/** * Author......: See docs/credits.txt * License.....: MIT */ //#define NEW_SIMD_CODE #ifdef KERNEL_STATIC #include M2S(INCLUDE_PATH/inc_vendor.h) #include M2S(INCLUDE_PATH/inc_types.h) #include M2S(INCLUDE_PATH/inc_platform.cl) #include M2S(INCLUDE_PATH/inc_common.cl) #include M2S(INCLUDE_PATH/inc_simd.cl) #include M2S(INCLUDE_PATH/inc_hash_streebog256.cl) #endif KERNEL_FQ void m11700_mxx (KERN_ATTR_VECTOR ()) { /** * modifier */ const u64 gid = get_global_id (0); const u64 lid = get_local_id (0); const u64 lsz = get_local_size (0); /** * shared lookup table */ #ifdef REAL_SHM LOCAL_VK u64a s_sbob_sl64[8][256]; for (u32 i = lid; i < 256; i += lsz) { s_sbob_sl64[0][i] = sbob256_sl64[0][i]; s_sbob_sl64[1][i] = sbob256_sl64[1][i]; s_sbob_sl64[2][i] = sbob256_sl64[2][i]; s_sbob_sl64[3][i] = sbob256_sl64[3][i]; s_sbob_sl64[4][i] = sbob256_sl64[4][i]; s_sbob_sl64[5][i] = sbob256_sl64[5][i]; s_sbob_sl64[6][i] = sbob256_sl64[6][i]; s_sbob_sl64[7][i] = sbob256_sl64[7][i]; } SYNC_THREADS (); #else CONSTANT_AS u64a (*s_sbob_sl64)[256] = sbob256_sl64; #endif if (gid >= GID_CNT) return; /** * base */ const u32 pw_len = pws[gid].pw_len; u32x w[64] = { 0 }; for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1) { w[idx] = pws[gid].i[idx]; } /** * loop */ u32x w0l = w[0]; for (u32 il_pos = 0; il_pos < IL_CNT; il_pos += VECT_SIZE) { const u32x w0r = words_buf_r[il_pos / VECT_SIZE]; const u32x w0 = w0l | w0r; w[0] = w0; streebog256_ctx_vector_t ctx; streebog256_init_vector (&ctx, s_sbob_sl64); streebog256_update_vector_swap (&ctx, w, pw_len); streebog256_final_vector (&ctx); const u32x r0 = l32_from_64 (ctx.h[0]); const u32x r1 = h32_from_64 (ctx.h[0]); const u32x r2 = l32_from_64 (ctx.h[1]); const u32x r3 = h32_from_64 (ctx.h[1]); COMPARE_M_SIMD (r0, r1, r2, r3); } } KERNEL_FQ void m11700_sxx (KERN_ATTR_VECTOR ()) { /** * modifier */ const u64 gid = get_global_id (0); const u64 lid = get_local_id (0); const u64 lsz = get_local_size (0); /** * shared lookup table */ #ifdef REAL_SHM LOCAL_VK u64a s_sbob_sl64[8][256]; for (u32 i = lid; i < 256; i += lsz) { s_sbob_sl64[0][i] = sbob256_sl64[0][i]; s_sbob_sl64[1][i] = sbob256_sl64[1][i]; s_sbob_sl64[2][i] = sbob256_sl64[2][i]; s_sbob_sl64[3][i] = sbob256_sl64[3][i]; s_sbob_sl64[4][i] = sbob256_sl64[4][i]; s_sbob_sl64[5][i] = sbob256_sl64[5][i]; s_sbob_sl64[6][i] = sbob256_sl64[6][i]; s_sbob_sl64[7][i] = sbob256_sl64[7][i]; } SYNC_THREADS (); #else CONSTANT_AS u64a (*s_sbob_sl64)[256] = sbob256_sl64; #endif if (gid >= GID_CNT) return; /** * digest */ const u32 search[4] = { digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R0], digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R1], digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R2], digests_buf[DIGESTS_OFFSET_HOST].digest_buf[DGST_R3] }; /** * base */ const u32 pw_len = pws[gid].pw_len; u32x w[64] = { 0 }; for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1) { w[idx] = pws[gid].i[idx]; } /** * loop */ u32x w0l = w[0]; for (u32 il_pos = 0; il_pos < IL_CNT; il_pos += VECT_SIZE) { const u32x w0r = words_buf_r[il_pos / VECT_SIZE]; const u32x w0 = w0l | w0r; w[0] = w0; streebog256_ctx_vector_t ctx; streebog256_init_vector (&ctx, s_sbob_sl64); streebog256_update_vector_swap (&ctx, w, pw_len); streebog256_final_vector (&ctx); const u32x r0 = l32_from_64 (ctx.h[0]); const u32x r1 = h32_from_64 (ctx.h[0]); const u32x r2 = l32_from_64 (ctx.h[1]); const u32x r3 = h32_from_64 (ctx.h[1]); COMPARE_S_SIMD (r0, r1, r2, r3); } }