shared_target= 
 $shared_cflag = 
 
+*** hpux-parisc1_1-cc
+$cc           = cc
+$cflags       = +DA1.1 +DS1.1 +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DMD32_XARRAY
+$unistd       = 
+$thread_cflag = -D_REENTRANT
+$lflags       = -ldl
+$bn_ops       = MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT
+$bn_obj       = 
+$des_obj      = 
+$bf_obj       = 
+$md5_obj      = 
+$sha1_obj     = 
+$cast_obj     = 
+$rc4_obj      = 
+$rmd160_obj   = 
+$rc5_obj      = 
+$dso_scheme   = dl
+$shared_target= 
+$shared_cflag = 
+
+*** hpux-parisc2-cc
+$cc           = cc
+$cflags       = +DA2.0 +DS2.0 +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DMD32_XARRAY
+$unistd       = 
+$thread_cflag = -D_REENTRANT
+$lflags       = -ldl
+$bn_ops       = SIXTY_FOUR_BIT MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT
+$bn_obj       = asm/pa-risc2.o
+$des_obj      = 
+$bf_obj       = 
+$md5_obj      = 
+$sha1_obj     = 
+$cast_obj     = 
+$rc4_obj      = 
+$rmd160_obj   = 
+$rc5_obj      = 
+$dso_scheme   = dl
+$shared_target= 
+$shared_cflag = 
+
 *** hpux10-brokencc
 $cc           = cc
 $cflags       = -DB_ENDIAN -DBN_DIV2W -Ae +ESlit +O2 -z
@@ -1320,6 +1360,26 @@ $dso_scheme   = dl
 $shared_target= 
 $shared_cflag = 
 
+*** hpux64-parisc2-cc
+$cc           = cc
+$cflags       = +DD64 +O3 +Optrs_strongly_typed +Olibcalls -Ae +ESlit -DB_ENDIAN -DMD32_XARRAY
+$unistd       = 
+$thread_cflag = -D_REENTRANT
+$lflags       = -ldl
+$bn_ops       = SIXTY_FOUR_BIT_LONG MD2_CHAR RC4_INDEX RC4_CHAR DES_UNROLL DES_RISC1 DES_INT
+$bn_obj       = asm/pa-risc2W.o
+$des_obj      = 
+$bf_obj       = 
+$md5_obj      = 
+$sha1_obj     = 
+$cast_obj     = 
+$rc4_obj      = 
+$rmd160_obj   = 
+$rc5_obj      = 
+$dso_scheme   = dl
+$shared_target= 
+$shared_cflag = 
+
 *** irix-cc
 $cc           = cc
 $cflags       = -O2 -use_readonly_const -DTERMIOS -DB_ENDIAN
diff --git a/apps/apps.c b/apps/apps.c
index 50ea09baca..167c319ebe 100644
--- a/apps/apps.c
+++ b/apps/apps.c
@@ -662,7 +662,7 @@ STACK_OF(X509) *load_certs(BIO *err, char *file, int format)
 
 	if      (format == FORMAT_PEM)
 		{
-		othercerts = sk_X509_new(NULL);
+		othercerts = sk_X509_new_null();
 		if(!othercerts)
 			{
 			sk_X509_free(othercerts);
diff --git a/apps/crl2p7.c b/apps/crl2p7.c
index ebf5fe90ec..d02862710d 100644
--- a/apps/crl2p7.c
+++ b/apps/crl2p7.c
@@ -141,7 +141,7 @@ int MAIN(int argc, char **argv)
 		else if (strcmp(*argv,"-certfile") == 0)
 			{
 			if (--argc < 1) goto bad;
-			if(!certflst) certflst = sk_new(NULL);
+			if(!certflst) certflst = sk_new_null();
 			sk_push(certflst,*(++argv));
 			}
 		else
@@ -215,7 +215,7 @@ bad:
 	p7s->contents->type=OBJ_nid2obj(NID_pkcs7_data);
 
 	if (!ASN1_INTEGER_set(p7s->version,1)) goto end;
-	if ((crl_stack=sk_X509_CRL_new(NULL)) == NULL) goto end;
+	if ((crl_stack=sk_X509_CRL_new_null()) == NULL) goto end;
 	p7s->crl=crl_stack;
 	if (crl != NULL)
 		{
@@ -223,7 +223,7 @@ bad:
 		crl=NULL; /* now part of p7 for OPENSSL_freeing */
 		}
 
-	if ((cert_stack=sk_X509_new(NULL)) == NULL) goto end;
+	if ((cert_stack=sk_X509_new_null()) == NULL) goto end;
 	p7s->cert=cert_stack;
 
 	if(certflst) for(i = 0; i < sk_num(certflst); i++) {
diff --git a/apps/dgst.c b/apps/dgst.c
index f0e5f954da..dbfa387f33 100644
--- a/apps/dgst.c
+++ b/apps/dgst.c
@@ -307,7 +307,7 @@ int MAIN(int argc, char **argv)
 		}
 		siglen = BIO_read(sigbio, sigbuf, siglen);
 		BIO_free(sigbio);
-		if(siglen <= 0) {
+		if(siglen == 0) {
 			BIO_printf(bio_err, "Error reading signature file %s\n",
 								sigfile);
 			ERR_print_errors(bio_err);
diff --git a/apps/nseq.c b/apps/nseq.c
index cc88d50ceb..7210fbdb5e 100644
--- a/apps/nseq.c
+++ b/apps/nseq.c
@@ -123,7 +123,7 @@ int MAIN(int argc, char **argv)
 
 	if (toseq) {
 		seq = NETSCAPE_CERT_SEQUENCE_new();
-		seq->certs = sk_X509_new(NULL);
+		seq->certs = sk_X509_new_null();
 		while((x509 = PEM_read_bio_X509(in, NULL, NULL, NULL))) 
 		    sk_X509_push(seq->certs,x509);
 
diff --git a/apps/pkcs12.c b/apps/pkcs12.c
index b57cf5f829..2cc9a0caee 100644
--- a/apps/pkcs12.c
+++ b/apps/pkcs12.c
@@ -200,7 +200,7 @@ int MAIN(int argc, char **argv)
 		} else if (!strcmp (*args, "-caname")) {
 		    if (args[1]) {
 			args++;	
-			if (!canames) canames = sk_new(NULL);
+			if (!canames) canames = sk_new_null();
 			sk_push(canames, *args);
 		    } else badarg = 1;
 		} else if (!strcmp (*args, "-in")) {
@@ -427,7 +427,7 @@ int MAIN(int argc, char **argv)
 	CRYPTO_push_info("reading certs from input");
 #endif
 
-	certs = sk_X509_new(NULL);
+	certs = sk_X509_new_null();
 
 	/* Load in all certs in input file */
 	if(!cert_load(in, certs)) {
@@ -459,7 +459,7 @@ int MAIN(int argc, char **argv)
 	CRYPTO_push_info("reading certs from certfile");
 #endif
 
-	bags = sk_PKCS12_SAFEBAG_new (NULL);
+	bags = sk_PKCS12_SAFEBAG_new_null ();
 
 	/* Add any more certificates asked for */
 	if (certsin) {
@@ -550,7 +550,7 @@ int MAIN(int argc, char **argv)
 		goto export_end;
 	}
 
-	safes = sk_PKCS7_new (NULL);
+	safes = sk_PKCS7_new_null ();
 	sk_PKCS7_push (safes, authsafe);
 
 #ifdef CRYPTO_MDEBUG
@@ -566,7 +566,7 @@ int MAIN(int argc, char **argv)
 	p8 = NULL;
         if (name) PKCS12_add_friendlyname (bag, name, -1);
 	PKCS12_add_localkeyid (bag, keyid, keyidlen);
-	bags = sk_PKCS12_SAFEBAG_new(NULL);
+	bags = sk_PKCS12_SAFEBAG_new_null();
 	sk_PKCS12_SAFEBAG_push (bags, bag);
 
 #ifdef CRYPTO_MDEBUG
diff --git a/crypto/asn1/a_set.c b/crypto/asn1/a_set.c
index 1921f5eaa1..caf5a1419c 100644
--- a/crypto/asn1/a_set.c
+++ b/crypto/asn1/a_set.c
@@ -158,7 +158,7 @@ STACK *d2i_ASN1_SET(STACK **a, unsigned char **pp, long length,
 	STACK *ret=NULL;
 
 	if ((a == NULL) || ((*a) == NULL))
-		{ if ((ret=sk_new(NULL)) == NULL) goto err; }
+		{ if ((ret=sk_new_null()) == NULL) goto err; }
 	else
 		ret=(*a);
 
diff --git a/crypto/bn/asm/README b/crypto/bn/asm/README
index 86bf64cfc2..a0fe58a677 100644
--- a/crypto/bn/asm/README
+++ b/crypto/bn/asm/README
@@ -15,9 +15,9 @@ On the 2 alpha C compilers I had access to, it was not possible to do
 were 64 bits).  So the hand assember gives access to the 128 bit result and
 a 2 times speedup :-).
 
-There are 2 versions of assember for the HP PA-RISC.
-pa-risc.s is the origional one which works fine.
-pa-risc2.s is a new version that often generates warnings but if the
-tests pass, it gives performance that is over 2 times faster than
-pa-risc.s.
-Both were generated using gcc :-)
+There are 3 versions of assember for the HP PA-RISC.
+
+pa-risc.s is the origional one which works fine and generated using gcc :-)
+
+pa-risc2W.s and pa-risc2.s are 64 and 32-bit PA-RISC 2.0 implementations
+by Chris Ruemmler from HP (with some help from the HP C compiler).
diff --git a/crypto/bn/asm/pa-risc2.s b/crypto/bn/asm/pa-risc2.s
index c2725996a4..7239aa2c76 100644
--- a/crypto/bn/asm/pa-risc2.s
+++ b/crypto/bn/asm/pa-risc2.s
@@ -1,416 +1,1618 @@
-	.SPACE $PRIVATE$
-	.SUBSPA $DATA$,QUAD=1,ALIGN=8,ACCESS=31
-	.SUBSPA $BSS$,QUAD=1,ALIGN=8,ACCESS=31,ZERO,SORT=82
-	.SPACE $TEXT$
-	.SUBSPA $LIT$,QUAD=0,ALIGN=8,ACCESS=44
-	.SUBSPA $CODE$,QUAD=0,ALIGN=8,ACCESS=44,CODE_ONLY
-	.IMPORT $global$,DATA
-	.IMPORT $$dyncall,MILLICODE
-; gcc_compiled.:
-	.SPACE $TEXT$
-	.SUBSPA $CODE$
-
-	.align 4
-	.EXPORT bn_mul_add_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR
+;
+; PA-RISC 2.0 implementation of bn_asm code, based on the
+; 64-bit version of the code.  This code is effectively the
+; same as the 64-bit version except the register model is
+; slightly different given all values must be 32-bit between
+; function calls.  Thus the 64-bit return values are returned
+; in %ret0 and %ret1 vs just %ret0 as is done in 64-bit
+;
+;
+; This code is approximately 2x faster than the C version
+; for RSA/DSA.
+;
+; See http://devresource.hp.com/  for more details on the PA-RISC
+; architecture.  Also see the book "PA-RISC 2.0 Architecture"
+; by Gerry Kane for information on the instruction set architecture.
+;
+; Code written by Chris Ruemmler (with some help from the HP C
+; compiler).
+;
+; The code compiles with HP's assembler
+;
+
+	.level	2.0N
+	.space	$TEXT$
+	.subspa	$CODE$,QUAD=0,ALIGN=8,ACCESS=0x2c,CODE_ONLY
+
+;
+; Global Register definitions used for the routines.
+;
+; Some information about HP's runtime architecture for 32-bits.
+;
+; "Caller save" means the calling function must save the register
+; if it wants the register to be preserved.
+; "Callee save" means if a function uses the register, it must save
+; the value before using it.
+;
+; For the floating point registers 
+;
+;    "caller save" registers: fr4-fr11, fr22-fr31
+;    "callee save" registers: fr12-fr21
+;    "special" registers: fr0-fr3 (status and exception registers)
+;
+; For the integer registers
+;     value zero             :  r0
+;     "caller save" registers: r1,r19-r26
+;     "callee save" registers: r3-r18
+;     return register        :  r2  (rp)
+;     return values          ; r28,r29  (ret0,ret1)
+;     Stack pointer          ; r30  (sp) 
+;     millicode return ptr   ; r31  (also a caller save register)
+
+
+;
+; Arguments to the routines
+;
+r_ptr       .reg %r26
+a_ptr       .reg %r25
+b_ptr       .reg %r24
+num         .reg %r24
+n           .reg %r23
+
+;
+; Note that the "w" argument for bn_mul_add_words and bn_mul_words
+; is passed on the stack at a delta of -56 from the top of stack
+; as the routine is entered.
+;
+
+;
+; Globals used in some routines
+;
+
+top_overflow .reg %r23
+high_mask    .reg %r22    ; value 0xffffffff80000000L
+
+
+;------------------------------------------------------------------------------
+;
+; bn_mul_add_words
+;
+;BN_ULONG bn_mul_add_words(BN_ULONG *r_ptr, BN_ULONG *a_ptr, 
+;								int num, BN_ULONG w)
+;
+; arg0 = r_ptr
+; arg1 = a_ptr
+; arg3 = num
+; -56(sp) =  w
+;
+; Local register definitions
+;
+
+fm1          .reg %fr22
+fm           .reg %fr23
+ht_temp      .reg %fr24
+ht_temp_1    .reg %fr25
+lt_temp      .reg %fr26
+lt_temp_1    .reg %fr27
+fm1_1        .reg %fr28
+fm_1         .reg %fr29
+
+fw_h         .reg %fr7L
+fw_l         .reg %fr7R
+fw           .reg %fr7
+
+fht_0        .reg %fr8L
+flt_0        .reg %fr8R
+t_float_0    .reg %fr8
+
+fht_1        .reg %fr9L
+flt_1        .reg %fr9R
+t_float_1    .reg %fr9
+
+tmp_0        .reg %r31
+tmp_1        .reg %r21
+m_0          .reg %r20 
+m_1          .reg %r19 
+ht_0         .reg %r1  
+ht_1         .reg %r3
+lt_0         .reg %r4
+lt_1         .reg %r5
+m1_0         .reg %r6 
+m1_1         .reg %r7 
+rp_val       .reg %r8
+rp_val_1     .reg %r9
+
 bn_mul_add_words
-	.PROC
-	.CALLINFO FRAME=64,CALLS,SAVE_RP,ENTRY_GR=4
-	.ENTRY
-	stw %r2,-20(0,%r30)
-	stwm %r4,64(0,%r30)
-	copy %r24,%r31
-	stw %r3,-60(0,%r30)
-	ldi 0,%r20
-	ldo 12(%r26),%r2
-	stw %r23,-16(0,%r30)
-	copy %r25,%r3
-	ldo 12(%r3),%r1
-	fldws -16(0,%r30),%fr8L
-L$0010
-	copy %r20,%r25
-	ldi 0,%r24
-	fldws 0(0,%r3),%fr9L
-	ldw 0(0,%r26),%r19
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r19,%r23
-	ldw -16(0,%r30),%r28
-	ldw -12(0,%r30),%r29
-	ldi 0,%r22
-	add %r23,%r29,%r29
-	addc %r22,%r28,%r28
-	add %r25,%r29,%r29
-	addc %r24,%r28,%r28
-	copy %r28,%r21
-	ldi 0,%r20
-	copy %r21,%r20
-	addib,= -1,%r31,L$0011
-	stw %r29,0(0,%r26)
-	copy %r20,%r25
-	ldi 0,%r24
-	fldws -8(0,%r1),%fr9L
-	ldw -8(0,%r2),%r19
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r19,%r23
-	ldw -16(0,%r30),%r28
-	ldw -12(0,%r30),%r29
-	ldi 0,%r22
-	add %r23,%r29,%r29
-	addc %r22,%r28,%r28
-	add %r25,%r29,%r29
-	addc %r24,%r28,%r28
-	copy %r28,%r21
-	ldi 0,%r20
-	copy %r21,%r20
-	addib,= -1,%r31,L$0011
-	stw %r29,-8(0,%r2)
-	copy %r20,%r25
-	ldi 0,%r24
-	fldws -4(0,%r1),%fr9L
-	ldw -4(0,%r2),%r19
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r19,%r23
-	ldw -16(0,%r30),%r28
-	ldw -12(0,%r30),%r29
-	ldi 0,%r22
-	add %r23,%r29,%r29
-	addc %r22,%r28,%r28
-	add %r25,%r29,%r29
-	addc %r24,%r28,%r28
-	copy %r28,%r21
-	ldi 0,%r20
-	copy %r21,%r20
-	addib,= -1,%r31,L$0011
-	stw %r29,-4(0,%r2)
-	copy %r20,%r25
-	ldi 0,%r24
-	fldws 0(0,%r1),%fr9L
-	ldw 0(0,%r2),%r19
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r19,%r23
-	ldw -16(0,%r30),%r28
-	ldw -12(0,%r30),%r29
-	ldi 0,%r22
-	add %r23,%r29,%r29
-	addc %r22,%r28,%r28
-	add %r25,%r29,%r29
-	addc %r24,%r28,%r28
-	copy %r28,%r21
-	ldi 0,%r20
-	copy %r21,%r20
-	addib,= -1,%r31,L$0011
-	stw %r29,0(0,%r2)
-	ldo 16(%r1),%r1
-	ldo 16(%r3),%r3
-	ldo 16(%r2),%r2
-	bl L$0010,0
-	ldo 16(%r26),%r26
-L$0011
-	copy %r20,%r28
-	ldw -84(0,%r30),%r2
-	ldw -60(0,%r30),%r3
-	bv 0(%r2)
-	ldwm -64(0,%r30),%r4
-	.EXIT
-	.PROCEND
-	.align 4
-	.EXPORT bn_mul_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR
+	.export	bn_mul_add_words,entry,NO_RELOCATION,LONG_RETURN
+	.proc
+	.callinfo frame=128
+    .entry
+	.align 64
+
+    STD     %r3,0(%sp)          ; save r3  
+    STD     %r4,8(%sp)          ; save r4  
+	NOP                         ; Needed to make the loop 16-byte aligned
+	NOP                         ; needed to make the loop 16-byte aligned
+
+    STD     %r5,16(%sp)         ; save r5  
+	NOP
+    STD     %r6,24(%sp)         ; save r6  
+    STD     %r7,32(%sp)         ; save r7  
+
+    STD     %r8,40(%sp)         ; save r8  
+    STD     %r9,48(%sp)         ; save r9  
+    COPY    %r0,%ret1           ; return 0 by default
+    DEPDI,Z 1,31,1,top_overflow ; top_overflow = 1 << 32    
+
+    CMPIB,>= 0,num,bn_mul_add_words_exit  ; if (num <= 0) then exit
+	LDO     128(%sp),%sp        ; bump stack
+
+	;
+	; The loop is unrolled twice, so if there is only 1 number
+    ; then go straight to the cleanup code.
+	;
+	CMPIB,= 1,num,bn_mul_add_words_single_top
+	FLDD    -184(%sp),fw        ; (-56-128) load up w into fw (fw_h/fw_l)
+
+	;
+	; This loop is unrolled 2 times (64-byte aligned as well)
+	;
+	; PA-RISC 2.0 chips have two fully pipelined multipliers, thus
+    ; two 32-bit mutiplies can be issued per cycle.
+    ; 
+bn_mul_add_words_unroll2
+
+    FLDD    0(a_ptr),t_float_0       ; load up 64-bit value (fr8L) ht(L)/lt(R)
+    FLDD    8(a_ptr),t_float_1       ; load up 64-bit value (fr8L) ht(L)/lt(R)
+    LDD     0(r_ptr),rp_val          ; rp[0]
+    LDD     8(r_ptr),rp_val_1        ; rp[1]
+
+    XMPYU   fht_0,fw_l,fm1           ; m1[0] = fht_0*fw_l
+    XMPYU   fht_1,fw_l,fm1_1         ; m1[1] = fht_1*fw_l
+    FSTD    fm1,-16(%sp)             ; -16(sp) = m1[0]
+    FSTD    fm1_1,-48(%sp)           ; -48(sp) = m1[1]
+
+    XMPYU   flt_0,fw_h,fm            ; m[0] = flt_0*fw_h
+    XMPYU   flt_1,fw_h,fm_1          ; m[1] = flt_1*fw_h
+    FSTD    fm,-8(%sp)               ; -8(sp) = m[0]
+    FSTD    fm_1,-40(%sp)            ; -40(sp) = m[1]
+
+    XMPYU   fht_0,fw_h,ht_temp       ; ht_temp   = fht_0*fw_h
+    XMPYU   fht_1,fw_h,ht_temp_1     ; ht_temp_1 = fht_1*fw_h
+    FSTD    ht_temp,-24(%sp)         ; -24(sp)   = ht_temp
+    FSTD    ht_temp_1,-56(%sp)       ; -56(sp)   = ht_temp_1
+
+    XMPYU   flt_0,fw_l,lt_temp       ; lt_temp = lt*fw_l
+    XMPYU   flt_1,fw_l,lt_temp_1     ; lt_temp = lt*fw_l
+    FSTD    lt_temp,-32(%sp)         ; -32(sp) = lt_temp 
+    FSTD    lt_temp_1,-64(%sp)       ; -64(sp) = lt_temp_1 
+
+    LDD     -8(%sp),m_0              ; m[0] 
+    LDD     -40(%sp),m_1             ; m[1]
+    LDD     -16(%sp),m1_0            ; m1[0]
+    LDD     -48(%sp),m1_1            ; m1[1]
+
+    LDD     -24(%sp),ht_0            ; ht[0]
+    LDD     -56(%sp),ht_1            ; ht[1]
+    ADD,L   m1_0,m_0,tmp_0           ; tmp_0 = m[0] + m1[0]; 
+    ADD,L   m1_1,m_1,tmp_1           ; tmp_1 = m[1] + m1[1]; 
+
+    LDD     -32(%sp),lt_0            
+    LDD     -64(%sp),lt_1            
+    CMPCLR,*>>= tmp_0,m1_0, %r0      ; if (m[0] < m1[0])
+    ADD,L   ht_0,top_overflow,ht_0   ; ht[0] += (1<<32)
+
+    CMPCLR,*>>= tmp_1,m1_1,%r0       ; if (m[1] < m1[1])
+    ADD,L   ht_1,top_overflow,ht_1   ; ht[1] += (1<<32)
+    EXTRD,U tmp_0,31,32,m_0          ; m[0]>>32  
+    DEPD,Z  tmp_0,31,32,m1_0         ; m1[0] = m[0]<<32 
+
+    EXTRD,U tmp_1,31,32,m_1          ; m[1]>>32  
+    DEPD,Z  tmp_1,31,32,m1_1         ; m1[1] = m[1]<<32 
+    ADD,L   ht_0,m_0,ht_0            ; ht[0]+= (m[0]>>32)
+    ADD,L   ht_1,m_1,ht_1            ; ht[1]+= (m[1]>>32)
+
+    ADD     lt_0,m1_0,lt_0           ; lt[0] = lt[0]+m1[0];
+	ADD,DC  ht_0,%r0,ht_0            ; ht[0]++
+    ADD     lt_1,m1_1,lt_1           ; lt[1] = lt[1]+m1[1];
+    ADD,DC  ht_1,%r0,ht_1            ; ht[1]++
+
+    ADD    %ret1,lt_0,lt_0           ; lt[0] = lt[0] + c;
+	ADD,DC  ht_0,%r0,ht_0            ; ht[0]++
+    ADD     lt_0,rp_val,lt_0         ; lt[0] = lt[0]+rp[0]
+    ADD,DC  ht_0,%r0,ht_0            ; ht[0]++
+
+	LDO    -2(num),num               ; num = num - 2;
+    ADD     ht_0,lt_1,lt_1           ; lt[1] = lt[1] + ht_0 (c);
+    ADD,DC  ht_1,%r0,ht_1            ; ht[1]++
+    STD     lt_0,0(r_ptr)            ; rp[0] = lt[0]
+
+    ADD     lt_1,rp_val_1,lt_1       ; lt[1] = lt[1]+rp[1]
+    ADD,DC  ht_1,%r0,%ret1           ; ht[1]++
+    LDO     16(a_ptr),a_ptr          ; a_ptr += 2
+
+    STD     lt_1,8(r_ptr)            ; rp[1] = lt[1]
+	CMPIB,<= 2,num,bn_mul_add_words_unroll2 ; go again if more to do
+    LDO     16(r_ptr),r_ptr          ; r_ptr += 2
+
+    CMPIB,=,N 0,num,bn_mul_add_words_exit ; are we done, or cleanup last one
+
+	;
+	; Top of loop aligned on 64-byte boundary
+	;
+bn_mul_add_words_single_top
+    FLDD    0(a_ptr),t_float_0        ; load up 64-bit value (fr8L) ht(L)/lt(R)
+    LDD     0(r_ptr),rp_val           ; rp[0]
+    LDO     8(a_ptr),a_ptr            ; a_ptr++
+    XMPYU   fht_0,fw_l,fm1            ; m1 = ht*fw_l
+    FSTD    fm1,-16(%sp)              ; -16(sp) = m1
+    XMPYU   flt_0,fw_h,fm             ; m = lt*fw_h
+    FSTD    fm,-8(%sp)                ; -8(sp) = m
+    XMPYU   fht_0,fw_h,ht_temp        ; ht_temp = ht*fw_h
+    FSTD    ht_temp,-24(%sp)          ; -24(sp) = ht
+    XMPYU   flt_0,fw_l,lt_temp        ; lt_temp = lt*fw_l
+    FSTD    lt_temp,-32(%sp)          ; -32(sp) = lt 
+
+    LDD     -8(%sp),m_0               
+    LDD    -16(%sp),m1_0              ; m1 = temp1 
+    ADD,L   m_0,m1_0,tmp_0            ; tmp_0 = m + m1; 
+    LDD     -24(%sp),ht_0             
+    LDD     -32(%sp),lt_0             
+
+    CMPCLR,*>>= tmp_0,m1_0,%r0        ; if (m < m1)
+    ADD,L   ht_0,top_overflow,ht_0    ; ht += (1<<32)
+
+    EXTRD,U tmp_0,31,32,m_0           ; m>>32  
+    DEPD,Z  tmp_0,31,32,m1_0          ; m1 = m<<32 
+
+    ADD,L   ht_0,m_0,ht_0             ; ht+= (m>>32)
+    ADD     lt_0,m1_0,tmp_0           ; tmp_0 = lt+m1;
+    ADD,DC  ht_0,%r0,ht_0             ; ht++
+    ADD     %ret1,tmp_0,lt_0          ; lt = lt + c;
+    ADD,DC  ht_0,%r0,ht_0             ; ht++
+    ADD     lt_0,rp_val,lt_0          ; lt = lt+rp[0]
+    ADD,DC  ht_0,%r0,%ret1            ; ht++
+    STD     lt_0,0(r_ptr)             ; rp[0] = lt
+
+bn_mul_add_words_exit
+    .EXIT
+	
+    EXTRD,U %ret1,31,32,%ret0         ; for 32-bit, return in ret0/ret1
+    LDD     -80(%sp),%r9              ; restore r9  
+    LDD     -88(%sp),%r8              ; restore r8  
+    LDD     -96(%sp),%r7              ; restore r7  
+    LDD     -104(%sp),%r6             ; restore r6  
+    LDD     -112(%sp),%r5             ; restore r5  
+    LDD     -120(%sp),%r4             ; restore r4  
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3             ; restore r3
+	.PROCEND	;in=23,24,25,26,29;out=28;
+
+;----------------------------------------------------------------------------
+;
+;BN_ULONG bn_mul_words(BN_ULONG *rp, BN_ULONG *ap, int num, BN_ULONG w)
+;
+; arg0 = rp
+; arg1 = ap
+; arg3 = num
+; w on stack at -56(sp)
+
 bn_mul_words
-	.PROC
-	.CALLINFO FRAME=64,CALLS,SAVE_RP,ENTRY_GR=3
-	.ENTRY
-	stw %r2,-20(0,%r30)
-	copy %r25,%r2
-	stwm %r4,64(0,%r30)
-	copy %r24,%r19
-	ldi 0,%r28
-	stw %r23,-16(0,%r30)
-	ldo 12(%r26),%r31
-	ldo 12(%r2),%r29
-	fldws -16(0,%r30),%fr8L
-L$0026
-	fldws 0(0,%r2),%fr9L
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r28,%r21
-	ldi 0,%r20
-	ldw -16(0,%r30),%r24
-	ldw -12(0,%r30),%r25
-	add %r21,%r25,%r25
-	addc %r20,%r24,%r24
-	copy %r24,%r23
-	ldi 0,%r22
-	copy %r23,%r28
-	addib,= -1,%r19,L$0027
-	stw %r25,0(0,%r26)
-	fldws -8(0,%r29),%fr9L
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r28,%r21
-	ldi 0,%r20
-	ldw -16(0,%r30),%r24
-	ldw -12(0,%r30),%r25
-	add %r21,%r25,%r25
-	addc %r20,%r24,%r24
-	copy %r24,%r23
-	ldi 0,%r22
-	copy %r23,%r28
-	addib,= -1,%r19,L$0027
-	stw %r25,-8(0,%r31)
-	fldws -4(0,%r29),%fr9L
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r28,%r21
-	ldi 0,%r20
-	ldw -16(0,%r30),%r24
-	ldw -12(0,%r30),%r25
-	add %r21,%r25,%r25
-	addc %r20,%r24,%r24
-	copy %r24,%r23
-	ldi 0,%r22
-	copy %r23,%r28
-	addib,= -1,%r19,L$0027
-	stw %r25,-4(0,%r31)
-	fldws 0(0,%r29),%fr9L
-	xmpyu %fr8L,%fr9L,%fr9
-	fstds %fr9,-16(0,%r30)
-	copy %r28,%r21
-	ldi 0,%r20
-	ldw -16(0,%r30),%r24
-	ldw -12(0,%r30),%r25
-	add %r21,%r25,%r25
-	addc %r20,%r24,%r24
-	copy %r24,%r23
-	ldi 0,%r22
-	copy %r23,%r28
-	addib,= -1,%r19,L$0027
-	stw %r25,0(0,%r31)
-	ldo 16(%r29),%r29
-	ldo 16(%r2),%r2
-	ldo 16(%r31),%r31
-	bl L$0026,0
-	ldo 16(%r26),%r26
-L$0027
-	ldw -84(0,%r30),%r2
-	bv 0(%r2)
-	ldwm -64(0,%r30),%r4
-	.EXIT
-	.PROCEND
-	.align 4
-	.EXPORT bn_sqr_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR
+	.proc
+	.callinfo frame=128
+    .entry
+	.EXPORT	bn_mul_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+	.align 64
+
+    STD     %r3,0(%sp)          ; save r3  
+    STD     %r4,8(%sp)          ; save r4  
+	NOP
+    STD     %r5,16(%sp)         ; save r5  
+
+    STD     %r6,24(%sp)         ; save r6  
+    STD     %r7,32(%sp)         ; save r7  
+    COPY    %r0,%ret1           ; return 0 by default
+    DEPDI,Z 1,31,1,top_overflow ; top_overflow = 1 << 32    
+
+    CMPIB,>= 0,num,bn_mul_words_exit
+	LDO     128(%sp),%sp    ; bump stack
+
+	;
+	; See if only 1 word to do, thus just do cleanup
+	;
+	CMPIB,= 1,num,bn_mul_words_single_top
+	FLDD    -184(%sp),fw        ; (-56-128) load up w into fw (fw_h/fw_l)
+
+	;
+	; This loop is unrolled 2 times (64-byte aligned as well)
+	;
+	; PA-RISC 2.0 chips have two fully pipelined multipliers, thus
+    ; two 32-bit mutiplies can be issued per cycle.
+    ; 
+bn_mul_words_unroll2
+
+    FLDD    0(a_ptr),t_float_0        ; load up 64-bit value (fr8L) ht(L)/lt(R)
+    FLDD    8(a_ptr),t_float_1        ; load up 64-bit value (fr8L) ht(L)/lt(R)
+    XMPYU   fht_0,fw_l,fm1            ; m1[0] = fht_0*fw_l
+    XMPYU   fht_1,fw_l,fm1_1          ; m1[1] = ht*fw_l
+
+    FSTD    fm1,-16(%sp)              ; -16(sp) = m1
+    FSTD    fm1_1,-48(%sp)            ; -48(sp) = m1
+    XMPYU   flt_0,fw_h,fm             ; m = lt*fw_h
+    XMPYU   flt_1,fw_h,fm_1           ; m = lt*fw_h
+
+    FSTD    fm,-8(%sp)                ; -8(sp) = m
+    FSTD    fm_1,-40(%sp)             ; -40(sp) = m
+    XMPYU   fht_0,fw_h,ht_temp        ; ht_temp = fht_0*fw_h
+    XMPYU   fht_1,fw_h,ht_temp_1      ; ht_temp = ht*fw_h
+
+    FSTD    ht_temp,-24(%sp)          ; -24(sp) = ht
+    FSTD    ht_temp_1,-56(%sp)        ; -56(sp) = ht
+    XMPYU   flt_0,fw_l,lt_temp        ; lt_temp = lt*fw_l
+    XMPYU   flt_1,fw_l,lt_temp_1      ; lt_temp = lt*fw_l
+
+    FSTD    lt_temp,-32(%sp)          ; -32(sp) = lt 
+    FSTD    lt_temp_1,-64(%sp)        ; -64(sp) = lt 
+    LDD     -8(%sp),m_0               
+    LDD     -40(%sp),m_1              
+
+    LDD    -16(%sp),m1_0              
+    LDD    -48(%sp),m1_1              
+    LDD     -24(%sp),ht_0             
+    LDD     -56(%sp),ht_1             
+
+    ADD,L   m1_0,m_0,tmp_0            ; tmp_0 = m + m1; 
+    ADD,L   m1_1,m_1,tmp_1            ; tmp_1 = m + m1; 
+    LDD     -32(%sp),lt_0             
+    LDD     -64(%sp),lt_1             
+
+    CMPCLR,*>>= tmp_0,m1_0, %r0       ; if (m < m1)
+    ADD,L   ht_0,top_overflow,ht_0    ; ht += (1<<32)
+    CMPCLR,*>>= tmp_1,m1_1,%r0        ; if (m < m1)
+    ADD,L   ht_1,top_overflow,ht_1    ; ht += (1<<32)
+
+    EXTRD,U tmp_0,31,32,m_0           ; m>>32  
+    DEPD,Z  tmp_0,31,32,m1_0          ; m1 = m<<32 
+    EXTRD,U tmp_1,31,32,m_1           ; m>>32  
+    DEPD,Z  tmp_1,31,32,m1_1          ; m1 = m<<32 
+
+    ADD,L   ht_0,m_0,ht_0             ; ht+= (m>>32)
+    ADD,L   ht_1,m_1,ht_1             ; ht+= (m>>32)
+    ADD     lt_0,m1_0,lt_0            ; lt = lt+m1;
+	ADD,DC  ht_0,%r0,ht_0             ; ht++
+
+    ADD     lt_1,m1_1,lt_1            ; lt = lt+m1;
+    ADD,DC  ht_1,%r0,ht_1             ; ht++
+    ADD    %ret1,lt_0,lt_0            ; lt = lt + c (ret1);
+	ADD,DC  ht_0,%r0,ht_0             ; ht++
+
+    ADD     ht_0,lt_1,lt_1            ; lt = lt + c (ht_0)
+    ADD,DC  ht_1,%r0,ht_1             ; ht++
+    STD     lt_0,0(r_ptr)             ; rp[0] = lt
+    STD     lt_1,8(r_ptr)             ; rp[1] = lt
+
+	COPY    ht_1,%ret1                ; carry = ht
+	LDO    -2(num),num                ; num = num - 2;
+    LDO     16(a_ptr),a_ptr           ; ap += 2
+	CMPIB,<= 2,num,bn_mul_words_unroll2
+    LDO     16(r_ptr),r_ptr           ; rp++
+
+    CMPIB,=,N 0,num,bn_mul_words_exit ; are we done?
+
+	;
+	; Top of loop aligned on 64-byte boundary
+	;
+bn_mul_words_single_top
+    FLDD    0(a_ptr),t_float_0        ; load up 64-bit value (fr8L) ht(L)/lt(R)
+
+    XMPYU   fht_0,fw_l,fm1            ; m1 = ht*fw_l
+    FSTD    fm1,-16(%sp)              ; -16(sp) = m1
+    XMPYU   flt_0,fw_h,fm             ; m = lt*fw_h
+    FSTD    fm,-8(%sp)                ; -8(sp) = m
+    XMPYU   fht_0,fw_h,ht_temp        ; ht_temp = ht*fw_h
+    FSTD    ht_temp,-24(%sp)          ; -24(sp) = ht
+    XMPYU   flt_0,fw_l,lt_temp        ; lt_temp = lt*fw_l
+    FSTD    lt_temp,-32(%sp)          ; -32(sp) = lt 
+
+    LDD     -8(%sp),m_0               
+    LDD    -16(%sp),m1_0              
+    ADD,L   m_0,m1_0,tmp_0            ; tmp_0 = m + m1; 
+    LDD     -24(%sp),ht_0             
+    LDD     -32(%sp),lt_0             
+
+    CMPCLR,*>>= tmp_0,m1_0,%r0        ; if (m < m1)
+    ADD,L   ht_0,top_overflow,ht_0    ; ht += (1<<32)
+
+    EXTRD,U tmp_0,31,32,m_0           ; m>>32  
+    DEPD,Z  tmp_0,31,32,m1_0          ; m1 = m<<32 
+
+    ADD,L   ht_0,m_0,ht_0             ; ht+= (m>>32)
+    ADD     lt_0,m1_0,lt_0            ; lt= lt+m1;
+    ADD,DC  ht_0,%r0,ht_0             ; ht++
+
+    ADD     %ret1,lt_0,lt_0           ; lt = lt + c;
+    ADD,DC  ht_0,%r0,ht_0             ; ht++
+
+    COPY    ht_0,%ret1                ; copy carry
+    STD     lt_0,0(r_ptr)             ; rp[0] = lt
+
+bn_mul_words_exit
+    .EXIT
+    EXTRD,U %ret1,31,32,%ret0           ; for 32-bit, return in ret0/ret1
+    LDD     -96(%sp),%r7              ; restore r7  
+    LDD     -104(%sp),%r6             ; restore r6  
+    LDD     -112(%sp),%r5             ; restore r5  
+    LDD     -120(%sp),%r4             ; restore r4  
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3             ; restore r3
+	.PROCEND	
+
+;----------------------------------------------------------------------------
+;
+;void bn_sqr_words(BN_ULONG *rp, BN_ULONG *ap, int num)
+;
+; arg0 = rp
+; arg1 = ap
+; arg2 = num
+;
+
 bn_sqr_words
+	.proc
+	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
+	.EXPORT	bn_sqr_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+    .entry
+	.align 64
+
+    STD     %r3,0(%sp)          ; save r3  
+    STD     %r4,8(%sp)          ; save r4  
+	NOP
+    STD     %r5,16(%sp)         ; save r5  
+
+    CMPIB,>= 0,num,bn_sqr_words_exit
+	LDO     128(%sp),%sp       ; bump stack
+
+	;
+	; If only 1, the goto straight to cleanup
+	;
+	CMPIB,= 1,num,bn_sqr_words_single_top
+    DEPDI,Z -1,32,33,high_mask   ; Create Mask 0xffffffff80000000L
+
+	;
+	; This loop is unrolled 2 times (64-byte aligned as well)
+	;
+
+bn_sqr_words_unroll2
+    FLDD    0(a_ptr),t_float_0        ; a[0]
+    FLDD    8(a_ptr),t_float_1        ; a[1]
+    XMPYU   fht_0,flt_0,fm            ; m[0]
+    XMPYU   fht_1,flt_1,fm_1          ; m[1]
+
+    FSTD    fm,-24(%sp)               ; store m[0]
+    FSTD    fm_1,-56(%sp)             ; store m[1]
+    XMPYU   flt_0,flt_0,lt_temp       ; lt[0]
+    XMPYU   flt_1,flt_1,lt_temp_1     ; lt[1]
+
+    FSTD    lt_temp,-16(%sp)          ; store lt[0]
+    FSTD    lt_temp_1,-48(%sp)        ; store lt[1]
+    XMPYU   fht_0,fht_0,ht_temp       ; ht[0]
+    XMPYU   fht_1,fht_1,ht_temp_1     ; ht[1]
+
+    FSTD    ht_temp,-8(%sp)           ; store ht[0]
+    FSTD    ht_temp_1,-40(%sp)        ; store ht[1]
+    LDD     -24(%sp),m_0             
+    LDD     -56(%sp),m_1              
+
+    AND     m_0,high_mask,tmp_0       ; m[0] & Mask
+    AND     m_1,high_mask,tmp_1       ; m[1] & Mask
+    DEPD,Z  m_0,30,31,m_0             ; m[0] << 32+1
+    DEPD,Z  m_1,30,31,m_1             ; m[1] << 32+1
+
+    LDD     -16(%sp),lt_0        
+    LDD     -48(%sp),lt_1        
+    EXTRD,U tmp_0,32,33,tmp_0         ; tmp_0 = m[0]&Mask >> 32-1
+    EXTRD,U tmp_1,32,33,tmp_1         ; tmp_1 = m[1]&Mask >> 32-1
+
+    LDD     -8(%sp),ht_0            
+    LDD     -40(%sp),ht_1           
+    ADD,L   ht_0,tmp_0,ht_0           ; ht[0] += tmp_0
+    ADD,L   ht_1,tmp_1,ht_1           ; ht[1] += tmp_1
+
+    ADD     lt_0,m_0,lt_0             ; lt = lt+m
+    ADD,DC  ht_0,%r0,ht_0             ; ht[0]++
+    STD     lt_0,0(r_ptr)             ; rp[0] = lt[0]
+    STD     ht_0,8(r_ptr)             ; rp[1] = ht[1]
+
+    ADD     lt_1,m_1,lt_1             ; lt = lt+m
+    ADD,DC  ht_1,%r0,ht_1             ; ht[1]++
+    STD     lt_1,16(r_ptr)            ; rp[2] = lt[1]
+    STD     ht_1,24(r_ptr)            ; rp[3] = ht[1]
+
+	LDO    -2(num),num                ; num = num - 2;
+    LDO     16(a_ptr),a_ptr           ; ap += 2
+	CMPIB,<= 2,num,bn_sqr_words_unroll2
+    LDO     32(r_ptr),r_ptr           ; rp += 4
+
+    CMPIB,=,N 0,num,bn_sqr_words_exit ; are we done?
+
+	;
+	; Top of loop aligned on 64-byte boundary
+	;
+bn_sqr_words_single_top
+    FLDD    0(a_ptr),t_float_0        ; load up 64-bit value (fr8L) ht(L)/lt(R)
+
+    XMPYU   fht_0,flt_0,fm            ; m
+    FSTD    fm,-24(%sp)               ; store m
+
+    XMPYU   flt_0,flt_0,lt_temp       ; lt
+    FSTD    lt_temp,-16(%sp)          ; store lt
+
+    XMPYU   fht_0,fht_0,ht_temp       ; ht
+    FSTD    ht_temp,-8(%sp)           ; store ht
+
+    LDD     -24(%sp),m_0              ; load m
+    AND     m_0,high_mask,tmp_0       ; m & Mask
+    DEPD,Z  m_0,30,31,m_0             ; m << 32+1
+    LDD     -16(%sp),lt_0             ; lt
+
+    LDD     -8(%sp),ht_0              ; ht
+    EXTRD,U tmp_0,32,33,tmp_0         ; tmp_0 = m&Mask >> 32-1
+    ADD     m_0,lt_0,lt_0             ; lt = lt+m
+    ADD,L   ht_0,tmp_0,ht_0           ; ht += tmp_0
+    ADD,DC  ht_0,%r0,ht_0             ; ht++
+
+    STD     lt_0,0(r_ptr)             ; rp[0] = lt
+    STD     ht_0,8(r_ptr)             ; rp[1] = ht
+
+bn_sqr_words_exit
+    .EXIT
+    LDD     -112(%sp),%r5       ; restore r5  
+    LDD     -120(%sp),%r4       ; restore r4  
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3 
+	.PROCEND	;in=23,24,25,26,29;out=28;
+
+
+;----------------------------------------------------------------------------
+;
+;BN_ULONG bn_add_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n)
+;
+; arg0 = rp 
+; arg1 = ap
+; arg2 = bp 
+; arg3 = n
+
+t  .reg %r22
+b  .reg %r21
+l  .reg %r20
+
+bn_add_words
+	.proc
+    .entry
+	.callinfo
+	.EXPORT	bn_add_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+	.align 64
+
+    CMPIB,>= 0,n,bn_add_words_exit
+    COPY    %r0,%ret1           ; return 0 by default
+
+	;
+	; If 2 or more numbers do the loop
+	;
+	CMPIB,= 1,n,bn_add_words_single_top
+	NOP
+
+	;
+	; This loop is unrolled 2 times (64-byte aligned as well)
+	;
+bn_add_words_unroll2
+	LDD     0(a_ptr),t
+	LDD     0(b_ptr),b
+	ADD     t,%ret1,t                    ; t = t+c;
+	ADD,DC  %r0,%r0,%ret1                ; set c to carry
+	ADD     t,b,l                        ; l = t + b[0]
+	ADD,DC  %ret1,%r0,%ret1              ; c+= carry
+	STD     l,0(r_ptr)
+
+	LDD     8(a_ptr),t
+	LDD     8(b_ptr),b
+	ADD     t,%ret1,t                     ; t = t+c;
+	ADD,DC  %r0,%r0,%ret1                 ; set c to carry
+	ADD     t,b,l                         ; l = t + b[0]
+	ADD,DC  %ret1,%r0,%ret1               ; c+= carry
+	STD     l,8(r_ptr)
+
+	LDO     -2(n),n
+	LDO     16(a_ptr),a_ptr
+	LDO     16(b_ptr),b_ptr
+
+	CMPIB,<= 2,n,bn_add_words_unroll2
+	LDO     16(r_ptr),r_ptr
+
+    CMPIB,=,N 0,n,bn_add_words_exit ; are we done?
+
+bn_add_words_single_top
+	LDD     0(a_ptr),t
+	LDD     0(b_ptr),b
+
+	ADD     t,%ret1,t                 ; t = t+c;
+	ADD,DC  %r0,%r0,%ret1             ; set c to carry (could use CMPCLR??)
+	ADD     t,b,l                     ; l = t + b[0]
+	ADD,DC  %ret1,%r0,%ret1           ; c+= carry
+	STD     l,0(r_ptr)
+
+bn_add_words_exit
+    .EXIT
+    BVE     (%rp)
+    EXTRD,U %ret1,31,32,%ret0           ; for 32-bit, return in ret0/ret1
+	.PROCEND	;in=23,24,25,26,29;out=28;
+
+;----------------------------------------------------------------------------
+;
+;BN_ULONG bn_sub_words(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b, int n)
+;
+; arg0 = rp 
+; arg1 = ap
+; arg2 = bp 
+; arg3 = n
+
+t1       .reg %r22
+t2       .reg %r21
+sub_tmp1 .reg %r20
+sub_tmp2 .reg %r19
+
+
+bn_sub_words
+	.proc
+	.callinfo 
+	.EXPORT	bn_sub_words,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+    .entry
+	.align 64
+
+    CMPIB,>=  0,n,bn_sub_words_exit
+    COPY    %r0,%ret1           ; return 0 by default
+
+	;
+	; If 2 or more numbers do the loop
+	;
+	CMPIB,= 1,n,bn_sub_words_single_top
+	NOP
+
+	;
+	; This loop is unrolled 2 times (64-byte aligned as well)
+	;
+bn_sub_words_unroll2
+	LDD     0(a_ptr),t1
+	LDD     0(b_ptr),t2
+	SUB     t1,t2,sub_tmp1           ; t3 = t1-t2; 
+	SUB     sub_tmp1,%ret1,sub_tmp1  ; t3 = t3- c; 
+
+	CMPCLR,*>> t1,t2,sub_tmp2        ; clear if t1 > t2
+	LDO      1(%r0),sub_tmp2
+	
+	CMPCLR,*= t1,t2,%r0
+	COPY    sub_tmp2,%ret1
+	STD     sub_tmp1,0(r_ptr)
+
+	LDD     8(a_ptr),t1
+	LDD     8(b_ptr),t2
+	SUB     t1,t2,sub_tmp1            ; t3 = t1-t2; 
+	SUB     sub_tmp1,%ret1,sub_tmp1   ; t3 = t3- c; 
+	CMPCLR,*>> t1,t2,sub_tmp2         ; clear if t1 > t2
+	LDO      1(%r0),sub_tmp2
+	
+	CMPCLR,*= t1,t2,%r0
+	COPY    sub_tmp2,%ret1
+	STD     sub_tmp1,8(r_ptr)
+
+	LDO     -2(n),n
+	LDO     16(a_ptr),a_ptr
+	LDO     16(b_ptr),b_ptr
+
+	CMPIB,<= 2,n,bn_sub_words_unroll2
+	LDO     16(r_ptr),r_ptr
+
+    CMPIB,=,N 0,n,bn_sub_words_exit ; are we done?
+
+bn_sub_words_single_top
+	LDD     0(a_ptr),t1
+	LDD     0(b_ptr),t2
+	SUB     t1,t2,sub_tmp1            ; t3 = t1-t2; 
+	SUB     sub_tmp1,%ret1,sub_tmp1   ; t3 = t3- c; 
+	CMPCLR,*>> t1,t2,sub_tmp2         ; clear if t1 > t2
+	LDO      1(%r0),sub_tmp2
+	
+	CMPCLR,*= t1,t2,%r0
+	COPY    sub_tmp2,%ret1
+
+	STD     sub_tmp1,0(r_ptr)
+
+bn_sub_words_exit
+    .EXIT
+    BVE     (%rp)
+    EXTRD,U %ret1,31,32,%ret0           ; for 32-bit, return in ret0/ret1
+	.PROCEND	;in=23,24,25,26,29;out=28;
+
+;------------------------------------------------------------------------------
+;
+; unsigned long bn_div_words(unsigned long h, unsigned long l, unsigned long d)
+;
+; arg0 = h
+; arg1 = l
+; arg2 = d
+;
+; This is mainly just output from the HP C compiler.  
+;
+;------------------------------------------------------------------------------
+bn_div_words
 	.PROC
-	.CALLINFO FRAME=0,NO_CALLS
-	.ENTRY
-	ldo 28(%r26),%r19
-	ldo 12(%r25),%r28
-L$0042
-	fldws 0(0,%r25),%fr8L
-	fldws 0(0,%r25),%fr8R
-	xmpyu %fr8L,%fr8R,%fr8
-	fstds %fr8,-16(0,%r30)
-	ldw -16(0,%r30),%r22
-	ldw -12(0,%r30),%r23
-	stw %r23,0(0,%r26)
-	copy %r22,%r21
-	ldi 0,%r20
-	addib,= -1,%r24,L$0049
-	stw %r21,-24(0,%r19)
-	fldws -8(0,%r28),%fr8L
-	fldws -8(0,%r28),%fr8R
-	xmpyu %fr8L,%fr8R,%fr8
-	fstds %fr8,-16(0,%r30)
-	ldw -16(0,%r30),%r22
-	ldw -12(0,%r30),%r23
-	stw %r23,-20(0,%r19)
-	copy %r22,%r21
-	ldi 0,%r20
-	addib,= -1,%r24,L$0049
-	stw %r21,-16(0,%r19)
-	fldws -4(0,%r28),%fr8L
-	fldws -4(0,%r28),%fr8R
-	xmpyu %fr8L,%fr8R,%fr8
-	fstds %fr8,-16(0,%r30)
-	ldw -16(0,%r30),%r22
-	ldw -12(0,%r30),%r23
-	stw %r23,-12(0,%r19)
-	copy %r22,%r21
-	ldi 0,%r20
-	addib,= -1,%r24,L$0049
-	stw %r21,-8(0,%r19)
-	fldws 0(0,%r28),%fr8L
-	fldws 0(0,%r28),%fr8R
-	xmpyu %fr8L,%fr8R,%fr8
-	fstds %fr8,-16(0,%r30)
-	ldw -16(0,%r30),%r22
-	ldw -12(0,%r30),%r23
-	stw %r23,-4(0,%r19)
-	copy %r22,%r21
-	ldi 0,%r20
-	addib,= -1,%r24,L$0049
-	stw %r21,0(0,%r19)
-	ldo 16(%r28),%r28
-	ldo 16(%r25),%r25
-	ldo 32(%r19),%r19
-	bl L$0042,0
-	ldo 32(%r26),%r26
-L$0049
-	bv,n 0(%r2)
-	.EXIT
-	.PROCEND
-	.IMPORT BN_num_bits_word,CODE
-	.IMPORT fprintf,CODE
-	.IMPORT __iob,DATA
-	.SPACE $TEXT$
-	.SUBSPA $LIT$
-
-	.align 4
-L$C0000
-	.STRING "Division would overflow (%d)\x0a\x00"
-	.IMPORT abort,CODE
-	.SPACE $TEXT$
-	.SUBSPA $CODE$
-
-	.align 4
-	.EXPORT bn_div64,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,RTNVAL=GR
-bn_div64
+	.EXPORT	bn_div_words,ENTRY,PRIV_LEV=3,ARGW0=GR,ARGW1=GR,ARGW2=GR,ARGW3=GR,RTNVAL=GR,LONG_RETURN
+	.IMPORT	BN_num_bits_word,CODE
+	.IMPORT	__iob,DATA
+	.IMPORT	fprintf,CODE
+	.IMPORT	abort,CODE
+	.IMPORT	$$div2U,MILLICODE
+	.CALLINFO CALLER,FRAME=144,ENTRY_GR=%r9,SAVE_RP,ARGS_SAVED,ORDERING_AWARE
+        .ENTRY
+        STW     %r2,-20(%r30)   ;offset 0x8ec
+        STW,MA  %r3,192(%r30)   ;offset 0x8f0
+        STW     %r4,-188(%r30)  ;offset 0x8f4
+        DEPD    %r5,31,32,%r6   ;offset 0x8f8
+        STD     %r6,-184(%r30)  ;offset 0x8fc
+        DEPD    %r7,31,32,%r8   ;offset 0x900
+        STD     %r8,-176(%r30)  ;offset 0x904
+        STW     %r9,-168(%r30)  ;offset 0x908
+        LDD     -248(%r30),%r3  ;offset 0x90c
+        COPY    %r26,%r4        ;offset 0x910
+        COPY    %r24,%r5        ;offset 0x914
+        DEPD    %r25,31,32,%r4  ;offset 0x918
+        CMPB,*<>        %r3,%r0,$0006000C       ;offset 0x91c
+        DEPD    %r23,31,32,%r5  ;offset 0x920
+        MOVIB,TR        -1,%r29,$00060002       ;offset 0x924
+        EXTRD,U %r29,31,32,%r28 ;offset 0x928
+$0006002A
+        LDO     -1(%r29),%r29   ;offset 0x92c
+        SUB     %r23,%r7,%r23   ;offset 0x930
+$00060024
+        SUB     %r4,%r31,%r25   ;offset 0x934
+        AND     %r25,%r19,%r26  ;offset 0x938
+        CMPB,*<>,N      %r0,%r26,$00060046      ;offset 0x93c
+        DEPD,Z  %r25,31,32,%r20 ;offset 0x940
+        OR      %r20,%r24,%r21  ;offset 0x944
+        CMPB,*<<,N      %r21,%r23,$0006002A     ;offset 0x948
+        SUB     %r31,%r2,%r31   ;offset 0x94c
+$00060046
+$0006002E
+        DEPD,Z  %r23,31,32,%r25 ;offset 0x950
+        EXTRD,U %r23,31,32,%r26 ;offset 0x954
+        AND     %r25,%r19,%r24  ;offset 0x958
+        ADD,L   %r31,%r26,%r31  ;offset 0x95c
+        CMPCLR,*>>=     %r5,%r24,%r0    ;offset 0x960
+        LDO     1(%r31),%r31    ;offset 0x964
+$00060032
+        CMPB,*<<=,N     %r31,%r4,$00060036      ;offset 0x968
+        LDO     -1(%r29),%r29   ;offset 0x96c
+        ADD,L   %r4,%r3,%r4     ;offset 0x970
+$00060036
+        ADDIB,=,N       -1,%r8,$D0      ;offset 0x974
+        SUB     %r5,%r24,%r28   ;offset 0x978
+$0006003A
+        SUB     %r4,%r31,%r24   ;offset 0x97c
+        SHRPD   %r24,%r28,32,%r4        ;offset 0x980
+        DEPD,Z  %r29,31,32,%r9  ;offset 0x984
+        DEPD,Z  %r28,31,32,%r5  ;offset 0x988
+$0006001C
+        EXTRD,U %r4,31,32,%r31  ;offset 0x98c
+        CMPB,*<>,N      %r31,%r2,$00060020      ;offset 0x990
+        MOVB,TR %r6,%r29,$D1    ;offset 0x994
+        STD     %r29,-152(%r30) ;offset 0x998
+$0006000C
+        EXTRD,U %r3,31,32,%r25  ;offset 0x99c
+        COPY    %r3,%r26        ;offset 0x9a0
+        EXTRD,U %r3,31,32,%r9   ;offset 0x9a4
+        EXTRD,U %r4,31,32,%r8   ;offset 0x9a8
+        .CALL   ARGW0=GR,ARGW1=GR,RTNVAL=GR     ;in=25,26;out=28;
+        B,L     BN_num_bits_word,%r2    ;offset 0x9ac
+        EXTRD,U %r5,31,32,%r7   ;offset 0x9b0
+        LDI     64,%r20 ;offset 0x9b4
+        DEPD    %r7,31,32,%r5   ;offset 0x9b8
+        DEPD    %r8,31,32,%r4   ;offset 0x9bc
+        DEPD    %r9,31,32,%r3   ;offset 0x9c0
+        CMPB,=  %r28,%r20,$00060012     ;offset 0x9c4
+        COPY    %r28,%r24       ;offset 0x9c8
+        MTSARCM %r24    ;offset 0x9cc
+        DEPDI,Z -1,%sar,1,%r19  ;offset 0x9d0
+        CMPB,*>>,N      %r4,%r19,$D2    ;offset 0x9d4
+$00060012
+        SUBI    64,%r24,%r31    ;offset 0x9d8
+        CMPCLR,*<<      %r4,%r3,%r0     ;offset 0x9dc
+        SUB     %r4,%r3,%r4     ;offset 0x9e0
+$00060016
+        CMPB,=  %r31,%r0,$0006001A      ;offset 0x9e4
+        COPY    %r0,%r9 ;offset 0x9e8
+        MTSARCM %r31    ;offset 0x9ec
+        DEPD,Z  %r3,%sar,64,%r3 ;offset 0x9f0
+        SUBI    64,%r31,%r26    ;offset 0x9f4
+        MTSAR   %r26    ;offset 0x9f8
+        SHRPD   %r4,%r5,%sar,%r4        ;offset 0x9fc
+        MTSARCM %r31    ;offset 0xa00
+        DEPD,Z  %r5,%sar,64,%r5 ;offset 0xa04
+$0006001A
+        DEPDI,Z -1,31,32,%r19   ;offset 0xa08
+        AND     %r3,%r19,%r29   ;offset 0xa0c
+        EXTRD,U %r29,31,32,%r2  ;offset 0xa10
+        DEPDI,Z -1,63,32,%r6    ;offset 0xa14
+        MOVIB,TR        2,%r8,$0006001C ;offset 0xa18
+        EXTRD,U %r3,63,32,%r7   ;offset 0xa1c
+$D2
+        ADDIL   LR'__iob-$global$,%r27,%r1      ;offset 0xa20
+        LDIL    LR'C$7,%r21     ;offset 0xa24
+        LDO     RR'__iob-$global$+32(%r1),%r26  ;offset 0xa28
+        .CALL   ARGW0=GR,ARGW1=GR,ARGW2=GR,RTNVAL=GR    ;in=24,25,26;out=28;
+        B,L     fprintf,%r2     ;offset 0xa2c
+        LDO     RR'C$7(%r21),%r25       ;offset 0xa30
+        .CALL           ;
+        B,L     abort,%r2       ;offset 0xa34
+        NOP             ;offset 0xa38
+        B       $D3     ;offset 0xa3c
+        LDW     -212(%r30),%r2  ;offset 0xa40
+$00060020
+        COPY    %r4,%r26        ;offset 0xa44
+        EXTRD,U %r4,31,32,%r25  ;offset 0xa48
+        COPY    %r2,%r24        ;offset 0xa4c
+        .CALL   ;in=23,24,25,26;out=20,21,22,28,29; (MILLICALL)
+        B,L     $$div2U,%r31    ;offset 0xa50
+        EXTRD,U %r2,31,32,%r23  ;offset 0xa54
+        DEPD    %r28,31,32,%r29 ;offset 0xa58
+$00060022
+        STD     %r29,-152(%r30) ;offset 0xa5c
+$D1
+        AND     %r5,%r19,%r24   ;offset 0xa60
+        EXTRD,U %r24,31,32,%r24 ;offset 0xa64
+        STW     %r2,-160(%r30)  ;offset 0xa68
+        STW     %r7,-128(%r30)  ;offset 0xa6c
+        FLDD    -152(%r30),%fr4 ;offset 0xa70
+        FLDD    -152(%r30),%fr7 ;offset 0xa74
+        FLDW    -160(%r30),%fr8L        ;offset 0xa78
+        FLDW    -128(%r30),%fr5L        ;offset 0xa7c
+        XMPYU   %fr8L,%fr7L,%fr10       ;offset 0xa80
+        FSTD    %fr10,-136(%r30)        ;offset 0xa84
+        XMPYU   %fr8L,%fr7R,%fr22       ;offset 0xa88
+        FSTD    %fr22,-144(%r30)        ;offset 0xa8c
+        XMPYU   %fr5L,%fr4L,%fr11       ;offset 0xa90
+        XMPYU   %fr5L,%fr4R,%fr23       ;offset 0xa94
+        FSTD    %fr11,-112(%r30)        ;offset 0xa98
+        FSTD    %fr23,-120(%r30)        ;offset 0xa9c
+        LDD     -136(%r30),%r28 ;offset 0xaa0
+        DEPD,Z  %r28,31,32,%r31 ;offset 0xaa4
+        LDD     -144(%r30),%r20 ;offset 0xaa8
+        ADD,L   %r20,%r31,%r31  ;offset 0xaac
+        LDD     -112(%r30),%r22 ;offset 0xab0
+        DEPD,Z  %r22,31,32,%r22 ;offset 0xab4
+        LDD     -120(%r30),%r21 ;offset 0xab8
+        B       $00060024       ;offset 0xabc
+        ADD,L   %r21,%r22,%r23  ;offset 0xac0
+$D0
+        OR      %r9,%r29,%r29   ;offset 0xac4
+$00060040
+        EXTRD,U %r29,31,32,%r28 ;offset 0xac8
+$00060002
+$L2
+        LDW     -212(%r30),%r2  ;offset 0xacc
+$D3
+        LDW     -168(%r30),%r9  ;offset 0xad0
+        LDD     -176(%r30),%r8  ;offset 0xad4
+        EXTRD,U %r8,31,32,%r7   ;offset 0xad8
+        LDD     -184(%r30),%r6  ;offset 0xadc
+        EXTRD,U %r6,31,32,%r5   ;offset 0xae0
+        LDW     -188(%r30),%r4  ;offset 0xae4
+        BVE     (%r2)   ;offset 0xae8
+        .EXIT
+        LDW,MB  -192(%r30),%r3  ;offset 0xaec
+	.PROCEND	;in=23,25;out=28,29;fpin=105,107;
+
+
+
+
+;----------------------------------------------------------------------------
+;
+; Registers to hold 64-bit values to manipulate.  The "L" part
+; of the register corresponds to the upper 32-bits, while the "R"
+; part corresponds to the lower 32-bits
+; 
+; Note, that when using b6 and b7, the code must save these before
+; using them because they are callee save registers 
+; 
+;
+; Floating point registers to use to save values that
+; are manipulated.  These don't collide with ftemp1-6 and
+; are all caller save registers
+;
+a0        .reg %fr22
+a0L       .reg %fr22L
+a0R       .reg %fr22R
+
+a1        .reg %fr23
+a1L       .reg %fr23L
+a1R       .reg %fr23R
+
+a2        .reg %fr24
+a2L       .reg %fr24L
+a2R       .reg %fr24R
+
+a3        .reg %fr25
+a3L       .reg %fr25L
+a3R       .reg %fr25R
+
+a4        .reg %fr26
+a4L       .reg %fr26L
+a4R       .reg %fr26R
+
+a5        .reg %fr27
+a5L       .reg %fr27L
+a5R       .reg %fr27R
+
+a6        .reg %fr28
+a6L       .reg %fr28L
+a6R       .reg %fr28R
+
+a7        .reg %fr29
+a7L       .reg %fr29L
+a7R       .reg %fr29R
+
+b0        .reg %fr30
+b0L       .reg %fr30L
+b0R       .reg %fr30R
+
+b1        .reg %fr31
+b1L       .reg %fr31L
+b1R       .reg %fr31R
+
+;
+; Temporary floating point variables, these are all caller save
+; registers
+;
+ftemp1    .reg %fr4
+ftemp2    .reg %fr5
+ftemp3    .reg %fr6
+ftemp4    .reg %fr7
+
+;
+; The B set of registers when used.
+;
+
+b2        .reg %fr8
+b2L       .reg %fr8L
+b2R       .reg %fr8R
+
+b3        .reg %fr9
+b3L       .reg %fr9L
+b3R       .reg %fr9R
+
+b4        .reg %fr10
+b4L       .reg %fr10L
+b4R       .reg %fr10R
+
+b5        .reg %fr11
+b5L       .reg %fr11L
+b5R       .reg %fr11R
+
+b6        .reg %fr12
+b6L       .reg %fr12L
+b6R       .reg %fr12R
+
+b7        .reg %fr13
+b7L       .reg %fr13L
+b7R       .reg %fr13R
+
+c1           .reg %r21   ; only reg
+temp1        .reg %r20   ; only reg
+temp2        .reg %r19   ; only reg
+temp3        .reg %r31   ; only reg
+
+m1           .reg %r28   
+c2           .reg %r23   
+high_one     .reg %r1
+ht           .reg %r6
+lt           .reg %r5
+m            .reg %r4
+c3           .reg %r3
+
+SQR_ADD_C  .macro  A0L,A0R,C1,C2,C3
+    XMPYU   A0L,A0R,ftemp1       ; m
+    FSTD    ftemp1,-24(%sp)      ; store m
+
+    XMPYU   A0R,A0R,ftemp2       ; lt
+    FSTD    ftemp2,-16(%sp)      ; store lt
+
+    XMPYU   A0L,A0L,ftemp3       ; ht
+    FSTD    ftemp3,-8(%sp)       ; store ht
+
+    LDD     -24(%sp),m           ; load m
+    AND     m,high_mask,temp2    ; m & Mask
+    DEPD,Z  m,30,31,temp3        ; m << 32+1
+    LDD     -16(%sp),lt          ; lt
+
+    LDD     -8(%sp),ht           ; ht
+    EXTRD,U temp2,32,33,temp1    ; temp1 = m&Mask >> 32-1
+    ADD     temp3,lt,lt          ; lt = lt+m
+    ADD,L   ht,temp1,ht          ; ht += temp1
+    ADD,DC  ht,%r0,ht            ; ht++
+
+    ADD     C1,lt,C1             ; c1=c1+lt
+    ADD,DC  ht,%r0,ht            ; ht++
+
+    ADD     C2,ht,C2             ; c2=c2+ht
+    ADD,DC  C3,%r0,C3            ; c3++
+.endm
+
+SQR_ADD_C2 .macro  A0L,A0R,A1L,A1R,C1,C2,C3
+    XMPYU   A0L,A1R,ftemp1          ; m1 = bl*ht
+    FSTD    ftemp1,-16(%sp)         ;
+    XMPYU   A0R,A1L,ftemp2          ; m = bh*lt
+    FSTD    ftemp2,-8(%sp)          ;
+    XMPYU   A0R,A1R,ftemp3          ; lt = bl*lt
+    FSTD    ftemp3,-32(%sp)
+    XMPYU   A0L,A1L,ftemp4          ; ht = bh*ht
+    FSTD    ftemp4,-24(%sp)         ;
+
+    LDD     -8(%sp),m               ; r21 = m
+    LDD     -16(%sp),m1             ; r19 = m1
+    ADD,L   m,m1,m                  ; m+m1
+
+    DEPD,Z  m,31,32,temp3           ; (m+m1<<32)
+    LDD     -24(%sp),ht             ; r24 = ht
+
+    CMPCLR,*>>= m,m1,%r0            ; if (m < m1)
+    ADD,L   ht,high_one,ht          ; ht+=high_one
+
+    EXTRD,U m,31,32,temp1           ; m >> 32
+    LDD     -32(%sp),lt             ; lt
+    ADD,L   ht,temp1,ht             ; ht+= m>>32
+    ADD     lt,temp3,lt             ; lt = lt+m1
+    ADD,DC  ht,%r0,ht               ; ht++
+
+    ADD     ht,ht,ht                ; ht=ht+ht;
+    ADD,DC  C3,%r0,C3               ; add in carry (c3++)
+
+    ADD     lt,lt,lt                ; lt=lt+lt;
+    ADD,DC  ht,%r0,ht               ; add in carry (ht++)
+
+    ADD     C1,lt,C1                ; c1=c1+lt
+    ADD,DC,*NUV ht,%r0,ht           ; add in carry (ht++)
+    LDO     1(C3),C3              ; bump c3 if overflow,nullify otherwise
+
+    ADD     C2,ht,C2                ; c2 = c2 + ht
+    ADD,DC  C3,%r0,C3             ; add in carry (c3++)
+.endm
+
+;
+;void bn_sqr_comba8(BN_ULONG *r, BN_ULONG *a)
+; arg0 = r_ptr
+; arg1 = a_ptr
+;
+
+bn_sqr_comba8
 	.PROC
-	.CALLINFO FRAME=128,CALLS,SAVE_RP,ENTRY_GR=8
-	.ENTRY
-	stw %r2,-20(0,%r30)
-	stwm %r8,128(0,%r30)
-	stw %r7,-124(0,%r30)
-	stw %r4,-112(0,%r30)
-	stw %r3,-108(0,%r30)
-	copy %r26,%r3
-	copy %r25,%r4
-	stw %r6,-120(0,%r30)
-	ldi 0,%r7
-	stw %r5,-116(0,%r30)
-	movb,<> %r24,%r5,L$0051
-	ldi 2,%r6
-	bl L$0068,0
-	ldi -1,%r28
-L$0051
-	.CALL ARGW0=GR
-	bl BN_num_bits_word,%r2
-	copy %r5,%r26
-	copy %r28,%r24
-	ldi 32,%r19
-	comb,= %r19,%r24,L$0052
-	subi 31,%r24,%r19
-	mtsar %r19
-	zvdepi 1,32,%r19
-	comb,>>= %r19,%r3,L$0052
-	addil LR'__iob-$global$+32,%r27
-	ldo RR'__iob-$global$+32(%r1),%r26
-	ldil LR'L$C0000,%r25
-	.CALL ARGW0=GR,ARGW1=GR,ARGW2=GR
-	bl fprintf,%r2
-	ldo RR'L$C0000(%r25),%r25
-	.CALL 
-	bl abort,%r2
-	nop
-L$0052
-	comb,>> %r5,%r3,L$0053
-	subi 32,%r24,%r24
-	sub %r3,%r5,%r3
-L$0053
-	comib,= 0,%r24,L$0054
-	subi 31,%r24,%r19
-	mtsar %r19
-	zvdep %r5,32,%r5
-	zvdep %r3,32,%r21
-	subi 32,%r24,%r20
-	mtsar %r20
-	vshd 0,%r4,%r20
-	or %r21,%r20,%r3
-	mtsar %r19
-	zvdep %r4,32,%r4
-L$0054
-	extru %r5,15,16,%r23
-	extru %r5,31,16,%r28
-L$0055
-	extru %r3,15,16,%r19
-	comb,<> %r23,%r19,L$0058
-	copy %r3,%r26
-	bl L$0059,0
-	zdepi -1,31,16,%r29
-L$0058
-	.IMPORT $$divU,MILLICODE
-	bl $$divU,%r31
-	copy %r23,%r25
-L$0059
-	stw %r29,-16(0,%r30)
-	fldws -16(0,%r30),%fr10L
-	stw %r28,-16(0,%r30)
-	fldws -16(0,%r30),%fr10R
-	stw %r23,-16(0,%r30)
-	xmpyu %fr10L,%fr10R,%fr8
-	fldws -16(0,%r30),%fr10R
-	fstws %fr8R,-16(0,%r30)
-	xmpyu %fr10L,%fr10R,%fr9
-	ldw -16(0,%r30),%r8
-	fstws %fr9R,-16(0,%r30)
-	copy %r8,%r22
-	ldw -16(0,%r30),%r8
-	extru %r4,15,16,%r24
-	copy %r8,%r21
-L$0060
-	sub %r3,%r21,%r20
-	copy %r20,%r19
-	depi 0,31,16,%r19
-	comib,<> 0,%r19,L$0061
-	zdep %r20,15,16,%r19
-	addl %r19,%r24,%r19
-	comb,>>= %r19,%r22,L$0061
-	sub %r22,%r28,%r22
-	sub %r21,%r23,%r21
-	bl L$0060,0
-	ldo -1(%r29),%r29
-L$0061
-	stw %r29,-16(0,%r30)
-	fldws -16(0,%r30),%fr10L
-	stw %r28,-16(0,%r30)
-	fldws -16(0,%r30),%fr10R
-	xmpyu %fr10L,%fr10R,%fr8
-	fstws %fr8R,-16(0,%r30)
-	ldw -16(0,%r30),%r8
-	stw %r23,-16(0,%r30)
-	fldws -16(0,%r30),%fr10R
-	copy %r8,%r19
-	xmpyu %fr10L,%fr10R,%fr8
-	fstws %fr8R,-16(0,%r30)
-	extru %r19,15,16,%r20
-	ldw -16(0,%r30),%r8
-	zdep %r19,15,16,%r19
-	addl %r8,%r20,%r20
-	comclr,<<= %r19,%r4,0
-	addi 1,%r20,%r20
-	comb,<<= %r20,%r3,L$0066
-	sub %r4,%r19,%r4
-	addl %r3,%r5,%r3
-	ldo -1(%r29),%r29
-L$0066
-	addib,= -1,%r6,L$0056
-	sub %r3,%r20,%r3
-	zdep %r29,15,16,%r7
-	shd %r3,%r4,16,%r3
-	bl L$0055,0
-	zdep %r4,15,16,%r4
-L$0056
-	or %r7,%r29,%r28
-L$0068
-	ldw -148(0,%r30),%r2
-	ldw -124(0,%r30),%r7
-	ldw -120(0,%r30),%r6
-	ldw -116(0,%r30),%r5
-	ldw -112(0,%r30),%r4
-	ldw -108(0,%r30),%r3
-	bv 0(%r2)
-	ldwm -128(0,%r30),%r8
-	.EXIT
-	.PROCEND
+	.CALLINFO FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
+	.EXPORT	bn_sqr_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+    .ENTRY
+	.align 64
+
+    STD     %r3,0(%sp)          ; save r3
+    STD     %r4,8(%sp)          ; save r4
+    STD     %r5,16(%sp)         ; save r5
+    STD     %r6,24(%sp)         ; save r6
+
+	;
+	; Zero out carries
+	;
+	COPY     %r0,c1
+	COPY     %r0,c2
+	COPY     %r0,c3
+
+	LDO      128(%sp),%sp       ; bump stack
+    DEPDI,Z -1,32,33,high_mask   ; Create Mask 0xffffffff80000000L
+    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32
+
+	;
+	; Load up all of the values we are going to use
+	;
+    FLDD     0(a_ptr),a0       
+    FLDD     8(a_ptr),a1       
+    FLDD    16(a_ptr),a2       
+    FLDD    24(a_ptr),a3       
+    FLDD    32(a_ptr),a4       
+    FLDD    40(a_ptr),a5       
+    FLDD    48(a_ptr),a6       
+    FLDD    56(a_ptr),a7       
+
+	SQR_ADD_C a0L,a0R,c1,c2,c3
+	STD     c1,0(r_ptr)          ; r[0] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1
+	STD     c2,8(r_ptr)          ; r[1] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C a1L,a1R,c3,c1,c2
+	SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2
+	STD     c3,16(r_ptr)            ; r[2] = c3;
+	COPY    %r0,c3
+
+	SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3
+	SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3
+	STD     c1,24(r_ptr)           ; r[3] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C a2L,a2R,c2,c3,c1
+	SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1
+	SQR_ADD_C2 a4L,a4R,a0L,a0R,c2,c3,c1
+	STD     c2,32(r_ptr)          ; r[4] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C2 a5L,a5R,a0L,a0R,c3,c1,c2
+	SQR_ADD_C2 a4L,a4R,a1L,a1R,c3,c1,c2
+	SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2
+	STD     c3,40(r_ptr)          ; r[5] = c3;
+	COPY    %r0,c3
+
+	SQR_ADD_C a3L,a3R,c1,c2,c3
+	SQR_ADD_C2 a4L,a4R,a2L,a2R,c1,c2,c3
+	SQR_ADD_C2 a5L,a5R,a1L,a1R,c1,c2,c3
+	SQR_ADD_C2 a6L,a6R,a0L,a0R,c1,c2,c3
+	STD     c1,48(r_ptr)          ; r[6] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C2 a7L,a7R,a0L,a0R,c2,c3,c1
+	SQR_ADD_C2 a6L,a6R,a1L,a1R,c2,c3,c1
+	SQR_ADD_C2 a5L,a5R,a2L,a2R,c2,c3,c1
+	SQR_ADD_C2 a4L,a4R,a3L,a3R,c2,c3,c1
+	STD     c2,56(r_ptr)          ; r[7] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C a4L,a4R,c3,c1,c2
+	SQR_ADD_C2 a5L,a5R,a3L,a3R,c3,c1,c2
+	SQR_ADD_C2 a6L,a6R,a2L,a2R,c3,c1,c2
+	SQR_ADD_C2 a7L,a7R,a1L,a1R,c3,c1,c2
+	STD     c3,64(r_ptr)          ; r[8] = c3;
+	COPY    %r0,c3
+
+	SQR_ADD_C2 a7L,a7R,a2L,a2R,c1,c2,c3
+	SQR_ADD_C2 a6L,a6R,a3L,a3R,c1,c2,c3
+	SQR_ADD_C2 a5L,a5R,a4L,a4R,c1,c2,c3
+	STD     c1,72(r_ptr)          ; r[9] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C a5L,a5R,c2,c3,c1
+	SQR_ADD_C2 a6L,a6R,a4L,a4R,c2,c3,c1
+	SQR_ADD_C2 a7L,a7R,a3L,a3R,c2,c3,c1
+	STD     c2,80(r_ptr)          ; r[10] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C2 a7L,a7R,a4L,a4R,c3,c1,c2
+	SQR_ADD_C2 a6L,a6R,a5L,a5R,c3,c1,c2
+	STD     c3,88(r_ptr)          ; r[11] = c3;
+	COPY    %r0,c3
+	
+	SQR_ADD_C a6L,a6R,c1,c2,c3
+	SQR_ADD_C2 a7L,a7R,a5L,a5R,c1,c2,c3
+	STD     c1,96(r_ptr)          ; r[12] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C2 a7L,a7R,a6L,a6R,c2,c3,c1
+	STD     c2,104(r_ptr)         ; r[13] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C a7L,a7R,c3,c1,c2
+	STD     c3, 112(r_ptr)       ; r[14] = c3
+	STD     c1, 120(r_ptr)       ; r[15] = c1
+
+    .EXIT
+    LDD     -104(%sp),%r6        ; restore r6
+    LDD     -112(%sp),%r5        ; restore r5
+    LDD     -120(%sp),%r4        ; restore r4
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3
+
+	.PROCEND	
+
+;-----------------------------------------------------------------------------
+;
+;void bn_sqr_comba4(BN_ULONG *r, BN_ULONG *a)
+; arg0 = r_ptr
+; arg1 = a_ptr
+;
+
+bn_sqr_comba4
+	.proc
+	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
+	.EXPORT	bn_sqr_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+    .entry
+	.align 64
+    STD     %r3,0(%sp)          ; save r3
+    STD     %r4,8(%sp)          ; save r4
+    STD     %r5,16(%sp)         ; save r5
+    STD     %r6,24(%sp)         ; save r6
+
+	;
+	; Zero out carries
+	;
+	COPY     %r0,c1
+	COPY     %r0,c2
+	COPY     %r0,c3
+
+	LDO      128(%sp),%sp       ; bump stack
+    DEPDI,Z -1,32,33,high_mask   ; Create Mask 0xffffffff80000000L
+    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32
+
+	;
+	; Load up all of the values we are going to use
+	;
+    FLDD     0(a_ptr),a0       
+    FLDD     8(a_ptr),a1       
+    FLDD    16(a_ptr),a2       
+    FLDD    24(a_ptr),a3       
+    FLDD    32(a_ptr),a4       
+    FLDD    40(a_ptr),a5       
+    FLDD    48(a_ptr),a6       
+    FLDD    56(a_ptr),a7       
+
+	SQR_ADD_C a0L,a0R,c1,c2,c3
+
+	STD     c1,0(r_ptr)          ; r[0] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C2 a1L,a1R,a0L,a0R,c2,c3,c1
+
+	STD     c2,8(r_ptr)          ; r[1] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C a1L,a1R,c3,c1,c2
+	SQR_ADD_C2 a2L,a2R,a0L,a0R,c3,c1,c2
+
+	STD     c3,16(r_ptr)            ; r[2] = c3;
+	COPY    %r0,c3
+
+	SQR_ADD_C2 a3L,a3R,a0L,a0R,c1,c2,c3
+	SQR_ADD_C2 a2L,a2R,a1L,a1R,c1,c2,c3
+
+	STD     c1,24(r_ptr)           ; r[3] = c1;
+	COPY    %r0,c1
+
+	SQR_ADD_C a2L,a2R,c2,c3,c1
+	SQR_ADD_C2 a3L,a3R,a1L,a1R,c2,c3,c1
+
+	STD     c2,32(r_ptr)           ; r[4] = c2;
+	COPY    %r0,c2
+
+	SQR_ADD_C2 a3L,a3R,a2L,a2R,c3,c1,c2
+	STD     c3,40(r_ptr)           ; r[5] = c3;
+	COPY    %r0,c3
+
+	SQR_ADD_C a3L,a3R,c1,c2,c3
+	STD     c1,48(r_ptr)           ; r[6] = c1;
+	STD     c2,56(r_ptr)           ; r[7] = c2;
+
+    .EXIT
+    LDD     -104(%sp),%r6        ; restore r6
+    LDD     -112(%sp),%r5        ; restore r5
+    LDD     -120(%sp),%r4        ; restore r4
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3
+
+	.PROCEND	
+
+
+;---------------------------------------------------------------------------
+
+MUL_ADD_C  .macro  A0L,A0R,B0L,B0R,C1,C2,C3
+    XMPYU   A0L,B0R,ftemp1        ; m1 = bl*ht
+    FSTD    ftemp1,-16(%sp)       ;
+    XMPYU   A0R,B0L,ftemp2        ; m = bh*lt
+    FSTD    ftemp2,-8(%sp)        ;
+    XMPYU   A0R,B0R,ftemp3        ; lt = bl*lt
+    FSTD    ftemp3,-32(%sp)
+    XMPYU   A0L,B0L,ftemp4        ; ht = bh*ht
+    FSTD    ftemp4,-24(%sp)       ;
+
+    LDD     -8(%sp),m             ; r21 = m
+    LDD     -16(%sp),m1           ; r19 = m1
+    ADD,L   m,m1,m                ; m+m1
+
+    DEPD,Z  m,31,32,temp3         ; (m+m1<<32)
+    LDD     -24(%sp),ht           ; r24 = ht
+
+    CMPCLR,*>>= m,m1,%r0          ; if (m < m1)
+    ADD,L   ht,high_one,ht        ; ht+=high_one
+
+    EXTRD,U m,31,32,temp1         ; m >> 32
+    LDD     -32(%sp),lt           ; lt
+    ADD,L   ht,temp1,ht           ; ht+= m>>32
+    ADD     lt,temp3,lt           ; lt = lt+m1
+    ADD,DC  ht,%r0,ht             ; ht++
+
+    ADD     C1,lt,C1              ; c1=c1+lt
+    ADD,DC  ht,%r0,ht             ; bump c3 if overflow,nullify otherwise
+
+    ADD     C2,ht,C2              ; c2 = c2 + ht
+    ADD,DC  C3,%r0,C3             ; add in carry (c3++)
+.endm
+
+
+;
+;void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
+; arg0 = r_ptr
+; arg1 = a_ptr
+; arg2 = b_ptr
+;
+
+bn_mul_comba8
+	.proc
+	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
+	.EXPORT	bn_mul_comba8,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+    .entry
+	.align 64
+
+    STD     %r3,0(%sp)          ; save r3
+    STD     %r4,8(%sp)          ; save r4
+    STD     %r5,16(%sp)         ; save r5
+    STD     %r6,24(%sp)         ; save r6
+    FSTD    %fr12,32(%sp)       ; save r6
+    FSTD    %fr13,40(%sp)       ; save r7
+
+	;
+	; Zero out carries
+	;
+	COPY     %r0,c1
+	COPY     %r0,c2
+	COPY     %r0,c3
+
+	LDO      128(%sp),%sp       ; bump stack
+    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32
+
+	;
+	; Load up all of the values we are going to use
+	;
+    FLDD      0(a_ptr),a0       
+    FLDD      8(a_ptr),a1       
+    FLDD     16(a_ptr),a2       
+    FLDD     24(a_ptr),a3       
+    FLDD     32(a_ptr),a4       
+    FLDD     40(a_ptr),a5       
+    FLDD     48(a_ptr),a6       
+    FLDD     56(a_ptr),a7       
+
+    FLDD      0(b_ptr),b0       
+    FLDD      8(b_ptr),b1       
+    FLDD     16(b_ptr),b2       
+    FLDD     24(b_ptr),b3       
+    FLDD     32(b_ptr),b4       
+    FLDD     40(b_ptr),b5       
+    FLDD     48(b_ptr),b6       
+    FLDD     56(b_ptr),b7       
+
+	MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3
+	STD       c1,0(r_ptr)
+	COPY      %r0,c1
+
+	MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1
+	MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1
+	STD       c2,8(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2
+	MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2
+	MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2
+	STD       c3,16(r_ptr)
+	COPY      %r0,c3
+
+	MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3
+	MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3
+	MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3
+	MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3
+	STD       c1,24(r_ptr)
+	COPY      %r0,c1
+
+	MUL_ADD_C a4L,a4R,b0L,b0R,c2,c3,c1
+	MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1
+	MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1
+	MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1
+	MUL_ADD_C a0L,a0R,b4L,b4R,c2,c3,c1
+	STD       c2,32(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a0L,a0R,b5L,b5R,c3,c1,c2
+	MUL_ADD_C a1L,a1R,b4L,b4R,c3,c1,c2
+	MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2
+	MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2
+	MUL_ADD_C a4L,a4R,b1L,b1R,c3,c1,c2
+	MUL_ADD_C a5L,a5R,b0L,b0R,c3,c1,c2
+	STD       c3,40(r_ptr)
+	COPY      %r0,c3
+
+	MUL_ADD_C a6L,a6R,b0L,b0R,c1,c2,c3
+	MUL_ADD_C a5L,a5R,b1L,b1R,c1,c2,c3
+	MUL_ADD_C a4L,a4R,b2L,b2R,c1,c2,c3
+	MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3
+	MUL_ADD_C a2L,a2R,b4L,b4R,c1,c2,c3
+	MUL_ADD_C a1L,a1R,b5L,b5R,c1,c2,c3
+	MUL_ADD_C a0L,a0R,b6L,b6R,c1,c2,c3
+	STD       c1,48(r_ptr)
+	COPY      %r0,c1
+	
+	MUL_ADD_C a0L,a0R,b7L,b7R,c2,c3,c1
+	MUL_ADD_C a1L,a1R,b6L,b6R,c2,c3,c1
+	MUL_ADD_C a2L,a2R,b5L,b5R,c2,c3,c1
+	MUL_ADD_C a3L,a3R,b4L,b4R,c2,c3,c1
+	MUL_ADD_C a4L,a4R,b3L,b3R,c2,c3,c1
+	MUL_ADD_C a5L,a5R,b2L,b2R,c2,c3,c1
+	MUL_ADD_C a6L,a6R,b1L,b1R,c2,c3,c1
+	MUL_ADD_C a7L,a7R,b0L,b0R,c2,c3,c1
+	STD       c2,56(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a7L,a7R,b1L,b1R,c3,c1,c2
+	MUL_ADD_C a6L,a6R,b2L,b2R,c3,c1,c2
+	MUL_ADD_C a5L,a5R,b3L,b3R,c3,c1,c2
+	MUL_ADD_C a4L,a4R,b4L,b4R,c3,c1,c2
+	MUL_ADD_C a3L,a3R,b5L,b5R,c3,c1,c2
+	MUL_ADD_C a2L,a2R,b6L,b6R,c3,c1,c2
+	MUL_ADD_C a1L,a1R,b7L,b7R,c3,c1,c2
+	STD       c3,64(r_ptr)
+	COPY      %r0,c3
+
+	MUL_ADD_C a2L,a2R,b7L,b7R,c1,c2,c3
+	MUL_ADD_C a3L,a3R,b6L,b6R,c1,c2,c3
+	MUL_ADD_C a4L,a4R,b5L,b5R,c1,c2,c3
+	MUL_ADD_C a5L,a5R,b4L,b4R,c1,c2,c3
+	MUL_ADD_C a6L,a6R,b3L,b3R,c1,c2,c3
+	MUL_ADD_C a7L,a7R,b2L,b2R,c1,c2,c3
+	STD       c1,72(r_ptr)
+	COPY      %r0,c1
+
+	MUL_ADD_C a7L,a7R,b3L,b3R,c2,c3,c1
+	MUL_ADD_C a6L,a6R,b4L,b4R,c2,c3,c1
+	MUL_ADD_C a5L,a5R,b5L,b5R,c2,c3,c1
+	MUL_ADD_C a4L,a4R,b6L,b6R,c2,c3,c1
+	MUL_ADD_C a3L,a3R,b7L,b7R,c2,c3,c1
+	STD       c2,80(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a4L,a4R,b7L,b7R,c3,c1,c2
+	MUL_ADD_C a5L,a5R,b6L,b6R,c3,c1,c2
+	MUL_ADD_C a6L,a6R,b5L,b5R,c3,c1,c2
+	MUL_ADD_C a7L,a7R,b4L,b4R,c3,c1,c2
+	STD       c3,88(r_ptr)
+	COPY      %r0,c3
+
+	MUL_ADD_C a7L,a7R,b5L,b5R,c1,c2,c3
+	MUL_ADD_C a6L,a6R,b6L,b6R,c1,c2,c3
+	MUL_ADD_C a5L,a5R,b7L,b7R,c1,c2,c3
+	STD       c1,96(r_ptr)
+	COPY      %r0,c1
+
+	MUL_ADD_C a6L,a6R,b7L,b7R,c2,c3,c1
+	MUL_ADD_C a7L,a7R,b6L,b6R,c2,c3,c1
+	STD       c2,104(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a7L,a7R,b7L,b7R,c3,c1,c2
+	STD       c3,112(r_ptr)
+	STD       c1,120(r_ptr)
+
+    .EXIT
+    FLDD    -88(%sp),%fr13 
+    FLDD    -96(%sp),%fr12 
+    LDD     -104(%sp),%r6        ; restore r6
+    LDD     -112(%sp),%r5        ; restore r5
+    LDD     -120(%sp),%r4        ; restore r4
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3
+
+	.PROCEND	
+
+;-----------------------------------------------------------------------------
+;
+;void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
+; arg0 = r_ptr
+; arg1 = a_ptr
+; arg2 = b_ptr
+;
+
+bn_mul_comba4
+	.proc
+	.callinfo FRAME=128,ENTRY_GR=%r3,ARGS_SAVED,ORDERING_AWARE
+	.EXPORT	bn_mul_comba4,ENTRY,PRIV_LEV=3,NO_RELOCATION,LONG_RETURN
+    .entry
+	.align 64
+
+    STD     %r3,0(%sp)          ; save r3
+    STD     %r4,8(%sp)          ; save r4
+    STD     %r5,16(%sp)         ; save r5
+    STD     %r6,24(%sp)         ; save r6
+    FSTD    %fr12,32(%sp)       ; save r6
+    FSTD    %fr13,40(%sp)       ; save r7
+
+	;
+	; Zero out carries
+	;
+	COPY     %r0,c1
+	COPY     %r0,c2
+	COPY     %r0,c3
+
+	LDO      128(%sp),%sp       ; bump stack
+    DEPDI,Z  1,31,1,high_one     ; Create Value  1 << 32
+
+	;
+	; Load up all of the values we are going to use
+	;
+    FLDD      0(a_ptr),a0       
+    FLDD      8(a_ptr),a1       
+    FLDD     16(a_ptr),a2       
+    FLDD     24(a_ptr),a3       
+
+    FLDD      0(b_ptr),b0       
+    FLDD      8(b_ptr),b1       
+    FLDD     16(b_ptr),b2       
+    FLDD     24(b_ptr),b3       
+
+	MUL_ADD_C a0L,a0R,b0L,b0R,c1,c2,c3
+	STD       c1,0(r_ptr)
+	COPY      %r0,c1
+
+	MUL_ADD_C a0L,a0R,b1L,b1R,c2,c3,c1
+	MUL_ADD_C a1L,a1R,b0L,b0R,c2,c3,c1
+	STD       c2,8(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a2L,a2R,b0L,b0R,c3,c1,c2
+	MUL_ADD_C a1L,a1R,b1L,b1R,c3,c1,c2
+	MUL_ADD_C a0L,a0R,b2L,b2R,c3,c1,c2
+	STD       c3,16(r_ptr)
+	COPY      %r0,c3
+
+	MUL_ADD_C a0L,a0R,b3L,b3R,c1,c2,c3
+	MUL_ADD_C a1L,a1R,b2L,b2R,c1,c2,c3
+	MUL_ADD_C a2L,a2R,b1L,b1R,c1,c2,c3
+	MUL_ADD_C a3L,a3R,b0L,b0R,c1,c2,c3
+	STD       c1,24(r_ptr)
+	COPY      %r0,c1
+
+	MUL_ADD_C a3L,a3R,b1L,b1R,c2,c3,c1
+	MUL_ADD_C a2L,a2R,b2L,b2R,c2,c3,c1
+	MUL_ADD_C a1L,a1R,b3L,b3R,c2,c3,c1
+	STD       c2,32(r_ptr)
+	COPY      %r0,c2
+
+	MUL_ADD_C a2L,a2R,b3L,b3R,c3,c1,c2
+	MUL_ADD_C a3L,a3R,b2L,b2R,c3,c1,c2
+	STD       c3,40(r_ptr)
+	COPY      %r0,c3
+
+	MUL_ADD_C a3L,a3R,b3L,b3R,c1,c2,c3
+	STD       c1,48(r_ptr)
+	STD       c2,56(r_ptr)
+
+    .EXIT
+    FLDD    -88(%sp),%fr13 
+    FLDD    -96(%sp),%fr12 
+    LDD     -104(%sp),%r6        ; restore r6
+    LDD     -112(%sp),%r5        ; restore r5
+    LDD     -120(%sp),%r4        ; restore r4
+    BVE     (%rp)
+    LDD,MB  -128(%sp),%r3
+
+	.PROCEND	
+
+
+	.SPACE	$TEXT$
+	.SUBSPA	$CODE$
+	.SPACE	$PRIVATE$,SORT=16
+	.IMPORT	$global$,DATA
+	.SPACE	$TEXT$
+	.SUBSPA	$CODE$
+	.SUBSPA	$LIT$,QUAD=0,ALIGN=8,ACCESS=0x2c,SORT=16
+C$7
+	.ALIGN	8
+	.STRINGZ	"Division would overflow (%d)\n"
+	.END
diff --git a/crypto/dso/dso_lib.c b/crypto/dso/dso_lib.c
index d1d50cc5e6..acd166697e 100644
--- a/crypto/dso/dso_lib.c
+++ b/crypto/dso/dso_lib.c
@@ -107,8 +107,8 @@ DSO *DSO_new_method(DSO_METHOD *meth)
 		return(NULL);
 		}
 	memset(ret, 0, sizeof(DSO));
-	ret->meth_data = sk_new(NULL);
-	if((ret->meth_data = sk_new(NULL)) == NULL)
+	ret->meth_data = sk_new_null();
+	if((ret->meth_data = sk_new_null()) == NULL)
 		{
 		/* sk_new doesn't generate any errors so we do */
 		DSOerr(DSO_F_DSO_NEW_METHOD,ERR_R_MALLOC_FAILURE);
diff --git a/crypto/opensslv.h b/crypto/opensslv.h
index c31d222831..4368f6bd51 100644
--- a/crypto/opensslv.h
+++ b/crypto/opensslv.h
@@ -25,8 +25,8 @@
  * (Prior to 0.9.5a beta1, a different scheme was used: MMNNFFRBB for
  *  major minor fix final patch/beta)
  */
-#define OPENSSL_VERSION_NUMBER	0x00906001L
-#define OPENSSL_VERSION_TEXT	"OpenSSL 0.9.6-beta1 [engine] 11 Sep 2000"
+#define OPENSSL_VERSION_NUMBER	0x00906002L
+#define OPENSSL_VERSION_TEXT	"OpenSSL 0.9.6-beta2 [engine] 17 Sep 2000"
 #define OPENSSL_VERSION_PTEXT	" part of " OPENSSL_VERSION_TEXT
 
 
diff --git a/crypto/pkcs12/p12_attr.c b/crypto/pkcs12/p12_attr.c
index b370c9cf3f..f1a210b5d2 100644
--- a/crypto/pkcs12/p12_attr.c
+++ b/crypto/pkcs12/p12_attr.c
@@ -87,13 +87,13 @@ int PKCS12_add_localkeyid (PKCS12_SAFEBAG *bag, unsigned char *name,
 		return 0;
 	}
 	attrib->object = OBJ_nid2obj(NID_localKeyID);
-	if (!(attrib->value.set = sk_ASN1_TYPE_new(NULL))) {
+	if (!(attrib->value.set = sk_ASN1_TYPE_new_null())) {
 		PKCS12err(PKCS12_F_PKCS12_ADD_LOCALKEYID, ERR_R_MALLOC_FAILURE);
 		return 0;
 	}
 	sk_ASN1_TYPE_push (attrib->value.set,keyid);
 	attrib->set = 1;
-	if (!bag->attrib && !(bag->attrib = sk_X509_ATTRIBUTE_new (NULL))) {
+	if (!bag->attrib && !(bag->attrib = sk_X509_ATTRIBUTE_new_null ())) {
 		PKCS12err(PKCS12_F_PKCS12_ADD_LOCALKEYID, ERR_R_MALLOC_FAILURE);
 		return 0;
 	}
@@ -129,14 +129,14 @@ int PKCS8_add_keyusage (PKCS8_PRIV_KEY_INFO *p8, int usage)
 		return 0;
 	}
 	attrib->object = OBJ_nid2obj(NID_key_usage);
-	if (!(attrib->value.set = sk_ASN1_TYPE_new(NULL))) {
+	if (!(attrib->value.set = sk_ASN1_TYPE_new_null())) {
 		PKCS12err(PKCS12_F_PKCS8_ADD_KEYUSAGE, ERR_R_MALLOC_FAILURE);
 		return 0;
 	}
 	sk_ASN1_TYPE_push (attrib->value.set,keyid);
 	attrib->set = 1;
 	if (!p8->attributes
-	    && !(p8->attributes = sk_X509_ATTRIBUTE_new (NULL))) {
+	    && !(p8->attributes = sk_X509_ATTRIBUTE_new_null ())) {
 		PKCS12err(PKCS12_F_PKCS8_ADD_KEYUSAGE, ERR_R_MALLOC_FAILURE);
 		return 0;
 	}
@@ -195,14 +195,14 @@ int PKCS12_add_friendlyname_uni (PKCS12_SAFEBAG *bag,
 		return 0;
 	}
 	attrib->object = OBJ_nid2obj(NID_friendlyName);
-	if (!(attrib->value.set = sk_ASN1_TYPE_new(NULL))) {
+	if (!(attrib->value.set = sk_ASN1_TYPE_new_null())) {
 		PKCS12err(PKCS12_F_PKCS12_ADD_FRIENDLYNAME,
 							ERR_R_MALLOC_FAILURE);
 		return 0;
 	}
 	sk_ASN1_TYPE_push (attrib->value.set,fname);
 	attrib->set = 1;
-	if (!bag->attrib && !(bag->attrib = sk_X509_ATTRIBUTE_new (NULL))) {
+	if (!bag->attrib && !(bag->attrib = sk_X509_ATTRIBUTE_new_null ())) {
 		PKCS12err(PKCS12_F_PKCS12_ADD_FRIENDLYNAME_UNI,
 							ERR_R_MALLOC_FAILURE);
 		return 0;
diff --git a/crypto/pkcs12/p12_crt.c b/crypto/pkcs12/p12_crt.c
index 5641a00898..a8f7b48882 100644
--- a/crypto/pkcs12/p12_crt.c
+++ b/crypto/pkcs12/p12_crt.c
@@ -88,7 +88,7 @@ PKCS12 *PKCS12_create(char *pass, char *name, EVP_PKEY *pkey, X509 *cert,
 
 	if(!X509_check_private_key(cert, pkey)) return NULL;
 
-	if(!(bags = sk_PKCS12_SAFEBAG_new (NULL))) {
+	if(!(bags = sk_PKCS12_SAFEBAG_new_null ())) {
 		PKCS12err(PKCS12_F_PKCS12_CREATE,ERR_R_MALLOC_FAILURE);
 		return NULL;
 	}
@@ -123,7 +123,7 @@ PKCS12 *PKCS12_create(char *pass, char *name, EVP_PKEY *pkey, X509 *cert,
 
 	if (!authsafe) return NULL;
 
-	if(!(safes = sk_PKCS7_new (NULL))
+	if(!(safes = sk_PKCS7_new_null ())
 	   || !sk_PKCS7_push(safes, authsafe)) {
 		PKCS12err(PKCS12_F_PKCS12_CREATE,ERR_R_MALLOC_FAILURE);
 		return NULL;
@@ -137,7 +137,7 @@ PKCS12 *PKCS12_create(char *pass, char *name, EVP_PKEY *pkey, X509 *cert,
 	PKCS8_PRIV_KEY_INFO_free(p8);
         if (name && !PKCS12_add_friendlyname (bag, name, -1)) return NULL;
 	if(!PKCS12_add_localkeyid (bag, keyid, keyidlen)) return NULL;
-	if(!(bags = sk_PKCS12_SAFEBAG_new(NULL))
+	if(!(bags = sk_PKCS12_SAFEBAG_new_null())
 	   || !sk_PKCS12_SAFEBAG_push (bags, bag)) {
 		PKCS12err(PKCS12_F_PKCS12_CREATE,ERR_R_MALLOC_FAILURE);
 		return NULL;
diff --git a/crypto/pkcs12/p12_npas.c b/crypto/pkcs12/p12_npas.c
index eed494a3f3..84e31a7f21 100644
--- a/crypto/pkcs12/p12_npas.c
+++ b/crypto/pkcs12/p12_npas.c
@@ -114,7 +114,7 @@ static int newpass_p12(PKCS12 *p12, char *oldpass, char *newpass)
 	unsigned int maclen;
 
 	if (!(asafes = M_PKCS12_unpack_authsafes(p12))) return 0;
-	if(!(newsafes = sk_PKCS7_new(NULL))) return 0;
+	if(!(newsafes = sk_PKCS7_new_null())) return 0;
 	for (i = 0; i < sk_PKCS7_num (asafes); i++) {
 		p7 = sk_PKCS7_value(asafes, i);
 		bagnid = OBJ_obj2nid(p7->type);
diff --git a/crypto/pkcs7/pk7_doit.c b/crypto/pkcs7/pk7_doit.c
index 2768247818..099e9651c1 100644
--- a/crypto/pkcs7/pk7_doit.c
+++ b/crypto/pkcs7/pk7_doit.c
@@ -909,7 +909,7 @@ static int add_attribute(STACK_OF(X509_ATTRIBUTE) **sk, int nid, int atrtype,
 
 	if (*sk == NULL)
 		{
-		*sk = sk_X509_ATTRIBUTE_new(NULL);
+		*sk = sk_X509_ATTRIBUTE_new_null();
 new_attrib:
 		attr=X509_ATTRIBUTE_create(nid,atrtype,value);
 		sk_X509_ATTRIBUTE_push(*sk,attr);
diff --git a/crypto/pkcs7/pk7_mime.c b/crypto/pkcs7/pk7_mime.c
index a7b6929436..994473c0bd 100644
--- a/crypto/pkcs7/pk7_mime.c
+++ b/crypto/pkcs7/pk7_mime.c
@@ -372,7 +372,7 @@ static int multi_split(BIO *bio, char *bound, STACK_OF(BIO) **ret)
 	part = 0;
 	state = 0;
 	first = 1;
-	parts = sk_BIO_new(NULL);
+	parts = sk_BIO_new_null();
 	*ret = parts;
 	while ((len = BIO_gets(bio, linebuf, MAX_SMLEN)) > 0) {
 		state = mime_bound_check(linebuf, len, bound, blen);
diff --git a/crypto/pkcs7/pk7_smime.c b/crypto/pkcs7/pk7_smime.c
index c8cd5a7f73..2ececcd07e 100644
--- a/crypto/pkcs7/pk7_smime.c
+++ b/crypto/pkcs7/pk7_smime.c
@@ -111,7 +111,7 @@ PKCS7 *PKCS7_sign(X509 *signcert, EVP_PKEY *pkey, STACK_OF(X509) *certs,
 		/* Add SMIMECapabilities */
 		if(!(flags & PKCS7_NOSMIMECAP))
 		{
-		if(!(smcap = sk_X509_ALGOR_new(NULL))) {
+		if(!(smcap = sk_X509_ALGOR_new_null())) {
 			PKCS7err(PKCS7_F_PKCS7_SIGN,ERR_R_MALLOC_FAILURE);
 			return NULL;
 		}
@@ -285,7 +285,7 @@ STACK_OF(X509) *PKCS7_get0_signers(PKCS7 *p7, STACK_OF(X509) *certs, int flags)
 		PKCS7err(PKCS7_F_PKCS7_GET0_SIGNERS,PKCS7_R_WRONG_CONTENT_TYPE);
 		return NULL;
 	}
-	if(!(signers = sk_X509_new(NULL))) {
+	if(!(signers = sk_X509_new_null())) {
 		PKCS7err(PKCS7_F_PKCS7_GET0_SIGNERS,ERR_R_MALLOC_FAILURE);
 		return NULL;
 	}
diff --git a/crypto/rand/rand_win.c b/crypto/rand/rand_win.c
index e9fc37c567..5ef197341f 100644
--- a/crypto/rand/rand_win.c
+++ b/crypto/rand/rand_win.c
@@ -169,6 +169,12 @@ typedef BOOL (WINAPI *PROCESS32)(HANDLE, LPPROCESSENTRY32);
 typedef BOOL (WINAPI *THREAD32)(HANDLE, LPTHREADENTRY32);
 typedef BOOL (WINAPI *MODULE32)(HANDLE, LPMODULEENTRY32);
 
+#include <lmcons.h>
+#include <lmstats.h>
+typedef NET_API_STATUS (NET_API_FUNCTION * NETSTATGET)
+        (LMSTR, LMSTR, DWORD, DWORD, LPBYTE*);
+typedef NET_API_STATUS (NET_API_FUNCTION * NETFREE)(LPBYTE);
+
 int RAND_poll(void)
 {
 	MEMORYSTATUS m;
@@ -177,15 +183,71 @@ int RAND_poll(void)
 	DWORD w;
 	HWND h;
 
-	HMODULE advapi, kernel, user;
+	HMODULE advapi, kernel, user, netapi;
 	CRYPTACQUIRECONTEXT acquire;
 	CRYPTGENRANDOM gen;
 	CRYPTRELEASECONTEXT release;
+	NETSTATGET netstatget;
+	NETFREE netfree;
 
 	/* load functions dynamically - not available on all systems */
 	advapi = GetModuleHandle("ADVAPI32.DLL");
 	kernel = GetModuleHandle("KERNEL32.DLL");
 	user = GetModuleHandle("USER32.DLL");
+	netapi = GetModuleHandle("NETAPI32.DLL");
+
+	if (netapi)
+		{
+		netstatget = (NETSTATGET) GetProcAddress(netapi,"NetStatisticsGet");
+		netfree = (NETFREE) GetProcAddress(netapi,"NetApiBufferFree");
+		}
+
+	if (netstatget && netfree)
+		{
+		LPBYTE outbuf;
+		/* NetStatisticsGet() is a Unicode only function */
+		if (netstatget(NULL, L"LanmanWorkstation", 0, 0, &outbuf) == 0)
+			{
+			RAND_add(outbuf, sizeof(STAT_WORKSTATION_0), 0);
+			netfree(outbuf);
+			}
+		if (netstatget(NULL, L"LanmanServer", 0, 0, &outbuf) == 0)
+			{
+			RAND_add(outbuf, sizeof(STAT_SERVER_0), 0);
+			netfree(outbuf);
+			}
+		}
+
+	/* Read Performance Statistics from NT/2000 registry */
+	/* The size of the performance data can vary from call to call */
+	/* so we must guess the size of the buffer to use and increase */
+	/* its size if we get an ERROR_MORE_DATA return instead of     */
+	/* ERROR_SUCCESS.                                              */
+	{
+	LONG   rc=ERROR_MORE_DATA;
+	char * buf=NULL;
+	DWORD bufsz=0;
+	DWORD length;
+
+	while (rc == ERROR_MORE_DATA)
+		{
+		buf = realloc(buf,bufsz+8192);
+		if (!buf)
+			break;
+		bufsz += 8192;
+
+		length = bufsz;
+		rc = RegQueryValueEx(HKEY_PERFORMANCE_DATA, "Global",
+			NULL, NULL, buf, &length);
+		}
+	if (rc == ERROR_SUCCESS)
+		{
+		RAND_add(&length, sizeof(length), 0);
+		RAND_add(buf, length, 0);
+		}
+	if ( buf )
+		free(buf);
+	}
 
 	if (advapi)
 		{
@@ -255,17 +317,29 @@ int RAND_poll(void)
 			RAND_add(&h, sizeof(h), 0);
 		}
 
-#if 0
+		if (cursor)
+			{
+			/* unfortunately, its not safe to call GetCursorInfo()
+			 * on NT4 even though it exists in SP3 (or SP6) and
+			 * higher.
+			 */
+			OSVERSIONINFO osverinfo ;
+			osverinfo.dwOSVersionInfoSize = sizeof(OSVERSIONINFO) ;
+			GetVersionEx( &osverinfo ) ;
+
+			if ( osverinfo.dwPlatformId == VER_PLATFORM_WIN32_NT &&
+				osverinfo.dwMajorVersion < 5)
+				cursor = 0;
+			}
+
 		if (cursor)
 			{
 			/* cursor position */
-                        PCURSORINFO p = (PCURSORINFO) buf;
-                        p->cbSize = sizeof(CURSORINFO);
-			if (cursor(p))
-				RAND_add(p+sizeof(p->cbSize),
-					p->cbSize-sizeof(p->cbSize), 0);
+			CURSORINFO ci;
+			ci.cbSize = sizeof(CURSORINFO);
+			if (cursor(&ci))
+				RAND_add(&ci, ci.cbSize, 0);
 			}
-#endif
 
 		if (queue)
 			{
@@ -277,7 +351,7 @@ int RAND_poll(void)
 
 	/* Toolhelp32 snapshot: enumerate processes, threads, modules and heap
 	 * http://msdn.microsoft.com/library/psdk/winbase/toolhelp_5pfd.htm
-	 * (Win 9x only, not available on NT)
+	 * (Win 9x and 2000 only, not available on NT)
 	 *
 	 * This seeding method was proposed in Peter Gutmann, Software
 	 * Generation of Practically Strong Random Numbers,
diff --git a/crypto/x509v3/v3_akey.c b/crypto/x509v3/v3_akey.c
index 877ae7da7b..0889a18993 100644
--- a/crypto/x509v3/v3_akey.c
+++ b/crypto/x509v3/v3_akey.c
@@ -224,7 +224,7 @@ if((issuer && !ikeyid) || (issuer == 2)) {
 if(!(akeyid = AUTHORITY_KEYID_new())) goto err;
 
 if(isname) {
-	if(!(gens = sk_GENERAL_NAME_new(NULL)) || !(gen = GENERAL_NAME_new())
+	if(!(gens = sk_GENERAL_NAME_new_null()) || !(gen = GENERAL_NAME_new())
 		|| !sk_GENERAL_NAME_push(gens, gen)) {
 		X509V3err(X509V3_F_V2I_AUTHORITY_KEYID,ERR_R_MALLOC_FAILURE);
 		goto err;
diff --git a/crypto/x509v3/v3_alt.c b/crypto/x509v3/v3_alt.c
index 5ccd1e0e3d..733919f250 100644
--- a/crypto/x509v3/v3_alt.c
+++ b/crypto/x509v3/v3_alt.c
@@ -160,7 +160,7 @@ static STACK_OF(GENERAL_NAME) *v2i_issuer_alt(X509V3_EXT_METHOD *method,
 	STACK_OF(GENERAL_NAME) *gens = NULL;
 	CONF_VALUE *cnf;
 	int i;
-	if(!(gens = sk_GENERAL_NAME_new(NULL))) {
+	if(!(gens = sk_GENERAL_NAME_new_null())) {
 		X509V3err(X509V3_F_V2I_GENERAL_NAMES,ERR_R_MALLOC_FAILURE);
 		return NULL;
 	}
@@ -225,7 +225,7 @@ static STACK_OF(GENERAL_NAME) *v2i_subject_alt(X509V3_EXT_METHOD *method,
 	STACK_OF(GENERAL_NAME) *gens = NULL;
 	CONF_VALUE *cnf;
 	int i;
-	if(!(gens = sk_GENERAL_NAME_new(NULL))) {
+	if(!(gens = sk_GENERAL_NAME_new_null())) {
 		X509V3err(X509V3_F_V2I_GENERAL_NAMES,ERR_R_MALLOC_FAILURE);
 		return NULL;
 	}
@@ -304,7 +304,7 @@ STACK_OF(GENERAL_NAME) *v2i_GENERAL_NAMES(X509V3_EXT_METHOD *method,
 	STACK_OF(GENERAL_NAME) *gens = NULL;
 	CONF_VALUE *cnf;
 	int i;
-	if(!(gens = sk_GENERAL_NAME_new(NULL))) {
+	if(!(gens = sk_GENERAL_NAME_new_null())) {
 		X509V3err(X509V3_F_V2I_GENERAL_NAMES,ERR_R_MALLOC_FAILURE);
 		return NULL;
 	}
diff --git a/crypto/x509v3/v3_crld.c b/crypto/x509v3/v3_crld.c
index 5ca5373f39..67feea4017 100644
--- a/crypto/x509v3/v3_crld.c
+++ b/crypto/x509v3/v3_crld.c
@@ -109,7 +109,7 @@ static STACK_OF(DIST_POINT) *v2i_crld(X509V3_EXT_METHOD *method,
 	GENERAL_NAME *gen = NULL;
 	CONF_VALUE *cnf;
 	int i;
-	if(!(crld = sk_DIST_POINT_new(NULL))) goto merr;
+	if(!(crld = sk_DIST_POINT_new_null())) goto merr;
 	for(i = 0; i < sk_CONF_VALUE_num(nval); i++) {
 		DIST_POINT *point;
 		cnf = sk_CONF_VALUE_value(nval, i);
diff --git a/crypto/x509v3/v3_extku.c b/crypto/x509v3/v3_extku.c
index e039d21cbf..53ec40a027 100644
--- a/crypto/x509v3/v3_extku.c
+++ b/crypto/x509v3/v3_extku.c
@@ -129,7 +129,7 @@ ASN1_OBJECT *objtmp;
 CONF_VALUE *val;
 int i;
 
-if(!(extku = sk_ASN1_OBJECT_new(NULL))) {
+if(!(extku = sk_ASN1_OBJECT_new_null())) {
 	X509V3err(X509V3_F_V2I_EXT_KU,ERR_R_MALLOC_FAILURE);
 	return NULL;
 }
diff --git a/crypto/x509v3/v3_genn.c b/crypto/x509v3/v3_genn.c
index 2de1c4b428..d44751458e 100644
--- a/crypto/x509v3/v3_genn.c
+++ b/crypto/x509v3/v3_genn.c
@@ -220,7 +220,7 @@ void GENERAL_NAME_free(GENERAL_NAME *a)
 
 STACK_OF(GENERAL_NAME) *GENERAL_NAMES_new()
 {
-	return sk_GENERAL_NAME_new(NULL);
+	return sk_GENERAL_NAME_new_null();
 }
 
 void GENERAL_NAMES_free(STACK_OF(GENERAL_NAME) *a)
diff --git a/crypto/x509v3/v3_info.c b/crypto/x509v3/v3_info.c
index 9ada0b310b..a045a629ee 100644
--- a/crypto/x509v3/v3_info.c
+++ b/crypto/x509v3/v3_info.c
@@ -119,7 +119,7 @@ static STACK_OF(ACCESS_DESCRIPTION) *v2i_AUTHORITY_INFO_ACCESS(X509V3_EXT_METHOD
 	ACCESS_DESCRIPTION *acc;
 	int i, objlen;
 	char *objtmp, *ptmp;
-	if(!(ainfo = sk_ACCESS_DESCRIPTION_new(NULL))) {
+	if(!(ainfo = sk_ACCESS_DESCRIPTION_new_null())) {
 		X509V3err(X509V3_F_V2I_ACCESS_DESCRIPTION,ERR_R_MALLOC_FAILURE);
 		return NULL;
 	}
@@ -209,7 +209,7 @@ void ACCESS_DESCRIPTION_free(ACCESS_DESCRIPTION *a)
 
 STACK_OF(ACCESS_DESCRIPTION) *AUTHORITY_INFO_ACCESS_new(void)
 {
-	return sk_ACCESS_DESCRIPTION_new(NULL);
+	return sk_ACCESS_DESCRIPTION_new_null();
 }
 
 void AUTHORITY_INFO_ACCESS_free(STACK_OF(ACCESS_DESCRIPTION) *a)
diff --git a/crypto/x509v3/v3_utl.c b/crypto/x509v3/v3_utl.c
index e6e99b2d5c..619f161b58 100644
--- a/crypto/x509v3/v3_utl.c
+++ b/crypto/x509v3/v3_utl.c
@@ -80,7 +80,7 @@ int X509V3_add_value(const char *name, const char *value,
 	if(name && !(tname = BUF_strdup(name))) goto err;
 	if(value && !(tvalue = BUF_strdup(value))) goto err;;
 	if(!(vtmp = (CONF_VALUE *)OPENSSL_malloc(sizeof(CONF_VALUE)))) goto err;
-	if(!*extlist && !(*extlist = sk_CONF_VALUE_new(NULL))) goto err;
+	if(!*extlist && !(*extlist = sk_CONF_VALUE_new_null())) goto err;
 	vtmp->section = NULL;
 	vtmp->name = tname;
 	vtmp->value = tvalue;
diff --git a/doc/crypto/BIO_s_fd.pod b/doc/crypto/BIO_s_fd.pod
index 61789fd2c1..b1de1d1015 100644
--- a/doc/crypto/BIO_s_fd.pod
+++ b/doc/crypto/BIO_s_fd.pod
@@ -20,6 +20,20 @@ BIO_s_fd, BIO_set_fd, BIO_get_fd, BIO_new_fd - file descriptor BIO
 BIO_s_fd() returns the file descriptor BIO method. This is a wrapper
 round the platforms file descriptor routines such as read() and write().
 
+BIO_read() and BIO_write() read or write the underlying descriptor.
+BIO_puts() is supported but BIO_gets() is not.
+
+If the close flag is set then then close() is called on the underlying
+file descriptor when the BIO is freed.
+
+BIO_reset() attempts to change the file pointer to the start of file
+using lseek(fd, 0, 0).
+
+BIO_seek() sets the file pointer to position B<ofs> from start of file
+using lseek(fd, ofs, 0).
+
+BIO_tell() returns the current file position by calling lseek(fd, 0, 1).
+
 BIO_set_fd() sets the file descriptor of BIO B<b> to B<fd> and the close
 flag to B<c>.
 
@@ -31,15 +45,6 @@ BIO_new_fd() returns a file descriptor BIO using B<fd> and B<close_flag>.
 
 =head1 NOTES
 
-If the close flag is set then then close() is called on the underlying
-file descriptor when the BIO is freed.
-
-BIO_reset() attempts to change the file pointer to the start of file
-using lseek(fd, 0, 0).
-
-BIO_read() and BIO_write() read or write the underlying descriptor.
-BIO_puts() is supported but BIO_gets() is not.
-
 The behaviour of BIO_read() and BIO_write() depends on the behavior of the
 platforms read() and write() calls on the descriptor. If the underlying 
 file descriptor is in a non blocking mode then the BIO will behave in the
@@ -53,6 +58,11 @@ instead.
 
 BIO_s_fd() returns the file descriptor BIO method.
 
+BIO_reset() returns zero for success and -1 if an error occurred.
+BIO_seek() and BIO_tell() return the current file position or -1
+is an error occurred. These values reflect the underlying lseek()
+behaviour.
+
 BIO_set_fd() always returns 1.
 
 BIO_get_fd() returns the file descriptor or -1 if the BIO has not
diff --git a/doc/crypto/BIO_s_file.pod b/doc/crypto/BIO_s_file.pod
index 166734b0bc..d449702bff 100644
--- a/doc/crypto/BIO_s_file.pod
+++ b/doc/crypto/BIO_s_file.pod
@@ -28,6 +28,23 @@ BIO_s_file() returns the BIO file method. As its name implies it
 is a wrapper round the stdio FILE structure and it is a
 source/sink BIO.
 
+Calls to BIO_read() and BIO_write() read and write data to the
+underlying stream. BIO_gets() and BIO_puts() are supported on file BIOs.
+
+BIO_flush() on a file BIO calls the fflush() function on the wrapped
+stream.
+
+BIO_reset() attempts to change the file pointer to the start of file
+using fseek(stream, 0, 0).
+
+BIO_seek() sets the file pointer to position B<ofs> from start of file
+using lseek(stream, ofs, 0).
+
+BIO_eof() calls feof().
+
+Setting the BIO_CLOSE flag calls fclose() on the stream when the BIO
+is freed.
+
 BIO_new_file() creates a new file BIO with mode B<mode> the meaning
 of B<mode> is the same as the stdio function fopen(). The BIO_CLOSE
 flag is set on the returned BIO.
@@ -42,26 +59,17 @@ meaning as in BIO_new_fp(), it is a macro.
 
 BIO_get_fp() retrieves the fp of a file BIO, it is a macro.
 
+BIO_seek() is a macro that sets the position pointer to B<offset> bytes
+from the start of file.
+
+BIO_tell() returns the value of the position pointer.
+
 BIO_read_filename(), BIO_write_filename(), BIO_append_filename() and
 BIO_rw_filename() set the file BIO B<b> to use file B<name> for
 reading, writing, append or read write respectively.
 
 =head1 NOTES
 
-Calls to BIO_read() and BIO_write() read and write data to the
-underlying stream. BIO_gets() and BIO_puts() are supported on file BIOs.
-
-BIO_flush() on a file BIO calls the fflush() function on the wrapped
-stream.
-
-BIO_reset() on a file BIO calls fseek() to reset the position indicator
-to the start of the file.
-
-BIO_eof() calls feof().
-
-Setting the BIO_CLOSE flag calls fclose() on the stream when the BIO
-is freed.
-
 When wrapping stdout, stdin or stderr the underlying stream should not
 normally be closed so the BIO_NOCLOSE flag should be set.
 
@@ -111,6 +119,11 @@ occurred.
 BIO_set_fp() and BIO_get_fp() return 1 for success or 0 for failure
 (although the current implementation never return 0).
 
+BIO_seek() returns the same value as the underlying fseek() function:
+0 for success or -1 for failure.
+
+BIO_tell() returns the current file position.
+
 BIO_read_filename(), BIO_write_filename(),  BIO_append_filename() and
 BIO_rw_filename() return 1 for success or 0 for failure.
 
diff --git a/ssl/ssl_cert.c b/ssl/ssl_cert.c
index 7579a3432d..c26df62c20 100644
--- a/ssl/ssl_cert.c
+++ b/ssl/ssl_cert.c
@@ -589,7 +589,7 @@ STACK_OF(X509_NAME) *SSL_load_client_CA_file(const char *file)
 	X509_NAME *xn=NULL;
 	STACK_OF(X509_NAME) *ret,*sk;
 
-	ret=sk_X509_NAME_new(NULL);
+	ret=sk_X509_NAME_new_null();
 	sk=sk_X509_NAME_new(xname_cmp);
 
 	in=BIO_new(BIO_s_file_internal());
diff --git a/ssl/ssl_ciph.c b/ssl/ssl_ciph.c
index ce73213b12..f63163f26c 100644
--- a/ssl/ssl_ciph.c
+++ b/ssl/ssl_ciph.c
@@ -799,7 +799,7 @@ STACK_OF(SSL_CIPHER) *ssl_create_cipher_list(const SSL_METHOD *ssl_method,
 	 * Allocate new "cipherstack" for the result, return with error
 	 * if we cannot get one.
 	 */
-	if ((cipherstack = sk_SSL_CIPHER_new(NULL)) == NULL)
+	if ((cipherstack = sk_SSL_CIPHER_new_null()) == NULL)
 		{
 		OPENSSL_free(list);
 		return(NULL);
diff --git a/ssl/ssl_lib.c b/ssl/ssl_lib.c
index 1483fad19a..635b25062e 100644
--- a/ssl/ssl_lib.c
+++ b/ssl/ssl_lib.c
@@ -1036,7 +1036,7 @@ STACK_OF(SSL_CIPHER) *ssl_bytes_to_cipher_list(SSL *s,unsigned char *p,int num,
 		return(NULL);
 		}
 	if ((skp == NULL) || (*skp == NULL))
-		sk=sk_SSL_CIPHER_new(NULL); /* change perhaps later */
+		sk=sk_SSL_CIPHER_new_null(); /* change perhaps later */
 	else
 		{
 		sk= *skp;