1
1
mirror of https://github.com/henrydcase/pqc.git synced 2024-11-24 00:11:27 +00:00
pqcrypto/crypto_kem/mceliece8192128/avx/vec128.c
Thom Wiggers b3f9d4f8d6
Classic McEliece (#259)
* 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
2020-02-05 13:09:56 +01:00

80 lines
1.9 KiB
C

#include "vec128.h"
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_set1_16b(uint16_t a) {
return _mm_set1_epi16(a);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_setzero(void) {
return _mm_setzero_si128();
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_and(vec128 a, vec128 b) {
return _mm_and_si128(a, b);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_xor(vec128 a, vec128 b) {
return _mm_xor_si128(a, b);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_or(vec128 a, vec128 b) {
return _mm_or_si128(a, b);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_sll_2x(vec128 a, int s) {
return _mm_slli_epi64(a, s);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_srl_2x(vec128 a, int s) {
return _mm_srli_epi64(a, s);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_set2x(uint64_t a0, uint64_t a1) {
return _mm_set_epi64x(a1, a0);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_unpack_low(vec128 a, vec128 b) {
return _mm_unpacklo_epi64(a, b);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_unpack_high(vec128 a, vec128 b) {
return _mm_unpackhi_epi64(a, b);
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_setbits(uint64_t a) {
return _mm_set1_epi64x(-a);
}
void PQCLEAN_MCELIECE8192128_AVX_vec128_copy(vec128 *dest, const vec128 *src) {
int i;
for (i = 0; i < GFBITS; i++) {
dest[i] = src[i];
}
}
void PQCLEAN_MCELIECE8192128_AVX_vec128_add(vec128 *c, const vec128 *a, const vec128 *b) {
int i;
for (i = 0; i < GFBITS; i++) {
c[i] = PQCLEAN_MCELIECE8192128_AVX_vec128_xor(a[i], b[i]);
}
}
vec128 PQCLEAN_MCELIECE8192128_AVX_vec128_or_reduce(const vec128 *a) {
int i;
vec128 ret;
ret = a[0];
for (i = 1; i < GFBITS; i++) {
ret = PQCLEAN_MCELIECE8192128_AVX_vec128_or(ret, a[i]);
}
return ret;
}
/* bitsliced field multiplications */
void PQCLEAN_MCELIECE8192128_AVX_vec128_mul(vec128 *h, vec128 *f, const vec128 *g) {
PQCLEAN_MCELIECE8192128_AVX_vec128_mul_asm(h, f, g, 16);
}