//const u32x w0 = w0l | w0r;
- w0[0] = w0l | w0r;
- w0[1] = w[1];
- w0[2] = w[2];
- w0[3] = w[3];
- w1[0] = w[4];
- w1[1] = w[5];
- w1[2] = w[6];
- w1[3] = w[7];
- w2[0] = w[8];
- w2[1] = w[9];
- w2[2] = w[10];
- w2[3] = w[11];
- w3[0] = w[12];
- w3[1] = w[13];
- w3[2] = w[14];
- w3[3] = w[15];
+ w0[0] = swap32 (w0l | w0r);
+ w0[1] = swap32 (w[1]);
+ w0[2] = swap32 (w[2]);
+ w0[3] = swap32 (w[3]);
+ w1[0] = swap32 (w[4]);
+ w1[1] = swap32 (w[5]);
+ w1[2] = swap32 (w[6]);
+ w1[3] = swap32 (w[7]);
+ w2[0] = swap32 (w[8]);
+ w2[1] = swap32 (w[9]);
+ w2[2] = swap32 (w[10]);
+ w2[3] = swap32 (w[11]);
+ w3[0] = swap32 (w[12]);
+ w3[1] = swap32 (w[13]);
+ w3[2] = swap32 (w[14]);
+ w3[3] = swap32 (w[15]);
const u32 out_len = mangle(w0, w1, pw_len);
append_0x80_2x4_VV (w0, w1, out_len);
w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w9_t);
wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wa_t);
wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, wb_t);
-
- //if (MATCHES_NONE_VS (e, e_rev)) continue;
-
wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wc_t);
wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, wd_t);
we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, we_t);
wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wf_t);
- COMPARE_S_SIMD (d, e, c, b);
+ COMPARE_M_SIMD (d, e, c, b);
}
}