kris/pqcrypto
1
1
Fork 0
mirror of https://github.com/henrydcase/pqc.git synced 2024-11-23 16:08:59 +00:00
Code Issues 1 Releases Wiki Activity
bff1ba5c74
pqcrypto/crypto_sign/dilithium4/avx2/poly.c
Thom Wiggers 00b17e0740
Use PQClean int guidelines
2019-12-09 12:05:25 +01:00

921 lines
36 KiB
C
Raw Blame History

#include <immintrin.h>
#include <stdint.h>
#include "fips202x4.h"
#include "ntt.h"
#include "nttconsts.h"
#include "params.h"
#include "poly.h"
#include "reduce.h"
#include "rejsample.h"
#include "rounding.h"
#include "symmetric.h"
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_reduce
*
* Description: Reduce all coefficients of input polynomial to representative
* in [0,2*Q[.
*
* Arguments: - poly *a: pointer to input/output polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_reduce(poly *a) {
PQCLEAN_DILITHIUM4_AVX2_reduce_avx(a->coeffs);
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_csubq
*
* Description: For all coefficients of input polynomial subtract Q if
* coefficient is bigger than Q.
*
* Arguments: - poly *a: pointer to input/output polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_csubq(poly *a) {
PQCLEAN_DILITHIUM4_AVX2_csubq_avx(a->coeffs);
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_freeze
*
* Description: Reduce all coefficients of the polynomial to standard
* representatives.
*
* Arguments: - poly *a: pointer to input/output polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_freeze(poly *a) {
PQCLEAN_DILITHIUM4_AVX2_reduce_avx(a->coeffs);
PQCLEAN_DILITHIUM4_AVX2_csubq_avx(a->coeffs);
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_add
*
* Description: Add polynomials. No modular reduction is performed.
*
* Arguments: - poly *c: pointer to output polynomial
* - const poly *a: pointer to first summand
* - const poly *b: pointer to second summand
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_add(poly *c, const poly *a, const poly *b) {
__m256i vec0, vec1;
for (size_t i = 0; i < N / 8; i++) {
vec0 = _mm256_load_si256(&a->coeffs_x8[i]);
vec1 = _mm256_load_si256(&b->coeffs_x8[i]);
vec0 = _mm256_add_epi32(vec0, vec1);
_mm256_store_si256(&c->coeffs_x8[i], vec0);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_sub
*
* Description: Subtract polynomials. Assumes coefficients of second input
* polynomial to be less than 2*Q. No modular reduction is
* performed.
*
* Arguments: - poly *c: pointer to output polynomial
* - const poly *a: pointer to first input polynomial
* - const poly *b: pointer to second input polynomial to be
* subtraced from first input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_sub(poly *c, const poly *a, const poly *b) {
__m256i vec0, vec1;
const __m256i twoq = _mm256_load_si256(_PQCLEAN_DILITHIUM4_AVX2_8x2q.as_vec);
for (size_t i = 0; i < N / 8; i++) {
vec0 = _mm256_load_si256(&a->coeffs_x8[i]);
vec1 = _mm256_load_si256(&b->coeffs_x8[i]);
vec0 = _mm256_add_epi32(vec0, twoq);
vec0 = _mm256_sub_epi32(vec0, vec1);
_mm256_store_si256(&c->coeffs_x8[i], vec0);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_shiftl
*
* Description: Multiply polynomial by 2^D without modular reduction. Assumes
* input coefficients to be less than 2^{32-D}.
*
* Arguments: - poly *a: pointer to input/output polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_shiftl(poly *a) {
__m256i vec;
for (size_t i = 0; i < N / 8; i++) {
vec = _mm256_load_si256(&a->coeffs_x8[i]);
vec = _mm256_slli_epi32(vec, D);
_mm256_store_si256(&a->coeffs_x8[i], vec);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_ntt
*
* Description: Forward NTT. Output coefficients can be up to 16*Q larger than
* input coefficients.
*
* Arguments: - poly *a: pointer to input/output polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_ntt(poly *a) {
ALIGNED_UINT64(N) tmp;
for (size_t i = 0; i < N / 32; ++i) {
PQCLEAN_DILITHIUM4_AVX2_ntt_levels0t2_avx(tmp.as_arr + 4 * i, a->coeffs + 4 * i, PQCLEAN_DILITHIUM4_AVX2_zetas.as_arr + 1);
}
for (size_t i = 0; i < N / 32; ++i) {
PQCLEAN_DILITHIUM4_AVX2_ntt_levels3t8_avx(a->coeffs + 32 * i, tmp.as_arr + 32 * i, PQCLEAN_DILITHIUM4_AVX2_zetas.as_arr + 8 + 31 * i);
}
}
/*************************************************
* Name: poly_invntt_montgomery
*
* Description: Inverse NTT and multiplication with 2^{32}. Input coefficients
* need to be less than 2*Q. Output coefficients are less than 2*Q.
*
* Arguments: - poly *a: pointer to input/output polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_invntt_montgomery(poly *a) {
ALIGNED_UINT64(N) tmp;
for (size_t i = 0; i < N / 32; i++) {
PQCLEAN_DILITHIUM4_AVX2_invntt_levels0t4_avx(tmp.as_arr + 32 * i, a->coeffs + 32 * i, PQCLEAN_DILITHIUM4_AVX2_zetas_inv.as_arr + 31 * i);
}
for (size_t i = 0; i < N / 32; i++) {
PQCLEAN_DILITHIUM4_AVX2_invntt_levels5t7_avx(a->coeffs + 4 * i, tmp.as_arr + 4 * i, PQCLEAN_DILITHIUM4_AVX2_zetas_inv.as_arr + 248);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_pointwise_invmontgomery
*
* Description: Pointwise multiplication of polynomials in NTT domain
* representation and multiplication of resulting polynomial
* with 2^{-32}. Output coefficients are less than 2*Q if input
* coefficient are less than 22*Q.
*
* Arguments: - poly *c: pointer to output polynomial
* - const poly *a: pointer to first input polynomial
* - const poly *b: pointer to second input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_pointwise_invmontgomery(poly *c, const poly *a, const poly *b) {
PQCLEAN_DILITHIUM4_AVX2_pointwise_avx(c->coeffs, a->coeffs, b->coeffs);
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_power2round
*
* Description: For all coefficients c of the input polynomial,
* compute c0, c1 such that c mod Q = c1*2^D + c0
* with -2^{D-1} < c0 <= 2^{D-1}. Assumes coefficients to be
* standard representatives.
*
* Arguments: - poly *a1: pointer to output polynomial with coefficients c1
* - poly *a0: pointer to output polynomial with coefficients Q + a0
* - const poly *v: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_power2round(poly *restrict a1,
poly *restrict a0,
const poly *restrict a) {
for (size_t i = 0; i < N; ++i) {
a1->coeffs[i] = PQCLEAN_DILITHIUM4_AVX2_power2round(a->coeffs[i], &a0->coeffs[i]);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_decompose
*
* Description: For all coefficients c of the input polynomial,
* compute high and low bits c0, c1 such c mod Q = c1*ALPHA + c0
* with -ALPHA/2 < c0 <= ALPHA/2 except c1 = (Q-1)/ALPHA where we
* set c1 = 0 and -ALPHA/2 <= c0 = c mod Q - Q < 0.
* Assumes coefficients to be standard representatives.
*
* Arguments: - poly *a1: pointer to output polynomial with coefficients c1
* - poly *a0: pointer to output polynomial with coefficients Q + a0
* - const poly *c: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_decompose(
poly *restrict a1,
poly *restrict a0,
const poly *restrict a) {
for (size_t i = 0; i < N; ++i) {
a1->coeffs[i] = PQCLEAN_DILITHIUM4_AVX2_decompose(a->coeffs[i], &a0->coeffs[i]);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_make_hint
*
* Description: Compute hint polynomial. The coefficients of which indicate
* whether the low bits of the corresponding coefficient of
* the input polynomial overflow into the high bits.
*
* Arguments: - poly *h: pointer to output hint polynomial
* - const poly *a0: pointer to low part of input polynomial
* - const poly *a1: pointer to high part of input polynomial
*
* Returns number of 1 bits.
**************************************************/
uint32_t PQCLEAN_DILITHIUM4_AVX2_poly_make_hint(
poly *restrict h,
const poly *restrict a0,
const poly *restrict a1) {
uint32_t s = 0;
for (size_t i = 0; i < N; ++i) {
h->coeffs[i] = PQCLEAN_DILITHIUM4_AVX2_make_hint(a0->coeffs[i], a1->coeffs[i]);
s += h->coeffs[i];
}
return s;
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_use_hint
*
* Description: Use hint polynomial to correct the high bits of a polynomial.
*
* Arguments: - poly *a: pointer to output polynomial with corrected high bits
* - const poly *b: pointer to input polynomial
* - const poly *h: pointer to input hint polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_poly_use_hint(
poly *restrict a,
const poly *restrict b,
const poly *restrict h) {
for (size_t i = 0; i < N; ++i) {
a->coeffs[i] = PQCLEAN_DILITHIUM4_AVX2_use_hint(b->coeffs[i], h->coeffs[i]);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_chknorm
*
* Description: Check infinity norm of polynomial against given bound.
* Assumes input coefficients to be standard representatives.
*
* Arguments: - const poly *a: pointer to polynomial
* - uint32_t B: norm bound
*
* Returns 0 if norm is strictly smaller than B and 1 otherwise.
**************************************************/
int PQCLEAN_DILITHIUM4_AVX2_poly_chknorm(const poly *a, uint32_t B) {
int32_t t;
/* It is ok to leak which coefficient violates the bound since
the probability for each coefficient is independent of secret
data but we must not leak the sign of the centralized representative. */
for (size_t i = 0; i < N; ++i) {
/* Absolute value of centralized representative */
t = (Q - 1) / 2 - a->coeffs[i];
t ^= (t >> 31);
t = (Q - 1) / 2 - t;
if ((uint32_t)t >= B) {
return 1;
}
}
return 0;
}
/*************************************************
* Name: rej_uniform_ref
*
* Description: Sample uniformly random coefficients in [0, Q-1] by
* performing rejection sampling using array of random bytes.
*
* Arguments: - uint32_t *a: pointer to output array (allocated)
* - size_t len: number of coefficients to be sampled
* - const uint8_t *buf: array of random bytes
* - size_t buflen: length of array of random bytes
*
* Returns number of sampled coefficients. Can be smaller than len if not enough
* random bytes were given.
**************************************************/
static size_t rej_uniform_ref(
uint32_t *a,
size_t len,
const uint8_t *buf,
size_t buflen) {
size_t ctr, pos;
uint32_t t;
ctr = pos = 0;
while (ctr < len && pos + 3 <= buflen) {
t = buf[pos++];
t |= (uint32_t)buf[pos++] << 8;
t |= (uint32_t)buf[pos++] << 16;
t &= 0x7FFFFF;
if (t < Q) {
a[ctr++] = t;
}
}
return ctr;
}
/*************************************************
* Name: poly_uniform
*
* Description: Sample polynomial with uniformly random coefficients
* in [0,Q-1] by performing rejection sampling using the
* output stream from SHAKE256(seed|nonce).
*
* Arguments: - poly *a: pointer to output polynomial
* - const uint8_t seed[]: byte array with seed of length
* SEEDBYTES
* - uint16_t nonce: 2-byte nonce
**************************************************/
#define POLY_UNIFORM_NBLOCKS ((769 + STREAM128_BLOCKBYTES) / STREAM128_BLOCKBYTES)
#define POLY_UNIFORM_BUFLEN (POLY_UNIFORM_NBLOCKS * STREAM128_BLOCKBYTES)
void PQCLEAN_DILITHIUM4_AVX2_poly_uniform(poly *a,
const uint8_t seed[SEEDBYTES],
uint16_t nonce) {
size_t ctr, off;
size_t nblocks = POLY_UNIFORM_NBLOCKS;
size_t buflen = POLY_UNIFORM_BUFLEN;
uint8_t buf[POLY_UNIFORM_BUFLEN + 2];
stream128_state state;
stream128_init(&state, seed, nonce);
stream128_squeezeblocks(buf, nblocks, &state);
ctr = PQCLEAN_DILITHIUM4_AVX2_rej_uniform(a->coeffs, N, buf, buflen);
while (ctr < N) {
off = buflen % 3;
for (size_t i = 0; i < off; ++i) {
buf[i] = buf[buflen - off + i];
}
buflen = STREAM128_BLOCKBYTES + off;
stream128_squeezeblocks(buf + off, 1, &state);
ctr += rej_uniform_ref(a->coeffs + ctr, N - ctr, buf, buflen);
}
}
void PQCLEAN_DILITHIUM4_AVX2_poly_uniform_4x(poly *a0,
poly *a1,
poly *a2,
poly *a3,
const uint8_t seed[SEEDBYTES],
uint16_t nonce0,
uint16_t nonce1,
uint16_t nonce2,
uint16_t nonce3) {
size_t ctr0, ctr1, ctr2, ctr3;
uint8_t inbuf[4][SEEDBYTES + 2];
uint8_t outbuf[4][5 * SHAKE128_RATE];
__m256i state[25];
for (size_t i = 0; i < SEEDBYTES; ++i) {
inbuf[0][i] = seed[i];
inbuf[1][i] = seed[i];
inbuf[2][i] = seed[i];
inbuf[3][i] = seed[i];
}
inbuf[0][SEEDBYTES + 0] = nonce0;
inbuf[0][SEEDBYTES + 1] = nonce0 >> 8;
inbuf[1][SEEDBYTES + 0] = nonce1;
inbuf[1][SEEDBYTES + 1] = nonce1 >> 8;
inbuf[2][SEEDBYTES + 0] = nonce2;
inbuf[2][SEEDBYTES + 1] = nonce2 >> 8;
inbuf[3][SEEDBYTES + 0] = nonce3;
inbuf[3][SEEDBYTES + 1] = nonce3 >> 8;
PQCLEAN_DILITHIUM4_AVX2_shake128_absorb4x(state, inbuf[0], inbuf[1], inbuf[2], inbuf[3],
SEEDBYTES + 2);
PQCLEAN_DILITHIUM4_AVX2_shake128_squeezeblocks4x(outbuf[0], outbuf[1], outbuf[2], outbuf[3], 5,
state);
ctr0 = PQCLEAN_DILITHIUM4_AVX2_rej_uniform(a0->coeffs, N, outbuf[0], 5 * SHAKE128_RATE);
ctr1 = PQCLEAN_DILITHIUM4_AVX2_rej_uniform(a1->coeffs, N, outbuf[1], 5 * SHAKE128_RATE);
ctr2 = PQCLEAN_DILITHIUM4_AVX2_rej_uniform(a2->coeffs, N, outbuf[2], 5 * SHAKE128_RATE);
ctr3 = PQCLEAN_DILITHIUM4_AVX2_rej_uniform(a3->coeffs, N, outbuf[3], 5 * SHAKE128_RATE);
while (ctr0 < N || ctr1 < N || ctr2 < N || ctr3 < N) {
PQCLEAN_DILITHIUM4_AVX2_shake128_squeezeblocks4x(outbuf[0], outbuf[1], outbuf[2], outbuf[3], 1,
state);
ctr0 += rej_uniform_ref(a0->coeffs + ctr0, N - ctr0, outbuf[0],
SHAKE128_RATE);
ctr1 += rej_uniform_ref(a1->coeffs + ctr1, N - ctr1, outbuf[1],
SHAKE128_RATE);
ctr2 += rej_uniform_ref(a2->coeffs + ctr2, N - ctr2, outbuf[2],
SHAKE128_RATE);
ctr3 += rej_uniform_ref(a3->coeffs + ctr3, N - ctr3, outbuf[3],
SHAKE128_RATE);
}
}
/*************************************************
* Name: rej_eta
*
* Description: Sample uniformly random coefficients in [-ETA, ETA] by
* performing rejection sampling using array of random bytes.
*
* Arguments: - uint32_t *a: pointer to output array (allocated)
* - size_t len: number of coefficients to be sampled
* - const uint8_t *buf: array of random bytes
* - size_t buflen: length of array of random bytes
*
* Returns number of sampled coefficients. Can be smaller than len if not enough
* random bytes were given.
**************************************************/
static size_t rej_eta_ref(
uint32_t *a,
size_t len,
const uint8_t *buf,
size_t buflen) {
size_t ctr, pos;
uint32_t t0, t1;
ctr = pos = 0;
while (ctr < len && pos < buflen) {
t0 = buf[pos] & 0x07;
t1 = buf[pos++] >> 5;
if (t0 <= 2 * ETA) {
a[ctr++] = Q + ETA - t0;
}
if (t1 <= 2 * ETA && ctr < len) {
a[ctr++] = Q + ETA - t1;
}
}
return ctr;
}
/*************************************************
* Name: poly_uniform_eta
*
* Description: Sample polynomial with uniformly random coefficients
* in [-ETA,ETA] by performing rejection sampling using the
* output stream from SHAKE256(seed|nonce).
*
* Arguments: - poly *a: pointer to output polynomial
* - const uint8_t seed[]: byte array with seed of length
* SEEDBYTES
* - uint16_t nonce: 2-byte nonce
**************************************************/
#define POLY_UNIFORM_ETA_NBLOCKS (((N / 2 * (1u << SETABITS)) / (2 * ETA + 1) + STREAM128_BLOCKBYTES) / STREAM128_BLOCKBYTES)
#define POLY_UNIFORM_ETA_BUFLEN (POLY_UNIFORM_ETA_NBLOCKS*STREAM128_BLOCKBYTES)
void PQCLEAN_DILITHIUM4_AVX2_poly_uniform_eta(
poly *a,
const uint8_t seed[SEEDBYTES],
uint16_t nonce) {
size_t ctr;
uint8_t buf[POLY_UNIFORM_ETA_BUFLEN];
stream128_state state;
stream128_init(&state, seed, nonce);
stream128_squeezeblocks(buf, POLY_UNIFORM_ETA_NBLOCKS, &state);
ctr = PQCLEAN_DILITHIUM4_AVX2_rej_eta(a->coeffs, N, buf, POLY_UNIFORM_ETA_BUFLEN);
while (ctr < N) {
stream128_squeezeblocks(buf, 1, &state);
ctr += rej_eta_ref(a->coeffs + ctr, N - ctr, buf, STREAM128_BLOCKBYTES);
}
}
void PQCLEAN_DILITHIUM4_AVX2_poly_uniform_eta_4x(
poly *a0,
poly *a1,
poly *a2,
poly *a3,
const uint8_t seed[SEEDBYTES],
uint16_t nonce0,
uint16_t nonce1,
uint16_t nonce2,
uint16_t nonce3) {
size_t ctr0, ctr1, ctr2, ctr3;
uint8_t inbuf[4][SEEDBYTES + 2];
uint8_t outbuf[4][2 * SHAKE128_RATE];
__m256i state[25];
for (size_t i = 0; i < SEEDBYTES; ++i) {
inbuf[0][i] = seed[i];
inbuf[1][i] = seed[i];
inbuf[2][i] = seed[i];
inbuf[3][i] = seed[i];
}
inbuf[0][SEEDBYTES + 0] = nonce0;
inbuf[0][SEEDBYTES + 1] = nonce0 >> 8;
inbuf[1][SEEDBYTES + 0] = nonce1;
inbuf[1][SEEDBYTES + 1] = nonce1 >> 8;
inbuf[2][SEEDBYTES + 0] = nonce2;
inbuf[2][SEEDBYTES + 1] = nonce2 >> 8;
inbuf[3][SEEDBYTES + 0] = nonce3;
inbuf[3][SEEDBYTES + 1] = nonce3 >> 8;
PQCLEAN_DILITHIUM4_AVX2_shake128_absorb4x(state, inbuf[0], inbuf[1], inbuf[2], inbuf[3],
SEEDBYTES + 2);
PQCLEAN_DILITHIUM4_AVX2_shake128_squeezeblocks4x(outbuf[0], outbuf[1], outbuf[2], outbuf[3], 2,
state);
ctr0 = PQCLEAN_DILITHIUM4_AVX2_rej_eta(a0->coeffs, N, outbuf[0], 2 * SHAKE128_RATE);
ctr1 = PQCLEAN_DILITHIUM4_AVX2_rej_eta(a1->coeffs, N, outbuf[1], 2 * SHAKE128_RATE);
ctr2 = PQCLEAN_DILITHIUM4_AVX2_rej_eta(a2->coeffs, N, outbuf[2], 2 * SHAKE128_RATE);
ctr3 = PQCLEAN_DILITHIUM4_AVX2_rej_eta(a3->coeffs, N, outbuf[3], 2 * SHAKE128_RATE);
while (ctr0 < N || ctr1 < N || ctr2 < N || ctr3 < N) {
PQCLEAN_DILITHIUM4_AVX2_shake128_squeezeblocks4x(outbuf[0], outbuf[1], outbuf[2], outbuf[3], 1,
state);
ctr0 += rej_eta_ref(a0->coeffs + ctr0, N - ctr0, outbuf[0], SHAKE128_RATE);
ctr1 += rej_eta_ref(a1->coeffs + ctr1, N - ctr1, outbuf[1], SHAKE128_RATE);
ctr2 += rej_eta_ref(a2->coeffs + ctr2, N - ctr2, outbuf[2], SHAKE128_RATE);
ctr3 += rej_eta_ref(a3->coeffs + ctr3, N - ctr3, outbuf[3], SHAKE128_RATE);
}
}
/*************************************************
* Name: rej_gamma1m1_ref
*
* Description: Sample uniformly random coefficients
* in [-(GAMMA1 - 1), GAMMA1 - 1] by performing rejection sampling
* using array of random bytes.
*
* Arguments: - uint32_t *a: pointer to output array (allocated)
* - size_t len: number of coefficients to be sampled
* - const uint8_t *buf: array of random bytes
* - size_t buflen: length of array of random bytes
*
* Returns number of sampled coefficients. Can be smaller than len if not enough
* random bytes were given.
**************************************************/
static size_t rej_gamma1m1_ref(
uint32_t *a,
size_t len,
const uint8_t *buf,
size_t buflen) {
size_t ctr, pos;
uint32_t t0, t1;
ctr = pos = 0;
while (ctr < len && pos + 5 <= buflen) {
t0 = buf[pos];
t0 |= (uint32_t)buf[pos + 1] << 8;
t0 |= (uint32_t)buf[pos + 2] << 16;
t0 &= 0xFFFFF;
t1 = buf[pos + 2] >> 4;
t1 |= (uint32_t)buf[pos + 3] << 4;
t1 |= (uint32_t)buf[pos + 4] << 12;
pos += 5;
if (t0 <= 2 * GAMMA1 - 2) {
a[ctr++] = Q + GAMMA1 - 1 - t0;
}
if (t1 <= 2 * GAMMA1 - 2 && ctr < len) {
a[ctr++] = Q + GAMMA1 - 1 - t1;
}
}
return ctr;
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_poly_uniform_gamma1m1
*
* Description: Sample polynomial with uniformly random coefficients
* in [-(GAMMA1 - 1), GAMMA1 - 1] by performing rejection
* sampling on output stream of SHAKE256(seed|nonce).
*
* Arguments: - poly *a: pointer to output polynomial
* - const uint8_t seed[]: byte array with seed of length
* CRHBYTES
* - uint16_t nonce: 16-bit nonce
**************************************************/
#define POLY_UNIFORM_GAMMA1M1_NBLOCKS ((641 + STREAM256_BLOCKBYTES) / STREAM256_BLOCKBYTES)
#define POLY_UNIFORM_GAMMA1M1_BUFLEN (POLY_UNIFORM_GAMMA1M1_NBLOCKS * STREAM256_BLOCKBYTES)
void PQCLEAN_DILITHIUM4_AVX2_poly_uniform_gamma1m1(
poly *a,
const uint8_t seed[CRHBYTES],
uint16_t nonce) {
size_t ctr, off;
size_t buflen = POLY_UNIFORM_GAMMA1M1_BUFLEN;
uint8_t buf[POLY_UNIFORM_GAMMA1M1_BUFLEN + 4];
stream256_state state;
stream256_init(&state, seed, nonce);
stream256_squeezeblocks(buf, POLY_UNIFORM_GAMMA1M1_NBLOCKS, &state);
ctr = PQCLEAN_DILITHIUM4_AVX2_rej_gamma1m1(a->coeffs, N, buf, POLY_UNIFORM_GAMMA1M1_BUFLEN);
while (ctr < N) {
off = buflen % 5;
for (size_t i = 0; i < off; ++i) {
buf[i] = buf[buflen - off + i];
}
buflen = STREAM256_BLOCKBYTES + off;
stream256_squeezeblocks(buf + off, 1, &state);
ctr += rej_gamma1m1_ref(a->coeffs + ctr, N - ctr, buf, buflen);
}
}
void PQCLEAN_DILITHIUM4_AVX2_poly_uniform_gamma1m1_4x(poly *a0,
poly *a1,
poly *a2,
poly *a3,
const uint8_t seed[CRHBYTES],
uint16_t nonce0,
uint16_t nonce1,
uint16_t nonce2,
uint16_t nonce3) {
size_t ctr0, ctr1, ctr2, ctr3;
uint8_t inbuf[4][CRHBYTES + 2];
uint8_t outbuf[4][5 * SHAKE256_RATE];
__m256i state[25];
for (size_t i = 0; i < CRHBYTES; ++i) {
inbuf[0][i] = seed[i];
inbuf[1][i] = seed[i];
inbuf[2][i] = seed[i];
inbuf[3][i] = seed[i];
}
inbuf[0][CRHBYTES + 0] = nonce0 & 0xFF;
inbuf[0][CRHBYTES + 1] = nonce0 >> 8;
inbuf[1][CRHBYTES + 0] = nonce1 & 0xFF;
inbuf[1][CRHBYTES + 1] = nonce1 >> 8;
inbuf[2][CRHBYTES + 0] = nonce2 & 0xFF;
inbuf[2][CRHBYTES + 1] = nonce2 >> 8;
inbuf[3][CRHBYTES + 0] = nonce3 & 0xFF;
inbuf[3][CRHBYTES + 1] = nonce3 >> 8;
PQCLEAN_DILITHIUM4_AVX2_shake256_absorb4x(state, inbuf[0], inbuf[1], inbuf[2], inbuf[3],
CRHBYTES + 2);
PQCLEAN_DILITHIUM4_AVX2_shake256_squeezeblocks4x(outbuf[0], outbuf[1], outbuf[2], outbuf[3], 5,
state);
ctr0 = rej_gamma1m1_ref(a0->coeffs, N, outbuf[0], 5 * SHAKE256_RATE);
ctr1 = rej_gamma1m1_ref(a1->coeffs, N, outbuf[1], 5 * SHAKE256_RATE);
ctr2 = rej_gamma1m1_ref(a2->coeffs, N, outbuf[2], 5 * SHAKE256_RATE);
ctr3 = rej_gamma1m1_ref(a3->coeffs, N, outbuf[3], 5 * SHAKE256_RATE);
while (ctr0 < N || ctr1 < N || ctr2 < N || ctr3 < N) {
PQCLEAN_DILITHIUM4_AVX2_shake256_squeezeblocks4x(outbuf[0], outbuf[1], outbuf[2], outbuf[3], 1,
state);
ctr0 += rej_gamma1m1_ref(a0->coeffs + ctr0, N - ctr0, outbuf[0],
SHAKE256_RATE);
ctr1 += rej_gamma1m1_ref(a1->coeffs + ctr1, N - ctr1, outbuf[1],
SHAKE256_RATE);
ctr2 += rej_gamma1m1_ref(a2->coeffs + ctr2, N - ctr2, outbuf[2],
SHAKE256_RATE);
ctr3 += rej_gamma1m1_ref(a3->coeffs + ctr3, N - ctr3, outbuf[3],
SHAKE256_RATE);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyeta_pack
*
* Description: Bit-pack polynomial with coefficients in [-ETA,ETA].
* Input coefficients are assumed to lie in [Q-ETA,Q+ETA].
*
* Arguments: - uint8_t *r: pointer to output byte array with at least
* POLETA_SIZE_PACKED bytes
* - const poly *a: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyeta_pack(uint8_t *restrict r, const poly *restrict a) {
uint8_t t[8];
for (size_t i = 0; i < N / 8; ++i) {
t[0] = Q + ETA - a->coeffs[8 * i + 0];
t[1] = Q + ETA - a->coeffs[8 * i + 1];
t[2] = Q + ETA - a->coeffs[8 * i + 2];
t[3] = Q + ETA - a->coeffs[8 * i + 3];
t[4] = Q + ETA - a->coeffs[8 * i + 4];
t[5] = Q + ETA - a->coeffs[8 * i + 5];
t[6] = Q + ETA - a->coeffs[8 * i + 6];
t[7] = Q + ETA - a->coeffs[8 * i + 7];
r[3 * i + 0] = (t[0] >> 0) | (t[1] << 3) | (t[2] << 6);
r[3 * i + 1] = (t[2] >> 2) | (t[3] << 1) | (t[4] << 4) | (t[5] << 7);
r[3 * i + 2] = (t[5] >> 1) | (t[6] << 2) | (t[7] << 5);
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyeta_unpack
*
* Description: Unpack polynomial with coefficients in [-ETA,ETA].
* Output coefficients lie in [Q-ETA,Q+ETA].
*
* Arguments: - poly *r: pointer to output polynomial
* - const uint8_t *a: byte array with bit-packed polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyeta_unpack(poly *restrict r, const uint8_t *restrict a) {
for (size_t i = 0; i < N / 8; ++i) {
r->coeffs[8 * i + 0] = a[3 * i + 0] & 0x07;
r->coeffs[8 * i + 1] = (a[3 * i + 0] >> 3) & 0x07;
r->coeffs[8 * i + 2] = (uint32_t)((a[3 * i + 0] >> 6) | (a[3 * i + 1] << 2)) & 0x07;
r->coeffs[8 * i + 3] = (a[3 * i + 1] >> 1) & 0x07;
r->coeffs[8 * i + 4] = (a[3 * i + 1] >> 4) & 0x07;
r->coeffs[8 * i + 5] = (uint32_t)((a[3 * i + 1] >> 7) | (a[3 * i + 2] << 1)) & 0x07;
r->coeffs[8 * i + 6] = (a[3 * i + 2] >> 2) & 0x07;
r->coeffs[8 * i + 7] = (a[3 * i + 2] >> 5) & 0x07;
r->coeffs[8 * i + 0] = Q + ETA - r->coeffs[8 * i + 0];
r->coeffs[8 * i + 1] = Q + ETA - r->coeffs[8 * i + 1];
r->coeffs[8 * i + 2] = Q + ETA - r->coeffs[8 * i + 2];
r->coeffs[8 * i + 3] = Q + ETA - r->coeffs[8 * i + 3];
r->coeffs[8 * i + 4] = Q + ETA - r->coeffs[8 * i + 4];
r->coeffs[8 * i + 5] = Q + ETA - r->coeffs[8 * i + 5];
r->coeffs[8 * i + 6] = Q + ETA - r->coeffs[8 * i + 6];
r->coeffs[8 * i + 7] = Q + ETA - r->coeffs[8 * i + 7];
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyt1_pack
*
* Description: Bit-pack polynomial t1 with coefficients fitting in 9 bits.
* Input coefficients are assumed to be standard representatives.
*
* Arguments: - uint8_t *r: pointer to output byte array with at least
* POLT1_SIZE_PACKED bytes
* - const poly *a: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyt1_pack(uint8_t *restrict r, const poly *restrict a) {
for (size_t i = 0; i < N / 8; ++i) {
r[9 * i + 0] = (uint8_t)((a->coeffs[8 * i + 0] >> 0));
r[9 * i + 1] = (uint8_t)((a->coeffs[8 * i + 0] >> 8) | (a->coeffs[8 * i + 1] << 1));
r[9 * i + 2] = (uint8_t)((a->coeffs[8 * i + 1] >> 7) | (a->coeffs[8 * i + 2] << 2));
r[9 * i + 3] = (uint8_t)((a->coeffs[8 * i + 2] >> 6) | (a->coeffs[8 * i + 3] << 3));
r[9 * i + 4] = (uint8_t)((a->coeffs[8 * i + 3] >> 5) | (a->coeffs[8 * i + 4] << 4));
r[9 * i + 5] = (uint8_t)((a->coeffs[8 * i + 4] >> 4) | (a->coeffs[8 * i + 5] << 5));
r[9 * i + 6] = (uint8_t)((a->coeffs[8 * i + 5] >> 3) | (a->coeffs[8 * i + 6] << 6));
r[9 * i + 7] = (uint8_t)((a->coeffs[8 * i + 6] >> 2) | (a->coeffs[8 * i + 7] << 7));
r[9 * i + 8] = (uint8_t)((a->coeffs[8 * i + 7] >> 1));
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyt1_unpack
*
* Description: Unpack polynomial t1 with 9-bit coefficients.
* Output coefficients are standard representatives.
*
* Arguments: - poly *r: pointer to output polynomial
* - const uint8_t *a: byte array with bit-packed polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyt1_unpack(poly *restrict r, const uint8_t *restrict a) {
for (size_t i = 0; i < N / 8; ++i) {
r->coeffs[8 * i + 0] = ((a[9 * i + 0] >> 0) | ((uint32_t)a[9 * i + 1] << 8)) & 0x1FF;
r->coeffs[8 * i + 1] = ((a[9 * i + 1] >> 1) | ((uint32_t)a[9 * i + 2] << 7)) & 0x1FF;
r->coeffs[8 * i + 2] = ((a[9 * i + 2] >> 2) | ((uint32_t)a[9 * i + 3] << 6)) & 0x1FF;
r->coeffs[8 * i + 3] = ((a[9 * i + 3] >> 3) | ((uint32_t)a[9 * i + 4] << 5)) & 0x1FF;
r->coeffs[8 * i + 4] = ((a[9 * i + 4] >> 4) | ((uint32_t)a[9 * i + 5] << 4)) & 0x1FF;
r->coeffs[8 * i + 5] = ((a[9 * i + 5] >> 5) | ((uint32_t)a[9 * i + 6] << 3)) & 0x1FF;
r->coeffs[8 * i + 6] = ((a[9 * i + 6] >> 6) | ((uint32_t)a[9 * i + 7] << 2)) & 0x1FF;
r->coeffs[8 * i + 7] = ((a[9 * i + 7] >> 7) | ((uint32_t)a[9 * i + 8] << 1)) & 0x1FF;
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyt0_pack
*
* Description: Bit-pack polynomial t0 with coefficients in ]-2^{D-1}, 2^{D-1}].
* Input coefficients are assumed to lie in ]Q-2^{D-1}, Q+2^{D-1}].
*
* Arguments: - uint8_t *r: pointer to output byte array with at least
* POLT0_SIZE_PACKED bytes
* - const poly *a: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyt0_pack(uint8_t *restrict r, const poly *restrict a) {
uint32_t t[4];
for (size_t i = 0; i < N / 4; ++i) {
t[0] = Q + (1U << (D - 1)) - a->coeffs[4 * i + 0];
t[1] = Q + (1U << (D - 1)) - a->coeffs[4 * i + 1];
t[2] = Q + (1U << (D - 1)) - a->coeffs[4 * i + 2];
t[3] = Q + (1U << (D - 1)) - a->coeffs[4 * i + 3];
r[7 * i + 0] = t[0];
r[7 * i + 1] = t[0] >> 8;
r[7 * i + 1] |= t[1] << 6;
r[7 * i + 2] = t[1] >> 2;
r[7 * i + 3] = t[1] >> 10;
r[7 * i + 3] |= t[2] << 4;
r[7 * i + 4] = t[2] >> 4;
r[7 * i + 5] = t[2] >> 12;
r[7 * i + 5] |= t[3] << 2;
r[7 * i + 6] = t[3] >> 6;
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyt0_unpack
*
* Description: Unpack polynomial t0 with coefficients in ]-2^{D-1}, 2^{D-1}].
* Output coefficients lie in ]Q-2^{D-1},Q+2^{D-1}].
*
* Arguments: - poly *r: pointer to output polynomial
* - const uint8_t *a: byte array with bit-packed polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyt0_unpack(poly *restrict r, const uint8_t *restrict a) {
for (size_t i = 0; i < N / 4; ++i) {
r->coeffs[4 * i + 0] = a[7 * i + 0];
r->coeffs[4 * i + 0] |= (uint32_t)(a[7 * i + 1] & 0x3F) << 8;
r->coeffs[4 * i + 1] = a[7 * i + 1] >> 6;
r->coeffs[4 * i + 1] |= (uint32_t)a[7 * i + 2] << 2;
r->coeffs[4 * i + 1] |= (uint32_t)(a[7 * i + 3] & 0x0F) << 10;
r->coeffs[4 * i + 2] = a[7 * i + 3] >> 4;
r->coeffs[4 * i + 2] |= (uint32_t)a[7 * i + 4] << 4;
r->coeffs[4 * i + 2] |= (uint32_t)(a[7 * i + 5] & 0x03) << 12;
r->coeffs[4 * i + 3] = a[7 * i + 5] >> 2;
r->coeffs[4 * i + 3] |= (uint32_t)a[7 * i + 6] << 6;
r->coeffs[4 * i + 0] = Q + (1U << (D - 1)) - r->coeffs[4 * i + 0];
r->coeffs[4 * i + 1] = Q + (1U << (D - 1)) - r->coeffs[4 * i + 1];
r->coeffs[4 * i + 2] = Q + (1U << (D - 1)) - r->coeffs[4 * i + 2];
r->coeffs[4 * i + 3] = Q + (1U << (D - 1)) - r->coeffs[4 * i + 3];
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyz_pack
*
* Description: Bit-pack polynomial z with coefficients
* in [-(GAMMA1 - 1), GAMMA1 - 1].
* Input coefficients are assumed to be standard representatives.
*
* Arguments: - uint8_t *r: pointer to output byte array with at least
* POLZ_SIZE_PACKED bytes
* - const poly *a: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyz_pack(uint8_t *restrict r, const poly *restrict a) {
uint32_t t[2];
for (size_t i = 0; i < N / 2; ++i) {
/* Map to {0,...,2*GAMMA1 - 2} */
t[0] = GAMMA1 - 1 - a->coeffs[2 * i + 0];
t[0] += ((int32_t)t[0] >> 31) & Q;
t[1] = GAMMA1 - 1 - a->coeffs[2 * i + 1];
t[1] += ((int32_t)t[1] >> 31) & Q;
r[5 * i + 0] = t[0];
r[5 * i + 1] = t[0] >> 8;
r[5 * i + 2] = t[0] >> 16;
r[5 * i + 2] |= t[1] << 4;
r[5 * i + 3] = t[1] >> 4;
r[5 * i + 4] = t[1] >> 12;
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyz_unpack
*
* Description: Unpack polynomial z with coefficients
* in [-(GAMMA1 - 1), GAMMA1 - 1].
* Output coefficients are standard representatives.
*
* Arguments: - poly *r: pointer to output polynomial
* - const uint8_t *a: byte array with bit-packed polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyz_unpack(poly *restrict r, const uint8_t *restrict a) {
for (size_t i = 0; i < N / 2; ++i) {
r->coeffs[2 * i + 0] = a[5 * i + 0];
r->coeffs[2 * i + 0] |= (uint32_t)a[5 * i + 1] << 8;
r->coeffs[2 * i + 0] |= (uint32_t)(a[5 * i + 2] & 0x0F) << 16;
r->coeffs[2 * i + 1] = a[5 * i + 2] >> 4;
r->coeffs[2 * i + 1] |= (uint32_t)a[5 * i + 3] << 4;
r->coeffs[2 * i + 1] |= (uint32_t)a[5 * i + 4] << 12;
r->coeffs[2 * i + 0] = GAMMA1 - 1 - r->coeffs[2 * i + 0];
r->coeffs[2 * i + 0] += ((int32_t)r->coeffs[2 * i + 0] >> 31) & Q;
r->coeffs[2 * i + 1] = GAMMA1 - 1 - r->coeffs[2 * i + 1];
r->coeffs[2 * i + 1] += ((int32_t)r->coeffs[2 * i + 1] >> 31) & Q;
}
}
/*************************************************
* Name: PQCLEAN_DILITHIUM4_AVX2_polyw1_pack
*
* Description: Bit-pack polynomial w1 with coefficients in [0, 15].
* Input coefficients are assumed to be standard representatives.
*
* Arguments: - uint8_t *r: pointer to output byte array with at least
* POLW1_SIZE_PACKED bytes
* - const poly *a: pointer to input polynomial
**************************************************/
void PQCLEAN_DILITHIUM4_AVX2_polyw1_pack(
uint8_t *restrict r,
const poly *restrict a) {
for (size_t i = 0; i < N / 2; ++i) {
r[i] = a->coeffs[2 * i + 0] | (a->coeffs[2 * i + 1] << 4);
}
}
Reference in New Issue View Git Blame Copy Permalink