diff --git a/OpenCL/inc_ecc_secp256k1.cl b/OpenCL/inc_ecc_secp256k1.cl index b3a70df78..a487152ec 100644 --- a/OpenCL/inc_ecc_secp256k1.cl +++ b/OpenCL/inc_ecc_secp256k1.cl @@ -124,7 +124,9 @@ DECLSPEC u32 sub (u32 *r, const u32 *a, const u32 *b) : "r"(a[0]), "r"(a[1]), "r"(a[2]), "r"(a[3]), "r"(a[4]), "r"(a[5]), "r"(a[6]), "r"(a[7]), "r"(b[0]), "r"(b[1]), "r"(b[2]), "r"(b[3]), "r"(b[4]), "r"(b[5]), "r"(b[6]), "r"(b[7]) ); - #elif (defined IS_AMD || defined IS_HIP) && HAS_VSUB == 1 && HAS_VSUBB == 1 + // HIP doesnt support these so we stick to OpenCL (aka IS_AMD) - is also faster without asm + //#elif (defined IS_AMD || defined IS_HIP) && HAS_VSUB == 1 && HAS_VSUBB == 1 + #elif 0 __asm__ __volatile__ ( "V_SUB_U32 %0, %9, %17;" @@ -176,7 +178,9 @@ DECLSPEC u32 add (u32 *r, const u32 *a, const u32 *b) : "r"(a[0]), "r"(a[1]), "r"(a[2]), "r"(a[3]), "r"(a[4]), "r"(a[5]), "r"(a[6]), "r"(a[7]), "r"(b[0]), "r"(b[1]), "r"(b[2]), "r"(b[3]), "r"(b[4]), "r"(b[5]), "r"(b[6]), "r"(b[7]) ); - #elif (defined IS_AMD || defined IS_HIP) && HAS_VADD == 1 && HAS_VADDC == 1 + // HIP doesnt support these so we stick to OpenCL (aka IS_AMD) - is also faster without asm + //#elif (defined IS_AMD || defined IS_HIP) && HAS_VSUB == 1 && HAS_VSUBB == 1 + #elif 0 __asm__ __volatile__ ( "V_ADD_U32 %0, %9, %17;" diff --git a/OpenCL/m09000-pure.cl b/OpenCL/m09000-pure.cl index 737adde4e..323cf8387 100644 --- a/OpenCL/m09000-pure.cl +++ b/OpenCL/m09000-pure.cl @@ -310,6 +310,51 @@ CONSTANT_VK u32a c_pbox[18] = 0x9216d5d9, 0x8979fb1b }; +// Yes, works only with CUDA atm + +#ifdef DYNAMIC_LOCAL +#define BCRYPT_AVOID_BANK_CONFLICTS +#endif + +#ifdef BCRYPT_AVOID_BANK_CONFLICTS + +// access pattern: minimize bank ID based on thread ID but thread ID is not saved from computation + +#define KEY32(lid,key) (((key) * FIXED_LOCAL_SIZE) + (lid)) + +DECLSPEC u32 GET_KEY32 (LOCAL_AS u32 *S, const u64 key) +{ + const u64 lid = get_local_id (0); + + return S[KEY32 (lid, key)]; +} + +DECLSPEC void SET_KEY32 (LOCAL_AS u32 *S, const u64 key, const u32 val) +{ + const u64 lid = get_local_id (0); + + S[KEY32 (lid, key)] = val; +} + +#undef KEY32 + +#else + +// access pattern: linear access with S offset already set to right offset based on thread ID saving it from compuation +// makes sense if there are not thread ID's (for instance on CPU) + +DECLSPEC inline u32 GET_KEY32 (LOCAL_AS u32 *S, const u64 key) +{ + return S[key]; +} + +DECLSPEC inline void SET_KEY32 (LOCAL_AS u32 *S, const u64 key, const u32 val) +{ + S[key] = val; +} + +#endif + #define BF_ROUND(L,R,N) \ { \ u32 tmp; \ @@ -319,10 +364,10 @@ CONSTANT_VK u32a c_pbox[18] = const u32 r2 = unpack_v8b_from_v32_S ((L)); \ const u32 r3 = unpack_v8a_from_v32_S ((L)); \ \ - tmp = S0[r0]; \ - tmp += S1[r1]; \ - tmp ^= S2[r2]; \ - tmp += S3[r3]; \ + tmp = GET_KEY32 (S0, r0); \ + tmp += GET_KEY32 (S1, r1); \ + tmp ^= GET_KEY32 (S2, r2); \ + tmp += GET_KEY32 (S3, r3); \ \ (R) ^= tmp ^ P[(N)]; \ } @@ -357,6 +402,10 @@ CONSTANT_VK u32a c_pbox[18] = L ^= P[17]; \ } +#ifdef DYNAMIC_LOCAL +extern __shared__ u32 S[]; +#endif + KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m09000_init (KERN_ATTR_TMPS (pwsafe2_tmp_t)) { /** @@ -471,22 +520,33 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m09000_init (KERN_ATTR_TMPS P[i] = c_pbox[i]; } + #ifdef DYNAMIC_LOCAL + // from host + #else LOCAL_VK u32 S0_all[FIXED_LOCAL_SIZE][256]; LOCAL_VK u32 S1_all[FIXED_LOCAL_SIZE][256]; LOCAL_VK u32 S2_all[FIXED_LOCAL_SIZE][256]; LOCAL_VK u32 S3_all[FIXED_LOCAL_SIZE][256]; + #endif + #ifdef BCRYPT_AVOID_BANK_CONFLICTS + LOCAL_AS u32 *S0 = S + (FIXED_LOCAL_SIZE * 256 * 0); + LOCAL_AS u32 *S1 = S + (FIXED_LOCAL_SIZE * 256 * 1); + LOCAL_AS u32 *S2 = S + (FIXED_LOCAL_SIZE * 256 * 2); + LOCAL_AS u32 *S3 = S + (FIXED_LOCAL_SIZE * 256 * 3); + #else LOCAL_AS u32 *S0 = S0_all[lid]; LOCAL_AS u32 *S1 = S1_all[lid]; LOCAL_AS u32 *S2 = S2_all[lid]; LOCAL_AS u32 *S3 = S3_all[lid]; + #endif for (u32 i = 0; i < 256; i++) { - S0[i] = c_sbox0[i]; - S1[i] = c_sbox1[i]; - S2[i] = c_sbox2[i]; - S3[i] = c_sbox3[i]; + SET_KEY32 (S0, i, c_sbox0[i]); + SET_KEY32 (S1, i, c_sbox1[i]); + SET_KEY32 (S2, i, c_sbox2[i]); + SET_KEY32 (S3, i, c_sbox3[i]); } for (u32 i = 0; i < 18; i++) @@ -509,59 +569,59 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m09000_init (KERN_ATTR_TMPS { BF_ENCRYPT (L0, R0); - S0[i + 0] = L0; - S0[i + 1] = R0; + SET_KEY32 (S0, i + 0, L0); + SET_KEY32 (S0, i + 1, R0); BF_ENCRYPT (L0, R0); - S0[i + 2] = L0; - S0[i + 3] = R0; + SET_KEY32 (S0, i + 2, L0); + SET_KEY32 (S0, i + 3, R0); } for (u32 i = 0; i < 256; i += 4) { BF_ENCRYPT (L0, R0); - S1[i + 0] = L0; - S1[i + 1] = R0; + SET_KEY32 (S1, i + 0, L0); + SET_KEY32 (S1, i + 1, R0); BF_ENCRYPT (L0, R0); - S1[i + 2] = L0; - S1[i + 3] = R0; + SET_KEY32 (S1, i + 2, L0); + SET_KEY32 (S1, i + 3, R0); } for (u32 i = 0; i < 256; i += 4) { BF_ENCRYPT (L0, R0); - S2[i + 0] = L0; - S2[i + 1] = R0; + SET_KEY32 (S2, i + 0, L0); + SET_KEY32 (S2, i + 1, R0); BF_ENCRYPT (L0, R0); - S2[i + 2] = L0; - S2[i + 3] = R0; + SET_KEY32 (S2, i + 2, L0); + SET_KEY32 (S2, i + 3, R0); } for (u32 i = 0; i < 256; i += 4) { BF_ENCRYPT (L0, R0); - S3[i + 0] = L0; - S3[i + 1] = R0; + SET_KEY32 (S3, i + 0, L0); + SET_KEY32 (S3, i + 1, R0); BF_ENCRYPT (L0, R0); - S3[i + 2] = L0; - S3[i + 3] = R0; + SET_KEY32 (S3, i + 2, L0); + SET_KEY32 (S3, i + 3, R0); } - // store - tmps[gid].digest[0] = salt_buf[0]; tmps[gid].digest[1] = salt_buf[1]; + // store + for (u32 i = 0; i < 18; i++) { tmps[gid].P[i] = P[i]; @@ -569,10 +629,10 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m09000_init (KERN_ATTR_TMPS for (u32 i = 0; i < 256; i++) { - tmps[gid].S0[i] = S0[i]; - tmps[gid].S1[i] = S1[i]; - tmps[gid].S2[i] = S2[i]; - tmps[gid].S3[i] = S3[i]; + tmps[gid].S0[i] = GET_KEY32 (S0, i); + tmps[gid].S1[i] = GET_KEY32 (S1, i); + tmps[gid].S2[i] = GET_KEY32 (S2, i); + tmps[gid].S3[i] = GET_KEY32 (S3, i); } } @@ -602,22 +662,33 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m09000_loop (KERN_ATTR_TMPS P[i] = tmps[gid].P[i]; } + #ifdef DYNAMIC_LOCAL + // from host + #else LOCAL_VK u32 S0_all[FIXED_LOCAL_SIZE][256]; LOCAL_VK u32 S1_all[FIXED_LOCAL_SIZE][256]; LOCAL_VK u32 S2_all[FIXED_LOCAL_SIZE][256]; LOCAL_VK u32 S3_all[FIXED_LOCAL_SIZE][256]; + #endif + #ifdef BCRYPT_AVOID_BANK_CONFLICTS + LOCAL_AS u32 *S0 = S + (FIXED_LOCAL_SIZE * 256 * 0); + LOCAL_AS u32 *S1 = S + (FIXED_LOCAL_SIZE * 256 * 1); + LOCAL_AS u32 *S2 = S + (FIXED_LOCAL_SIZE * 256 * 2); + LOCAL_AS u32 *S3 = S + (FIXED_LOCAL_SIZE * 256 * 3); + #else LOCAL_AS u32 *S0 = S0_all[lid]; LOCAL_AS u32 *S1 = S1_all[lid]; LOCAL_AS u32 *S2 = S2_all[lid]; LOCAL_AS u32 *S3 = S3_all[lid]; + #endif for (u32 i = 0; i < 256; i++) { - S0[i] = tmps[gid].S0[i]; - S1[i] = tmps[gid].S1[i]; - S2[i] = tmps[gid].S2[i]; - S3[i] = tmps[gid].S3[i]; + SET_KEY32 (S0, i, tmps[gid].S0[i]); + SET_KEY32 (S1, i, tmps[gid].S1[i]); + SET_KEY32 (S2, i, tmps[gid].S2[i]); + SET_KEY32 (S3, i, tmps[gid].S3[i]); } // loop @@ -630,8 +701,6 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m09000_loop (KERN_ATTR_TMPS BF_ENCRYPT (L0, R0); } - // store - tmps[gid].digest[0] = L0; tmps[gid].digest[1] = R0; } diff --git a/OpenCL/m18600-pure.cl b/OpenCL/m18600-pure.cl index 061c61bd8..f98aca9c0 100644 --- a/OpenCL/m18600-pure.cl +++ b/OpenCL/m18600-pure.cl @@ -319,6 +319,51 @@ CONSTANT_VK u32a c_pbox[18] = 0x9216d5d9, 0x8979fb1b }; +// Yes, works only with CUDA atm + +#ifdef DYNAMIC_LOCAL +#define BCRYPT_AVOID_BANK_CONFLICTS +#endif + +#ifdef BCRYPT_AVOID_BANK_CONFLICTS + +// access pattern: minimize bank ID based on thread ID but thread ID is not saved from computation + +#define KEY32(lid,key) (((key) * FIXED_LOCAL_SIZE_COMP) + (lid)) + +DECLSPEC u32 GET_KEY32 (LOCAL_AS u32 *S, const u64 key) +{ + const u64 lid = get_local_id (0); + + return S[KEY32 (lid, key)]; +} + +DECLSPEC void SET_KEY32 (LOCAL_AS u32 *S, const u64 key, const u32 val) +{ + const u64 lid = get_local_id (0); + + S[KEY32 (lid, key)] = val; +} + +#undef KEY32 + +#else + +// access pattern: linear access with S offset already set to right offset based on thread ID saving it from compuation +// makes sense if there are not thread ID's (for instance on CPU) + +DECLSPEC inline u32 GET_KEY32 (LOCAL_AS u32 *S, const u64 key) +{ + return S[key]; +} + +DECLSPEC inline void SET_KEY32 (LOCAL_AS u32 *S, const u64 key, const u32 val) +{ + S[key] = val; +} + +#endif + #define BF_ROUND(L,R,N) \ { \ u32 tmp; \ @@ -328,10 +373,10 @@ CONSTANT_VK u32a c_pbox[18] = const u32 r2 = unpack_v8b_from_v32_S ((L)); \ const u32 r3 = unpack_v8a_from_v32_S ((L)); \ \ - tmp = S0[r0]; \ - tmp += S1[r1]; \ - tmp ^= S2[r2]; \ - tmp += S3[r3]; \ + tmp = GET_KEY32 (S0, r0); \ + tmp += GET_KEY32 (S1, r1); \ + tmp ^= GET_KEY32 (S2, r2); \ + tmp += GET_KEY32 (S3, r3); \ \ (R) ^= tmp ^ P[(N)]; \ } @@ -366,6 +411,10 @@ CONSTANT_VK u32a c_pbox[18] = L ^= P[17]; \ } +#ifdef DYNAMIC_LOCAL +extern __shared__ u32 S[]; +#endif + DECLSPEC void hmac_sha1_run_V (u32x *w0, u32x *w1, u32x *w2, u32x *w3, u32x *ipad, u32x *opad, u32x *digest) { digest[0] = ipad[0]; @@ -586,7 +635,7 @@ KERNEL_FQ void m18600_loop (KERN_ATTR_TMPS_ESALT (odf11_tmp_t, odf11_t)) } } -KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m18600_comp (KERN_ATTR_TMPS_ESALT (odf11_tmp_t, odf11_t)) +KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE_COMP) m18600_comp (KERN_ATTR_TMPS_ESALT (odf11_tmp_t, odf11_t)) { const u64 gid = get_global_id (0); const u64 lid = get_local_id (0); @@ -616,22 +665,33 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m18600_comp (KERN_ATTR_TMPS_ P[i] = c_pbox[i] ^ ukey[i % 4]; } - LOCAL_VK u32 S0_all[FIXED_LOCAL_SIZE][256]; - LOCAL_VK u32 S1_all[FIXED_LOCAL_SIZE][256]; - LOCAL_VK u32 S2_all[FIXED_LOCAL_SIZE][256]; - LOCAL_VK u32 S3_all[FIXED_LOCAL_SIZE][256]; + #ifdef DYNAMIC_LOCAL + // from host + #else + LOCAL_VK u32 S0_all[FIXED_LOCAL_SIZE_COMP][256]; + LOCAL_VK u32 S1_all[FIXED_LOCAL_SIZE_COMP][256]; + LOCAL_VK u32 S2_all[FIXED_LOCAL_SIZE_COMP][256]; + LOCAL_VK u32 S3_all[FIXED_LOCAL_SIZE_COMP][256]; + #endif + #ifdef BCRYPT_AVOID_BANK_CONFLICTS + LOCAL_AS u32 *S0 = S + (FIXED_LOCAL_SIZE_COMP * 256 * 0); + LOCAL_AS u32 *S1 = S + (FIXED_LOCAL_SIZE_COMP * 256 * 1); + LOCAL_AS u32 *S2 = S + (FIXED_LOCAL_SIZE_COMP * 256 * 2); + LOCAL_AS u32 *S3 = S + (FIXED_LOCAL_SIZE_COMP * 256 * 3); + #else LOCAL_AS u32 *S0 = S0_all[lid]; LOCAL_AS u32 *S1 = S1_all[lid]; LOCAL_AS u32 *S2 = S2_all[lid]; LOCAL_AS u32 *S3 = S3_all[lid]; + #endif for (u32 i = 0; i < 256; i++) { - S0[i] = c_sbox0[i]; - S1[i] = c_sbox1[i]; - S2[i] = c_sbox2[i]; - S3[i] = c_sbox3[i]; + SET_KEY32 (S0, i, c_sbox0[i]); + SET_KEY32 (S1, i, c_sbox1[i]); + SET_KEY32 (S2, i, c_sbox2[i]); + SET_KEY32 (S3, i, c_sbox3[i]); } u32 L0 = 0; @@ -649,52 +709,52 @@ KERNEL_FQ void FIXED_THREAD_COUNT(FIXED_LOCAL_SIZE) m18600_comp (KERN_ATTR_TMPS_ { BF_ENCRYPT (L0, R0); - S0[i + 0] = L0; - S0[i + 1] = R0; + SET_KEY32 (S0, i + 0, L0); + SET_KEY32 (S0, i + 1, R0); BF_ENCRYPT (L0, R0); - S0[i + 2] = L0; - S0[i + 3] = R0; + SET_KEY32 (S0, i + 2, L0); + SET_KEY32 (S0, i + 3, R0); } for (u32 i = 0; i < 256; i += 4) { BF_ENCRYPT (L0, R0); - S1[i + 0] = L0; - S1[i + 1] = R0; + SET_KEY32 (S1, i + 0, L0); + SET_KEY32 (S1, i + 1, R0); BF_ENCRYPT (L0, R0); - S1[i + 2] = L0; - S1[i + 3] = R0; + SET_KEY32 (S1, i + 2, L0); + SET_KEY32 (S1, i + 3, R0); } for (u32 i = 0; i < 256; i += 4) { BF_ENCRYPT (L0, R0); - S2[i + 0] = L0; - S2[i + 1] = R0; + SET_KEY32 (S2, i + 0, L0); + SET_KEY32 (S2, i + 1, R0); BF_ENCRYPT (L0, R0); - S2[i + 2] = L0; - S2[i + 3] = R0; + SET_KEY32 (S2, i + 2, L0); + SET_KEY32 (S2, i + 3, R0); } for (u32 i = 0; i < 256; i += 4) { BF_ENCRYPT (L0, R0); - S3[i + 0] = L0; - S3[i + 1] = R0; + SET_KEY32 (S3, i + 0, L0); + SET_KEY32 (S3, i + 1, R0); BF_ENCRYPT (L0, R0); - S3[i + 2] = L0; - S3[i + 3] = R0; + SET_KEY32 (S3, i + 2, L0); + SET_KEY32 (S3, i + 3, R0); } GLOBAL_AS const odf11_t *es = &esalt_bufs[DIGESTS_OFFSET]; diff --git a/OpenCL/m25000-pure.cl b/OpenCL/m25000-pure.cl new file mode 100644 index 000000000..249aa95fd --- /dev/null +++ b/OpenCL/m25000-pure.cl @@ -0,0 +1,590 @@ +/** + * Author......: See docs/credits.txt + * License.....: MIT + */ + +//#define NEW_SIMD_CODE + +#ifdef KERNEL_STATIC +#include "inc_vendor.h" +#include "inc_types.h" +#include "inc_platform.cl" +#include "inc_common.cl" +#include "inc_simd.cl" +#include "inc_hash_md5.cl" +#include "inc_hash_sha1.cl" +#endif + +#define COMPARE_S "inc_comp_single.cl" +#define COMPARE_M "inc_comp_multi.cl" + +#define SNMPV3_SALT_MAX 1500 +#define SNMPV3_ENGINEID_MAX 34 +#define SNMPV3_MSG_AUTH_PARAMS_LEN 12 +#define SNMPV3_ROUNDS 1048576 +#define SNMPV3_MAX_PW_LENGTH 64 + +#define SNMPV3_TMP_ELEMS 4096 // 4096 = (256 (max pw length) * 64) / sizeof (u32) +#define SNMPV3_HASH_ELEMS_MD5 4 +#define SNMPV3_HASH_ELEMS_SHA1 8 // 8 = aligned 5 + +#define SNMPV3_MAX_SALT_ELEMS 512 // 512 * 4 = 2048 > 1500, also has to be multiple of 64 +#define SNMPV3_MAX_ENGINE_ELEMS 16 // 16 * 4 = 64 > 32, also has to be multiple of 64 +#define SNMPV3_MAX_PNUM_ELEMS 4 // 4 * 4 = 16 > 9 + +typedef struct hmac_md5_tmp +{ + u32 tmp_md5[SNMPV3_TMP_ELEMS]; + u32 tmp_sha1[SNMPV3_TMP_ELEMS]; + + u32 h_md5[SNMPV3_HASH_ELEMS_MD5]; + u32 h_sha1[SNMPV3_HASH_ELEMS_SHA1]; + +} hmac_md5_tmp_t; + +typedef struct snmpv3 +{ + u32 salt_buf[SNMPV3_MAX_SALT_ELEMS]; + u32 salt_len; + + u32 engineID_buf[SNMPV3_MAX_ENGINE_ELEMS]; + u32 engineID_len; + + u32 packet_number[SNMPV3_MAX_PNUM_ELEMS]; + +} snmpv3_t; + +KERNEL_FQ void m25000_init (KERN_ATTR_TMPS_ESALT (hmac_md5_tmp_t, snmpv3_t)) +{ + /** + * modifier + */ + + const u64 gid = get_global_id (0); + + if (gid >= gid_max) return; + + /** + * base + */ + + const u32 pw_len = pws[gid].pw_len; + + u32 w[64] = { 0 }; + + for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1) + { + w[idx] = pws[gid].i[idx]; + } + + u8 *src_ptr = (u8 *) w; + + // password 64 times, also swapped + + u32 dst_buf[16]; + + u8 *dst_ptr = (u8 *) dst_buf; + + int tmp_idx = 0; + + for (int i = 0; i < 64; i++) + { + for (int j = 0; j < pw_len; j++) + { + const int dst_idx = tmp_idx & 63; + + dst_ptr[dst_idx] = src_ptr[j]; + + // write to global memory every time 64 byte are written into cache + + if (dst_idx == 63) + { + const int tmp_idx4 = (tmp_idx - 63) / 4; + + // md5 + + tmps[gid].tmp_md5[tmp_idx4 + 0] = dst_buf[ 0]; + tmps[gid].tmp_md5[tmp_idx4 + 1] = dst_buf[ 1]; + tmps[gid].tmp_md5[tmp_idx4 + 2] = dst_buf[ 2]; + tmps[gid].tmp_md5[tmp_idx4 + 3] = dst_buf[ 3]; + tmps[gid].tmp_md5[tmp_idx4 + 4] = dst_buf[ 4]; + tmps[gid].tmp_md5[tmp_idx4 + 5] = dst_buf[ 5]; + tmps[gid].tmp_md5[tmp_idx4 + 6] = dst_buf[ 6]; + tmps[gid].tmp_md5[tmp_idx4 + 7] = dst_buf[ 7]; + tmps[gid].tmp_md5[tmp_idx4 + 8] = dst_buf[ 8]; + tmps[gid].tmp_md5[tmp_idx4 + 9] = dst_buf[ 9]; + tmps[gid].tmp_md5[tmp_idx4 + 10] = dst_buf[10]; + tmps[gid].tmp_md5[tmp_idx4 + 11] = dst_buf[11]; + tmps[gid].tmp_md5[tmp_idx4 + 12] = dst_buf[12]; + tmps[gid].tmp_md5[tmp_idx4 + 13] = dst_buf[13]; + tmps[gid].tmp_md5[tmp_idx4 + 14] = dst_buf[14]; + tmps[gid].tmp_md5[tmp_idx4 + 15] = dst_buf[15]; + + // sha1 + + tmps[gid].tmp_sha1[tmp_idx4 + 0] = hc_swap32_S (dst_buf[ 0]); + tmps[gid].tmp_sha1[tmp_idx4 + 1] = hc_swap32_S (dst_buf[ 1]); + tmps[gid].tmp_sha1[tmp_idx4 + 2] = hc_swap32_S (dst_buf[ 2]); + tmps[gid].tmp_sha1[tmp_idx4 + 3] = hc_swap32_S (dst_buf[ 3]); + tmps[gid].tmp_sha1[tmp_idx4 + 4] = hc_swap32_S (dst_buf[ 4]); + tmps[gid].tmp_sha1[tmp_idx4 + 5] = hc_swap32_S (dst_buf[ 5]); + tmps[gid].tmp_sha1[tmp_idx4 + 6] = hc_swap32_S (dst_buf[ 6]); + tmps[gid].tmp_sha1[tmp_idx4 + 7] = hc_swap32_S (dst_buf[ 7]); + tmps[gid].tmp_sha1[tmp_idx4 + 8] = hc_swap32_S (dst_buf[ 8]); + tmps[gid].tmp_sha1[tmp_idx4 + 9] = hc_swap32_S (dst_buf[ 9]); + tmps[gid].tmp_sha1[tmp_idx4 + 10] = hc_swap32_S (dst_buf[10]); + tmps[gid].tmp_sha1[tmp_idx4 + 11] = hc_swap32_S (dst_buf[11]); + tmps[gid].tmp_sha1[tmp_idx4 + 12] = hc_swap32_S (dst_buf[12]); + tmps[gid].tmp_sha1[tmp_idx4 + 13] = hc_swap32_S (dst_buf[13]); + tmps[gid].tmp_sha1[tmp_idx4 + 14] = hc_swap32_S (dst_buf[14]); + tmps[gid].tmp_sha1[tmp_idx4 + 15] = hc_swap32_S (dst_buf[15]); + } + + tmp_idx++; + } + } + + // hash md5 + + tmps[gid].h_md5[0] = MD5M_A; + tmps[gid].h_md5[1] = MD5M_B; + tmps[gid].h_md5[2] = MD5M_C; + tmps[gid].h_md5[3] = MD5M_D; + + // hash sha1 + + tmps[gid].h_sha1[0] = SHA1M_A; + tmps[gid].h_sha1[1] = SHA1M_B; + tmps[gid].h_sha1[2] = SHA1M_C; + tmps[gid].h_sha1[3] = SHA1M_D; + tmps[gid].h_sha1[4] = SHA1M_E; +} + +KERNEL_FQ void m25000_loop (KERN_ATTR_TMPS_ESALT (hmac_md5_tmp_t, snmpv3_t)) +{ + /** + * base + */ + + const u64 gid = get_global_id (0); + + if (gid >= gid_max) return; + + u32 h_md5[4]; + + h_md5[0] = tmps[gid].h_md5[0]; + h_md5[1] = tmps[gid].h_md5[1]; + h_md5[2] = tmps[gid].h_md5[2]; + h_md5[3] = tmps[gid].h_md5[3]; + + u32 h_sha1[5]; + + h_sha1[0] = tmps[gid].h_sha1[0]; + h_sha1[1] = tmps[gid].h_sha1[1]; + h_sha1[2] = tmps[gid].h_sha1[2]; + h_sha1[3] = tmps[gid].h_sha1[3]; + h_sha1[4] = tmps[gid].h_sha1[4]; + + const u32 pw_len = pws[gid].pw_len; + + const int pw_len64 = pw_len * 64; + + #define SNMPV3_TMP_ELEMS_OPT 1024 // 1024 = (64 max pw length * 64) / sizeof (u32) + // for pw length > 64 we use global memory reads + + if (pw_len < 64) + { + u32 tmp_shared[SNMPV3_TMP_ELEMS_OPT]; + + // md5 + + for (int i = 0; i < pw_len64 / 4; i++) + { + tmp_shared[i] = tmps[gid].tmp_md5[i]; + } + + for (int i = 0, j = loop_pos; i < loop_cnt; i += 64, j += 64) + { + const int idx = (j % pw_len64) / 4; // the optimization trick is to be able to do this + + u32 w0[4]; + u32 w1[4]; + u32 w2[4]; + u32 w3[4]; + + w0[0] = tmp_shared[idx + 0]; + w0[1] = tmp_shared[idx + 1]; + w0[2] = tmp_shared[idx + 2]; + w0[3] = tmp_shared[idx + 3]; + w1[0] = tmp_shared[idx + 4]; + w1[1] = tmp_shared[idx + 5]; + w1[2] = tmp_shared[idx + 6]; + w1[3] = tmp_shared[idx + 7]; + w2[0] = tmp_shared[idx + 8]; + w2[1] = tmp_shared[idx + 9]; + w2[2] = tmp_shared[idx + 10]; + w2[3] = tmp_shared[idx + 11]; + w3[0] = tmp_shared[idx + 12]; + w3[1] = tmp_shared[idx + 13]; + w3[2] = tmp_shared[idx + 14]; + w3[3] = tmp_shared[idx + 15]; + + md5_transform (w0, w1, w2, w3, h_md5); + } + + // sha1 + + for (int i = 0; i < pw_len64 / 4; i++) + { + tmp_shared[i] = tmps[gid].tmp_sha1[i]; + } + + for (int i = 0, j = loop_pos; i < loop_cnt; i += 64, j += 64) + { + const int idx = (j % pw_len64) / 4; // the optimization trick is to be able to do this + + u32 w0[4]; + u32 w1[4]; + u32 w2[4]; + u32 w3[4]; + + w0[0] = tmp_shared[idx + 0]; + w0[1] = tmp_shared[idx + 1]; + w0[2] = tmp_shared[idx + 2]; + w0[3] = tmp_shared[idx + 3]; + w1[0] = tmp_shared[idx + 4]; + w1[1] = tmp_shared[idx + 5]; + w1[2] = tmp_shared[idx + 6]; + w1[3] = tmp_shared[idx + 7]; + w2[0] = tmp_shared[idx + 8]; + w2[1] = tmp_shared[idx + 9]; + w2[2] = tmp_shared[idx + 10]; + w2[3] = tmp_shared[idx + 11]; + w3[0] = tmp_shared[idx + 12]; + w3[1] = tmp_shared[idx + 13]; + w3[2] = tmp_shared[idx + 14]; + w3[3] = tmp_shared[idx + 15]; + + sha1_transform (w0, w1, w2, w3, h_sha1); + } + } + else + { + for (int i = 0, j = loop_pos; i < loop_cnt; i += 64, j += 64) + { + const int idx = (j % pw_len64) / 4; // the optimization trick is to be able to do this + + u32 w0[4]; + u32 w1[4]; + u32 w2[4]; + u32 w3[4]; + + // md5 + + w0[0] = tmps[gid].tmp_md5[idx + 0]; + w0[1] = tmps[gid].tmp_md5[idx + 1]; + w0[2] = tmps[gid].tmp_md5[idx + 2]; + w0[3] = tmps[gid].tmp_md5[idx + 3]; + w1[0] = tmps[gid].tmp_md5[idx + 4]; + w1[1] = tmps[gid].tmp_md5[idx + 5]; + w1[2] = tmps[gid].tmp_md5[idx + 6]; + w1[3] = tmps[gid].tmp_md5[idx + 7]; + w2[0] = tmps[gid].tmp_md5[idx + 8]; + w2[1] = tmps[gid].tmp_md5[idx + 9]; + w2[2] = tmps[gid].tmp_md5[idx + 10]; + w2[3] = tmps[gid].tmp_md5[idx + 11]; + w3[0] = tmps[gid].tmp_md5[idx + 12]; + w3[1] = tmps[gid].tmp_md5[idx + 13]; + w3[2] = tmps[gid].tmp_md5[idx + 14]; + w3[3] = tmps[gid].tmp_md5[idx + 15]; + + md5_transform (w0, w1, w2, w3, h_md5); + + // sha1 + + w0[0] = tmps[gid].tmp_sha1[idx + 0]; + w0[1] = tmps[gid].tmp_sha1[idx + 1]; + w0[2] = tmps[gid].tmp_sha1[idx + 2]; + w0[3] = tmps[gid].tmp_sha1[idx + 3]; + w1[0] = tmps[gid].tmp_sha1[idx + 4]; + w1[1] = tmps[gid].tmp_sha1[idx + 5]; + w1[2] = tmps[gid].tmp_sha1[idx + 6]; + w1[3] = tmps[gid].tmp_sha1[idx + 7]; + w2[0] = tmps[gid].tmp_sha1[idx + 8]; + w2[1] = tmps[gid].tmp_sha1[idx + 9]; + w2[2] = tmps[gid].tmp_sha1[idx + 10]; + w2[3] = tmps[gid].tmp_sha1[idx + 11]; + w3[0] = tmps[gid].tmp_sha1[idx + 12]; + w3[1] = tmps[gid].tmp_sha1[idx + 13]; + w3[2] = tmps[gid].tmp_sha1[idx + 14]; + w3[3] = tmps[gid].tmp_sha1[idx + 15]; + + sha1_transform (w0, w1, w2, w3, h_sha1); + } + } + + tmps[gid].h_md5[0] = h_md5[0]; + tmps[gid].h_md5[1] = h_md5[1]; + tmps[gid].h_md5[2] = h_md5[2]; + tmps[gid].h_md5[3] = h_md5[3]; + + tmps[gid].h_sha1[0] = h_sha1[0]; + tmps[gid].h_sha1[1] = h_sha1[1]; + tmps[gid].h_sha1[2] = h_sha1[2]; + tmps[gid].h_sha1[3] = h_sha1[3]; + tmps[gid].h_sha1[4] = h_sha1[4]; +} + +KERNEL_FQ void m25000_comp (KERN_ATTR_TMPS_ESALT (hmac_md5_tmp_t, snmpv3_t)) +{ + /** + * modifier + */ + + const u64 gid = get_global_id (0); + + if (gid >= gid_max) return; + + u32 w0[4]; + u32 w1[4]; + u32 w2[4]; + u32 w3[4]; + + // md5 + + w0[0] = 0x00000080; + 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] = 1048576 * 8; + w3[3] = 0; + + u32 h_md5[4]; + + h_md5[0] = tmps[gid].h_md5[0]; + h_md5[1] = tmps[gid].h_md5[1]; + h_md5[2] = tmps[gid].h_md5[2]; + h_md5[3] = tmps[gid].h_md5[3]; + + md5_transform (w0, w1, w2, w3, h_md5); + + // sha1 + + w0[0] = 0x80000000; + 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] = 1048576 * 8; + + u32 h_sha1[5]; + + h_sha1[0] = tmps[gid].h_sha1[0]; + h_sha1[1] = tmps[gid].h_sha1[1]; + h_sha1[2] = tmps[gid].h_sha1[2]; + h_sha1[3] = tmps[gid].h_sha1[3]; + h_sha1[4] = tmps[gid].h_sha1[4]; + + sha1_transform (w0, w1, w2, w3, h_sha1); + + md5_ctx_t md5_ctx; + sha1_ctx_t sha1_ctx; + + md5_init (&md5_ctx); + sha1_init (&sha1_ctx); + + u32 w[16]; + + // md5 + + w[ 0] = h_md5[0]; + w[ 1] = h_md5[1]; + w[ 2] = h_md5[2]; + w[ 3] = h_md5[3]; + w[ 4] = 0; + w[ 5] = 0; + w[ 6] = 0; + w[ 7] = 0; + w[ 8] = 0; + w[ 9] = 0; + w[10] = 0; + w[11] = 0; + w[12] = 0; + w[13] = 0; + w[14] = 0; + w[15] = 0; + + md5_update (&md5_ctx, w, 16); + + // sha1 + + w[ 0] = h_sha1[0]; + w[ 1] = h_sha1[1]; + w[ 2] = h_sha1[2]; + w[ 3] = h_sha1[3]; + w[ 4] = h_sha1[4]; + w[ 5] = 0; + w[ 6] = 0; + w[ 7] = 0; + w[ 8] = 0; + w[ 9] = 0; + w[10] = 0; + w[11] = 0; + w[12] = 0; + w[13] = 0; + w[14] = 0; + w[15] = 0; + + sha1_update (&sha1_ctx, w, 20); + + // engineID + + md5_update_global (&md5_ctx, esalt_bufs[DIGESTS_OFFSET].engineID_buf, esalt_bufs[DIGESTS_OFFSET].engineID_len); + + sha1_update_global_swap (&sha1_ctx, esalt_bufs[DIGESTS_OFFSET].engineID_buf, esalt_bufs[DIGESTS_OFFSET].engineID_len); + + // md5 + + w[ 0] = h_md5[0]; + w[ 1] = h_md5[1]; + w[ 2] = h_md5[2]; + w[ 3] = h_md5[3]; + w[ 4] = 0; + w[ 5] = 0; + w[ 6] = 0; + w[ 7] = 0; + w[ 8] = 0; + w[ 9] = 0; + w[10] = 0; + w[11] = 0; + w[12] = 0; + w[13] = 0; + w[14] = 0; + w[15] = 0; + + md5_update (&md5_ctx, w, 16); + + // sha1 + + w[ 0] = h_sha1[0]; + w[ 1] = h_sha1[1]; + w[ 2] = h_sha1[2]; + w[ 3] = h_sha1[3]; + w[ 4] = h_sha1[4]; + w[ 5] = 0; + w[ 6] = 0; + w[ 7] = 0; + w[ 8] = 0; + w[ 9] = 0; + w[10] = 0; + w[11] = 0; + w[12] = 0; + w[13] = 0; + w[14] = 0; + w[15] = 0; + + sha1_update (&sha1_ctx, w, 20); + + md5_final (&md5_ctx); + sha1_final (&sha1_ctx); + + // md5 + + w[ 0] = md5_ctx.h[0]; + w[ 1] = md5_ctx.h[1]; + w[ 2] = md5_ctx.h[2]; + w[ 3] = md5_ctx.h[3]; + w[ 4] = 0; + w[ 5] = 0; + w[ 6] = 0; + w[ 7] = 0; + w[ 8] = 0; + w[ 9] = 0; + w[10] = 0; + w[11] = 0; + w[12] = 0; + w[13] = 0; + w[14] = 0; + w[15] = 0; + + md5_hmac_ctx_t md5_hmac_ctx; + + md5_hmac_init (&md5_hmac_ctx, w, 16); + + md5_hmac_update_global (&md5_hmac_ctx, esalt_bufs[DIGESTS_OFFSET].salt_buf, esalt_bufs[DIGESTS_OFFSET].salt_len); + + md5_hmac_final (&md5_hmac_ctx); + + { + const u32 r0 = hc_swap32_S (md5_hmac_ctx.opad.h[DGST_R0]); + const u32 r1 = hc_swap32_S (md5_hmac_ctx.opad.h[DGST_R1]); + const u32 r2 = hc_swap32_S (md5_hmac_ctx.opad.h[DGST_R2]); + const u32 r3 = 0; + + #define il_pos 0 + + #ifdef KERNEL_STATIC + #include COMPARE_M + #endif + } + + // sha1 + + w[ 0] = sha1_ctx.h[0]; + w[ 1] = sha1_ctx.h[1]; + w[ 2] = sha1_ctx.h[2]; + w[ 3] = sha1_ctx.h[3]; + w[ 4] = sha1_ctx.h[4]; + w[ 5] = 0; + w[ 6] = 0; + w[ 7] = 0; + w[ 8] = 0; + w[ 9] = 0; + w[10] = 0; + w[11] = 0; + w[12] = 0; + w[13] = 0; + w[14] = 0; + w[15] = 0; + + sha1_hmac_ctx_t sha1_hmac_ctx; + + sha1_hmac_init (&sha1_hmac_ctx, w, 20); + + sha1_hmac_update_global_swap (&sha1_hmac_ctx, esalt_bufs[DIGESTS_OFFSET].salt_buf, esalt_bufs[DIGESTS_OFFSET].salt_len); + + sha1_hmac_final (&sha1_hmac_ctx); + + { + const u32 r0 = sha1_hmac_ctx.opad.h[DGST_R0]; + const u32 r1 = sha1_hmac_ctx.opad.h[DGST_R1]; + const u32 r2 = sha1_hmac_ctx.opad.h[DGST_R2]; + const u32 r3 = 0; + + #define il_pos 0 + + #ifdef KERNEL_STATIC + #include COMPARE_M + #endif + } +} diff --git a/OpenCL/m25100-pure.cl b/OpenCL/m25100-pure.cl index 72cb0ba47..1fc28c664 100644 --- a/OpenCL/m25100-pure.cl +++ b/OpenCL/m25100-pure.cl @@ -18,8 +18,8 @@ #define COMPARE_M "inc_comp_multi.cl" #define SNMPV3_SALT_MAX 1500 -#define SNMPV3_ENGINEID_MAX 32 -#define SNMPV3_MSG_AUTH_PARAMS_MAX 12 +#define SNMPV3_ENGINEID_MAX 34 +#define SNMPV3_MSG_AUTH_PARAMS_LEN 12 #define SNMPV3_ROUNDS 1048576 #define SNMPV3_MAX_PW_LENGTH 64 diff --git a/OpenCL/m25200-pure.cl b/OpenCL/m25200-pure.cl index f72fce044..e36caaf6e 100644 --- a/OpenCL/m25200-pure.cl +++ b/OpenCL/m25200-pure.cl @@ -18,13 +18,17 @@ #define COMPARE_M "inc_comp_multi.cl" #define SNMPV3_SALT_MAX 1500 -#define SNMPV3_ENGINEID_MAX 32 -#define SNMPV3_MSG_AUTH_PARAMS_MAX 12 +#define SNMPV3_ENGINEID_MAX 34 +#define SNMPV3_MSG_AUTH_PARAMS_LEN 12 #define SNMPV3_ROUNDS 1048576 #define SNMPV3_MAX_PW_LENGTH 64 -#define SNMPV3_TMP_ELEMS 4096 // 4096 = (256 (max pw length) * 64) / sizeof (u32) -#define SNMPV3_HASH_ELEMS 8 // 8 = aligned 5 +#define SNMPV3_TMP_ELEMS 4096 // 4096 = (256 (max pw length) * 64) / sizeof (u32) +#define SNMPV3_HASH_ELEMS 8 // 8 = aligned 5 + +#define SNMPV3_MAX_SALT_ELEMS 512 // 512 * 4 = 2048 > 1500, also has to be multiple of 64 +#define SNMPV3_MAX_ENGINE_ELEMS 16 // 16 * 4 = 64 > 32, also has to be multiple of 64 +#define SNMPV3_MAX_PNUM_ELEMS 4 // 4 * 4 = 16 > 9 typedef struct hmac_sha1_tmp { @@ -33,10 +37,6 @@ typedef struct hmac_sha1_tmp } hmac_sha1_tmp_t; -#define SNMPV3_MAX_SALT_ELEMS 512 // 512 * 4 = 2048 > 1500, also has to be multiple of 64 -#define SNMPV3_MAX_ENGINE_ELEMS 16 // 16 * 4 = 64 > 32, also has to be multiple of 64 -#define SNMPV3_MAX_PNUM_ELEMS 4 // 4 * 4 = 16 > 9 - typedef struct snmpv3 { u32 salt_buf[SNMPV3_MAX_SALT_ELEMS]; diff --git a/docs/changes.txt b/docs/changes.txt index 51353f661..0ca93c599 100644 --- a/docs/changes.txt +++ b/docs/changes.txt @@ -21,6 +21,7 @@ - AMD GPUs: Add inline assembly code for md5crypt/sha256crypt, PDF 1.7, 7-Zip, RAR3, Samsung Android and Windows Phone 8+ - Blake Kernels: Optimize BLAKE2B_ROUND() 64 bit rotates giving a 5% performance increase +- Blowfish Kernels: Backport optimizations reducing bank conflicts from bcrypt to Password Safe v2 and Open Document Format (ODF) 1.1 - Brain Session: Adds hashconfig specific opti_type and opts_type parameters to hashcat session computation to cover features like -O and -M - Kernel Threads: Use warp size / wavefront size query instead of hardcoded values as base for kernel threads - Shared Memory: Calculate kernel dynamic memory size based on CU_DEVICE_ATTRIBUTE_MAX_SHARED_MEMORY_PER_BLOCK_OPTIN @@ -32,8 +33,8 @@ - ADL: Updated support for AMD Display Library to 14.0, updated datatypes and added support for OverDrive 7 and 8 based GPUs - Commandline: Throw an error if separator character given by the user with -p option is not exactly 1 byte -- Kernel Cache: Add kernel threads into hash computation which is later used in the kernel cache filename - HIP Kernels: Got rid of hip/hip_runtime.h dependancy to enable more easy integration of the HIP backend on Windows +- Kernel Cache: Add kernel threads into hash computation which is later used in the kernel cache filename - SCRYPT Kernels: Add more optimized values for some new NV/AMD GPUs ## @@ -42,6 +43,7 @@ - Added hash-mode: SNMPv3 HMAC-SHA1-96 - Added hash-mode: SNMPv3 HMAC-MD5-96 +- Added hash-mode: SNMPv3 HMAC-MD5-96/HMAC-SHA1-96 * changes v6.2.2 -> v6.2.3 diff --git a/docs/readme.txt b/docs/readme.txt index cff519531..ca10a4644 100644 --- a/docs/readme.txt +++ b/docs/readme.txt @@ -156,6 +156,7 @@ NVIDIA GPUs require "NVIDIA Driver" (440.64 or later) and "CUDA Toolkit" (9.0 or - IKE-PSK MD5 - IKE-PSK SHA1 - SNMPv3 HMAC-MD5-96 +- SNMPv3 HMAC-MD5-96/HMAC-SHA1-96 - SNMPv3 HMAC-SHA1-96 - WPA-EAPOL-PBKDF2 - WPA-EAPOL-PMK diff --git a/src/backend.c b/src/backend.c index 9f544589e..cddd8cb5d 100644 --- a/src/backend.c +++ b/src/backend.c @@ -9097,6 +9097,19 @@ int backend_ctx_devices_init (hashcat_ctx_t *hashcat_ctx, const int comptime) if (device_param->opencl_device_type & CL_DEVICE_TYPE_GPU) { + if ((device_param->opencl_platform_vendor_id == VENDOR_ID_APPLE) && (device_param->opencl_device_vendor_id == VENDOR_ID_AMD)) + { + // from https://www.khronos.org/registry/OpenCL/extensions/amd/cl_amd_device_attribute_query.txt + #define CL_DEVICE_WAVEFRONT_WIDTH_AMD 0x4043 + + // crazy, but apple does not support this query! + // the best alternative is "Preferred work group size multiple (kernel)", but requires to specify a kernel. + // so we will set kernel_preferred_wgs_multiple intentionally to 0 because otherwise it it set to 8 by default. + // we then assign the value kernel_preferred_wgs_multiple a small kernel like bzero after test if this was set to 0. + + device_param->kernel_preferred_wgs_multiple = 0; + } + if ((device_param->opencl_platform_vendor_id == VENDOR_ID_AMD) && (device_param->opencl_device_vendor_id == VENDOR_ID_AMD)) { cl_uint device_wavefront_width_amd; @@ -11628,6 +11641,10 @@ int backend_session_begin (hashcat_ctx_t *hashcat_ctx) if (get_opencl_kernel_preferred_wgs_multiple (hashcat_ctx, device_param, device_param->opencl_kernel_bzero, &device_param->kernel_preferred_wgs_multiple_bzero) == -1) return -1; + // apple hack, but perhaps also an alternative for other vendors + + if (device_param->kernel_preferred_wgs_multiple == 0) device_param->kernel_preferred_wgs_multiple = device_param->kernel_preferred_wgs_multiple_bzero; + // GPU autotune init if (hc_clCreateKernel (hashcat_ctx, device_param->opencl_program_shared, "gpu_atinit", &device_param->opencl_kernel_atinit) == -1) return -1; diff --git a/src/modules/module_18600.c b/src/modules/module_18600.c index c24ea37fe..8f2d05f4c 100644 --- a/src/modules/module_18600.c +++ b/src/modules/module_18600.c @@ -21,7 +21,8 @@ static const char *HASH_NAME = "Open Document Format (ODF) 1.1 (SHA-1, Blow static const u64 KERN_TYPE = 18600; static const u32 OPTI_TYPE = OPTI_TYPE_ZERO_BYTE | OPTI_TYPE_SLOW_HASH_SIMD_LOOP; -static const u64 OPTS_TYPE = OPTS_TYPE_PT_GENERATE_LE; +static const u64 OPTS_TYPE = OPTS_TYPE_PT_GENERATE_LE + | OPTS_TYPE_DYNAMIC_SHARED; static const u32 SALT_TYPE = SALT_TYPE_EMBEDDED; static const char *ST_PASS = "hashcat"; static const char *ST_HASH = "$odf$*0*0*1024*16*bff753835f4ea15644b8a2f8e4b5be3d147b9576*8*ee371da34333b69d*16*a902eff54a4d782a26a899a31f97bef4*0*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"; @@ -66,16 +67,25 @@ char *module_jit_build_options (MAYBE_UNUSED const hashconfig_t *hashconfig, MAY { char *jit_build_options = NULL; + // this mode heavily depends on the available shared memory size + // note the kernel need to have some special code changes in order to make use to use post-48k memory region + // we need to set some macros + + bool use_dynamic = false; + + if (device_param->is_cuda == true) + { + use_dynamic = true; + } + // this uses some nice feedback effect. // based on the device_local_mem_size the reqd_work_group_size in the kernel is set to some value // which is then is read from the opencl host in the kernel_preferred_wgs_multiple1/2/3 result. // therefore we do not need to set module_kernel_threads_min/max except for CPU, where the threads are set to fixed 1. - u32 fixed_local_size = 0; - if (device_param->opencl_device_type & CL_DEVICE_TYPE_CPU) { - fixed_local_size = 1; + hc_asprintf (&jit_build_options, "-D FIXED_LOCAL_SIZE_COMP=%u", 1); } else { @@ -91,29 +101,58 @@ char *module_jit_build_options (MAYBE_UNUSED const hashconfig_t *hashconfig, MAY if (device_param->is_opencl == true) { - overhead = 4; + overhead = 1; } } if (user_options->kernel_threads_chgd == true) { - fixed_local_size = user_options->kernel_threads; + u32 fixed_local_size = user_options->kernel_threads; - // otherwise out-of-bound reads - - if ((fixed_local_size * 4096) > (device_param->device_local_mem_size - overhead)) + if (use_dynamic == true) { - fixed_local_size = (device_param->device_local_mem_size - overhead) / 4096; + if ((fixed_local_size * 4096) > device_param->kernel_dynamic_local_mem_size_memset) + { + // otherwise out-of-bound reads + + fixed_local_size = device_param->kernel_dynamic_local_mem_size_memset / 4096; + } + + hc_asprintf (&jit_build_options, "-D FIXED_LOCAL_SIZE_COMP=%u -D DYNAMIC_LOCAL", fixed_local_size); + } + else + { + if ((fixed_local_size * 4096) > (device_param->device_local_mem_size - overhead)) + { + // otherwise out-of-bound reads + + fixed_local_size = (device_param->device_local_mem_size - overhead) / 4096; + } + + hc_asprintf (&jit_build_options, "-D FIXED_LOCAL_SIZE_COMP=%u", fixed_local_size); } } else { - fixed_local_size = (device_param->device_local_mem_size - overhead) / 4096; + if (use_dynamic == true) + { + // using kernel_dynamic_local_mem_size_memset is a bit hackish. + // we had to brute-force this value out of an already loaded CUDA function. + // there's no official way to query for this value. + + const u32 fixed_local_size = device_param->kernel_dynamic_local_mem_size_memset / 4096; + + hc_asprintf (&jit_build_options, "-D FIXED_LOCAL_SIZE_COMP=%u -D DYNAMIC_LOCAL", fixed_local_size); + } + else + { + const u32 fixed_local_size = (device_param->device_local_mem_size - overhead) / 4096; + + hc_asprintf (&jit_build_options, "-D FIXED_LOCAL_SIZE_COMP=%u", fixed_local_size); + } } } - hc_asprintf (&jit_build_options, "-D FIXED_LOCAL_SIZE=%u", fixed_local_size); - return jit_build_options; } diff --git a/src/modules/module_25000.c b/src/modules/module_25000.c new file mode 100644 index 000000000..deac74055 --- /dev/null +++ b/src/modules/module_25000.c @@ -0,0 +1,342 @@ +/** + * Author......: See docs/credits.txt + * License.....: MIT + */ + +#include "common.h" +#include "types.h" +#include "modules.h" +#include "bitops.h" +#include "convert.h" +#include "shared.h" +#include "memory.h" +#include "emu_inc_hash_md5.h" + +static const u32 ATTACK_EXEC = ATTACK_EXEC_OUTSIDE_KERNEL; +static const u32 DGST_POS0 = 0; +static const u32 DGST_POS1 = 1; +static const u32 DGST_POS2 = 2; +static const u32 DGST_POS3 = 3; +static const u32 DGST_SIZE = DGST_SIZE_4_4; // 4_3 +static const u32 HASH_CATEGORY = HASH_CATEGORY_NETWORK_PROTOCOL; +static const char *HASH_NAME = "SNMPv3 HMAC-MD5-96/HMAC-SHA1-96"; +static const u64 KERN_TYPE = 25000; +static const u32 OPTI_TYPE = OPTI_TYPE_ZERO_BYTE; +static const u64 OPTS_TYPE = OPTS_TYPE_PT_GENERATE_LE; +static const u32 SALT_TYPE = SALT_TYPE_EMBEDDED; +static const char *ST_PASS = "hashcat1"; +static const char *ST_HASH = "$SNMPv3$0$45889431$30818f0201033011020409242fc0020300ffe304010102010304383036041180001f88808106d566db57fd600000000002011002020118040a6d61747269785f4d4435040c0000000000000000000000000400303d041180001f88808106d566db57fd60000000000400a226020411f319300201000201003018301606082b06010201010200060a2b06010401bf0803020a$80001f88808106d566db57fd6000000000$1b37c3ea872731f922959e90"; + +u32 module_attack_exec (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ATTACK_EXEC; } +u32 module_dgst_pos0 (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS0; } +u32 module_dgst_pos1 (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS1; } +u32 module_dgst_pos2 (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS2; } +u32 module_dgst_pos3 (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS3; } +u32 module_dgst_size (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_SIZE; } +u32 module_hash_category (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return HASH_CATEGORY; } +const char *module_hash_name (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return HASH_NAME; } +u64 module_kern_type (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return KERN_TYPE; } +u32 module_opti_type (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return OPTI_TYPE; } +u64 module_opts_type (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return OPTS_TYPE; } +u32 module_salt_type (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return SALT_TYPE; } +const char *module_st_hash (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ST_HASH; } +const char *module_st_pass (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ST_PASS; } + +static const char *SIGNATURE_SNMPV3 = "$SNMPv3$0$"; + +#define SNMPV3_SALT_MAX 1500 +#define SNMPV3_ENGINEID_MAX 34 +#define SNMPV3_MSG_AUTH_PARAMS_LEN 12 +#define SNMPV3_ROUNDS 1048576 +#define SNMPV3_MAX_PW_LENGTH 64 + +#define SNMPV3_TMP_ELEMS 4096 // 4096 = (256 (max pw length) * 64) / sizeof (u32) +#define SNMPV3_HASH_ELEMS_MD5 4 +#define SNMPV3_HASH_ELEMS_SHA1 8 + +#define SNMPV3_MAX_SALT_ELEMS 512 // 512 * 4 = 2048 > 1500, also has to be multiple of 64 +#define SNMPV3_MAX_ENGINE_ELEMS 16 // 16 * 4 = 64 > 32, also has to be multiple of 64 +#define SNMPV3_MAX_PNUM_ELEMS 4 // 4 * 4 = 16 > 9 + +typedef struct hmac_md5_tmp +{ + u32 tmp_md5[SNMPV3_TMP_ELEMS]; + u32 tmp_sha1[SNMPV3_TMP_ELEMS]; + + u32 h_md5[SNMPV3_HASH_ELEMS_MD5]; + u32 h_sha1[SNMPV3_HASH_ELEMS_SHA1]; + +} hmac_md5_tmp_t; + +typedef struct snmpv3 +{ + u32 salt_buf[SNMPV3_MAX_SALT_ELEMS]; + u32 salt_len; + + u32 engineID_buf[SNMPV3_MAX_ENGINE_ELEMS]; + u32 engineID_len; + + u32 packet_number[SNMPV3_MAX_PNUM_ELEMS]; + +} snmpv3_t; + +u32 module_pw_min (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + const u32 pw_min = 8; + + return pw_min; +} + +u64 module_esalt_size (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + const u64 esalt_size = (const u64) sizeof (snmpv3_t); + + return esalt_size; +} + +u64 module_tmp_size (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + const u64 tmp_size = (const u64) sizeof (hmac_md5_tmp_t); + + return tmp_size; +} + +u32 module_kernel_loops_min (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + // we need to fix iteration count to guarantee the loop count is a multiple of 64 + // 2k calls to md5_transform/sha1_transform typically is enough to overtime pcie bottleneck + + const u32 kernel_loops_min = 2048 * 64; + + return kernel_loops_min; +} + +u32 module_kernel_loops_max (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + const u32 kernel_loops_max = 2048 * 64; + + return kernel_loops_max; +} + +int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED void *digest_buf, MAYBE_UNUSED salt_t *salt, MAYBE_UNUSED void *esalt_buf, MAYBE_UNUSED void *hook_salt_buf, MAYBE_UNUSED hashinfo_t *hash_info, const char *line_buf, MAYBE_UNUSED const int line_len) +{ + u32 *digest = (u32 *) digest_buf; + + snmpv3_t *snmpv3 = (snmpv3_t *) esalt_buf; + + token_t token; + + token.token_cnt = 5; + token.signatures_cnt = 1; + token.signatures_buf[0] = SIGNATURE_SNMPV3; + + token.len[0] = 10; + token.attr[0] = TOKEN_ATTR_FIXED_LENGTH + | TOKEN_ATTR_VERIFY_SIGNATURE; + + // packet number + token.len_min[1] = 1; + token.len_max[1] = 8; + token.sep[1] = '$'; + token.attr[1] = TOKEN_ATTR_VERIFY_LENGTH + | TOKEN_ATTR_VERIFY_DIGIT; + // salt + token.len_min[2] = SNMPV3_MSG_AUTH_PARAMS_LEN * 2; + token.len_max[2] = SNMPV3_SALT_MAX * 2; + token.sep[2] = '$'; + token.attr[2] = TOKEN_ATTR_VERIFY_LENGTH + | TOKEN_ATTR_VERIFY_HEX; + + // engineid + token.len_min[3] = 26; + token.len_max[3] = SNMPV3_ENGINEID_MAX; + token.sep[3] = '$'; + token.attr[3] = TOKEN_ATTR_VERIFY_LENGTH + | TOKEN_ATTR_VERIFY_HEX; + + // digest + token.len[4] = SNMPV3_MSG_AUTH_PARAMS_LEN * 2; + token.sep[4] = '$'; + token.attr[4] = TOKEN_ATTR_FIXED_LENGTH + | TOKEN_ATTR_VERIFY_HEX; + + const int rc_tokenizer = input_tokenizer ((const u8 *) line_buf, line_len, &token); + + if (rc_tokenizer != PARSER_OK) return (rc_tokenizer); + + // packet number + + const u8 *packet_number_pos = token.buf[1]; + const int packet_number_len = token.len[1]; + + memset (snmpv3->packet_number, 0, sizeof (snmpv3->packet_number)); + + strncpy ((char *) snmpv3->packet_number, (char *) packet_number_pos, packet_number_len); + + // salt + + const u8 *salt_pos = token.buf[2]; + const int salt_len = token.len[2]; + + u8 *salt_ptr = (u8 *) snmpv3->salt_buf; + + snmpv3->salt_len = hex_decode (salt_pos, salt_len, salt_ptr); + + salt->salt_iter = SNMPV3_ROUNDS; + + // handle unique salts detection + + md5_ctx_t md5_ctx; + + md5_init (&md5_ctx); + md5_update (&md5_ctx, snmpv3->salt_buf, snmpv3->salt_len); + md5_final (&md5_ctx); + + // store md5(snmpv3->salt_buf) in salt_buf + + salt->salt_len = 16; + + memcpy (salt->salt_buf, md5_ctx.h, salt->salt_len); + + // engineid + + const u8 *engineID_pos = token.buf[3]; + const int engineID_len = token.len[3]; + + u8 *engineID_ptr = (u8 *) snmpv3->engineID_buf; + + snmpv3->engineID_len = hex_decode (engineID_pos, engineID_len, engineID_ptr); + + // digest + + const u8 *hash_pos = token.buf[4]; + + digest[0] = hex_to_u32 (hash_pos + 0); + digest[1] = hex_to_u32 (hash_pos + 8); + digest[2] = hex_to_u32 (hash_pos + 16); + + // prefer sha1 due to speed + + digest[0] = byte_swap_32 (digest[0]); + digest[1] = byte_swap_32 (digest[1]); + digest[2] = byte_swap_32 (digest[2]); + + digest[3] = 0; + + return (PARSER_OK); +} + +int module_hash_encode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const void *digest_buf, MAYBE_UNUSED const salt_t *salt, MAYBE_UNUSED const void *esalt_buf, MAYBE_UNUSED const void *hook_salt_buf, MAYBE_UNUSED const hashinfo_t *hash_info, char *line_buf, MAYBE_UNUSED const int line_size) +{ + const u32 *digest = (const u32 *) digest_buf; + + snmpv3_t *snmpv3 = (snmpv3_t *) esalt_buf; + + u8 *out_buf = (u8 *) line_buf; + + int out_len = snprintf (line_buf, line_size, "%s%s$", SIGNATURE_SNMPV3, (char *) snmpv3->packet_number); + + out_len += hex_encode ((u8 *) snmpv3->salt_buf, snmpv3->salt_len, out_buf + out_len); + + out_buf[out_len] = '$'; + + out_len++; + + out_len += hex_encode ((u8 *) snmpv3->engineID_buf, snmpv3->engineID_len, out_buf + out_len); + + out_buf[out_len] = '$'; + + out_len++; + + // prefer sha1 due to speed + + u32 digest_tmp[3]; + + digest_tmp[0] = byte_swap_32 (digest[0]); + digest_tmp[1] = byte_swap_32 (digest[1]); + digest_tmp[2] = byte_swap_32 (digest[2]); + + u32_to_hex (digest_tmp[0], out_buf + out_len); out_len += 8; + u32_to_hex (digest_tmp[1], out_buf + out_len); out_len += 8; + u32_to_hex (digest_tmp[2], out_buf + out_len); out_len += 8; + + out_buf[out_len] = 0; + + return out_len; +} + +void module_init (module_ctx_t *module_ctx) +{ + module_ctx->module_context_size = MODULE_CONTEXT_SIZE_CURRENT; + module_ctx->module_interface_version = MODULE_INTERFACE_VERSION_CURRENT; + + module_ctx->module_attack_exec = module_attack_exec; + module_ctx->module_benchmark_esalt = MODULE_DEFAULT; + module_ctx->module_benchmark_hook_salt = MODULE_DEFAULT; + module_ctx->module_benchmark_mask = MODULE_DEFAULT; + module_ctx->module_benchmark_salt = MODULE_DEFAULT; + module_ctx->module_build_plain_postprocess = MODULE_DEFAULT; + module_ctx->module_deep_comp_kernel = MODULE_DEFAULT; + module_ctx->module_dgst_pos0 = module_dgst_pos0; + module_ctx->module_dgst_pos1 = module_dgst_pos1; + module_ctx->module_dgst_pos2 = module_dgst_pos2; + module_ctx->module_dgst_pos3 = module_dgst_pos3; + module_ctx->module_dgst_size = module_dgst_size; + module_ctx->module_dictstat_disable = MODULE_DEFAULT; + module_ctx->module_esalt_size = module_esalt_size; + module_ctx->module_extra_buffer_size = MODULE_DEFAULT; + module_ctx->module_extra_tmp_size = MODULE_DEFAULT; + module_ctx->module_forced_outfile_format = MODULE_DEFAULT; + module_ctx->module_hash_binary_count = MODULE_DEFAULT; + module_ctx->module_hash_binary_parse = MODULE_DEFAULT; + module_ctx->module_hash_binary_save = MODULE_DEFAULT; + module_ctx->module_hash_decode_potfile = MODULE_DEFAULT; + module_ctx->module_hash_decode_zero_hash = MODULE_DEFAULT; + module_ctx->module_hash_decode = module_hash_decode; + module_ctx->module_hash_encode_status = MODULE_DEFAULT; + module_ctx->module_hash_encode_potfile = MODULE_DEFAULT; + module_ctx->module_hash_encode = module_hash_encode; + module_ctx->module_hash_init_selftest = MODULE_DEFAULT; + module_ctx->module_hash_mode = MODULE_DEFAULT; + module_ctx->module_hash_category = module_hash_category; + module_ctx->module_hash_name = module_hash_name; + module_ctx->module_hashes_count_min = MODULE_DEFAULT; + module_ctx->module_hashes_count_max = MODULE_DEFAULT; + module_ctx->module_hlfmt_disable = MODULE_DEFAULT; + module_ctx->module_hook_extra_param_size = MODULE_DEFAULT; + module_ctx->module_hook_extra_param_init = MODULE_DEFAULT; + module_ctx->module_hook_extra_param_term = MODULE_DEFAULT; + module_ctx->module_hook12 = MODULE_DEFAULT; + module_ctx->module_hook23 = MODULE_DEFAULT; + module_ctx->module_hook_salt_size = MODULE_DEFAULT; + module_ctx->module_hook_size = MODULE_DEFAULT; + module_ctx->module_jit_build_options = MODULE_DEFAULT; + module_ctx->module_jit_cache_disable = MODULE_DEFAULT; + module_ctx->module_kernel_accel_max = MODULE_DEFAULT; + module_ctx->module_kernel_accel_min = MODULE_DEFAULT; + module_ctx->module_kernel_loops_max = module_kernel_loops_max; + module_ctx->module_kernel_loops_min = module_kernel_loops_min; + module_ctx->module_kernel_threads_max = MODULE_DEFAULT; + module_ctx->module_kernel_threads_min = MODULE_DEFAULT; + module_ctx->module_kern_type = module_kern_type; + module_ctx->module_kern_type_dynamic = MODULE_DEFAULT; + module_ctx->module_opti_type = module_opti_type; + module_ctx->module_opts_type = module_opts_type; + module_ctx->module_outfile_check_disable = MODULE_DEFAULT; + module_ctx->module_outfile_check_nocomp = MODULE_DEFAULT; + module_ctx->module_potfile_custom_check = MODULE_DEFAULT; + module_ctx->module_potfile_disable = MODULE_DEFAULT; + module_ctx->module_potfile_keep_all_hashes = MODULE_DEFAULT; + module_ctx->module_pwdump_column = MODULE_DEFAULT; + module_ctx->module_pw_max = MODULE_DEFAULT; + module_ctx->module_pw_min = module_pw_min; + module_ctx->module_salt_max = MODULE_DEFAULT; + module_ctx->module_salt_min = MODULE_DEFAULT; + module_ctx->module_salt_type = module_salt_type; + module_ctx->module_separator = MODULE_DEFAULT; + module_ctx->module_st_hash = module_st_hash; + module_ctx->module_st_pass = module_st_pass; + module_ctx->module_tmp_size = module_tmp_size; + module_ctx->module_unstable_warning = MODULE_DEFAULT; + module_ctx->module_warmup_disable = MODULE_DEFAULT; +} diff --git a/src/modules/module_25100.c b/src/modules/module_25100.c index 22155ce9a..4fdbdf828 100644 --- a/src/modules/module_25100.c +++ b/src/modules/module_25100.c @@ -24,8 +24,8 @@ static const u64 KERN_TYPE = 25100; static const u32 OPTI_TYPE = OPTI_TYPE_ZERO_BYTE; static const u64 OPTS_TYPE = OPTS_TYPE_PT_GENERATE_LE; static const u32 SALT_TYPE = SALT_TYPE_EMBEDDED; -static const char *ST_PASS = "hashcat"; -static const char *ST_HASH = "$SNMPv3$1$76$3081b10201033011020430f6f3d5020300ffe304010702010304373035040d80001f888059dc486145a2632202010802020ab90405706970706f040c00000000000000000000000004080000000103d5321a0460826ecf6443956d4c364bfc6f6ffc8ee0df000ffd0955af12d2c0f3c60fadea417d2bb80c0b2c1fa7a46ce44f9f16e15ee830a49881f60ecfa757d2f04000eb39a94058121d88ca20eeef4e6bf06784c67c15f144915d9bc2c6a0461da92a4abe$80001f888059dc486145a26322$c51ba677ad96869c1cb32196"; +static const char *ST_PASS = "hashcat1"; +static const char *ST_HASH = "$SNMPv3$1$45889431$30818f0201033011020409242fc0020300ffe304010102010304383036041180001f88808106d566db57fd600000000002011002020118040a6d61747269785f4d4435040c0000000000000000000000000400303d041180001f88808106d566db57fd60000000000400a226020411f319300201000201003018301606082b06010201010200060a2b06010401bf0803020a$80001f88808106d566db57fd6000000000$1b37c3ea872731f922959e90"; u32 module_attack_exec (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ATTACK_EXEC; } u32 module_dgst_pos0 (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS0; } @@ -45,8 +45,8 @@ const char *module_st_pass (MAYBE_UNUSED const hashconfig_t *hashconfig, static const char *SIGNATURE_SNMPV3 = "$SNMPv3$1$"; #define SNMPV3_SALT_MAX 1500 -#define SNMPV3_ENGINEID_MAX 32 -#define SNMPV3_MSG_AUTH_PARAMS_MAX 12 +#define SNMPV3_ENGINEID_MAX 34 +#define SNMPV3_MSG_AUTH_PARAMS_LEN 12 #define SNMPV3_ROUNDS 1048576 #define SNMPV3_MAX_PW_LENGTH 64 @@ -76,6 +76,13 @@ typedef struct snmpv3 } snmpv3_t; +u32 module_pw_min (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + const u32 pw_min = 8; + + return pw_min; +} + u64 module_esalt_size (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { const u64 esalt_size = (const u64) sizeof (snmpv3_t); @@ -130,23 +137,23 @@ int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSE token.attr[1] = TOKEN_ATTR_VERIFY_LENGTH | TOKEN_ATTR_VERIFY_DIGIT; // salt - token.len_min[2] = 12 * 2; + token.len_min[2] = SNMPV3_MSG_AUTH_PARAMS_LEN * 2; token.len_max[2] = SNMPV3_SALT_MAX * 2; token.sep[2] = '$'; token.attr[2] = TOKEN_ATTR_VERIFY_LENGTH | TOKEN_ATTR_VERIFY_HEX; // engineid - token.len_min[3] = 5; + token.len_min[3] = 26; token.len_max[3] = SNMPV3_ENGINEID_MAX; token.sep[3] = '$'; - token.attr[3] = TOKEN_ATTR_VERIFY_LENGTH; + token.attr[3] = TOKEN_ATTR_VERIFY_LENGTH + | TOKEN_ATTR_VERIFY_HEX; // digest - token.len_min[4] = SNMPV3_MSG_AUTH_PARAMS_MAX * 2; - token.len_max[4] = SNMPV3_MSG_AUTH_PARAMS_MAX * 2; + token.len[4] = SNMPV3_MSG_AUTH_PARAMS_LEN * 2; token.sep[4] = '$'; - token.attr[4] = TOKEN_ATTR_VERIFY_LENGTH + token.attr[4] = TOKEN_ATTR_FIXED_LENGTH | TOKEN_ATTR_VERIFY_HEX; const int rc_tokenizer = input_tokenizer ((const u8 *) line_buf, line_len, &token); @@ -303,7 +310,7 @@ void module_init (module_ctx_t *module_ctx) module_ctx->module_potfile_keep_all_hashes = MODULE_DEFAULT; module_ctx->module_pwdump_column = MODULE_DEFAULT; module_ctx->module_pw_max = MODULE_DEFAULT; - module_ctx->module_pw_min = MODULE_DEFAULT; + module_ctx->module_pw_min = module_pw_min; module_ctx->module_salt_max = MODULE_DEFAULT; module_ctx->module_salt_min = MODULE_DEFAULT; module_ctx->module_salt_type = module_salt_type; diff --git a/src/modules/module_25200.c b/src/modules/module_25200.c index c98a347ef..66573cb5a 100644 --- a/src/modules/module_25200.c +++ b/src/modules/module_25200.c @@ -24,8 +24,8 @@ static const u64 KERN_TYPE = 25200; static const u32 OPTI_TYPE = OPTI_TYPE_ZERO_BYTE; static const u64 OPTS_TYPE = OPTS_TYPE_PT_GENERATE_LE; static const u32 SALT_TYPE = SALT_TYPE_EMBEDDED; -static const char *ST_PASS = "hashcat"; -static const char *ST_HASH = "$SNMPv3$2$66763052$13981919518623358902340156831753173612320956749283824166083320737667668557830898783481876963136410266762758410322896320705075044221495960812100760230106803899899467077793703068392752686845035561487927252457444567685389901239388468830507087105054207914325254376053788152029716918450770264047103676562621965276752797029332926039166807829108367446173251908238116020942421323633620301312478670302264165059728208402342845743839533979473825394866704960428648622730299023225638967097578710279784722583947877561544154219162080289188160001741612377820114739093961409809862173307722539556954826052612794054060797358016549602977742745078911393042420821004243620362464971828700104979572910001640083882586179153483503492341163054930853321963503411228241996417991605003371264529827508426941919673592574025732354318435733211018917539824570724324796232199960952117561108106623865308577977944499366806697863259301760429786001824121720055893438673268643594146796410437039466462606490272723136671298529920486664067752007564122205089571790718437001200506203464426405927405102300269665189637001279369690218157456566218400534722049383049029139069701182053729830585217732347396312967325628046845068493719801191260136945971516486442056102815519090214442808707545803919529217103430588641187558031052830941742920355893755319896626873275796534820394248837050567688575113833311009595128372820474678989203565094681918285106102363272728922586582037066265522397748326630668375500179630717875844561081542915676557961288028298248995547031274515608973804660067065502484039882958958452781062725550260382637592283691962996228392332833626159043179186189904614052189303508782635840692436969244901198720814518$79f7b1$57e964c7cb117647004cf132"; +static const char *ST_PASS = "hashcat1"; +static const char *ST_HASH = "$SNMPv3$2$45889431$30818f02010330110204371780f3020300ffe304010102010304383036041180001f88808106d566db57fd600000000002011002020118040a6d61747269785f534841040c0000000000000000000000000400303d041180001f88808106d566db57fd60000000000400a2260204073557d50201000201003018301606082b06010201010200060a2b06010401bf0803020a$80001f88808106d566db57fd6000000000$81f14f1930589f26f6755f6b"; u32 module_attack_exec (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ATTACK_EXEC; } u32 module_dgst_pos0 (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS0; } @@ -45,8 +45,8 @@ const char *module_st_pass (MAYBE_UNUSED const hashconfig_t *hashconfig, static const char *SIGNATURE_SNMPV3 = "$SNMPv3$2$"; #define SNMPV3_SALT_MAX 1500 -#define SNMPV3_ENGINEID_MAX 32 -#define SNMPV3_MSG_AUTH_PARAMS_MAX 12 +#define SNMPV3_ENGINEID_MAX 34 +#define SNMPV3_MSG_AUTH_PARAMS_LEN 12 #define SNMPV3_ROUNDS 1048576 #define SNMPV3_MAX_PW_LENGTH 64 @@ -76,6 +76,13 @@ typedef struct snmpv3 } snmpv3_t; +u32 module_pw_min (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) +{ + const u32 pw_min = 8; + + return pw_min; +} + u64 module_esalt_size (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { const u64 esalt_size = (const u64) sizeof (snmpv3_t); @@ -130,23 +137,23 @@ int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSE token.attr[1] = TOKEN_ATTR_VERIFY_LENGTH | TOKEN_ATTR_VERIFY_DIGIT; // salt - token.len_min[2] = 12 * 2; + token.len_min[2] = SNMPV3_MSG_AUTH_PARAMS_LEN * 2; token.len_max[2] = SNMPV3_SALT_MAX * 2; token.sep[2] = '$'; token.attr[2] = TOKEN_ATTR_VERIFY_LENGTH | TOKEN_ATTR_VERIFY_HEX; // engineid - token.len_min[3] = 5; + token.len_min[3] = 26; token.len_max[3] = SNMPV3_ENGINEID_MAX; token.sep[3] = '$'; - token.attr[3] = TOKEN_ATTR_VERIFY_LENGTH; + token.attr[3] = TOKEN_ATTR_VERIFY_LENGTH + | TOKEN_ATTR_VERIFY_HEX; // digest - token.len_min[4] = SNMPV3_MSG_AUTH_PARAMS_MAX * 2; - token.len_max[4] = SNMPV3_MSG_AUTH_PARAMS_MAX * 2; + token.len[4] = SNMPV3_MSG_AUTH_PARAMS_LEN * 2; token.sep[4] = '$'; - token.attr[4] = TOKEN_ATTR_VERIFY_LENGTH + token.attr[4] = TOKEN_ATTR_FIXED_LENGTH | TOKEN_ATTR_VERIFY_HEX; const int rc_tokenizer = input_tokenizer ((const u8 *) line_buf, line_len, &token); @@ -183,10 +190,10 @@ int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSE // store sha1(snmpv3->salt_buf) in salt_buf - memcpy (salt->salt_buf, sha1_ctx.h, 20); - salt->salt_len = 20; + memcpy (salt->salt_buf, sha1_ctx.h, salt->salt_len); + // engineid const u8 *engineID_pos = token.buf[3]; @@ -203,12 +210,13 @@ int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSE digest[0] = hex_to_u32 (hash_pos + 0); digest[1] = hex_to_u32 (hash_pos + 8); digest[2] = hex_to_u32 (hash_pos + 16); - digest[3] = 0; digest[0] = byte_swap_32 (digest[0]); digest[1] = byte_swap_32 (digest[1]); digest[2] = byte_swap_32 (digest[2]); + digest[3] = 0; + return (PARSER_OK); } @@ -313,7 +321,7 @@ void module_init (module_ctx_t *module_ctx) module_ctx->module_potfile_keep_all_hashes = MODULE_DEFAULT; module_ctx->module_pwdump_column = MODULE_DEFAULT; module_ctx->module_pw_max = MODULE_DEFAULT; - module_ctx->module_pw_min = MODULE_DEFAULT; + module_ctx->module_pw_min = module_pw_min; module_ctx->module_salt_max = MODULE_DEFAULT; module_ctx->module_salt_min = MODULE_DEFAULT; module_ctx->module_salt_type = module_salt_type; diff --git a/tools/test_modules/m25000.pm b/tools/test_modules/m25000.pm new file mode 100644 index 000000000..71cc1c512 --- /dev/null +++ b/tools/test_modules/m25000.pm @@ -0,0 +1,116 @@ +#!/usr/bin/env perl + +## +## Author......: See docs/credits.txt +## License.....: MIT +## + +use strict; +use warnings; + +use Digest::MD5 qw (md5 md5_hex); +use Digest::SHA qw (sha1 sha1_hex); +use Digest::HMAC qw (hmac hmac_hex); + +sub module_constraints { [[8, 256], [24, 3000], [-1, -1], [-1, -1], [-1, -1]] } + +sub module_generate_hash +{ + my $word = shift; + my $salt = shift; + my $pkt_num = shift // int(rand(100000000)); + my $engineID = shift // random_hex_string(26, 34); + my $mode = shift // int(rand(1)) + 1; + + # make even if needed + + if (length($salt) %2 == 1) + { + $salt = $salt . "8"; + } + + my $string1 = $word x 1048576; + + $string1 = substr ($string1, 0, 1048576); + + my $digest1 = ''; + + if ($mode eq 2) + { + $digest1 = sha1_hex ($string1); + } + elsif ($mode eq 1) + { + $digest1 = md5_hex ($string1); + } + + my $buf = join '', $digest1, $engineID, $digest1; + + my $digest = ''; + + if ($mode eq 2) + { + my $digest2 = sha1(pack("H*", $buf)); + + $digest = hmac_hex (pack("H*", $salt), $digest2, \&sha1); + } + elsif ($mode eq 1) + { + my $digest2 = md5(pack("H*", $buf)); + + $digest = hmac_hex (pack("H*", $salt), $digest2, \&md5); + } + + $digest = substr ($digest, 0, 24); + + my $hash = sprintf ("\$SNMPv3\$0\$%s\$%s\$%s\$%s", $pkt_num, $salt, $engineID, $digest); + + return $hash; +} + +sub module_verify_hash +{ + my $line = shift; + + my $idx = index ($line, ':'); + + return unless $idx >= 0; + + my $hash = substr ($line, 0, $idx); + my $word = substr ($line, $idx + 1); + + return unless length ($word) gt 0; + return unless substr ($hash, 0, 10) eq '$SNMPv3$0$'; + + my (undef, $signature, $version, $pkt_num, $salt, $engineID, $digest) = split '\$', $hash; + + return unless defined $signature; + return unless defined $version; + return unless defined $pkt_num; + return unless defined $salt; + return unless defined $engineID; + return unless defined $digest; + + my $word_packed = pack_if_HEX_notation ($word); + + # gen md5 & sha1 hashes + + my $new_hash_md5 = module_generate_hash ($word_packed, $salt, $pkt_num, $engineID, 1); + my $new_hash_sha1 = module_generate_hash ($word_packed, $salt, $pkt_num, $engineID, 2); + + # parse digests + + my (undef, undef, undef, undef, undef, undef, $digest_md5) = split '\$', $new_hash_md5; + my (undef, undef, undef, undef, undef, undef, $digest_sha1) = split '\$', $new_hash_sha1; + + if ($digest eq $digest_md5) + { + return ($new_hash_md5, $word); + } + else + { + return ($new_hash_sha1, $word); + } +} + +1; diff --git a/tools/test_modules/m25100.pm b/tools/test_modules/m25100.pm index 6249df9a8..2335f7f2f 100644 --- a/tools/test_modules/m25100.pm +++ b/tools/test_modules/m25100.pm @@ -11,14 +11,14 @@ use warnings; use Digest::MD5 qw (md5 md5_hex); use Digest::HMAC qw (hmac hmac_hex); -sub module_constraints { [[1, 256], [24, 3000], [-1, -1], [-1, -1], [-1, -1]] } +sub module_constraints { [[8, 256], [24, 3000], [-1, -1], [-1, -1], [-1, -1]] } sub module_generate_hash { my $word = shift; my $salt = shift; - my $pkt_num = shift // int(rand(99999999)); - my $engineID = shift // random_hex_string(6); + my $pkt_num = shift // int(rand(100000000)); + my $engineID = shift // random_hex_string(26, 34); # make even if needed @@ -71,7 +71,7 @@ sub module_verify_hash my $word_packed = pack_if_HEX_notation ($word); - my $new_hash = module_generate_hash ($word_packed, $salt, $pkt_num, $engineID); #, $digest); + my $new_hash = module_generate_hash ($word_packed, $salt, $pkt_num, $engineID); return ($new_hash, $word); } diff --git a/tools/test_modules/m25200.pm b/tools/test_modules/m25200.pm index c44212825..d27908255 100644 --- a/tools/test_modules/m25200.pm +++ b/tools/test_modules/m25200.pm @@ -11,14 +11,14 @@ use warnings; use Digest::SHA qw (sha1 sha1_hex); use Digest::HMAC qw (hmac hmac_hex); -sub module_constraints { [[1, 256], [24, 3000], [-1, -1], [-1, -1], [-1, -1]] } +sub module_constraints { [[8, 256], [24, 3000], [-1, -1], [-1, -1], [-1, -1]] } sub module_generate_hash { my $word = shift; my $salt = shift; - my $pkt_num = shift // int(rand(99999999)); - my $engineID = shift // random_hex_string(6); + my $pkt_num = shift // int(rand(100000000)); + my $engineID = shift // random_hex_string(26, 34); # make even if needed