blob: e951b9a055f602e8c8871cca74f4e9ac2c6693c5 [file] [log] [blame]
/*
* Copyright © 2012 Siarhei Siamashka <siarhei.siamashka@gmail.com>
*
* Based on the public domain implementation of small noncryptographic PRNG
* authored by Bob Jenkins: http://burtleburtle.net/bob/rand/smallprng.html
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
* DEALINGS IN THE SOFTWARE.
*/
#include "utils.h"
#include "utils-prng.h"
#if defined(HAVE_GCC_VECTOR_EXTENSIONS) && defined(__SSE2__)
#include <xmmintrin.h>
#endif
void smallprng_srand_r (smallprng_t *x, uint32_t seed)
{
uint32_t i;
x->a = 0xf1ea5eed, x->b = x->c = x->d = seed;
for (i = 0; i < 20; ++i)
smallprng_rand_r (x);
}
/*
* Set a 32-bit seed for PRNG
*
* LCG is used here for generating independent seeds for different
* smallprng instances (in the case if smallprng is also used for
* generating these seeds, "Big Crush" test from TestU01 detects
* some problems in the glued 'prng_rand_128_r' output data).
* Actually we might be even better using some cryptographic
* hash for this purpose, but LCG seems to be also enough for
* passing "Big Crush".
*/
void prng_srand_r (prng_t *x, uint32_t seed)
{
#ifdef HAVE_GCC_VECTOR_EXTENSIONS
int i;
prng_rand_128_data_t dummy;
smallprng_srand_r (&x->p0, seed);
x->a[0] = x->a[1] = x->a[2] = x->a[3] = 0xf1ea5eed;
x->b[0] = x->c[0] = x->d[0] = (seed = seed * 1103515245 + 12345);
x->b[1] = x->c[1] = x->d[1] = (seed = seed * 1103515245 + 12345);
x->b[2] = x->c[2] = x->d[2] = (seed = seed * 1103515245 + 12345);
x->b[3] = x->c[3] = x->d[3] = (seed = seed * 1103515245 + 12345);
for (i = 0; i < 20; ++i)
prng_rand_128_r (x, &dummy);
#else
smallprng_srand_r (&x->p0, seed);
smallprng_srand_r (&x->p1, (seed = seed * 1103515245 + 12345));
smallprng_srand_r (&x->p2, (seed = seed * 1103515245 + 12345));
smallprng_srand_r (&x->p3, (seed = seed * 1103515245 + 12345));
smallprng_srand_r (&x->p4, (seed = seed * 1103515245 + 12345));
#endif
}
static force_inline void
store_rand_128_data (void *addr, prng_rand_128_data_t *d, int aligned)
{
#ifdef HAVE_GCC_VECTOR_EXTENSIONS
if (aligned)
{
*(uint8x16 *)addr = d->vb;
return;
}
else
{
#ifdef __SSE2__
/* workaround for http://gcc.gnu.org/PR55614 */
_mm_storeu_si128 (addr, _mm_loadu_si128 ((__m128i *)d));
return;
#endif
}
#endif
/* we could try something better for unaligned writes (packed attribute),
* but GCC is not very reliable: http://gcc.gnu.org/PR55454 */
memcpy (addr, d, 16);
}
/*
* Helper function and the actual code for "prng_randmemset_r" function
*/
static force_inline void
randmemset_internal (prng_t *prng,
uint8_t *buf,
size_t size,
prng_randmemset_flags_t flags,
int aligned)
{
prng_t local_prng = *prng;
prng_rand_128_data_t randdata;
size_t i;
while (size >= 16)
{
prng_rand_128_data_t t;
if (flags == 0)
{
prng_rand_128_r (&local_prng, &randdata);
}
else
{
prng_rand_128_r (&local_prng, &t);
prng_rand_128_r (&local_prng, &randdata);
#ifdef HAVE_GCC_VECTOR_EXTENSIONS
if (flags & RANDMEMSET_MORE_FF)
{
const uint8x16 const_C0 =
{
0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0,
0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0, 0xC0
};
randdata.vb |= (t.vb >= const_C0);
}
if (flags & RANDMEMSET_MORE_00)
{
const uint8x16 const_40 =
{
0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,
0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40
};
randdata.vb &= (t.vb >= const_40);
}
if (flags & RANDMEMSET_MORE_FFFFFFFF)
{
const uint32x4 const_C0000000 =
{
0xC0000000, 0xC0000000, 0xC0000000, 0xC0000000
};
randdata.vw |= ((t.vw << 30) >= const_C0000000);
}
if (flags & RANDMEMSET_MORE_00000000)
{
const uint32x4 const_40000000 =
{
0x40000000, 0x40000000, 0x40000000, 0x40000000
};
randdata.vw &= ((t.vw << 30) >= const_40000000);
}
#else
#define PROCESS_ONE_LANE(i) \
if (flags & RANDMEMSET_MORE_FF) \
{ \
uint32_t mask_ff = (t.w[i] & (t.w[i] << 1)) & 0x80808080; \
mask_ff |= mask_ff >> 1; \
mask_ff |= mask_ff >> 2; \
mask_ff |= mask_ff >> 4; \
randdata.w[i] |= mask_ff; \
} \
if (flags & RANDMEMSET_MORE_00) \
{ \
uint32_t mask_00 = (t.w[i] | (t.w[i] << 1)) & 0x80808080; \
mask_00 |= mask_00 >> 1; \
mask_00 |= mask_00 >> 2; \
mask_00 |= mask_00 >> 4; \
randdata.w[i] &= mask_00; \
} \
if (flags & RANDMEMSET_MORE_FFFFFFFF) \
{ \
int32_t mask_ff = ((t.w[i] << 30) & (t.w[i] << 31)) & \
0x80000000; \
randdata.w[i] |= mask_ff >> 31; \
} \
if (flags & RANDMEMSET_MORE_00000000) \
{ \
int32_t mask_00 = ((t.w[i] << 30) | (t.w[i] << 31)) & \
0x80000000; \
randdata.w[i] &= mask_00 >> 31; \
}
PROCESS_ONE_LANE (0)
PROCESS_ONE_LANE (1)
PROCESS_ONE_LANE (2)
PROCESS_ONE_LANE (3)
#endif
}
if (is_little_endian ())
{
store_rand_128_data (buf, &randdata, aligned);
buf += 16;
}
else
{
#ifdef HAVE_GCC_VECTOR_EXTENSIONS
#define BSWAP_SHUFFLEMASK 3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12
#if defined(__clang__)
randdata.vb = __builtin_shufflevector (randdata.vb, randdata.vb, BSWAP_SHUFFLEMASK);
#else
const uint8x16 bswap_shufflemask =
{
BSWAP_SHUFFLEMASK
};
randdata.vb = __builtin_shuffle (randdata.vb, bswap_shufflemask);
#endif
store_rand_128_data (buf, &randdata, aligned);
buf += 16;
#else
uint8_t t1, t2, t3, t4;
#define STORE_ONE_LANE(i) \
t1 = randdata.b[i * 4 + 3]; \
t2 = randdata.b[i * 4 + 2]; \
t3 = randdata.b[i * 4 + 1]; \
t4 = randdata.b[i * 4 + 0]; \
*buf++ = t1; \
*buf++ = t2; \
*buf++ = t3; \
*buf++ = t4;
STORE_ONE_LANE (0)
STORE_ONE_LANE (1)
STORE_ONE_LANE (2)
STORE_ONE_LANE (3)
#endif
}
size -= 16;
}
i = 0;
while (i < size)
{
uint8_t randbyte = prng_rand_r (&local_prng) & 0xFF;
if (flags != 0)
{
uint8_t t = prng_rand_r (&local_prng) & 0xFF;
if ((flags & RANDMEMSET_MORE_FF) && (t >= 0xC0))
randbyte = 0xFF;
if ((flags & RANDMEMSET_MORE_00) && (t < 0x40))
randbyte = 0x00;
if (i % 4 == 0 && i + 4 <= size)
{
t = prng_rand_r (&local_prng) & 0xFF;
if ((flags & RANDMEMSET_MORE_FFFFFFFF) && (t >= 0xC0))
{
memset(&buf[i], 0xFF, 4);
i += 4;
continue;
}
if ((flags & RANDMEMSET_MORE_00000000) && (t < 0x40))
{
memset(&buf[i], 0x00, 4);
i += 4;
continue;
}
}
}
buf[i] = randbyte;
i++;
}
*prng = local_prng;
}
/*
* Fill memory buffer with random data. Flags argument may be used
* to tweak some statistics properties:
* RANDMEMSET_MORE_00 - set ~25% of bytes to 0x00
* RANDMEMSET_MORE_FF - set ~25% of bytes to 0xFF
* RANDMEMSET_MORE_00000000 - ~25% chance for 00000000 4-byte clusters
* RANDMEMSET_MORE_FFFFFFFF - ~25% chance for FFFFFFFF 4-byte clusters
*/
void prng_randmemset_r (prng_t *prng,
void *voidbuf,
size_t size,
prng_randmemset_flags_t flags)
{
uint8_t *buf = (uint8_t *)voidbuf;
if ((uintptr_t)buf & 15)
{
/* unaligned buffer */
if (flags == 0)
randmemset_internal (prng, buf, size, 0, 0);
else if (flags == RANDMEMSET_MORE_00_AND_FF)
randmemset_internal (prng, buf, size, RANDMEMSET_MORE_00_AND_FF, 0);
else
randmemset_internal (prng, buf, size, flags, 0);
}
else
{
/* aligned buffer */
if (flags == 0)
randmemset_internal (prng, buf, size, 0, 1);
else if (flags == RANDMEMSET_MORE_00_AND_FF)
randmemset_internal (prng, buf, size, RANDMEMSET_MORE_00_AND_FF, 1);
else
randmemset_internal (prng, buf, size, flags, 1);
}
}