Initial Commit
This commit is contained in:
root
209
openssl-1.0.2f/crypto/des/asm/crypt586.pl
Normal file
209
openssl-1.0.2f/crypto/des/asm/crypt586.pl
Normal file
@@ -0,0 +1,209 @@
|
||||
#!/usr/local/bin/perl
|
||||
#
|
||||
# The inner loop instruction sequence and the IP/FP modifications are from
|
||||
# Svend Olaf Mikkelsen <svolaf@inet.uni-c.dk>
|
||||
# I've added the stuff needed for crypt() but I've not worried about making
|
||||
# things perfect.
|
||||
#
|
||||
|
||||
$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
|
||||
push(@INC,"${dir}","${dir}../../perlasm");
|
||||
require "x86asm.pl";
|
||||
|
||||
&asm_init($ARGV[0],"crypt586.pl");
|
||||
|
||||
$L="edi";
|
||||
$R="esi";
|
||||
|
||||
&external_label("DES_SPtrans");
|
||||
&fcrypt_body("fcrypt_body");
|
||||
&asm_finish();
|
||||
|
||||
sub fcrypt_body
|
||||
{
|
||||
local($name,$do_ip)=@_;
|
||||
|
||||
&function_begin($name);
|
||||
|
||||
&comment("");
|
||||
&comment("Load the 2 words");
|
||||
$trans="ebp";
|
||||
|
||||
&xor( $L, $L);
|
||||
&xor( $R, $R);
|
||||
|
||||
# PIC-ification:-)
|
||||
&picmeup("edx","DES_SPtrans");
|
||||
#if ($cpp) { &picmeup("edx","DES_SPtrans"); }
|
||||
#else { &lea("edx",&DWP("DES_SPtrans")); }
|
||||
&push("edx"); # becomes &swtmp(1)
|
||||
#
|
||||
&mov($trans,&wparam(1)); # reloaded with DES_SPtrans in D_ENCRYPT
|
||||
|
||||
&push(&DWC(25)); # add a variable
|
||||
|
||||
&set_label("start");
|
||||
for ($i=0; $i<16; $i+=2)
|
||||
{
|
||||
&comment("");
|
||||
&comment("Round $i");
|
||||
&D_ENCRYPT($i,$L,$R,$i*2,$trans,"eax","ebx","ecx","edx");
|
||||
|
||||
&comment("");
|
||||
&comment("Round ".sprintf("%d",$i+1));
|
||||
&D_ENCRYPT($i+1,$R,$L,($i+1)*2,$trans,"eax","ebx","ecx","edx");
|
||||
}
|
||||
&mov("ebx", &swtmp(0));
|
||||
&mov("eax", $L);
|
||||
&dec("ebx");
|
||||
&mov($L, $R);
|
||||
&mov($R, "eax");
|
||||
&mov(&swtmp(0), "ebx");
|
||||
&jnz(&label("start"));
|
||||
|
||||
&comment("");
|
||||
&comment("FP");
|
||||
&mov("edx",&wparam(0));
|
||||
|
||||
&FP_new($R,$L,"eax",3);
|
||||
&mov(&DWP(0,"edx","",0),"eax");
|
||||
&mov(&DWP(4,"edx","",0),$L);
|
||||
|
||||
&add("esp",8); # remove variables
|
||||
|
||||
&function_end($name);
|
||||
}
|
||||
|
||||
sub D_ENCRYPT
|
||||
{
|
||||
local($r,$L,$R,$S,$trans,$u,$tmp1,$tmp2,$t)=@_;
|
||||
|
||||
&mov( $u, &wparam(2)); # 2
|
||||
&mov( $t, $R);
|
||||
&shr( $t, 16); # 1
|
||||
&mov( $tmp2, &wparam(3)); # 2
|
||||
&xor( $t, $R); # 1
|
||||
|
||||
&and( $u, $t); # 2
|
||||
&and( $t, $tmp2); # 2
|
||||
|
||||
&mov( $tmp1, $u);
|
||||
&shl( $tmp1, 16); # 1
|
||||
&mov( $tmp2, $t);
|
||||
&shl( $tmp2, 16); # 1
|
||||
&xor( $u, $tmp1); # 2
|
||||
&xor( $t, $tmp2); # 2
|
||||
&mov( $tmp1, &DWP(&n2a($S*4),$trans,"",0)); # 2
|
||||
&xor( $u, $tmp1);
|
||||
&mov( $tmp2, &DWP(&n2a(($S+1)*4),$trans,"",0)); # 2
|
||||
&xor( $u, $R);
|
||||
&xor( $t, $R);
|
||||
&xor( $t, $tmp2);
|
||||
|
||||
&and( $u, "0xfcfcfcfc" ); # 2
|
||||
&xor( $tmp1, $tmp1); # 1
|
||||
&and( $t, "0xcfcfcfcf" ); # 2
|
||||
&xor( $tmp2, $tmp2);
|
||||
&movb( &LB($tmp1), &LB($u) );
|
||||
&movb( &LB($tmp2), &HB($u) );
|
||||
&rotr( $t, 4 );
|
||||
&mov( $trans, &swtmp(1));
|
||||
&xor( $L, &DWP(" ",$trans,$tmp1,0));
|
||||
&movb( &LB($tmp1), &LB($t) );
|
||||
&xor( $L, &DWP("0x200",$trans,$tmp2,0));
|
||||
&movb( &LB($tmp2), &HB($t) );
|
||||
&shr( $u, 16);
|
||||
&xor( $L, &DWP("0x100",$trans,$tmp1,0));
|
||||
&movb( &LB($tmp1), &HB($u) );
|
||||
&shr( $t, 16);
|
||||
&xor( $L, &DWP("0x300",$trans,$tmp2,0));
|
||||
&movb( &LB($tmp2), &HB($t) );
|
||||
&and( $u, "0xff" );
|
||||
&and( $t, "0xff" );
|
||||
&mov( $tmp1, &DWP("0x600",$trans,$tmp1,0));
|
||||
&xor( $L, $tmp1);
|
||||
&mov( $tmp1, &DWP("0x700",$trans,$tmp2,0));
|
||||
&xor( $L, $tmp1);
|
||||
&mov( $tmp1, &DWP("0x400",$trans,$u,0));
|
||||
&xor( $L, $tmp1);
|
||||
&mov( $tmp1, &DWP("0x500",$trans,$t,0));
|
||||
&xor( $L, $tmp1);
|
||||
&mov( $trans, &wparam(1));
|
||||
}
|
||||
|
||||
sub n2a
|
||||
{
|
||||
sprintf("%d",$_[0]);
|
||||
}
|
||||
|
||||
# now has a side affect of rotating $a by $shift
|
||||
sub R_PERM_OP
|
||||
{
|
||||
local($a,$b,$tt,$shift,$mask,$last)=@_;
|
||||
|
||||
&rotl( $a, $shift ) if ($shift != 0);
|
||||
&mov( $tt, $a );
|
||||
&xor( $a, $b );
|
||||
&and( $a, $mask );
|
||||
if ($notlast eq $b)
|
||||
{
|
||||
&xor( $b, $a );
|
||||
&xor( $tt, $a );
|
||||
}
|
||||
else
|
||||
{
|
||||
&xor( $tt, $a );
|
||||
&xor( $b, $a );
|
||||
}
|
||||
&comment("");
|
||||
}
|
||||
|
||||
sub IP_new
|
||||
{
|
||||
local($l,$r,$tt,$lr)=@_;
|
||||
|
||||
&R_PERM_OP($l,$r,$tt, 4,"0xf0f0f0f0",$l);
|
||||
&R_PERM_OP($r,$tt,$l,20,"0xfff0000f",$l);
|
||||
&R_PERM_OP($l,$tt,$r,14,"0x33333333",$r);
|
||||
&R_PERM_OP($tt,$r,$l,22,"0x03fc03fc",$r);
|
||||
&R_PERM_OP($l,$r,$tt, 9,"0xaaaaaaaa",$r);
|
||||
|
||||
if ($lr != 3)
|
||||
{
|
||||
if (($lr-3) < 0)
|
||||
{ &rotr($tt, 3-$lr); }
|
||||
else { &rotl($tt, $lr-3); }
|
||||
}
|
||||
if ($lr != 2)
|
||||
{
|
||||
if (($lr-2) < 0)
|
||||
{ &rotr($r, 2-$lr); }
|
||||
else { &rotl($r, $lr-2); }
|
||||
}
|
||||
}
|
||||
|
||||
sub FP_new
|
||||
{
|
||||
local($l,$r,$tt,$lr)=@_;
|
||||
|
||||
if ($lr != 2)
|
||||
{
|
||||
if (($lr-2) < 0)
|
||||
{ &rotl($r, 2-$lr); }
|
||||
else { &rotr($r, $lr-2); }
|
||||
}
|
||||
if ($lr != 3)
|
||||
{
|
||||
if (($lr-3) < 0)
|
||||
{ &rotl($l, 3-$lr); }
|
||||
else { &rotr($l, $lr-3); }
|
||||
}
|
||||
|
||||
&R_PERM_OP($l,$r,$tt, 0,"0xaaaaaaaa",$r);
|
||||
&R_PERM_OP($tt,$r,$l,23,"0x03fc03fc",$r);
|
||||
&R_PERM_OP($l,$r,$tt,10,"0x33333333",$l);
|
||||
&R_PERM_OP($r,$tt,$l,18,"0xfff0000f",$l);
|
||||
&R_PERM_OP($l,$tt,$r,12,"0xf0f0f0f0",$r);
|
||||
&rotr($tt , 4);
|
||||
}
|
||||
|
||||
455
openssl-1.0.2f/crypto/des/asm/des-586.pl
Normal file
455
openssl-1.0.2f/crypto/des/asm/des-586.pl
Normal file
@@ -0,0 +1,455 @@
|
||||
#!/usr/local/bin/perl
|
||||
#
|
||||
# The inner loop instruction sequence and the IP/FP modifications are from
|
||||
# Svend Olaf Mikkelsen <svolaf@inet.uni-c.dk>
|
||||
#
|
||||
|
||||
$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
|
||||
push(@INC,"${dir}","${dir}../../perlasm");
|
||||
require "x86asm.pl";
|
||||
require "cbc.pl";
|
||||
require "desboth.pl";
|
||||
|
||||
# base code is in microsft
|
||||
# op dest, source
|
||||
# format.
|
||||
#
|
||||
|
||||
&asm_init($ARGV[0],"des-586.pl");
|
||||
|
||||
$L="edi";
|
||||
$R="esi";
|
||||
$trans="ebp";
|
||||
$small_footprint=1 if (grep(/\-DOPENSSL_SMALL_FOOTPRINT/,@ARGV));
|
||||
# one can discuss setting this variable to 1 unconditionally, as
|
||||
# the folded loop is only 3% slower than unrolled, but >7 times smaller
|
||||
|
||||
&public_label("DES_SPtrans");
|
||||
&static_label("des_sptrans");
|
||||
|
||||
&DES_encrypt_internal();
|
||||
&DES_decrypt_internal();
|
||||
&DES_encrypt("DES_encrypt1",1);
|
||||
&DES_encrypt("DES_encrypt2",0);
|
||||
&DES_encrypt3("DES_encrypt3",1);
|
||||
&DES_encrypt3("DES_decrypt3",0);
|
||||
&cbc("DES_ncbc_encrypt","DES_encrypt1","DES_encrypt1",0,4,5,3,5,-1);
|
||||
&cbc("DES_ede3_cbc_encrypt","DES_encrypt3","DES_decrypt3",0,6,7,3,4,5);
|
||||
&DES_SPtrans();
|
||||
|
||||
&asm_finish();
|
||||
|
||||
sub DES_encrypt_internal()
|
||||
{
|
||||
&function_begin_B("_x86_DES_encrypt");
|
||||
|
||||
if ($small_footprint)
|
||||
{
|
||||
&lea("edx",&DWP(128,"ecx"));
|
||||
&push("edx");
|
||||
&push("ecx");
|
||||
&set_label("eloop");
|
||||
&D_ENCRYPT(0,$L,$R,0,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
&comment("");
|
||||
&D_ENCRYPT(1,$R,$L,2,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
&comment("");
|
||||
&add("ecx",16);
|
||||
&cmp("ecx",&swtmp(1));
|
||||
&mov(&swtmp(0),"ecx");
|
||||
&jb(&label("eloop"));
|
||||
&add("esp",8);
|
||||
}
|
||||
else
|
||||
{
|
||||
&push("ecx");
|
||||
for ($i=0; $i<16; $i+=2)
|
||||
{
|
||||
&comment("Round $i");
|
||||
&D_ENCRYPT($i,$L,$R,$i*2,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
&comment("Round ".sprintf("%d",$i+1));
|
||||
&D_ENCRYPT($i+1,$R,$L,($i+1)*2,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
}
|
||||
&add("esp",4);
|
||||
}
|
||||
&ret();
|
||||
|
||||
&function_end_B("_x86_DES_encrypt");
|
||||
}
|
||||
|
||||
sub DES_decrypt_internal()
|
||||
{
|
||||
&function_begin_B("_x86_DES_decrypt");
|
||||
|
||||
if ($small_footprint)
|
||||
{
|
||||
&push("ecx");
|
||||
&lea("ecx",&DWP(128,"ecx"));
|
||||
&push("ecx");
|
||||
&set_label("dloop");
|
||||
&D_ENCRYPT(0,$L,$R,-2,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
&comment("");
|
||||
&D_ENCRYPT(1,$R,$L,-4,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
&comment("");
|
||||
&sub("ecx",16);
|
||||
&cmp("ecx",&swtmp(1));
|
||||
&mov(&swtmp(0),"ecx");
|
||||
&ja(&label("dloop"));
|
||||
&add("esp",8);
|
||||
}
|
||||
else
|
||||
{
|
||||
&push("ecx");
|
||||
for ($i=15; $i>0; $i-=2)
|
||||
{
|
||||
&comment("Round $i");
|
||||
&D_ENCRYPT(15-$i,$L,$R,$i*2,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
&comment("Round ".sprintf("%d",$i-1));
|
||||
&D_ENCRYPT(15-$i+1,$R,$L,($i-1)*2,$trans,"eax","ebx","ecx","edx",&swtmp(0));
|
||||
}
|
||||
&add("esp",4);
|
||||
}
|
||||
&ret();
|
||||
|
||||
&function_end_B("_x86_DES_decrypt");
|
||||
}
|
||||
|
||||
sub DES_encrypt
|
||||
{
|
||||
local($name,$do_ip)=@_;
|
||||
|
||||
&function_begin_B($name);
|
||||
|
||||
&push("esi");
|
||||
&push("edi");
|
||||
|
||||
&comment("");
|
||||
&comment("Load the 2 words");
|
||||
|
||||
if ($do_ip)
|
||||
{
|
||||
&mov($R,&wparam(0));
|
||||
&xor( "ecx", "ecx" );
|
||||
|
||||
&push("ebx");
|
||||
&push("ebp");
|
||||
|
||||
&mov("eax",&DWP(0,$R,"",0));
|
||||
&mov("ebx",&wparam(2)); # get encrypt flag
|
||||
&mov($L,&DWP(4,$R,"",0));
|
||||
&comment("");
|
||||
&comment("IP");
|
||||
&IP_new("eax",$L,$R,3);
|
||||
}
|
||||
else
|
||||
{
|
||||
&mov("eax",&wparam(0));
|
||||
&xor( "ecx", "ecx" );
|
||||
|
||||
&push("ebx");
|
||||
&push("ebp");
|
||||
|
||||
&mov($R,&DWP(0,"eax","",0));
|
||||
&mov("ebx",&wparam(2)); # get encrypt flag
|
||||
&rotl($R,3);
|
||||
&mov($L,&DWP(4,"eax","",0));
|
||||
&rotl($L,3);
|
||||
}
|
||||
|
||||
# PIC-ification:-)
|
||||
&call (&label("pic_point"));
|
||||
&set_label("pic_point");
|
||||
&blindpop($trans);
|
||||
&lea ($trans,&DWP(&label("des_sptrans")."-".&label("pic_point"),$trans));
|
||||
|
||||
&mov( "ecx", &wparam(1) );
|
||||
|
||||
&cmp("ebx","0");
|
||||
&je(&label("decrypt"));
|
||||
&call("_x86_DES_encrypt");
|
||||
&jmp(&label("done"));
|
||||
&set_label("decrypt");
|
||||
&call("_x86_DES_decrypt");
|
||||
&set_label("done");
|
||||
|
||||
if ($do_ip)
|
||||
{
|
||||
&comment("");
|
||||
&comment("FP");
|
||||
&mov("edx",&wparam(0));
|
||||
&FP_new($L,$R,"eax",3);
|
||||
|
||||
&mov(&DWP(0,"edx","",0),"eax");
|
||||
&mov(&DWP(4,"edx","",0),$R);
|
||||
}
|
||||
else
|
||||
{
|
||||
&comment("");
|
||||
&comment("Fixup");
|
||||
&rotr($L,3); # r
|
||||
&mov("eax",&wparam(0));
|
||||
&rotr($R,3); # l
|
||||
&mov(&DWP(0,"eax","",0),$L);
|
||||
&mov(&DWP(4,"eax","",0),$R);
|
||||
}
|
||||
|
||||
&pop("ebp");
|
||||
&pop("ebx");
|
||||
&pop("edi");
|
||||
&pop("esi");
|
||||
&ret();
|
||||
|
||||
&function_end_B($name);
|
||||
}
|
||||
|
||||
sub D_ENCRYPT
|
||||
{
|
||||
local($r,$L,$R,$S,$trans,$u,$tmp1,$tmp2,$t,$wp1)=@_;
|
||||
|
||||
&mov( $u, &DWP(&n2a($S*4),$tmp2,"",0));
|
||||
&xor( $tmp1, $tmp1);
|
||||
&mov( $t, &DWP(&n2a(($S+1)*4),$tmp2,"",0));
|
||||
&xor( $u, $R);
|
||||
&xor( $tmp2, $tmp2);
|
||||
&xor( $t, $R);
|
||||
&and( $u, "0xfcfcfcfc" );
|
||||
&and( $t, "0xcfcfcfcf" );
|
||||
&movb( &LB($tmp1), &LB($u) );
|
||||
&movb( &LB($tmp2), &HB($u) );
|
||||
&rotr( $t, 4 );
|
||||
&xor( $L, &DWP(" ",$trans,$tmp1,0));
|
||||
&movb( &LB($tmp1), &LB($t) );
|
||||
&xor( $L, &DWP("0x200",$trans,$tmp2,0));
|
||||
&movb( &LB($tmp2), &HB($t) );
|
||||
&shr( $u, 16);
|
||||
&xor( $L, &DWP("0x100",$trans,$tmp1,0));
|
||||
&movb( &LB($tmp1), &HB($u) );
|
||||
&shr( $t, 16);
|
||||
&xor( $L, &DWP("0x300",$trans,$tmp2,0));
|
||||
&movb( &LB($tmp2), &HB($t) );
|
||||
&and( $u, "0xff" );
|
||||
&and( $t, "0xff" );
|
||||
&xor( $L, &DWP("0x600",$trans,$tmp1,0));
|
||||
&xor( $L, &DWP("0x700",$trans,$tmp2,0));
|
||||
&mov( $tmp2, $wp1 );
|
||||
&xor( $L, &DWP("0x400",$trans,$u,0));
|
||||
&xor( $L, &DWP("0x500",$trans,$t,0));
|
||||
}
|
||||
|
||||
sub n2a
|
||||
{
|
||||
sprintf("%d",$_[0]);
|
||||
}
|
||||
|
||||
# now has a side affect of rotating $a by $shift
|
||||
sub R_PERM_OP
|
||||
{
|
||||
local($a,$b,$tt,$shift,$mask,$last)=@_;
|
||||
|
||||
&rotl( $a, $shift ) if ($shift != 0);
|
||||
&mov( $tt, $a );
|
||||
&xor( $a, $b );
|
||||
&and( $a, $mask );
|
||||
# This can never succeed, and besides it is difficult to see what the
|
||||
# idea was - Ben 13 Feb 99
|
||||
if (!$last eq $b)
|
||||
{
|
||||
&xor( $b, $a );
|
||||
&xor( $tt, $a );
|
||||
}
|
||||
else
|
||||
{
|
||||
&xor( $tt, $a );
|
||||
&xor( $b, $a );
|
||||
}
|
||||
&comment("");
|
||||
}
|
||||
|
||||
sub IP_new
|
||||
{
|
||||
local($l,$r,$tt,$lr)=@_;
|
||||
|
||||
&R_PERM_OP($l,$r,$tt, 4,"0xf0f0f0f0",$l);
|
||||
&R_PERM_OP($r,$tt,$l,20,"0xfff0000f",$l);
|
||||
&R_PERM_OP($l,$tt,$r,14,"0x33333333",$r);
|
||||
&R_PERM_OP($tt,$r,$l,22,"0x03fc03fc",$r);
|
||||
&R_PERM_OP($l,$r,$tt, 9,"0xaaaaaaaa",$r);
|
||||
|
||||
if ($lr != 3)
|
||||
{
|
||||
if (($lr-3) < 0)
|
||||
{ &rotr($tt, 3-$lr); }
|
||||
else { &rotl($tt, $lr-3); }
|
||||
}
|
||||
if ($lr != 2)
|
||||
{
|
||||
if (($lr-2) < 0)
|
||||
{ &rotr($r, 2-$lr); }
|
||||
else { &rotl($r, $lr-2); }
|
||||
}
|
||||
}
|
||||
|
||||
sub FP_new
|
||||
{
|
||||
local($l,$r,$tt,$lr)=@_;
|
||||
|
||||
if ($lr != 2)
|
||||
{
|
||||
if (($lr-2) < 0)
|
||||
{ &rotl($r, 2-$lr); }
|
||||
else { &rotr($r, $lr-2); }
|
||||
}
|
||||
if ($lr != 3)
|
||||
{
|
||||
if (($lr-3) < 0)
|
||||
{ &rotl($l, 3-$lr); }
|
||||
else { &rotr($l, $lr-3); }
|
||||
}
|
||||
|
||||
&R_PERM_OP($l,$r,$tt, 0,"0xaaaaaaaa",$r);
|
||||
&R_PERM_OP($tt,$r,$l,23,"0x03fc03fc",$r);
|
||||
&R_PERM_OP($l,$r,$tt,10,"0x33333333",$l);
|
||||
&R_PERM_OP($r,$tt,$l,18,"0xfff0000f",$l);
|
||||
&R_PERM_OP($l,$tt,$r,12,"0xf0f0f0f0",$r);
|
||||
&rotr($tt , 4);
|
||||
}
|
||||
|
||||
sub DES_SPtrans
|
||||
{
|
||||
&set_label("DES_SPtrans",64);
|
||||
&set_label("des_sptrans");
|
||||
&data_word(0x02080800, 0x00080000, 0x02000002, 0x02080802);
|
||||
&data_word(0x02000000, 0x00080802, 0x00080002, 0x02000002);
|
||||
&data_word(0x00080802, 0x02080800, 0x02080000, 0x00000802);
|
||||
&data_word(0x02000802, 0x02000000, 0x00000000, 0x00080002);
|
||||
&data_word(0x00080000, 0x00000002, 0x02000800, 0x00080800);
|
||||
&data_word(0x02080802, 0x02080000, 0x00000802, 0x02000800);
|
||||
&data_word(0x00000002, 0x00000800, 0x00080800, 0x02080002);
|
||||
&data_word(0x00000800, 0x02000802, 0x02080002, 0x00000000);
|
||||
&data_word(0x00000000, 0x02080802, 0x02000800, 0x00080002);
|
||||
&data_word(0x02080800, 0x00080000, 0x00000802, 0x02000800);
|
||||
&data_word(0x02080002, 0x00000800, 0x00080800, 0x02000002);
|
||||
&data_word(0x00080802, 0x00000002, 0x02000002, 0x02080000);
|
||||
&data_word(0x02080802, 0x00080800, 0x02080000, 0x02000802);
|
||||
&data_word(0x02000000, 0x00000802, 0x00080002, 0x00000000);
|
||||
&data_word(0x00080000, 0x02000000, 0x02000802, 0x02080800);
|
||||
&data_word(0x00000002, 0x02080002, 0x00000800, 0x00080802);
|
||||
# nibble 1
|
||||
&data_word(0x40108010, 0x00000000, 0x00108000, 0x40100000);
|
||||
&data_word(0x40000010, 0x00008010, 0x40008000, 0x00108000);
|
||||
&data_word(0x00008000, 0x40100010, 0x00000010, 0x40008000);
|
||||
&data_word(0x00100010, 0x40108000, 0x40100000, 0x00000010);
|
||||
&data_word(0x00100000, 0x40008010, 0x40100010, 0x00008000);
|
||||
&data_word(0x00108010, 0x40000000, 0x00000000, 0x00100010);
|
||||
&data_word(0x40008010, 0x00108010, 0x40108000, 0x40000010);
|
||||
&data_word(0x40000000, 0x00100000, 0x00008010, 0x40108010);
|
||||
&data_word(0x00100010, 0x40108000, 0x40008000, 0x00108010);
|
||||
&data_word(0x40108010, 0x00100010, 0x40000010, 0x00000000);
|
||||
&data_word(0x40000000, 0x00008010, 0x00100000, 0x40100010);
|
||||
&data_word(0x00008000, 0x40000000, 0x00108010, 0x40008010);
|
||||
&data_word(0x40108000, 0x00008000, 0x00000000, 0x40000010);
|
||||
&data_word(0x00000010, 0x40108010, 0x00108000, 0x40100000);
|
||||
&data_word(0x40100010, 0x00100000, 0x00008010, 0x40008000);
|
||||
&data_word(0x40008010, 0x00000010, 0x40100000, 0x00108000);
|
||||
# nibble 2
|
||||
&data_word(0x04000001, 0x04040100, 0x00000100, 0x04000101);
|
||||
&data_word(0x00040001, 0x04000000, 0x04000101, 0x00040100);
|
||||
&data_word(0x04000100, 0x00040000, 0x04040000, 0x00000001);
|
||||
&data_word(0x04040101, 0x00000101, 0x00000001, 0x04040001);
|
||||
&data_word(0x00000000, 0x00040001, 0x04040100, 0x00000100);
|
||||
&data_word(0x00000101, 0x04040101, 0x00040000, 0x04000001);
|
||||
&data_word(0x04040001, 0x04000100, 0x00040101, 0x04040000);
|
||||
&data_word(0x00040100, 0x00000000, 0x04000000, 0x00040101);
|
||||
&data_word(0x04040100, 0x00000100, 0x00000001, 0x00040000);
|
||||
&data_word(0x00000101, 0x00040001, 0x04040000, 0x04000101);
|
||||
&data_word(0x00000000, 0x04040100, 0x00040100, 0x04040001);
|
||||
&data_word(0x00040001, 0x04000000, 0x04040101, 0x00000001);
|
||||
&data_word(0x00040101, 0x04000001, 0x04000000, 0x04040101);
|
||||
&data_word(0x00040000, 0x04000100, 0x04000101, 0x00040100);
|
||||
&data_word(0x04000100, 0x00000000, 0x04040001, 0x00000101);
|
||||
&data_word(0x04000001, 0x00040101, 0x00000100, 0x04040000);
|
||||
# nibble 3
|
||||
&data_word(0x00401008, 0x10001000, 0x00000008, 0x10401008);
|
||||
&data_word(0x00000000, 0x10400000, 0x10001008, 0x00400008);
|
||||
&data_word(0x10401000, 0x10000008, 0x10000000, 0x00001008);
|
||||
&data_word(0x10000008, 0x00401008, 0x00400000, 0x10000000);
|
||||
&data_word(0x10400008, 0x00401000, 0x00001000, 0x00000008);
|
||||
&data_word(0x00401000, 0x10001008, 0x10400000, 0x00001000);
|
||||
&data_word(0x00001008, 0x00000000, 0x00400008, 0x10401000);
|
||||
&data_word(0x10001000, 0x10400008, 0x10401008, 0x00400000);
|
||||
&data_word(0x10400008, 0x00001008, 0x00400000, 0x10000008);
|
||||
&data_word(0x00401000, 0x10001000, 0x00000008, 0x10400000);
|
||||
&data_word(0x10001008, 0x00000000, 0x00001000, 0x00400008);
|
||||
&data_word(0x00000000, 0x10400008, 0x10401000, 0x00001000);
|
||||
&data_word(0x10000000, 0x10401008, 0x00401008, 0x00400000);
|
||||
&data_word(0x10401008, 0x00000008, 0x10001000, 0x00401008);
|
||||
&data_word(0x00400008, 0x00401000, 0x10400000, 0x10001008);
|
||||
&data_word(0x00001008, 0x10000000, 0x10000008, 0x10401000);
|
||||
# nibble 4
|
||||
&data_word(0x08000000, 0x00010000, 0x00000400, 0x08010420);
|
||||
&data_word(0x08010020, 0x08000400, 0x00010420, 0x08010000);
|
||||
&data_word(0x00010000, 0x00000020, 0x08000020, 0x00010400);
|
||||
&data_word(0x08000420, 0x08010020, 0x08010400, 0x00000000);
|
||||
&data_word(0x00010400, 0x08000000, 0x00010020, 0x00000420);
|
||||
&data_word(0x08000400, 0x00010420, 0x00000000, 0x08000020);
|
||||
&data_word(0x00000020, 0x08000420, 0x08010420, 0x00010020);
|
||||
&data_word(0x08010000, 0x00000400, 0x00000420, 0x08010400);
|
||||
&data_word(0x08010400, 0x08000420, 0x00010020, 0x08010000);
|
||||
&data_word(0x00010000, 0x00000020, 0x08000020, 0x08000400);
|
||||
&data_word(0x08000000, 0x00010400, 0x08010420, 0x00000000);
|
||||
&data_word(0x00010420, 0x08000000, 0x00000400, 0x00010020);
|
||||
&data_word(0x08000420, 0x00000400, 0x00000000, 0x08010420);
|
||||
&data_word(0x08010020, 0x08010400, 0x00000420, 0x00010000);
|
||||
&data_word(0x00010400, 0x08010020, 0x08000400, 0x00000420);
|
||||
&data_word(0x00000020, 0x00010420, 0x08010000, 0x08000020);
|
||||
# nibble 5
|
||||
&data_word(0x80000040, 0x00200040, 0x00000000, 0x80202000);
|
||||
&data_word(0x00200040, 0x00002000, 0x80002040, 0x00200000);
|
||||
&data_word(0x00002040, 0x80202040, 0x00202000, 0x80000000);
|
||||
&data_word(0x80002000, 0x80000040, 0x80200000, 0x00202040);
|
||||
&data_word(0x00200000, 0x80002040, 0x80200040, 0x00000000);
|
||||
&data_word(0x00002000, 0x00000040, 0x80202000, 0x80200040);
|
||||
&data_word(0x80202040, 0x80200000, 0x80000000, 0x00002040);
|
||||
&data_word(0x00000040, 0x00202000, 0x00202040, 0x80002000);
|
||||
&data_word(0x00002040, 0x80000000, 0x80002000, 0x00202040);
|
||||
&data_word(0x80202000, 0x00200040, 0x00000000, 0x80002000);
|
||||
&data_word(0x80000000, 0x00002000, 0x80200040, 0x00200000);
|
||||
&data_word(0x00200040, 0x80202040, 0x00202000, 0x00000040);
|
||||
&data_word(0x80202040, 0x00202000, 0x00200000, 0x80002040);
|
||||
&data_word(0x80000040, 0x80200000, 0x00202040, 0x00000000);
|
||||
&data_word(0x00002000, 0x80000040, 0x80002040, 0x80202000);
|
||||
&data_word(0x80200000, 0x00002040, 0x00000040, 0x80200040);
|
||||
# nibble 6
|
||||
&data_word(0x00004000, 0x00000200, 0x01000200, 0x01000004);
|
||||
&data_word(0x01004204, 0x00004004, 0x00004200, 0x00000000);
|
||||
&data_word(0x01000000, 0x01000204, 0x00000204, 0x01004000);
|
||||
&data_word(0x00000004, 0x01004200, 0x01004000, 0x00000204);
|
||||
&data_word(0x01000204, 0x00004000, 0x00004004, 0x01004204);
|
||||
&data_word(0x00000000, 0x01000200, 0x01000004, 0x00004200);
|
||||
&data_word(0x01004004, 0x00004204, 0x01004200, 0x00000004);
|
||||
&data_word(0x00004204, 0x01004004, 0x00000200, 0x01000000);
|
||||
&data_word(0x00004204, 0x01004000, 0x01004004, 0x00000204);
|
||||
&data_word(0x00004000, 0x00000200, 0x01000000, 0x01004004);
|
||||
&data_word(0x01000204, 0x00004204, 0x00004200, 0x00000000);
|
||||
&data_word(0x00000200, 0x01000004, 0x00000004, 0x01000200);
|
||||
&data_word(0x00000000, 0x01000204, 0x01000200, 0x00004200);
|
||||
&data_word(0x00000204, 0x00004000, 0x01004204, 0x01000000);
|
||||
&data_word(0x01004200, 0x00000004, 0x00004004, 0x01004204);
|
||||
&data_word(0x01000004, 0x01004200, 0x01004000, 0x00004004);
|
||||
# nibble 7
|
||||
&data_word(0x20800080, 0x20820000, 0x00020080, 0x00000000);
|
||||
&data_word(0x20020000, 0x00800080, 0x20800000, 0x20820080);
|
||||
&data_word(0x00000080, 0x20000000, 0x00820000, 0x00020080);
|
||||
&data_word(0x00820080, 0x20020080, 0x20000080, 0x20800000);
|
||||
&data_word(0x00020000, 0x00820080, 0x00800080, 0x20020000);
|
||||
&data_word(0x20820080, 0x20000080, 0x00000000, 0x00820000);
|
||||
&data_word(0x20000000, 0x00800000, 0x20020080, 0x20800080);
|
||||
&data_word(0x00800000, 0x00020000, 0x20820000, 0x00000080);
|
||||
&data_word(0x00800000, 0x00020000, 0x20000080, 0x20820080);
|
||||
&data_word(0x00020080, 0x20000000, 0x00000000, 0x00820000);
|
||||
&data_word(0x20800080, 0x20020080, 0x20020000, 0x00800080);
|
||||
&data_word(0x20820000, 0x00000080, 0x00800080, 0x20020000);
|
||||
&data_word(0x20820080, 0x00800000, 0x20800000, 0x20000080);
|
||||
&data_word(0x00820000, 0x00020080, 0x20020080, 0x20800000);
|
||||
&data_word(0x00000080, 0x20820000, 0x00820080, 0x00000000);
|
||||
&data_word(0x20000000, 0x20800080, 0x00020000, 0x00820080);
|
||||
}
|
||||
2101
openssl-1.0.2f/crypto/des/asm/des_enc.m4
Normal file
2101
openssl-1.0.2f/crypto/des/asm/des_enc.m4
Normal file
File diff suppressed because it is too large
Load Diff
79
openssl-1.0.2f/crypto/des/asm/desboth.pl
Normal file
79
openssl-1.0.2f/crypto/des/asm/desboth.pl
Normal file
@@ -0,0 +1,79 @@
|
||||
#!/usr/local/bin/perl
|
||||
|
||||
$L="edi";
|
||||
$R="esi";
|
||||
|
||||
sub DES_encrypt3
|
||||
{
|
||||
local($name,$enc)=@_;
|
||||
|
||||
&function_begin_B($name,"");
|
||||
&push("ebx");
|
||||
&mov("ebx",&wparam(0));
|
||||
|
||||
&push("ebp");
|
||||
&push("esi");
|
||||
|
||||
&push("edi");
|
||||
|
||||
&comment("");
|
||||
&comment("Load the data words");
|
||||
&mov($L,&DWP(0,"ebx","",0));
|
||||
&mov($R,&DWP(4,"ebx","",0));
|
||||
&stack_push(3);
|
||||
|
||||
&comment("");
|
||||
&comment("IP");
|
||||
&IP_new($L,$R,"edx",0);
|
||||
|
||||
# put them back
|
||||
|
||||
if ($enc)
|
||||
{
|
||||
&mov(&DWP(4,"ebx","",0),$R);
|
||||
&mov("eax",&wparam(1));
|
||||
&mov(&DWP(0,"ebx","",0),"edx");
|
||||
&mov("edi",&wparam(2));
|
||||
&mov("esi",&wparam(3));
|
||||
}
|
||||
else
|
||||
{
|
||||
&mov(&DWP(4,"ebx","",0),$R);
|
||||
&mov("esi",&wparam(1));
|
||||
&mov(&DWP(0,"ebx","",0),"edx");
|
||||
&mov("edi",&wparam(2));
|
||||
&mov("eax",&wparam(3));
|
||||
}
|
||||
&mov(&swtmp(2), (DWC(($enc)?"1":"0")));
|
||||
&mov(&swtmp(1), "eax");
|
||||
&mov(&swtmp(0), "ebx");
|
||||
&call("DES_encrypt2");
|
||||
&mov(&swtmp(2), (DWC(($enc)?"0":"1")));
|
||||
&mov(&swtmp(1), "edi");
|
||||
&mov(&swtmp(0), "ebx");
|
||||
&call("DES_encrypt2");
|
||||
&mov(&swtmp(2), (DWC(($enc)?"1":"0")));
|
||||
&mov(&swtmp(1), "esi");
|
||||
&mov(&swtmp(0), "ebx");
|
||||
&call("DES_encrypt2");
|
||||
|
||||
&stack_pop(3);
|
||||
&mov($L,&DWP(0,"ebx","",0));
|
||||
&mov($R,&DWP(4,"ebx","",0));
|
||||
|
||||
&comment("");
|
||||
&comment("FP");
|
||||
&FP_new($L,$R,"eax",0);
|
||||
|
||||
&mov(&DWP(0,"ebx","",0),"eax");
|
||||
&mov(&DWP(4,"ebx","",0),$R);
|
||||
|
||||
&pop("edi");
|
||||
&pop("esi");
|
||||
&pop("ebp");
|
||||
&pop("ebx");
|
||||
&ret();
|
||||
&function_end_B($name);
|
||||
}
|
||||
|
||||
|
||||
617
openssl-1.0.2f/crypto/des/asm/dest4-sparcv9.pl
Normal file
617
openssl-1.0.2f/crypto/des/asm/dest4-sparcv9.pl
Normal file
@@ -0,0 +1,617 @@
|
||||
#!/usr/bin/env perl
|
||||
|
||||
# ====================================================================
|
||||
# Written by David S. Miller <davem@devemloft.net> and Andy Polyakov
|
||||
# <appro@openssl.org>. The module is licensed under 2-clause BSD
|
||||
# license. March 2013. All rights reserved.
|
||||
# ====================================================================
|
||||
|
||||
######################################################################
|
||||
# DES for SPARC T4.
|
||||
#
|
||||
# As with other hardware-assisted ciphers CBC encrypt results [for
|
||||
# aligned data] are virtually identical to critical path lengths:
|
||||
#
|
||||
# DES Triple-DES
|
||||
# CBC encrypt 4.14/4.15(*) 11.7/11.7
|
||||
# CBC decrypt 1.77/4.11(**) 6.42/7.47
|
||||
#
|
||||
# (*) numbers after slash are for
|
||||
# misaligned data;
|
||||
# (**) this is result for largest
|
||||
# block size, unlike all other
|
||||
# cases smaller blocks results
|
||||
# are better[?];
|
||||
|
||||
$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1;
|
||||
push(@INC,"${dir}","${dir}../../perlasm");
|
||||
require "sparcv9_modes.pl";
|
||||
|
||||
&asm_init(@ARGV);
|
||||
|
||||
$code.=<<___ if ($::abibits==64);
|
||||
.register %g2,#scratch
|
||||
.register %g3,#scratch
|
||||
___
|
||||
|
||||
$code.=<<___;
|
||||
.text
|
||||
___
|
||||
|
||||
{ my ($inp,$out)=("%o0","%o1");
|
||||
|
||||
$code.=<<___;
|
||||
.align 32
|
||||
.globl des_t4_key_expand
|
||||
.type des_t4_key_expand,#function
|
||||
des_t4_key_expand:
|
||||
andcc $inp, 0x7, %g0
|
||||
alignaddr $inp, %g0, $inp
|
||||
bz,pt %icc, 1f
|
||||
ldd [$inp + 0x00], %f0
|
||||
ldd [$inp + 0x08], %f2
|
||||
faligndata %f0, %f2, %f0
|
||||
1: des_kexpand %f0, 0, %f0
|
||||
des_kexpand %f0, 1, %f2
|
||||
std %f0, [$out + 0x00]
|
||||
des_kexpand %f2, 3, %f6
|
||||
std %f2, [$out + 0x08]
|
||||
des_kexpand %f2, 2, %f4
|
||||
des_kexpand %f6, 3, %f10
|
||||
std %f6, [$out + 0x18]
|
||||
des_kexpand %f6, 2, %f8
|
||||
std %f4, [$out + 0x10]
|
||||
des_kexpand %f10, 3, %f14
|
||||
std %f10, [$out + 0x28]
|
||||
des_kexpand %f10, 2, %f12
|
||||
std %f8, [$out + 0x20]
|
||||
des_kexpand %f14, 1, %f16
|
||||
std %f14, [$out + 0x38]
|
||||
des_kexpand %f16, 3, %f20
|
||||
std %f12, [$out + 0x30]
|
||||
des_kexpand %f16, 2, %f18
|
||||
std %f16, [$out + 0x40]
|
||||
des_kexpand %f20, 3, %f24
|
||||
std %f20, [$out + 0x50]
|
||||
des_kexpand %f20, 2, %f22
|
||||
std %f18, [$out + 0x48]
|
||||
des_kexpand %f24, 3, %f28
|
||||
std %f24, [$out + 0x60]
|
||||
des_kexpand %f24, 2, %f26
|
||||
std %f22, [$out + 0x58]
|
||||
des_kexpand %f28, 1, %f30
|
||||
std %f28, [$out + 0x70]
|
||||
std %f26, [$out + 0x68]
|
||||
retl
|
||||
std %f30, [$out + 0x78]
|
||||
.size des_t4_key_expand,.-des_t4_key_expand
|
||||
___
|
||||
}
|
||||
{ my ($inp,$out,$len,$key,$ivec) = map("%o$_",(0..4));
|
||||
my ($ileft,$iright,$omask) = map("%g$_",(1..3));
|
||||
|
||||
$code.=<<___;
|
||||
.globl des_t4_cbc_encrypt
|
||||
.align 32
|
||||
des_t4_cbc_encrypt:
|
||||
cmp $len, 0
|
||||
be,pn $::size_t_cc, .Lcbc_abort
|
||||
nop
|
||||
ld [$ivec + 0], %f0 ! load ivec
|
||||
ld [$ivec + 4], %f1
|
||||
|
||||
and $inp, 7, $ileft
|
||||
andn $inp, 7, $inp
|
||||
sll $ileft, 3, $ileft
|
||||
mov 0xff, $omask
|
||||
prefetch [$inp], 20
|
||||
prefetch [$inp + 63], 20
|
||||
sub %g0, $ileft, $iright
|
||||
and $out, 7, %g4
|
||||
alignaddrl $out, %g0, $out
|
||||
srl $omask, %g4, $omask
|
||||
srlx $len, 3, $len
|
||||
movrz %g4, 0, $omask
|
||||
prefetch [$out], 22
|
||||
|
||||
ldd [$key + 0x00], %f4 ! load key schedule
|
||||
ldd [$key + 0x08], %f6
|
||||
ldd [$key + 0x10], %f8
|
||||
ldd [$key + 0x18], %f10
|
||||
ldd [$key + 0x20], %f12
|
||||
ldd [$key + 0x28], %f14
|
||||
ldd [$key + 0x30], %f16
|
||||
ldd [$key + 0x38], %f18
|
||||
ldd [$key + 0x40], %f20
|
||||
ldd [$key + 0x48], %f22
|
||||
ldd [$key + 0x50], %f24
|
||||
ldd [$key + 0x58], %f26
|
||||
ldd [$key + 0x60], %f28
|
||||
ldd [$key + 0x68], %f30
|
||||
ldd [$key + 0x70], %f32
|
||||
ldd [$key + 0x78], %f34
|
||||
|
||||
.Ldes_cbc_enc_loop:
|
||||
ldx [$inp + 0], %g4
|
||||
brz,pt $ileft, 4f
|
||||
nop
|
||||
|
||||
ldx [$inp + 8], %g5
|
||||
sllx %g4, $ileft, %g4
|
||||
srlx %g5, $iright, %g5
|
||||
or %g5, %g4, %g4
|
||||
4:
|
||||
movxtod %g4, %f2
|
||||
prefetch [$inp + 8+63], 20
|
||||
add $inp, 8, $inp
|
||||
fxor %f2, %f0, %f0 ! ^= ivec
|
||||
prefetch [$out + 63], 22
|
||||
|
||||
des_ip %f0, %f0
|
||||
des_round %f4, %f6, %f0, %f0
|
||||
des_round %f8, %f10, %f0, %f0
|
||||
des_round %f12, %f14, %f0, %f0
|
||||
des_round %f16, %f18, %f0, %f0
|
||||
des_round %f20, %f22, %f0, %f0
|
||||
des_round %f24, %f26, %f0, %f0
|
||||
des_round %f28, %f30, %f0, %f0
|
||||
des_round %f32, %f34, %f0, %f0
|
||||
des_iip %f0, %f0
|
||||
|
||||
brnz,pn $omask, 2f
|
||||
sub $len, 1, $len
|
||||
|
||||
std %f0, [$out + 0]
|
||||
brnz,pt $len, .Ldes_cbc_enc_loop
|
||||
add $out, 8, $out
|
||||
|
||||
st %f0, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f1, [$ivec + 4]
|
||||
.Lcbc_abort:
|
||||
retl
|
||||
nop
|
||||
|
||||
.align 16
|
||||
2: ldxa [$inp]0x82, %g4 ! avoid read-after-write hazard
|
||||
! and ~4x deterioration
|
||||
! in inp==out case
|
||||
faligndata %f0, %f0, %f2 ! handle unaligned output
|
||||
|
||||
stda %f2, [$out + $omask]0xc0 ! partial store
|
||||
add $out, 8, $out
|
||||
orn %g0, $omask, $omask
|
||||
stda %f2, [$out + $omask]0xc0 ! partial store
|
||||
|
||||
brnz,pt $len, .Ldes_cbc_enc_loop+4
|
||||
orn %g0, $omask, $omask
|
||||
|
||||
st %f0, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f1, [$ivec + 4]
|
||||
.type des_t4_cbc_encrypt,#function
|
||||
.size des_t4_cbc_encrypt,.-des_t4_cbc_encrypt
|
||||
|
||||
.globl des_t4_cbc_decrypt
|
||||
.align 32
|
||||
des_t4_cbc_decrypt:
|
||||
cmp $len, 0
|
||||
be,pn $::size_t_cc, .Lcbc_abort
|
||||
nop
|
||||
ld [$ivec + 0], %f2 ! load ivec
|
||||
ld [$ivec + 4], %f3
|
||||
|
||||
and $inp, 7, $ileft
|
||||
andn $inp, 7, $inp
|
||||
sll $ileft, 3, $ileft
|
||||
mov 0xff, $omask
|
||||
prefetch [$inp], 20
|
||||
prefetch [$inp + 63], 20
|
||||
sub %g0, $ileft, $iright
|
||||
and $out, 7, %g4
|
||||
alignaddrl $out, %g0, $out
|
||||
srl $omask, %g4, $omask
|
||||
srlx $len, 3, $len
|
||||
movrz %g4, 0, $omask
|
||||
prefetch [$out], 22
|
||||
|
||||
ldd [$key + 0x78], %f4 ! load key schedule
|
||||
ldd [$key + 0x70], %f6
|
||||
ldd [$key + 0x68], %f8
|
||||
ldd [$key + 0x60], %f10
|
||||
ldd [$key + 0x58], %f12
|
||||
ldd [$key + 0x50], %f14
|
||||
ldd [$key + 0x48], %f16
|
||||
ldd [$key + 0x40], %f18
|
||||
ldd [$key + 0x38], %f20
|
||||
ldd [$key + 0x30], %f22
|
||||
ldd [$key + 0x28], %f24
|
||||
ldd [$key + 0x20], %f26
|
||||
ldd [$key + 0x18], %f28
|
||||
ldd [$key + 0x10], %f30
|
||||
ldd [$key + 0x08], %f32
|
||||
ldd [$key + 0x00], %f34
|
||||
|
||||
.Ldes_cbc_dec_loop:
|
||||
ldx [$inp + 0], %g4
|
||||
brz,pt $ileft, 4f
|
||||
nop
|
||||
|
||||
ldx [$inp + 8], %g5
|
||||
sllx %g4, $ileft, %g4
|
||||
srlx %g5, $iright, %g5
|
||||
or %g5, %g4, %g4
|
||||
4:
|
||||
movxtod %g4, %f0
|
||||
prefetch [$inp + 8+63], 20
|
||||
add $inp, 8, $inp
|
||||
prefetch [$out + 63], 22
|
||||
|
||||
des_ip %f0, %f0
|
||||
des_round %f4, %f6, %f0, %f0
|
||||
des_round %f8, %f10, %f0, %f0
|
||||
des_round %f12, %f14, %f0, %f0
|
||||
des_round %f16, %f18, %f0, %f0
|
||||
des_round %f20, %f22, %f0, %f0
|
||||
des_round %f24, %f26, %f0, %f0
|
||||
des_round %f28, %f30, %f0, %f0
|
||||
des_round %f32, %f34, %f0, %f0
|
||||
des_iip %f0, %f0
|
||||
|
||||
fxor %f2, %f0, %f0 ! ^= ivec
|
||||
movxtod %g4, %f2
|
||||
|
||||
brnz,pn $omask, 2f
|
||||
sub $len, 1, $len
|
||||
|
||||
std %f0, [$out + 0]
|
||||
brnz,pt $len, .Ldes_cbc_dec_loop
|
||||
add $out, 8, $out
|
||||
|
||||
st %f2, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f3, [$ivec + 4]
|
||||
|
||||
.align 16
|
||||
2: ldxa [$inp]0x82, %g4 ! avoid read-after-write hazard
|
||||
! and ~4x deterioration
|
||||
! in inp==out case
|
||||
faligndata %f0, %f0, %f0 ! handle unaligned output
|
||||
|
||||
stda %f0, [$out + $omask]0xc0 ! partial store
|
||||
add $out, 8, $out
|
||||
orn %g0, $omask, $omask
|
||||
stda %f0, [$out + $omask]0xc0 ! partial store
|
||||
|
||||
brnz,pt $len, .Ldes_cbc_dec_loop+4
|
||||
orn %g0, $omask, $omask
|
||||
|
||||
st %f2, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f3, [$ivec + 4]
|
||||
.type des_t4_cbc_decrypt,#function
|
||||
.size des_t4_cbc_decrypt,.-des_t4_cbc_decrypt
|
||||
___
|
||||
|
||||
# One might wonder why does one have back-to-back des_iip/des_ip
|
||||
# pairs between EDE passes. Indeed, aren't they inverse of each other?
|
||||
# They almost are. Outcome of the pair is 32-bit words being swapped
|
||||
# in target register. Consider pair of des_iip/des_ip as a way to
|
||||
# perform the due swap, it's actually fastest way in this case.
|
||||
|
||||
$code.=<<___;
|
||||
.globl des_t4_ede3_cbc_encrypt
|
||||
.align 32
|
||||
des_t4_ede3_cbc_encrypt:
|
||||
cmp $len, 0
|
||||
be,pn $::size_t_cc, .Lcbc_abort
|
||||
nop
|
||||
ld [$ivec + 0], %f0 ! load ivec
|
||||
ld [$ivec + 4], %f1
|
||||
|
||||
and $inp, 7, $ileft
|
||||
andn $inp, 7, $inp
|
||||
sll $ileft, 3, $ileft
|
||||
mov 0xff, $omask
|
||||
prefetch [$inp], 20
|
||||
prefetch [$inp + 63], 20
|
||||
sub %g0, $ileft, $iright
|
||||
and $out, 7, %g4
|
||||
alignaddrl $out, %g0, $out
|
||||
srl $omask, %g4, $omask
|
||||
srlx $len, 3, $len
|
||||
movrz %g4, 0, $omask
|
||||
prefetch [$out], 22
|
||||
|
||||
ldd [$key + 0x00], %f4 ! load key schedule
|
||||
ldd [$key + 0x08], %f6
|
||||
ldd [$key + 0x10], %f8
|
||||
ldd [$key + 0x18], %f10
|
||||
ldd [$key + 0x20], %f12
|
||||
ldd [$key + 0x28], %f14
|
||||
ldd [$key + 0x30], %f16
|
||||
ldd [$key + 0x38], %f18
|
||||
ldd [$key + 0x40], %f20
|
||||
ldd [$key + 0x48], %f22
|
||||
ldd [$key + 0x50], %f24
|
||||
ldd [$key + 0x58], %f26
|
||||
ldd [$key + 0x60], %f28
|
||||
ldd [$key + 0x68], %f30
|
||||
ldd [$key + 0x70], %f32
|
||||
ldd [$key + 0x78], %f34
|
||||
|
||||
.Ldes_ede3_cbc_enc_loop:
|
||||
ldx [$inp + 0], %g4
|
||||
brz,pt $ileft, 4f
|
||||
nop
|
||||
|
||||
ldx [$inp + 8], %g5
|
||||
sllx %g4, $ileft, %g4
|
||||
srlx %g5, $iright, %g5
|
||||
or %g5, %g4, %g4
|
||||
4:
|
||||
movxtod %g4, %f2
|
||||
prefetch [$inp + 8+63], 20
|
||||
add $inp, 8, $inp
|
||||
fxor %f2, %f0, %f0 ! ^= ivec
|
||||
prefetch [$out + 63], 22
|
||||
|
||||
des_ip %f0, %f0
|
||||
des_round %f4, %f6, %f0, %f0
|
||||
des_round %f8, %f10, %f0, %f0
|
||||
des_round %f12, %f14, %f0, %f0
|
||||
des_round %f16, %f18, %f0, %f0
|
||||
ldd [$key + 0x100-0x08], %f36
|
||||
ldd [$key + 0x100-0x10], %f38
|
||||
des_round %f20, %f22, %f0, %f0
|
||||
ldd [$key + 0x100-0x18], %f40
|
||||
ldd [$key + 0x100-0x20], %f42
|
||||
des_round %f24, %f26, %f0, %f0
|
||||
ldd [$key + 0x100-0x28], %f44
|
||||
ldd [$key + 0x100-0x30], %f46
|
||||
des_round %f28, %f30, %f0, %f0
|
||||
ldd [$key + 0x100-0x38], %f48
|
||||
ldd [$key + 0x100-0x40], %f50
|
||||
des_round %f32, %f34, %f0, %f0
|
||||
ldd [$key + 0x100-0x48], %f52
|
||||
ldd [$key + 0x100-0x50], %f54
|
||||
des_iip %f0, %f0
|
||||
|
||||
ldd [$key + 0x100-0x58], %f56
|
||||
ldd [$key + 0x100-0x60], %f58
|
||||
des_ip %f0, %f0
|
||||
ldd [$key + 0x100-0x68], %f60
|
||||
ldd [$key + 0x100-0x70], %f62
|
||||
des_round %f36, %f38, %f0, %f0
|
||||
ldd [$key + 0x100-0x78], %f36
|
||||
ldd [$key + 0x100-0x80], %f38
|
||||
des_round %f40, %f42, %f0, %f0
|
||||
des_round %f44, %f46, %f0, %f0
|
||||
des_round %f48, %f50, %f0, %f0
|
||||
ldd [$key + 0x100+0x00], %f40
|
||||
ldd [$key + 0x100+0x08], %f42
|
||||
des_round %f52, %f54, %f0, %f0
|
||||
ldd [$key + 0x100+0x10], %f44
|
||||
ldd [$key + 0x100+0x18], %f46
|
||||
des_round %f56, %f58, %f0, %f0
|
||||
ldd [$key + 0x100+0x20], %f48
|
||||
ldd [$key + 0x100+0x28], %f50
|
||||
des_round %f60, %f62, %f0, %f0
|
||||
ldd [$key + 0x100+0x30], %f52
|
||||
ldd [$key + 0x100+0x38], %f54
|
||||
des_round %f36, %f38, %f0, %f0
|
||||
ldd [$key + 0x100+0x40], %f56
|
||||
ldd [$key + 0x100+0x48], %f58
|
||||
des_iip %f0, %f0
|
||||
|
||||
ldd [$key + 0x100+0x50], %f60
|
||||
ldd [$key + 0x100+0x58], %f62
|
||||
des_ip %f0, %f0
|
||||
ldd [$key + 0x100+0x60], %f36
|
||||
ldd [$key + 0x100+0x68], %f38
|
||||
des_round %f40, %f42, %f0, %f0
|
||||
ldd [$key + 0x100+0x70], %f40
|
||||
ldd [$key + 0x100+0x78], %f42
|
||||
des_round %f44, %f46, %f0, %f0
|
||||
des_round %f48, %f50, %f0, %f0
|
||||
des_round %f52, %f54, %f0, %f0
|
||||
des_round %f56, %f58, %f0, %f0
|
||||
des_round %f60, %f62, %f0, %f0
|
||||
des_round %f36, %f38, %f0, %f0
|
||||
des_round %f40, %f42, %f0, %f0
|
||||
des_iip %f0, %f0
|
||||
|
||||
brnz,pn $omask, 2f
|
||||
sub $len, 1, $len
|
||||
|
||||
std %f0, [$out + 0]
|
||||
brnz,pt $len, .Ldes_ede3_cbc_enc_loop
|
||||
add $out, 8, $out
|
||||
|
||||
st %f0, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f1, [$ivec + 4]
|
||||
|
||||
.align 16
|
||||
2: ldxa [$inp]0x82, %g4 ! avoid read-after-write hazard
|
||||
! and ~2x deterioration
|
||||
! in inp==out case
|
||||
faligndata %f0, %f0, %f2 ! handle unaligned output
|
||||
|
||||
stda %f2, [$out + $omask]0xc0 ! partial store
|
||||
add $out, 8, $out
|
||||
orn %g0, $omask, $omask
|
||||
stda %f2, [$out + $omask]0xc0 ! partial store
|
||||
|
||||
brnz,pt $len, .Ldes_ede3_cbc_enc_loop+4
|
||||
orn %g0, $omask, $omask
|
||||
|
||||
st %f0, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f1, [$ivec + 4]
|
||||
.type des_t4_ede3_cbc_encrypt,#function
|
||||
.size des_t4_ede3_cbc_encrypt,.-des_t4_ede3_cbc_encrypt
|
||||
|
||||
.globl des_t4_ede3_cbc_decrypt
|
||||
.align 32
|
||||
des_t4_ede3_cbc_decrypt:
|
||||
cmp $len, 0
|
||||
be,pn $::size_t_cc, .Lcbc_abort
|
||||
nop
|
||||
ld [$ivec + 0], %f2 ! load ivec
|
||||
ld [$ivec + 4], %f3
|
||||
|
||||
and $inp, 7, $ileft
|
||||
andn $inp, 7, $inp
|
||||
sll $ileft, 3, $ileft
|
||||
mov 0xff, $omask
|
||||
prefetch [$inp], 20
|
||||
prefetch [$inp + 63], 20
|
||||
sub %g0, $ileft, $iright
|
||||
and $out, 7, %g4
|
||||
alignaddrl $out, %g0, $out
|
||||
srl $omask, %g4, $omask
|
||||
srlx $len, 3, $len
|
||||
movrz %g4, 0, $omask
|
||||
prefetch [$out], 22
|
||||
|
||||
ldd [$key + 0x100+0x78], %f4 ! load key schedule
|
||||
ldd [$key + 0x100+0x70], %f6
|
||||
ldd [$key + 0x100+0x68], %f8
|
||||
ldd [$key + 0x100+0x60], %f10
|
||||
ldd [$key + 0x100+0x58], %f12
|
||||
ldd [$key + 0x100+0x50], %f14
|
||||
ldd [$key + 0x100+0x48], %f16
|
||||
ldd [$key + 0x100+0x40], %f18
|
||||
ldd [$key + 0x100+0x38], %f20
|
||||
ldd [$key + 0x100+0x30], %f22
|
||||
ldd [$key + 0x100+0x28], %f24
|
||||
ldd [$key + 0x100+0x20], %f26
|
||||
ldd [$key + 0x100+0x18], %f28
|
||||
ldd [$key + 0x100+0x10], %f30
|
||||
ldd [$key + 0x100+0x08], %f32
|
||||
ldd [$key + 0x100+0x00], %f34
|
||||
|
||||
.Ldes_ede3_cbc_dec_loop:
|
||||
ldx [$inp + 0], %g4
|
||||
brz,pt $ileft, 4f
|
||||
nop
|
||||
|
||||
ldx [$inp + 8], %g5
|
||||
sllx %g4, $ileft, %g4
|
||||
srlx %g5, $iright, %g5
|
||||
or %g5, %g4, %g4
|
||||
4:
|
||||
movxtod %g4, %f0
|
||||
prefetch [$inp + 8+63], 20
|
||||
add $inp, 8, $inp
|
||||
prefetch [$out + 63], 22
|
||||
|
||||
des_ip %f0, %f0
|
||||
des_round %f4, %f6, %f0, %f0
|
||||
des_round %f8, %f10, %f0, %f0
|
||||
des_round %f12, %f14, %f0, %f0
|
||||
des_round %f16, %f18, %f0, %f0
|
||||
ldd [$key + 0x80+0x00], %f36
|
||||
ldd [$key + 0x80+0x08], %f38
|
||||
des_round %f20, %f22, %f0, %f0
|
||||
ldd [$key + 0x80+0x10], %f40
|
||||
ldd [$key + 0x80+0x18], %f42
|
||||
des_round %f24, %f26, %f0, %f0
|
||||
ldd [$key + 0x80+0x20], %f44
|
||||
ldd [$key + 0x80+0x28], %f46
|
||||
des_round %f28, %f30, %f0, %f0
|
||||
ldd [$key + 0x80+0x30], %f48
|
||||
ldd [$key + 0x80+0x38], %f50
|
||||
des_round %f32, %f34, %f0, %f0
|
||||
ldd [$key + 0x80+0x40], %f52
|
||||
ldd [$key + 0x80+0x48], %f54
|
||||
des_iip %f0, %f0
|
||||
|
||||
ldd [$key + 0x80+0x50], %f56
|
||||
ldd [$key + 0x80+0x58], %f58
|
||||
des_ip %f0, %f0
|
||||
ldd [$key + 0x80+0x60], %f60
|
||||
ldd [$key + 0x80+0x68], %f62
|
||||
des_round %f36, %f38, %f0, %f0
|
||||
ldd [$key + 0x80+0x70], %f36
|
||||
ldd [$key + 0x80+0x78], %f38
|
||||
des_round %f40, %f42, %f0, %f0
|
||||
des_round %f44, %f46, %f0, %f0
|
||||
des_round %f48, %f50, %f0, %f0
|
||||
ldd [$key + 0x80-0x08], %f40
|
||||
ldd [$key + 0x80-0x10], %f42
|
||||
des_round %f52, %f54, %f0, %f0
|
||||
ldd [$key + 0x80-0x18], %f44
|
||||
ldd [$key + 0x80-0x20], %f46
|
||||
des_round %f56, %f58, %f0, %f0
|
||||
ldd [$key + 0x80-0x28], %f48
|
||||
ldd [$key + 0x80-0x30], %f50
|
||||
des_round %f60, %f62, %f0, %f0
|
||||
ldd [$key + 0x80-0x38], %f52
|
||||
ldd [$key + 0x80-0x40], %f54
|
||||
des_round %f36, %f38, %f0, %f0
|
||||
ldd [$key + 0x80-0x48], %f56
|
||||
ldd [$key + 0x80-0x50], %f58
|
||||
des_iip %f0, %f0
|
||||
|
||||
ldd [$key + 0x80-0x58], %f60
|
||||
ldd [$key + 0x80-0x60], %f62
|
||||
des_ip %f0, %f0
|
||||
ldd [$key + 0x80-0x68], %f36
|
||||
ldd [$key + 0x80-0x70], %f38
|
||||
des_round %f40, %f42, %f0, %f0
|
||||
ldd [$key + 0x80-0x78], %f40
|
||||
ldd [$key + 0x80-0x80], %f42
|
||||
des_round %f44, %f46, %f0, %f0
|
||||
des_round %f48, %f50, %f0, %f0
|
||||
des_round %f52, %f54, %f0, %f0
|
||||
des_round %f56, %f58, %f0, %f0
|
||||
des_round %f60, %f62, %f0, %f0
|
||||
des_round %f36, %f38, %f0, %f0
|
||||
des_round %f40, %f42, %f0, %f0
|
||||
des_iip %f0, %f0
|
||||
|
||||
fxor %f2, %f0, %f0 ! ^= ivec
|
||||
movxtod %g4, %f2
|
||||
|
||||
brnz,pn $omask, 2f
|
||||
sub $len, 1, $len
|
||||
|
||||
std %f0, [$out + 0]
|
||||
brnz,pt $len, .Ldes_ede3_cbc_dec_loop
|
||||
add $out, 8, $out
|
||||
|
||||
st %f2, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f3, [$ivec + 4]
|
||||
|
||||
.align 16
|
||||
2: ldxa [$inp]0x82, %g4 ! avoid read-after-write hazard
|
||||
! and ~3x deterioration
|
||||
! in inp==out case
|
||||
faligndata %f0, %f0, %f0 ! handle unaligned output
|
||||
|
||||
stda %f0, [$out + $omask]0xc0 ! partial store
|
||||
add $out, 8, $out
|
||||
orn %g0, $omask, $omask
|
||||
stda %f0, [$out + $omask]0xc0 ! partial store
|
||||
|
||||
brnz,pt $len, .Ldes_ede3_cbc_dec_loop+4
|
||||
orn %g0, $omask, $omask
|
||||
|
||||
st %f2, [$ivec + 0] ! write out ivec
|
||||
retl
|
||||
st %f3, [$ivec + 4]
|
||||
.type des_t4_ede3_cbc_decrypt,#function
|
||||
.size des_t4_ede3_cbc_decrypt,.-des_t4_ede3_cbc_decrypt
|
||||
___
|
||||
}
|
||||
$code.=<<___;
|
||||
.asciz "DES for SPARC T4, David S. Miller, Andy Polyakov"
|
||||
.align 4
|
||||
___
|
||||
|
||||
&emit_assembler();
|
||||
|
||||
close STDOUT;
|
||||
131
openssl-1.0.2f/crypto/des/asm/readme
Normal file
131
openssl-1.0.2f/crypto/des/asm/readme
Normal file
@@ -0,0 +1,131 @@
|
||||
First up, let me say I don't like writing in assembler. It is not portable,
|
||||
dependant on the particular CPU architecture release and is generally a pig
|
||||
to debug and get right. Having said that, the x86 architecture is probably
|
||||
the most important for speed due to number of boxes and since
|
||||
it appears to be the worst architecture to to get
|
||||
good C compilers for. So due to this, I have lowered myself to do
|
||||
assembler for the inner DES routines in libdes :-).
|
||||
|
||||
The file to implement in assembler is des_enc.c. Replace the following
|
||||
4 functions
|
||||
des_encrypt1(DES_LONG data[2],des_key_schedule ks, int encrypt);
|
||||
des_encrypt2(DES_LONG data[2],des_key_schedule ks, int encrypt);
|
||||
des_encrypt3(DES_LONG data[2],des_key_schedule ks1,ks2,ks3);
|
||||
des_decrypt3(DES_LONG data[2],des_key_schedule ks1,ks2,ks3);
|
||||
|
||||
They encrypt/decrypt the 64 bits held in 'data' using
|
||||
the 'ks' key schedules. The only difference between the 4 functions is that
|
||||
des_encrypt2() does not perform IP() or FP() on the data (this is an
|
||||
optimization for when doing triple DES and des_encrypt3() and des_decrypt3()
|
||||
perform triple des. The triple DES routines are in here because it does
|
||||
make a big difference to have them located near the des_encrypt2 function
|
||||
at link time..
|
||||
|
||||
Now as we all know, there are lots of different operating systems running on
|
||||
x86 boxes, and unfortunately they normally try to make sure their assembler
|
||||
formating is not the same as the other peoples.
|
||||
The 4 main formats I know of are
|
||||
Microsoft Windows 95/Windows NT
|
||||
Elf Includes Linux and FreeBSD(?).
|
||||
a.out The older Linux.
|
||||
Solaris Same as Elf but different comments :-(.
|
||||
|
||||
Now I was not overly keen to write 4 different copies of the same code,
|
||||
so I wrote a few perl routines to output the correct assembler, given
|
||||
a target assembler type. This code is ugly and is just a hack.
|
||||
The libraries are x86unix.pl and x86ms.pl.
|
||||
des586.pl, des686.pl and des-som[23].pl are the programs to actually
|
||||
generate the assembler.
|
||||
|
||||
So to generate elf assembler
|
||||
perl des-som3.pl elf >dx86-elf.s
|
||||
For Windows 95/NT
|
||||
perl des-som2.pl win32 >win32.asm
|
||||
|
||||
[ update 4 Jan 1996 ]
|
||||
I have added another way to do things.
|
||||
perl des-som3.pl cpp >dx86-cpp.s
|
||||
generates a file that will be included by dx86unix.cpp when it is compiled.
|
||||
To build for elf, a.out, solaris, bsdi etc,
|
||||
cc -E -DELF asm/dx86unix.cpp | as -o asm/dx86-elf.o
|
||||
cc -E -DSOL asm/dx86unix.cpp | as -o asm/dx86-sol.o
|
||||
cc -E -DOUT asm/dx86unix.cpp | as -o asm/dx86-out.o
|
||||
cc -E -DBSDI asm/dx86unix.cpp | as -o asm/dx86bsdi.o
|
||||
This was done to cut down the number of files in the distribution.
|
||||
|
||||
Now the ugly part. I acquired my copy of Intels
|
||||
"Optimization's For Intel's 32-Bit Processors" and found a few interesting
|
||||
things. First, the aim of the exersize is to 'extract' one byte at a time
|
||||
from a word and do an array lookup. This involves getting the byte from
|
||||
the 4 locations in the word and moving it to a new word and doing the lookup.
|
||||
The most obvious way to do this is
|
||||
xor eax, eax # clear word
|
||||
movb al, cl # get low byte
|
||||
xor edi DWORD PTR 0x100+des_SP[eax] # xor in word
|
||||
movb al, ch # get next byte
|
||||
xor edi DWORD PTR 0x300+des_SP[eax] # xor in word
|
||||
shr ecx 16
|
||||
which seems ok. For the pentium, this system appears to be the best.
|
||||
One has to do instruction interleaving to keep both functional units
|
||||
operating, but it is basically very efficient.
|
||||
|
||||
Now the crunch. When a full register is used after a partial write, eg.
|
||||
mov al, cl
|
||||
xor edi, DWORD PTR 0x100+des_SP[eax]
|
||||
386 - 1 cycle stall
|
||||
486 - 1 cycle stall
|
||||
586 - 0 cycle stall
|
||||
686 - at least 7 cycle stall (page 22 of the above mentioned document).
|
||||
|
||||
So the technique that produces the best results on a pentium, according to
|
||||
the documentation, will produce hideous results on a pentium pro.
|
||||
|
||||
To get around this, des686.pl will generate code that is not as fast on
|
||||
a pentium, should be very good on a pentium pro.
|
||||
mov eax, ecx # copy word
|
||||
shr ecx, 8 # line up next byte
|
||||
and eax, 0fch # mask byte
|
||||
xor edi DWORD PTR 0x100+des_SP[eax] # xor in array lookup
|
||||
mov eax, ecx # get word
|
||||
shr ecx 8 # line up next byte
|
||||
and eax, 0fch # mask byte
|
||||
xor edi DWORD PTR 0x300+des_SP[eax] # xor in array lookup
|
||||
|
||||
Due to the execution units in the pentium, this actually works quite well.
|
||||
For a pentium pro it should be very good. This is the type of output
|
||||
Visual C++ generates.
|
||||
|
||||
There is a third option. instead of using
|
||||
mov al, ch
|
||||
which is bad on the pentium pro, one may be able to use
|
||||
movzx eax, ch
|
||||
which may not incur the partial write penalty. On the pentium,
|
||||
this instruction takes 4 cycles so is not worth using but on the
|
||||
pentium pro it appears it may be worth while. I need access to one to
|
||||
experiment :-).
|
||||
|
||||
eric (20 Oct 1996)
|
||||
|
||||
22 Nov 1996 - I have asked people to run the 2 different version on pentium
|
||||
pros and it appears that the intel documentation is wrong. The
|
||||
mov al,bh is still faster on a pentium pro, so just use the des586.pl
|
||||
install des686.pl
|
||||
|
||||
3 Dec 1996 - I added des_encrypt3/des_decrypt3 because I have moved these
|
||||
functions into des_enc.c because it does make a massive performance
|
||||
difference on some boxes to have the functions code located close to
|
||||
the des_encrypt2() function.
|
||||
|
||||
9 Jan 1997 - des-som2.pl is now the correct perl script to use for
|
||||
pentiums. It contains an inner loop from
|
||||
Svend Olaf Mikkelsen <svolaf@inet.uni-c.dk> which does raw ecb DES calls at
|
||||
273,000 per second. He had a previous version at 250,000 and the best
|
||||
I was able to get was 203,000. The content has not changed, this is all
|
||||
due to instruction sequencing (and actual instructions choice) which is able
|
||||
to keep both functional units of the pentium going.
|
||||
We may have lost the ugly register usage restrictions when x86 went 32 bit
|
||||
but for the pentium it has been replaced by evil instruction ordering tricks.
|
||||
|
||||
13 Jan 1997 - des-som3.pl, more optimizations from Svend Olaf.
|
||||
raw DES at 281,000 per second on a pentium 100.
|
||||
|
||||
Reference in New Issue
Block a user