ac2c20045c
* Add McEliece reference implementations * Add Vec implementations of McEliece * Add sse implementations * Add AVX2 implementations * Get rid of stuff not supported by Mac ABI * restrict to two cores * Ditch .data files * Remove .hidden from all .S files * speed up duplicate consistency tests by batching * make cpuinfo more robust * Hope to stabilize macos cpuinfo without ccache * Revert "Hope to stabilize macos cpuinfo without ccache" This reverts commit 6129c3cabe1abbc8b956bc87e902a698e32bf322. * Just hardcode what's available at travis * Fixed-size types in api.h * namespace all header files in mceliece * Ditch operations.h * Get rid of static inline functions * fixup! Ditch operations.h
291 lines
8.9 KiB
C
291 lines
8.9 KiB
C
/*
|
|
This file is for public-key generation
|
|
*/
|
|
|
|
#include "pk_gen.h"
|
|
|
|
#include "benes.h"
|
|
#include "controlbits.h"
|
|
#include "fft.h"
|
|
#include "params.h"
|
|
#include "util.h"
|
|
|
|
#include <stdint.h>
|
|
|
|
static void de_bitslicing(uint64_t *out, vec256 in[][GFBITS]) {
|
|
int i, j, r;
|
|
uint64_t u = 0;
|
|
|
|
for (i = 0; i < (1 << GFBITS); i++) {
|
|
out[i] = 0 ;
|
|
}
|
|
|
|
for (i = 0; i < 32; i++) {
|
|
for (j = GFBITS - 1; j >= 0; j--) {
|
|
u = PQCLEAN_MCELIECE460896_AVX_vec256_extract(in[i][j], 0);
|
|
for (r = 0; r < 64; r++) {
|
|
out[i * 256 + 0 * 64 + r] <<= 1;
|
|
out[i * 256 + 0 * 64 + r] |= (u >> r) & 1;
|
|
}
|
|
u = PQCLEAN_MCELIECE460896_AVX_vec256_extract(in[i][j], 1);
|
|
for (r = 0; r < 64; r++) {
|
|
out[i * 256 + 1 * 64 + r] <<= 1;
|
|
out[i * 256 + 1 * 64 + r] |= (u >> r) & 1;
|
|
}
|
|
u = PQCLEAN_MCELIECE460896_AVX_vec256_extract(in[i][j], 2);
|
|
for (r = 0; r < 64; r++) {
|
|
out[i * 256 + 2 * 64 + r] <<= 1;
|
|
out[i * 256 + 2 * 64 + r] |= (u >> r) & 1;
|
|
}
|
|
u = PQCLEAN_MCELIECE460896_AVX_vec256_extract(in[i][j], 3);
|
|
for (r = 0; r < 64; r++) {
|
|
out[i * 256 + 3 * 64 + r] <<= 1;
|
|
out[i * 256 + 3 * 64 + r] |= (u >> r) & 1;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
static void to_bitslicing_2x(vec256 out0[][GFBITS], vec256 out1[][GFBITS], const uint64_t *in) {
|
|
int i, j, k, r;
|
|
uint64_t u[4] = {0};
|
|
|
|
for (i = 0; i < 32; i++) {
|
|
for (j = GFBITS - 1; j >= 0; j--) {
|
|
for (k = 0; k < 4; k++) {
|
|
for (r = 63; r >= 0; r--) {
|
|
u[k] <<= 1;
|
|
u[k] |= (in[i * 256 + k * 64 + r] >> (j + GFBITS)) & 1;
|
|
}
|
|
}
|
|
|
|
out1[i][j] = PQCLEAN_MCELIECE460896_AVX_vec256_set4x(u[0], u[1], u[2], u[3]);
|
|
}
|
|
|
|
for (j = GFBITS - 1; j >= 0; j--) {
|
|
for (k = 0; k < 4; k++) {
|
|
for (r = 63; r >= 0; r--) {
|
|
u[k] <<= 1;
|
|
u[k] |= (in[i * 256 + k * 64 + r] >> j) & 1;
|
|
}
|
|
}
|
|
|
|
out0[i][GFBITS - 1 - j] = PQCLEAN_MCELIECE460896_AVX_vec256_set4x(u[0], u[1], u[2], u[3]);
|
|
}
|
|
}
|
|
}
|
|
|
|
#define NBLOCKS1_H ((SYS_N + 63) / 64)
|
|
#define NBLOCKS2_H ((SYS_N + 255) / 256)
|
|
#define NBLOCKS1_I ((GFBITS * SYS_T + 63) / 64)
|
|
#define NBLOCKS2_I ((GFBITS * SYS_T + 255) / 256)
|
|
int PQCLEAN_MCELIECE460896_AVX_pk_gen(unsigned char *pk, uint32_t *perm, const unsigned char *sk) {
|
|
const int block_idx = NBLOCKS1_I - 1;
|
|
int tail = (GFBITS * SYS_T) % 64;
|
|
|
|
int i, j, k;
|
|
int row, c;
|
|
|
|
uint64_t mat[ GFBITS * SYS_T ][ NBLOCKS2_H * 4 ];
|
|
uint64_t ops[ GFBITS * SYS_T ][ NBLOCKS1_I ];
|
|
|
|
uint64_t mask;
|
|
|
|
vec128 sk_int[ GFBITS ];
|
|
|
|
vec256 consts[ 32 ][ GFBITS ];
|
|
vec256 eval[ 32 ][ GFBITS ];
|
|
vec256 prod[ 32 ][ GFBITS ];
|
|
vec256 tmp[ GFBITS ];
|
|
|
|
uint64_t list[1 << GFBITS];
|
|
uint64_t one_row[ 128 ];
|
|
|
|
// compute the inverses
|
|
|
|
PQCLEAN_MCELIECE460896_AVX_irr_load(sk_int, sk);
|
|
|
|
PQCLEAN_MCELIECE460896_AVX_fft(eval, sk_int);
|
|
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_copy(prod[0], eval[0]);
|
|
|
|
for (i = 1; i < 32; i++) {
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_mul(prod[i], prod[i - 1], eval[i]);
|
|
}
|
|
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_inv(tmp, prod[31]);
|
|
|
|
for (i = 30; i >= 0; i--) {
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_mul(prod[i + 1], prod[i], tmp);
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_mul(tmp, tmp, eval[i + 1]);
|
|
}
|
|
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_copy(prod[0], tmp);
|
|
|
|
// fill matrix
|
|
|
|
de_bitslicing(list, prod);
|
|
|
|
for (i = 0; i < (1 << GFBITS); i++) {
|
|
list[i] <<= GFBITS;
|
|
list[i] |= i;
|
|
list[i] |= ((uint64_t) perm[i]) << 31;
|
|
}
|
|
|
|
PQCLEAN_MCELIECE460896_AVX_sort_63b(1 << GFBITS, list);
|
|
|
|
to_bitslicing_2x(consts, prod, list);
|
|
|
|
for (i = 0; i < (1 << GFBITS); i++) {
|
|
perm[i] = list[i] & GFMASK;
|
|
}
|
|
|
|
for (j = 0; j < NBLOCKS2_I; j++) {
|
|
for (k = 0; k < GFBITS; k++) {
|
|
mat[ k ][ 4 * j + 0 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 0);
|
|
mat[ k ][ 4 * j + 1 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 1);
|
|
mat[ k ][ 4 * j + 2 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 2);
|
|
mat[ k ][ 4 * j + 3 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 3);
|
|
}
|
|
}
|
|
|
|
for (i = 1; i < SYS_T; i++) {
|
|
for (j = 0; j < NBLOCKS2_I; j++) {
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_mul(prod[j], prod[j], consts[j]);
|
|
|
|
for (k = 0; k < GFBITS; k++) {
|
|
mat[ i * GFBITS + k ][ 4 * j + 0 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 0);
|
|
mat[ i * GFBITS + k ][ 4 * j + 1 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 1);
|
|
mat[ i * GFBITS + k ][ 4 * j + 2 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 2);
|
|
mat[ i * GFBITS + k ][ 4 * j + 3 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 3);
|
|
}
|
|
}
|
|
}
|
|
|
|
// gaussian elimination to obtain an upper triangular matrix
|
|
// and keep track of the operations in ops
|
|
|
|
for (i = 0; i < PK_NROWS; i++) {
|
|
for (j = 0; j < NBLOCKS1_I; j++) {
|
|
ops[ i ][ j ] = 0;
|
|
}
|
|
}
|
|
|
|
for (i = 0; i < PK_NROWS; i++) {
|
|
ops[ i ][ i / 64 ] = 1;
|
|
ops[ i ][ i / 64 ] <<= (i % 64);
|
|
}
|
|
|
|
uint64_t column[ PK_NROWS ];
|
|
|
|
for (i = 0; i < PK_NROWS; i++) {
|
|
column[i] = mat[ i ][ block_idx ];
|
|
}
|
|
|
|
for (row = 0; row < PK_NROWS; row++) {
|
|
i = row >> 6;
|
|
j = row & 63;
|
|
|
|
for (k = row + 1; k < PK_NROWS; k++) {
|
|
mask = mat[ row ][ i ] >> j;
|
|
mask &= 1;
|
|
mask -= 1;
|
|
|
|
for (c = 0; c < NBLOCKS1_I; c++) {
|
|
mat[ row ][ c ] ^= mat[ k ][ c ] & mask;
|
|
ops[ row ][ c ] ^= ops[ k ][ c ] & mask;
|
|
}
|
|
}
|
|
|
|
if ( ((mat[ row ][ i ] >> j) & 1) == 0 ) { // return if not systematic
|
|
return -1;
|
|
}
|
|
|
|
for (k = row + 1; k < PK_NROWS; k++) {
|
|
mask = mat[ k ][ i ] >> j;
|
|
mask &= 1;
|
|
mask = -mask;
|
|
|
|
for (c = 0; c < NBLOCKS1_I; c++) {
|
|
mat[ k ][ c ] ^= mat[ row ][ c ] & mask;
|
|
ops[ k ][ c ] ^= ops[ row ][ c ] & mask;
|
|
}
|
|
}
|
|
}
|
|
|
|
// computing the lineaer map required to obatin the systematic form
|
|
|
|
for (row = PK_NROWS - 1; row >= 0; row--) {
|
|
for (k = 0; k < row; k++) {
|
|
mask = mat[ k ][ row / 64 ] >> (row & 63);
|
|
mask &= 1;
|
|
mask = -mask;
|
|
|
|
for (c = 0; c < NBLOCKS1_I; c++) {
|
|
ops[ k ][ c ] ^= ops[ row ][ c ] & mask;
|
|
}
|
|
}
|
|
}
|
|
|
|
// apply the linear map to the non-systematic part
|
|
|
|
for (j = NBLOCKS2_I; j < NBLOCKS2_H; j++) {
|
|
for (k = 0; k < GFBITS; k++) {
|
|
mat[ k ][ 4 * j + 0 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 0);
|
|
mat[ k ][ 4 * j + 1 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 1);
|
|
mat[ k ][ 4 * j + 2 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 2);
|
|
mat[ k ][ 4 * j + 3 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 3);
|
|
}
|
|
}
|
|
|
|
for (i = 1; i < SYS_T; i++) {
|
|
for (j = NBLOCKS2_I; j < NBLOCKS2_H; j++) {
|
|
PQCLEAN_MCELIECE460896_AVX_vec256_mul(prod[j], prod[j], consts[j]);
|
|
|
|
for (k = 0; k < GFBITS; k++) {
|
|
mat[ i * GFBITS + k ][ 4 * j + 0 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 0);
|
|
mat[ i * GFBITS + k ][ 4 * j + 1 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 1);
|
|
mat[ i * GFBITS + k ][ 4 * j + 2 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 2);
|
|
mat[ i * GFBITS + k ][ 4 * j + 3 ] = PQCLEAN_MCELIECE460896_AVX_vec256_extract(prod[ j ][ k ], 3);
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
for (i = 0; i < PK_NROWS; i++) {
|
|
mat[ i ][ block_idx ] = column[i];
|
|
}
|
|
|
|
for (row = 0; row < PK_NROWS; row++) {
|
|
for (k = 0; k < NBLOCKS1_H; k++) {
|
|
one_row[ k ] = 0;
|
|
}
|
|
|
|
for (c = 0; c < PK_NROWS; c++) {
|
|
mask = ops[ row ][ c >> 6 ] >> (c & 63);
|
|
mask &= 1;
|
|
mask = -mask;
|
|
|
|
for (k = block_idx; k < NBLOCKS1_H; k++) {
|
|
one_row[ k ] ^= mat[ c ][ k ] & mask;
|
|
}
|
|
}
|
|
|
|
for (k = block_idx; k < NBLOCKS1_H - 1; k++) {
|
|
one_row[k] = (one_row[k] >> tail) | (one_row[k + 1] << (64 - tail));
|
|
PQCLEAN_MCELIECE460896_AVX_store8(pk, one_row[k]);
|
|
pk += 8;
|
|
}
|
|
|
|
one_row[k] >>= tail;
|
|
PQCLEAN_MCELIECE460896_AVX_store_i(pk, one_row[k], PK_ROW_BYTES % 8);
|
|
|
|
pk += PK_ROW_BYTES % 8;
|
|
}
|
|
|
|
//
|
|
|
|
return 0;
|
|
}
|
|
|