1
0
mirror of https://github.com/trezor/trezor-firmware.git synced 2024-12-05 05:58:23 +00:00
trezor-firmware/crypto/aes/aescrypt.c
2024-09-12 14:44:33 +02:00

308 lines
10 KiB
C

/*
---------------------------------------------------------------------------
Copyright (c) 1998-2013, Brian Gladman, Worcester, UK. All rights reserved.
The redistribution and use of this software (with or without changes)
is allowed without the payment of fees or royalties provided that:
source code distributions include the above copyright notice, this
list of conditions and the following disclaimer;
binary distributions include the above copyright notice, this list
of conditions and the following disclaimer in their documentation.
This software is provided 'as is' with no explicit or implied warranties
in respect of its operation, including, but not limited to, correctness
and fitness for purpose.
---------------------------------------------------------------------------
Issue Date: 20/12/2007
*/
#include "aesopt.h"
#include "aestab.h"
#if defined( USE_INTEL_AES_IF_PRESENT )
# include "aes_ni.h"
#else
/* map names here to provide the external API ('name' -> 'aes_name') */
# define aes_xi(x) aes_ ## x
#endif
#if defined(__cplusplus)
extern "C"
{
#endif
#define si(y,x,k,c) (s(y,c) = word_in(x, c) ^ (k)[c])
#define so(y,x,c) word_out(y, c, s(x,c))
#if defined(ARRAYS)
#define locals(y,x) x[4] = {0}, y[4] = {0}
#else
#define locals(y,x) x##0=0,x##1=0,x##2=0,x##3=0,y##0=0,y##1=0,y##2=0,y##3=0
#endif
#define l_copy(y, x) s(y,0) = s(x,0); s(y,1) = s(x,1); \
s(y,2) = s(x,2); s(y,3) = s(x,3);
#define state_in(y,x,k) si(y,x,k,0); si(y,x,k,1); si(y,x,k,2); si(y,x,k,3)
#define state_out(y,x) so(y,x,0); so(y,x,1); so(y,x,2); so(y,x,3)
#define round(rm,y,x,k) rm(y,x,k,0); rm(y,x,k,1); rm(y,x,k,2); rm(y,x,k,3)
#if ( FUNCS_IN_C & ENCRYPTION_IN_C )
/* Visual C++ .Net v7.1 provides the fastest encryption code when using
Pentium optimiation with small code but this is poor for decryption
so we need to control this with the following VC++ pragmas
*/
#if defined( _MSC_VER ) && !defined( _WIN64 ) && !defined( __clang__ )
#pragma optimize( "s", on )
#endif
/* Given the column (c) of the output state variable, the following
macros give the input state variables which are needed in its
computation for each row (r) of the state. All the alternative
macros give the same end values but expand into different ways
of calculating these values. In particular the complex macro
used for dynamically variable block sizes is designed to expand
to a compile time constant whenever possible but will expand to
conditional clauses on some branches (I am grateful to Frank
Yellin for this construction)
*/
#define fwd_var(x,r,c)\
( r == 0 ? ( c == 0 ? s(x,0) : c == 1 ? s(x,1) : c == 2 ? s(x,2) : s(x,3))\
: r == 1 ? ( c == 0 ? s(x,1) : c == 1 ? s(x,2) : c == 2 ? s(x,3) : s(x,0))\
: r == 2 ? ( c == 0 ? s(x,2) : c == 1 ? s(x,3) : c == 2 ? s(x,0) : s(x,1))\
: ( c == 0 ? s(x,3) : c == 1 ? s(x,0) : c == 2 ? s(x,1) : s(x,2)))
#if defined(FT4_SET)
#undef dec_fmvars
#define fwd_rnd(y,x,k,c) (s(y,c) = (k)[c] ^ four_tables(x,t_use(f,n),fwd_var,rf1,c))
#elif defined(FT1_SET)
#undef dec_fmvars
#define fwd_rnd(y,x,k,c) (s(y,c) = (k)[c] ^ one_table(x,upr,t_use(f,n),fwd_var,rf1,c))
#else
#define fwd_rnd(y,x,k,c) (s(y,c) = (k)[c] ^ fwd_mcol(no_table(x,t_use(s,box),fwd_var,rf1,c)))
#endif
#if defined(FL4_SET)
#define fwd_lrnd(y,x,k,c) (s(y,c) = (k)[c] ^ four_tables(x,t_use(f,l),fwd_var,rf1,c))
#elif defined(FL1_SET)
#define fwd_lrnd(y,x,k,c) (s(y,c) = (k)[c] ^ one_table(x,ups,t_use(f,l),fwd_var,rf1,c))
#else
#define fwd_lrnd(y,x,k,c) (s(y,c) = (k)[c] ^ no_table(x,t_use(s,box),fwd_var,rf1,c))
#endif
AES_RETURN aes_xi(encrypt)(const unsigned char *in, unsigned char *out, const aes_encrypt_ctx cx[1])
{ uint32_t locals(b0, b1);
const uint32_t *kp = NULL;
#if defined( dec_fmvars )
dec_fmvars; /* declare variables for fwd_mcol() if needed */
#endif
if(cx->inf.b[0] != 10 * AES_BLOCK_SIZE && cx->inf.b[0] != 12 * AES_BLOCK_SIZE && cx->inf.b[0] != 14 * AES_BLOCK_SIZE)
return EXIT_FAILURE;
kp = cx->ks;
state_in(b0, in, kp);
#if (ENC_UNROLL == FULL)
switch(cx->inf.b[0])
{
case 14 * AES_BLOCK_SIZE:
round(fwd_rnd, b1, b0, kp + 1 * N_COLS);
round(fwd_rnd, b0, b1, kp + 2 * N_COLS);
kp += 2 * N_COLS;
//-fallthrough
case 12 * AES_BLOCK_SIZE:
round(fwd_rnd, b1, b0, kp + 1 * N_COLS);
round(fwd_rnd, b0, b1, kp + 2 * N_COLS);
kp += 2 * N_COLS;
//-fallthrough
case 10 * AES_BLOCK_SIZE:
round(fwd_rnd, b1, b0, kp + 1 * N_COLS);
round(fwd_rnd, b0, b1, kp + 2 * N_COLS);
round(fwd_rnd, b1, b0, kp + 3 * N_COLS);
round(fwd_rnd, b0, b1, kp + 4 * N_COLS);
round(fwd_rnd, b1, b0, kp + 5 * N_COLS);
round(fwd_rnd, b0, b1, kp + 6 * N_COLS);
round(fwd_rnd, b1, b0, kp + 7 * N_COLS);
round(fwd_rnd, b0, b1, kp + 8 * N_COLS);
round(fwd_rnd, b1, b0, kp + 9 * N_COLS);
round(fwd_lrnd, b0, b1, kp +10 * N_COLS);
//-fallthrough
}
#else
#if (ENC_UNROLL == PARTIAL)
{ uint32_t rnd = 0;
for(rnd = 0; rnd < (cx->inf.b[0] >> 5) - 1; ++rnd)
{
kp += N_COLS;
round(fwd_rnd, b1, b0, kp);
kp += N_COLS;
round(fwd_rnd, b0, b1, kp);
}
kp += N_COLS;
round(fwd_rnd, b1, b0, kp);
#else
{ int rnd = 0;
for(rnd = 0; rnd < (cx->inf.b[0] >> 4) - 1; ++rnd)
{
kp += N_COLS;
round(fwd_rnd, b1, b0, kp);
l_copy(b0, b1);
}
#endif
kp += N_COLS;
round(fwd_lrnd, b0, b1, kp);
}
#endif
state_out(out, b0);
return EXIT_SUCCESS;
}
#endif
#if ( FUNCS_IN_C & DECRYPTION_IN_C)
/* Visual C++ .Net v7.1 provides the fastest encryption code when using
Pentium optimiation with small code but this is poor for decryption
so we need to control this with the following VC++ pragmas
*/
#if defined( _MSC_VER ) && !defined( _WIN64 ) && !defined( __clang__ )
#pragma optimize( "t", on )
#endif
/* Given the column (c) of the output state variable, the following
macros give the input state variables which are needed in its
computation for each row (r) of the state. All the alternative
macros give the same end values but expand into different ways
of calculating these values. In particular the complex macro
used for dynamically variable block sizes is designed to expand
to a compile time constant whenever possible but will expand to
conditional clauses on some branches (I am grateful to Frank
Yellin for this construction)
*/
#define inv_var(x,r,c)\
( r == 0 ? ( c == 0 ? s(x,0) : c == 1 ? s(x,1) : c == 2 ? s(x,2) : s(x,3))\
: r == 1 ? ( c == 0 ? s(x,3) : c == 1 ? s(x,0) : c == 2 ? s(x,1) : s(x,2))\
: r == 2 ? ( c == 0 ? s(x,2) : c == 1 ? s(x,3) : c == 2 ? s(x,0) : s(x,1))\
: ( c == 0 ? s(x,1) : c == 1 ? s(x,2) : c == 2 ? s(x,3) : s(x,0)))
#if defined(IT4_SET)
#undef dec_imvars
#define inv_rnd(y,x,k,c) (s(y,c) = (k)[c] ^ four_tables(x,t_use(i,n),inv_var,rf1,c))
#elif defined(IT1_SET)
#undef dec_imvars
#define inv_rnd(y,x,k,c) (s(y,c) = (k)[c] ^ one_table(x,upr,t_use(i,n),inv_var,rf1,c))
#else
#define inv_rnd(y,x,k,c) (s(y,c) = inv_mcol((k)[c] ^ no_table(x,t_use(i,box),inv_var,rf1,c)))
#endif
#if defined(IL4_SET)
#define inv_lrnd(y,x,k,c) (s(y,c) = (k)[c] ^ four_tables(x,t_use(i,l),inv_var,rf1,c))
#elif defined(IL1_SET)
#define inv_lrnd(y,x,k,c) (s(y,c) = (k)[c] ^ one_table(x,ups,t_use(i,l),inv_var,rf1,c))
#else
#define inv_lrnd(y,x,k,c) (s(y,c) = (k)[c] ^ no_table(x,t_use(i,box),inv_var,rf1,c))
#endif
/* This code can work with the decryption key schedule in the */
/* order that is used for encrytpion (where the 1st decryption */
/* round key is at the high end ot the schedule) or with a key */
/* schedule that has been reversed to put the 1st decryption */
/* round key at the low end of the schedule in memory (when */
/* AES_REV_DKS is defined) */
#ifdef AES_REV_DKS
#define key_ofs 0
#define rnd_key(n) (kp + n * N_COLS)
#else
#define key_ofs 1
#define rnd_key(n) (kp - n * N_COLS)
#endif
AES_RETURN aes_xi(decrypt)(const unsigned char *in, unsigned char *out, const aes_decrypt_ctx cx[1])
{ uint32_t locals(b0, b1);
#if defined( dec_imvars )
dec_imvars; /* declare variables for inv_mcol() if needed */
#endif
const uint32_t *kp = NULL;
if(cx->inf.b[0] != 10 * AES_BLOCK_SIZE && cx->inf.b[0] != 12 * AES_BLOCK_SIZE && cx->inf.b[0] != 14 * AES_BLOCK_SIZE)
return EXIT_FAILURE;
kp = cx->ks + (key_ofs ? (cx->inf.b[0] >> 2) : 0);
state_in(b0, in, kp);
#if (DEC_UNROLL == FULL)
kp = cx->ks + (key_ofs ? 0 : (cx->inf.b[0] >> 2));
switch(cx->inf.b[0])
{
case 14 * AES_BLOCK_SIZE:
round(inv_rnd, b1, b0, rnd_key(-13));
round(inv_rnd, b0, b1, rnd_key(-12));
//-fallthrough
case 12 * AES_BLOCK_SIZE:
round(inv_rnd, b1, b0, rnd_key(-11));
round(inv_rnd, b0, b1, rnd_key(-10));
//-fallthrough
case 10 * AES_BLOCK_SIZE:
round(inv_rnd, b1, b0, rnd_key(-9));
round(inv_rnd, b0, b1, rnd_key(-8));
round(inv_rnd, b1, b0, rnd_key(-7));
round(inv_rnd, b0, b1, rnd_key(-6));
round(inv_rnd, b1, b0, rnd_key(-5));
round(inv_rnd, b0, b1, rnd_key(-4));
round(inv_rnd, b1, b0, rnd_key(-3));
round(inv_rnd, b0, b1, rnd_key(-2));
round(inv_rnd, b1, b0, rnd_key(-1));
round(inv_lrnd, b0, b1, rnd_key( 0));
//-fallthrough
}
#else
#if (DEC_UNROLL == PARTIAL)
{ uint32_t rnd = 0;
for(rnd = 0; rnd < (cx->inf.b[0] >> 5) - 1; ++rnd)
{
kp = rnd_key(1);
round(inv_rnd, b1, b0, kp);
kp = rnd_key(1);
round(inv_rnd, b0, b1, kp);
}
kp = rnd_key(1);
round(inv_rnd, b1, b0, kp);
#else
{ int rnd = 0;
for(rnd = 0; rnd < (cx->inf.b[0] >> 4) - 1; ++rnd)
{
kp = rnd_key(1);
round(inv_rnd, b1, b0, kp);
l_copy(b0, b1);
}
#endif
kp = rnd_key(1);
round(inv_lrnd, b0, b1, kp);
}
#endif
state_out(out, b0);
return EXIT_SUCCESS;
}
#endif
#if defined(__cplusplus)
}
#endif