* Originally written by Mike Hamburg
*/
-#include "f_field.h"
+#include "field.h"
#if (defined(__OPTIMIZE__) && !defined(__OPTIMIZE_SIZE__) && !I_HATE_UNROLLED_LOOPS) \
- || defined(DECAF_FORCE_UNROLL)
+ || defined(C448_FORCE_UNROLL)
# define REPEAT8(_x) _x _x _x _x _x _x _x _x
# define FOR_LIMB(_i,_start,_end,_x) do { _i=_start; REPEAT8( if (_i<_end) { _x; } _i++;) } while (0)
#else
# define FOR_LIMB(_i,_start,_end,_x) do { for (_i=_start; _i<_end; _i++) _x; } while (0)
#endif
-void gf_mul(gf_s * __restrict__ cs, const gf as, const gf bs)
+void gf_mul(gf_s * RESTRICT cs, const gf as, const gf bs)
{
const uint32_t *a = as->limb, *b = bs->limb;
uint32_t *c = cs->limb;
-
uint64_t accum0 = 0, accum1 = 0, accum2 = 0;
uint32_t mask = (1 << 28) - 1;
-
uint32_t aa[8], bb[8];
-
int i, j;
+
for (i = 0; i < 8; i++) {
aa[i] = a[i] + a[i + 8];
bb[i] = b[i] + b[i + 8];
c[1] += ((uint32_t)(accum1));
}
-void gf_mulw_unsigned(gf_s * __restrict__ cs, const gf as, uint32_t b)
+void gf_mulw_unsigned(gf_s * RESTRICT cs, const gf as, uint32_t b)
{
const uint32_t *a = as->limb;
uint32_t *c = cs->limb;
});
accum0 += accum8 + c[8];
- c[8] = accum0 & mask;
- c[9] += accum0 >> 28;
+ c[8] = ((uint32_t)accum0) & mask;
+ c[9] += (uint32_t)(accum0 >> 28);
accum8 += c[0];
- c[0] = accum8 & mask;
- c[1] += accum8 >> 28;
+ c[0] = ((uint32_t)accum8) & mask;
+ c[1] += (uint32_t)(accum8 >> 28);
}
-void gf_sqr(gf_s * __restrict__ cs, const gf as)
+void gf_sqr(gf_s * RESTRICT cs, const gf as)
{
gf_mul(cs, as, as); /* Performs better with a dedicated square */
}