diff options
Diffstat (limited to 'scripts/conf-w32brg/cipher')
| -rw-r--r-- | scripts/conf-w32brg/cipher/aes.h | 158 | ||||
| -rw-r--r-- | scripts/conf-w32brg/cipher/aescrypt.asm | 404 | ||||
| -rw-r--r-- | scripts/conf-w32brg/cipher/aescrypt.c | 311 | ||||
| -rw-r--r-- | scripts/conf-w32brg/cipher/aeskey.c | 463 | ||||
| -rw-r--r-- | scripts/conf-w32brg/cipher/aesopt.h | 1042 | ||||
| -rw-r--r-- | scripts/conf-w32brg/cipher/aestab.c | 232 | ||||
| -rw-r--r-- | scripts/conf-w32brg/cipher/rijndael2.c | 279 |
7 files changed, 2889 insertions, 0 deletions
diff --git a/scripts/conf-w32brg/cipher/aes.h b/scripts/conf-w32brg/cipher/aes.h new file mode 100644 index 000000000..1bec72059 --- /dev/null +++ b/scripts/conf-w32brg/cipher/aes.h @@ -0,0 +1,158 @@ +/* + --------------------------------------------------------------------------- + Copyright (c) 2003, Dr Brian Gladman <[email protected]>, Worcester, UK. + All rights reserved. + + LICENSE TERMS + + The free distribution and use of this software in both source and binary + form is allowed (with or without changes) provided that: + + 1. distributions of this source code include the above copyright + notice, this list of conditions and the following disclaimer; + + 2. distributions in binary form include the above copyright + notice, this list of conditions and the following disclaimer + in the documentation and/or other associated materials; + + 3. the copyright holder's name is not used to endorse products + built using this software without specific written permission. + + ALTERNATIVELY, provided that this notice is retained in full, this product + may be distributed under the terms of the GNU General Public License (GPL), + in which case the provisions of the GPL apply INSTEAD OF those given above. + + DISCLAIMER + + This software is provided 'as is' with no explicit or implied warranties + in respect of its properties, including, but not limited to, correctness + and/or fitness for purpose. + --------------------------------------------------------------------------- + Issue Date: 1/06/2003 + + This file contains the definitions required to use AES in C. See aesopt.h + for optimisation details. +*/ + +#ifndef _AES_H +#define _AES_H + +#if defined(__cplusplus) +extern "C" +{ +#endif + +#define AES_128 /* define if AES with 128 bit keys is needed */ +#define AES_192 /* define if AES with 192 bit keys is needed */ +#define AES_256 /* define if AES with 256 bit keys is needed */ +#define AES_VAR /* define if a variable key size is needed */ + +/* The following must also be set in assembler files if being used */ + +#define AES_ENCRYPT /* if support for encryption is needed */ +#define AES_DECRYPT /* if support for decryption is needed */ +#define AES_ERR_CHK /* for parameter checks & error return codes */ + +/* This include is used to find 8 & 32 bit unsigned integer types */ +#include "limits.h" + +#if UCHAR_MAX == 0xff /* an unsigned 8 bit type */ + typedef unsigned char aes_08t; +#else +#error Please define aes_08t as an 8-bit unsigned integer type in aes.h +#endif + +#if UINT_MAX == 0xffffffff /* an unsigned 32 bit type */ + typedef unsigned int aes_32t; +#elif ULONG_MAX == 0xffffffff + typedef unsigned long aes_32t; +#else +#error Please define aes_32t as a 32-bit unsigned integer type in aes.h +#endif + +#define AES_BLOCK_SIZE 16 /* the AES block size in bytes */ +#define N_COLS 4 /* the number of columns in the state */ + +/* a maximum of 60 32-bit words are needed for the key schedule but */ +/* 64 are claimed to allow space at the top for a CBC xor buffer. */ +/* If this is not needed, this value can be reduced to 60. A value */ +/* of 64 may also help in maintaining alignment in some situations */ +#define KS_LENGTH 64 + +#ifdef AES_ERR_CHK +#define aes_ret int +#define aes_good 0 +#define aes_error -1 +#else +#define aes_ret void +#endif + +#ifndef AES_DLL /* implement normal/DLL functions */ +#define aes_rval aes_ret +#else +#define aes_rval aes_ret __declspec(dllexport) _stdcall +#endif + +/* This routine must be called before first use if non-static */ +/* tables are being used */ + +void gen_tabs(void); + +/* The key length (klen) is input in bytes when it is in the range */ +/* 16 <= klen <= 32 or in bits when in the range 128 <= klen <= 256 */ + +#ifdef AES_ENCRYPT + +typedef struct +{ aes_32t ks[KS_LENGTH]; +} aes_encrypt_ctx; + +#if defined(AES_128) || defined(AES_VAR) +aes_rval aes_encrypt_key128(const void *in_key, aes_encrypt_ctx cx[1]); +#endif + +#if defined(AES_192) || defined(AES_VAR) +aes_rval aes_encrypt_key192(const void *in_key, aes_encrypt_ctx cx[1]); +#endif + +#if defined(AES_256) || defined(AES_VAR) +aes_rval aes_encrypt_key256(const void *in_key, aes_encrypt_ctx cx[1]); +#endif + +#if defined(AES_VAR) +aes_rval aes_encrypt_key(const void *in_key, int key_len, aes_encrypt_ctx cx[1]); +#endif + +aes_rval aes_encrypt(const void *in_blk, void *out_blk, const aes_encrypt_ctx cx[1]); +#endif + +#ifdef AES_DECRYPT + +typedef struct +{ aes_32t ks[KS_LENGTH]; +} aes_decrypt_ctx; + +#if defined(AES_128) || defined(AES_VAR) +aes_rval aes_decrypt_key128(const void *in_key, aes_decrypt_ctx cx[1]); +#endif + +#if defined(AES_192) || defined(AES_VAR) +aes_rval aes_decrypt_key192(const void *in_key, aes_decrypt_ctx cx[1]); +#endif + +#if defined(AES_256) || defined(AES_VAR) +aes_rval aes_decrypt_key256(const void *in_key, aes_decrypt_ctx cx[1]); +#endif + +#if defined(AES_VAR) +aes_rval aes_decrypt_key(const void *in_key, int key_len, aes_decrypt_ctx cx[1]); +#endif + +aes_rval aes_decrypt(const void *in_blk, void *out_blk, const aes_decrypt_ctx cx[1]); +#endif + +#if defined(__cplusplus) +} +#endif + +#endif diff --git a/scripts/conf-w32brg/cipher/aescrypt.asm b/scripts/conf-w32brg/cipher/aescrypt.asm new file mode 100644 index 000000000..90beaaabf --- /dev/null +++ b/scripts/conf-w32brg/cipher/aescrypt.asm @@ -0,0 +1,404 @@ + +; --------------------------------------------------------------------------- +; Copyright (c) 2002, Dr Brian Gladman <[email protected]>, Worcester, UK. +; All rights reserved. +; +; LICENSE TERMS +; +; The free distribution and use of this software in both source and binary +; form is allowed (with or without changes) provided that: +; +; 1. distributions of this source code include the above copyright +; notice, this list of conditions and the following disclaimer; +; +; 2. distributions in binary form include the above copyright +; notice, this list of conditions and the following disclaimer +; in the documentation and/or other associated materials; +; +; 3. the copyright holder's name is not used to endorse products +; built using this software without specific written permission. +; +; ALTERNATIVELY, provided that this notice is retained in full, this product +; may be distributed under the terms of the GNU General Public License (GPL), +; in which case the provisions of the GPL apply INSTEAD OF those given above. +; +; DISCLAIMER +; +; This software is provided 'as is' with no explicit or implied warranties +; in respect of its properties, including, but not limited to, correctness +; and/or fitness for purpose. +; --------------------------------------------------------------------------- +; Issue Date: 1/06/2003 + +; An AES implementation for Pentium processors using the NASM assembler (see +; <http://sourceforge.net/projects/nasm>).This version provides the standard +; AES block length (128 bits, 16 bytes) with the same interface as that used +; in my C implementation. The eax, ecx and edx registers and the artihmetic +; status flags are not preserved. The ebx, esi, edi, and ebp registers are +; preserved across calls. Only encryption and decryption are provided here, +; here, the key scheduling code being that in aeskey.c compiled with USE_ASM +; defined. This code uses the VC++ register saving conentions; if it is used +; with another compiler, its conventions for using and saving registers will +; need to be checked (and calling conventions). The NASM command line for +; the VC++ custom build step is: +; +; nasm -O2 -f win32 -o "$(TargetDir)\$(InputName).obj" "$(InputPath)" + + section .text ; use32 + +; aes_rval aes_encrypt(const unsigned char in_blk[], +; unsigned char out_blk[], const aes_encrypt_ctx cx[1]); +; aes_rval aes_decrypt(const unsigned char in_blk[], +; unsigned char out_blk[], const aes_decrypt_ctx cx[1]); +; +; comment in/out the following lines to obtain the desired subroutines + +%define ENCRYPTION ; define if encryption is needed +%define DECRYPTION ; define if decryption is needed + +; The DLL interface must use the _stdcall convention in which the number +; of bytes of parameter space is added after an @ to the sutine's name. +; We must also remove our parameters from the stack before return (see +; the do_ret macro). Define AES_DLL for the Dynamic Link Library version. + +;%define AES_DLL + +tlen: equ 1024 ; length of each of 4 'xor' arrays (256 32-bit words) + +; offsets to parameters with one register pushed onto stack + +in_blk: equ 4 ; input byte array address parameter +out_blk:equ 8 ; output byte array address parameter +ctx: equ 12 ; AES context structure +stk_spc:equ 24 ; stack space + +; register mapping for encrypt and decrypt subroutines + +%define r0 eax +%define r1 ebx +%define r2 esi +%define r3 edi +%define r4 ecx +%define r5 edx +%define r6 ebp + +%define eaxl al +%define eaxh ah +%define ebxl bl +%define ebxh bh +%define ecxl cl +%define ecxh ch +%define edxl dl +%define edxh dh + +; These macros take a 32-bit word representing a column and use each +; of its 4 bytes to index a table of 256 32-bit words which are xored +; into each of the four output columns. The output values are in the +; registers %1, %2, %3 and %4 and the column input is in %5 with %6 +; as a scratch register. + +; Parameters: +; %1 out_state[0] +; %2 out_state[1] +; %3 out_state[2] +; %4 out_state[3] +; %5 input register for the round (destroyed) +; %6 scratch register for the round +; %7 key schedule address for round (in form r6 + offset) + +%macro do_fcol 8 ; first column forward round + + movzx %6,%5l + mov %1,[%8] + xor %1,[4*%6+%7] + movzx %6,%5h + shr %5,16 + mov %2,[%8+12] + xor %2,[4*%6+%7+tlen] + movzx %6,%5l + mov %3,[%8+ 8] + xor %3,[4*%6+%7+2*tlen] + movzx %6,%5h + mov %5,%4 ; save an input register value + mov %4,[%8+ 4] + xor %4,[4*%6+%7+3*tlen] + +%endmacro + +%macro do_icol 8 ; first column for inverse round + + movzx %6,%5l + mov %1,[%8] + xor %1,[4*%6+%7] + movzx %6,%5h + shr %5,16 + mov %2,[%8+ 4] + xor %2,[4*%6+%7+tlen] + movzx %6,%5l + mov %3,[%8+ 8] + xor %3,[4*%6+%7+2*tlen] + movzx %6,%5h + mov %5,%4 ; save an input register value + mov %4,[%8+12] + xor %4,[4*%6+%7+3*tlen] + +%endmacro + +%macro do_col 7 ; other columns for forward and inverse rounds + + movzx %6,%5l + xor %1,[4*%6+%7] + movzx %6,%5h + shr %5,16 + xor %2,[4*%6+%7+tlen] + movzx %6,%5l + xor %3,[4*%6+%7+2*tlen] + movzx %6,%5h + xor %4,[4*%6+%7+3*tlen] + +%endmacro + +; These macros implement stack based local variables + +%macro save 2 + mov [esp+4*%1],%2 +%endmacro + +%macro restore 2 + mov %1,[esp+4*%2] +%endmacro + +; This macro performs a forward encryption cycle. It is entered with +; the first previous round column values in r0, r1, r2 and r3 and +; exits with the final values in the same registers. + +%macro fwd_rnd 1-2 _t_fn ; normal forward rounds + + mov r4,r0 + save 0,r2 + save 1,r3 + +; compute new column values + + do_fcol r0,r3,r2,r1, r4,r5, %2, %1 ; r4 = input r0 + do_col r1,r0,r3,r2, r4,r5, %2 ; r4 = input r1 (saved in fcol_f) + restore r4,0 + do_col r2,r1,r0,r3, r4,r5, %2 ; r4 = input r2 + restore r4,1 + do_col r3,r2,r1,r0, r4,r5, %2 ; r4 = input r3 + +%endmacro + +; This macro performs an inverse encryption cycle. It is entered with +; the first previous round column values in r0, r1, r2 and r3 and +; exits with the final values in the same registers. + +%macro inv_rnd 1-2 _t_in ; normal inverse round + + mov r4,r0 + save 0,r1 + save 1,r2 + +; compute new column values + + do_icol r0,r1,r2,r3, r4,r5, %2, %1 ; r4 = r0 + do_col r3,r0,r1,r2, r4,r5, %2 ; r4 = r3 (saved in icol_f) + restore r4,1 + do_col r2,r3,r0,r1, r4,r5, %2 ; r4 = r2 + restore r4,0 + do_col r1,r2,r3,r0, r4,r5, %2 ; r4 = r1 + +%endmacro + +; the DLL has to implement the _stdcall calling interface on return +; In this case we have to take our parameters (3 4-byte pointers) +; off the stack + +%macro do_ret 0 +%ifdef AES_DLL + ret 12 +%else + ret +%endif +%endmacro + +%macro do_name 1 +%ifndef AES_DLL + global %1 +%1: +%else + global %1@12 + export %1@12 +%1@12: +%endif +%endmacro + +; AES Encryption Subroutine + +%ifdef ENCRYPTION + + extern _t_fn + extern _t_fl + + do_name _aes_encrypt + + sub esp,stk_spc + mov [esp+20],ebp + mov [esp+16],ebx + mov [esp+12],esi + mov [esp+ 8],edi + mov r4,[esp+in_blk+stk_spc] ; input pointer + mov r6,[esp+ctx+stk_spc] ; key pointer + +; input four columns and xor in first round key + + mov r0,[r4 ] + mov r1,[r4+ 4] + xor r0,[r6 ] + xor r1,[r6+ 4] + mov r2,[r4+ 8] + mov r3,[r4+12] + xor r2,[r6+ 8] + xor r3,[r6+12] + +; determine the number of rounds + + mov r4,[r6+4*45] + mov r5,[r6+4*52] + xor r4,[r6+4*53] + xor r4,r5 + je .1 + cmp r5,10 + je .3 + cmp r5,12 + je .2 + mov ebp,[esp+20] + mov ebx,[esp+16] + mov esi,[esp+12] + mov edi,[esp+ 8] + lea esp,[esp+stk_spc] + mov eax,-1 + do_ret + +.1: fwd_rnd r6+ 16 ; 14 rounds for 256-bit key + fwd_rnd r6+ 32 + lea r6,[r6+32] +.2: fwd_rnd r6+ 16 ; 12 rounds for 192-bit key + fwd_rnd r6+ 32 + lea r6,[r6+32] +.3: fwd_rnd r6+ 16 ; 10 rounds for 128-bit key + fwd_rnd r6+ 32 + fwd_rnd r6+ 48 + fwd_rnd r6+ 64 + fwd_rnd r6+ 80 + fwd_rnd r6+ 96 + fwd_rnd r6+112 + fwd_rnd r6+128 + fwd_rnd r6+144 + fwd_rnd r6+160, _t_fl ; last round uses a different table + +; move final values to the output array + + mov r6,[esp+out_blk+stk_spc] + mov [r6+12],r3 + mov [r6+8],r2 + mov [r6+4],r1 + mov [r6],r0 + mov ebp,[esp+20] + mov ebx,[esp+16] + mov esi,[esp+12] + mov edi,[esp+ 8] + lea esp,[esp+stk_spc] + xor eax,eax + do_ret + +%endif + +; AES Decryption Subroutine + +%ifdef DECRYPTION + + extern _t_in + extern _t_il + + do_name _aes_decrypt + + sub esp,stk_spc + mov [esp+20],ebp + mov [esp+16],ebx + mov [esp+12],esi + mov [esp+ 8],edi + mov r4,[esp+in_blk+stk_spc] ; input pointer + mov r6,[esp+ctx+stk_spc] ; context pointer + +; input four columns + + mov r0,[r4] + mov r1,[r4+4] + mov r2,[r4+8] + mov r3,[r4+12] + +; determine the number of rounds + + mov r5,[r6+4*52] + mov r4,[r6+4*45] + xor r4,[r6+4*53] + xor r4,r5 + jne .1 + mov r5,14 + +; xor in initial keys + +.1: lea r4,[4*r5] + xor r0,[r6+4*r4 ] + xor r1,[r6+4*r4+ 4] + xor r2,[r6+4*r4+ 8] + xor r3,[r6+4*r4+12] + cmp r5,10 + je .3 + cmp r5,12 + je .2 + cmp r5,14 + jne .4 + + inv_rnd r6+208 ; 14 rounds for 256-bit key + inv_rnd r6+192 +.2: inv_rnd r6+176 ; 12 rounds for 192-bit key + inv_rnd r6+160 +.3: inv_rnd r6+144 ; 10 rounds for 128-bit key + inv_rnd r6+128 + inv_rnd r6+112 + inv_rnd r6+ 96 + inv_rnd r6+ 80 + inv_rnd r6+ 64 + inv_rnd r6+ 48 + inv_rnd r6+ 32 + inv_rnd r6+ 16 + inv_rnd r6, _t_il ; last round uses a different table + +; move final values to the output array. + + mov r6,[esp+out_blk+stk_spc] + mov [r6+12],r3 + mov [r6+8],r2 + mov [r6+4],r1 + mov [r6],r0 + mov ebp,[esp+20] + mov ebx,[esp+16] + mov esi,[esp+12] + mov edi,[esp+ 8] + lea esp,[esp+stk_spc] + xor eax,eax + do_ret + +.4: mov ebp,[esp+20] + mov ebx,[esp+16] + mov esi,[esp+12] + mov edi,[esp+ 8] + lea esp,[esp+stk_spc] + mov eax,-1 + do_ret + +%endif + + end diff --git a/scripts/conf-w32brg/cipher/aescrypt.c b/scripts/conf-w32brg/cipher/aescrypt.c new file mode 100644 index 000000000..944b0890a --- /dev/null +++ b/scripts/conf-w32brg/cipher/aescrypt.c @@ -0,0 +1,311 @@ +/* + --------------------------------------------------------------------------- + Copyright (c) 2003, Dr Brian Gladman <[email protected]>, Worcester, UK. + All rights reserved. + + LICENSE TERMS + + The free distribution and use of this software in both source and binary + form is allowed (with or without changes) provided that: + + 1. distributions of this source code include the above copyright + notice, this list of conditions and the following disclaimer; + + 2. distributions in binary form include the above copyright + notice, this list of conditions and the following disclaimer + in the documentation and/or other associated materials; + + 3. the copyright holder's name is not used to endorse products + built using this software without specific written permission. + + ALTERNATIVELY, provided that this notice is retained in full, this product + may be distributed under the terms of the GNU General Public License (GPL), + in which case the provisions of the GPL apply INSTEAD OF those given above. + + DISCLAIMER + + This software is provided 'as is' with no explicit or implied warranties + in respect of its properties, including, but not limited to, correctness + and/or fitness for purpose. + --------------------------------------------------------------------------- + Issue Date: 1/06/2003 + + This file contains the code for implementing encryption and decryption + for AES (Rijndael) for block and key sizes of 16, 24 and 32 bytes. It + can optionally be replaced by code written in assembler using NASM. For + further details see the file aesopt.h +*/ + +#include "aesopt.h" + +#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],y[4] +#else +#define locals(y,x) x##0,x##1,x##2,x##3,y##0,y##1,y##2,y##3 +#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 defined(ENCRYPTION) && !defined(AES_ASM) + +/* 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) +#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_rval aes_encrypt(const void *in_blk, void *out_blk, const aes_encrypt_ctx cx[1]) +{ aes_32t locals(b0, b1); + const aes_32t *kp = cx->ks; +#ifdef dec_fmvars + dec_fmvars; /* declare variables for fwd_mcol() if needed */ +#endif + + aes_32t nr = (kp[45] ^ kp[52] ^ kp[53] ? kp[52] : 14); + +#ifdef AES_ERR_CHK + if( (nr != 10 || !(kp[0] | kp[3] | kp[4])) + && (nr != 12 || !(kp[0] | kp[5] | kp[6])) + && (nr != 14 || !(kp[0] | kp[7] | kp[8])) ) + return aes_error; +#endif + + state_in(b0, in_blk, kp); + +#if (ENC_UNROLL == FULL) + + switch(nr) + { + case 14: + round(fwd_rnd, b1, b0, kp + 1 * N_COLS); + round(fwd_rnd, b0, b1, kp + 2 * N_COLS); + kp += 2 * N_COLS; + case 12: + round(fwd_rnd, b1, b0, kp + 1 * N_COLS); + round(fwd_rnd, b0, b1, kp + 2 * N_COLS); + kp += 2 * N_COLS; + case 10: + 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); + } + +#else + +#if (ENC_UNROLL == PARTIAL) + { aes_32t rnd; + for(rnd = 0; rnd < (nr >> 1) - 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 + { aes_32t rnd; + for(rnd = 0; rnd < nr - 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_blk, b0); +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(DECRYPTION) && !defined(AES_ASM) + +/* 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) +#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 + +aes_rval aes_decrypt(const void *in_blk, void *out_blk, const aes_decrypt_ctx cx[1]) +{ aes_32t locals(b0, b1); +#ifdef dec_imvars + dec_imvars; /* declare variables for inv_mcol() if needed */ +#endif + + aes_32t nr = (cx->ks[45] ^ cx->ks[52] ^ cx->ks[53] ? cx->ks[52] : 14); + const aes_32t *kp = cx->ks + nr * N_COLS; + +#ifdef AES_ERR_CHK + if( (nr != 10 || !(cx->ks[0] | cx->ks[3] | cx->ks[4])) + && (nr != 12 || !(cx->ks[0] | cx->ks[5] | cx->ks[6])) + && (nr != 14 || !(cx->ks[0] | cx->ks[7] | cx->ks[8])) ) + return aes_error; +#endif + + state_in(b0, in_blk, kp); + +#if (DEC_UNROLL == FULL) + + switch(nr) + { + case 14: + round(inv_rnd, b1, b0, kp - 1 * N_COLS); + round(inv_rnd, b0, b1, kp - 2 * N_COLS); + kp -= 2 * N_COLS; + case 12: + round(inv_rnd, b1, b0, kp - 1 * N_COLS); + round(inv_rnd, b0, b1, kp - 2 * N_COLS); + kp -= 2 * N_COLS; + case 10: + round(inv_rnd, b1, b0, kp - 1 * N_COLS); + round(inv_rnd, b0, b1, kp - 2 * N_COLS); + round(inv_rnd, b1, b0, kp - 3 * N_COLS); + round(inv_rnd, b0, b1, kp - 4 * N_COLS); + round(inv_rnd, b1, b0, kp - 5 * N_COLS); + round(inv_rnd, b0, b1, kp - 6 * N_COLS); + round(inv_rnd, b1, b0, kp - 7 * N_COLS); + round(inv_rnd, b0, b1, kp - 8 * N_COLS); + round(inv_rnd, b1, b0, kp - 9 * N_COLS); + round(inv_lrnd, b0, b1, kp - 10 * N_COLS); + } + +#else + +#if (DEC_UNROLL == PARTIAL) + { aes_32t rnd; + for(rnd = 0; rnd < (nr >> 1) - 1; ++rnd) + { + kp -= N_COLS; + round(inv_rnd, b1, b0, kp); + kp -= N_COLS; + round(inv_rnd, b0, b1, kp); + } + kp -= N_COLS; + round(inv_rnd, b1, b0, kp); +#else + { aes_32t rnd; + for(rnd = 0; rnd < nr - 1; ++rnd) + { + kp -= N_COLS; + round(inv_rnd, b1, b0, kp); + l_copy(b0, b1); + } +#endif + kp -= N_COLS; + round(inv_lrnd, b0, b1, kp); + } +#endif + + state_out(out_blk, b0); +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(__cplusplus) +} +#endif diff --git a/scripts/conf-w32brg/cipher/aeskey.c b/scripts/conf-w32brg/cipher/aeskey.c new file mode 100644 index 000000000..acc1b8899 --- /dev/null +++ b/scripts/conf-w32brg/cipher/aeskey.c @@ -0,0 +1,463 @@ +/* + --------------------------------------------------------------------------- + Copyright (c) 2003, Dr Brian Gladman <[email protected]>, Worcester, UK. + All rights reserved. + + LICENSE TERMS + + The free distribution and use of this software in both source and binary + form is allowed (with or without changes) provided that: + + 1. distributions of this source code include the above copyright + notice, this list of conditions and the following disclaimer; + + 2. distributions in binary form include the above copyright + notice, this list of conditions and the following disclaimer + in the documentation and/or other associated materials; + + 3. the copyright holder's name is not used to endorse products + built using this software without specific written permission. + + ALTERNATIVELY, provided that this notice is retained in full, this product + may be distributed under the terms of the GNU General Public License (GPL), + in which case the provisions of the GPL apply INSTEAD OF those given above. + + DISCLAIMER + + This software is provided 'as is' with no explicit or implied warranties + in respect of its properties, including, but not limited to, correctness + and/or fitness for purpose. + --------------------------------------------------------------------------- + Issue Date: 1/06/2003 + + This file contains the code for implementing the key schedule for AES + (Rijndael) for block and key sizes of 16, 24, and 32 bytes. See aesopt.h + for further details including optimisation. +*/ + +#include "aesopt.h" + +#if defined(__cplusplus) +extern "C" +{ +#endif + +/* Initialise the key schedule from the user supplied key. The key + length can be specified in bytes, with legal values of 16, 24 + and 32, or in bits, with legal values of 128, 192 and 256. These + values correspond with Nk values of 4, 6 and 8 respectively. + + The following macros implement a single cycle in the key + schedule generation process. The number of cycles needed + for each cx->n_col and nk value is: + + nk = 4 5 6 7 8 + ------------------------------ + cx->n_col = 4 10 9 8 7 7 + cx->n_col = 5 14 11 10 9 9 + cx->n_col = 6 19 15 12 11 11 + cx->n_col = 7 21 19 16 13 14 + cx->n_col = 8 29 23 19 17 14 +*/ + +#define ke4(k,i) \ +{ k[4*(i)+4] = ss[0] ^= ls_box(ss[3],3) ^ t_use(r,c)[i]; k[4*(i)+5] = ss[1] ^= ss[0]; \ + k[4*(i)+6] = ss[2] ^= ss[1]; k[4*(i)+7] = ss[3] ^= ss[2]; \ +} +#define kel4(k,i) \ +{ k[4*(i)+4] = ss[0] ^= ls_box(ss[3],3) ^ t_use(r,c)[i]; k[4*(i)+5] = ss[1] ^= ss[0]; \ + k[4*(i)+6] = ss[2] ^= ss[1]; k[4*(i)+7] = ss[3] ^= ss[2]; \ +} + +#define ke6(k,i) \ +{ k[6*(i)+ 6] = ss[0] ^= ls_box(ss[5],3) ^ t_use(r,c)[i]; k[6*(i)+ 7] = ss[1] ^= ss[0]; \ + k[6*(i)+ 8] = ss[2] ^= ss[1]; k[6*(i)+ 9] = ss[3] ^= ss[2]; \ + k[6*(i)+10] = ss[4] ^= ss[3]; k[6*(i)+11] = ss[5] ^= ss[4]; \ +} +#define kel6(k,i) \ +{ k[6*(i)+ 6] = ss[0] ^= ls_box(ss[5],3) ^ t_use(r,c)[i]; k[6*(i)+ 7] = ss[1] ^= ss[0]; \ + k[6*(i)+ 8] = ss[2] ^= ss[1]; k[6*(i)+ 9] = ss[3] ^= ss[2]; \ +} + +#define ke8(k,i) \ +{ k[8*(i)+ 8] = ss[0] ^= ls_box(ss[7],3) ^ t_use(r,c)[i]; k[8*(i)+ 9] = ss[1] ^= ss[0]; \ + k[8*(i)+10] = ss[2] ^= ss[1]; k[8*(i)+11] = ss[3] ^= ss[2]; \ + k[8*(i)+12] = ss[4] ^= ls_box(ss[3],0); k[8*(i)+13] = ss[5] ^= ss[4]; \ + k[8*(i)+14] = ss[6] ^= ss[5]; k[8*(i)+15] = ss[7] ^= ss[6]; \ +} +#define kel8(k,i) \ +{ k[8*(i)+ 8] = ss[0] ^= ls_box(ss[7],3) ^ t_use(r,c)[i]; k[8*(i)+ 9] = ss[1] ^= ss[0]; \ + k[8*(i)+10] = ss[2] ^= ss[1]; k[8*(i)+11] = ss[3] ^= ss[2]; \ +} + +#if defined(ENCRYPTION_KEY_SCHEDULE) + +#if defined(AES_128) || defined(AES_VAR) + +aes_rval aes_encrypt_key128(const void *in_key, aes_encrypt_ctx cx[1]) +{ aes_32t ss[4]; + + cx->ks[0] = ss[0] = word_in(in_key, 0); + cx->ks[1] = ss[1] = word_in(in_key, 1); + cx->ks[2] = ss[2] = word_in(in_key, 2); + cx->ks[3] = ss[3] = word_in(in_key, 3); + +#if ENC_UNROLL == NONE + { aes_32t i; + + for(i = 0; i < ((11 * N_COLS - 1) / 4); ++i) + ke4(cx->ks, i); + } +#else + ke4(cx->ks, 0); ke4(cx->ks, 1); + ke4(cx->ks, 2); ke4(cx->ks, 3); + ke4(cx->ks, 4); ke4(cx->ks, 5); + ke4(cx->ks, 6); ke4(cx->ks, 7); + ke4(cx->ks, 8); kel4(cx->ks, 9); +#endif + + /* cx->ks[45] ^ cx->ks[52] ^ cx->ks[53] is zero for a 256 bit */ + /* key and must be non-zero for 128 and 192 bits keys */ + cx->ks[53] = cx->ks[45] = 0; + cx->ks[52] = 10; +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(AES_192) || defined(AES_VAR) + +aes_rval aes_encrypt_key192(const void *in_key, aes_encrypt_ctx cx[1]) +{ aes_32t ss[6]; + + cx->ks[0] = ss[0] = word_in(in_key, 0); + cx->ks[1] = ss[1] = word_in(in_key, 1); + cx->ks[2] = ss[2] = word_in(in_key, 2); + cx->ks[3] = ss[3] = word_in(in_key, 3); + cx->ks[4] = ss[4] = word_in(in_key, 4); + cx->ks[5] = ss[5] = word_in(in_key, 5); + +#if ENC_UNROLL == NONE + { aes_32t i; + + for(i = 0; i < (13 * N_COLS - 1) / 6; ++i) + ke6(cx->ks, i); + } +#else + ke6(cx->ks, 0); ke6(cx->ks, 1); + ke6(cx->ks, 2); ke6(cx->ks, 3); + ke6(cx->ks, 4); ke6(cx->ks, 5); + ke6(cx->ks, 6); kel6(cx->ks, 7); +#endif + + /* cx->ks[45] ^ cx->ks[52] ^ cx->ks[53] is zero for a 256 bit */ + /* key and must be non-zero for 128 and 192 bits keys */ + cx->ks[53] = cx->ks[45]; + cx->ks[52] = 12; +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(AES_256) || defined(AES_VAR) + +aes_rval aes_encrypt_key256(const void *in_key, aes_encrypt_ctx cx[1]) +{ aes_32t ss[8]; + + cx->ks[0] = ss[0] = word_in(in_key, 0); + cx->ks[1] = ss[1] = word_in(in_key, 1); + cx->ks[2] = ss[2] = word_in(in_key, 2); + cx->ks[3] = ss[3] = word_in(in_key, 3); + cx->ks[4] = ss[4] = word_in(in_key, 4); + cx->ks[5] = ss[5] = word_in(in_key, 5); + cx->ks[6] = ss[6] = word_in(in_key, 6); + cx->ks[7] = ss[7] = word_in(in_key, 7); + +#if ENC_UNROLL == NONE + { aes_32t i; + + for(i = 0; i < (15 * N_COLS - 1) / 8; ++i) + ke8(cx->ks, i); + } +#else + ke8(cx->ks, 0); ke8(cx->ks, 1); + ke8(cx->ks, 2); ke8(cx->ks, 3); + ke8(cx->ks, 4); ke8(cx->ks, 5); + kel8(cx->ks, 6); +#endif +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(AES_VAR) + +aes_rval aes_encrypt_key(const void *in_key, int key_len, aes_encrypt_ctx cx[1]) +{ + switch(key_len) + { +#ifdef AES_ERR_CHK + case 16: case 128: return aes_encrypt_key128(in_key, cx); + case 24: case 192: return aes_encrypt_key192(in_key, cx); + case 32: case 256: return aes_encrypt_key256(in_key, cx); + default: return aes_error; +#else + case 16: case 128: aes_encrypt_key128(in_key, cx); return; + case 24: case 192: aes_encrypt_key192(in_key, cx); return; + case 32: case 256: aes_encrypt_key256(in_key, cx); return; +#endif + } +} + +#endif + +#endif + +#if defined(DECRYPTION_KEY_SCHEDULE) + +#if DEC_ROUND == NO_TABLES +#define ff(x) (x) +#else +#define ff(x) inv_mcol(x) +#ifdef dec_imvars +#define d_vars dec_imvars +#endif +#endif + +#if 1 +#define kdf4(k,i) \ +{ ss[0] = ss[0] ^ ss[2] ^ ss[1] ^ ss[3]; ss[1] = ss[1] ^ ss[3]; ss[2] = ss[2] ^ ss[3]; ss[3] = ss[3]; \ + ss[4] = ls_box(ss[(i+3) % 4], 3) ^ t_use(r,c)[i]; ss[i % 4] ^= ss[4]; \ + ss[4] ^= k[4*(i)]; k[4*(i)+4] = ff(ss[4]); ss[4] ^= k[4*(i)+1]; k[4*(i)+5] = ff(ss[4]); \ + ss[4] ^= k[4*(i)+2]; k[4*(i)+6] = ff(ss[4]); ss[4] ^= k[4*(i)+3]; k[4*(i)+7] = ff(ss[4]); \ +} +#define kd4(k,i) \ +{ ss[4] = ls_box(ss[(i+3) % 4], 3) ^ t_use(r,c)[i]; ss[i % 4] ^= ss[4]; ss[4] = ff(ss[4]); \ + k[4*(i)+4] = ss[4] ^= k[4*(i)]; k[4*(i)+5] = ss[4] ^= k[4*(i)+1]; \ + k[4*(i)+6] = ss[4] ^= k[4*(i)+2]; k[4*(i)+7] = ss[4] ^= k[4*(i)+3]; \ +} +#define kdl4(k,i) \ +{ ss[4] = ls_box(ss[(i+3) % 4], 3) ^ t_use(r,c)[i]; ss[i % 4] ^= ss[4]; \ + k[4*(i)+4] = (ss[0] ^= ss[1]) ^ ss[2] ^ ss[3]; k[4*(i)+5] = ss[1] ^ ss[3]; \ + k[4*(i)+6] = ss[0]; k[4*(i)+7] = ss[1]; \ +} +#else +#define kdf4(k,i) \ +{ ss[0] ^= ls_box(ss[3],3) ^ t_use(r,c)[i]; k[4*(i)+ 4] = ff(ss[0]); ss[1] ^= ss[0]; k[4*(i)+ 5] = ff(ss[1]); \ + ss[2] ^= ss[1]; k[4*(i)+ 6] = ff(ss[2]); ss[3] ^= ss[2]; k[4*(i)+ 7] = ff(ss[3]); \ +} +#define kd4(k,i) \ +{ ss[4] = ls_box(ss[3],3) ^ t_use(r,c)[i]; \ + ss[0] ^= ss[4]; ss[4] = ff(ss[4]); k[4*(i)+ 4] = ss[4] ^= k[4*(i)]; \ + ss[1] ^= ss[0]; k[4*(i)+ 5] = ss[4] ^= k[4*(i)+ 1]; \ + ss[2] ^= ss[1]; k[4*(i)+ 6] = ss[4] ^= k[4*(i)+ 2]; \ + ss[3] ^= ss[2]; k[4*(i)+ 7] = ss[4] ^= k[4*(i)+ 3]; \ +} +#define kdl4(k,i) \ +{ ss[0] ^= ls_box(ss[3],3) ^ t_use(r,c)[i]; k[4*(i)+ 4] = ss[0]; ss[1] ^= ss[0]; k[4*(i)+ 5] = ss[1]; \ + ss[2] ^= ss[1]; k[4*(i)+ 6] = ss[2]; ss[3] ^= ss[2]; k[4*(i)+ 7] = ss[3]; \ +} +#endif + +#define kdf6(k,i) \ +{ ss[0] ^= ls_box(ss[5],3) ^ t_use(r,c)[i]; k[6*(i)+ 6] = ff(ss[0]); ss[1] ^= ss[0]; k[6*(i)+ 7] = ff(ss[1]); \ + ss[2] ^= ss[1]; k[6*(i)+ 8] = ff(ss[2]); ss[3] ^= ss[2]; k[6*(i)+ 9] = ff(ss[3]); \ + ss[4] ^= ss[3]; k[6*(i)+10] = ff(ss[4]); ss[5] ^= ss[4]; k[6*(i)+11] = ff(ss[5]); \ +} +#define kd6(k,i) \ +{ ss[6] = ls_box(ss[5],3) ^ t_use(r,c)[i]; \ + ss[0] ^= ss[6]; ss[6] = ff(ss[6]); k[6*(i)+ 6] = ss[6] ^= k[6*(i)]; \ + ss[1] ^= ss[0]; k[6*(i)+ 7] = ss[6] ^= k[6*(i)+ 1]; \ + ss[2] ^= ss[1]; k[6*(i)+ 8] = ss[6] ^= k[6*(i)+ 2]; \ + ss[3] ^= ss[2]; k[6*(i)+ 9] = ss[6] ^= k[6*(i)+ 3]; \ + ss[4] ^= ss[3]; k[6*(i)+10] = ss[6] ^= k[6*(i)+ 4]; \ + ss[5] ^= ss[4]; k[6*(i)+11] = ss[6] ^= k[6*(i)+ 5]; \ +} +#define kdl6(k,i) \ +{ ss[0] ^= ls_box(ss[5],3) ^ t_use(r,c)[i]; k[6*(i)+ 6] = ss[0]; ss[1] ^= ss[0]; k[6*(i)+ 7] = ss[1]; \ + ss[2] ^= ss[1]; k[6*(i)+ 8] = ss[2]; ss[3] ^= ss[2]; k[6*(i)+ 9] = ss[3]; \ +} + +#define kdf8(k,i) \ +{ ss[0] ^= ls_box(ss[7],3) ^ t_use(r,c)[i]; k[8*(i)+ 8] = ff(ss[0]); ss[1] ^= ss[0]; k[8*(i)+ 9] = ff(ss[1]); \ + ss[2] ^= ss[1]; k[8*(i)+10] = ff(ss[2]); ss[3] ^= ss[2]; k[8*(i)+11] = ff(ss[3]); \ + ss[4] ^= ls_box(ss[3],0); k[8*(i)+12] = ff(ss[4]); ss[5] ^= ss[4]; k[8*(i)+13] = ff(ss[5]); \ + ss[6] ^= ss[5]; k[8*(i)+14] = ff(ss[6]); ss[7] ^= ss[6]; k[8*(i)+15] = ff(ss[7]); \ +} +#define kd8(k,i) \ +{ aes_32t g = ls_box(ss[7],3) ^ t_use(r,c)[i]; \ + ss[0] ^= g; g = ff(g); k[8*(i)+ 8] = g ^= k[8*(i)]; \ + ss[1] ^= ss[0]; k[8*(i)+ 9] = g ^= k[8*(i)+ 1]; \ + ss[2] ^= ss[1]; k[8*(i)+10] = g ^= k[8*(i)+ 2]; \ + ss[3] ^= ss[2]; k[8*(i)+11] = g ^= k[8*(i)+ 3]; \ + g = ls_box(ss[3],0); \ + ss[4] ^= g; g = ff(g); k[8*(i)+12] = g ^= k[8*(i)+ 4]; \ + ss[5] ^= ss[4]; k[8*(i)+13] = g ^= k[8*(i)+ 5]; \ + ss[6] ^= ss[5]; k[8*(i)+14] = g ^= k[8*(i)+ 6]; \ + ss[7] ^= ss[6]; k[8*(i)+15] = g ^= k[8*(i)+ 7]; \ +} +#define kdl8(k,i) \ +{ ss[0] ^= ls_box(ss[7],3) ^ t_use(r,c)[i]; k[8*(i)+ 8] = ss[0]; ss[1] ^= ss[0]; k[8*(i)+ 9] = ss[1]; \ + ss[2] ^= ss[1]; k[8*(i)+10] = ss[2]; ss[3] ^= ss[2]; k[8*(i)+11] = ss[3]; \ +} + +#if defined(AES_128) || defined(AES_VAR) + +aes_rval aes_decrypt_key128(const void *in_key, aes_decrypt_ctx cx[1]) +{ aes_32t ss[5]; +#ifdef d_vars + d_vars; +#endif + cx->ks[0] = ss[0] = word_in(in_key, 0); + cx->ks[1] = ss[1] = word_in(in_key, 1); + cx->ks[2] = ss[2] = word_in(in_key, 2); + cx->ks[3] = ss[3] = word_in(in_key, 3); + +#if DEC_UNROLL == NONE + { aes_32t i; + + for(i = 0; i < (11 * N_COLS - 1) / 4; ++i) + ke4(cx->ks, i); +#if !(DEC_ROUND == NO_TABLES) + for(i = N_COLS; i < 10 * N_COLS; ++i) + cx->ks[i] = inv_mcol(cx->ks[i]); +#endif + } +#else + kdf4(cx->ks, 0); kd4(cx->ks, 1); + kd4(cx->ks, 2); kd4(cx->ks, 3); + kd4(cx->ks, 4); kd4(cx->ks, 5); + kd4(cx->ks, 6); kd4(cx->ks, 7); + kd4(cx->ks, 8); kdl4(cx->ks, 9); +#endif + + /* cx->ks[45] ^ cx->ks[52] ^ cx->ks[53] is zero for a 256 bit */ + /* key and must be non-zero for 128 and 192 bits keys */ + cx->ks[53] = cx->ks[45] = 0; + cx->ks[52] = 10; +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(AES_192) || defined(AES_VAR) + +aes_rval aes_decrypt_key192(const void *in_key, aes_decrypt_ctx cx[1]) +{ aes_32t ss[7]; +#ifdef d_vars + d_vars; +#endif + cx->ks[0] = ss[0] = word_in(in_key, 0); + cx->ks[1] = ss[1] = word_in(in_key, 1); + cx->ks[2] = ss[2] = word_in(in_key, 2); + cx->ks[3] = ss[3] = word_in(in_key, 3); + +#if DEC_UNROLL == NONE + cx->ks[4] = ss[4] = word_in(in_key, 4); + cx->ks[5] = ss[5] = word_in(in_key, 5); + { aes_32t i; + + for(i = 0; i < (13 * N_COLS - 1) / 6; ++i) + ke6(cx->ks, i); +#if !(DEC_ROUND == NO_TABLES) + for(i = N_COLS; i < 12 * N_COLS; ++i) + cx->ks[i] = inv_mcol(cx->ks[i]); +#endif + } +#else + cx->ks[4] = ff(ss[4] = word_in(in_key, 4)); + cx->ks[5] = ff(ss[5] = word_in(in_key, 5)); + kdf6(cx->ks, 0); kd6(cx->ks, 1); + kd6(cx->ks, 2); kd6(cx->ks, 3); + kd6(cx->ks, 4); kd6(cx->ks, 5); + kd6(cx->ks, 6); kdl6(cx->ks, 7); +#endif + + /* cx->ks[45] ^ cx->ks[52] ^ cx->ks[53] is zero for a 256 bit */ + /* key and must be non-zero for 128 and 192 bits keys */ + cx->ks[53] = cx->ks[45]; + cx->ks[52] = 12; +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(AES_256) || defined(AES_VAR) + +aes_rval aes_decrypt_key256(const void *in_key, aes_decrypt_ctx cx[1]) +{ aes_32t ss[8]; +#ifdef d_vars + d_vars; +#endif + cx->ks[0] = ss[0] = word_in(in_key, 0); + cx->ks[1] = ss[1] = word_in(in_key, 1); + cx->ks[2] = ss[2] = word_in(in_key, 2); + cx->ks[3] = ss[3] = word_in(in_key, 3); + +#if DEC_UNROLL == NONE + cx->ks[4] = ss[4] = word_in(in_key, 4); + cx->ks[5] = ss[5] = word_in(in_key, 5); + cx->ks[6] = ss[6] = word_in(in_key, 6); + cx->ks[7] = ss[7] = word_in(in_key, 7); + { aes_32t i; + + for(i = 0; i < (15 * N_COLS - 1) / 8; ++i) + ke8(cx->ks, i); +#if !(DEC_ROUND == NO_TABLES) + for(i = N_COLS; i < 14 * N_COLS; ++i) + cx->ks[i] = inv_mcol(cx->ks[i]); +#endif + } +#else + cx->ks[4] = ff(ss[4] = word_in(in_key, 4)); + cx->ks[5] = ff(ss[5] = word_in(in_key, 5)); + cx->ks[6] = ff(ss[6] = word_in(in_key, 6)); + cx->ks[7] = ff(ss[7] = word_in(in_key, 7)); + kdf8(cx->ks, 0); kd8(cx->ks, 1); + kd8(cx->ks, 2); kd8(cx->ks, 3); + kd8(cx->ks, 4); kd8(cx->ks, 5); + kdl8(cx->ks, 6); +#endif +#ifdef AES_ERR_CHK + return aes_good; +#endif +} + +#endif + +#if defined(AES_VAR) + +aes_rval aes_decrypt_key(const void *in_key, int key_len, aes_decrypt_ctx cx[1]) +{ + switch(key_len) + { +#ifdef AES_ERR_CHK + case 16: case 128: return aes_decrypt_key128(in_key, cx); + case 24: case 192: return aes_decrypt_key192(in_key, cx); + case 32: case 256: return aes_decrypt_key256(in_key, cx); + default: return aes_error; +#else + case 16: case 128: aes_decrypt_key128(in_key, cx); return; + case 24: case 192: aes_decrypt_key192(in_key, cx); return; + case 32: case 256: aes_decrypt_key256(in_key, cx); return; +#endif + } +} + +#endif + +#endif + +#if defined(__cplusplus) +} +#endif diff --git a/scripts/conf-w32brg/cipher/aesopt.h b/scripts/conf-w32brg/cipher/aesopt.h new file mode 100644 index 000000000..ba94f97b3 --- /dev/null +++ b/scripts/conf-w32brg/cipher/aesopt.h @@ -0,0 +1,1042 @@ +/* + --------------------------------------------------------------------------- + Copyright (c) 2003, Dr Brian Gladman <[email protected]>, Worcester, UK. + All rights reserved. + + LICENSE TERMS + + The free distribution and use of this software in both source and binary + form is allowed (with or without changes) provided that: + + 1. distributions of this source code include the above copyright + notice, this list of conditions and the following disclaimer; + + 2. distributions in binary form include the above copyright + notice, this list of conditions and the following disclaimer + in the documentation and/or other associated materials; + + 3. the copyright holder's name is not used to endorse products + built using this software without specific written permission. + + ALTERNATIVELY, provided that this notice is retained in full, this product + may be distributed under the terms of the GNU General Public License (GPL), + in which case the provisions of the GPL apply INSTEAD OF those given above. + + DISCLAIMER + + This software is provided 'as is' with no explicit or implied warranties + in respect of its properties, including, but not limited to, correctness + and/or fitness for purpose. + --------------------------------------------------------------------------- + Issue Date: 1/06/2003 + + My thanks go to Dag Arne Osvik for devising the schemes used here for key + length derivation from the form of the key schedule + + This file contains the compilation options for AES (Rijndael) and code + that is common across encryption, key scheduling and table generation. + + OPERATION + + These source code files implement the AES algorithm Rijndael designed by + Joan Daemen and Vincent Rijmen. This version is designed for the standard + block size of 16 bytes and for key sizes of 128, 192 and 256 bits (16, 24 + and 32 bytes). + + This version is designed for flexibility and speed using operations on + 32-bit words rather than operations on bytes. It can be compiled with + either big or little endian internal byte order but is faster when the + native byte order for the processor is used. + + THE CIPHER INTERFACE + + The cipher interface is implemented as an array of bytes in which lower + AES bit sequence indexes map to higher numeric significance within bytes. + + aes_08t (an unsigned 8-bit type) + aes_32t (an unsigned 32-bit type) + struct aes_encrypt_ctx (structure for the cipher encryption context) + struct aes_decrypt_ctx (structure for the cipher decryption context) + aes_rval the function return type + + C subroutine calls: + + aes_rval aes_encrypt_key128(const void *in_key, aes_encrypt_ctx cx[1]); + aes_rval aes_encrypt_key192(const void *in_key, aes_encrypt_ctx cx[1]); + aes_rval aes_encrypt_key256(const void *in_key, aes_encrypt_ctx cx[1]); + aes_rval aes_encrypt(const void *in_blk, + void *out_blk, const aes_encrypt_ctx cx[1]); + + aes_rval aes_decrypt_key128(const void *in_key, aes_decrypt_ctx cx[1]); + aes_rval aes_decrypt_key192(const void *in_key, aes_decrypt_ctx cx[1]); + aes_rval aes_decrypt_key256(const void *in_key, aes_decrypt_ctx cx[1]); + aes_rval aes_decrypt(const void *in_blk, + void *out_blk, const aes_decrypt_ctx cx[1]); + + IMPORTANT NOTE: If you are using this C interface with dynamic tables make sure that + you call genTabs() before AES is used so that the tables are initialised. + + C++ aes class subroutines: + + Class AESencrypt for encryption + + Construtors: + AESencrypt(void) + AESencrypt(const void *in_key) - 128 bit key + Members: + void key128(const void *in_key) + void key192(const void *in_key) + void key256(const void *in_key) + void encrypt(const void *in_blk, void *out_blk) const + + Class AESdecrypt for encryption + Construtors: + AESdecrypt(void) + AESdecrypt(const void *in_key) - 128 bit key + Members: + void key128(const void *in_key) + void key192(const void *in_key) + void key256(const void *in_key) + void decrypt(const void *in_blk, void *out_blk) const + + COMPILATION + + The files used to provide AES (Rijndael) are + + a. aes.h for the definitions needed for use in C. + b. aescpp.h for the definitions needed for use in C++. + c. aesopt.h for setting compilation options (also includes common code). + d. aescrypt.c for encryption and decrytpion, or + e. aeskey.c for key scheduling. + f. aestab.c for table loading or generation. + g. aescrypt.asm for encryption and decryption using assembler code. + h. aescrypt.mmx.asm for encryption and decryption using MMX assembler. + + To compile AES (Rijndael) for use in C code use aes.h and set the + defines here for the facilities you need (key lengths, encryption + and/or decryption). Do not define AES_DLL or AES_CPP. Set the options + for optimisations and table sizes here. + + To compile AES (Rijndael) for use in in C++ code use aescpp.h but do + not define AES_DLL + + To compile AES (Rijndael) in C as a Dynamic Link Library DLL) use + aes.h and include the AES_DLL define. + + CONFIGURATION OPTIONS (here and in aes.h) + + a. set AES_DLL in aes.h if AES (Rijndael) is to be compiled as a DLL + b. You may need to set PLATFORM_BYTE_ORDER to define the byte order. + c. If you want the code to run in a specific internal byte order, then + INTERNAL_BYTE_ORDER must be set accordingly. + d. set other configuration options decribed below. +*/ + +#ifndef _AESOPT_H +#define _AESOPT_H + +#if defined(__cplusplus) +extern "C" +{ +#endif + +/* START OF CONFIGURATION OPTIONS + + USE OF DEFINES + + Later in this section there are a number of defines that control the + operation of the code. In each section, the purpose of each define is + explained so that the relevant form can be included or excluded by + setting either 1's or 0's respectively on the branches of the related + #if clauses. +*/ + +/* DO NOT CHANGE THE FOLLOWING EIGHT DEFINES */ + +#define NO_TABLES 0 +#define ONE_TABLE 1 +#define FOUR_TABLES 4 +#define NONE 0 +#define PARTIAL 1 +#define FULL 2 +#define AES_LITTLE_ENDIAN 1234 /* byte 0 is least significant (i386) */ +#define AES_BIG_ENDIAN 4321 /* byte 0 is most significant (mc68k) */ + +/* 1. PLATFORM SPECIFIC INCLUDES */ + +#if defined( __CRYPTLIB__ ) && !defined( INC_ALL ) && !defined( INC_CHILD ) +#include "crypt/aes.h" +#else + #include "aes.h" +#endif + +#if defined(__GNUC__) || defined(__GNU_LIBRARY__) +# if defined( __FreeBSD__ ) || defined( __OpenBSD__ ) +# include <sys/endian.h> +# elif defined( __APPLE__ ) +# if defined( __BIG_ENDIAN__ ) && !defined( BIG_ENDIAN ) +# define BIG_ENDIAN +# elif defined( __LITTLE_ENDIAN__ ) && !defined( LITTLE_ENDIAN ) +# define LITTLE_ENDIAN +# else +# error Need to define CPU endianness for OS X +# endif +# else +# include <endian.h> +# include <byteswap.h> +# endif /* *BSDs don't use standard Gnu setup */ +#elif defined(__CRYPTLIB__) +# if defined( INC_ALL ) +# include "crypt.h" +# elif defined( INC_CHILD ) +# include "../crypt.h" +# else +# include "crypt.h" +# endif +# if defined(DATA_LITTLEENDIAN) +# define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +# else +# define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +# endif +#elif defined(_MSC_VER) +# include <stdlib.h> +#elif !defined(WIN32) +# include <stdlib.h> +# if defined (_ENDIAN_H) +# include "endian.h" +# else +# include <sys/param.h> +# endif +#endif + +#if defined(bswap32) +#define aes_sw32 bswap32 +#elif defined(bswap_32) +#define aes_sw32 bswap_32 +#endif + +/* 2. BYTE ORDER IN 32-BIT WORDS + + To obtain the highest speed on processors with 32-bit words, this code + needs to determine the order in which bytes are packed into such words. + The following block of code is an attempt to capture the most obvious + ways in which various environemnts define byte order. It may well fail, + in which case the definitions will need to be set by editing at the + points marked **** EDIT HERE IF NECESSARY **** below. +*/ +#if !defined(PLATFORM_BYTE_ORDER) +#if defined(LITTLE_ENDIAN) || defined(BIG_ENDIAN) +# if defined(LITTLE_ENDIAN) && defined(BIG_ENDIAN) +# if defined(BYTE_ORDER) +# if (BYTE_ORDER == LITTLE_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +# elif (BYTE_ORDER == BIG_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +# endif +# endif +# elif defined(LITTLE_ENDIAN) && !defined(BIG_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +# elif !defined(LITTLE_ENDIAN) && defined(BIG_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +# endif +#elif defined(_LITTLE_ENDIAN) || defined(_BIG_ENDIAN) +# if defined(_LITTLE_ENDIAN) && defined(_BIG_ENDIAN) +# if defined(_BYTE_ORDER) +# if (_BYTE_ORDER == _LITTLE_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +# elif (_BYTE_ORDER == _BIG_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +# endif +# endif +# elif defined(_LITTLE_ENDIAN) && !defined(_BIG_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +# elif !defined(_LITTLE_ENDIAN) && defined(_BIG_ENDIAN) +# define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +# endif +#elif 0 /* **** EDIT HERE IF NECESSARY **** */ +#define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +#elif 0 /* **** EDIT HERE IF NECESSARY **** */ +#define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +#elif (('1234' >> 24) == '1') +# define PLATFORM_BYTE_ORDER AES_LITTLE_ENDIAN +#elif (('4321' >> 24) == '1') +# define PLATFORM_BYTE_ORDER AES_BIG_ENDIAN +#endif +#endif + +#if !defined(PLATFORM_BYTE_ORDER) +# error Please set undetermined byte order (lines 241 or 243 of aesopt.h). +#endif + +/* 3. FUNCTIONS REQUIRED + + This implementation provides subroutines for encryption, decryption + and for setting the three key lengths (separately) for encryption + and decryption. When the assembler code is not being used the following + definition blocks allow the selection of the routines that are to be + included in the compilation. +*/ +#ifdef AES_ENCRYPT +#define ENCRYPTION +#define ENCRYPTION_KEY_SCHEDULE +#endif + +#ifdef AES_DECRYPT +#define DECRYPTION +#define DECRYPTION_KEY_SCHEDULE +#endif + +/* 4. ASSEMBLER SUPPORT + + This define (which can be on the command line) enables the use of the + assembler code routines for encryption and decryption with the C code + only providing key scheduling +*/ +#if 0 +#define AES_ASM +#endif + +/* 5. BYTE ORDER WITHIN 32 BIT WORDS + + The fundamental data processing units in Rijndael are 8-bit bytes. The + input, output and key input are all enumerated arrays of bytes in which + bytes are numbered starting at zero and increasing to one less than the + number of bytes in the array in question. This enumeration is only used + for naming bytes and does not imply any adjacency or order relationship + from one byte to another. When these inputs and outputs are considered + as bit sequences, bits 8*n to 8*n+7 of the bit sequence are mapped to + byte[n] with bit 8n+i in the sequence mapped to bit 7-i within the byte. + In this implementation bits are numbered from 0 to 7 starting at the + numerically least significant end of each byte (bit n represents 2^n). + + However, Rijndael can be implemented more efficiently using 32-bit + words by packing bytes into words so that bytes 4*n to 4*n+3 are placed + into word[n]. While in principle these bytes can be assembled into words + in any positions, this implementation only supports the two formats in + which bytes in adjacent positions within words also have adjacent byte + numbers. This order is called big-endian if the lowest numbered bytes + in words have the highest numeric significance and little-endian if the + opposite applies. + + This code can work in either order irrespective of the order used by the + machine on which it runs. Normally the internal byte order will be set + to the order of the processor on which the code is to be run but this + define can be used to reverse this in special situations + + NOTE: Assembler code versions rely on PLATFORM_BYTE_ORDER being set +*/ +#if 1 || defined(AES_ASM) +#define INTERNAL_BYTE_ORDER PLATFORM_BYTE_ORDER +#elif 0 +#define INTERNAL_BYTE_ORDER AES_LITTLE_ENDIAN +#elif 0 +#define INTERNAL_BYTE_ORDER AES_BIG_ENDIAN +#else +#error The internal byte order is not defined +#endif + +/* 6. FAST INPUT/OUTPUT OPERATIONS. + + On some machines it is possible to improve speed by transferring the + bytes in the input and output arrays to and from the internal 32-bit + variables by addressing these arrays as if they are arrays of 32-bit + words. On some machines this will always be possible but there may + be a large performance penalty if the byte arrays are not aligned on + the normal word boundaries. On other machines this technique will + lead to memory access errors when such 32-bit word accesses are not + properly aligned. The option SAFE_IO avoids such problems but will + often be slower on those machines that support misaligned access + (especially so if care is taken to align the input and output byte + arrays on 32-bit word boundaries). If SAFE_IO is not defined it is + assumed that access to byte arrays as if they are arrays of 32-bit + words will not cause problems when such accesses are misaligned. +*/ +#if 1 && !defined(_MSC_VER) +#define SAFE_IO +#endif + +/* 7. LOOP UNROLLING + + The code for encryption and decrytpion cycles through a number of rounds + that can be implemented either in a loop or by expanding the code into a + long sequence of instructions, the latter producing a larger program but + one that will often be much faster. The latter is called loop unrolling. + There are also potential speed advantages in expanding two iterations in + a loop with half the number of iterations, which is called partial loop + unrolling. The following options allow partial or full loop unrolling + to be set independently for encryption and decryption +*/ +#if 1 +#define ENC_UNROLL FULL +#elif 0 +#define ENC_UNROLL PARTIAL +#else +#define ENC_UNROLL NONE +#endif + +#if 1 +#define DEC_UNROLL FULL +#elif 0 +#define DEC_UNROLL PARTIAL +#else +#define DEC_UNROLL NONE +#endif + +/* 8. FAST FINITE FIELD OPERATIONS + + If this section is included, tables are used to provide faster finite + field arithmetic (this has no effect if FIXED_TABLES is defined). +*/ +#if 1 +#define FF_TABLES +#endif + +/* 9. INTERNAL STATE VARIABLE FORMAT + + The internal state of Rijndael is stored in a number of local 32-bit + word varaibles which can be defined either as an array or as individual + names variables. Include this section if you want to store these local + varaibles in arrays. Otherwise individual local variables will be used. +*/ +#if 1 +#define ARRAYS +#endif + +/* In this implementation the columns of the state array are each held in + 32-bit words. The state array can be held in various ways: in an array + of words, in a number of individual word variables or in a number of + processor registers. The following define maps a variable name x and + a column number c to the way the state array variable is to be held. + The first define below maps the state into an array x[c] whereas the + second form maps the state into a number of individual variables x0, + x1, etc. Another form could map individual state colums to machine + register names. +*/ + +#if defined(ARRAYS) +#define s(x,c) x[c] +#else +#define s(x,c) x##c +#endif + +/* 10. FIXED OR DYNAMIC TABLES + + When this section is included the tables used by the code are compiled + statically into the binary file. Otherwise the subroutine gen_tabs() + must be called to compute them before the code is first used. +*/ +#if 1 +#define FIXED_TABLES +#endif + +/* 11. TABLE ALIGNMENT + + On some sytsems speed will be improved by aligning the AES large lookup + tables on particular boundaries. This define should be set to a power of + two giving the desired alignment. It can be left undefined if alignment + is not needed. This option is specific to the Microsft VC++ compiler. +*/ + +#define TABLE_ALIGN 64 + +/* 12. INTERNAL TABLE CONFIGURATION + + This cipher proceeds by repeating in a number of cycles known as 'rounds' + which are implemented by a round function which can optionally be speeded + up using tables. The basic tables are each 256 32-bit words, with either + one or four tables being required for each round function depending on + how much speed is required. The encryption and decryption round functions + are different and the last encryption and decrytpion round functions are + different again making four different round functions in all. + + This means that: + 1. Normal encryption and decryption rounds can each use either 0, 1 + or 4 tables and table spaces of 0, 1024 or 4096 bytes each. + 2. The last encryption and decryption rounds can also use either 0, 1 + or 4 tables and table spaces of 0, 1024 or 4096 bytes each. + + Include or exclude the appropriate definitions below to set the number + of tables used by this implementation. +*/ + +#if 1 /* set tables for the normal encryption round */ +#define ENC_ROUND FOUR_TABLES +#elif 0 +#define ENC_ROUND ONE_TABLE +#else +#define ENC_ROUND NO_TABLES +#endif + +#if 1 /* set tables for the last encryption round */ +#define LAST_ENC_ROUND FOUR_TABLES +#elif 0 +#define LAST_ENC_ROUND ONE_TABLE +#else +#define LAST_ENC_ROUND NO_TABLES +#endif + +#if 1 /* set tables for the normal decryption round */ +#define DEC_ROUND FOUR_TABLES +#elif 0 +#define DEC_ROUND ONE_TABLE +#else +#define DEC_ROUND NO_TABLES +#endif + +#if 1 /* set tables for the last decryption round */ +#define LAST_DEC_ROUND FOUR_TABLES +#elif 0 +#define LAST_DEC_ROUND ONE_TABLE +#else +#define LAST_DEC_ROUND NO_TABLES +#endif + +/* The decryption key schedule can be speeded up with tables in the same + way that the round functions can. Include or exclude the following + defines to set this requirement. +*/ +#if 1 +#define KEY_SCHED FOUR_TABLES +#elif 0 +#define KEY_SCHED ONE_TABLE +#else +#define KEY_SCHED NO_TABLES +#endif + +/* END OF CONFIGURATION OPTIONS */ + +#define RC_LENGTH (5 * (AES_BLOCK_SIZE / 4 - 2)) + +/* Disable or report errors on some combinations of options */ + +#if ENC_ROUND == NO_TABLES && LAST_ENC_ROUND != NO_TABLES +#undef LAST_ENC_ROUND +#define LAST_ENC_ROUND NO_TABLES +#elif ENC_ROUND == ONE_TABLE && LAST_ENC_ROUND == FOUR_TABLES +#undef LAST_ENC_ROUND +#define LAST_ENC_ROUND ONE_TABLE +#endif + +#if ENC_ROUND == NO_TABLES && ENC_UNROLL != NONE +#undef ENC_UNROLL +#define ENC_UNROLL NONE +#endif + +#if DEC_ROUND == NO_TABLES && LAST_DEC_ROUND != NO_TABLES +#undef LAST_DEC_ROUND +#define LAST_DEC_ROUND NO_TABLES +#elif DEC_ROUND == ONE_TABLE && LAST_DEC_ROUND == FOUR_TABLES +#undef LAST_DEC_ROUND +#define LAST_DEC_ROUND ONE_TABLE +#endif + +#if DEC_ROUND == NO_TABLES && DEC_UNROLL != NONE +#undef DEC_UNROLL +#define DEC_UNROLL NONE +#endif + +/* upr(x,n): rotates bytes within words by n positions, moving bytes to + higher index positions with wrap around into low positions + ups(x,n): moves bytes by n positions to higher index positions in + words but without wrap around + bval(x,n): extracts a byte from a word + + NOTE: The definitions given here are intended only for use with + unsigned variables and with shift counts that are compile + time constants +*/ + +#if (INTERNAL_BYTE_ORDER == AES_LITTLE_ENDIAN) +#define upr(x,n) (((aes_32t)(x) << (8 * (n))) | ((aes_32t)(x) >> (32 - 8 * (n)))) +#define ups(x,n) ((aes_32t) (x) << (8 * (n))) +#define bval(x,n) ((aes_08t)((x) >> (8 * (n)))) +#define bytes2word(b0, b1, b2, b3) \ + (((aes_32t)(b3) << 24) | ((aes_32t)(b2) << 16) | ((aes_32t)(b1) << 8) | (b0)) +#endif + +#if (INTERNAL_BYTE_ORDER == AES_BIG_ENDIAN) +#define upr(x,n) (((aes_32t)(x) >> (8 * (n))) | ((aes_32t)(x) << (32 - 8 * (n)))) +#define ups(x,n) ((aes_32t) (x) >> (8 * (n)))) +#define bval(x,n) ((aes_08t)((x) >> (24 - 8 * (n)))) +#define bytes2word(b0, b1, b2, b3) \ + (((aes_32t)(b0) << 24) | ((aes_32t)(b1) << 16) | ((aes_32t)(b2) << 8) | (b3)) +#endif + +#if defined(SAFE_IO) + +#define word_in(x,c) bytes2word(((aes_08t*)(x)+4*c)[0], ((aes_08t*)(x)+4*c)[1], \ + ((aes_08t*)(x)+4*c)[2], ((aes_08t*)(x)+4*c)[3]) +#define word_out(x,c,v) { ((aes_08t*)(x)+4*c)[0] = bval(v,0); ((aes_08t*)(x)+4*c)[1] = bval(v,1); \ + ((aes_08t*)(x)+4*c)[2] = bval(v,2); ((aes_08t*)(x)+4*c)[3] = bval(v,3); } + +#elif (INTERNAL_BYTE_ORDER == PLATFORM_BYTE_ORDER) + +#define word_in(x,c) (*((aes_32t*)(x)+(c))) +#define word_out(x,c,v) (*((aes_32t*)(x)+(c)) = (v)) + +#else + +#ifndef aes_sw32 +#define brot(x,n) (((aes_32t)(x) << n) | ((aes_32t)(x) >> (32 - n))) +#define aes_sw32(x) ((brot((x),8) & 0x00ff00ff) | (brot((x),24) & 0xff00ff00)) +#endif + +#define word_in(x,c) aes_sw32(*((aes_32t*)(x)+(c))) +#define word_out(x,c,v) (*((aes_32t*)(x)+(c)) = aes_sw32(v)) + +#endif + +/* the finite field modular polynomial and elements */ + +#define WPOLY 0x011b +#define BPOLY 0x1b + +/* multiply four bytes in GF(2^8) by 'x' {02} in parallel */ + +#define m1 0x80808080 +#define m2 0x7f7f7f7f +#define gf_mulx(x) ((((x) & m2) << 1) ^ ((((x) & m1) >> 7) * BPOLY)) + +/* The following defines provide alternative definitions of gf_mulx that might + give improved performance if a fast 32-bit multiply is not available. Note + that a temporary variable u needs to be defined where gf_mulx is used. + +#define gf_mulx(x) (u = (x) & m1, u |= (u >> 1), ((x) & m2) << 1) ^ ((u >> 3) | (u >> 6)) +#define m4 (0x01010101 * BPOLY) +#define gf_mulx(x) (u = (x) & m1, ((x) & m2) << 1) ^ ((u - (u >> 7)) & m4) +*/ + +/* Work out which tables are needed for the different options */ + +#ifdef AES_ASM +#ifdef ENC_ROUND +#undef ENC_ROUND +#endif +#define ENC_ROUND FOUR_TABLES +#ifdef LAST_ENC_ROUND +#undef LAST_ENC_ROUND +#endif +#define LAST_ENC_ROUND FOUR_TABLES +#ifdef DEC_ROUND +#undef DEC_ROUND +#endif +#define DEC_ROUND FOUR_TABLES +#ifdef LAST_DEC_ROUND +#undef LAST_DEC_ROUND +#endif +#define LAST_DEC_ROUND FOUR_TABLES +#ifdef KEY_SCHED +#undef KEY_SCHED +#define KEY_SCHED FOUR_TABLES +#endif +#endif + +#if defined(ENCRYPTION) || defined(AES_ASM) +#if ENC_ROUND == ONE_TABLE +#define FT1_SET +#elif ENC_ROUND == FOUR_TABLES +#define FT4_SET +#else +#define SBX_SET +#endif +#if LAST_ENC_ROUND == ONE_TABLE +#define FL1_SET +#elif LAST_ENC_ROUND == FOUR_TABLES +#define FL4_SET +#elif !defined(SBX_SET) +#define SBX_SET +#endif +#endif + +#if defined(DECRYPTION) || defined(AES_ASM) +#if DEC_ROUND == ONE_TABLE +#define IT1_SET +#elif DEC_ROUND == FOUR_TABLES +#define IT4_SET +#else +#define ISB_SET +#endif +#if LAST_DEC_ROUND == ONE_TABLE +#define IL1_SET +#elif LAST_DEC_ROUND == FOUR_TABLES +#define IL4_SET +#elif !defined(ISB_SET) +#define ISB_SET +#endif +#endif + +#if defined(ENCRYPTION_KEY_SCHEDULE) || defined(DECRYPTION_KEY_SCHEDULE) +#if KEY_SCHED == ONE_TABLE +#define LS1_SET +#define IM1_SET +#elif KEY_SCHED == FOUR_TABLES +#define LS4_SET +#define IM4_SET +#elif !defined(SBX_SET) +#define SBX_SET +#endif +#endif + +/* If there are no global variables, the AES tables are placed in + a structure and a pointer is added to the AES context. If this + facility is used, the calling program has to ensure that this + pointer is managed appropriately. In particular, the value of + the t_dec(in,it) item in the table structure must be set to zero + in order to ensure that the tables are initialised. In practice + the three code sequences in aeskey.c that control the calls to + gen_tabs() and the gen_tabs() routine itself will require some + changes for a specific implementation. If global variables are + available it will generally be preferable to use them with the + precomputed FIXED_TABLES option that uses static global tables. + + The following defines can be used to control the way the tables + are defined, initialised and used in embedded environments that + require special features for these purposes + + the 't_dec' construction is used to declare fixed table arrays + the 't_set' construction is used to set fixed table values + the 't_use' construction is used to access fixed table values + + 256 byte tables: + + t_xxx(s,box) => forward S box + t_xxx(i,box) => inverse S box + + 256 32-bit word OR 4 x 256 32-bit word tables: + + t_xxx(f,n) => forward normal round + t_xxx(f,l) => forward last round + t_xxx(i,n) => inverse normal round + t_xxx(i,l) => inverse last round + t_xxx(l,s) => key schedule table + t_xxx(i,m) => key schedule table + + Other variables and tables: + + t_xxx(r,c) => the rcon table +*/ + +#define t_dec(m,n) t_##m##n +#define t_set(m,n) t_##m##n +#define t_use(m,n) t_##m##n + +#if defined(DO_TABLES) /* declare and instantiate tables */ + +/* finite field arithmetic operations for table generation */ + +#if defined(FIXED_TABLES) || !defined(FF_TABLES) + +#define f2(x) ((x<<1) ^ (((x>>7) & 1) * WPOLY)) +#define f4(x) ((x<<2) ^ (((x>>6) & 1) * WPOLY) ^ (((x>>6) & 2) * WPOLY)) +#define f8(x) ((x<<3) ^ (((x>>5) & 1) * WPOLY) ^ (((x>>5) & 2) * WPOLY) \ + ^ (((x>>5) & 4) * WPOLY)) +#define f3(x) (f2(x) ^ x) +#define f9(x) (f8(x) ^ x) +#define fb(x) (f8(x) ^ f2(x) ^ x) +#define fd(x) (f8(x) ^ f4(x) ^ x) +#define fe(x) (f8(x) ^ f4(x) ^ f2(x)) + +#else + +#define f2(x) ((x) ? pow[log[x] + 0x19] : 0) +#define f3(x) ((x) ? pow[log[x] + 0x01] : 0) +#define f9(x) ((x) ? pow[log[x] + 0xc7] : 0) +#define fb(x) ((x) ? pow[log[x] + 0x68] : 0) +#define fd(x) ((x) ? pow[log[x] + 0xee] : 0) +#define fe(x) ((x) ? pow[log[x] + 0xdf] : 0) +#define fi(x) ((x) ? pow[ 255 - log[x]] : 0) + +#endif + +#if defined(FIXED_TABLES) /* declare and set values for static tables */ + +#define sb_data(w) \ + w(0x63), w(0x7c), w(0x77), w(0x7b), w(0xf2), w(0x6b), w(0x6f), w(0xc5),\ + w(0x30), w(0x01), w(0x67), w(0x2b), w(0xfe), w(0xd7), w(0xab), w(0x76),\ + w(0xca), w(0x82), w(0xc9), w(0x7d), w(0xfa), w(0x59), w(0x47), w(0xf0),\ + w(0xad), w(0xd4), w(0xa2), w(0xaf), w(0x9c), w(0xa4), w(0x72), w(0xc0),\ + w(0xb7), w(0xfd), w(0x93), w(0x26), w(0x36), w(0x3f), w(0xf7), w(0xcc),\ + w(0x34), w(0xa5), w(0xe5), w(0xf1), w(0x71), w(0xd8), w(0x31), w(0x15),\ + w(0x04), w(0xc7), w(0x23), w(0xc3), w(0x18), w(0x96), w(0x05), w(0x9a),\ + w(0x07), w(0x12), w(0x80), w(0xe2), w(0xeb), w(0x27), w(0xb2), w(0x75),\ + w(0x09), w(0x83), w(0x2c), w(0x1a), w(0x1b), w(0x6e), w(0x5a), w(0xa0),\ + w(0x52), w(0x3b), w(0xd6), w(0xb3), w(0x29), w(0xe3), w(0x2f), w(0x84),\ + w(0x53), w(0xd1), w(0x00), w(0xed), w(0x20), w(0xfc), w(0xb1), w(0x5b),\ + w(0x6a), w(0xcb), w(0xbe), w(0x39), w(0x4a), w(0x4c), w(0x58), w(0xcf),\ + w(0xd0), w(0xef), w(0xaa), w(0xfb), w(0x43), w(0x4d), w(0x33), w(0x85),\ + w(0x45), w(0xf9), w(0x02), w(0x7f), w(0x50), w(0x3c), w(0x9f), w(0xa8),\ + w(0x51), w(0xa3), w(0x40), w(0x8f), w(0x92), w(0x9d), w(0x38), w(0xf5),\ + w(0xbc), w(0xb6), w(0xda), w(0x21), w(0x10), w(0xff), w(0xf3), w(0xd2),\ + w(0xcd), w(0x0c), w(0x13), w(0xec), w(0x5f), w(0x97), w(0x44), w(0x17),\ + w(0xc4), w(0xa7), w(0x7e), w(0x3d), w(0x64), w(0x5d), w(0x19), w(0x73),\ + w(0x60), w(0x81), w(0x4f), w(0xdc), w(0x22), w(0x2a), w(0x90), w(0x88),\ + w(0x46), w(0xee), w(0xb8), w(0x14), w(0xde), w(0x5e), w(0x0b), w(0xdb),\ + w(0xe0), w(0x32), w(0x3a), w(0x0a), w(0x49), w(0x06), w(0x24), w(0x5c),\ + w(0xc2), w(0xd3), w(0xac), w(0x62), w(0x91), w(0x95), w(0xe4), w(0x79),\ + w(0xe7), w(0xc8), w(0x37), w(0x6d), w(0x8d), w(0xd5), w(0x4e), w(0xa9),\ + w(0x6c), w(0x56), w(0xf4), w(0xea), w(0x65), w(0x7a), w(0xae), w(0x08),\ + w(0xba), w(0x78), w(0x25), w(0x2e), w(0x1c), w(0xa6), w(0xb4), w(0xc6),\ + w(0xe8), w(0xdd), w(0x74), w(0x1f), w(0x4b), w(0xbd), w(0x8b), w(0x8a),\ + w(0x70), w(0x3e), w(0xb5), w(0x66), w(0x48), w(0x03), w(0xf6), w(0x0e),\ + w(0x61), w(0x35), w(0x57), w(0xb9), w(0x86), w(0xc1), w(0x1d), w(0x9e),\ + w(0xe1), w(0xf8), w(0x98), w(0x11), w(0x69), w(0xd9), w(0x8e), w(0x94),\ + w(0x9b), w(0x1e), w(0x87), w(0xe9), w(0xce), w(0x55), w(0x28), w(0xdf),\ + w(0x8c), w(0xa1), w(0x89), w(0x0d), w(0xbf), w(0xe6), w(0x42), w(0x68),\ + w(0x41), w(0x99), w(0x2d), w(0x0f), w(0xb0), w(0x54), w(0xbb), w(0x16) + +#define isb_data(w) \ + w(0x52), w(0x09), w(0x6a), w(0xd5), w(0x30), w(0x36), w(0xa5), w(0x38),\ + w(0xbf), w(0x40), w(0xa3), w(0x9e), w(0x81), w(0xf3), w(0xd7), w(0xfb),\ + w(0x7c), w(0xe3), w(0x39), w(0x82), w(0x9b), w(0x2f), w(0xff), w(0x87),\ + w(0x34), w(0x8e), w(0x43), w(0x44), w(0xc4), w(0xde), w(0xe9), w(0xcb),\ + w(0x54), w(0x7b), w(0x94), w(0x32), w(0xa6), w(0xc2), w(0x23), w(0x3d),\ + w(0xee), w(0x4c), w(0x95), w(0x0b), w(0x42), w(0xfa), w(0xc3), w(0x4e),\ + w(0x08), w(0x2e), w(0xa1), w(0x66), w(0x28), w(0xd9), w(0x24), w(0xb2),\ + w(0x76), w(0x5b), w(0xa2), w(0x49), w(0x6d), w(0x8b), w(0xd1), w(0x25),\ + w(0x72), w(0xf8), w(0xf6), w(0x64), w(0x86), w(0x68), w(0x98), w(0x16),\ + w(0xd4), w(0xa4), w(0x5c), w(0xcc), w(0x5d), w(0x65), w(0xb6), w(0x92),\ + w(0x6c), w(0x70), w(0x48), w(0x50), w(0xfd), w(0xed), w(0xb9), w(0xda),\ + w(0x5e), w(0x15), w(0x46), w(0x57), w(0xa7), w(0x8d), w(0x9d), w(0x84),\ + w(0x90), w(0xd8), w(0xab), w(0x00), w(0x8c), w(0xbc), w(0xd3), w(0x0a),\ + w(0xf7), w(0xe4), w(0x58), w(0x05), w(0xb8), w(0xb3), w(0x45), w(0x06),\ + w(0xd0), w(0x2c), w(0x1e), w(0x8f), w(0xca), w(0x3f), w(0x0f), w(0x02),\ + w(0xc1), w(0xaf), w(0xbd), w(0x03), w(0x01), w(0x13), w(0x8a), w(0x6b),\ + w(0x3a), w(0x91), w(0x11), w(0x41), w(0x4f), w(0x67), w(0xdc), w(0xea),\ + w(0x97), w(0xf2), w(0xcf), w(0xce), w(0xf0), w(0xb4), w(0xe6), w(0x73),\ + w(0x96), w(0xac), w(0x74), w(0x22), w(0xe7), w(0xad), w(0x35), w(0x85),\ + w(0xe2), w(0xf9), w(0x37), w(0xe8), w(0x1c), w(0x75), w(0xdf), w(0x6e),\ + w(0x47), w(0xf1), w(0x1a), w(0x71), w(0x1d), w(0x29), w(0xc5), w(0x89),\ + w(0x6f), w(0xb7), w(0x62), w(0x0e), w(0xaa), w(0x18), w(0xbe), w(0x1b),\ + w(0xfc), w(0x56), w(0x3e), w(0x4b), w(0xc6), w(0xd2), w(0x79), w(0x20),\ + w(0x9a), w(0xdb), w(0xc0), w(0xfe), w(0x78), w(0xcd), w(0x5a), w(0xf4),\ + w(0x1f), w(0xdd), w(0xa8), w(0x33), w(0x88), w(0x07), w(0xc7), w(0x31),\ + w(0xb1), w(0x12), w(0x10), w(0x59), w(0x27), w(0x80), w(0xec), w(0x5f),\ + w(0x60), w(0x51), w(0x7f), w(0xa9), w(0x19), w(0xb5), w(0x4a), w(0x0d),\ + w(0x2d), w(0xe5), w(0x7a), w(0x9f), w(0x93), w(0xc9), w(0x9c), w(0xef),\ + w(0xa0), w(0xe0), w(0x3b), w(0x4d), w(0xae), w(0x2a), w(0xf5), w(0xb0),\ + w(0xc8), w(0xeb), w(0xbb), w(0x3c), w(0x83), w(0x53), w(0x99), w(0x61),\ + w(0x17), w(0x2b), w(0x04), w(0x7e), w(0xba), w(0x77), w(0xd6), w(0x26),\ + w(0xe1), w(0x69), w(0x14), w(0x63), w(0x55), w(0x21), w(0x0c), w(0x7d), + +#define mm_data(w) \ + w(0x00), w(0x01), w(0x02), w(0x03), w(0x04), w(0x05), w(0x06), w(0x07),\ + w(0x08), w(0x09), w(0x0a), w(0x0b), w(0x0c), w(0x0d), w(0x0e), w(0x0f),\ + w(0x10), w(0x11), w(0x12), w(0x13), w(0x14), w(0x15), w(0x16), w(0x17),\ + w(0x18), w(0x19), w(0x1a), w(0x1b), w(0x1c), w(0x1d), w(0x1e), w(0x1f),\ + w(0x20), w(0x21), w(0x22), w(0x23), w(0x24), w(0x25), w(0x26), w(0x27),\ + w(0x28), w(0x29), w(0x2a), w(0x2b), w(0x2c), w(0x2d), w(0x2e), w(0x2f),\ + w(0x30), w(0x31), w(0x32), w(0x33), w(0x34), w(0x35), w(0x36), w(0x37),\ + w(0x38), w(0x39), w(0x3a), w(0x3b), w(0x3c), w(0x3d), w(0x3e), w(0x3f),\ + w(0x40), w(0x41), w(0x42), w(0x43), w(0x44), w(0x45), w(0x46), w(0x47),\ + w(0x48), w(0x49), w(0x4a), w(0x4b), w(0x4c), w(0x4d), w(0x4e), w(0x4f),\ + w(0x50), w(0x51), w(0x52), w(0x53), w(0x54), w(0x55), w(0x56), w(0x57),\ + w(0x58), w(0x59), w(0x5a), w(0x5b), w(0x5c), w(0x5d), w(0x5e), w(0x5f),\ + w(0x60), w(0x61), w(0x62), w(0x63), w(0x64), w(0x65), w(0x66), w(0x67),\ + w(0x68), w(0x69), w(0x6a), w(0x6b), w(0x6c), w(0x6d), w(0x6e), w(0x6f),\ + w(0x70), w(0x71), w(0x72), w(0x73), w(0x74), w(0x75), w(0x76), w(0x77),\ + w(0x78), w(0x79), w(0x7a), w(0x7b), w(0x7c), w(0x7d), w(0x7e), w(0x7f),\ + w(0x80), w(0x81), w(0x82), w(0x83), w(0x84), w(0x85), w(0x86), w(0x87),\ + w(0x88), w(0x89), w(0x8a), w(0x8b), w(0x8c), w(0x8d), w(0x8e), w(0x8f),\ + w(0x90), w(0x91), w(0x92), w(0x93), w(0x94), w(0x95), w(0x96), w(0x97),\ + w(0x98), w(0x99), w(0x9a), w(0x9b), w(0x9c), w(0x9d), w(0x9e), w(0x9f),\ + w(0xa0), w(0xa1), w(0xa2), w(0xa3), w(0xa4), w(0xa5), w(0xa6), w(0xa7),\ + w(0xa8), w(0xa9), w(0xaa), w(0xab), w(0xac), w(0xad), w(0xae), w(0xaf),\ + w(0xb0), w(0xb1), w(0xb2), w(0xb3), w(0xb4), w(0xb5), w(0xb6), w(0xb7),\ + w(0xb8), w(0xb9), w(0xba), w(0xbb), w(0xbc), w(0xbd), w(0xbe), w(0xbf),\ + w(0xc0), w(0xc1), w(0xc2), w(0xc3), w(0xc4), w(0xc5), w(0xc6), w(0xc7),\ + w(0xc8), w(0xc9), w(0xca), w(0xcb), w(0xcc), w(0xcd), w(0xce), w(0xcf),\ + w(0xd0), w(0xd1), w(0xd2), w(0xd3), w(0xd4), w(0xd5), w(0xd6), w(0xd7),\ + w(0xd8), w(0xd9), w(0xda), w(0xdb), w(0xdc), w(0xdd), w(0xde), w(0xdf),\ + w(0xe0), w(0xe1), w(0xe2), w(0xe3), w(0xe4), w(0xe5), w(0xe6), w(0xe7),\ + w(0xe8), w(0xe9), w(0xea), w(0xeb), w(0xec), w(0xed), w(0xee), w(0xef),\ + w(0xf0), w(0xf1), w(0xf2), w(0xf3), w(0xf4), w(0xf5), w(0xf6), w(0xf7),\ + w(0xf8), w(0xf9), w(0xfa), w(0xfb), w(0xfc), w(0xfd), w(0xfe), w(0xff) + +#define h0(x) (x) + +/* These defines are used to ensure tables are generated in the + right format depending on the internal byte order required +*/ + +#define w0(p) bytes2word(p, 0, 0, 0) +#define w1(p) bytes2word(0, p, 0, 0) +#define w2(p) bytes2word(0, 0, p, 0) +#define w3(p) bytes2word(0, 0, 0, p) + +#define u0(p) bytes2word(f2(p), p, p, f3(p)) +#define u1(p) bytes2word(f3(p), f2(p), p, p) +#define u2(p) bytes2word(p, f3(p), f2(p), p) +#define u3(p) bytes2word(p, p, f3(p), f2(p)) + +#define v0(p) bytes2word(fe(p), f9(p), fd(p), fb(p)) +#define v1(p) bytes2word(fb(p), fe(p), f9(p), fd(p)) +#define v2(p) bytes2word(fd(p), fb(p), fe(p), f9(p)) +#define v3(p) bytes2word(f9(p), fd(p), fb(p), fe(p)) + +const aes_32t t_dec(r,c)[RC_LENGTH] = +{ + w0(0x01), w0(0x02), w0(0x04), w0(0x08), w0(0x10), + w0(0x20), w0(0x40), w0(0x80), w0(0x1b), w0(0x36) +}; + +#define d_1(t,n,b,v) const t n[256] = { b(v##0) } +#define d_4(t,n,b,v) const t n[4][256] = { { b(v##0) }, { b(v##1) }, { b(v##2) }, { b(v##3) } } + +#else /* declare and instantiate tables for dynamic value generation in in tab.c */ + +aes_32t t_dec(r,c)[RC_LENGTH]; + +#define d_1(t,n,b,v) t n[256] +#define d_4(t,n,b,v) t n[4][256] + +#endif + +#else /* declare tables without instantiation */ + +#if defined(FIXED_TABLES) + +extern const aes_32t t_dec(r,c)[RC_LENGTH]; + +#if defined(_MSC_VER) && defined(TABLE_ALIGN) +#define d_1(t,n,b,v) extern __declspec(align(TABLE_ALIGN)) const t n[256] +#define d_4(t,n,b,v) extern __declspec(align(TABLE_ALIGN)) const t n[4][256] +#else +#define d_1(t,n,b,v) extern const t n[256] +#define d_4(t,n,b,v) extern const t n[4][256] +#endif +#else + +extern aes_32t t_dec(r,c)[RC_LENGTH]; + +#if defined(_MSC_VER) && defined(TABLE_ALIGN) +#define d_1(t,n,b,v) extern __declspec(align(TABLE_ALIGN)) t n[256] +#define d_4(t,n,b,v) extern __declspec(align(TABLE_ALIGN)) t n[4][256] +#else +#define d_1(t,n,b,v) extern t n[256] +#define d_4(t,n,b,v) extern t n[4][256] +#endif +#endif + +#endif + +#ifdef SBX_SET + d_1(aes_08t, t_dec(s,box), sb_data, h); +#endif +#ifdef ISB_SET + d_1(aes_08t, t_dec(i,box), isb_data, h); +#endif + +#ifdef FT1_SET + d_1(aes_32t, t_dec(f,n), sb_data, u); +#endif +#ifdef FT4_SET + d_4(aes_32t, t_dec(f,n), sb_data, u); +#endif + +#ifdef FL1_SET + d_1(aes_32t, t_dec(f,l), sb_data, w); +#endif +#ifdef FL4_SET + d_4(aes_32t, t_dec(f,l), sb_data, w); +#endif + +#ifdef IT1_SET + d_1(aes_32t, t_dec(i,n), isb_data, v); +#endif +#ifdef IT4_SET + d_4(aes_32t, t_dec(i,n), isb_data, v); +#endif + +#ifdef IL1_SET + d_1(aes_32t, t_dec(i,l), isb_data, w); +#endif +#ifdef IL4_SET + d_4(aes_32t, t_dec(i,l), isb_data, w); +#endif + +#ifdef LS1_SET +#ifdef FL1_SET +#undef LS1_SET +#else + d_1(aes_32t, t_dec(l,s), sb_data, w); +#endif +#endif + +#ifdef LS4_SET +#ifdef FL4_SET +#undef LS4_SET +#else + d_4(aes_32t, t_dec(l,s), sb_data, w); +#endif +#endif + +#ifdef IM1_SET + d_1(aes_32t, t_dec(i,m), mm_data, v); +#endif +#ifdef IM4_SET + d_4(aes_32t, t_dec(i,m), mm_data, v); +#endif + +/* generic definitions of Rijndael macros that use tables */ + +#define no_table(x,box,vf,rf,c) bytes2word( \ + box[bval(vf(x,0,c),rf(0,c))], \ + box[bval(vf(x,1,c),rf(1,c))], \ + box[bval(vf(x,2,c),rf(2,c))], \ + box[bval(vf(x,3,c),rf(3,c))]) + +#define one_table(x,op,tab,vf,rf,c) \ + ( tab[bval(vf(x,0,c),rf(0,c))] \ + ^ op(tab[bval(vf(x,1,c),rf(1,c))],1) \ + ^ op(tab[bval(vf(x,2,c),rf(2,c))],2) \ + ^ op(tab[bval(vf(x,3,c),rf(3,c))],3)) + +#define four_tables(x,tab,vf,rf,c) \ + ( tab[0][bval(vf(x,0,c),rf(0,c))] \ + ^ tab[1][bval(vf(x,1,c),rf(1,c))] \ + ^ tab[2][bval(vf(x,2,c),rf(2,c))] \ + ^ tab[3][bval(vf(x,3,c),rf(3,c))]) + +#define vf1(x,r,c) (x) +#define rf1(r,c) (r) +#define rf2(r,c) ((8+r-c)&3) + +/* perform forward and inverse column mix operation on four bytes in long word x in */ +/* parallel. NOTE: x must be a simple variable, NOT an expression in these macros. */ + +#if defined(FM4_SET) /* not currently used */ +#define fwd_mcol(x) four_tables(x,t_use(f,m),vf1,rf1,0) +#elif defined(FM1_SET) /* not currently used */ +#define fwd_mcol(x) one_table(x,upr,t_use(f,m),vf1,rf1,0) +#else +#define dec_fmvars aes_32t g2 +#define fwd_mcol(x) (g2 = gf_mulx(x), g2 ^ upr((x) ^ g2, 3) ^ upr((x), 2) ^ upr((x), 1)) +#endif + +#if defined(IM4_SET) +#define inv_mcol(x) four_tables(x,t_use(i,m),vf1,rf1,0) +#elif defined(IM1_SET) +#define inv_mcol(x) one_table(x,upr,t_use(i,m),vf1,rf1,0) +#else +#define dec_imvars aes_32t g2, g4, g9 +#define inv_mcol(x) (g2 = gf_mulx(x), g4 = gf_mulx(g2), g9 = (x) ^ gf_mulx(g4), g4 ^= g9, \ + (x) ^ g2 ^ g4 ^ upr(g2 ^ g9, 3) ^ upr(g4, 2) ^ upr(g9, 1)) +#endif + +#if defined(FL4_SET) +#define ls_box(x,c) four_tables(x,t_use(f,l),vf1,rf2,c) +#elif defined(LS4_SET) +#define ls_box(x,c) four_tables(x,t_use(l,s),vf1,rf2,c) +#elif defined(FL1_SET) +#define ls_box(x,c) one_table(x,upr,t_use(f,l),vf1,rf2,c) +#elif defined(LS1_SET) +#define ls_box(x,c) one_table(x,upr,t_use(l,s),vf1,rf2,c) +#else +#define ls_box(x,c) no_table(x,t_use(s,box),vf1,rf2,c) +#endif + +#if defined(__cplusplus) +} +#endif + +#endif diff --git a/scripts/conf-w32brg/cipher/aestab.c b/scripts/conf-w32brg/cipher/aestab.c new file mode 100644 index 000000000..cf051f6c6 --- /dev/null +++ b/scripts/conf-w32brg/cipher/aestab.c @@ -0,0 +1,232 @@ +/* + --------------------------------------------------------------------------- + Copyright (c) 2003, Dr Brian Gladman <[email protected]>, Worcester, UK. + All rights reserved. + + LICENSE TERMS + + The free distribution and use of this software in both source and binary + form is allowed (with or without changes) provided that: + + 1. distributions of this source code include the above copyright + notice, this list of conditions and the following disclaimer; + + 2. distributions in binary form include the above copyright + notice, this list of conditions and the following disclaimer + in the documentation and/or other associated materials; + + 3. the copyright holder's name is not used to endorse products + built using this software without specific written permission. + + ALTERNATIVELY, provided that this notice is retained in full, this product + may be distributed under the terms of the GNU General Public License (GPL), + in which case the provisions of the GPL apply INSTEAD OF those given above. + + DISCLAIMER + + This software is provided 'as is' with no explicit or implied warranties + in respect of its properties, including, but not limited to, correctness + and/or fitness for purpose. + --------------------------------------------------------------------------- + Issue Date: 1/06/2003 + +*/ + +#if defined(__cplusplus) +extern "C" +{ +#endif + +#define DO_TABLES + +#include "aesopt.h" + +#if defined(FIXED_TABLES) + +/* implemented in case of wrong call for fixed tables */ + +void gen_tabs(void) +{ +} + +#else /* dynamic table generation */ + +#if !defined(FF_TABLES) + +/* Generate the tables for the dynamic table option + + It will generally be sensible to use tables to compute finite + field multiplies and inverses but where memory is scarse this + code might sometimes be better. But it only has effect during + initialisation so its pretty unimportant in overall terms. +*/ + +/* return 2 ^ (n - 1) where n is the bit number of the highest bit + set in x with x in the range 1 < x < 0x00000200. This form is + used so that locals within fi can be bytes rather than words +*/ + +static aes_08t hibit(const aes_32t x) +{ aes_08t r = (aes_08t)((x >> 1) | (x >> 2)); + + r |= (r >> 2); + r |= (r >> 4); + return (r + 1) >> 1; +} + +/* return the inverse of the finite field element x */ + +static aes_08t fi(const aes_08t x) +{ aes_08t p1 = x, p2 = BPOLY, n1 = hibit(x), n2 = 0x80, v1 = 1, v2 = 0; + + if(x < 2) return x; + + for(;;) + { + if(!n1) return v1; + + while(n2 >= n1) + { + n2 /= n1; p2 ^= p1 * n2; v2 ^= v1 * n2; n2 = hibit(p2); + } + + if(!n2) return v2; + + while(n1 >= n2) + { + n1 /= n2; p1 ^= p2 * n1; v1 ^= v2 * n1; n1 = hibit(p1); + } + } +} + +#endif + +/* The forward and inverse affine transformations used in the S-box */ + +#define fwd_affine(x) \ + (w = (aes_32t)x, w ^= (w<<1)^(w<<2)^(w<<3)^(w<<4), 0x63^(aes_08t)(w^(w>>8))) + +#define inv_affine(x) \ + (w = (aes_32t)x, w = (w<<1)^(w<<3)^(w<<6), 0x05^(aes_08t)(w^(w>>8))) + +static int init = 0; + +void gen_tabs(void) +{ aes_32t i, w; + +#if defined(FF_TABLES) + + aes_08t pow[512], log[256]; + + if(init) return; + /* log and power tables for GF(2^8) finite field with + WPOLY as modular polynomial - the simplest primitive + root is 0x03, used here to generate the tables + */ + + i = 0; w = 1; + do + { + pow[i] = (aes_08t)w; + pow[i + 255] = (aes_08t)w; + log[w] = (aes_08t)i++; + w ^= (w << 1) ^ (w & 0x80 ? WPOLY : 0); + } + while (w != 1); + +#else + if(init) return; +#endif + + for(i = 0, w = 1; i < RC_LENGTH; ++i) + { + t_set(r,c)[i] = bytes2word(w, 0, 0, 0); + w = f2(w); + } + + for(i = 0; i < 256; ++i) + { aes_08t b; + + b = fwd_affine(fi((aes_08t)i)); + w = bytes2word(f2(b), b, b, f3(b)); + +#ifdef SBX_SET + t_set(s,box)[i] = b; +#endif + +#ifdef FT1_SET /* tables for a normal encryption round */ + t_set(f,n)[i] = w; +#endif +#ifdef FT4_SET + t_set(f,n)[0][i] = w; + t_set(f,n)[1][i] = upr(w,1); + t_set(f,n)[2][i] = upr(w,2); + t_set(f,n)[3][i] = upr(w,3); +#endif + w = bytes2word(b, 0, 0, 0); + +#ifdef FL1_SET /* tables for last encryption round (may also */ + t_set(f,l)[i] = w; /* be used in the key schedule) */ +#endif +#ifdef FL4_SET + t_set(f,l)[0][i] = w; + t_set(f,l)[1][i] = upr(w,1); + t_set(f,l)[2][i] = upr(w,2); + t_set(f,l)[3][i] = upr(w,3); +#endif + +#ifdef LS1_SET /* table for key schedule if t_set(f,l) above is */ + t_set(l,s)[i] = w; /* not of the required form */ +#endif +#ifdef LS4_SET + t_set(l,s)[0][i] = w; + t_set(l,s)[1][i] = upr(w,1); + t_set(l,s)[2][i] = upr(w,2); + t_set(l,s)[3][i] = upr(w,3); +#endif + + b = fi(inv_affine((aes_08t)i)); + w = bytes2word(fe(b), f9(b), fd(b), fb(b)); + +#ifdef IM1_SET /* tables for the inverse mix column operation */ + t_set(i,m)[b] = w; +#endif +#ifdef IM4_SET + t_set(i,m)[0][b] = w; + t_set(i,m)[1][b] = upr(w,1); + t_set(i,m)[2][b] = upr(w,2); + t_set(i,m)[3][b] = upr(w,3); +#endif + +#ifdef ISB_SET + t_set(i,box)[i] = b; +#endif +#ifdef IT1_SET /* tables for a normal decryption round */ + t_set(i,n)[i] = w; +#endif +#ifdef IT4_SET + t_set(i,n)[0][i] = w; + t_set(i,n)[1][i] = upr(w,1); + t_set(i,n)[2][i] = upr(w,2); + t_set(i,n)[3][i] = upr(w,3); +#endif + w = bytes2word(b, 0, 0, 0); +#ifdef IL1_SET /* tables for last decryption round */ + t_set(i,l)[i] = w; +#endif +#ifdef IL4_SET + t_set(i,l)[0][i] = w; + t_set(i,l)[1][i] = upr(w,1); + t_set(i,l)[2][i] = upr(w,2); + t_set(i,l)[3][i] = upr(w,3); +#endif + } + init = 1; +} + +#endif + +#if defined(__cplusplus) +} +#endif + diff --git a/scripts/conf-w32brg/cipher/rijndael2.c b/scripts/conf-w32brg/cipher/rijndael2.c new file mode 100644 index 000000000..421f993b2 --- /dev/null +++ b/scripts/conf-w32brg/cipher/rijndael2.c @@ -0,0 +1,279 @@ +/* Rijndael (AES) for GnuPG + * Copyright (C) 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA + ******************************************************************* + A version of rijndael.c modified by Brian Gladman to use his AES code + */ + +#include <stdlib.h> +#include <string.h> /* for memcmp() */ + +#include "util.h" + +#include "aes.h" + +typedef struct +{ aes_encrypt_ctx ectx[1]; + aes_decrypt_ctx dctx[1]; + unsigned int klen; + unsigned int dkey; +} RIJNDAEL_context; + +static const char *selftest(void); +static int tested = 0; + +static void +burn_stack (int bytes) +{ + char buf[64]; + + wipememory(buf,sizeof buf); + bytes -= sizeof buf; + if (bytes > 0) + burn_stack (bytes); +} + +static int +rijndael_setkey (RIJNDAEL_context *ctx, const byte *key, const unsigned keylen) +{ int rc; + + if(!tested) + { const char *tr; + tested = 1; + tr = selftest(); + if(tr) + { + fprintf(stderr, "%s\n", tr ); + return G10ERR_SELFTEST_FAILED; + } + } + + ctx->klen = keylen; + ctx->dkey = 0; + rc = 0; + if(keylen == 16 || keylen == 24 || keylen == 32) + aes_encrypt_key(key, keylen, ctx->ectx); + else + rc = 1; + burn_stack ( 100 + 16*sizeof(int)); + return rc; +} + +static void +rijndael_encrypt (const RIJNDAEL_context *ctx, byte *b, const byte *a) +{ + aes_encrypt(a, b, ctx->ectx); + burn_stack (16 + 2*sizeof(int)); +} + +static void +rijndael_decrypt (RIJNDAEL_context *ctx, byte *b, const byte *a) +{ + if(!(ctx->dkey)) + { + aes_decrypt_key((byte*)ctx->ectx, ctx->klen, ctx->dctx); + ctx->dkey = 1; + } + aes_decrypt(a, b, ctx->dctx); + burn_stack (16+2*sizeof(int)); +} + +/* Test a single encryption and decryption with each key size. */ + +static const char* +selftest (void) +{ + RIJNDAEL_context ctx; + byte scratch[16]; + + /* The test vectors are from the AES supplied ones; more or less + * randomly taken from ecb_tbl.txt (I=42,81,14) + */ + static const byte plaintext[16] = { + 0x01,0x4B,0xAF,0x22,0x78,0xA6,0x9D,0x33, + 0x1D,0x51,0x80,0x10,0x36,0x43,0xE9,0x9A + }; + static const byte key[16] = { + 0xE8,0xE9,0xEA,0xEB,0xED,0xEE,0xEF,0xF0, + 0xF2,0xF3,0xF4,0xF5,0xF7,0xF8,0xF9,0xFA + }; + static const byte ciphertext[16] = { + 0x67,0x43,0xC3,0xD1,0x51,0x9A,0xB4,0xF2, + 0xCD,0x9A,0x78,0xAB,0x09,0xA5,0x11,0xBD + }; + + static const byte plaintext_192[16] = { + 0x76,0x77,0x74,0x75,0xF1,0xF2,0xF3,0xF4, + 0xF8,0xF9,0xE6,0xE7,0x77,0x70,0x71,0x72 + }; + static const byte key_192[24] = { + 0x04,0x05,0x06,0x07,0x09,0x0A,0x0B,0x0C, + 0x0E,0x0F,0x10,0x11,0x13,0x14,0x15,0x16, + 0x18,0x19,0x1A,0x1B,0x1D,0x1E,0x1F,0x20 + }; + static const byte ciphertext_192[16] = { + 0x5D,0x1E,0xF2,0x0D,0xCE,0xD6,0xBC,0xBC, + 0x12,0x13,0x1A,0xC7,0xC5,0x47,0x88,0xAA + }; + + static const byte plaintext_256[16] = { + 0x06,0x9A,0x00,0x7F,0xC7,0x6A,0x45,0x9F, + 0x98,0xBA,0xF9,0x17,0xFE,0xDF,0x95,0x21 + }; + static const byte key_256[32] = { + 0x08,0x09,0x0A,0x0B,0x0D,0x0E,0x0F,0x10, + 0x12,0x13,0x14,0x15,0x17,0x18,0x19,0x1A, + 0x1C,0x1D,0x1E,0x1F,0x21,0x22,0x23,0x24, + 0x26,0x27,0x28,0x29,0x2B,0x2C,0x2D,0x2E + }; + static const byte ciphertext_256[16] = { + 0x08,0x0E,0x95,0x17,0xEB,0x16,0x77,0x71, + 0x9A,0xCF,0x72,0x80,0x86,0x04,0x0A,0xE3 + }; + + rijndael_setkey (&ctx, key, sizeof(key)); + rijndael_encrypt (&ctx, scratch, plaintext); + if (memcmp (scratch, ciphertext, sizeof (ciphertext))) + return "Rijndael-128 test encryption failed."; + rijndael_decrypt (&ctx, scratch, scratch); + if (memcmp (scratch, plaintext, sizeof (plaintext))) + return "Rijndael-128 test decryption failed."; + + rijndael_setkey (&ctx, key_192, sizeof(key_192)); + rijndael_encrypt (&ctx, scratch, plaintext_192); + if (memcmp (scratch, ciphertext_192, sizeof (ciphertext_192))) + return "Rijndael-192 test encryption failed."; + rijndael_decrypt (&ctx, scratch, scratch); + if (memcmp (scratch, plaintext_192, sizeof (plaintext_192))) + return "Rijndael-192 test decryption failed."; + + rijndael_setkey (&ctx, key_256, sizeof(key_256)); + rijndael_encrypt (&ctx, scratch, plaintext_256); + if (memcmp (scratch, ciphertext_256, sizeof (ciphertext_256))) + return "Rijndael-256 test encryption failed."; + rijndael_decrypt (&ctx, scratch, scratch); + if (memcmp (scratch, plaintext_256, sizeof (plaintext_256))) + return "Rijndael-256 test decryption failed."; + + return NULL; +} + +#ifdef IS_MODULE +static +#endif + const char * + rijndael_get_info (int algo, size_t *keylen, + size_t *blocksize, size_t *contextsize, + int (**r_setkey) (void *c, byte *key, unsigned keylen), + void (**r_encrypt) (void *c, byte *outbuf, byte *inbuf), + void (**r_decrypt) (void *c, byte *outbuf, byte *inbuf) + ) +{ + *keylen = algo==7? 128 : algo==8? 192 : 256; + *blocksize = 16; + *contextsize = sizeof (RIJNDAEL_context); + + *(int (**)(RIJNDAEL_context*, const byte*, const unsigned))r_setkey + = rijndael_setkey; + *(void (**)(const RIJNDAEL_context*, byte*, const byte*))r_encrypt + = rijndael_encrypt; + *(void (**)(RIJNDAEL_context*, byte*, const byte*))r_decrypt + = rijndael_decrypt; + + if( algo == 7 ) + return "AES"; + if (algo == 8) + return "AES192"; + if (algo == 9) + return "AES256"; + return NULL; +} + + +#ifdef IS_MODULE +static +const char * const gnupgext_version = "RIJNDAEL ($Revision$)"; + +static struct { + int class; + int version; + int value; + void (*func)(void); +} func_table[] = { + { 20, 1, 0, (void*)rijndael_get_info }, + { 21, 1, 7 }, + { 21, 1, 8 }, + { 21, 1, 9 }, +}; + + + +/**************** + * Enumerate the names of the functions together with information about + * this function. Set sequence to an integer with a initial value of 0 and + * do not change it. + * If what is 0 all kind of functions are returned. + * Return values: class := class of function: + * 10 = message digest algorithm info function + * 11 = integer with available md algorithms + * 20 = cipher algorithm info function + * 21 = integer with available cipher algorithms + * 30 = public key algorithm info function + * 31 = integer with available pubkey algorithms + * version = interface version of the function/pointer + * (currently this is 1 for all functions) + */ +static +void * +gnupgext_enum_func ( int what, int *sequence, int *class, int *vers ) +{ + void *ret; + int i = *sequence; + + do { + if ( i >= DIM(func_table) || i < 0 ) { + return NULL; + } + *class = func_table[i].class; + *vers = func_table[i].version; + switch( *class ) { + case 11: + case 21: + case 31: + ret = &func_table[i].value; + break; + default: + ret = func_table[i].func; + break; + } + i++; + } while ( what && what != *class ); + + *sequence = i; + return ret; +} +#endif + + + + + + + + + |