65 lines
1.8 KiB
C
65 lines
1.8 KiB
C
#include <stdint.h>
|
|
|
|
#define CAML_NAME_SPACE
|
|
#include <caml/mlvalues.h>
|
|
#include <caml/bigarray.h>
|
|
|
|
static inline uint32_t r(uint32_t a, int b) {
|
|
int rs = 32 - b;
|
|
return (a << b) | (a >> rs);
|
|
}
|
|
|
|
static inline uint32_t combine(uint32_t y0, uint32_t y1, uint32_t y2, int shift) {
|
|
return r(y1 + y2, shift) ^ y0;
|
|
}
|
|
|
|
static inline void quarterround(uint32_t *x, int y0, int y1, int y2, int y3) {
|
|
x[y1] = combine(x[y1], x[y0], x[y3], 7);
|
|
x[y2] = combine(x[y2], x[y1], x[y0], 9);
|
|
x[y3] = combine(x[y3], x[y2], x[y1], 13);
|
|
x[y0] = combine(x[y0], x[y3], x[y2], 18);
|
|
}
|
|
|
|
static inline uint32_t get_u32_le(const uint8_t *input, int offset) {
|
|
return input[offset]
|
|
| (input[offset + 1] << 8)
|
|
| (input[offset + 2] << 16)
|
|
| (input[offset + 3] << 24);
|
|
}
|
|
|
|
static inline void set_u32_le(uint8_t *input, int offset, uint32_t value) {
|
|
input[offset] = (uint8_t) value;
|
|
input[offset + 1] = (uint8_t) (value >> 8);
|
|
input[offset + 2] = (uint8_t) (value >> 16);
|
|
input[offset + 3] = (uint8_t) (value >> 24);
|
|
}
|
|
|
|
static void salsa_core(int count, const uint8_t *src, uint8_t *dst) {
|
|
uint32_t x[16];
|
|
for (int i = 0; i < 16; i++) {
|
|
x[i] = get_u32_le(src, i * 4);
|
|
}
|
|
for (int i = 0; i < count; i++) {
|
|
quarterround(x, 0, 4, 8, 12);
|
|
quarterround(x, 5, 9, 13, 1);
|
|
quarterround(x, 10, 14, 2, 6);
|
|
quarterround(x, 15, 3, 7, 11);
|
|
|
|
quarterround(x, 0, 1, 2, 3);
|
|
quarterround(x, 5, 6, 7, 4);
|
|
quarterround(x, 10, 11, 8, 9);
|
|
quarterround(x, 15, 12, 13, 14);
|
|
}
|
|
for (int i = 0; i < 16; i++) {
|
|
uint32_t xi = x[i];
|
|
uint32_t hj = get_u32_le(src, i * 4);
|
|
set_u32_le(dst, i * 4, xi + hj);
|
|
}
|
|
}
|
|
|
|
CAMLprim value
|
|
caml_salsa_core(value count, value src, value dst)
|
|
{
|
|
salsa_core(Int_val(count), (const uint8_t*)(String_val(src)), Bytes_val(dst));
|
|
return Val_unit;
|
|
}
|