/** * 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_sha256.cl" #define COMPARE_S "inc_comp_single.cl" #define COMPARE_M "inc_comp_multi.cl" void hmac_sha256_run_V (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x ipad[8], u32x opad[8], u32x digest[8]) { digest[0] = ipad[0]; digest[1] = ipad[1]; digest[2] = ipad[2]; digest[3] = ipad[3]; digest[4] = ipad[4]; digest[5] = ipad[5]; digest[6] = ipad[6]; digest[7] = ipad[7]; sha256_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] = digest[4]; w1[1] = digest[5]; w1[2] = digest[6]; w1[3] = digest[7]; w2[0] = 0x80000000; w2[1] = 0; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = (64 + 32) * 8; digest[0] = opad[0]; digest[1] = opad[1]; digest[2] = opad[2]; digest[3] = opad[3]; digest[4] = opad[4]; digest[5] = opad[5]; digest[6] = opad[6]; digest[7] = opad[7]; sha256_transform_vector (w0, w1, w2, w3, digest); } __kernel void m13000_init (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha256_tmp_t *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 pbkdf2_sha256_t *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) { /** * base */ const u64 gid = get_global_id (0); if (gid >= gid_max) return; sha256_hmac_ctx_t sha256_hmac_ctx; sha256_hmac_init_global_swap (&sha256_hmac_ctx, pws[gid].i, pws[gid].pw_len); tmps[gid].ipad[0] = sha256_hmac_ctx.ipad.h[0]; tmps[gid].ipad[1] = sha256_hmac_ctx.ipad.h[1]; tmps[gid].ipad[2] = sha256_hmac_ctx.ipad.h[2]; tmps[gid].ipad[3] = sha256_hmac_ctx.ipad.h[3]; tmps[gid].ipad[4] = sha256_hmac_ctx.ipad.h[4]; tmps[gid].ipad[5] = sha256_hmac_ctx.ipad.h[5]; tmps[gid].ipad[6] = sha256_hmac_ctx.ipad.h[6]; tmps[gid].ipad[7] = sha256_hmac_ctx.ipad.h[7]; tmps[gid].opad[0] = sha256_hmac_ctx.opad.h[0]; tmps[gid].opad[1] = sha256_hmac_ctx.opad.h[1]; tmps[gid].opad[2] = sha256_hmac_ctx.opad.h[2]; tmps[gid].opad[3] = sha256_hmac_ctx.opad.h[3]; tmps[gid].opad[4] = sha256_hmac_ctx.opad.h[4]; tmps[gid].opad[5] = sha256_hmac_ctx.opad.h[5]; tmps[gid].opad[6] = sha256_hmac_ctx.opad.h[6]; tmps[gid].opad[7] = sha256_hmac_ctx.opad.h[7]; sha256_hmac_update_global (&sha256_hmac_ctx, salt_bufs[digests_offset].salt_buf, salt_bufs[salt_pos].salt_len); for (u32 i = 0, j = 1; i < 8; i += 8, j += 1) { sha256_hmac_ctx_t sha256_hmac_ctx2 = sha256_hmac_ctx; u32 w0[4]; u32 w1[4]; u32 w2[4]; u32 w3[4]; w0[0] = j; w0[1] = 0; w0[2] = 0; w0[3] = 0; w1[0] = 0; 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] = 0; w3[3] = 0; sha256_hmac_update_64 (&sha256_hmac_ctx2, w0, w1, w2, w3, 4); sha256_hmac_final (&sha256_hmac_ctx2); tmps[gid].dgst[i + 0] = sha256_hmac_ctx2.opad.h[0]; tmps[gid].dgst[i + 1] = sha256_hmac_ctx2.opad.h[1]; tmps[gid].dgst[i + 2] = sha256_hmac_ctx2.opad.h[2]; tmps[gid].dgst[i + 3] = sha256_hmac_ctx2.opad.h[3]; tmps[gid].dgst[i + 4] = sha256_hmac_ctx2.opad.h[4]; tmps[gid].dgst[i + 5] = sha256_hmac_ctx2.opad.h[5]; tmps[gid].dgst[i + 6] = sha256_hmac_ctx2.opad.h[6]; tmps[gid].dgst[i + 7] = sha256_hmac_ctx2.opad.h[7]; tmps[gid].out[i + 0] = tmps[gid].dgst[i + 0]; tmps[gid].out[i + 1] = tmps[gid].dgst[i + 1]; tmps[gid].out[i + 2] = tmps[gid].dgst[i + 2]; tmps[gid].out[i + 3] = tmps[gid].dgst[i + 3]; tmps[gid].out[i + 4] = tmps[gid].dgst[i + 4]; tmps[gid].out[i + 5] = tmps[gid].dgst[i + 5]; tmps[gid].out[i + 6] = tmps[gid].dgst[i + 6]; tmps[gid].out[i + 7] = tmps[gid].dgst[i + 7]; } } __kernel void m13000_loop (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha256_tmp_t *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 pbkdf2_sha256_t *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) { const u64 gid = get_global_id (0); if ((gid * VECT_SIZE) >= gid_max) return; u32x ipad[8]; u32x opad[8]; ipad[0] = packv (tmps, ipad, gid, 0); ipad[1] = packv (tmps, ipad, gid, 1); ipad[2] = packv (tmps, ipad, gid, 2); ipad[3] = packv (tmps, ipad, gid, 3); ipad[4] = packv (tmps, ipad, gid, 4); ipad[5] = packv (tmps, ipad, gid, 5); ipad[6] = packv (tmps, ipad, gid, 6); ipad[7] = packv (tmps, ipad, gid, 7); opad[0] = packv (tmps, opad, gid, 0); opad[1] = packv (tmps, opad, gid, 1); opad[2] = packv (tmps, opad, gid, 2); opad[3] = packv (tmps, opad, gid, 3); opad[4] = packv (tmps, opad, gid, 4); opad[5] = packv (tmps, opad, gid, 5); opad[6] = packv (tmps, opad, gid, 6); opad[7] = packv (tmps, opad, gid, 7); for (u32 i = 0; i < 8; i += 8) { u32x dgst[8]; u32x out[8]; dgst[0] = packv (tmps, dgst, gid, i + 0); dgst[1] = packv (tmps, dgst, gid, i + 1); dgst[2] = packv (tmps, dgst, gid, i + 2); dgst[3] = packv (tmps, dgst, gid, i + 3); dgst[4] = packv (tmps, dgst, gid, i + 4); dgst[5] = packv (tmps, dgst, gid, i + 5); dgst[6] = packv (tmps, dgst, gid, i + 6); dgst[7] = packv (tmps, dgst, gid, i + 7); out[0] = packv (tmps, out, gid, i + 0); out[1] = packv (tmps, out, gid, i + 1); out[2] = packv (tmps, out, gid, i + 2); out[3] = packv (tmps, out, gid, i + 3); out[4] = packv (tmps, out, gid, i + 4); out[5] = packv (tmps, out, gid, i + 5); out[6] = packv (tmps, out, gid, i + 6); out[7] = packv (tmps, out, gid, i + 7); for (u32 j = 0; j < loop_cnt; j++) { u32x w0[4]; u32x w1[4]; u32x w2[4]; u32x w3[4]; w0[0] = dgst[0]; w0[1] = dgst[1]; w0[2] = dgst[2]; w0[3] = dgst[3]; w1[0] = dgst[4]; w1[1] = dgst[5]; w1[2] = dgst[6]; w1[3] = dgst[7]; w2[0] = 0x80000000; w2[1] = 0; w2[2] = 0; w2[3] = 0; w3[0] = 0; w3[1] = 0; w3[2] = 0; w3[3] = (64 + 32) * 8; hmac_sha256_run_V (w0, w1, w2, w3, ipad, opad, dgst); out[0] ^= dgst[0]; out[1] ^= dgst[1]; out[2] ^= dgst[2]; out[3] ^= dgst[3]; out[4] ^= dgst[4]; out[5] ^= dgst[5]; out[6] ^= dgst[6]; out[7] ^= dgst[7]; } unpackv (tmps, dgst, gid, i + 0, dgst[0]); unpackv (tmps, dgst, gid, i + 1, dgst[1]); unpackv (tmps, dgst, gid, i + 2, dgst[2]); unpackv (tmps, dgst, gid, i + 3, dgst[3]); unpackv (tmps, dgst, gid, i + 4, dgst[4]); unpackv (tmps, dgst, gid, i + 5, dgst[5]); unpackv (tmps, dgst, gid, i + 6, dgst[6]); unpackv (tmps, dgst, gid, i + 7, dgst[7]); unpackv (tmps, out, gid, i + 0, out[0]); unpackv (tmps, out, gid, i + 1, out[1]); unpackv (tmps, out, gid, i + 2, out[2]); unpackv (tmps, out, gid, i + 3, out[3]); unpackv (tmps, out, gid, i + 4, out[4]); unpackv (tmps, out, gid, i + 5, out[5]); unpackv (tmps, out, gid, i + 6, out[6]); unpackv (tmps, out, gid, i + 7, out[7]); } } __kernel void m13000_comp (__global pw_t *pws, __constant const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha256_tmp_t *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 pbkdf2_sha256_t *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) { /** * base */ const u64 gid = get_global_id (0); if (gid >= gid_max) return; const u64 lid = get_local_id (0); u32 outl4[4]; outl4[0] = tmps[gid].out[0]; outl4[1] = tmps[gid].out[2]; outl4[2] = tmps[gid].out[4]; outl4[3] = tmps[gid].out[6]; u32 outr4[4]; outr4[0] = tmps[gid].out[1]; outr4[1] = tmps[gid].out[3]; outr4[2] = tmps[gid].out[5]; outr4[3] = tmps[gid].out[7]; u32 outl = outl4[0] ^ outl4[1] ^ outl4[2] ^ outl4[3]; u32 outr = outr4[0] ^ outr4[1] ^ outr4[2] ^ outr4[3]; const u32 r0 = outl; const u32 r1 = outr; const u32 r2 = 0; const u32 r3 = 0; #define il_pos 0 #include COMPARE_M }