/** * Author......: See docs/credits.txt * License.....: MIT */ #define NEW_SIMD_CODE #include "inc_vendor.cl" #include "inc_hash_constants.h" #include "inc_hash_functions.cl" #include "inc_types.cl" #include "inc_common.cl" #include "inc_simd.cl" #include "inc_hash_md5.cl" void hmac_md5_pad (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[4], u32x opad[4]) { w0[0] = w0[0] ^ 0x36363636; w0[1] = w0[1] ^ 0x36363636; w0[2] = w0[2] ^ 0x36363636; w0[3] = w0[3] ^ 0x36363636; w1[0] = w1[0] ^ 0x36363636; w1[1] = w1[1] ^ 0x36363636; w1[2] = w1[2] ^ 0x36363636; w1[3] = w1[3] ^ 0x36363636; w2[0] = w2[0] ^ 0x36363636; w2[1] = w2[1] ^ 0x36363636; w2[2] = w2[2] ^ 0x36363636; w2[3] = w2[3] ^ 0x36363636; w3[0] = w3[0] ^ 0x36363636; w3[1] = w3[1] ^ 0x36363636; w3[2] = w3[2] ^ 0x36363636; w3[3] = w3[3] ^ 0x36363636; ipad[0] = MD5M_A; ipad[1] = MD5M_B; ipad[2] = MD5M_C; ipad[3] = MD5M_D; md5_transform_vector (w0, w1, w2, w3, ipad); w0[0] = w0[0] ^ 0x6a6a6a6a; w0[1] = w0[1] ^ 0x6a6a6a6a; w0[2] = w0[2] ^ 0x6a6a6a6a; w0[3] = w0[3] ^ 0x6a6a6a6a; w1[0] = w1[0] ^ 0x6a6a6a6a; w1[1] = w1[1] ^ 0x6a6a6a6a; w1[2] = w1[2] ^ 0x6a6a6a6a; w1[3] = w1[3] ^ 0x6a6a6a6a; w2[0] = w2[0] ^ 0x6a6a6a6a; w2[1] = w2[1] ^ 0x6a6a6a6a; w2[2] = w2[2] ^ 0x6a6a6a6a; w2[3] = w2[3] ^ 0x6a6a6a6a; w3[0] = w3[0] ^ 0x6a6a6a6a; w3[1] = w3[1] ^ 0x6a6a6a6a; w3[2] = w3[2] ^ 0x6a6a6a6a; w3[3] = w3[3] ^ 0x6a6a6a6a; opad[0] = MD5M_A; opad[1] = MD5M_B; opad[2] = MD5M_C; opad[3] = MD5M_D; md5_transform_vector (w0, w1, w2, w3, opad); } void hmac_md5_run (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[4], u32x opad[4], u32x digest[4]) { digest[0] = ipad[0]; digest[1] = ipad[1]; digest[2] = ipad[2]; digest[3] = ipad[3]; md5_transform_vector (w0, w1, w2, w3, digest); w0[0] = digest[0]; w0[1] = digest[1]; w0[2] = digest[2]; w0[3] = digest[3]; w1[0] = 0x80; w1[1] = 0; w1[2] = 0; w1[3] = 0; w2[0] = 0; w2[1] = 0; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = (64 + 16) * 8; w3[3] = 0; digest[0] = opad[0]; digest[1] = opad[1]; digest[2] = opad[2]; digest[3] = opad[3]; md5_transform_vector (w0, w1, w2, w3, digest); } __kernel void m00060_m04 (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u64 gid_max) { /** * 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; /** * salt */ u32 salt_buf0[4]; u32 salt_buf1[4]; u32 salt_buf2[4]; u32 salt_buf3[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] = salt_bufs[salt_pos].salt_buf[ 3]; salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4]; salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5]; salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6]; salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7]; salt_buf2[0] = salt_bufs[salt_pos].salt_buf[ 8]; salt_buf2[1] = salt_bufs[salt_pos].salt_buf[ 9]; salt_buf2[2] = salt_bufs[salt_pos].salt_buf[10]; salt_buf2[3] = salt_bufs[salt_pos].salt_buf[11]; salt_buf3[0] = salt_bufs[salt_pos].salt_buf[12]; salt_buf3[1] = salt_bufs[salt_pos].salt_buf[13]; salt_buf3[2] = salt_bufs[salt_pos].salt_buf[14]; salt_buf3[3] = salt_bufs[salt_pos].salt_buf[15]; const u32 salt_len = salt_bufs[salt_pos].salt_len; /** * pads */ u32x w0_t[4]; u32x w1_t[4]; u32x w2_t[4]; u32x w3_t[4]; w0_t[0] = salt_buf0[0]; w0_t[1] = salt_buf0[1]; w0_t[2] = salt_buf0[2]; w0_t[3] = salt_buf0[3]; w1_t[0] = salt_buf1[0]; w1_t[1] = salt_buf1[1]; w1_t[2] = salt_buf1[2]; w1_t[3] = salt_buf1[3]; w2_t[0] = salt_buf2[0]; w2_t[1] = salt_buf2[1]; w2_t[2] = salt_buf2[2]; w2_t[3] = salt_buf2[3]; w3_t[0] = salt_buf3[0]; w3_t[1] = salt_buf3[1]; w3_t[2] = salt_buf3[2]; w3_t[3] = salt_buf3[3]; u32x ipad[4]; u32x opad[4]; hmac_md5_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad); /** * 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); const u32x pw_len = pw_l_len + pw_r_len; /** * 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] = (64 + pw_len) * 8; w3[3] = 0; u32x digest[4]; hmac_md5_run (w0, w1, w2, w3, ipad, opad, digest); COMPARE_M_SIMD (digest[0], digest[3], digest[2], digest[1]); } } __kernel void m00060_m08 (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u64 gid_max) { } __kernel void m00060_m16 (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u64 gid_max) { } __kernel void m00060_s04 (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u64 gid_max) { /** * 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; /** * salt */ u32 salt_buf0[4]; u32 salt_buf1[4]; u32 salt_buf2[4]; u32 salt_buf3[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] = salt_bufs[salt_pos].salt_buf[ 3]; salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4]; salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5]; salt_buf1[2] = salt_bufs[salt_pos].salt_buf[ 6]; salt_buf1[3] = salt_bufs[salt_pos].salt_buf[ 7]; salt_buf2[0] = salt_bufs[salt_pos].salt_buf[ 8]; salt_buf2[1] = salt_bufs[salt_pos].salt_buf[ 9]; salt_buf2[2] = salt_bufs[salt_pos].salt_buf[10]; salt_buf2[3] = salt_bufs[salt_pos].salt_buf[11]; salt_buf3[0] = salt_bufs[salt_pos].salt_buf[12]; salt_buf3[1] = salt_bufs[salt_pos].salt_buf[13]; salt_buf3[2] = salt_bufs[salt_pos].salt_buf[14]; salt_buf3[3] = salt_bufs[salt_pos].salt_buf[15]; const u32 salt_len = salt_bufs[salt_pos].salt_len; /** * pads */ u32x w0_t[4]; u32x w1_t[4]; u32x w2_t[4]; u32x w3_t[4]; w0_t[0] = salt_buf0[0]; w0_t[1] = salt_buf0[1]; w0_t[2] = salt_buf0[2]; w0_t[3] = salt_buf0[3]; w1_t[0] = salt_buf1[0]; w1_t[1] = salt_buf1[1]; w1_t[2] = salt_buf1[2]; w1_t[3] = salt_buf1[3]; w2_t[0] = salt_buf2[0]; w2_t[1] = salt_buf2[1]; w2_t[2] = salt_buf2[2]; w2_t[3] = salt_buf2[3]; w3_t[0] = salt_buf3[0]; w3_t[1] = salt_buf3[1]; w3_t[2] = salt_buf3[2]; w3_t[3] = salt_buf3[3]; u32x ipad[4]; u32x opad[4]; hmac_md5_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad); /** * 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] }; /** * 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); const u32x pw_len = pw_l_len + pw_r_len; /** * 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] = (64 + pw_len) * 8; w3[3] = 0; u32x digest[4]; hmac_md5_run (w0, w1, w2, w3, ipad, opad, digest); COMPARE_S_SIMD (digest[0], digest[3], digest[2], digest[1]); } } __kernel void m00060_s08 (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u64 gid_max) { } __kernel void m00060_s16 (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u64 gid_max) { }