diff --git a/OpenCL/inc_hash_sha1.cl b/OpenCL/inc_hash_sha1.cl index 786554bf1..a4ca265c4 100644 --- a/OpenCL/inc_hash_sha1.cl +++ b/OpenCL/inc_hash_sha1.cl @@ -42,7 +42,6 @@ void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u3 u32 we_t = w3[2]; u32 wf_t = w3[3]; - #undef K #define K SHA1C00 SHA1_STEP_S (SHA1_F0o, a, b, c, d, e, w0_t); @@ -138,6 +137,8 @@ void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u3 we_t = rotl32_S ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP_S (SHA1_F1, c, d, e, a, b, we_t); wf_t = rotl32_S ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP_S (SHA1_F1, b, c, d, e, a, wf_t); + #undef K + digest[0] += a; digest[1] += b; digest[2] += c; @@ -844,7 +845,6 @@ void sha1_transform_vector (const u32x w0[4], const u32x w1[4], const u32x w2[4] u32x we_t = w3[2]; u32x wf_t = w3[3]; - #undef K #define K SHA1C00 SHA1_STEP (SHA1_F0o, a, b, c, d, e, w0_t); @@ -940,6 +940,8 @@ void sha1_transform_vector (const u32x w0[4], const u32x w1[4], const u32x w2[4] we_t = rotl32 ((wb_t ^ w6_t ^ w0_t ^ we_t), 1u); SHA1_STEP (SHA1_F1, c, d, e, a, b, we_t); wf_t = rotl32 ((wc_t ^ w7_t ^ w1_t ^ wf_t), 1u); SHA1_STEP (SHA1_F1, b, c, d, e, a, wf_t); + #undef K + digest[0] += a; digest[1] += b; digest[2] += c; diff --git a/OpenCL/inc_hash_sha256.cl b/OpenCL/inc_hash_sha256.cl index a5f60a5b3..3da164324 100644 --- a/OpenCL/inc_hash_sha256.cl +++ b/OpenCL/inc_hash_sha256.cl @@ -115,6 +115,9 @@ void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const ROUND_EXPAND_S (); ROUND_STEP_S (i); } + #undef ROUND_EXPAND_S + #undef ROUND_STEP_S + digest[0] += a; digest[1] += b; digest[2] += c; @@ -880,6 +883,9 @@ void sha256_transform_vector (const u32x w0[4], const u32x w1[4], const u32x w2[ ROUND_EXPAND (); ROUND_STEP (i); } + #undef ROUND_EXPAND + #undef ROUND_STEP + digest[0] += a; digest[1] += b; digest[2] += c; diff --git a/OpenCL/inc_hash_sha512.cl b/OpenCL/inc_hash_sha512.cl index d847ac65f..610da0f34 100644 --- a/OpenCL/inc_hash_sha512.cl +++ b/OpenCL/inc_hash_sha512.cl @@ -123,6 +123,9 @@ void sha512_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const ROUND_EXPAND_S (); ROUND_STEP_S (i); } + #undef ROUND_EXPAND_S + #undef ROUND_STEP_S + digest[0] += a; digest[1] += b; digest[2] += c; @@ -1315,6 +1318,9 @@ void sha512_transform_vector (const u32x w0[4], const u32x w1[4], const u32x w2[ ROUND_EXPAND (); ROUND_STEP (i); } + #undef ROUND_EXPAND + #undef ROUND_STEP + digest[0] += a; digest[1] += b; digest[2] += c; diff --git a/OpenCL/inc_hash_whirlpool.cl b/OpenCL/inc_hash_whirlpool.cl index e2c57e0de..5bcef30aa 100644 --- a/OpenCL/inc_hash_whirlpool.cl +++ b/OpenCL/inc_hash_whirlpool.cl @@ -2351,5 +2351,6 @@ void whirlpool_final_vector (whirlpool_ctx_vector_t *ctx, __local u32 (*s_Ch)[25 whirlpool_transform_vector (ctx->w0, ctx->w1, ctx->w2, ctx->w3, ctx->h, s_Ch, s_Cl); } +#undef R #undef BOX #undef BOX_S diff --git a/OpenCL/m08200.cl b/OpenCL/m08200.cl index 7d6d702f2..7a7a25897 100644 --- a/OpenCL/m08200.cl +++ b/OpenCL/m08200.cl @@ -3,207 +3,21 @@ * 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" +#include "inc_hash_sha512.cl" #define COMPARE_S "inc_comp_single.cl" #define COMPARE_M "inc_comp_multi.cl" -__constant u32a k_sha256[64] = -{ - SHA256C00, SHA256C01, SHA256C02, SHA256C03, - SHA256C04, SHA256C05, SHA256C06, SHA256C07, - SHA256C08, SHA256C09, SHA256C0a, SHA256C0b, - SHA256C0c, SHA256C0d, SHA256C0e, SHA256C0f, - SHA256C10, SHA256C11, SHA256C12, SHA256C13, - SHA256C14, SHA256C15, SHA256C16, SHA256C17, - SHA256C18, SHA256C19, SHA256C1a, SHA256C1b, - SHA256C1c, SHA256C1d, SHA256C1e, SHA256C1f, - SHA256C20, SHA256C21, SHA256C22, SHA256C23, - SHA256C24, SHA256C25, SHA256C26, SHA256C27, - SHA256C28, SHA256C29, SHA256C2a, SHA256C2b, - SHA256C2c, SHA256C2d, SHA256C2e, SHA256C2f, - SHA256C30, SHA256C31, SHA256C32, SHA256C33, - SHA256C34, SHA256C35, SHA256C36, SHA256C37, - SHA256C38, SHA256C39, SHA256C3a, SHA256C3b, - SHA256C3c, SHA256C3d, SHA256C3e, SHA256C3f, -}; - -__constant u64a k_sha512[80] = -{ - SHA512C00, SHA512C01, SHA512C02, SHA512C03, - SHA512C04, SHA512C05, SHA512C06, SHA512C07, - SHA512C08, SHA512C09, SHA512C0a, SHA512C0b, - SHA512C0c, SHA512C0d, SHA512C0e, SHA512C0f, - SHA512C10, SHA512C11, SHA512C12, SHA512C13, - SHA512C14, SHA512C15, SHA512C16, SHA512C17, - SHA512C18, SHA512C19, SHA512C1a, SHA512C1b, - SHA512C1c, SHA512C1d, SHA512C1e, SHA512C1f, - SHA512C20, SHA512C21, SHA512C22, SHA512C23, - SHA512C24, SHA512C25, SHA512C26, SHA512C27, - SHA512C28, SHA512C29, SHA512C2a, SHA512C2b, - SHA512C2c, SHA512C2d, SHA512C2e, SHA512C2f, - SHA512C30, SHA512C31, SHA512C32, SHA512C33, - SHA512C34, SHA512C35, SHA512C36, SHA512C37, - SHA512C38, SHA512C39, SHA512C3a, SHA512C3b, - SHA512C3c, SHA512C3d, SHA512C3e, SHA512C3f, - SHA512C40, SHA512C41, SHA512C42, SHA512C43, - SHA512C44, SHA512C45, SHA512C46, SHA512C47, - SHA512C48, SHA512C49, SHA512C4a, SHA512C4b, - SHA512C4c, SHA512C4d, SHA512C4e, SHA512C4f, -}; - -void sha256_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[8]) -{ - u32 a = digest[0]; - u32 b = digest[1]; - u32 c = digest[2]; - u32 d = digest[3]; - u32 e = digest[4]; - u32 f = digest[5]; - u32 g = digest[6]; - u32 h = digest[7]; - - u32 w0_t = w0[0]; - u32 w1_t = w0[1]; - u32 w2_t = w0[2]; - u32 w3_t = w0[3]; - u32 w4_t = w1[0]; - u32 w5_t = w1[1]; - u32 w6_t = w1[2]; - u32 w7_t = w1[3]; - u32 w8_t = w2[0]; - u32 w9_t = w2[1]; - u32 wa_t = w2[2]; - u32 wb_t = w2[3]; - u32 wc_t = w3[0]; - u32 wd_t = w3[1]; - u32 we_t = w3[2]; - u32 wf_t = w3[3]; - - #define ROUND_EXPAND() \ - { \ - w0_t = SHA256_EXPAND (we_t, w9_t, w1_t, w0_t); \ - w1_t = SHA256_EXPAND (wf_t, wa_t, w2_t, w1_t); \ - w2_t = SHA256_EXPAND (w0_t, wb_t, w3_t, w2_t); \ - w3_t = SHA256_EXPAND (w1_t, wc_t, w4_t, w3_t); \ - w4_t = SHA256_EXPAND (w2_t, wd_t, w5_t, w4_t); \ - w5_t = SHA256_EXPAND (w3_t, we_t, w6_t, w5_t); \ - w6_t = SHA256_EXPAND (w4_t, wf_t, w7_t, w6_t); \ - w7_t = SHA256_EXPAND (w5_t, w0_t, w8_t, w7_t); \ - w8_t = SHA256_EXPAND (w6_t, w1_t, w9_t, w8_t); \ - w9_t = SHA256_EXPAND (w7_t, w2_t, wa_t, w9_t); \ - wa_t = SHA256_EXPAND (w8_t, w3_t, wb_t, wa_t); \ - wb_t = SHA256_EXPAND (w9_t, w4_t, wc_t, wb_t); \ - wc_t = SHA256_EXPAND (wa_t, w5_t, wd_t, wc_t); \ - wd_t = SHA256_EXPAND (wb_t, w6_t, we_t, wd_t); \ - we_t = SHA256_EXPAND (wc_t, w7_t, wf_t, we_t); \ - wf_t = SHA256_EXPAND (wd_t, w8_t, w0_t, wf_t); \ - } - - #define ROUND_STEP(i) \ - { \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha256[i + 0]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha256[i + 1]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha256[i + 2]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha256[i + 3]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha256[i + 4]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha256[i + 5]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha256[i + 6]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha256[i + 7]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha256[i + 8]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha256[i + 9]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha256[i + 10]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha256[i + 11]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha256[i + 12]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha256[i + 13]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, c, d, e, f, g, h, a, b, we_t, k_sha256[i + 14]); \ - SHA256_STEP (SHA256_F0o, SHA256_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha256[i + 15]); \ - } - - ROUND_STEP (0); - - #ifdef _unroll - #pragma unroll - #endif - for (int i = 16; i < 64; i += 16) - { - ROUND_EXPAND (); ROUND_STEP (i); - } - - digest[0] += a; - digest[1] += b; - digest[2] += c; - digest[3] += d; - digest[4] += e; - digest[5] += f; - digest[6] += g; - digest[7] += h; -} - -void hmac_sha256_pad (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[8], u32 opad[8]) -{ - 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] = SHA256M_A; - ipad[1] = SHA256M_B; - ipad[2] = SHA256M_C; - ipad[3] = SHA256M_D; - ipad[4] = SHA256M_E; - ipad[5] = SHA256M_F; - ipad[6] = SHA256M_G; - ipad[7] = SHA256M_H; - - sha256_transform (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] = SHA256M_A; - opad[1] = SHA256M_B; - opad[2] = SHA256M_C; - opad[3] = SHA256M_D; - opad[4] = SHA256M_E; - opad[5] = SHA256M_F; - opad[6] = SHA256M_G; - opad[7] = SHA256M_H; - - sha256_transform (w0, w1, w2, w3, opad); -} - -void hmac_sha256_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[8], u32 opad[8], u32 digest[8]) +void hmac_sha512_run_V (u32x w0[4], u32x w1[4], u32x w2[4], u32x w3[4], u32x w4[4], u32x w5[4], u32x w6[4], u32x w7[4], u64x ipad[8], u64x opad[8], u64x digest[8]) { digest[0] = ipad[0]; digest[1] = ipad[1]; @@ -214,24 +28,40 @@ void hmac_sha256_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[8], u digest[6] = ipad[6]; digest[7] = ipad[7]; - sha256_transform (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; + sha512_transform_vector (w0, w1, w2, w3, w4, w5, w6, w7, digest); + + w0[0] = h32_from_64 (digest[0]); + w0[1] = l32_from_64 (digest[0]); + w0[2] = h32_from_64 (digest[1]); + w0[3] = l32_from_64 (digest[1]); + w1[0] = h32_from_64 (digest[2]); + w1[1] = l32_from_64 (digest[2]); + w1[2] = h32_from_64 (digest[3]); + w1[3] = l32_from_64 (digest[3]); + w2[0] = h32_from_64 (digest[4]); + w2[1] = l32_from_64 (digest[4]); + w2[2] = h32_from_64 (digest[5]); + w2[3] = l32_from_64 (digest[5]); + w3[0] = h32_from_64 (digest[6]); + w3[1] = l32_from_64 (digest[6]); + w3[2] = h32_from_64 (digest[7]); + w3[3] = l32_from_64 (digest[7]); + w4[0] = 0x80000000; + w4[1] = 0; + w4[2] = 0; + w4[3] = 0; + w5[0] = 0; + w5[1] = 0; + w5[2] = 0; + w5[3] = 0; + w6[0] = 0; + w6[1] = 0; + w6[2] = 0; + w6[3] = 0; + w7[0] = 0; + w7[1] = 0; + w7[2] = 0; + w7[3] = (128 + 64) * 8; digest[0] = opad[0]; digest[1] = opad[1]; @@ -242,254 +72,10 @@ void hmac_sha256_run (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], u32 ipad[8], u digest[6] = opad[6]; digest[7] = opad[7]; - sha256_transform (w0, w1, w2, w3, digest); + sha512_transform_vector (w0, w1, w2, w3, w4, w5, w6, w7, digest); } -void sha512_transform (const u64 w[16], u64 dgst[8]) -{ - u64 a = dgst[0]; - u64 b = dgst[1]; - u64 c = dgst[2]; - u64 d = dgst[3]; - u64 e = dgst[4]; - u64 f = dgst[5]; - u64 g = dgst[6]; - u64 h = dgst[7]; - - u64 w0_t = w[ 0]; - u64 w1_t = w[ 1]; - u64 w2_t = w[ 2]; - u64 w3_t = w[ 3]; - u64 w4_t = w[ 4]; - u64 w5_t = w[ 5]; - u64 w6_t = w[ 6]; - u64 w7_t = w[ 7]; - u64 w8_t = w[ 8]; - u64 w9_t = w[ 9]; - u64 wa_t = w[10]; - u64 wb_t = w[11]; - u64 wc_t = w[12]; - u64 wd_t = w[13]; - u64 we_t = w[14]; - u64 wf_t = w[15]; - - #define ROUND512_EXPAND() \ - { \ - w0_t = SHA512_EXPAND (we_t, w9_t, w1_t, w0_t); \ - w1_t = SHA512_EXPAND (wf_t, wa_t, w2_t, w1_t); \ - w2_t = SHA512_EXPAND (w0_t, wb_t, w3_t, w2_t); \ - w3_t = SHA512_EXPAND (w1_t, wc_t, w4_t, w3_t); \ - w4_t = SHA512_EXPAND (w2_t, wd_t, w5_t, w4_t); \ - w5_t = SHA512_EXPAND (w3_t, we_t, w6_t, w5_t); \ - w6_t = SHA512_EXPAND (w4_t, wf_t, w7_t, w6_t); \ - w7_t = SHA512_EXPAND (w5_t, w0_t, w8_t, w7_t); \ - w8_t = SHA512_EXPAND (w6_t, w1_t, w9_t, w8_t); \ - w9_t = SHA512_EXPAND (w7_t, w2_t, wa_t, w9_t); \ - wa_t = SHA512_EXPAND (w8_t, w3_t, wb_t, wa_t); \ - wb_t = SHA512_EXPAND (w9_t, w4_t, wc_t, wb_t); \ - wc_t = SHA512_EXPAND (wa_t, w5_t, wd_t, wc_t); \ - wd_t = SHA512_EXPAND (wb_t, w6_t, we_t, wd_t); \ - we_t = SHA512_EXPAND (wc_t, w7_t, wf_t, we_t); \ - wf_t = SHA512_EXPAND (wd_t, w8_t, w0_t, wf_t); \ - } - - #define ROUND512_STEP(i) \ - { \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha512[i + 0]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha512[i + 1]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha512[i + 2]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha512[i + 3]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha512[i + 4]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha512[i + 5]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha512[i + 6]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha512[i + 7]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha512[i + 8]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha512[i + 9]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha512[i + 10]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha512[i + 11]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha512[i + 12]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha512[i + 13]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, c, d, e, f, g, h, a, b, we_t, k_sha512[i + 14]); \ - SHA512_STEP (SHA512_F0o, SHA512_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha512[i + 15]); \ - } - - ROUND512_STEP (0); - - #ifdef _unroll - #pragma unroll - #endif - for (int i = 16; i < 80; i += 16) - { - ROUND512_EXPAND (); ROUND512_STEP (i); - } - - dgst[0] += a; - dgst[1] += b; - dgst[2] += c; - dgst[3] += d; - dgst[4] += e; - dgst[5] += f; - dgst[6] += g; - dgst[7] += h; -} - -void hmac_sha512_run (const u64 w1[16], const u64 ipad[8], const u64 opad[8], u64 dgst[8]) -{ - dgst[0] = ipad[0]; - dgst[1] = ipad[1]; - dgst[2] = ipad[2]; - dgst[3] = ipad[3]; - dgst[4] = ipad[4]; - dgst[5] = ipad[5]; - dgst[6] = ipad[6]; - dgst[7] = ipad[7]; - - sha512_transform (w1, dgst); - - u64 w[16]; - - w[ 0] = dgst[0]; - w[ 1] = dgst[1]; - w[ 2] = dgst[2]; - w[ 3] = dgst[3]; - w[ 4] = dgst[4]; - w[ 5] = dgst[5]; - w[ 6] = dgst[6]; - w[ 7] = dgst[7]; - w[ 8] = 0x8000000000000000; - w[ 9] = 0; - w[10] = 0; - w[11] = 0; - w[12] = 0; - w[13] = 0; - w[14] = 0; - w[15] = (128 + 64) * 8; - - dgst[0] = opad[0]; - dgst[1] = opad[1]; - dgst[2] = opad[2]; - dgst[3] = opad[3]; - dgst[4] = opad[4]; - dgst[5] = opad[5]; - dgst[6] = opad[6]; - dgst[7] = opad[7]; - - sha512_transform (w, dgst); -} - -void hmac_sha512_run_x (const u64 ipad[8], const u64 opad[8], u64 dgst[8]) -{ - u64 w[16]; - - w[ 0] = dgst[0]; - w[ 1] = dgst[1]; - w[ 2] = dgst[2]; - w[ 3] = dgst[3]; - w[ 4] = dgst[4]; - w[ 5] = dgst[5]; - w[ 6] = dgst[6]; - w[ 7] = dgst[7]; - w[ 8] = 0x8000000000000000; - w[ 9] = 0; - w[10] = 0; - w[11] = 0; - w[12] = 0; - w[13] = 0; - w[14] = 0; - w[15] = (128 + 64) * 8; - - dgst[0] = ipad[0]; - dgst[1] = ipad[1]; - dgst[2] = ipad[2]; - dgst[3] = ipad[3]; - dgst[4] = ipad[4]; - dgst[5] = ipad[5]; - dgst[6] = ipad[6]; - dgst[7] = ipad[7]; - - sha512_transform (w, dgst); - - w[ 0] = dgst[0]; - w[ 1] = dgst[1]; - w[ 2] = dgst[2]; - w[ 3] = dgst[3]; - w[ 4] = dgst[4]; - w[ 5] = dgst[5]; - w[ 6] = dgst[6]; - w[ 7] = dgst[7]; - - dgst[0] = opad[0]; - dgst[1] = opad[1]; - dgst[2] = opad[2]; - dgst[3] = opad[3]; - dgst[4] = opad[4]; - dgst[5] = opad[5]; - dgst[6] = opad[6]; - dgst[7] = opad[7]; - - sha512_transform (w, dgst); -} - -void hmac_sha512_init (u64 w[16], u64 ipad[8], u64 opad[8]) -{ - w[ 0] ^= 0x3636363636363636; - w[ 1] ^= 0x3636363636363636; - w[ 2] ^= 0x3636363636363636; - w[ 3] ^= 0x3636363636363636; - w[ 4] ^= 0x3636363636363636; - w[ 5] ^= 0x3636363636363636; - w[ 6] ^= 0x3636363636363636; - w[ 7] ^= 0x3636363636363636; - w[ 8] ^= 0x3636363636363636; - w[ 9] ^= 0x3636363636363636; - w[10] ^= 0x3636363636363636; - w[11] ^= 0x3636363636363636; - w[12] ^= 0x3636363636363636; - w[13] ^= 0x3636363636363636; - w[14] ^= 0x3636363636363636; - w[15] ^= 0x3636363636363636; - - ipad[0] = SHA512M_A; - ipad[1] = SHA512M_B; - ipad[2] = SHA512M_C; - ipad[3] = SHA512M_D; - ipad[4] = SHA512M_E; - ipad[5] = SHA512M_F; - ipad[6] = SHA512M_G; - ipad[7] = SHA512M_H; - - sha512_transform (w, ipad); - - w[ 0] ^= 0x6a6a6a6a6a6a6a6a; - w[ 1] ^= 0x6a6a6a6a6a6a6a6a; - w[ 2] ^= 0x6a6a6a6a6a6a6a6a; - w[ 3] ^= 0x6a6a6a6a6a6a6a6a; - w[ 4] ^= 0x6a6a6a6a6a6a6a6a; - w[ 5] ^= 0x6a6a6a6a6a6a6a6a; - w[ 6] ^= 0x6a6a6a6a6a6a6a6a; - w[ 7] ^= 0x6a6a6a6a6a6a6a6a; - w[ 8] ^= 0x6a6a6a6a6a6a6a6a; - w[ 9] ^= 0x6a6a6a6a6a6a6a6a; - w[10] ^= 0x6a6a6a6a6a6a6a6a; - w[11] ^= 0x6a6a6a6a6a6a6a6a; - w[12] ^= 0x6a6a6a6a6a6a6a6a; - w[13] ^= 0x6a6a6a6a6a6a6a6a; - w[14] ^= 0x6a6a6a6a6a6a6a6a; - w[15] ^= 0x6a6a6a6a6a6a6a6a; - - opad[0] = SHA512M_A; - opad[1] = SHA512M_B; - opad[2] = SHA512M_C; - opad[3] = SHA512M_D; - opad[4] = SHA512M_E; - opad[5] = SHA512M_F; - opad[6] = SHA512M_G; - opad[7] = SHA512M_H; - - sha512_transform (w, opad); -} - -__kernel void m08200_init (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha512_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 cloudkey_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 u32 gid_max) +__kernel void m08200_init (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha512_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 cloudkey_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 u32 gid_max) { /** * base @@ -500,204 +86,292 @@ __kernel void m08200_init (__global pw_t *pws, __global const kernel_rule_t *rul if (gid >= gid_max) return; u32 w0[4]; + u32 w1[4]; + u32 w2[4]; + u32 w3[4]; + u32 w4[4]; + u32 w5[4]; + u32 w6[4]; + u32 w7[4]; w0[0] = pws[gid].i[ 0]; w0[1] = pws[gid].i[ 1]; w0[2] = pws[gid].i[ 2]; w0[3] = pws[gid].i[ 3]; - - u32 w1[4]; - w1[0] = pws[gid].i[ 4]; w1[1] = pws[gid].i[ 5]; w1[2] = pws[gid].i[ 6]; w1[3] = pws[gid].i[ 7]; - - u32 w2[4]; - w2[0] = pws[gid].i[ 8]; w2[1] = pws[gid].i[ 9]; w2[2] = pws[gid].i[10]; w2[3] = pws[gid].i[11]; - - u32 w3[4]; - w3[0] = pws[gid].i[12]; w3[1] = pws[gid].i[13]; w3[2] = pws[gid].i[14]; w3[3] = pws[gid].i[15]; - - /** - * salt - */ - - u32 salt_buf0[4]; - - salt_buf0[0] = salt_bufs[salt_pos].salt_buf[ 0]; - salt_buf0[1] = salt_bufs[salt_pos].salt_buf[ 1]; - salt_buf0[2] = salt_bufs[salt_pos].salt_buf[ 2]; - salt_buf0[3] = salt_bufs[salt_pos].salt_buf[ 3]; - - u32 salt_buf1[4]; - - salt_buf1[0] = salt_bufs[salt_pos].salt_buf[ 4]; - salt_buf1[1] = salt_bufs[salt_pos].salt_buf[ 5]; - salt_buf1[2] = 0; - salt_buf1[3] = 0; - - const u32 salt_len = salt_bufs[salt_pos].salt_len; - - u64 data_buf[16]; - - data_buf[ 0] = ((u64) swap32 (salt_buf0[ 0])) << 32 | (u64) swap32 (salt_buf0[ 1]); - data_buf[ 1] = ((u64) swap32 (salt_buf0[ 2])) << 32 | (u64) swap32 (salt_buf0[ 3]); - data_buf[ 2] = ((u64) swap32 (salt_buf1[ 0])) << 32 | (u64) swap32 (salt_buf1[ 1]); - data_buf[ 3] = 0; - data_buf[ 4] = 0; - data_buf[ 5] = 0; - data_buf[ 6] = 0; - data_buf[ 7] = 0; - data_buf[ 8] = 0; - data_buf[ 9] = 0; - data_buf[10] = 0; - data_buf[11] = 0; - data_buf[12] = 0; - data_buf[13] = 0; - data_buf[14] = 0; - data_buf[15] = (128 + salt_len + 4) * 8; - - u64 w[16]; - - w[ 0] = ((u64) swap32 (w0[0])) << 32 | (u64) swap32 (w0[1]); - w[ 1] = ((u64) swap32 (w0[2])) << 32 | (u64) swap32 (w0[3]); - w[ 2] = ((u64) swap32 (w1[0])) << 32 | (u64) swap32 (w1[1]); - w[ 3] = ((u64) swap32 (w1[2])) << 32 | (u64) swap32 (w1[3]); - w[ 4] = ((u64) swap32 (w2[0])) << 32 | (u64) swap32 (w2[1]); - w[ 5] = ((u64) swap32 (w2[2])) << 32 | (u64) swap32 (w2[3]); - w[ 6] = ((u64) swap32 (w3[0])) << 32 | (u64) swap32 (w3[1]); - w[ 7] = ((u64) swap32 (w3[2])) << 32 | (u64) swap32 (w3[3]); - w[ 8] = 0; - w[ 9] = 0; - w[10] = 0; - w[11] = 0; - w[12] = 0; - w[13] = 0; - w[14] = 0; - w[15] = 0; - - u64 ipad[8]; - u64 opad[8]; - - hmac_sha512_init (w, ipad, opad); - - tmps[gid].ipad[0] = ipad[0]; - tmps[gid].ipad[1] = ipad[1]; - tmps[gid].ipad[2] = ipad[2]; - tmps[gid].ipad[3] = ipad[3]; - tmps[gid].ipad[4] = ipad[4]; - tmps[gid].ipad[5] = ipad[5]; - tmps[gid].ipad[6] = ipad[6]; - tmps[gid].ipad[7] = ipad[7]; - - tmps[gid].opad[0] = opad[0]; - tmps[gid].opad[1] = opad[1]; - tmps[gid].opad[2] = opad[2]; - tmps[gid].opad[3] = opad[3]; - tmps[gid].opad[4] = opad[4]; - tmps[gid].opad[5] = opad[5]; - tmps[gid].opad[6] = opad[6]; - tmps[gid].opad[7] = opad[7]; + w4[0] = pws[gid].i[16]; + w4[1] = pws[gid].i[17]; + w4[2] = pws[gid].i[18]; + w4[3] = pws[gid].i[19]; + w5[0] = pws[gid].i[20]; + w5[1] = pws[gid].i[21]; + w5[2] = pws[gid].i[22]; + w5[3] = pws[gid].i[23]; + w6[0] = pws[gid].i[24]; + w6[1] = pws[gid].i[25]; + w6[2] = pws[gid].i[26]; + w6[3] = pws[gid].i[27]; + w7[0] = pws[gid].i[28]; + w7[1] = pws[gid].i[29]; + w7[2] = pws[gid].i[30]; + w7[3] = pws[gid].i[31]; + + w0[0] = swap32_S (w0[0]); + w0[1] = swap32_S (w0[1]); + w0[2] = swap32_S (w0[2]); + w0[3] = swap32_S (w0[3]); + w1[0] = swap32_S (w1[0]); + w1[1] = swap32_S (w1[1]); + w1[2] = swap32_S (w1[2]); + w1[3] = swap32_S (w1[3]); + w2[0] = swap32_S (w2[0]); + w2[1] = swap32_S (w2[1]); + w2[2] = swap32_S (w2[2]); + w2[3] = swap32_S (w2[3]); + w3[0] = swap32_S (w3[0]); + w3[1] = swap32_S (w3[1]); + w3[2] = swap32_S (w3[2]); + w3[3] = swap32_S (w3[3]); + w4[0] = swap32_S (w4[0]); + w4[1] = swap32_S (w4[1]); + w4[2] = swap32_S (w4[2]); + w4[3] = swap32_S (w4[3]); + w5[0] = swap32_S (w5[0]); + w5[1] = swap32_S (w5[1]); + w5[2] = swap32_S (w5[2]); + w5[3] = swap32_S (w5[3]); + w6[0] = swap32_S (w6[0]); + w6[1] = swap32_S (w6[1]); + w6[2] = swap32_S (w6[2]); + w6[3] = swap32_S (w6[3]); + w7[0] = swap32_S (w7[0]); + w7[1] = swap32_S (w7[1]); + w7[2] = swap32_S (w7[2]); + w7[3] = swap32_S (w7[3]); + + sha512_hmac_ctx_t sha512_hmac_ctx; + + sha512_hmac_init (&sha512_hmac_ctx, w0, w1, w2, w3, w5, w5, w6, w7); + + tmps[gid].ipad[0] = sha512_hmac_ctx.ipad.h[0]; + tmps[gid].ipad[1] = sha512_hmac_ctx.ipad.h[1]; + tmps[gid].ipad[2] = sha512_hmac_ctx.ipad.h[2]; + tmps[gid].ipad[3] = sha512_hmac_ctx.ipad.h[3]; + tmps[gid].ipad[4] = sha512_hmac_ctx.ipad.h[4]; + tmps[gid].ipad[5] = sha512_hmac_ctx.ipad.h[5]; + tmps[gid].ipad[6] = sha512_hmac_ctx.ipad.h[6]; + tmps[gid].ipad[7] = sha512_hmac_ctx.ipad.h[7]; + + tmps[gid].opad[0] = sha512_hmac_ctx.opad.h[0]; + tmps[gid].opad[1] = sha512_hmac_ctx.opad.h[1]; + tmps[gid].opad[2] = sha512_hmac_ctx.opad.h[2]; + tmps[gid].opad[3] = sha512_hmac_ctx.opad.h[3]; + tmps[gid].opad[4] = sha512_hmac_ctx.opad.h[4]; + tmps[gid].opad[5] = sha512_hmac_ctx.opad.h[5]; + tmps[gid].opad[6] = sha512_hmac_ctx.opad.h[6]; + tmps[gid].opad[7] = sha512_hmac_ctx.opad.h[7]; + + sha512_hmac_update_global_swap (&sha512_hmac_ctx, salt_bufs[salt_pos].salt_buf, salt_bufs[salt_pos].salt_len); for (u32 i = 0, j = 1; i < 8; i += 8, j += 1) { - u64 dgst[8]; - - hmac_sha512_run (data_buf, ipad, opad, dgst); - - tmps[gid].dgst[i + 0] = dgst[0]; - tmps[gid].dgst[i + 1] = dgst[1]; - tmps[gid].dgst[i + 2] = dgst[2]; - tmps[gid].dgst[i + 3] = dgst[3]; - tmps[gid].dgst[i + 4] = dgst[4]; - tmps[gid].dgst[i + 5] = dgst[5]; - tmps[gid].dgst[i + 6] = dgst[6]; - tmps[gid].dgst[i + 7] = dgst[7]; - - tmps[gid].out[i + 0] = dgst[0]; - tmps[gid].out[i + 1] = dgst[1]; - tmps[gid].out[i + 2] = dgst[2]; - tmps[gid].out[i + 3] = dgst[3]; - tmps[gid].out[i + 4] = dgst[4]; - tmps[gid].out[i + 5] = dgst[5]; - tmps[gid].out[i + 6] = dgst[6]; - tmps[gid].out[i + 7] = dgst[7]; + sha512_hmac_ctx_t sha512_hmac_ctx2 = sha512_hmac_ctx; + + 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; + w4[0] = 0; + w4[1] = 0; + w4[2] = 0; + w4[3] = 0; + w5[0] = 0; + w5[1] = 0; + w5[2] = 0; + w5[3] = 0; + w6[0] = 0; + w6[1] = 0; + w6[2] = 0; + w6[3] = 0; + w7[0] = 0; + w7[1] = 0; + w7[2] = 0; + w7[3] = 0; + + sha512_hmac_update_128 (&sha512_hmac_ctx2, w0, w1, w2, w3, w4, w5, w6, w7, 4); + + sha512_hmac_final (&sha512_hmac_ctx2); + + tmps[gid].dgst[i + 0] = sha512_hmac_ctx2.opad.h[0]; + tmps[gid].dgst[i + 1] = sha512_hmac_ctx2.opad.h[1]; + tmps[gid].dgst[i + 2] = sha512_hmac_ctx2.opad.h[2]; + tmps[gid].dgst[i + 3] = sha512_hmac_ctx2.opad.h[3]; + tmps[gid].dgst[i + 4] = sha512_hmac_ctx2.opad.h[4]; + tmps[gid].dgst[i + 5] = sha512_hmac_ctx2.opad.h[5]; + tmps[gid].dgst[i + 6] = sha512_hmac_ctx2.opad.h[6]; + tmps[gid].dgst[i + 7] = sha512_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 m08200_loop (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha512_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 cloudkey_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 u32 gid_max) +__kernel void m08200_loop (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha512_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 cloudkey_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 u32 gid_max) { const u32 gid = get_global_id (0); - if (gid >= gid_max) return; - - u64 ipad[8]; - - ipad[0] = tmps[gid].ipad[0]; - ipad[1] = tmps[gid].ipad[1]; - ipad[2] = tmps[gid].ipad[2]; - ipad[3] = tmps[gid].ipad[3]; - ipad[4] = tmps[gid].ipad[4]; - ipad[5] = tmps[gid].ipad[5]; - ipad[6] = tmps[gid].ipad[6]; - ipad[7] = tmps[gid].ipad[7]; - - u64 opad[8]; - - opad[0] = tmps[gid].opad[0]; - opad[1] = tmps[gid].opad[1]; - opad[2] = tmps[gid].opad[2]; - opad[3] = tmps[gid].opad[3]; - opad[4] = tmps[gid].opad[4]; - opad[5] = tmps[gid].opad[5]; - opad[6] = tmps[gid].opad[6]; - opad[7] = tmps[gid].opad[7]; - - u64 dgst[8]; - - dgst[0] = tmps[gid].dgst[0]; - dgst[1] = tmps[gid].dgst[1]; - dgst[2] = tmps[gid].dgst[2]; - dgst[3] = tmps[gid].dgst[3]; - dgst[4] = tmps[gid].dgst[4]; - dgst[5] = tmps[gid].dgst[5]; - dgst[6] = tmps[gid].dgst[6]; - dgst[7] = tmps[gid].dgst[7]; - - for (u32 j = 0; j < loop_cnt; j++) + if ((gid * VECT_SIZE) >= gid_max) return; + + u64x ipad[8]; + u64x opad[8]; + + ipad[0] = pack64v (tmps, ipad, gid, 0); + ipad[1] = pack64v (tmps, ipad, gid, 1); + ipad[2] = pack64v (tmps, ipad, gid, 2); + ipad[3] = pack64v (tmps, ipad, gid, 3); + ipad[4] = pack64v (tmps, ipad, gid, 4); + ipad[5] = pack64v (tmps, ipad, gid, 5); + ipad[6] = pack64v (tmps, ipad, gid, 6); + ipad[7] = pack64v (tmps, ipad, gid, 7); + + opad[0] = pack64v (tmps, opad, gid, 0); + opad[1] = pack64v (tmps, opad, gid, 1); + opad[2] = pack64v (tmps, opad, gid, 2); + opad[3] = pack64v (tmps, opad, gid, 3); + opad[4] = pack64v (tmps, opad, gid, 4); + opad[5] = pack64v (tmps, opad, gid, 5); + opad[6] = pack64v (tmps, opad, gid, 6); + opad[7] = pack64v (tmps, opad, gid, 7); + + for (u32 i = 0; i < 8; i += 8) { - hmac_sha512_run_x (ipad, opad, dgst); - - tmps[gid].out[0] ^= dgst[0]; - tmps[gid].out[1] ^= dgst[1]; - tmps[gid].out[2] ^= dgst[2]; - tmps[gid].out[3] ^= dgst[3]; - tmps[gid].out[4] ^= dgst[4]; - tmps[gid].out[5] ^= dgst[5]; - tmps[gid].out[6] ^= dgst[6]; - tmps[gid].out[7] ^= dgst[7]; + u64x dgst[8]; + u64x out[8]; + + dgst[0] = pack64v (tmps, dgst, gid, i + 0); + dgst[1] = pack64v (tmps, dgst, gid, i + 1); + dgst[2] = pack64v (tmps, dgst, gid, i + 2); + dgst[3] = pack64v (tmps, dgst, gid, i + 3); + dgst[4] = pack64v (tmps, dgst, gid, i + 4); + dgst[5] = pack64v (tmps, dgst, gid, i + 5); + dgst[6] = pack64v (tmps, dgst, gid, i + 6); + dgst[7] = pack64v (tmps, dgst, gid, i + 7); + + out[0] = pack64v (tmps, out, gid, i + 0); + out[1] = pack64v (tmps, out, gid, i + 1); + out[2] = pack64v (tmps, out, gid, i + 2); + out[3] = pack64v (tmps, out, gid, i + 3); + out[4] = pack64v (tmps, out, gid, i + 4); + out[5] = pack64v (tmps, out, gid, i + 5); + out[6] = pack64v (tmps, out, gid, i + 6); + out[7] = pack64v (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]; + u32x w4[4]; + u32x w5[4]; + u32x w6[4]; + u32x w7[4]; + + w0[0] = h32_from_64 (dgst[0]); + w0[1] = l32_from_64 (dgst[0]); + w0[2] = h32_from_64 (dgst[1]); + w0[3] = l32_from_64 (dgst[1]); + w1[0] = h32_from_64 (dgst[2]); + w1[1] = l32_from_64 (dgst[2]); + w1[2] = h32_from_64 (dgst[3]); + w1[3] = l32_from_64 (dgst[3]); + w2[0] = h32_from_64 (dgst[4]); + w2[1] = l32_from_64 (dgst[4]); + w2[2] = h32_from_64 (dgst[5]); + w2[3] = l32_from_64 (dgst[5]); + w3[0] = h32_from_64 (dgst[6]); + w3[1] = l32_from_64 (dgst[6]); + w3[2] = h32_from_64 (dgst[7]); + w3[3] = l32_from_64 (dgst[7]); + w4[0] = 0x80000000; + w4[1] = 0; + w4[2] = 0; + w4[3] = 0; + w5[0] = 0; + w5[1] = 0; + w5[2] = 0; + w5[3] = 0; + w6[0] = 0; + w6[1] = 0; + w6[2] = 0; + w6[3] = 0; + w7[0] = 0; + w7[1] = 0; + w7[2] = 0; + w7[3] = (128 + 64) * 8; + + hmac_sha512_run_V (w0, w1, w2, w3, w4, w5, w6, w7, 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]; + } + + unpack64v (tmps, dgst, gid, i + 0, dgst[0]); + unpack64v (tmps, dgst, gid, i + 1, dgst[1]); + unpack64v (tmps, dgst, gid, i + 2, dgst[2]); + unpack64v (tmps, dgst, gid, i + 3, dgst[3]); + unpack64v (tmps, dgst, gid, i + 4, dgst[4]); + unpack64v (tmps, dgst, gid, i + 5, dgst[5]); + unpack64v (tmps, dgst, gid, i + 6, dgst[6]); + unpack64v (tmps, dgst, gid, i + 7, dgst[7]); + + unpack64v (tmps, out, gid, i + 0, out[0]); + unpack64v (tmps, out, gid, i + 1, out[1]); + unpack64v (tmps, out, gid, i + 2, out[2]); + unpack64v (tmps, out, gid, i + 3, out[3]); + unpack64v (tmps, out, gid, i + 4, out[4]); + unpack64v (tmps, out, gid, i + 5, out[5]); + unpack64v (tmps, out, gid, i + 6, out[6]); + unpack64v (tmps, out, gid, i + 7, out[7]); } - - tmps[gid].dgst[0] = dgst[0]; - tmps[gid].dgst[1] = dgst[1]; - tmps[gid].dgst[2] = dgst[2]; - tmps[gid].dgst[3] = dgst[3]; - tmps[gid].dgst[4] = dgst[4]; - tmps[gid].dgst[5] = dgst[5]; - tmps[gid].dgst[6] = dgst[6]; - tmps[gid].dgst[7] = dgst[7]; } -__kernel void m08200_comp (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha512_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 cloudkey_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 u32 gid_max) +__kernel void m08200_comp (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global pbkdf2_sha512_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 cloudkey_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 u32 gid_max) { /** * base @@ -731,63 +405,18 @@ __kernel void m08200_comp (__global pw_t *pws, __global const kernel_rule_t *rul w3[2] = 0; w3[3] = 0; - u32 ipad[8]; - u32 opad[8]; - - hmac_sha256_pad (w0, w1, w2, w3, ipad, opad); - - int size = esalt_bufs[digests_offset].data_len; - - int left; - int off; + sha256_hmac_ctx_t ctx; - for (left = size, off = 0; left >= 56; left -= 64, off += 16) - { - w0[0] = esalt_bufs[digests_offset].data_buf[off + 0]; - w0[1] = esalt_bufs[digests_offset].data_buf[off + 1]; - w0[2] = esalt_bufs[digests_offset].data_buf[off + 2]; - w0[3] = esalt_bufs[digests_offset].data_buf[off + 3]; - w1[0] = esalt_bufs[digests_offset].data_buf[off + 4]; - w1[1] = esalt_bufs[digests_offset].data_buf[off + 5]; - w1[2] = esalt_bufs[digests_offset].data_buf[off + 6]; - w1[3] = esalt_bufs[digests_offset].data_buf[off + 7]; - w2[0] = esalt_bufs[digests_offset].data_buf[off + 8]; - w2[1] = esalt_bufs[digests_offset].data_buf[off + 9]; - w2[2] = esalt_bufs[digests_offset].data_buf[off + 10]; - w2[3] = esalt_bufs[digests_offset].data_buf[off + 11]; - w3[0] = esalt_bufs[digests_offset].data_buf[off + 12]; - w3[1] = esalt_bufs[digests_offset].data_buf[off + 13]; - w3[2] = esalt_bufs[digests_offset].data_buf[off + 14]; - w3[3] = esalt_bufs[digests_offset].data_buf[off + 15]; - - sha256_transform (w0, w1, w2, w3, ipad); - } - - w0[0] = esalt_bufs[digests_offset].data_buf[off + 0]; - w0[1] = esalt_bufs[digests_offset].data_buf[off + 1]; - w0[2] = esalt_bufs[digests_offset].data_buf[off + 2]; - w0[3] = esalt_bufs[digests_offset].data_buf[off + 3]; - w1[0] = esalt_bufs[digests_offset].data_buf[off + 4]; - w1[1] = esalt_bufs[digests_offset].data_buf[off + 5]; - w1[2] = esalt_bufs[digests_offset].data_buf[off + 6]; - w1[3] = esalt_bufs[digests_offset].data_buf[off + 7]; - w2[0] = esalt_bufs[digests_offset].data_buf[off + 8]; - w2[1] = esalt_bufs[digests_offset].data_buf[off + 9]; - w2[2] = esalt_bufs[digests_offset].data_buf[off + 10]; - w2[3] = esalt_bufs[digests_offset].data_buf[off + 11]; - w3[0] = esalt_bufs[digests_offset].data_buf[off + 12]; - w3[1] = esalt_bufs[digests_offset].data_buf[off + 13]; - w3[2] = 0; - w3[3] = (64 + size) * 8; + sha256_hmac_init (&ctx, w0, w1, w2, w3); - u32 digest[8]; + sha256_hmac_update_global (&ctx, esalt_bufs[digests_offset].data_buf, esalt_bufs[digests_offset].data_len); - hmac_sha256_run (w0, w1, w2, w3, ipad, opad, digest); + sha256_hmac_final (&ctx); - const u32 r0 = digest[0]; - const u32 r1 = digest[1]; - const u32 r2 = digest[2]; - const u32 r3 = digest[3]; + const u32 r0 = ctx.opad.h[0]; + const u32 r1 = ctx.opad.h[1]; + const u32 r2 = ctx.opad.h[2]; + const u32 r3 = ctx.opad.h[3]; #define il_pos 0 diff --git a/obj/.lock b/obj/.lock deleted file mode 100644 index e69de29bb..000000000 diff --git a/src/interface.c b/src/interface.c index 7c2f74638..a98a2358f 100644 --- a/src/interface.c +++ b/src/interface.c @@ -24698,6 +24698,8 @@ int hashconfig_init (hashcat_ctx_t *hashcat_ctx) break; case 7100: hashconfig->pw_max = 128; // PBKDF2-HMAC-SHA512 max break; + case 8200: hashconfig->pw_max = 128; // PBKDF2-HMAC-SHA512 max + break; case 8500: hashconfig->pw_max = 8; // DES max break; case 8800: hashconfig->pw_max = 64; // PBKDF2-HMAC-SHA1 max