From 8df278fc2014064ff7e96acc860edc65d2747beb Mon Sep 17 00:00:00 2001 From: Jens Steube Date: Sat, 16 Apr 2016 17:30:50 +0200 Subject: [PATCH] Cleanup -m 84xx kernels to latest standard --- OpenCL/m08400_a0.cl | 518 ++++++++++++------------- OpenCL/m08400_a1.cl | 894 ++++++++++++++++++++------------------------ OpenCL/m08400_a3.cl | 76 ++-- 3 files changed, 691 insertions(+), 797 deletions(-) diff --git a/OpenCL/m08400_a0.cl b/OpenCL/m08400_a0.cl index 182fa4e75..6712f58ad 100644 --- a/OpenCL/m08400_a0.cl +++ b/OpenCL/m08400_a0.cl @@ -198,18 +198,16 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, */ u32 pw_buf0[4]; - - pw_buf0[0] = pws[gid].i[ 0]; - pw_buf0[1] = pws[gid].i[ 1]; - pw_buf0[2] = pws[gid].i[ 2]; - pw_buf0[3] = pws[gid].i[ 3]; - u32 pw_buf1[4]; - pw_buf1[0] = pws[gid].i[ 4]; - pw_buf1[1] = pws[gid].i[ 5]; - pw_buf1[2] = pws[gid].i[ 6]; - pw_buf1[3] = pws[gid].i[ 7]; + pw_buf0[0] = pws[gid].i[0]; + pw_buf0[1] = pws[gid].i[1]; + pw_buf0[2] = pws[gid].i[2]; + pw_buf0[3] = pws[gid].i[3]; + pw_buf1[0] = pws[gid].i[4]; + pw_buf1[1] = pws[gid].i[5]; + pw_buf1[2] = pws[gid].i[6]; + pw_buf1[3] = pws[gid].i[7]; const u32 pw_len = pws[gid].pw_len; @@ -218,21 +216,17 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, */ u32 salt_buf0[4]; + u32 salt_buf1[4]; + u32 salt_buf2[4]; salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]); salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]); salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]); salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]); - - u32 salt_buf1[4]; - salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]); salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]); salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]); salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]); - - u32 salt_buf2[4]; - salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]); salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]); salt_buf2[2] = 0; @@ -255,33 +249,26 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, append_0x80_2x4_VV (w0, w1, out_len); - u32x w0_t[4]; - - w0_t[0] = swap32 (w0[0]); - w0_t[1] = swap32 (w0[1]); - w0_t[2] = swap32 (w0[2]); - w0_t[3] = swap32 (w0[3]); - - u32x w1_t[4]; - - w1_t[0] = swap32 (w1[0]); - w1_t[1] = swap32 (w1[1]); - w1_t[2] = swap32 (w1[2]); - w1_t[3] = swap32 (w1[3]); - - u32x w2_t[4]; - - w2_t[0] = swap32 (w2[0]); - w2_t[1] = swap32 (w2[1]); - w2_t[2] = swap32 (w2[2]); - w2_t[3] = swap32 (w2[3]); - - u32x w3_t[4]; - - w3_t[0] = swap32 (w3[0]); - w3_t[1] = swap32 (w3[1]); - w3_t[2] = 0; - w3_t[3] = out_len * 8; + /** + * SHA1 + */ + + w0[0] = swap32 (w0[0]); + w0[1] = swap32 (w0[1]); + w0[2] = swap32 (w0[2]); + w0[3] = swap32 (w0[3]); + w1[0] = swap32 (w1[0]); + w1[1] = swap32 (w1[1]); + w1[2] = swap32 (w1[2]); + w1[3] = swap32 (w1[3]); + w2[0] = swap32 (w2[0]); + w2[1] = swap32 (w2[1]); + w2[2] = swap32 (w2[2]); + w2[3] = swap32 (w2[3]); + w3[0] = swap32 (w3[0]); + w3[1] = swap32 (w3[1]); + w3[2] = 0; + w3[3] = out_len * 8; u32x digest[5]; @@ -291,7 +278,7 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); u32x a; u32x b; @@ -305,28 +292,28 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -334,30 +321,30 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = 80 * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); a = digest[0]; b = digest[1]; @@ -365,28 +352,28 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -394,30 +381,30 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = 80 * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]); } @@ -465,18 +452,16 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, */ u32 pw_buf0[4]; - - pw_buf0[0] = pws[gid].i[ 0]; - pw_buf0[1] = pws[gid].i[ 1]; - pw_buf0[2] = pws[gid].i[ 2]; - pw_buf0[3] = pws[gid].i[ 3]; - u32 pw_buf1[4]; - pw_buf1[0] = pws[gid].i[ 4]; - pw_buf1[1] = pws[gid].i[ 5]; - pw_buf1[2] = pws[gid].i[ 6]; - pw_buf1[3] = pws[gid].i[ 7]; + pw_buf0[0] = pws[gid].i[0]; + pw_buf0[1] = pws[gid].i[1]; + pw_buf0[2] = pws[gid].i[2]; + pw_buf0[3] = pws[gid].i[3]; + pw_buf1[0] = pws[gid].i[4]; + pw_buf1[1] = pws[gid].i[5]; + pw_buf1[2] = pws[gid].i[6]; + pw_buf1[3] = pws[gid].i[7]; const u32 pw_len = pws[gid].pw_len; @@ -485,21 +470,17 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, */ u32 salt_buf0[4]; + u32 salt_buf1[4]; + u32 salt_buf2[4]; salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]); salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]); salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]); salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]); - - u32 salt_buf1[4]; - salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]); salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]); salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]); salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]); - - u32 salt_buf2[4]; - salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]); salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]); salt_buf2[2] = 0; @@ -534,33 +515,26 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, append_0x80_2x4_VV (w0, w1, out_len); - u32x w0_t[4]; - - w0_t[0] = swap32 (w0[0]); - w0_t[1] = swap32 (w0[1]); - w0_t[2] = swap32 (w0[2]); - w0_t[3] = swap32 (w0[3]); - - u32x w1_t[4]; - - w1_t[0] = swap32 (w1[0]); - w1_t[1] = swap32 (w1[1]); - w1_t[2] = swap32 (w1[2]); - w1_t[3] = swap32 (w1[3]); - - u32x w2_t[4]; - - w2_t[0] = swap32 (w2[0]); - w2_t[1] = swap32 (w2[1]); - w2_t[2] = swap32 (w2[2]); - w2_t[3] = swap32 (w2[3]); - - u32x w3_t[4]; - - w3_t[0] = swap32 (w3[0]); - w3_t[1] = swap32 (w3[1]); - w3_t[2] = 0; - w3_t[3] = out_len * 8; + /** + * SHA1 + */ + + w0[0] = swap32 (w0[0]); + w0[1] = swap32 (w0[1]); + w0[2] = swap32 (w0[2]); + w0[3] = swap32 (w0[3]); + w1[0] = swap32 (w1[0]); + w1[1] = swap32 (w1[1]); + w1[2] = swap32 (w1[2]); + w1[3] = swap32 (w1[3]); + w2[0] = swap32 (w2[0]); + w2[1] = swap32 (w2[1]); + w2[2] = swap32 (w2[2]); + w2[3] = swap32 (w2[3]); + w3[0] = swap32 (w3[0]); + w3[1] = swap32 (w3[1]); + w3[2] = 0; + w3[3] = out_len * 8; u32x digest[5]; @@ -570,7 +544,7 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); u32x a; u32x b; @@ -584,28 +558,28 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -613,30 +587,30 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = (salt_len + 40) * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); a = digest[0]; b = digest[1]; @@ -644,28 +618,28 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -673,30 +647,30 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = (salt_len + 40) * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]); } diff --git a/OpenCL/m08400_a1.cl b/OpenCL/m08400_a1.cl index 80312bb96..e5528f8bf 100644 --- a/OpenCL/m08400_a1.cl +++ b/OpenCL/m08400_a1.cl @@ -5,6 +5,8 @@ #define _SHA1_ +#define NEW_SIMD_CODE + #include "include/constants.h" #include "include/kernel_vendor.h" @@ -16,36 +18,44 @@ #include "include/kernel_functions.c" #include "OpenCL/types_ocl.c" #include "OpenCL/common.c" - -#define COMPARE_S "OpenCL/check_single_comp4.c" -#define COMPARE_M "OpenCL/check_multi_comp4.c" - -#define uint_to_hex_lower8_le(i) l_bin2asc[(i)] - -static void sha1_transform (const u32 w0[4], const u32 w1[4], const u32 w2[4], const u32 w3[4], u32 digest[5]) +#include "OpenCL/simd.c" + +#if VECT_SIZE == 1 +#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i)]) +#elif VECT_SIZE == 2 +#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1]) +#elif VECT_SIZE == 4 +#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3]) +#elif VECT_SIZE == 8 +#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3], l_bin2asc[(i).s4], l_bin2asc[(i).s5], l_bin2asc[(i).s6], l_bin2asc[(i).s7]) +#elif VECT_SIZE == 16 +#define uint_to_hex_lower8_le(i) (u32x) (l_bin2asc[(i).s0], l_bin2asc[(i).s1], l_bin2asc[(i).s2], l_bin2asc[(i).s3], l_bin2asc[(i).s4], l_bin2asc[(i).s5], l_bin2asc[(i).s6], l_bin2asc[(i).s7], l_bin2asc[(i).s8], l_bin2asc[(i).s9], l_bin2asc[(i).sa], l_bin2asc[(i).sb], l_bin2asc[(i).sc], l_bin2asc[(i).sd], l_bin2asc[(i).se], l_bin2asc[(i).sf]) +#endif + +static void sha1_transform (const u32x w0[4], const u32x w1[4], const u32x w2[4], const u32x w3[4], u32x digest[5]) { - u32 A = digest[0]; - u32 B = digest[1]; - u32 C = digest[2]; - u32 D = digest[3]; - u32 E = digest[4]; - - 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]; + u32x A = digest[0]; + u32x B = digest[1]; + u32x C = digest[2]; + u32x D = digest[3]; + u32x E = digest[4]; + + u32x w0_t = w0[0]; + u32x w1_t = w0[1]; + u32x w2_t = w0[2]; + u32x w3_t = w0[3]; + u32x w4_t = w1[0]; + u32x w5_t = w1[1]; + u32x w6_t = w1[2]; + u32x w7_t = w1[3]; + u32x w8_t = w2[0]; + u32x w9_t = w2[1]; + u32x wa_t = w2[2]; + u32x wb_t = w2[3]; + u32x wc_t = w3[0]; + u32x wd_t = w3[1]; + u32x we_t = w3[2]; + u32x wf_t = w3[3]; #undef K #define K SHA1C00 @@ -183,65 +193,38 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * base */ - u32 wordl0[4]; - - wordl0[0] = pws[gid].i[ 0]; - wordl0[1] = pws[gid].i[ 1]; - wordl0[2] = pws[gid].i[ 2]; - wordl0[3] = pws[gid].i[ 3]; - - u32 wordl1[4]; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - wordl1[0] = pws[gid].i[ 4]; - wordl1[1] = pws[gid].i[ 5]; - wordl1[2] = pws[gid].i[ 6]; - wordl1[3] = pws[gid].i[ 7]; - - u32 wordl2[4]; - - wordl2[0] = 0; - wordl2[1] = 0; - wordl2[2] = 0; - wordl2[3] = 0; - - u32 wordl3[4]; - - wordl3[0] = 0; - wordl3[1] = 0; - wordl3[2] = 0; - wordl3[3] = 0; + pw_buf0[0] = pws[gid].i[0]; + pw_buf0[1] = pws[gid].i[1]; + pw_buf0[2] = pws[gid].i[2]; + pw_buf0[3] = pws[gid].i[3]; + pw_buf1[0] = pws[gid].i[4]; + pw_buf1[1] = pws[gid].i[5]; + pw_buf1[2] = pws[gid].i[6]; + pw_buf1[3] = pws[gid].i[7]; const u32 pw_l_len = pws[gid].pw_len; - if (combs_mode == COMBINATOR_MODE_BASE_RIGHT) - { - append_0x80_2x4 (wordl0, wordl1, pw_l_len); - - switch_buffer_by_offset_le (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len); - } - /** * salt */ u32 salt_buf0[4]; - - salt_buf0[0] = swap32 (salt_bufs[salt_pos].salt_buf[ 0]); - salt_buf0[1] = swap32 (salt_bufs[salt_pos].salt_buf[ 1]); - salt_buf0[2] = swap32 (salt_bufs[salt_pos].salt_buf[ 2]); - salt_buf0[3] = swap32 (salt_bufs[salt_pos].salt_buf[ 3]); - u32 salt_buf1[4]; - - salt_buf1[0] = swap32 (salt_bufs[salt_pos].salt_buf[ 4]); - salt_buf1[1] = swap32 (salt_bufs[salt_pos].salt_buf[ 5]); - salt_buf1[2] = swap32 (salt_bufs[salt_pos].salt_buf[ 6]); - salt_buf1[3] = swap32 (salt_bufs[salt_pos].salt_buf[ 7]); - u32 salt_buf2[4]; - salt_buf2[0] = swap32 (salt_bufs[salt_pos].salt_buf[ 8]); - salt_buf2[1] = swap32 (salt_bufs[salt_pos].salt_buf[ 9]); + salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]); + salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]); + salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]); + salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]); + salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]); + salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]); + salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]); + salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]); + salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]); + salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]); salt_buf2[2] = 0; salt_buf2[3] = 0; @@ -251,104 +234,97 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * loop */ - for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) + for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE) { - const u32 pw_r_len = combs_buf[il_pos].pw_len; - - const u32 pw_len = pw_l_len + pw_r_len; - - u32 wordr0[4]; - - wordr0[0] = combs_buf[il_pos].i[0]; - wordr0[1] = combs_buf[il_pos].i[1]; - wordr0[2] = combs_buf[il_pos].i[2]; - wordr0[3] = combs_buf[il_pos].i[3]; - - u32 wordr1[4]; - - wordr1[0] = combs_buf[il_pos].i[4]; - wordr1[1] = combs_buf[il_pos].i[5]; - wordr1[2] = combs_buf[il_pos].i[6]; - wordr1[3] = combs_buf[il_pos].i[7]; - - u32 wordr2[4]; - - wordr2[0] = 0; - wordr2[1] = 0; - wordr2[2] = 0; - wordr2[3] = 0; - - u32 wordr3[4]; - - wordr3[0] = 0; - wordr3[1] = 0; - wordr3[2] = 0; - wordr3[3] = 0; + const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos); + + const u32x pw_len = pw_l_len + pw_r_len; + + /** + * concat password candidate + */ + + u32x wordl0[4] = { 0 }; + u32x wordl1[4] = { 0 }; + u32x wordl2[4] = { 0 }; + u32x wordl3[4] = { 0 }; + + wordl0[0] = pw_buf0[0]; + wordl0[1] = pw_buf0[1]; + wordl0[2] = pw_buf0[2]; + wordl0[3] = pw_buf0[3]; + wordl1[0] = pw_buf1[0]; + wordl1[1] = pw_buf1[1]; + wordl1[2] = pw_buf1[2]; + wordl1[3] = pw_buf1[3]; + + u32x wordr0[4] = { 0 }; + u32x wordr1[4] = { 0 }; + u32x wordr2[4] = { 0 }; + u32x wordr3[4] = { 0 }; + + wordr0[0] = ix_create_combt (combs_buf, il_pos, 0); + wordr0[1] = ix_create_combt (combs_buf, il_pos, 1); + wordr0[2] = ix_create_combt (combs_buf, il_pos, 2); + wordr0[3] = ix_create_combt (combs_buf, il_pos, 3); + wordr1[0] = ix_create_combt (combs_buf, il_pos, 4); + wordr1[1] = ix_create_combt (combs_buf, il_pos, 5); + wordr1[2] = ix_create_combt (combs_buf, il_pos, 6); + wordr1[3] = ix_create_combt (combs_buf, il_pos, 7); if (combs_mode == COMBINATOR_MODE_BASE_LEFT) { - append_0x80_2x4 (wordr0, wordr1, pw_r_len); - - switch_buffer_by_offset_le (wordr0, wordr1, wordr2, wordr3, pw_l_len); + switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len); + } + else + { + switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len); } - u32 w0[4]; + u32x w0[4]; + u32x w1[4]; + u32x w2[4]; + u32x w3[4]; w0[0] = wordl0[0] | wordr0[0]; w0[1] = wordl0[1] | wordr0[1]; w0[2] = wordl0[2] | wordr0[2]; w0[3] = wordl0[3] | wordr0[3]; - - u32 w1[4]; - w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; w1[3] = wordl1[3] | wordr1[3]; - - u32 w2[4]; - w2[0] = wordl2[0] | wordr2[0]; w2[1] = wordl2[1] | wordr2[1]; w2[2] = wordl2[2] | wordr2[2]; w2[3] = wordl2[3] | wordr2[3]; - - u32 w3[4]; - w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; w3[2] = wordl3[2] | wordr3[2]; w3[3] = wordl3[3] | wordr3[3]; - u32 w0_t[4]; - - w0_t[0] = swap32 (w0[0]); - w0_t[1] = swap32 (w0[1]); - w0_t[2] = swap32 (w0[2]); - w0_t[3] = swap32 (w0[3]); - - u32 w1_t[4]; - - w1_t[0] = swap32 (w1[0]); - w1_t[1] = swap32 (w1[1]); - w1_t[2] = swap32 (w1[2]); - w1_t[3] = swap32 (w1[3]); - - u32 w2_t[4]; - - w2_t[0] = swap32 (w2[0]); - w2_t[1] = swap32 (w2[1]); - w2_t[2] = swap32 (w2[2]); - w2_t[3] = swap32 (w2[3]); - - u32 w3_t[4]; - - w3_t[0] = swap32 (w3[0]); - w3_t[1] = swap32 (w3[1]); - w3_t[2] = 0; - w3_t[3] = pw_len * 8; - - u32 digest[5]; + /** + * SHA1 + */ + + w0[0] = swap32 (w0[0]); + w0[1] = swap32 (w0[1]); + w0[2] = swap32 (w0[2]); + w0[3] = swap32 (w0[3]); + w1[0] = swap32 (w1[0]); + w1[1] = swap32 (w1[1]); + w1[2] = swap32 (w1[2]); + w1[3] = swap32 (w1[3]); + w2[0] = swap32 (w2[0]); + w2[1] = swap32 (w2[1]); + w2[2] = swap32 (w2[2]); + w2[3] = swap32 (w2[3]); + w3[0] = swap32 (w3[0]); + w3[1] = swap32 (w3[1]); + w3[2] = 0; + w3[3] = pw_len * 8; + + u32x digest[5]; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -356,13 +332,13 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); - u32 a; - u32 b; - u32 c; - u32 d; - u32 e; + u32x a; + u32x b; + u32x c; + u32x d; + u32x e; a = digest[0]; b = digest[1]; @@ -370,28 +346,28 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -399,30 +375,30 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = 80 * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); a = digest[0]; b = digest[1]; @@ -430,28 +406,28 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -459,37 +435,32 @@ __kernel void m08400_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = 80 * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - const u32 r0 = digest[3]; - const u32 r1 = digest[4]; - const u32 r2 = digest[2]; - const u32 r3 = digest[1]; - - #include COMPARE_M + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); + + COMPARE_M_SIMD (digest[3], digest[4], digest[2], digest[1]); } } @@ -534,65 +505,38 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * base */ - u32 wordl0[4]; - - wordl0[0] = pws[gid].i[ 0]; - wordl0[1] = pws[gid].i[ 1]; - wordl0[2] = pws[gid].i[ 2]; - wordl0[3] = pws[gid].i[ 3]; - - u32 wordl1[4]; - - wordl1[0] = pws[gid].i[ 4]; - wordl1[1] = pws[gid].i[ 5]; - wordl1[2] = pws[gid].i[ 6]; - wordl1[3] = pws[gid].i[ 7]; - - u32 wordl2[4]; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - wordl2[0] = 0; - wordl2[1] = 0; - wordl2[2] = 0; - wordl2[3] = 0; - - u32 wordl3[4]; - - wordl3[0] = 0; - wordl3[1] = 0; - wordl3[2] = 0; - wordl3[3] = 0; + pw_buf0[0] = pws[gid].i[0]; + pw_buf0[1] = pws[gid].i[1]; + pw_buf0[2] = pws[gid].i[2]; + pw_buf0[3] = pws[gid].i[3]; + pw_buf1[0] = pws[gid].i[4]; + pw_buf1[1] = pws[gid].i[5]; + pw_buf1[2] = pws[gid].i[6]; + pw_buf1[3] = pws[gid].i[7]; const u32 pw_l_len = pws[gid].pw_len; - if (combs_mode == COMBINATOR_MODE_BASE_RIGHT) - { - append_0x80_2x4 (wordl0, wordl1, pw_l_len); - - switch_buffer_by_offset_le (wordl0, wordl1, wordl2, wordl3, combs_buf[0].pw_len); - } - /** * salt */ u32 salt_buf0[4]; - - salt_buf0[0] = swap32 (salt_bufs[salt_pos].salt_buf[ 0]); - salt_buf0[1] = swap32 (salt_bufs[salt_pos].salt_buf[ 1]); - salt_buf0[2] = swap32 (salt_bufs[salt_pos].salt_buf[ 2]); - salt_buf0[3] = swap32 (salt_bufs[salt_pos].salt_buf[ 3]); - u32 salt_buf1[4]; - - salt_buf1[0] = swap32 (salt_bufs[salt_pos].salt_buf[ 4]); - salt_buf1[1] = swap32 (salt_bufs[salt_pos].salt_buf[ 5]); - salt_buf1[2] = swap32 (salt_bufs[salt_pos].salt_buf[ 6]); - salt_buf1[3] = swap32 (salt_bufs[salt_pos].salt_buf[ 7]); - u32 salt_buf2[4]; - salt_buf2[0] = swap32 (salt_bufs[salt_pos].salt_buf[ 8]); - salt_buf2[1] = swap32 (salt_bufs[salt_pos].salt_buf[ 9]); + salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]); + salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]); + salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]); + salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]); + salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]); + salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]); + salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]); + salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]); + salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]); + salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]); salt_buf2[2] = 0; salt_buf2[3] = 0; @@ -614,104 +558,97 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * loop */ - for (u32 il_pos = 0; il_pos < il_cnt; il_pos++) + for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE) { - const u32 pw_r_len = combs_buf[il_pos].pw_len; - - const u32 pw_len = pw_l_len + pw_r_len; - - u32 wordr0[4]; - - wordr0[0] = combs_buf[il_pos].i[0]; - wordr0[1] = combs_buf[il_pos].i[1]; - wordr0[2] = combs_buf[il_pos].i[2]; - wordr0[3] = combs_buf[il_pos].i[3]; - - u32 wordr1[4]; - - wordr1[0] = combs_buf[il_pos].i[4]; - wordr1[1] = combs_buf[il_pos].i[5]; - wordr1[2] = combs_buf[il_pos].i[6]; - wordr1[3] = combs_buf[il_pos].i[7]; - - u32 wordr2[4]; - - wordr2[0] = 0; - wordr2[1] = 0; - wordr2[2] = 0; - wordr2[3] = 0; - - u32 wordr3[4]; - - wordr3[0] = 0; - wordr3[1] = 0; - wordr3[2] = 0; - wordr3[3] = 0; + const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos); + + const u32x pw_len = pw_l_len + pw_r_len; + + /** + * concat password candidate + */ + + u32x wordl0[4] = { 0 }; + u32x wordl1[4] = { 0 }; + u32x wordl2[4] = { 0 }; + u32x wordl3[4] = { 0 }; + + wordl0[0] = pw_buf0[0]; + wordl0[1] = pw_buf0[1]; + wordl0[2] = pw_buf0[2]; + wordl0[3] = pw_buf0[3]; + wordl1[0] = pw_buf1[0]; + wordl1[1] = pw_buf1[1]; + wordl1[2] = pw_buf1[2]; + wordl1[3] = pw_buf1[3]; + + u32x wordr0[4] = { 0 }; + u32x wordr1[4] = { 0 }; + u32x wordr2[4] = { 0 }; + u32x wordr3[4] = { 0 }; + + wordr0[0] = ix_create_combt (combs_buf, il_pos, 0); + wordr0[1] = ix_create_combt (combs_buf, il_pos, 1); + wordr0[2] = ix_create_combt (combs_buf, il_pos, 2); + wordr0[3] = ix_create_combt (combs_buf, il_pos, 3); + wordr1[0] = ix_create_combt (combs_buf, il_pos, 4); + wordr1[1] = ix_create_combt (combs_buf, il_pos, 5); + wordr1[2] = ix_create_combt (combs_buf, il_pos, 6); + wordr1[3] = ix_create_combt (combs_buf, il_pos, 7); if (combs_mode == COMBINATOR_MODE_BASE_LEFT) { - append_0x80_2x4 (wordr0, wordr1, pw_r_len); - - switch_buffer_by_offset_le (wordr0, wordr1, wordr2, wordr3, pw_l_len); + switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len); + } + else + { + switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len); } - u32 w0[4]; + u32x w0[4]; + u32x w1[4]; + u32x w2[4]; + u32x w3[4]; w0[0] = wordl0[0] | wordr0[0]; w0[1] = wordl0[1] | wordr0[1]; w0[2] = wordl0[2] | wordr0[2]; w0[3] = wordl0[3] | wordr0[3]; - - u32 w1[4]; - w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; w1[3] = wordl1[3] | wordr1[3]; - - u32 w2[4]; - w2[0] = wordl2[0] | wordr2[0]; w2[1] = wordl2[1] | wordr2[1]; w2[2] = wordl2[2] | wordr2[2]; w2[3] = wordl2[3] | wordr2[3]; - - u32 w3[4]; - w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; w3[2] = wordl3[2] | wordr3[2]; w3[3] = wordl3[3] | wordr3[3]; - u32 w0_t[4]; - - w0_t[0] = swap32 (w0[0]); - w0_t[1] = swap32 (w0[1]); - w0_t[2] = swap32 (w0[2]); - w0_t[3] = swap32 (w0[3]); - - u32 w1_t[4]; - - w1_t[0] = swap32 (w1[0]); - w1_t[1] = swap32 (w1[1]); - w1_t[2] = swap32 (w1[2]); - w1_t[3] = swap32 (w1[3]); - - u32 w2_t[4]; - - w2_t[0] = swap32 (w2[0]); - w2_t[1] = swap32 (w2[1]); - w2_t[2] = swap32 (w2[2]); - w2_t[3] = swap32 (w2[3]); - - u32 w3_t[4]; - - w3_t[0] = swap32 (w3[0]); - w3_t[1] = swap32 (w3[1]); - w3_t[2] = 0; - w3_t[3] = pw_len * 8; - - u32 digest[5]; + /** + * SHA1 + */ + + w0[0] = swap32 (w0[0]); + w0[1] = swap32 (w0[1]); + w0[2] = swap32 (w0[2]); + w0[3] = swap32 (w0[3]); + w1[0] = swap32 (w1[0]); + w1[1] = swap32 (w1[1]); + w1[2] = swap32 (w1[2]); + w1[3] = swap32 (w1[3]); + w2[0] = swap32 (w2[0]); + w2[1] = swap32 (w2[1]); + w2[2] = swap32 (w2[2]); + w2[3] = swap32 (w2[3]); + w3[0] = swap32 (w3[0]); + w3[1] = swap32 (w3[1]); + w3[2] = 0; + w3[3] = pw_len * 8; + + u32x digest[5]; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -719,13 +656,13 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); - u32 a; - u32 b; - u32 c; - u32 d; - u32 e; + u32x a; + u32x b; + u32x c; + u32x d; + u32x e; a = digest[0]; b = digest[1]; @@ -733,28 +670,28 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -762,30 +699,30 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = (salt_len + 40) * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); a = digest[0]; b = digest[1]; @@ -793,28 +730,28 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, d = digest[3]; e = digest[4]; - w0_t[0] = salt_buf0[0]; - w0_t[1] = salt_buf0[1]; - w0_t[2] = salt_buf0[2]; - w0_t[3] = salt_buf0[3]; - w1_t[0] = salt_buf1[0]; - w1_t[1] = salt_buf1[1]; - w1_t[2] = salt_buf1[2]; - w1_t[3] = salt_buf1[3]; - w2_t[0] = salt_buf2[0]; - w2_t[1] = salt_buf2[1]; - w2_t[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; - w2_t[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; - w3_t[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; - w3_t[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; - w3_t[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; - w3_t[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; + w0[0] = salt_buf0[0]; + w0[1] = salt_buf0[1]; + w0[2] = salt_buf0[2]; + w0[3] = salt_buf0[3]; + w1[0] = salt_buf1[0]; + w1[1] = salt_buf1[1]; + w1[2] = salt_buf1[2]; + w1[3] = salt_buf1[3]; + w2[0] = salt_buf2[0]; + w2[1] = salt_buf2[1]; + w2[2] = uint_to_hex_lower8_le ((a >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 24) & 255) << 16; + w2[3] = uint_to_hex_lower8_le ((a >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((a >> 8) & 255) << 16; + w3[0] = uint_to_hex_lower8_le ((b >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 24) & 255) << 16; + w3[1] = uint_to_hex_lower8_le ((b >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((b >> 8) & 255) << 16; + w3[2] = uint_to_hex_lower8_le ((c >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 24) & 255) << 16; + w3[3] = uint_to_hex_lower8_le ((c >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((c >> 8) & 255) << 16; digest[0] = SHA1M_A; digest[1] = SHA1M_B; @@ -822,37 +759,32 @@ __kernel void m08400_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, digest[3] = SHA1M_D; digest[4] = SHA1M_E; - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - w0_t[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; - w0_t[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; - w0_t[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; - w0_t[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 - | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; - w1_t[0] = 0x80000000; - w1_t[1] = 0; - w1_t[2] = 0; - w1_t[3] = 0; - w2_t[0] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = (salt_len + 40) * 8; - - sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); - - const u32 r0 = digest[3]; - const u32 r1 = digest[4]; - const u32 r2 = digest[2]; - const u32 r3 = digest[1]; - - #include COMPARE_S + sha1_transform (w0, w1, w2, w3, digest); + + w0[0] = uint_to_hex_lower8_le ((d >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 24) & 255) << 16; + w0[1] = uint_to_hex_lower8_le ((d >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((d >> 8) & 255) << 16; + w0[2] = uint_to_hex_lower8_le ((e >> 16) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 24) & 255) << 16; + w0[3] = uint_to_hex_lower8_le ((e >> 0) & 255) << 0 + | uint_to_hex_lower8_le ((e >> 8) & 255) << 16; + w1[0] = 0x80000000; + 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] = 80 * 8; + + sha1_transform (w0, w1, w2, w3, digest); + + COMPARE_S_SIMD (digest[3], digest[4], digest[2], digest[1]); } } diff --git a/OpenCL/m08400_a3.cl b/OpenCL/m08400_a3.cl index 64e3f36f5..41d5b6315 100644 --- a/OpenCL/m08400_a3.cl +++ b/OpenCL/m08400_a3.cl @@ -177,21 +177,17 @@ static void m08400m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le */ u32 salt_buf0[4]; + u32 salt_buf1[4]; + u32 salt_buf2[4]; salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]); salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]); salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]); salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]); - - u32 salt_buf1[4]; - salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]); salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]); salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]); salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]); - - u32 salt_buf2[4]; - salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]); salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]); salt_buf2[2] = 0; @@ -212,33 +208,31 @@ static void m08400m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le const u32x w0lr = w0l | w0r; u32x w0_t[4]; + u32x w1_t[4]; + u32x w2_t[4]; + u32x w3_t[4]; w0_t[0] = w0lr; w0_t[1] = w0[1]; w0_t[2] = w0[2]; w0_t[3] = w0[3]; - - u32x w1_t[4]; - w1_t[0] = w1[0]; w1_t[1] = w1[1]; w1_t[2] = w1[2]; w1_t[3] = w1[3]; - - u32x w2_t[4]; - w2_t[0] = w2[0]; w2_t[1] = w2[1]; w2_t[2] = w2[2]; w2_t[3] = w2[3]; - - u32x w3_t[4]; - w3_t[0] = w3[0]; w3_t[1] = w3[1]; w3_t[2] = 0; w3_t[3] = pw_len * 8; + /** + * SHA1 + */ + u32x digest[5]; digest[0] = SHA1M_A; @@ -311,7 +305,7 @@ static void m08400m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 0; - w3_t[3] = 80 * 8; + w3_t[3] = (salt_len + 40) * 8; sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); @@ -371,7 +365,7 @@ static void m08400m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le w3_t[0] = 0; w3_t[1] = 0; w3_t[2] = 0; - w3_t[3] = 80 * 8; + w3_t[3] = (salt_len + 40) * 8; sha1_transform (w0_t, w1_t, w2_t, w3_t, digest); @@ -388,38 +382,22 @@ static void m08400s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le const u32 gid = get_global_id (0); const u32 lid = get_local_id (0); - /** - * digest - */ - - const u32 search[4] = - { - digests_buf[digests_offset].digest_buf[DGST_R0], - digests_buf[digests_offset].digest_buf[DGST_R1], - digests_buf[digests_offset].digest_buf[DGST_R2], - digests_buf[digests_offset].digest_buf[DGST_R3] - }; - /** * salt */ u32 salt_buf0[4]; + u32 salt_buf1[4]; + u32 salt_buf2[4]; salt_buf0[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 0]); salt_buf0[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 1]); salt_buf0[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 2]); salt_buf0[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 3]); - - u32 salt_buf1[4]; - salt_buf1[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 4]); salt_buf1[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 5]); salt_buf1[2] = swap32_S (salt_bufs[salt_pos].salt_buf[ 6]); salt_buf1[3] = swap32_S (salt_bufs[salt_pos].salt_buf[ 7]); - - u32 salt_buf2[4]; - salt_buf2[0] = swap32_S (salt_bufs[salt_pos].salt_buf[ 8]); salt_buf2[1] = swap32_S (salt_bufs[salt_pos].salt_buf[ 9]); salt_buf2[2] = 0; @@ -427,6 +405,18 @@ static void m08400s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le const u32 salt_len = salt_bufs[salt_pos].salt_len; + /** + * digest + */ + + const u32 search[4] = + { + digests_buf[digests_offset].digest_buf[DGST_R0], + digests_buf[digests_offset].digest_buf[DGST_R1], + digests_buf[digests_offset].digest_buf[DGST_R2], + digests_buf[digests_offset].digest_buf[DGST_R3] + }; + /** * loop */ @@ -440,33 +430,31 @@ static void m08400s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le const u32x w0lr = w0l | w0r; u32x w0_t[4]; + u32x w1_t[4]; + u32x w2_t[4]; + u32x w3_t[4]; w0_t[0] = w0lr; w0_t[1] = w0[1]; w0_t[2] = w0[2]; w0_t[3] = w0[3]; - - u32x w1_t[4]; - w1_t[0] = w1[0]; w1_t[1] = w1[1]; w1_t[2] = w1[2]; w1_t[3] = w1[3]; - - u32x w2_t[4]; - w2_t[0] = w2[0]; w2_t[1] = w2[1]; w2_t[2] = w2[2]; w2_t[3] = w2[3]; - - u32x w3_t[4]; - w3_t[0] = w3[0]; w3_t[1] = w3[1]; w3_t[2] = 0; w3_t[3] = pw_len * 8; + /** + * SHA1 + */ + u32x digest[5]; digest[0] = SHA1M_A;