From 6b62d1fdd8a4fd35acfcc0c4bb1bf8b757fa8cda Mon Sep 17 00:00:00 2001 From: djm <> Date: Sat, 6 Sep 2008 12:17:54 +0000 Subject: resolve conflicts --- src/lib/libcrypto/sha/asm/sha1-586.pl | 472 ++++++++++------------------------ src/lib/libcrypto/sha/sha.h | 90 ++++++- src/lib/libcrypto/sha/sha1_one.c | 4 +- src/lib/libcrypto/sha/sha1dgst.c | 11 +- src/lib/libcrypto/sha/sha1test.c | 12 +- src/lib/libcrypto/sha/sha_dgst.c | 3 +- src/lib/libcrypto/sha/sha_locl.h | 306 ++++++++++------------ src/lib/libcrypto/sha/sha_one.c | 2 +- src/lib/libcrypto/sha/shatest.c | 12 +- 9 files changed, 369 insertions(+), 543 deletions(-) (limited to 'src/lib/libcrypto/sha') diff --git a/src/lib/libcrypto/sha/asm/sha1-586.pl b/src/lib/libcrypto/sha/asm/sha1-586.pl index 041acc0348..0b4dab2bd5 100644 --- a/src/lib/libcrypto/sha/asm/sha1-586.pl +++ b/src/lib/libcrypto/sha/asm/sha1-586.pl @@ -1,4 +1,16 @@ -#!/usr/local/bin/perl +#!/usr/bin/env perl + +# ==================================================================== +# [Re]written by Andy Polyakov 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* @@ -9,7 +21,7 @@ # # compared with original compared with Intel cc # assembler impl. generated code -# Pentium -25% +37% +# Pentium -16% +48% # PIII/AMD +8% +16% # P4 +85%(!) +45% # @@ -17,174 +29,115 @@ # 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. -# ---------------------------------------------------------------- # -$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 $n"); - &mov($tmp1,$a); - &mov($f,$c); # f to hold F_00_19(b,c,d) + &mov($f,$c); # f to hold F_00_19(b,c,d) + if ($n==0) { &mov($tmp1,$a); } + else { &mov($a,$tmp1); } &rotl($tmp1,5); # tmp1=ROTATE(a,5) &xor($f,$d); - &and($f,$b); - &rotr($b,2); # b=ROTATE(b,30) &add($tmp1,$e); # tmp1+=e; - &mov($e,&swtmp($n)); # e becomes volatile and - # is loaded with xi - &xor($f,$d); # f holds F_00_19(b,c,d) - &lea($tmp1,&DWP($K,$tmp1,$e,1));# tmp1+=K_00_19+xi - - &add($f,$tmp1); # f+=tmp1 + &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(0x5a827999,$tmp1,$e)); # tmp1+=K_00_19+xi + + if ($n==15) { &add($f,$tmp1); } # f+=tmp1 + else { &add($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,$b); # tmp1 holds F_00_19(b,c,d) - &xor($f,&swtmp($n3)); # f holds xa^xb^xc^xd - &rotr($b,2); # b=ROTATE(b,30) - &xor($tmp1,$d); # tmp1=F_00_19(b,c,d) - &rotl($f,1); # f=ROATE(f,1) - &mov(&swtmp($n0),$f); # xi=f - &lea($f,&DWP($K,$f,$e,1)); # f+=K_00_19+e + &rotr($b,2); # b=ROTATE(b,30) + &xor($f,&swtmp(($n+13)%16)); # f holds xa^xb^xc^xd + &rotl($f,1); # f=ROTATE(f,1) + &xor($tmp1,$d); # tmp1=F_00_19(b,c,d) + &mov(&swtmp($n%16),$f); # xi=f + &lea($f,&DWP(0x5a827999,$f,$e));# f+=K_00_19+e &mov($e,$a); # e becomes volatile - &add($f,$tmp1); # f+=F_00_19(b,c,d) - &rotl($e,5); # e=ROTATE(a,5) + &rotl($e,5); # e=ROTATE(a,5) + &add($f,$tmp1); # f+=F_00_19(b,c,d) &add($f,$e); # f+=ROTATE(a,5) } 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($n); - &mov($f,&swtmp($n0)); # f to hold Xupdate(xi,xa,xb,xc,xd) - &mov($tmp1,$b); # tmp1 to hold F_20_39(b,c,d) - &xor($f,&swtmp($n1)); - &rotr($b,2); # b=ROTATE(b,30) - &xor($f,&swtmp($n2)); - &xor($tmp1,$c); - &xor($f,&swtmp($n3)); # f holds xa^xb^xc^xd - &xor($tmp1,$d); # tmp1 holds F_20_39(b,c,d) + &mov($tmp1,$b); # tmp1 to hold F_20_39(b,c,d) + &mov($f,&swtmp($n%16)); # f to hold Xupdate(xi,xa,xb,xc,xd) + &rotr($b,2); # b=ROTATE(b,30) + &xor($f,&swtmp(($n+2)%16)); + &xor($tmp1,$c); + &xor($f,&swtmp(($n+8)%16)); + &xor($tmp1,$d); # tmp1 holds F_20_39(b,c,d) + &xor($f,&swtmp(($n+13)%16)); # f holds xa^xb^xc^xd &rotl($f,1); # f=ROTATE(f,1) - &mov(&swtmp($n0),$f); # xi=f - &lea($f,&DWP($K,$f,$e,1)); # f+=K_20_39+e + &add($tmp1,$e); + &mov(&swtmp($n%16),$f); # xi=f &mov($e,$a); # e becomes volatile &rotl($e,5); # e=ROTATE(a,5) - &add($f,$tmp1); # f+=F_20_39(b,c,d) + &lea($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($n0)); # f to hold Xupdate(xi,xa,xb,xc,xd) + &mov($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(($n+8)%16)); + &xor($f,$tmp1); + &mov($tmp1,&swtmp(($n+13)%16)); + &xor($f,$tmp1); # f holds xa^xb^xc^xd &mov($tmp1,$b); # tmp1 to hold F_40_59(b,c,d) - &xor($f,&swtmp($n1)); + &rotl($f,1); # f=ROTATE(f,1) &or($tmp1,$c); - &xor($f,&swtmp($n2)); + &mov(&swtmp($n%16),$f); # xi=f &and($tmp1,$d); - &xor($f,&swtmp($n3)); # f holds xa^xb^xc^xd - &rotl($f,1); # f=ROTATE(f,1) - &mov(&swtmp($n0),$f); # xi=f - &lea($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($b,2); # b=ROTATE(b,30) @@ -192,234 +145,75 @@ sub BODY_40_59 &or($tmp1,$e); # tmp1 holds F_40_59(b,c,d) &mov($e,$a); &rotl($e,5); # e=ROTATE(a,5) - &add($tmp1,$e); # tmp1+=ROTATE(a,5) - &add($f,$tmp1); # f+=tmp1; - } - -sub BODY_60_79 - { - &BODY_20_39(@_); + &add($f,$tmp1); # f+=tmp1; + &add($f,$e); # f+=ROTATE(a,5) } -sub sha1_block_host - { - local($name, $sclabel)=@_; +&function_begin("sha1_block_data_order",16); + &mov($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 - &function_begin_B($name,""); + &set_label("loop",16); - # parameter 1 is the MD5_CTX structure. - # A 0 - # B 4 - # C 8 - # D 12 - # E 16 - - &mov("ecx", &wparam(2)); - &push("esi"); - &shl("ecx",6); - &mov("esi", &wparam(1)); - &push("ebp"); - &add("ecx","esi"); # offset to leave on - &push("ebx"); - &mov("ebp", &wparam(0)); - &push("edi"); - &mov($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"); - - 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/src/lib/libcrypto/sha/sha.h b/src/lib/libcrypto/sha/sha.h index 79c07b0fd1..eed44d7f94 100644 --- a/src/lib/libcrypto/sha/sha.h +++ b/src/lib/libcrypto/sha/sha.h @@ -60,6 +60,7 @@ #define HEADER_SHA_H #include +#include #ifdef __cplusplus extern "C" { @@ -70,7 +71,7 @@ extern "C" { #endif #if defined(OPENSSL_FIPS) -#define FIPS_SHA_SIZE_T unsigned long +#define FIPS_SHA_SIZE_T size_t #endif /* @@ -101,26 +102,97 @@ typedef struct SHAstate_st SHA_LONG h0,h1,h2,h3,h4; SHA_LONG Nl,Nh; SHA_LONG data[SHA_LBLOCK]; - int num; + unsigned int num; } SHA_CTX; #ifndef OPENSSL_NO_SHA0 -#ifdef OPENSSL_FIPS -int private_SHA_Init(SHA_CTX *c); -#endif int SHA_Init(SHA_CTX *c); -int SHA_Update(SHA_CTX *c, const void *data, unsigned long len); +int SHA_Update(SHA_CTX *c, const void *data, size_t len); int SHA_Final(unsigned char *md, SHA_CTX *c); -unsigned char *SHA(const unsigned char *d, unsigned long n,unsigned char *md); +unsigned char *SHA(const unsigned char *d, size_t n, unsigned char *md); void SHA_Transform(SHA_CTX *c, const unsigned char *data); #endif #ifndef OPENSSL_NO_SHA1 int SHA1_Init(SHA_CTX *c); -int SHA1_Update(SHA_CTX *c, const void *data, unsigned long len); +int SHA1_Update(SHA_CTX *c, const void *data, size_t len); int SHA1_Final(unsigned char *md, SHA_CTX *c); -unsigned char *SHA1(const unsigned char *d, unsigned long n,unsigned char *md); +unsigned char *SHA1(const unsigned char *d, size_t n, unsigned char *md); void SHA1_Transform(SHA_CTX *c, const unsigned char *data); #endif + +#define SHA256_CBLOCK (SHA_LBLOCK*4) /* SHA-256 treats input data as a + * contiguous array of 32 bit + * wide big-endian values. */ +#define SHA224_DIGEST_LENGTH 28 +#define SHA256_DIGEST_LENGTH 32 + +typedef struct SHA256state_st + { + SHA_LONG h[8]; + SHA_LONG Nl,Nh; + SHA_LONG data[SHA_LBLOCK]; + unsigned int num,md_len; + } SHA256_CTX; + +#ifndef OPENSSL_NO_SHA256 +int SHA224_Init(SHA256_CTX *c); +int SHA224_Update(SHA256_CTX *c, const void *data, size_t len); +int SHA224_Final(unsigned char *md, SHA256_CTX *c); +unsigned char *SHA224(const unsigned char *d, size_t n,unsigned char *md); +int SHA256_Init(SHA256_CTX *c); +int SHA256_Update(SHA256_CTX *c, const void *data, size_t len); +int SHA256_Final(unsigned char *md, SHA256_CTX *c); +unsigned char *SHA256(const unsigned char *d, size_t n,unsigned char *md); +void SHA256_Transform(SHA256_CTX *c, const unsigned char *data); +#endif + +#define SHA384_DIGEST_LENGTH 48 +#define SHA512_DIGEST_LENGTH 64 + +#ifndef OPENSSL_NO_SHA512 +/* + * Unlike 32-bit digest algorithms, SHA-512 *relies* on SHA_LONG64 + * being exactly 64-bit wide. See Implementation Notes in sha512.c + * for further details. + */ +#define SHA512_CBLOCK (SHA_LBLOCK*8) /* SHA-512 treats input data as a + * contiguous array of 64 bit + * wide big-endian values. */ +#if (defined(_WIN32) || defined(_WIN64)) && !defined(__MINGW32__) +#define SHA_LONG64 unsigned __int64 +#define U64(C) C##UI64 +#elif defined(__arch64__) +#define SHA_LONG64 unsigned long +#define U64(C) C##UL +#else +#define SHA_LONG64 unsigned long long +#define U64(C) C##ULL +#endif + +typedef struct SHA512state_st + { + SHA_LONG64 h[8]; + SHA_LONG64 Nl,Nh; + union { + SHA_LONG64 d[SHA_LBLOCK]; + unsigned char p[SHA512_CBLOCK]; + } u; + unsigned int num,md_len; + } SHA512_CTX; +#endif + +#ifndef OPENSSL_NO_SHA512 +int SHA384_Init(SHA512_CTX *c); +int SHA384_Update(SHA512_CTX *c, const void *data, size_t len); +int SHA384_Final(unsigned char *md, SHA512_CTX *c); +unsigned char *SHA384(const unsigned char *d, size_t n,unsigned char *md); +int SHA512_Init(SHA512_CTX *c); +int SHA512_Update(SHA512_CTX *c, const void *data, size_t len); +int SHA512_Final(unsigned char *md, SHA512_CTX *c); +unsigned char *SHA512(const unsigned char *d, size_t n,unsigned char *md); +void SHA512_Transform(SHA512_CTX *c, const unsigned char *data); +#endif + #ifdef __cplusplus } #endif diff --git a/src/lib/libcrypto/sha/sha1_one.c b/src/lib/libcrypto/sha/sha1_one.c index f4694b701b..7c65b60276 100644 --- a/src/lib/libcrypto/sha/sha1_one.c +++ b/src/lib/libcrypto/sha/sha1_one.c @@ -61,8 +61,8 @@ #include #include -#if !defined(OPENSSL_NO_SHA1) && !defined(OPENSSL_FIPS) -unsigned char *SHA1(const unsigned char *d, unsigned long n, unsigned char *md) +#ifndef OPENSSL_NO_SHA1 +unsigned char *SHA1(const unsigned char *d, size_t n, unsigned char *md) { SHA_CTX c; static unsigned char m[SHA_DIGEST_LENGTH]; diff --git a/src/lib/libcrypto/sha/sha1dgst.c b/src/lib/libcrypto/sha/sha1dgst.c index 1e2009b760..50d1925cde 100644 --- a/src/lib/libcrypto/sha/sha1dgst.c +++ b/src/lib/libcrypto/sha/sha1dgst.c @@ -56,26 +56,19 @@ * [including the GNU Public Licence.] */ +#include #if !defined(OPENSSL_NO_SHA1) && !defined(OPENSSL_NO_SHA) #undef SHA_0 #define SHA_1 #include -#include -#ifndef OPENSSL_FIPS -const char *SHA1_version="SHA1" OPENSSL_VERSION_PTEXT; +const char SHA1_version[]="SHA1" OPENSSL_VERSION_PTEXT; /* The implementation is in ../md32_common.h */ #include "sha_locl.h" -#else /* ndef OPENSSL_FIPS */ - -static void *dummy=&dummy; - -#endif /* ndef OPENSSL_FIPS */ - #endif diff --git a/src/lib/libcrypto/sha/sha1test.c b/src/lib/libcrypto/sha/sha1test.c index 4f2e4ada2d..6feb3964c7 100644 --- a/src/lib/libcrypto/sha/sha1test.c +++ b/src/lib/libcrypto/sha/sha1test.c @@ -106,7 +106,7 @@ static char *pt(unsigned char *md); int main(int argc, char *argv[]) { int i,err=0; - unsigned char **P,**R; + char **P,**R; static unsigned char buf[1000]; char *p,*r; EVP_MD_CTX c; @@ -118,12 +118,12 @@ int main(int argc, char *argv[]) #endif EVP_MD_CTX_init(&c); - P=(unsigned char **)test; - R=(unsigned char **)ret; + P=test; + R=ret; i=1; while (*P != NULL) { - EVP_Digest(*P,(unsigned long)strlen((char *)*P),md,NULL,EVP_sha1(), NULL); + EVP_Digest(*P,strlen((char *)*P),md,NULL,EVP_sha1(), NULL); p=pt(md); if (strcmp(p,(char *)*R) != 0) { @@ -157,6 +157,10 @@ int main(int argc, char *argv[]) } else printf("test 3 ok\n"); + +#ifdef OPENSSL_SYS_NETWARE + if (err) printf("ERROR: %d\n", err); +#endif EXIT(err); EVP_MD_CTX_cleanup(&c); return(0); diff --git a/src/lib/libcrypto/sha/sha_dgst.c b/src/lib/libcrypto/sha/sha_dgst.c index 5a4b3ab204..70eb56032c 100644 --- a/src/lib/libcrypto/sha/sha_dgst.c +++ b/src/lib/libcrypto/sha/sha_dgst.c @@ -56,6 +56,7 @@ * [including the GNU Public Licence.] */ +#include #if !defined(OPENSSL_NO_SHA0) && !defined(OPENSSL_NO_SHA) #undef SHA_1 @@ -63,7 +64,7 @@ #include -const char *SHA_version="SHA" OPENSSL_VERSION_PTEXT; +const char SHA_version[]="SHA" OPENSSL_VERSION_PTEXT; /* The implementation is in ../md32_common.h */ diff --git a/src/lib/libcrypto/sha/sha_locl.h b/src/lib/libcrypto/sha/sha_locl.h index a3623f72da..e37e5726e3 100644 --- a/src/lib/libcrypto/sha/sha_locl.h +++ b/src/lib/libcrypto/sha/sha_locl.h @@ -62,17 +62,11 @@ #include #include -#ifndef SHA_LONG_LOG2 -#define SHA_LONG_LOG2 2 /* default to 32 bits */ -#endif - #define DATA_ORDER_IS_BIG_ENDIAN #define HASH_LONG SHA_LONG -#define HASH_LONG_LOG2 SHA_LONG_LOG2 #define HASH_CTX SHA_CTX #define HASH_CBLOCK SHA_CBLOCK -#define HASH_LBLOCK SHA_LBLOCK #define HASH_MAKE_STRING(c,s) do { \ unsigned long ll; \ ll=(c)->h0; HOST_l2c(ll,(s)); \ @@ -88,12 +82,10 @@ # define HASH_TRANSFORM SHA_Transform # define HASH_FINAL SHA_Final # define HASH_INIT SHA_Init -# define HASH_BLOCK_HOST_ORDER sha_block_host_order # define HASH_BLOCK_DATA_ORDER sha_block_data_order # define Xupdate(a,ix,ia,ib,ic,id) (ix=(a)=(ia^ib^ic^id)) - void sha_block_host_order (SHA_CTX *c, const void *p,int num); - void sha_block_data_order (SHA_CTX *c, const void *p,int num); +static void sha_block_data_order (SHA_CTX *c, const void *p,size_t num); #elif defined(SHA_1) @@ -101,7 +93,6 @@ # define HASH_TRANSFORM SHA1_Transform # define HASH_FINAL SHA1_Final # define HASH_INIT SHA1_Init -# define HASH_BLOCK_HOST_ORDER sha1_block_host_order # define HASH_BLOCK_DATA_ORDER sha1_block_data_order # if defined(__MWERKS__) && defined(__MC68K__) /* Metrowerks for Motorola fails otherwise:-( */ @@ -114,22 +105,10 @@ ) # endif -# ifdef SHA1_ASM -# if defined(__i386) || defined(__i386__) || defined(_M_IX86) || defined(__INTEL__) -# define sha1_block_host_order sha1_block_asm_host_order -# define DONT_IMPLEMENT_BLOCK_HOST_ORDER -# define sha1_block_data_order sha1_block_asm_data_order -# define DONT_IMPLEMENT_BLOCK_DATA_ORDER -# define HASH_BLOCK_DATA_ORDER_ALIGNED sha1_block_asm_data_order -# elif defined(__ia64) || defined(__ia64__) || defined(_M_IA64) -# define sha1_block_host_order sha1_block_asm_host_order -# define DONT_IMPLEMENT_BLOCK_HOST_ORDER -# define sha1_block_data_order sha1_block_asm_data_order -# define DONT_IMPLEMENT_BLOCK_DATA_ORDER -# endif -# endif - void sha1_block_host_order (SHA_CTX *c, const void *p,int num); - void sha1_block_data_order (SHA_CTX *c, const void *p,int num); +#ifndef SHA1_ASM +static +#endif +void sha1_block_data_order (SHA_CTX *c, const void *p,size_t num); #else # error "Either SHA_0 or SHA_1 must be defined." @@ -143,11 +122,7 @@ #define INIT_DATA_h3 0x10325476UL #define INIT_DATA_h4 0xc3d2e1f0UL -#if defined(SHA_0) && defined(OPENSSL_FIPS) -FIPS_NON_FIPS_MD_Init(SHA) -#else int HASH_INIT (SHA_CTX *c) -#endif { c->h0=INIT_DATA_h0; c->h1=INIT_DATA_h1; @@ -177,6 +152,8 @@ int HASH_INIT (SHA_CTX *c) #define F_40_59(b,c,d) (((b) & (c)) | (((b)|(c)) & (d))) #define F_60_79(b,c,d) F_20_39(b,c,d) +#ifndef OPENSSL_SMALL_FOOTPRINT + #define BODY_00_15(i,a,b,c,d,e,f,xi) \ (f)=xi+(e)+K_00_19+ROTATE((a),5)+F_00_19((b),(c),(d)); \ (b)=ROTATE((b),30); @@ -229,11 +206,11 @@ int HASH_INIT (SHA_CTX *c) # define X(i) XX[i] #endif -#ifndef DONT_IMPLEMENT_BLOCK_HOST_ORDER -void HASH_BLOCK_HOST_ORDER (SHA_CTX *c, const void *d, int num) +#if !defined(SHA_1) || !defined(SHA1_ASM) +static void HASH_BLOCK_DATA_ORDER (SHA_CTX *c, const void *p, size_t num) { - const SHA_LONG *W=d; - register unsigned MD32_REG_T A,B,C,D,E,T; + const unsigned char *data=p; + register unsigned MD32_REG_T A,B,C,D,E,T,l; #ifndef MD32_XARRAY unsigned MD32_REG_T XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7, XX8, XX9,XX10,XX11,XX12,XX13,XX14,XX15; @@ -248,41 +225,71 @@ void HASH_BLOCK_HOST_ORDER (SHA_CTX *c, const void *d, int num) E=c->h4; for (;;) + { + const union { long one; char little; } is_endian = {1}; + + if (!is_endian.little && sizeof(SHA_LONG)==4 && ((size_t)p%4)==0) + { + const SHA_LONG *W=(const SHA_LONG *)data; + + X( 0) = W[0]; X( 1) = W[ 1]; + BODY_00_15( 0,A,B,C,D,E,T,X( 0)); X( 2) = W[ 2]; + BODY_00_15( 1,T,A,B,C,D,E,X( 1)); X( 3) = W[ 3]; + BODY_00_15( 2,E,T,A,B,C,D,X( 2)); X( 4) = W[ 4]; + BODY_00_15( 3,D,E,T,A,B,C,X( 3)); X( 5) = W[ 5]; + BODY_00_15( 4,C,D,E,T,A,B,X( 4)); X( 6) = W[ 6]; + BODY_00_15( 5,B,C,D,E,T,A,X( 5)); X( 7) = W[ 7]; + BODY_00_15( 6,A,B,C,D,E,T,X( 6)); X( 8) = W[ 8]; + BODY_00_15( 7,T,A,B,C,D,E,X( 7)); X( 9) = W[ 9]; + BODY_00_15( 8,E,T,A,B,C,D,X( 8)); X(10) = W[10]; + BODY_00_15( 9,D,E,T,A,B,C,X( 9)); X(11) = W[11]; + BODY_00_15(10,C,D,E,T,A,B,X(10)); X(12) = W[12]; + BODY_00_15(11,B,C,D,E,T,A,X(11)); X(13) = W[13]; + BODY_00_15(12,A,B,C,D,E,T,X(12)); X(14) = W[14]; + BODY_00_15(13,T,A,B,C,D,E,X(13)); X(15) = W[15]; + BODY_00_15(14,E,T,A,B,C,D,X(14)); + BODY_00_15(15,D,E,T,A,B,C,X(15)); + + data += SHA_CBLOCK; + } + else { - BODY_00_15( 0,A,B,C,D,E,T,W[ 0]); - BODY_00_15( 1,T,A,B,C,D,E,W[ 1]); - BODY_00_15( 2,E,T,A,B,C,D,W[ 2]); - BODY_00_15( 3,D,E,T,A,B,C,W[ 3]); - BODY_00_15( 4,C,D,E,T,A,B,W[ 4]); - BODY_00_15( 5,B,C,D,E,T,A,W[ 5]); - BODY_00_15( 6,A,B,C,D,E,T,W[ 6]); - BODY_00_15( 7,T,A,B,C,D,E,W[ 7]); - BODY_00_15( 8,E,T,A,B,C,D,W[ 8]); - BODY_00_15( 9,D,E,T,A,B,C,W[ 9]); - BODY_00_15(10,C,D,E,T,A,B,W[10]); - BODY_00_15(11,B,C,D,E,T,A,W[11]); - BODY_00_15(12,A,B,C,D,E,T,W[12]); - BODY_00_15(13,T,A,B,C,D,E,W[13]); - BODY_00_15(14,E,T,A,B,C,D,W[14]); - BODY_00_15(15,D,E,T,A,B,C,W[15]); - - BODY_16_19(16,C,D,E,T,A,B,X( 0),W[ 0],W[ 2],W[ 8],W[13]); - BODY_16_19(17,B,C,D,E,T,A,X( 1),W[ 1],W[ 3],W[ 9],W[14]); - BODY_16_19(18,A,B,C,D,E,T,X( 2),W[ 2],W[ 4],W[10],W[15]); - BODY_16_19(19,T,A,B,C,D,E,X( 3),W[ 3],W[ 5],W[11],X( 0)); - - BODY_20_31(20,E,T,A,B,C,D,X( 4),W[ 4],W[ 6],W[12],X( 1)); - BODY_20_31(21,D,E,T,A,B,C,X( 5),W[ 5],W[ 7],W[13],X( 2)); - BODY_20_31(22,C,D,E,T,A,B,X( 6),W[ 6],W[ 8],W[14],X( 3)); - BODY_20_31(23,B,C,D,E,T,A,X( 7),W[ 7],W[ 9],W[15],X( 4)); - BODY_20_31(24,A,B,C,D,E,T,X( 8),W[ 8],W[10],X( 0),X( 5)); - BODY_20_31(25,T,A,B,C,D,E,X( 9),W[ 9],W[11],X( 1),X( 6)); - BODY_20_31(26,E,T,A,B,C,D,X(10),W[10],W[12],X( 2),X( 7)); - BODY_20_31(27,D,E,T,A,B,C,X(11),W[11],W[13],X( 3),X( 8)); - BODY_20_31(28,C,D,E,T,A,B,X(12),W[12],W[14],X( 4),X( 9)); - BODY_20_31(29,B,C,D,E,T,A,X(13),W[13],W[15],X( 5),X(10)); - BODY_20_31(30,A,B,C,D,E,T,X(14),W[14],X( 0),X( 6),X(11)); - BODY_20_31(31,T,A,B,C,D,E,X(15),W[15],X( 1),X( 7),X(12)); + HOST_c2l(data,l); X( 0)=l; HOST_c2l(data,l); X( 1)=l; + BODY_00_15( 0,A,B,C,D,E,T,X( 0)); HOST_c2l(data,l); X( 2)=l; + BODY_00_15( 1,T,A,B,C,D,E,X( 1)); HOST_c2l(data,l); X( 3)=l; + BODY_00_15( 2,E,T,A,B,C,D,X( 2)); HOST_c2l(data,l); X( 4)=l; + BODY_00_15( 3,D,E,T,A,B,C,X( 3)); HOST_c2l(data,l); X( 5)=l; + BODY_00_15( 4,C,D,E,T,A,B,X( 4)); HOST_c2l(data,l); X( 6)=l; + BODY_00_15( 5,B,C,D,E,T,A,X( 5)); HOST_c2l(data,l); X( 7)=l; + BODY_00_15( 6,A,B,C,D,E,T,X( 6)); HOST_c2l(data,l); X( 8)=l; + BODY_00_15( 7,T,A,B,C,D,E,X( 7)); HOST_c2l(data,l); X( 9)=l; + BODY_00_15( 8,E,T,A,B,C,D,X( 8)); HOST_c2l(data,l); X(10)=l; + BODY_00_15( 9,D,E,T,A,B,C,X( 9)); HOST_c2l(data,l); X(11)=l; + BODY_00_15(10,C,D,E,T,A,B,X(10)); HOST_c2l(data,l); X(12)=l; + BODY_00_15(11,B,C,D,E,T,A,X(11)); HOST_c2l(data,l); X(13)=l; + BODY_00_15(12,A,B,C,D,E,T,X(12)); HOST_c2l(data,l); X(14)=l; + BODY_00_15(13,T,A,B,C,D,E,X(13)); HOST_c2l(data,l); X(15)=l; + BODY_00_15(14,E,T,A,B,C,D,X(14)); + BODY_00_15(15,D,E,T,A,B,C,X(15)); + } + + BODY_16_19(16,C,D,E,T,A,B,X( 0),X( 0),X( 2),X( 8),X(13)); + BODY_16_19(17,B,C,D,E,T,A,X( 1),X( 1),X( 3),X( 9),X(14)); + BODY_16_19(18,A,B,C,D,E,T,X( 2),X( 2),X( 4),X(10),X(15)); + BODY_16_19(19,T,A,B,C,D,E,X( 3),X( 3),X( 5),X(11),X( 0)); + + BODY_20_31(20,E,T,A,B,C,D,X( 4),X( 4),X( 6),X(12),X( 1)); + BODY_20_31(21,D,E,T,A,B,C,X( 5),X( 5),X( 7),X(13),X( 2)); + BODY_20_31(22,C,D,E,T,A,B,X( 6),X( 6),X( 8),X(14),X( 3)); + BODY_20_31(23,B,C,D,E,T,A,X( 7),X( 7),X( 9),X(15),X( 4)); + BODY_20_31(24,A,B,C,D,E,T,X( 8),X( 8),X(10),X( 0),X( 5)); + BODY_20_31(25,T,A,B,C,D,E,X( 9),X( 9),X(11),X( 1),X( 6)); + BODY_20_31(26,E,T,A,B,C,D,X(10),X(10),X(12),X( 2),X( 7)); + BODY_20_31(27,D,E,T,A,B,C,X(11),X(11),X(13),X( 3),X( 8)); + BODY_20_31(28,C,D,E,T,A,B,X(12),X(12),X(14),X( 4),X( 9)); + BODY_20_31(29,B,C,D,E,T,A,X(13),X(13),X(15),X( 5),X(10)); + BODY_20_31(30,A,B,C,D,E,T,X(14),X(14),X( 0),X( 6),X(11)); + BODY_20_31(31,T,A,B,C,D,E,X(15),X(15),X( 1),X( 7),X(12)); BODY_32_39(32,E,T,A,B,C,D,X( 0),X( 2),X( 8),X(13)); BODY_32_39(33,D,E,T,A,B,C,X( 1),X( 3),X( 9),X(14)); @@ -341,7 +348,7 @@ void HASH_BLOCK_HOST_ORDER (SHA_CTX *c, const void *d, int num) c->h3=(c->h3+B)&0xffffffffL; c->h4=(c->h4+C)&0xffffffffL; - if (--num <= 0) break; + if (--num == 0) break; A=c->h0; B=c->h1; @@ -349,22 +356,48 @@ void HASH_BLOCK_HOST_ORDER (SHA_CTX *c, const void *d, int num) D=c->h3; E=c->h4; - W+=SHA_LBLOCK; - } + } } #endif -#ifndef DONT_IMPLEMENT_BLOCK_DATA_ORDER -void HASH_BLOCK_DATA_ORDER (SHA_CTX *c, const void *p, int num) +#else /* OPENSSL_SMALL_FOOTPRINT */ + +#define BODY_00_15(xi) do { \ + T=E+K_00_19+F_00_19(B,C,D); \ + E=D, D=C, C=ROTATE(B,30), B=A; \ + A=ROTATE(A,5)+T+xi; } while(0) + +#define BODY_16_19(xa,xb,xc,xd) do { \ + Xupdate(T,xa,xa,xb,xc,xd); \ + T+=E+K_00_19+F_00_19(B,C,D); \ + E=D, D=C, C=ROTATE(B,30), B=A; \ + A=ROTATE(A,5)+T; } while(0) + +#define BODY_20_39(xa,xb,xc,xd) do { \ + Xupdate(T,xa,xa,xb,xc,xd); \ + T+=E+K_20_39+F_20_39(B,C,D); \ + E=D, D=C, C=ROTATE(B,30), B=A; \ + A=ROTATE(A,5)+T; } while(0) + +#define BODY_40_59(xa,xb,xc,xd) do { \ + Xupdate(T,xa,xa,xb,xc,xd); \ + T+=E+K_40_59+F_40_59(B,C,D); \ + E=D, D=C, C=ROTATE(B,30), B=A; \ + A=ROTATE(A,5)+T; } while(0) + +#define BODY_60_79(xa,xb,xc,xd) do { \ + Xupdate(T,xa,xa,xb,xc,xd); \ + T=E+K_60_79+F_60_79(B,C,D); \ + E=D, D=C, C=ROTATE(B,30), B=A; \ + A=ROTATE(A,5)+T+xa; } while(0) + +#if !defined(SHA_1) || !defined(SHA1_ASM) +static void HASH_BLOCK_DATA_ORDER (SHA_CTX *c, const void *p, size_t num) { const unsigned char *data=p; register unsigned MD32_REG_T A,B,C,D,E,T,l; -#ifndef MD32_XARRAY - unsigned MD32_REG_T XX0, XX1, XX2, XX3, XX4, XX5, XX6, XX7, - XX8, XX9,XX10,XX11,XX12,XX13,XX14,XX15; -#else - SHA_LONG XX[16]; -#endif + int i; + SHA_LONG X[16]; A=c->h0; B=c->h1; @@ -374,101 +407,24 @@ void HASH_BLOCK_DATA_ORDER (SHA_CTX *c, const void *p, int num) for (;;) { - - HOST_c2l(data,l); X( 0)=l; HOST_c2l(data,l); X( 1)=l; - BODY_00_15( 0,A,B,C,D,E,T,X( 0)); HOST_c2l(data,l); X( 2)=l; - BODY_00_15( 1,T,A,B,C,D,E,X( 1)); HOST_c2l(data,l); X( 3)=l; - BODY_00_15( 2,E,T,A,B,C,D,X( 2)); HOST_c2l(data,l); X( 4)=l; - BODY_00_15( 3,D,E,T,A,B,C,X( 3)); HOST_c2l(data,l); X( 5)=l; - BODY_00_15( 4,C,D,E,T,A,B,X( 4)); HOST_c2l(data,l); X( 6)=l; - BODY_00_15( 5,B,C,D,E,T,A,X( 5)); HOST_c2l(data,l); X( 7)=l; - BODY_00_15( 6,A,B,C,D,E,T,X( 6)); HOST_c2l(data,l); X( 8)=l; - BODY_00_15( 7,T,A,B,C,D,E,X( 7)); HOST_c2l(data,l); X( 9)=l; - BODY_00_15( 8,E,T,A,B,C,D,X( 8)); HOST_c2l(data,l); X(10)=l; - BODY_00_15( 9,D,E,T,A,B,C,X( 9)); HOST_c2l(data,l); X(11)=l; - BODY_00_15(10,C,D,E,T,A,B,X(10)); HOST_c2l(data,l); X(12)=l; - BODY_00_15(11,B,C,D,E,T,A,X(11)); HOST_c2l(data,l); X(13)=l; - BODY_00_15(12,A,B,C,D,E,T,X(12)); HOST_c2l(data,l); X(14)=l; - BODY_00_15(13,T,A,B,C,D,E,X(13)); HOST_c2l(data,l); X(15)=l; - BODY_00_15(14,E,T,A,B,C,D,X(14)); - BODY_00_15(15,D,E,T,A,B,C,X(15)); - - BODY_16_19(16,C,D,E,T,A,B,X( 0),X( 0),X( 2),X( 8),X(13)); - BODY_16_19(17,B,C,D,E,T,A,X( 1),X( 1),X( 3),X( 9),X(14)); - BODY_16_19(18,A,B,C,D,E,T,X( 2),X( 2),X( 4),X(10),X(15)); - BODY_16_19(19,T,A,B,C,D,E,X( 3),X( 3),X( 5),X(11),X( 0)); - - BODY_20_31(20,E,T,A,B,C,D,X( 4),X( 4),X( 6),X(12),X( 1)); - BODY_20_31(21,D,E,T,A,B,C,X( 5),X( 5),X( 7),X(13),X( 2)); - BODY_20_31(22,C,D,E,T,A,B,X( 6),X( 6),X( 8),X(14),X( 3)); - BODY_20_31(23,B,C,D,E,T,A,X( 7),X( 7),X( 9),X(15),X( 4)); - BODY_20_31(24,A,B,C,D,E,T,X( 8),X( 8),X(10),X( 0),X( 5)); - BODY_20_31(25,T,A,B,C,D,E,X( 9),X( 9),X(11),X( 1),X( 6)); - BODY_20_31(26,E,T,A,B,C,D,X(10),X(10),X(12),X( 2),X( 7)); - BODY_20_31(27,D,E,T,A,B,C,X(11),X(11),X(13),X( 3),X( 8)); - BODY_20_31(28,C,D,E,T,A,B,X(12),X(12),X(14),X( 4),X( 9)); - BODY_20_31(29,B,C,D,E,T,A,X(13),X(13),X(15),X( 5),X(10)); - BODY_20_31(30,A,B,C,D,E,T,X(14),X(14),X( 0),X( 6),X(11)); - BODY_20_31(31,T,A,B,C,D,E,X(15),X(15),X( 1),X( 7),X(12)); - - BODY_32_39(32,E,T,A,B,C,D,X( 0),X( 2),X( 8),X(13)); - BODY_32_39(33,D,E,T,A,B,C,X( 1),X( 3),X( 9),X(14)); - BODY_32_39(34,C,D,E,T,A,B,X( 2),X( 4),X(10),X(15)); - BODY_32_39(35,B,C,D,E,T,A,X( 3),X( 5),X(11),X( 0)); - BODY_32_39(36,A,B,C,D,E,T,X( 4),X( 6),X(12),X( 1)); - BODY_32_39(37,T,A,B,C,D,E,X( 5),X( 7),X(13),X( 2)); - BODY_32_39(38,E,T,A,B,C,D,X( 6),X( 8),X(14),X( 3)); - BODY_32_39(39,D,E,T,A,B,C,X( 7),X( 9),X(15),X( 4)); - - BODY_40_59(40,C,D,E,T,A,B,X( 8),X(10),X( 0),X( 5)); - BODY_40_59(41,B,C,D,E,T,A,X( 9),X(11),X( 1),X( 6)); - BODY_40_59(42,A,B,C,D,E,T,X(10),X(12),X( 2),X( 7)); - BODY_40_59(43,T,A,B,C,D,E,X(11),X(13),X( 3),X( 8)); - BODY_40_59(44,E,T,A,B,C,D,X(12),X(14),X( 4),X( 9)); - BODY_40_59(45,D,E,T,A,B,C,X(13),X(15),X( 5),X(10)); - BODY_40_59(46,C,D,E,T,A,B,X(14),X( 0),X( 6),X(11)); - BODY_40_59(47,B,C,D,E,T,A,X(15),X( 1),X( 7),X(12)); - BODY_40_59(48,A,B,C,D,E,T,X( 0),X( 2),X( 8),X(13)); - BODY_40_59(49,T,A,B,C,D,E,X( 1),X( 3),X( 9),X(14)); - BODY_40_59(50,E,T,A,B,C,D,X( 2),X( 4),X(10),X(15)); - BODY_40_59(51,D,E,T,A,B,C,X( 3),X( 5),X(11),X( 0)); - BODY_40_59(52,C,D,E,T,A,B,X( 4),X( 6),X(12),X( 1)); - BODY_40_59(53,B,C,D,E,T,A,X( 5),X( 7),X(13),X( 2)); - BODY_40_59(54,A,B,C,D,E,T,X( 6),X( 8),X(14),X( 3)); - BODY_40_59(55,T,A,B,C,D,E,X( 7),X( 9),X(15),X( 4)); - BODY_40_59(56,E,T,A,B,C,D,X( 8),X(10),X( 0),X( 5)); - BODY_40_59(57,D,E,T,A,B,C,X( 9),X(11),X( 1),X( 6)); - BODY_40_59(58,C,D,E,T,A,B,X(10),X(12),X( 2),X( 7)); - BODY_40_59(59,B,C,D,E,T,A,X(11),X(13),X( 3),X( 8)); - - BODY_60_79(60,A,B,C,D,E,T,X(12),X(14),X( 4),X( 9)); - BODY_60_79(61,T,A,B,C,D,E,X(13),X(15),X( 5),X(10)); - BODY_60_79(62,E,T,A,B,C,D,X(14),X( 0),X( 6),X(11)); - BODY_60_79(63,D,E,T,A,B,C,X(15),X( 1),X( 7),X(12)); - BODY_60_79(64,C,D,E,T,A,B,X( 0),X( 2),X( 8),X(13)); - BODY_60_79(65,B,C,D,E,T,A,X( 1),X( 3),X( 9),X(14)); - BODY_60_79(66,A,B,C,D,E,T,X( 2),X( 4),X(10),X(15)); - BODY_60_79(67,T,A,B,C,D,E,X( 3),X( 5),X(11),X( 0)); - BODY_60_79(68,E,T,A,B,C,D,X( 4),X( 6),X(12),X( 1)); - BODY_60_79(69,D,E,T,A,B,C,X( 5),X( 7),X(13),X( 2)); - BODY_60_79(70,C,D,E,T,A,B,X( 6),X( 8),X(14),X( 3)); - BODY_60_79(71,B,C,D,E,T,A,X( 7),X( 9),X(15),X( 4)); - BODY_60_79(72,A,B,C,D,E,T,X( 8),X(10),X( 0),X( 5)); - BODY_60_79(73,T,A,B,C,D,E,X( 9),X(11),X( 1),X( 6)); - BODY_60_79(74,E,T,A,B,C,D,X(10),X(12),X( 2),X( 7)); - BODY_60_79(75,D,E,T,A,B,C,X(11),X(13),X( 3),X( 8)); - BODY_60_79(76,C,D,E,T,A,B,X(12),X(14),X( 4),X( 9)); - BODY_60_79(77,B,C,D,E,T,A,X(13),X(15),X( 5),X(10)); - BODY_60_79(78,A,B,C,D,E,T,X(14),X( 0),X( 6),X(11)); - BODY_60_79(79,T,A,B,C,D,E,X(15),X( 1),X( 7),X(12)); - - c->h0=(c->h0+E)&0xffffffffL; - c->h1=(c->h1+T)&0xffffffffL; - c->h2=(c->h2+A)&0xffffffffL; - c->h3=(c->h3+B)&0xffffffffL; - c->h4=(c->h4+C)&0xffffffffL; - - if (--num <= 0) break; + for (i=0;i<16;i++) + { HOST_c2l(data,l); X[i]=l; BODY_00_15(X[i]); } + for (i=0;i<4;i++) + { BODY_16_19(X[i], X[i+2], X[i+8], X[(i+13)&15]); } + for (;i<24;i++) + { BODY_20_39(X[i&15], X[(i+2)&15], X[(i+8)&15],X[(i+13)&15]); } + for (i=0;i<20;i++) + { BODY_40_59(X[(i+8)&15],X[(i+10)&15],X[i&15], X[(i+5)&15]); } + for (i=4;i<24;i++) + { BODY_60_79(X[(i+8)&15],X[(i+10)&15],X[i&15], X[(i+5)&15]); } + + c->h0=(c->h0+A)&0xffffffffL; + c->h1=(c->h1+B)&0xffffffffL; + c->h2=(c->h2+C)&0xffffffffL; + c->h3=(c->h3+D)&0xffffffffL; + c->h4=(c->h4+E)&0xffffffffL; + + if (--num == 0) break; A=c->h0; B=c->h1; @@ -479,3 +435,5 @@ void HASH_BLOCK_DATA_ORDER (SHA_CTX *c, const void *p, int num) } } #endif + +#endif diff --git a/src/lib/libcrypto/sha/sha_one.c b/src/lib/libcrypto/sha/sha_one.c index d4f4d344df..3bae623ce8 100644 --- a/src/lib/libcrypto/sha/sha_one.c +++ b/src/lib/libcrypto/sha/sha_one.c @@ -62,7 +62,7 @@ #include #ifndef OPENSSL_NO_SHA0 -unsigned char *SHA(const unsigned char *d, unsigned long n, unsigned char *md) +unsigned char *SHA(const unsigned char *d, size_t n, unsigned char *md) { SHA_CTX c; static unsigned char m[SHA_DIGEST_LENGTH]; diff --git a/src/lib/libcrypto/sha/shatest.c b/src/lib/libcrypto/sha/shatest.c index ff702aa53e..ed0fe06a7b 100644 --- a/src/lib/libcrypto/sha/shatest.c +++ b/src/lib/libcrypto/sha/shatest.c @@ -106,7 +106,7 @@ static char *pt(unsigned char *md); int main(int argc, char *argv[]) { int i,err=0; - unsigned char **P,**R; + char **P,**R; static unsigned char buf[1000]; char *p,*r; EVP_MD_CTX c; @@ -118,12 +118,12 @@ int main(int argc, char *argv[]) #endif EVP_MD_CTX_init(&c); - P=(unsigned char **)test; - R=(unsigned char **)ret; + P=test; + R=ret; i=1; while (*P != NULL) { - EVP_Digest(*P,(unsigned long)strlen((char *)*P),md,NULL,EVP_sha(), NULL); + EVP_Digest(*P,strlen((char *)*P),md,NULL,EVP_sha(), NULL); p=pt(md); if (strcmp(p,(char *)*R) != 0) { @@ -157,6 +157,10 @@ int main(int argc, char *argv[]) } else printf("test 3 ok\n"); + +#ifdef OPENSSL_SYS_NETWARE + if (err) printf("ERROR: %d\n", err); +#endif EVP_MD_CTX_cleanup(&c); EXIT(err); return(0); -- cgit v1.2.3-55-g6feb