modes/ocb128.c: Reset nonce-dependent variables on setiv

[openssl.git] / crypto / aes / asm / aes-sparcv9.pl

diff --git a/crypto/aes/asm/aes-sparcv9.pl b/crypto/aes/asm/aes-sparcv9.pl
index 30f38d7a5a2ec060b97513cf16939c3ca146434b..40d1f94ccd7213f150a43136ef3355ddbb4d5b6f 100755 (executable)
--- a/crypto/aes/asm/aes-sparcv9.pl
+++ b/crypto/aes/asm/aes-sparcv9.pl
@@ -1,12 +1,19 @@
-#!/usr/bin/env perl
+#! /usr/bin/env perl
+# Copyright 2005-2016 The OpenSSL Project Authors. All Rights Reserved.
+#
+# Licensed under the OpenSSL license (the "License").  You may not use
+# this file except in compliance with the License.  You can obtain a copy
+# in the file LICENSE in the source distribution or at
+# https://www.openssl.org/source/license.html
+
 #
 # ====================================================================
-# Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
+# Written by Andy Polyakov <appro@openssl.org> for the OpenSSL
 # project. Rights for redistribution and usage in source and binary
 # forms are granted according to the OpenSSL license.
 # ====================================================================
 #
-# Version 1.0
+# Version 1.1
 #
 # The major reason for undertaken effort was to mitigate the hazard of
 # cache-timing attack. This is [currently and initially!] addressed in
@@ -16,6 +23,9 @@
 # is an initial draft and one should expect more countermeasures to
 # be implemented...
 #
+# Version 1.1 prefetches T[ed]4 in order to mitigate attack on last
+# round.
+#
 # Even though performance was not the primary goal [on the contrary,
 # extra shifts "induced" by compressed S-box and longer loop epilogue
 # "induced" by scheduling for L2 have negative effect on performance],
@@ -27,10 +37,11 @@
 # optimal decrypt procedure]. Compared to GNU C generated code both
 # procedures are more than 60% faster:-)
 
-$bits=32;
-for (@ARGV)    { $bits=64 if (/\-m64/ || /\-xarch\=v9/); }
-if ($bits==64) { $bias=2047; $frame=192; }
-else           { $bias=0;    $frame=112; }
+$output = pop;
+open STDOUT,">$output";
+
+$frame="STACK_FRAME";
+$bias="STACK_BIAS";
 $locals=16;
 
 $acc0="%l0";
@@ -71,14 +82,16 @@ sub _data_word()
     while(defined($i=shift)) { $code.=sprintf"\t.long\t0x%08x,0x%08x\n",$i,$i; }
 }
 
-$code.=<<___ if ($bits==64);
+$code.=<<___;
+#include "sparc_arch.h"
+
+#ifdef  __arch64__
 .register      %g2,#scratch
 .register      %g3,#scratch
-___
-$code.=<<___;
+#endif
 .section       ".text",#alloc,#execinstr
 
-.align 64
+.align 256
 AES_Te:
 ___
 &_data_word(
@@ -364,20 +377,28 @@ _sparcv9_AES_encrypt:
        ld      [$key+28],$t3                   !
                srlx    $acc9,8,$acc9
                xor     $acc5,$s1,$s1
+       ldx     [$tbl+2048+0],%g0               ! prefetch te4
                srlx    $acc10,16,$acc10
                xor     $acc6,$s1,$s1
+       ldx     [$tbl+2048+32],%g0              ! prefetch te4
                srlx    $acc11,24,$acc11
                xor     $acc7,$s1,$s1
+       ldx     [$tbl+2048+64],%g0              ! prefetch te4
                srlx    $acc13,8,$acc13
                xor     $acc8,$s2,$s2
+       ldx     [$tbl+2048+96],%g0              ! prefetch te4
                srlx    $acc14,16,$acc14        !
                xor     $acc9,$s2,$s2
+       ldx     [$tbl+2048+128],%g0             ! prefetch te4
                srlx    $acc15,24,$acc15
                xor     $acc10,$s2,$s2
+       ldx     [$tbl+2048+160],%g0             ! prefetch te4
        srl     $s0,21,$acc0
                xor     $acc11,$s2,$s2
+       ldx     [$tbl+2048+192],%g0             ! prefetch te4
                xor     $acc12,$acc14,$acc14
                xor     $acc13,$s3,$s3
+       ldx     [$tbl+2048+224],%g0             ! prefetch te4
        srl     $s1,13,$acc1                    !
                xor     $acc14,$s3,$s3
                xor     $acc15,$s3,$s3
@@ -512,10 +533,9 @@ AES_encrypt:
        ld      [%i0+12],%o3
 
 1:     call    .+8
-       mov     %i2,%o5
-       sub     %o7,1b-AES_Te,%o4
+       add     %o7,AES_Te-1b,%o4
        call    _sparcv9_AES_encrypt
-       nop
+       mov     %i2,%o5
 
        st      %o0,[%i1+0]
        st      %o1,[%i1+4]
@@ -573,10 +593,9 @@ AES_encrypt:
        or      %l4,%l6,%o3
 
 1:     call    .+8
-       mov     %i2,%o5
-       sub     %o7,1b-AES_Te,%o4
+       add     %o7,AES_Te-1b,%o4
        call    _sparcv9_AES_encrypt
-       nop
+       mov     %i2,%o5
 
        srl     %o0,24,%l0
        srl     %o0,16,%l1
@@ -618,7 +637,7 @@ AES_encrypt:
 ___
 
 $code.=<<___;
-.align 64
+.align 256
 AES_Td:
 ___
 &_data_word(
@@ -904,20 +923,28 @@ _sparcv9_AES_decrypt:
        ld      [$key+28],$t3                   !
                srlx    $acc9,8,$acc9
                xor     $acc5,$s1,$s1
+       ldx     [$tbl+2048+0],%g0               ! prefetch td4
                srlx    $acc10,16,$acc10
                xor     $acc6,$s1,$s1
+       ldx     [$tbl+2048+32],%g0              ! prefetch td4
                srlx    $acc11,24,$acc11
                xor     $acc7,$s1,$s1
+       ldx     [$tbl+2048+64],%g0              ! prefetch td4
                srlx    $acc13,8,$acc13
                xor     $acc8,$s2,$s2
+       ldx     [$tbl+2048+96],%g0              ! prefetch td4
                srlx    $acc14,16,$acc14        !
                xor     $acc9,$s2,$s2
+       ldx     [$tbl+2048+128],%g0             ! prefetch td4
                srlx    $acc15,24,$acc15
                xor     $acc10,$s2,$s2
+       ldx     [$tbl+2048+160],%g0             ! prefetch td4
        srl     $s0,21,$acc0
                xor     $acc11,$s2,$s2
+       ldx     [$tbl+2048+192],%g0             ! prefetch td4
                xor     $acc12,$acc14,$acc14
                xor     $acc13,$s3,$s3
+       ldx     [$tbl+2048+224],%g0             ! prefetch td4
        and     $acc0,2040,$acc0                !
                xor     $acc14,$s3,$s3
                xor     $acc15,$s3,$s3
@@ -1052,10 +1079,9 @@ AES_decrypt:
        ld      [%i0+12],%o3
 
 1:     call    .+8
-       mov     %i2,%o5
-       sub     %o7,1b-AES_Td,%o4
+       add     %o7,AES_Td-1b,%o4
        call    _sparcv9_AES_decrypt
-       nop
+       mov     %i2,%o5
 
        st      %o0,[%i1+0]
        st      %o1,[%i1+4]
@@ -1113,10 +1139,9 @@ AES_decrypt:
        or      %l4,%l6,%o3
 
 1:     call    .+8
-       mov     %i2,%o5
-       sub     %o7,1b-AES_Td,%o4
+       add     %o7,AES_Td-1b,%o4
        call    _sparcv9_AES_decrypt
-       nop
+       mov     %i2,%o5
 
        srl     %o0,24,%l0
        srl     %o0,16,%l1
@@ -1161,6 +1186,7 @@ ___
 # As UltraSPARC T1, a.k.a. Niagara, has shared FPU, FP nops can have
 # undesired effect, so just omit them and sacrifice some portion of
 # percent in performance...
-$code =~ s/fmovs.*$//gem;
+$code =~ s/fmovs.*$//gm;
 
 print $code;
+close STDOUT;  # ensure flush