mirror of
https://github.com/yuzu-emu/mbedtls.git
synced 2024-11-24 01:45:41 +01:00
1e14827beb
As a result, the copyright of contributors other than Arm is now acknowledged, and the years of publishing are no longer tracked in the source files. Also remove the now-redundant lines declaring that the files are part of MbedTLS. This commit was generated using the following script: # ======================== #!/bin/sh # Find files find '(' -path './.git' -o -path './3rdparty' ')' -prune -o -type f -print | xargs sed -bi ' # Replace copyright attribution line s/Copyright.*Arm.*/Copyright The Mbed TLS Contributors/I # Remove redundant declaration and the preceding line $!N /This file is part of Mbed TLS/Id P D ' # ======================== Signed-off-by: Bence Szépkúti <bence.szepkuti@arm.com>
238 lines
8.7 KiB
C
238 lines
8.7 KiB
C
/**
|
|
* \brief HAVEGE: HArdware Volatile Entropy Gathering and Expansion
|
|
*
|
|
* Copyright The Mbed TLS Contributors
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
*
|
|
* Licensed under the Apache License, Version 2.0 (the "License"); you may
|
|
* not use this file except in compliance with the License.
|
|
* You may obtain a copy of the License at
|
|
*
|
|
* http://www.apache.org/licenses/LICENSE-2.0
|
|
*
|
|
* Unless required by applicable law or agreed to in writing, software
|
|
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
|
|
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
* See the License for the specific language governing permissions and
|
|
* limitations under the License.
|
|
*/
|
|
/*
|
|
* The HAVEGE RNG was designed by Andre Seznec in 2002.
|
|
*
|
|
* http://www.irisa.fr/caps/projects/hipsor/publi.php
|
|
*
|
|
* Contact: seznec(at)irisa_dot_fr - orocheco(at)irisa_dot_fr
|
|
*/
|
|
|
|
#include "common.h"
|
|
|
|
#if defined(MBEDTLS_HAVEGE_C)
|
|
|
|
#include "mbedtls/havege.h"
|
|
#include "mbedtls/timing.h"
|
|
#include "mbedtls/platform_util.h"
|
|
|
|
#include <stdint.h>
|
|
#include <string.h>
|
|
|
|
/* ------------------------------------------------------------------------
|
|
* On average, one iteration accesses two 8-word blocks in the havege WALK
|
|
* table, and generates 16 words in the RES array.
|
|
*
|
|
* The data read in the WALK table is updated and permuted after each use.
|
|
* The result of the hardware clock counter read is used for this update.
|
|
*
|
|
* 25 conditional tests are present. The conditional tests are grouped in
|
|
* two nested groups of 12 conditional tests and 1 test that controls the
|
|
* permutation; on average, there should be 6 tests executed and 3 of them
|
|
* should be mispredicted.
|
|
* ------------------------------------------------------------------------
|
|
*/
|
|
|
|
#define SWAP(X,Y) { uint32_t *T = (X); (X) = (Y); (Y) = T; }
|
|
|
|
#define TST1_ENTER if( PTEST & 1 ) { PTEST ^= 3; PTEST >>= 1;
|
|
#define TST2_ENTER if( PTEST & 1 ) { PTEST ^= 3; PTEST >>= 1;
|
|
|
|
#define TST1_LEAVE U1++; }
|
|
#define TST2_LEAVE U2++; }
|
|
|
|
#define ONE_ITERATION \
|
|
\
|
|
PTEST = PT1 >> 20; \
|
|
\
|
|
TST1_ENTER TST1_ENTER TST1_ENTER TST1_ENTER \
|
|
TST1_ENTER TST1_ENTER TST1_ENTER TST1_ENTER \
|
|
TST1_ENTER TST1_ENTER TST1_ENTER TST1_ENTER \
|
|
\
|
|
TST1_LEAVE TST1_LEAVE TST1_LEAVE TST1_LEAVE \
|
|
TST1_LEAVE TST1_LEAVE TST1_LEAVE TST1_LEAVE \
|
|
TST1_LEAVE TST1_LEAVE TST1_LEAVE TST1_LEAVE \
|
|
\
|
|
PTX = (PT1 >> 18) & 7; \
|
|
PT1 &= 0x1FFF; \
|
|
PT2 &= 0x1FFF; \
|
|
CLK = (uint32_t) mbedtls_timing_hardclock(); \
|
|
\
|
|
i = 0; \
|
|
A = &WALK[PT1 ]; RES[i++] ^= *A; \
|
|
B = &WALK[PT2 ]; RES[i++] ^= *B; \
|
|
C = &WALK[PT1 ^ 1]; RES[i++] ^= *C; \
|
|
D = &WALK[PT2 ^ 4]; RES[i++] ^= *D; \
|
|
\
|
|
IN = (*A >> (1)) ^ (*A << (31)) ^ CLK; \
|
|
*A = (*B >> (2)) ^ (*B << (30)) ^ CLK; \
|
|
*B = IN ^ U1; \
|
|
*C = (*C >> (3)) ^ (*C << (29)) ^ CLK; \
|
|
*D = (*D >> (4)) ^ (*D << (28)) ^ CLK; \
|
|
\
|
|
A = &WALK[PT1 ^ 2]; RES[i++] ^= *A; \
|
|
B = &WALK[PT2 ^ 2]; RES[i++] ^= *B; \
|
|
C = &WALK[PT1 ^ 3]; RES[i++] ^= *C; \
|
|
D = &WALK[PT2 ^ 6]; RES[i++] ^= *D; \
|
|
\
|
|
if( PTEST & 1 ) SWAP( A, C ); \
|
|
\
|
|
IN = (*A >> (5)) ^ (*A << (27)) ^ CLK; \
|
|
*A = (*B >> (6)) ^ (*B << (26)) ^ CLK; \
|
|
*B = IN; CLK = (uint32_t) mbedtls_timing_hardclock(); \
|
|
*C = (*C >> (7)) ^ (*C << (25)) ^ CLK; \
|
|
*D = (*D >> (8)) ^ (*D << (24)) ^ CLK; \
|
|
\
|
|
A = &WALK[PT1 ^ 4]; \
|
|
B = &WALK[PT2 ^ 1]; \
|
|
\
|
|
PTEST = PT2 >> 1; \
|
|
\
|
|
PT2 = (RES[(i - 8) ^ PTY] ^ WALK[PT2 ^ PTY ^ 7]); \
|
|
PT2 = ((PT2 & 0x1FFF) & (~8)) ^ ((PT1 ^ 8) & 0x8); \
|
|
PTY = (PT2 >> 10) & 7; \
|
|
\
|
|
TST2_ENTER TST2_ENTER TST2_ENTER TST2_ENTER \
|
|
TST2_ENTER TST2_ENTER TST2_ENTER TST2_ENTER \
|
|
TST2_ENTER TST2_ENTER TST2_ENTER TST2_ENTER \
|
|
\
|
|
TST2_LEAVE TST2_LEAVE TST2_LEAVE TST2_LEAVE \
|
|
TST2_LEAVE TST2_LEAVE TST2_LEAVE TST2_LEAVE \
|
|
TST2_LEAVE TST2_LEAVE TST2_LEAVE TST2_LEAVE \
|
|
\
|
|
C = &WALK[PT1 ^ 5]; \
|
|
D = &WALK[PT2 ^ 5]; \
|
|
\
|
|
RES[i++] ^= *A; \
|
|
RES[i++] ^= *B; \
|
|
RES[i++] ^= *C; \
|
|
RES[i++] ^= *D; \
|
|
\
|
|
IN = (*A >> ( 9)) ^ (*A << (23)) ^ CLK; \
|
|
*A = (*B >> (10)) ^ (*B << (22)) ^ CLK; \
|
|
*B = IN ^ U2; \
|
|
*C = (*C >> (11)) ^ (*C << (21)) ^ CLK; \
|
|
*D = (*D >> (12)) ^ (*D << (20)) ^ CLK; \
|
|
\
|
|
A = &WALK[PT1 ^ 6]; RES[i++] ^= *A; \
|
|
B = &WALK[PT2 ^ 3]; RES[i++] ^= *B; \
|
|
C = &WALK[PT1 ^ 7]; RES[i++] ^= *C; \
|
|
D = &WALK[PT2 ^ 7]; RES[i++] ^= *D; \
|
|
\
|
|
IN = (*A >> (13)) ^ (*A << (19)) ^ CLK; \
|
|
*A = (*B >> (14)) ^ (*B << (18)) ^ CLK; \
|
|
*B = IN; \
|
|
*C = (*C >> (15)) ^ (*C << (17)) ^ CLK; \
|
|
*D = (*D >> (16)) ^ (*D << (16)) ^ CLK; \
|
|
\
|
|
PT1 = ( RES[( i - 8 ) ^ PTX] ^ \
|
|
WALK[PT1 ^ PTX ^ 7] ) & (~1); \
|
|
PT1 ^= (PT2 ^ 0x10) & 0x10; \
|
|
\
|
|
for( n++, i = 0; i < 16; i++ ) \
|
|
hs->pool[n % MBEDTLS_HAVEGE_COLLECT_SIZE] ^= RES[i];
|
|
|
|
/*
|
|
* Entropy gathering function
|
|
*/
|
|
static void havege_fill( mbedtls_havege_state *hs )
|
|
{
|
|
size_t n = 0;
|
|
size_t i;
|
|
uint32_t U1, U2, *A, *B, *C, *D;
|
|
uint32_t PT1, PT2, *WALK, RES[16];
|
|
uint32_t PTX, PTY, CLK, PTEST, IN;
|
|
|
|
WALK = hs->WALK;
|
|
PT1 = hs->PT1;
|
|
PT2 = hs->PT2;
|
|
|
|
PTX = U1 = 0;
|
|
PTY = U2 = 0;
|
|
|
|
(void)PTX;
|
|
|
|
memset( RES, 0, sizeof( RES ) );
|
|
|
|
while( n < MBEDTLS_HAVEGE_COLLECT_SIZE * 4 )
|
|
{
|
|
ONE_ITERATION
|
|
ONE_ITERATION
|
|
ONE_ITERATION
|
|
ONE_ITERATION
|
|
}
|
|
|
|
hs->PT1 = PT1;
|
|
hs->PT2 = PT2;
|
|
|
|
hs->offset[0] = 0;
|
|
hs->offset[1] = MBEDTLS_HAVEGE_COLLECT_SIZE / 2;
|
|
}
|
|
|
|
/*
|
|
* HAVEGE initialization
|
|
*/
|
|
void mbedtls_havege_init( mbedtls_havege_state *hs )
|
|
{
|
|
memset( hs, 0, sizeof( mbedtls_havege_state ) );
|
|
|
|
havege_fill( hs );
|
|
}
|
|
|
|
void mbedtls_havege_free( mbedtls_havege_state *hs )
|
|
{
|
|
if( hs == NULL )
|
|
return;
|
|
|
|
mbedtls_platform_zeroize( hs, sizeof( mbedtls_havege_state ) );
|
|
}
|
|
|
|
/*
|
|
* HAVEGE rand function
|
|
*/
|
|
int mbedtls_havege_random( void *p_rng, unsigned char *buf, size_t len )
|
|
{
|
|
uint32_t val;
|
|
size_t use_len;
|
|
mbedtls_havege_state *hs = (mbedtls_havege_state *) p_rng;
|
|
unsigned char *p = buf;
|
|
|
|
while( len > 0 )
|
|
{
|
|
use_len = len;
|
|
if( use_len > sizeof( val ) )
|
|
use_len = sizeof( val );
|
|
|
|
if( hs->offset[1] >= MBEDTLS_HAVEGE_COLLECT_SIZE )
|
|
havege_fill( hs );
|
|
|
|
val = hs->pool[hs->offset[0]++];
|
|
val ^= hs->pool[hs->offset[1]++];
|
|
|
|
memcpy( p, &val, use_len );
|
|
|
|
len -= use_len;
|
|
p += use_len;
|
|
}
|
|
|
|
return( 0 );
|
|
}
|
|
|
|
#endif /* MBEDTLS_HAVEGE_C */
|