b3f9d4f8d6
* 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
148 lines
3.6 KiB
C
148 lines
3.6 KiB
C
/*
|
|
This file is for Benes network related functions
|
|
*/
|
|
|
|
#include "benes.h"
|
|
|
|
#include "params.h"
|
|
#include "transpose.h"
|
|
#include "util.h"
|
|
#include "vec.h"
|
|
|
|
/* middle layers of the benes network */
|
|
static void layer_in(uint64_t data[2][64], uint64_t *bits, int lgs) {
|
|
int i, j, s;
|
|
|
|
uint64_t d;
|
|
|
|
s = 1 << lgs;
|
|
|
|
for (i = 0; i < 64; i += s * 2) {
|
|
for (j = i; j < i + s; j++) {
|
|
|
|
d = (data[0][j + 0] ^ data[0][j + s]);
|
|
d &= (*bits++);
|
|
data[0][j + 0] ^= d;
|
|
data[0][j + s] ^= d;
|
|
|
|
d = (data[1][j + 0] ^ data[1][j + s]);
|
|
d &= (*bits++);
|
|
data[1][j + 0] ^= d;
|
|
data[1][j + s] ^= d;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* first and last layers of the benes network */
|
|
static void layer_ex(uint64_t *data, uint64_t *bits, int lgs) {
|
|
int i, j, s;
|
|
|
|
uint64_t d;
|
|
|
|
s = 1 << lgs;
|
|
|
|
for (i = 0; i < 128; i += s * 2) {
|
|
for (j = i; j < i + s; j++) {
|
|
|
|
d = (data[j + 0] ^ data[j + s]);
|
|
d &= (*bits++);
|
|
data[j + 0] ^= d;
|
|
data[j + s] ^= d;
|
|
}
|
|
}
|
|
}
|
|
|
|
/* input: r, sequence of bits to be permuted */
|
|
/* bits, condition bits of the Benes network */
|
|
/* rev, 0 for normal application; !0 for inverse */
|
|
/* output: r, permuted bits */
|
|
void PQCLEAN_MCELIECE8192128_VEC_benes(vec *r, const unsigned char *bits, int rev) {
|
|
int i, iter, inc;
|
|
|
|
const unsigned char *bits_ptr;
|
|
|
|
uint64_t r_int_v[2][64];
|
|
uint64_t r_int_h[2][64];
|
|
uint64_t b_int_v[64];
|
|
uint64_t b_int_h[64];
|
|
|
|
//
|
|
|
|
if (rev) {
|
|
bits_ptr = bits + 12288;
|
|
inc = -1024;
|
|
} else {
|
|
bits_ptr = bits;
|
|
inc = 0;
|
|
}
|
|
|
|
for (i = 0; i < 64; i++) {
|
|
r_int_v[0][i] = r[i * 2 + 0];
|
|
r_int_v[1][i] = r[i * 2 + 1];
|
|
}
|
|
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_h[0], r_int_v[0]);
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_h[1], r_int_v[1]);
|
|
|
|
for (iter = 0; iter <= 6; iter++) {
|
|
for (i = 0; i < 64; i++) {
|
|
b_int_v[i] = PQCLEAN_MCELIECE8192128_VEC_load8(bits_ptr);
|
|
bits_ptr += 8;
|
|
}
|
|
|
|
bits_ptr += inc;
|
|
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(b_int_h, b_int_v);
|
|
|
|
layer_ex(r_int_h[0], b_int_h, iter);
|
|
}
|
|
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_v[0], r_int_h[0]);
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_v[1], r_int_h[1]);
|
|
|
|
for (iter = 0; iter <= 5; iter++) {
|
|
for (i = 0; i < 64; i++) {
|
|
b_int_v[i] = PQCLEAN_MCELIECE8192128_VEC_load8(bits_ptr);
|
|
bits_ptr += 8;
|
|
}
|
|
bits_ptr += inc;
|
|
|
|
layer_in(r_int_v, b_int_v, iter);
|
|
}
|
|
|
|
for (iter = 4; iter >= 0; iter--) {
|
|
for (i = 0; i < 64; i++) {
|
|
b_int_v[i] = PQCLEAN_MCELIECE8192128_VEC_load8(bits_ptr);
|
|
bits_ptr += 8;
|
|
}
|
|
bits_ptr += inc;
|
|
|
|
layer_in(r_int_v, b_int_v, iter);
|
|
}
|
|
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_h[0], r_int_v[0]);
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_h[1], r_int_v[1]);
|
|
|
|
for (iter = 6; iter >= 0; iter--) {
|
|
for (i = 0; i < 64; i++) {
|
|
b_int_v[i] = PQCLEAN_MCELIECE8192128_VEC_load8(bits_ptr);
|
|
bits_ptr += 8;
|
|
}
|
|
|
|
bits_ptr += inc;
|
|
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(b_int_h, b_int_v);
|
|
|
|
layer_ex(r_int_h[0], b_int_h, iter);
|
|
}
|
|
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_v[0], r_int_h[0]);
|
|
PQCLEAN_MCELIECE8192128_VEC_transpose_64x64(r_int_v[1], r_int_h[1]);
|
|
|
|
for (i = 0; i < 64; i++) {
|
|
r[i * 2 + 0] = r_int_v[0][i];
|
|
r[i * 2 + 1] = r_int_v[1][i];
|
|
}
|
|
}
|
|
|