Synchronize message digests in 098-fips with 098.

6 files changed, 635 insertions, 1007 deletions

diff --git a/crypto/sha/asm/sha1-586.pl b/crypto/sha/asm/sha1-586.pl
index 4f8521f1e2..0b4dab2bd5 100644
--- a/crypto/sha/asm/sha1-586.pl
+++ b/crypto/sha/asm/sha1-586.pl
@@ -1,4 +1,16 @@
-#!/usr/local/bin/perl
+#!/usr/bin/env perl
+
+# ====================================================================
+# [Re]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/.
+# ====================================================================
+
+# "[Re]written" was achieved in two major overhauls. In 2004 BODY_*
+# functions were re-implemented to address P4 performance issue [see
+# commentary below], and in 2006 the rest was rewritten in order to
+# gain freedom to liberate licensing terms.
 
 # It was noted that Intel IA-32 C compiler generates code which
 # performs ~30% *faster* on P4 CPU than original *hand-coded*
@@ -17,90 +29,27 @@
 # improvement on P4 outweights the loss and incorporate this
 # re-tuned code to 0.9.7 and later.
 # ----------------------------------------------------------------
-# Those who for any particular reason absolutely must score on
-# Pentium can replace this module with one from 0.9.6 distribution.
-# This "offer" shall be revoked the moment programming interface to
-# this module is changed, in which case this paragraph should be
-# removed.
-# ----------------------------------------------------------------
 #					<appro@fy.chalmers.se>
 
-$normal=0;
-
-push(@INC,"perlasm","../../perlasm");
+$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
+push(@INC,"${dir}","${dir}../../perlasm");
 require "x86asm.pl";
 
 &asm_init($ARGV[0],"sha1-586.pl",$ARGV[$#ARGV] eq "386");
 
 $A="eax";
-$B="ecx";
-$C="ebx";
+$B="ebx";
+$C="ecx";
 $D="edx";
 $E="edi";
 $T="esi";
 $tmp1="ebp";
 
-$off=9*4;
-
-@K=(0x5a827999,0x6ed9eba1,0x8f1bbcdc,0xca62c1d6);
-
-&sha1_block_data("sha1_block_asm_data_order");
-
-&asm_finish();
-
-sub Nn
-	{
-	local($p)=@_;
-	local(%n)=($A,$T,$B,$A,$C,$B,$D,$C,$E,$D,$T,$E);
-	return($n{$p});
-	}
-
-sub Np
-	{
-	local($p)=@_;
-	local(%n)=($A,$T,$B,$A,$C,$B,$D,$C,$E,$D,$T,$E);
-	local(%n)=($A,$B,$B,$C,$C,$D,$D,$E,$E,$T,$T,$A);
-	return($n{$p});
-	}
-
-sub Na
-	{
-	local($n)=@_;
-	return( (($n   )&0x0f),
-		(($n+ 2)&0x0f),
-		(($n+ 8)&0x0f),
-		(($n+13)&0x0f),
-		(($n+ 1)&0x0f));
-	}
-
-sub X_expand
-	{
-	local($in)=@_;
-
-	&comment("First, load the words onto the stack in network byte order");
-	for ($i=0; $i<16; $i+=2)
-		{
-		&mov($A,&DWP(($i+0)*4,$in,"",0));# unless $i == 0;
-		 &mov($B,&DWP(($i+1)*4,$in,"",0));
-		&bswap($A);
-		 &bswap($B);
-		&mov(&swtmp($i+0),$A);
-		 &mov(&swtmp($i+1),$B);
-		}
-
-	&comment("We now have the X array on the stack");
-	&comment("starting at sp-4");
-	}
-
-# Rules of engagement
-# F is always trashable at the start, the running total.
-# E becomes the next F so it can be trashed after it has been 'accumulated'
-# F becomes A in the next round.  We don't need to access it much.
-# During the X update part, the result ends up in $X[$n0].
+@V=($A,$B,$C,$D,$E,$T);
 
 sub BODY_00_15
 	{
-	local($pos,$K,$X,$n,$a,$b,$c,$d,$e,$f)=@_;
+	local($n,$a,$b,$c,$d,$e,$f)=@_;
 
 	&comment("00_15 $n");
 
@@ -109,37 +58,37 @@ sub BODY_00_15
 	 else        { &mov($a,$tmp1); }
 	&rotl($tmp1,5);			# tmp1=ROTATE(a,5)
 	 &xor($f,$d);
-	&and($f,$b);
-	 &add($tmp1,$e);		# tmp1+=e;
-	&mov($e,&swtmp($n));		# e becomes volatile and
-	 				# is loaded with xi
+	&add($tmp1,$e);			# tmp1+=e;
+	 &and($f,$b);
+	&mov($e,&swtmp($n%16));		# e becomes volatile and is loaded
+	 				# with xi, also note that e becomes
+					# f in next round...
 	 &xor($f,$d);			# f holds F_00_19(b,c,d)
 	&rotr($b,2);			# b=ROTATE(b,30)
-	 &lea($tmp1,&DWP($K,$tmp1,$e,1));# tmp1+=K_00_19+xi
+	 &lea($tmp1,&DWP(0x5a827999,$tmp1,$e));	# tmp1+=K_00_19+xi
 
 	if ($n==15) { &add($f,$tmp1); }	# f+=tmp1
-	else        { &add($tmp1,$f); }
+	else        { &add($tmp1,$f); }	# f becomes a in next round
 	}
 
 sub BODY_16_19
 	{
-	local($pos,$K,$X,$n,$a,$b,$c,$d,$e,$f)=@_;
-	local($n0,$n1,$n2,$n3,$np)=&Na($n);
+	local($n,$a,$b,$c,$d,$e,$f)=@_;
 
 	&comment("16_19 $n");
 
-	&mov($f,&swtmp($n1));		# f to hold Xupdate(xi,xa,xb,xc,xd)
+	&mov($f,&swtmp($n%16));		# f to hold Xupdate(xi,xa,xb,xc,xd)
 	 &mov($tmp1,$c);		# tmp1 to hold F_00_19(b,c,d)
-	&xor($f,&swtmp($n0));
+	&xor($f,&swtmp(($n+2)%16));
 	 &xor($tmp1,$d);
-	&xor($f,&swtmp($n2));
+	&xor($f,&swtmp(($n+8)%16));
 	 &and($tmp1,$b);		# tmp1 holds F_00_19(b,c,d)
 	&rotr($b,2);			# b=ROTATE(b,30)
-	 &xor($f,&swtmp($n3));		# f holds xa^xb^xc^xd
-	&rotl($f,1);			# f=ROATE(f,1)
+	 &xor($f,&swtmp(($n+13)%16));	# f holds xa^xb^xc^xd
+	&rotl($f,1);			# f=ROTATE(f,1)
 	 &xor($tmp1,$d);		# tmp1=F_00_19(b,c,d)
-	&mov(&swtmp($n0),$f);		# xi=f
-	&lea($f,&DWP($K,$f,$e,1));	# f+=K_00_19+e
+	&mov(&swtmp($n%16),$f);		# xi=f
+	&lea($f,&DWP(0x5a827999,$f,$e));# f+=K_00_19+e
 	 &mov($e,$a);			# e becomes volatile
 	&rotl($e,5);			# e=ROTATE(a,5)
 	 &add($f,$tmp1);		# f+=F_00_19(b,c,d)
@@ -148,48 +97,47 @@ sub BODY_16_19
 
 sub BODY_20_39
 	{
-	local($pos,$K,$X,$n,$a,$b,$c,$d,$e,$f)=@_;
+	local($n,$a,$b,$c,$d,$e,$f)=@_;
+	local $K=($n<40)?0x6ed9eba1:0xca62c1d6;
 
 	&comment("20_39 $n");
-	local($n0,$n1,$n2,$n3,$np)=&Na($n);
 
 	&mov($tmp1,$b);			# tmp1 to hold F_20_39(b,c,d)
-	 &mov($f,&swtmp($n0));		# f to hold Xupdate(xi,xa,xb,xc,xd)
+	 &mov($f,&swtmp($n%16));	# f to hold Xupdate(xi,xa,xb,xc,xd)
 	&rotr($b,2);			# b=ROTATE(b,30)
-	 &xor($f,&swtmp($n1));
+	 &xor($f,&swtmp(($n+2)%16));
 	&xor($tmp1,$c);
-	 &xor($f,&swtmp($n2));
+	 &xor($f,&swtmp(($n+8)%16));
 	&xor($tmp1,$d);			# tmp1 holds F_20_39(b,c,d)
-	 &xor($f,&swtmp($n3));		# f holds xa^xb^xc^xd
+	 &xor($f,&swtmp(($n+13)%16));	# f holds xa^xb^xc^xd
 	&rotl($f,1);			# f=ROTATE(f,1)
 	 &add($tmp1,$e);
-	&mov(&swtmp($n0),$f);		# xi=f
+	&mov(&swtmp($n%16),$f);		# xi=f
 	 &mov($e,$a);			# e becomes volatile
 	&rotl($e,5);			# e=ROTATE(a,5)
-	 &lea($f,&DWP($K,$f,$tmp1,1));	# f+=K_20_39+e
+	 &lea($f,&DWP($K,$f,$tmp1));	# f+=K_20_39+e
 	&add($f,$e);			# f+=ROTATE(a,5)
 	}
 
 sub BODY_40_59
 	{
-	local($pos,$K,$X,$n,$a,$b,$c,$d,$e,$f)=@_;
+	local($n,$a,$b,$c,$d,$e,$f)=@_;
 
 	&comment("40_59 $n");
-	local($n0,$n1,$n2,$n3,$np)=&Na($n);
 
-	&mov($f,&swtmp($n0));		# f to hold Xupdate(xi,xa,xb,xc,xd)
-	 &mov($tmp1,&swtmp($n1));
+	&mov($f,&swtmp($n%16));		# f to hold Xupdate(xi,xa,xb,xc,xd)
+	 &mov($tmp1,&swtmp(($n+2)%16));
 	&xor($f,$tmp1);
-	 &mov($tmp1,&swtmp($n2));
+	 &mov($tmp1,&swtmp(($n+8)%16));
 	&xor($f,$tmp1);
-	 &mov($tmp1,&swtmp($n3));
+	 &mov($tmp1,&swtmp(($n+13)%16));
 	&xor($f,$tmp1);			# f holds xa^xb^xc^xd
 	 &mov($tmp1,$b);		# tmp1 to hold F_40_59(b,c,d)
 	&rotl($f,1);			# f=ROTATE(f,1)
 	 &or($tmp1,$c);
-	&mov(&swtmp($n0),$f);		# xi=f
+	&mov(&swtmp($n%16),$f);		# xi=f
 	 &and($tmp1,$d);
-	&lea($f,&DWP($K,$f,$e,1));	# f+=K_40_59+e
+	&lea($f,&DWP(0x8f1bbcdc,$f,$e));# f+=K_40_59+e
 	 &mov($e,$b);			# e becomes volatile and is used
 					# to calculate F_40_59(b,c,d)
 	&rotr($b,2);			# b=ROTATE(b,30)
@@ -201,230 +149,71 @@ sub BODY_40_59
 	&add($f,$e);			# f+=ROTATE(a,5)
 	}
 
-sub BODY_60_79
-	{
-	&BODY_20_39(@_);
-	}
-
-sub sha1_block_host
-	{
-	local($name, $sclabel)=@_;
-
-	&function_begin_B($name,"");
-
-	# parameter 1 is the MD5_CTX structure.
-	# A	0
-	# B	4
-	# C	8
-	# D 	12
-	# E 	16
-
-	&mov("ecx",	&wparam(2));
-	 &push("esi");
-	&shl("ecx",6);
-	 &mov("esi",	&wparam(1));
-	&push("ebp");
-	 &add("ecx","esi");	# offset to leave on
-	&push("ebx");
-	 &mov("ebp",	&wparam(0));
-	&push("edi");
-	 &mov($D,	&DWP(12,"ebp","",0));
-	&stack_push(18+9);
-	 &mov($E,	&DWP(16,"ebp","",0));
-	&mov($C,	&DWP( 8,"ebp","",0));
-	 &mov(&swtmp(17),"ecx");
+&function_begin("sha1_block_data_order",16);
+	&mov($tmp1,&wparam(0));	# SHA_CTX *c
+	&mov($T,&wparam(1));	# const void *input
+	&mov($A,&wparam(2));	# size_t num
+	&stack_push(16);	# allocate X[16]
+	&shl($A,6);
+	&add($A,$T);
+	&mov(&wparam(2),$A);	# pointer beyond the end of input
+	&mov($E,&DWP(16,$tmp1));# pre-load E
 
-	&comment("First we need to setup the X array");
+	&set_label("loop",16);
 
-	for ($i=0; $i<16; $i+=2)
+	# copy input chunk to X, but reversing byte order!
+	for ($i=0; $i<16; $i+=4)
 		{
-		&mov($A,&DWP(($i+0)*4,"esi","",0));# unless $i == 0;
-		 &mov($B,&DWP(($i+1)*4,"esi","",0));
+		&mov($A,&DWP(4*($i+0),$T));
+		&mov($B,&DWP(4*($i+1),$T));
+		&mov($C,&DWP(4*($i+2),$T));
+		&mov($D,&DWP(4*($i+3),$T));
+		&bswap($A);
+		&bswap($B);
+		&bswap($C);
+		&bswap($D);
 		&mov(&swtmp($i+0),$A);
-		 &mov(&swtmp($i+1),$B);
+		&mov(&swtmp($i+1),$B);
+		&mov(&swtmp($i+2),$C);
+		&mov(&swtmp($i+3),$D);
 		}
-	&jmp($sclabel);
-	&function_end_B($name);
-	}
-
-
-sub sha1_block_data
-	{
-	local($name)=@_;
-
-	&function_begin_B($name,"");
-
-	# parameter 1 is the MD5_CTX structure.
-	# A	0
-	# B	4
-	# C	8
-	# D 	12
-	# E 	16
-
-	&mov("ecx",	&wparam(2));
-	 &push("esi");
-	&shl("ecx",6);
-	 &mov("esi",	&wparam(1));
-	&push("ebp");
-	 &add("ecx","esi");	# offset to leave on
-	&push("ebx");
-	 &mov("ebp",	&wparam(0));
-	&push("edi");
-	 &mov($D,	&DWP(12,"ebp","",0));
-	&stack_push(18+9);
-	 &mov($E,	&DWP(16,"ebp","",0));
-	&mov($C,	&DWP( 8,"ebp","",0));
-	 &mov(&swtmp(17),"ecx");
-
-	&comment("First we need to setup the X array");
-
-	&set_label("start") unless $normal;
-
-	&X_expand("esi");
-	 &mov(&wparam(1),"esi");
-
-	&set_label("shortcut", 0, 1);
-	&comment("");
-	&comment("Start processing");
-
-	# odd start
-	&mov($A,	&DWP( 0,"ebp","",0));
-	 &mov($B,	&DWP( 4,"ebp","",0));
-	$X="esp";
-	&BODY_00_15(-2,$K[0],$X, 0,$A,$B,$C,$D,$E,$T);
-	&BODY_00_15( 0,$K[0],$X, 1,$T,$A,$B,$C,$D,$E);
-	&BODY_00_15( 0,$K[0],$X, 2,$E,$T,$A,$B,$C,$D);
-	&BODY_00_15( 0,$K[0],$X, 3,$D,$E,$T,$A,$B,$C);
-	&BODY_00_15( 0,$K[0],$X, 4,$C,$D,$E,$T,$A,$B);
-	&BODY_00_15( 0,$K[0],$X, 5,$B,$C,$D,$E,$T,$A);
-	&BODY_00_15( 0,$K[0],$X, 6,$A,$B,$C,$D,$E,$T);
-	&BODY_00_15( 0,$K[0],$X, 7,$T,$A,$B,$C,$D,$E);
-	&BODY_00_15( 0,$K[0],$X, 8,$E,$T,$A,$B,$C,$D);
-	&BODY_00_15( 0,$K[0],$X, 9,$D,$E,$T,$A,$B,$C);
-	&BODY_00_15( 0,$K[0],$X,10,$C,$D,$E,$T,$A,$B);
-	&BODY_00_15( 0,$K[0],$X,11,$B,$C,$D,$E,$T,$A);
-	&BODY_00_15( 0,$K[0],$X,12,$A,$B,$C,$D,$E,$T);
-	&BODY_00_15( 0,$K[0],$X,13,$T,$A,$B,$C,$D,$E);
-	&BODY_00_15( 0,$K[0],$X,14,$E,$T,$A,$B,$C,$D);
-	&BODY_00_15( 1,$K[0],$X,15,$D,$E,$T,$A,$B,$C);
-	&BODY_16_19(-1,$K[0],$X,16,$C,$D,$E,$T,$A,$B);
-	&BODY_16_19( 0,$K[0],$X,17,$B,$C,$D,$E,$T,$A);
-	&BODY_16_19( 0,$K[0],$X,18,$A,$B,$C,$D,$E,$T);
-	&BODY_16_19( 1,$K[0],$X,19,$T,$A,$B,$C,$D,$E);
-
-	&BODY_20_39(-1,$K[1],$X,20,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39( 0,$K[1],$X,21,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39( 0,$K[1],$X,22,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39( 0,$K[1],$X,23,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39( 0,$K[1],$X,24,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39( 0,$K[1],$X,25,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39( 0,$K[1],$X,26,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39( 0,$K[1],$X,27,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39( 0,$K[1],$X,28,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39( 0,$K[1],$X,29,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39( 0,$K[1],$X,30,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39( 0,$K[1],$X,31,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39( 0,$K[1],$X,32,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39( 0,$K[1],$X,33,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39( 0,$K[1],$X,34,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39( 0,$K[1],$X,35,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39( 0,$K[1],$X,36,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39( 0,$K[1],$X,37,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39( 0,$K[1],$X,38,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39( 1,$K[1],$X,39,$D,$E,$T,$A,$B,$C);
-
-	&BODY_40_59(-1,$K[2],$X,40,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59( 0,$K[2],$X,41,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59( 0,$K[2],$X,42,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59( 0,$K[2],$X,43,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59( 0,$K[2],$X,44,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59( 0,$K[2],$X,45,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59( 0,$K[2],$X,46,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59( 0,$K[2],$X,47,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59( 0,$K[2],$X,48,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59( 0,$K[2],$X,49,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59( 0,$K[2],$X,50,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59( 0,$K[2],$X,51,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59( 0,$K[2],$X,52,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59( 0,$K[2],$X,53,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59( 0,$K[2],$X,54,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59( 0,$K[2],$X,55,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59( 0,$K[2],$X,56,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59( 0,$K[2],$X,57,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59( 0,$K[2],$X,58,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59( 1,$K[2],$X,59,$B,$C,$D,$E,$T,$A);
-
-	&BODY_60_79(-1,$K[3],$X,60,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79( 0,$K[3],$X,61,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79( 0,$K[3],$X,62,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79( 0,$K[3],$X,63,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79( 0,$K[3],$X,64,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79( 0,$K[3],$X,65,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79( 0,$K[3],$X,66,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79( 0,$K[3],$X,67,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79( 0,$K[3],$X,68,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79( 0,$K[3],$X,69,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79( 0,$K[3],$X,70,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79( 0,$K[3],$X,71,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79( 0,$K[3],$X,72,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79( 0,$K[3],$X,73,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79( 0,$K[3],$X,74,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79( 0,$K[3],$X,75,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79( 0,$K[3],$X,76,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79( 0,$K[3],$X,77,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79( 0,$K[3],$X,78,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79( 2,$K[3],$X,79,$T,$A,$B,$C,$D,$E);
-
-	&comment("End processing");
-	&comment("");
-	# D is the tmp value
-
-	# E -> A
-	# T -> B
-	# A -> C
-	# B -> D
-	# C -> E
-	# D -> T
-
-	&mov($tmp1,&wparam(0));
-
-	 &mov($D,	&DWP(12,$tmp1,"",0));
-	&add($D,$B);
-	 &mov($B,	&DWP( 4,$tmp1,"",0));
-	&add($B,$T);
-	 &mov($T,	$A);
-	&mov($A,	&DWP( 0,$tmp1,"",0));
-	 &mov(&DWP(12,$tmp1,"",0),$D);
-
-	&add($A,$E);
-	 &mov($E,	&DWP(16,$tmp1,"",0));
-	&add($E,$C);
-	 &mov($C,	&DWP( 8,$tmp1,"",0));
-	&add($C,$T);
-
-	 &mov(&DWP( 0,$tmp1,"",0),$A);
-	&mov("esi",&wparam(1));
-	 &mov(&DWP( 8,$tmp1,"",0),$C);
- 	&add("esi",64);
-	 &mov("eax",&swtmp(17));
-	&mov(&DWP(16,$tmp1,"",0),$E);
-	 &cmp("esi","eax");
-	&mov(&DWP( 4,$tmp1,"",0),$B);
-	 &jb(&label("start"));
-
-	&stack_pop(18+9);
-	 &pop("edi");
-	&pop("ebx");
-	 &pop("ebp");
-	&pop("esi");
-	 &ret();
-
-	# keep a note of shortcut label so it can be used outside
-	# block.
-	my $sclabel = &label("shortcut");
-
-	&function_end_B($name);
-	# Putting this here avoids problems with MASM in debugging mode
-	&sha1_block_host("sha1_block_asm_host_order", $sclabel);
-	}
+	&mov(&wparam(1),$T);	# redundant in 1st spin
+
+	&mov($A,&DWP(0,$tmp1));	# load SHA_CTX
+	&mov($B,&DWP(4,$tmp1));
+	&mov($C,&DWP(8,$tmp1));
+	&mov($D,&DWP(12,$tmp1));
+	# E is pre-loaded
+
+	for($i=0;$i<16;$i++)	{ &BODY_00_15($i,@V); unshift(@V,pop(@V)); }
+	for(;$i<20;$i++)	{ &BODY_16_19($i,@V); unshift(@V,pop(@V)); }
+	for(;$i<40;$i++)	{ &BODY_20_39($i,@V); unshift(@V,pop(@V)); }
+	for(;$i<60;$i++)	{ &BODY_40_59($i,@V); unshift(@V,pop(@V)); }
+	for(;$i<80;$i++)	{ &BODY_20_39($i,@V); unshift(@V,pop(@V)); }
+
+	(($V[5] eq $D) and ($V[0] eq $E)) or die;	# double-check
+
+	&mov($tmp1,&wparam(0));	# re-load SHA_CTX*
+	&mov($D,&wparam(1));	# D is last "T" and is discarded
+
+	&add($E,&DWP(0,$tmp1));	# E is last "A"...
+	&add($T,&DWP(4,$tmp1));
+	&add($A,&DWP(8,$tmp1));
+	&add($B,&DWP(12,$tmp1));
+	&add($C,&DWP(16,$tmp1));
+
+	&mov(&DWP(0,$tmp1),$E);	# update SHA_CTX
+	 &add($D,64);		# advance input pointer
+	&mov(&DWP(4,$tmp1),$T);
+	 &cmp($D,&wparam(2));	# have we reached the end yet?
+	&mov(&DWP(8,$tmp1),$A);
+	 &mov($E,$C);		# C is last "E" which needs to be "pre-loaded"
+	&mov(&DWP(12,$tmp1),$B);
+	 &mov($T,$D);		# input pointer
+	&mov(&DWP(16,$tmp1),$C);
+	&jb(&label("loop"));
+
+	&stack_pop(16);
+&function_end("sha1_block_data_order");
 
+&asm_finish();
diff --git a/crypto/sha/asm/sha1-ia64.pl b/crypto/sha/asm/sha1-ia64.pl
index 9478f5dd5d..aa18c1089b 100644
--- a/crypto/sha/asm/sha1-ia64.pl
+++ b/crypto/sha/asm/sha1-ia64.pl
@@ -2,8 +2,9 @@
 #
 # ====================================================================
 # Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
-# project. Rights for redistribution and usage in source and binary
-# forms are granted according to the OpenSSL license.
+# 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/.
 # ====================================================================
 #
 # Eternal question is what's wrong with compiler generated code? The
@@ -11,15 +12,10 @@
 # to perform rotations by maintaining copy of 32-bit value in upper
 # bits of 64-bit register. Just follow mux2 and shrp instructions...
 # Performance under big-endian OS such as HP-UX is 179MBps*1GHz, which
-# is >50% better than HP C and >2x better than gcc. As of this moment
-# performance under little-endian OS such as Linux and Windows will be
-# a bit lower, because data has to be picked in reverse byte-order.
-# It's possible to resolve this issue by implementing third function,
-# sha1_block_asm_data_order_aligned, which would temporarily flip
-# BE field in User Mask register...
+# is >50% better than HP C and >2x better than gcc.
 
 $code=<<___;
-.ident  \"sha1-ia64.s, version 1.0\"
+.ident  \"sha1-ia64.s, version 1.2\"
 .ident  \"IA-64 ISA artwork by Andy Polyakov <appro\@fy.chalmers.se>\"
 .explicit
 
@@ -55,63 +51,55 @@ else {
 
 sub BODY_00_15 {
 local	*code=shift;
-local	($i,$a,$b,$c,$d,$e,$f,$unaligned)=@_;
+local	($i,$a,$b,$c,$d,$e,$f)=@_;
 
-if ($unaligned) {
-	$code.=<<___;
-{ .mmi;	ld1	tmp0=[inp],2		    // MSB
-	ld1	tmp1=[tmp3],2		};;
-{ .mmi;	ld1	tmp2=[inp],2
-	ld1	$X[$i&0xf]=[tmp3],2	    // LSB
-	dep	tmp1=tmp0,tmp1,8,8	};;
-{ .mii;	cmp.ne	p16,p0=r0,r0		    // no misaligned prefetch
-	dep	$X[$i&0xf]=tmp2,$X[$i&0xf],8,8;;
-	dep	$X[$i&0xf]=tmp1,$X[$i&0xf],16,16	};;
-{ .mmi;	nop.m	0
-___
-	}
-elsif ($i<15) {
-	$code.=<<___;
-{ .mmi;	ld4	$X[($i+1)&0xf]=[inp],4	// prefetch
-___
-	}
-else	{
-	$code.=<<___;
-{ .mmi;	nop.m	0
+$code.=<<___ if ($i==0);
+{ .mmi;	ld1	$X[$i&0xf]=[inp],2	    // MSB
+	ld1	tmp2=[tmp3],2		};;
+{ .mmi;	ld1	tmp0=[inp],2
+	ld1	tmp4=[tmp3],2		    // LSB
+	dep	$X[$i&0xf]=$X[$i&0xf],tmp2,8,8	};;
 ___
-	}
 if ($i<15) {
 	$code.=<<___;
-	and	tmp0=$c,$b
-	dep.z	tmp5=$a,5,27		}   // a<<5
+{ .mmi;	ld1	$X[($i+1)&0xf]=[inp],2	    // +1
+	dep	tmp1=tmp0,tmp4,8,8	};;
+{ .mmi;	ld1	tmp2=[tmp3],2		    // +1
+	and	tmp4=$c,$b
+	dep	$X[$i&0xf]=$X[$i&0xf],tmp1,16,16	} //;;
 { .mmi;	andcm	tmp1=$d,$b
-	add	tmp4=$e,$K_00_19	};;
-{ .mmi;	or	tmp0=tmp0,tmp1		    // F_00_19(b,c,d)=(b&c)|(~b&d)
-	add	$f=tmp4,$X[$i&0xf]	    // f=xi+e+K_00_19
+	add	tmp0=$e,$K_00_19
+	dep.z	tmp5=$a,5,27		};; // a<<5
+{ .mmi;	or	tmp4=tmp4,tmp1		    // F_00_19(b,c,d)=(b&c)|(~b&d)
+	add	$f=tmp0,$X[$i&0xf]	    // f=xi+e+K_00_19
 	extr.u	tmp1=$a,27,5		};; // a>>27
-{ .mib;	add	$f=$f,tmp0		    // f+=F_00_19(b,c,d)
+{ .mmi;	ld1	tmp0=[inp],2		    // +1
+	add	$f=$f,tmp4		    // f+=F_00_19(b,c,d)
 	shrp	$b=tmp6,tmp6,2		}   // b=ROTATE(b,30)
-{ .mib;	or	tmp1=tmp1,tmp5		    // ROTATE(a,5)
+{ .mmi;	ld1	tmp4=[tmp3],2		    // +1
+	or	tmp5=tmp1,tmp5		    // ROTATE(a,5)
 	mux2	tmp6=$a,0x44		};; // see b in next iteration
-{ .mii;	add	$f=$f,tmp1		    // f+=ROTATE(a,5)
-	mux2	$X[$i&0xf]=$X[$i&0xf],0x44
-	nop.i	0			};;
+{ .mii;	add	$f=$f,tmp5		    // f+=ROTATE(a,5)
+	dep	$X[($i+1)&0xf]=$X[($i+1)&0xf],tmp2,8,8	// +1
+	mux2	$X[$i&0xf]=$X[$i&0xf],0x44	} //;;
 
 ___
 	}
 else	{
 	$code.=<<___;
-	and	tmp0=$c,$b
-	dep.z	tmp5=$a,5,27		}   // a<<5 ;;?
+{ .mii;	and	tmp3=$c,$b
+	dep	tmp1=tmp0,tmp4,8,8;;
+	dep	$X[$i&0xf]=$X[$i&0xf],tmp1,16,16	} //;;
 { .mmi;	andcm	tmp1=$d,$b
-	add	tmp4=$e,$K_00_19	};;
-{ .mmi;	or	tmp0=tmp0,tmp1		    // F_00_19(b,c,d)=(b&c)|(~b&d)
-	add	$f=tmp4,$X[$i&0xf]	    // f=xi+e+K_00_19
+	add	tmp0=$e,$K_00_19
+	dep.z	tmp5=$a,5,27		};; // a<<5
+{ .mmi;	or	tmp4=tmp3,tmp1		    // F_00_19(b,c,d)=(b&c)|(~b&d)
+	add	$f=tmp0,$X[$i&0xf]	    // f=xi+e+K_00_19
 	extr.u	tmp1=$a,27,5		}   // a>>27
 { .mmi;	xor	tmp2=$X[($i+0+1)&0xf],$X[($i+2+1)&0xf]	// +1
 	xor	tmp3=$X[($i+8+1)&0xf],$X[($i+13+1)&0xf] // +1
 	nop.i	0			};;
-{ .mmi;	add	$f=$f,tmp0		    // f+=F_00_19(b,c,d)
+{ .mmi;	add	$f=$f,tmp4		    // f+=F_00_19(b,c,d)
 	xor	tmp2=tmp2,tmp3		    // +1
 	shrp	$b=tmp6,tmp6,2		}   // b=ROTATE(b,30)
 { .mmi; or	tmp1=tmp1,tmp5		    // ROTATE(a,5)
@@ -190,9 +178,7 @@ $code.=<<___;
 	extr.u	tmp1=$a,27,5		}   // a>>27
 { .mib;	add	$f=$f,tmp4		    // f+=e+K_20_39
 	add	$h1=$h1,$a		};; // wrap up
-{ .mmi;
-(p16)	ld4.s	$X[0]=[inp],4		    // non-faulting prefetch
-	add	$f=$f,tmp0		    // f+=F_20_39(b,c,d)
+{ .mmi;	add	$f=$f,tmp0		    // f+=F_20_39(b,c,d)
 	shrp	$b=tmp6,tmp6,2		}   // b=ROTATE(b,30) ;;?
 { .mmi;	or	tmp1=tmp1,tmp5		    // ROTATE(a,5)
 	add	$h3=$h3,$c		};; // wrap up
@@ -245,164 +231,11 @@ tmp3=r11;
 ctx=r32;	// in0
 inp=r33;	// in1
 
-// void sha1_block_asm_host_order(SHA_CTX *c,const void *p,size_t num);
-.global	sha1_block_asm_host_order#
-.proc	sha1_block_asm_host_order#
+// void sha1_block_data_order(SHA_CTX *c,const void *p,size_t num);
+.global	sha1_block_data_order#
+.proc	sha1_block_data_order#
 .align	32
-sha1_block_asm_host_order:
-	.prologue
-{ .mmi;	alloc	tmp1=ar.pfs,3,15,0,0
-	$ADDP	tmp0=4,ctx
-	.save	ar.lc,r3
-	mov	r3=ar.lc		}
-{ .mmi;	$ADDP	ctx=0,ctx
-	$ADDP	inp=0,inp
-	mov	r2=pr			};;
-tmp4=in2;
-tmp5=loc13;
-tmp6=loc14;
-	.body
-{ .mlx;	ld4	$h0=[ctx],8
-	movl	$K_00_19=0x5a827999	}
-{ .mlx;	ld4	$h1=[tmp0],8
-	movl	$K_20_39=0x6ed9eba1	};;
-{ .mlx;	ld4	$h2=[ctx],8
-	movl	$K_40_59=0x8f1bbcdc	}
-{ .mlx;	ld4	$h3=[tmp0]
-	movl	$K_60_79=0xca62c1d6	};;
-{ .mmi;	ld4	$h4=[ctx],-16
-	add	in2=-1,in2		    // adjust num for ar.lc
-	mov	ar.ec=1			};;
-{ .mmi;	ld4	$X[0]=[inp],4		    // prefetch
-	cmp.ne	p16,p0=r0,in2		    // prefecth at loop end
-	mov	ar.lc=in2		};; // brp.loop.imp: too far
-
-.Lhtop:
-{ .mmi;	mov	$A=$h0
-	mov	$B=$h1
-	mux2	tmp6=$h1,0x44		}
-{ .mmi;	mov	$C=$h2
-	mov	$D=$h3
-	mov	$E=$h4			};;
-
-___
-
-	&BODY_00_15(\$code, 0,$A,$B,$C,$D,$E,$T);
-	&BODY_00_15(\$code, 1,$T,$A,$B,$C,$D,$E);
-	&BODY_00_15(\$code, 2,$E,$T,$A,$B,$C,$D);
-	&BODY_00_15(\$code, 3,$D,$E,$T,$A,$B,$C);
-	&BODY_00_15(\$code, 4,$C,$D,$E,$T,$A,$B);
-	&BODY_00_15(\$code, 5,$B,$C,$D,$E,$T,$A);
-	&BODY_00_15(\$code, 6,$A,$B,$C,$D,$E,$T);
-	&BODY_00_15(\$code, 7,$T,$A,$B,$C,$D,$E);
-	&BODY_00_15(\$code, 8,$E,$T,$A,$B,$C,$D);
-	&BODY_00_15(\$code, 9,$D,$E,$T,$A,$B,$C);
-	&BODY_00_15(\$code,10,$C,$D,$E,$T,$A,$B);
-	&BODY_00_15(\$code,11,$B,$C,$D,$E,$T,$A);
-	&BODY_00_15(\$code,12,$A,$B,$C,$D,$E,$T);
-	&BODY_00_15(\$code,13,$T,$A,$B,$C,$D,$E);
-	&BODY_00_15(\$code,14,$E,$T,$A,$B,$C,$D);
-	&BODY_00_15(\$code,15,$D,$E,$T,$A,$B,$C);
-
-	&BODY_16_19(\$code,16,$C,$D,$E,$T,$A,$B);
-	&BODY_16_19(\$code,17,$B,$C,$D,$E,$T,$A);
-	&BODY_16_19(\$code,18,$A,$B,$C,$D,$E,$T);
-	&BODY_16_19(\$code,19,$T,$A,$B,$C,$D,$E);
-
-	&BODY_20_39(\$code,20,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,21,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39(\$code,22,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39(\$code,23,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39(\$code,24,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39(\$code,25,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39(\$code,26,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,27,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39(\$code,28,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39(\$code,29,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39(\$code,30,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39(\$code,31,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39(\$code,32,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,33,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39(\$code,34,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39(\$code,35,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39(\$code,36,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39(\$code,37,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39(\$code,38,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,39,$D,$E,$T,$A,$B,$C);
-
-	&BODY_40_59(\$code,40,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,41,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59(\$code,42,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59(\$code,43,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59(\$code,44,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59(\$code,45,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59(\$code,46,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,47,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59(\$code,48,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59(\$code,49,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59(\$code,50,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59(\$code,51,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59(\$code,52,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,53,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59(\$code,54,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59(\$code,55,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59(\$code,56,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59(\$code,57,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59(\$code,58,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,59,$B,$C,$D,$E,$T,$A);
-
-	&BODY_60_79(\$code,60,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,61,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79(\$code,62,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79(\$code,63,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79(\$code,64,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79(\$code,65,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79(\$code,66,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,67,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79(\$code,68,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79(\$code,69,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79(\$code,70,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79(\$code,71,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79(\$code,72,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,73,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79(\$code,74,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79(\$code,75,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79(\$code,76,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79(\$code,77,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79(\$code,78,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,79,$T,$A,$B,$C,$D,$E);
-
-$code.=<<___;
-{ .mmb;	add	$h0=$h0,$E
-	nop.m	0
-	br.ctop.dptk.many	.Lhtop	};;
-.Lhend:
-{ .mmi;	add	tmp0=4,ctx
-	mov	ar.lc=r3		};;
-{ .mmi;	st4	[ctx]=$h0,8
-	st4	[tmp0]=$h1,8		};;
-{ .mmi;	st4	[ctx]=$h2,8
-	st4	[tmp0]=$h3		};;
-{ .mib;	st4	[ctx]=$h4,-16
-	mov	pr=r2,0x1ffff
-	br.ret.sptk.many	b0	};;
-.endp	sha1_block_asm_host_order#
-___
-
-
-$code.=<<___;
-// void sha1_block_asm_data_order(SHA_CTX *c,const void *p,size_t num);
-.global	sha1_block_asm_data_order#
-.proc	sha1_block_asm_data_order#
-.align	32
-sha1_block_asm_data_order:
-___
-$code.=<<___ if ($big_endian);
-{ .mmi;	and	r2=3,inp				};;
-{ .mib;	cmp.eq	p6,p0=r0,r2
-(p6)	br.dptk.many	sha1_block_asm_host_order	};;
-___
-$code.=<<___;
+sha1_block_data_order:
 	.prologue
 { .mmi;	alloc	tmp1=ar.pfs,3,15,0,0
 	$ADDP	tmp0=4,ctx
@@ -440,90 +273,16 @@ tmp6=loc14;
 
 ___
 
-	&BODY_00_15(\$code, 0,$A,$B,$C,$D,$E,$T,1);
-	&BODY_00_15(\$code, 1,$T,$A,$B,$C,$D,$E,1);
-	&BODY_00_15(\$code, 2,$E,$T,$A,$B,$C,$D,1);
-	&BODY_00_15(\$code, 3,$D,$E,$T,$A,$B,$C,1);
-	&BODY_00_15(\$code, 4,$C,$D,$E,$T,$A,$B,1);
-	&BODY_00_15(\$code, 5,$B,$C,$D,$E,$T,$A,1);
-	&BODY_00_15(\$code, 6,$A,$B,$C,$D,$E,$T,1);
-	&BODY_00_15(\$code, 7,$T,$A,$B,$C,$D,$E,1);
-	&BODY_00_15(\$code, 8,$E,$T,$A,$B,$C,$D,1);
-	&BODY_00_15(\$code, 9,$D,$E,$T,$A,$B,$C,1);
-	&BODY_00_15(\$code,10,$C,$D,$E,$T,$A,$B,1);
-	&BODY_00_15(\$code,11,$B,$C,$D,$E,$T,$A,1);
-	&BODY_00_15(\$code,12,$A,$B,$C,$D,$E,$T,1);
-	&BODY_00_15(\$code,13,$T,$A,$B,$C,$D,$E,1);
-	&BODY_00_15(\$code,14,$E,$T,$A,$B,$C,$D,1);
-	&BODY_00_15(\$code,15,$D,$E,$T,$A,$B,$C,1);
-
-	&BODY_16_19(\$code,16,$C,$D,$E,$T,$A,$B);
-	&BODY_16_19(\$code,17,$B,$C,$D,$E,$T,$A);
-	&BODY_16_19(\$code,18,$A,$B,$C,$D,$E,$T);
-	&BODY_16_19(\$code,19,$T,$A,$B,$C,$D,$E);
+{ my $i,@V=($A,$B,$C,$D,$E,$T);
 
-	&BODY_20_39(\$code,20,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,21,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39(\$code,22,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39(\$code,23,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39(\$code,24,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39(\$code,25,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39(\$code,26,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,27,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39(\$code,28,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39(\$code,29,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39(\$code,30,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39(\$code,31,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39(\$code,32,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,33,$D,$E,$T,$A,$B,$C);
-	&BODY_20_39(\$code,34,$C,$D,$E,$T,$A,$B);
-	&BODY_20_39(\$code,35,$B,$C,$D,$E,$T,$A);
-	&BODY_20_39(\$code,36,$A,$B,$C,$D,$E,$T);
-	&BODY_20_39(\$code,37,$T,$A,$B,$C,$D,$E);
-	&BODY_20_39(\$code,38,$E,$T,$A,$B,$C,$D);
-	&BODY_20_39(\$code,39,$D,$E,$T,$A,$B,$C);
+	for($i=0;$i<16;$i++)	{ &BODY_00_15(\$code,$i,@V); unshift(@V,pop(@V)); }
+	for(;$i<20;$i++)	{ &BODY_16_19(\$code,$i,@V); unshift(@V,pop(@V)); }
+	for(;$i<40;$i++)	{ &BODY_20_39(\$code,$i,@V); unshift(@V,pop(@V)); }
+	for(;$i<60;$i++)	{ &BODY_40_59(\$code,$i,@V); unshift(@V,pop(@V)); }
+	for(;$i<80;$i++)	{ &BODY_60_79(\$code,$i,@V); unshift(@V,pop(@V)); }
 
-	&BODY_40_59(\$code,40,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,41,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59(\$code,42,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59(\$code,43,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59(\$code,44,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59(\$code,45,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59(\$code,46,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,47,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59(\$code,48,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59(\$code,49,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59(\$code,50,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59(\$code,51,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59(\$code,52,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,53,$B,$C,$D,$E,$T,$A);
-	&BODY_40_59(\$code,54,$A,$B,$C,$D,$E,$T);
-	&BODY_40_59(\$code,55,$T,$A,$B,$C,$D,$E);
-	&BODY_40_59(\$code,56,$E,$T,$A,$B,$C,$D);
-	&BODY_40_59(\$code,57,$D,$E,$T,$A,$B,$C);
-	&BODY_40_59(\$code,58,$C,$D,$E,$T,$A,$B);
-	&BODY_40_59(\$code,59,$B,$C,$D,$E,$T,$A);
-
-	&BODY_60_79(\$code,60,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,61,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79(\$code,62,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79(\$code,63,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79(\$code,64,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79(\$code,65,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79(\$code,66,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,67,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79(\$code,68,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79(\$code,69,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79(\$code,70,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79(\$code,71,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79(\$code,72,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,73,$T,$A,$B,$C,$D,$E);
-	&BODY_60_79(\$code,74,$E,$T,$A,$B,$C,$D);
-	&BODY_60_79(\$code,75,$D,$E,$T,$A,$B,$C);
-	&BODY_60_79(\$code,76,$C,$D,$E,$T,$A,$B);
-	&BODY_60_79(\$code,77,$B,$C,$D,$E,$T,$A);
-	&BODY_60_79(\$code,78,$A,$B,$C,$D,$E,$T);
-	&BODY_60_79(\$code,79,$T,$A,$B,$C,$D,$E);
+	(($V[5] eq $D) and ($V[0] eq $E)) or die;	# double-check
+}
 
 $code.=<<___;
 { .mmb;	add	$h0=$h0,$E
@@ -539,7 +298,8 @@ $code.=<<___;
 { .mib;	st4	[ctx]=$h4,-16
 	mov	pr=r2,0x1ffff
 	br.ret.sptk.many	b0	};;
-.endp	sha1_block_asm_data_order#
+.endp	sha1_block_data_order#
+stringz	"SHA1 block transform for IA64, CRYPTOGAMS by <appro\@openssl.org>"
 ___
 
 print $code;
diff --git a/crypto/sha/asm/sha512-ia64.pl b/crypto/sha/asm/sha512-ia64.pl
index 9de9174992..1c6ce56522 100755
--- a/crypto/sha/asm/sha512-ia64.pl
+++ b/crypto/sha/asm/sha512-ia64.pl
@@ -2,8 +2,9 @@
 #
 # ====================================================================
 # Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
-# project. Rights for redistribution and usage in source and binary
-# forms are granted according to the OpenSSL license.