/** * Author......: See docs/credits.txt * License.....: MIT */ #define NEW_SIMD_CODE #include "inc_vendor.h" #include "inc_types.h" #include "inc_common.cl" #include "inc_simd.cl" #include "inc_hash_sha1.cl" __kernel void m07000_m04 (KERN_ATTR_BASIC ()) { /** * modifier */ const u64 lid = get_local_id (0); /** * base */ const u64 gid = get_global_id (0); if (gid >= gid_max) return; u32 pw_buf0[4]; u32 pw_buf1[4]; pw_buf0[0] = pws[gid].i[0]; pw_buf0[1] = pws[gid].i[1]; pw_buf0[2] = pws[gid].i[2]; pw_buf0[3] = pws[gid].i[3]; pw_buf1[0] = pws[gid].i[4]; pw_buf1[1] = pws[gid].i[5]; pw_buf1[2] = pws[gid].i[6]; pw_buf1[3] = pws[gid].i[7]; const u32 pw_l_len = pws[gid].pw_len & 63; /** * salt */ u32 salt_buf0[4]; salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0]; salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1]; salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2]; salt_buf0[3] = 0; const u32 salt_len = 12; // salt_bufs[salt_pos].salt_len; const u32 magic_len = 24; /** * loop */ for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE) { const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos) & 63; const u32x pw_len = (pw_l_len + pw_r_len) & 63; /** * concat password candidate */ u32x wordl0[4] = { 0 }; u32x wordl1[4] = { 0 }; u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; wordl0[0] = pw_buf0[0]; wordl0[1] = pw_buf0[1]; wordl0[2] = pw_buf0[2]; wordl0[3] = pw_buf0[3]; wordl1[0] = pw_buf1[0]; wordl1[1] = pw_buf1[1]; wordl1[2] = pw_buf1[2]; wordl1[3] = pw_buf1[3]; u32x wordr0[4] = { 0 }; u32x wordr1[4] = { 0 }; u32x wordr2[4] = { 0 }; u32x wordr3[4] = { 0 }; wordr0[0] = ix_create_combt (combs_buf, il_pos, 0); wordr0[1] = ix_create_combt (combs_buf, il_pos, 1); wordr0[2] = ix_create_combt (combs_buf, il_pos, 2); wordr0[3] = ix_create_combt (combs_buf, il_pos, 3); wordr1[0] = ix_create_combt (combs_buf, il_pos, 4); wordr1[1] = ix_create_combt (combs_buf, il_pos, 5); wordr1[2] = ix_create_combt (combs_buf, il_pos, 6); wordr1[3] = ix_create_combt (combs_buf, il_pos, 7); if (combs_mode == COMBINATOR_MODE_BASE_LEFT) { switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len); } else { switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len); } u32x w0[4]; u32x w1[4]; u32x w2[4]; u32x w3[4]; w0[0] = wordl0[0] | wordr0[0]; w0[1] = wordl0[1] | wordr0[1]; w0[2] = wordl0[2] | wordr0[2]; w0[3] = wordl0[3] | wordr0[3]; w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; w1[3] = wordl1[3] | wordr1[3]; w2[0] = wordl2[0] | wordr2[0]; w2[1] = wordl2[1] | wordr2[1]; w2[2] = wordl2[2] | wordr2[2]; w2[3] = wordl2[3] | wordr2[3]; w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; w3[2] = wordl3[2] | wordr3[2]; w3[3] = wordl3[3] | wordr3[3]; /** * combine pw + magic */ u32x s0[4]; u32x s1[4]; u32x s2[4]; u32x s3[4]; s0[0] = FORTIGATE_A; s0[1] = FORTIGATE_B; s0[2] = FORTIGATE_C; s0[3] = FORTIGATE_D; s1[0] = FORTIGATE_E; s1[1] = FORTIGATE_F; s1[2] = 0; s1[3] = 0; s2[0] = 0; s2[1] = 0; s2[2] = 0; s2[3] = 0; s3[0] = 0; s3[1] = 0; s3[2] = 0; s3[3] = 0; switch_buffer_by_offset_le_VV (s0, s1, s2, s3, pw_len); s0[0] |= w0[0]; s0[1] |= w0[1]; s0[2] |= w0[2]; s0[3] |= w0[3]; s1[0] |= w1[0]; s1[1] |= w1[1]; s1[2] |= w1[2]; s1[3] |= w1[3]; s2[0] |= w2[0]; s2[1] |= w2[1]; s2[2] |= w2[2]; s2[3] |= w2[3]; s3[0] |= w3[0]; s3[1] |= w3[1]; s3[2] |= w3[2]; s3[3] |= w3[3]; /** * prepend salt */ w0[0] = salt_buf0[0]; w0[1] = salt_buf0[1]; w0[2] = salt_buf0[2]; w0[3] = s0[0]; w1[0] = s0[1]; w1[1] = s0[2]; w1[2] = s0[3]; w1[3] = s1[0]; w2[0] = s1[1]; w2[1] = s1[2]; w2[2] = s1[3]; w2[3] = s2[0]; w3[0] = s2[1]; w3[1] = s2[2]; w3[2] = s2[3]; w3[3] = s3[0]; const u32x final_len = salt_len + pw_len + magic_len; append_0x80_4x4_VV (w0, w1, w2, w3, final_len); /** * sha1 */ u32x w0_t = swap32 (w0[0]); u32x w1_t = swap32 (w0[1]); u32x w2_t = swap32 (w0[2]); u32x w3_t = swap32 (w0[3]); u32x w4_t = swap32 (w1[0]); u32x w5_t = swap32 (w1[1]); u32x w6_t = swap32 (w1[2]); u32x w7_t = swap32 (w1[3]); u32x w8_t = swap32 (w2[0]); u32x w9_t = swap32 (w2[1]); u32x wa_t = swap32 (w2[2]); u32x wb_t = swap32 (w2[3]); u32x wc_t = swap32 (w3[0]); u32x wd_t = swap32 (w3[1]); u32x we_t = 0; u32x wf_t = final_len * 8; u32x a = SHA1M_A; u32x b = SHA1M_B; u32x c = SHA1M_C; u32x d = SHA1M_D; u32x e = SHA1M_E; #undef K #define K SHA1C00 SHA1_STEP (SHA1_F0o, a, b, c, d, e, w0_t); SHA1_STEP (SHA1_F0o, e, a, b, c, d, w1_t); SHA1_STEP (SHA1_F0o, d, e, a, b, c, w2_t); SHA1_STEP (SHA1_F0o, c, d, e, a, b, w3_t); SHA1_STEP (SHA1_F0o, b, c, d, e, a, w4_t); SHA1_STEP (SHA1_F0o, a, b, c, d, e, w5_t); SHA1_STEP (SHA1_F0o, e, a, b, c, d, w6_t); SHA1_STEP (SHA1_F0o, d, e, a, b, c, w7_t); SHA1_STEP (SHA1_F0o, c, d, e, a, b, w8_t); SHA1_STEP (SHA1_F0o, b, c, d, e, a, w9_t); SHA1_STEP (SHA1_F0o, a, b, c, d, e, wa_t); SHA1_STEP (SHA1_F0o, e, a, b, c, d, wb_t); SHA1_STEP (SHA1_F0o, d, e, a, b, c, wc_t); SHA1_STEP (SHA1_F0o, c, d, e, a, b, wd_t); SHA1_STEP (SHA1_F0o, b, c, d, e, a, we_t); SHA1_STEP (SHA1_F0o, a, b, c, d, e, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, e, a, b, c, d, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, d, e, a, b, c, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, c, d, e, a, b, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, b, c, d, e, a, w3_t); #undef K #define K SHA1C01 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w7_t); #undef K #define K SHA1C02 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, wb_t); #undef K #define K SHA1C03 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w8_t); 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); 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_M_SIMD (d, e, c, b); } } __kernel void m07000_m08 (KERN_ATTR_BASIC ()) { } __kernel void m07000_m16 (KERN_ATTR_BASIC ()) { } __kernel void m07000_s04 (KERN_ATTR_BASIC ()) { /** * modifier */ const u64 lid = get_local_id (0); /** * base */ const u64 gid = get_global_id (0); if (gid >= gid_max) return; u32 pw_buf0[4]; u32 pw_buf1[4]; pw_buf0[0] = pws[gid].i[0]; pw_buf0[1] = pws[gid].i[1]; pw_buf0[2] = pws[gid].i[2]; pw_buf0[3] = pws[gid].i[3]; pw_buf1[0] = pws[gid].i[4]; pw_buf1[1] = pws[gid].i[5]; pw_buf1[2] = pws[gid].i[6]; pw_buf1[3] = pws[gid].i[7]; const u32 pw_l_len = pws[gid].pw_len & 63; /** * salt */ u32 salt_buf0[4]; salt_buf0[0] = salt_bufs[salt_pos].salt_buf[0]; salt_buf0[1] = salt_bufs[salt_pos].salt_buf[1]; salt_buf0[2] = salt_bufs[salt_pos].salt_buf[2]; salt_buf0[3] = 0; const u32 salt_len = 12; // salt_bufs[salt_pos].salt_len; const u32 magic_len = 24; /** * digest */ const u32 search[4] = { digests_buf[digests_offset].digest_buf[DGST_R0], digests_buf[digests_offset].digest_buf[DGST_R1], digests_buf[digests_offset].digest_buf[DGST_R2], digests_buf[digests_offset].digest_buf[DGST_R3] }; /** * reverse */ const u32 e_rev = rotl32_S (search[1], 2u); /** * loop */ for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE) { const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos) & 63; const u32x pw_len = (pw_l_len + pw_r_len) & 63; /** * concat password candidate */ u32x wordl0[4] = { 0 }; u32x wordl1[4] = { 0 }; u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; wordl0[0] = pw_buf0[0]; wordl0[1] = pw_buf0[1]; wordl0[2] = pw_buf0[2]; wordl0[3] = pw_buf0[3]; wordl1[0] = pw_buf1[0]; wordl1[1] = pw_buf1[1]; wordl1[2] = pw_buf1[2]; wordl1[3] = pw_buf1[3]; u32x wordr0[4] = { 0 }; u32x wordr1[4] = { 0 }; u32x wordr2[4] = { 0 }; u32x wordr3[4] = { 0 }; wordr0[0] = ix_create_combt (combs_buf, il_pos, 0); wordr0[1] = ix_create_combt (combs_buf, il_pos, 1); wordr0[2] = ix_create_combt (combs_buf, il_pos, 2); wordr0[3] = ix_create_combt (combs_buf, il_pos, 3); wordr1[0] = ix_create_combt (combs_buf, il_pos, 4); wordr1[1] = ix_create_combt (combs_buf, il_pos, 5); wordr1[2] = ix_create_combt (combs_buf, il_pos, 6); wordr1[3] = ix_create_combt (combs_buf, il_pos, 7); if (combs_mode == COMBINATOR_MODE_BASE_LEFT) { switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len); } else { switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len); } u32x w0[4]; u32x w1[4]; u32x w2[4]; u32x w3[4]; w0[0] = wordl0[0] | wordr0[0]; w0[1] = wordl0[1] | wordr0[1]; w0[2] = wordl0[2] | wordr0[2]; w0[3] = wordl0[3] | wordr0[3]; w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; w1[3] = wordl1[3] | wordr1[3]; w2[0] = wordl2[0] | wordr2[0]; w2[1] = wordl2[1] | wordr2[1]; w2[2] = wordl2[2] | wordr2[2]; w2[3] = wordl2[3] | wordr2[3]; w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; w3[2] = wordl3[2] | wordr3[2]; w3[3] = wordl3[3] | wordr3[3]; /** * combine pw + magic */ u32x s0[4]; u32x s1[4]; u32x s2[4]; u32x s3[4]; s0[0] = FORTIGATE_A; s0[1] = FORTIGATE_B; s0[2] = FORTIGATE_C; s0[3] = FORTIGATE_D; s1[0] = FORTIGATE_E; s1[1] = FORTIGATE_F; s1[2] = 0; s1[3] = 0; s2[0] = 0; s2[1] = 0; s2[2] = 0; s2[3] = 0; s3[0] = 0; s3[1] = 0; s3[2] = 0; s3[3] = 0; switch_buffer_by_offset_le_VV (s0, s1, s2, s3, pw_len); s0[0] |= w0[0]; s0[1] |= w0[1]; s0[2] |= w0[2]; s0[3] |= w0[3]; s1[0] |= w1[0]; s1[1] |= w1[1]; s1[2] |= w1[2]; s1[3] |= w1[3]; s2[0] |= w2[0]; s2[1] |= w2[1]; s2[2] |= w2[2]; s2[3] |= w2[3]; s3[0] |= w3[0]; s3[1] |= w3[1]; s3[2] |= w3[2]; s3[3] |= w3[3]; /** * prepend salt */ w0[0] = salt_buf0[0]; w0[1] = salt_buf0[1]; w0[2] = salt_buf0[2]; w0[3] = s0[0]; w1[0] = s0[1]; w1[1] = s0[2]; w1[2] = s0[3]; w1[3] = s1[0]; w2[0] = s1[1]; w2[1] = s1[2]; w2[2] = s1[3]; w2[3] = s2[0]; w3[0] = s2[1]; w3[1] = s2[2]; w3[2] = s2[3]; w3[3] = s3[0]; const u32x final_len = salt_len + pw_len + magic_len; append_0x80_4x4_VV (w0, w1, w2, w3, final_len); /** * sha1 */ u32x w0_t = swap32 (w0[0]); u32x w1_t = swap32 (w0[1]); u32x w2_t = swap32 (w0[2]); u32x w3_t = swap32 (w0[3]); u32x w4_t = swap32 (w1[0]); u32x w5_t = swap32 (w1[1]); u32x w6_t = swap32 (w1[2]); u32x w7_t = swap32 (w1[3]); u32x w8_t = swap32 (w2[0]); u32x w9_t = swap32 (w2[1]); u32x wa_t = swap32 (w2[2]); u32x wb_t = swap32 (w2[3]); u32x wc_t = swap32 (w3[0]); u32x wd_t = swap32 (w3[1]); u32x we_t = 0; u32x wf_t = final_len * 8; u32x a = SHA1M_A; u32x b = SHA1M_B; u32x c = SHA1M_C; u32x d = SHA1M_D; u32x e = SHA1M_E; #undef K #define K SHA1C00 SHA1_STEP (SHA1_F0o, a, b, c, d, e, w0_t); SHA1_STEP (SHA1_F0o, e, a, b, c, d, w1_t); SHA1_STEP (SHA1_F0o, d, e, a, b, c, w2_t); SHA1_STEP (SHA1_F0o, c, d, e, a, b, w3_t); SHA1_STEP (SHA1_F0o, b, c, d, e, a, w4_t); SHA1_STEP (SHA1_F0o, a, b, c, d, e, w5_t); SHA1_STEP (SHA1_F0o, e, a, b, c, d, w6_t); SHA1_STEP (SHA1_F0o, d, e, a, b, c, w7_t); SHA1_STEP (SHA1_F0o, c, d, e, a, b, w8_t); SHA1_STEP (SHA1_F0o, b, c, d, e, a, w9_t); SHA1_STEP (SHA1_F0o, a, b, c, d, e, wa_t); SHA1_STEP (SHA1_F0o, e, a, b, c, d, wb_t); SHA1_STEP (SHA1_F0o, d, e, a, b, c, wc_t); SHA1_STEP (SHA1_F0o, c, d, e, a, b, wd_t); SHA1_STEP (SHA1_F0o, b, c, d, e, a, we_t); SHA1_STEP (SHA1_F0o, a, b, c, d, e, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F0o, e, a, b, c, d, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F0o, d, e, a, b, c, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F0o, c, d, e, a, b, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F0o, b, c, d, e, a, w3_t); #undef K #define K SHA1C01 w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w7_t); #undef K #define K SHA1C02 w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, wb_t); wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F2o, a, b, c, d, e, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F2o, e, a, b, c, d, w8_t); w9_t = rotl32 ((w6_t ^ w1_t ^ wb_t ^ w9_t), 1u); SHA1_STEP (SHA1_F2o, d, e, a, b, c, w9_t); wa_t = rotl32 ((w7_t ^ w2_t ^ wc_t ^ wa_t), 1u); SHA1_STEP (SHA1_F2o, c, d, e, a, b, wa_t); wb_t = rotl32 ((w8_t ^ w3_t ^ wd_t ^ wb_t), 1u); SHA1_STEP (SHA1_F2o, b, c, d, e, a, wb_t); #undef K #define K SHA1C03 wc_t = rotl32 ((w9_t ^ w4_t ^ we_t ^ wc_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, wc_t); wd_t = rotl32 ((wa_t ^ w5_t ^ wf_t ^ wd_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, wd_t); we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, wf_t); w0_t = rotl32 ((wd_t ^ w8_t ^ w2_t ^ w0_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w0_t); w1_t = rotl32 ((we_t ^ w9_t ^ w3_t ^ w1_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w1_t); w2_t = rotl32 ((wf_t ^ wa_t ^ w4_t ^ w2_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w2_t); w3_t = rotl32 ((w0_t ^ wb_t ^ w5_t ^ w3_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w3_t); w4_t = rotl32 ((w1_t ^ wc_t ^ w6_t ^ w4_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, w4_t); w5_t = rotl32 ((w2_t ^ wd_t ^ w7_t ^ w5_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, w5_t); w6_t = rotl32 ((w3_t ^ we_t ^ w8_t ^ w6_t), 1u); SHA1_STEP (SHA1_F1, a, b, c, d, e, w6_t); w7_t = rotl32 ((w4_t ^ wf_t ^ w9_t ^ w7_t), 1u); SHA1_STEP (SHA1_F1, e, a, b, c, d, w7_t); w8_t = rotl32 ((w5_t ^ w0_t ^ wa_t ^ w8_t), 1u); SHA1_STEP (SHA1_F1, d, e, a, b, c, w8_t); 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); } } __kernel void m07000_s08 (KERN_ATTR_BASIC ()) { } __kernel void m07000_s16 (KERN_ATTR_BASIC ()) { }