diff --git a/OpenCL/m00000_a0.cl b/OpenCL/m00000_a0.cl index 0b946fc44..c44d3e912 100644 --- a/OpenCL/m00000_a0.cl +++ b/OpenCL/m00000_a0.cl @@ -39,18 +39,16 @@ __kernel void m00000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -70,6 +68,7 @@ __kernel void m00000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, append_0x80_2x4 (w0, w1, out_len); w3[2] = out_len * 8; + w3[3] = 0; u32x a = MD5M_A; u32x b = MD5M_B; @@ -173,18 +172,16 @@ __kernel void m00000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -216,6 +213,7 @@ __kernel void m00000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, append_0x80_2x4 (w0, w1, out_len); w3[2] = out_len * 8; + w3[3] = 0; u32x a = MD5M_A; u32x b = MD5M_B; diff --git a/OpenCL/m00000_a1.cl b/OpenCL/m00000_a1.cl index e7e2d58f6..385973e4e 100644 --- a/OpenCL/m00000_a1.cl +++ b/OpenCL/m00000_a1.cl @@ -36,17 +36,17 @@ __kernel void m00000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -69,14 +69,14 @@ __kernel void m00000_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -228,17 +228,17 @@ __kernel void m00000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -273,14 +273,14 @@ __kernel void m00000_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; diff --git a/OpenCL/m00010_a0.cl b/OpenCL/m00010_a0.cl index 85260a952..b769056d4 100644 --- a/OpenCL/m00010_a0.cl +++ b/OpenCL/m00010_a0.cl @@ -39,18 +39,16 @@ __kernel void m00010_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -99,10 +97,10 @@ __kernel void m00010_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * append salt */ - u32x s0[4] = { 0 }; - u32x s1[4] = { 0 }; - u32x s2[4] = { 0 }; - u32x s3[4] = { 0 }; + u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; @@ -248,18 +246,16 @@ __kernel void m00010_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -320,10 +316,10 @@ __kernel void m00010_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * append salt */ - u32x s0[4] = { 0 }; - u32x s1[4] = { 0 }; - u32x s2[4] = { 0 }; - u32x s3[4] = { 0 }; + u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; diff --git a/OpenCL/m00010_a1.cl b/OpenCL/m00010_a1.cl index 3ce73f1c3..d50bd2dab 100644 --- a/OpenCL/m00010_a1.cl +++ b/OpenCL/m00010_a1.cl @@ -36,17 +36,17 @@ __kernel void m00010_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -97,14 +97,14 @@ __kernel void m00010_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -155,10 +155,10 @@ __kernel void m00010_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * append salt */ - u32x s0[4] = { 0 }; - u32x s1[4] = { 0 }; - u32x s2[4] = { 0 }; - u32x s3[4] = { 0 }; + u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; @@ -303,17 +303,17 @@ __kernel void m00010_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -376,14 +376,14 @@ __kernel void m00010_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -434,10 +434,10 @@ __kernel void m00010_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * append salt */ - u32x s0[4] = { 0 }; - u32x s1[4] = { 0 }; - u32x s2[4] = { 0 }; - u32x s3[4] = { 0 }; + u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; diff --git a/OpenCL/m00010_a3.cl b/OpenCL/m00010_a3.cl index f2cc12a18..a96f60bff 100644 --- a/OpenCL/m00010_a3.cl +++ b/OpenCL/m00010_a3.cl @@ -74,6 +74,8 @@ static void m00010m (u32 w[16], const u32 pw_len, __global pw_t *pws, __global k const u32 salt_len = salt_bufs[salt_pos].salt_len; + const u32 pw_salt_len = pw_len + salt_len; + switch_buffer_by_offset_le_S (salt_buf0, salt_buf1, salt_buf2, salt_buf3, pw_len); w[ 0] |= salt_buf0[0]; @@ -90,13 +92,8 @@ static void m00010m (u32 w[16], const u32 pw_len, __global pw_t *pws, __global k w[11] |= salt_buf2[3]; w[12] |= salt_buf3[0]; w[13] |= salt_buf3[1]; - w[14] |= salt_buf3[2]; - w[15] |= salt_buf3[3]; - - const u32 pw_salt_len = pw_len + salt_len; - - w[14] = pw_salt_len * 8; - w[15] = 0; + w[14] = pw_salt_len * 8; + w[15] = 0; /** * base @@ -509,7 +506,7 @@ __kernel void m00010_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w[11] = 0; w[12] = 0; w[13] = 0; - w[14] = pws[gid].i[14]; + w[14] = 0; w[15] = 0; const u32 pw_len = pws[gid].pw_len; @@ -547,7 +544,7 @@ __kernel void m00010_m08 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w[11] = 0; w[12] = 0; w[13] = 0; - w[14] = pws[gid].i[14]; + w[14] = 0; w[15] = 0; const u32 pw_len = pws[gid].pw_len; diff --git a/OpenCL/m00020_a0.cl b/OpenCL/m00020_a0.cl index c10513cf1..dcaaf57e4 100644 --- a/OpenCL/m00020_a0.cl +++ b/OpenCL/m00020_a0.cl @@ -39,18 +39,16 @@ __kernel void m00020_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -101,51 +99,26 @@ __kernel void m00020_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, const u32 out_salt_len = out_len + salt_len; - u32x w0_t[4]; - u32x w1_t[4]; - u32x w2_t[4]; - u32x w3_t[4]; + switch_buffer_by_offset_le (w0, w1, w2, w3, salt_len); - w0_t[0] = w0[0]; - w0_t[1] = w0[1]; - w0_t[2] = w0[2]; - w0_t[3] = w0[3]; - w1_t[0] = w1[0]; - w1_t[1] = w1[1]; - w1_t[2] = w1[2]; - w1_t[3] = w1[3]; - w2_t[0] = w2[0]; - w2_t[1] = w2[1]; - w2_t[2] = w2[2]; - w2_t[3] = w2[3]; - w3_t[0] = w3[0]; - w3_t[1] = w3[1]; - w3_t[2] = w3[2]; - w3_t[3] = w3[3]; + 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] |= salt_buf2[2]; + w2[3] |= salt_buf2[3]; + w3[0] |= salt_buf3[0]; + w3[1] |= salt_buf3[1]; + w3[2] = out_salt_len * 8; + w3[3] = 0; - switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len); - - 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] |= salt_buf2[2]; - w2_t[3] |= salt_buf2[3]; - w3_t[0] |= salt_buf3[0]; - w3_t[1] |= salt_buf3[1]; - w3_t[2] |= salt_buf3[2]; - w3_t[3] |= salt_buf3[3]; - - append_0x80_4x4 (w0_t, w1_t, w2_t, w3_t, out_salt_len); - - w3_t[2] = out_salt_len * 8; - w3_t[3] = 0; + append_0x80_4x4 (w0, w1, w2, w3, out_salt_len); /** * md5 @@ -156,73 +129,73 @@ __kernel void m00020_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x c = MD5M_C; u32x d = MD5M_D; - MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); COMPARE_M_SIMD (a, d, c, b); } @@ -253,18 +226,16 @@ __kernel void m00020_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -327,51 +298,26 @@ __kernel void m00020_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, const u32 out_salt_len = out_len + salt_len; - u32x w0_t[4]; - u32x w1_t[4]; - u32x w2_t[4]; - u32x w3_t[4]; + switch_buffer_by_offset_le (w0, w1, w2, w3, salt_len); - w0_t[0] = w0[0]; - w0_t[1] = w0[1]; - w0_t[2] = w0[2]; - w0_t[3] = w0[3]; - w1_t[0] = w1[0]; - w1_t[1] = w1[1]; - w1_t[2] = w1[2]; - w1_t[3] = w1[3]; - w2_t[0] = w2[0]; - w2_t[1] = w2[1]; - w2_t[2] = w2[2]; - w2_t[3] = w2[3]; - w3_t[0] = w3[0]; - w3_t[1] = w3[1]; - w3_t[2] = w3[2]; - w3_t[3] = w3[3]; + 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] |= salt_buf2[2]; + w2[3] |= salt_buf2[3]; + w3[0] |= salt_buf3[0]; + w3[1] |= salt_buf3[1]; + w3[2] = out_salt_len * 8; + w3[3] = 0; - switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len); - - 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] |= salt_buf2[2]; - w2_t[3] |= salt_buf2[3]; - w3_t[0] |= salt_buf3[0]; - w3_t[1] |= salt_buf3[1]; - w3_t[2] |= salt_buf3[2]; - w3_t[3] |= salt_buf3[3]; - - append_0x80_4x4 (w0_t, w1_t, w2_t, w3_t, out_salt_len); - - w3_t[2] = out_salt_len * 8; - w3_t[3] = 0; + append_0x80_4x4 (w0, w1, w2, w3, out_salt_len); /** * md5 @@ -382,76 +328,76 @@ __kernel void m00020_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x c = MD5M_C; u32x d = MD5M_D; - MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); if (MATCHES_NONE_VS (a, search[0])) continue; - MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); + MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); COMPARE_S_SIMD (a, d, c, b); } diff --git a/OpenCL/m00020_a1.cl b/OpenCL/m00020_a1.cl index 86ca18e64..f70fadc95 100644 --- a/OpenCL/m00020_a1.cl +++ b/OpenCL/m00020_a1.cl @@ -36,17 +36,17 @@ __kernel void m00020_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -97,14 +97,14 @@ __kernel void m00020_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -148,8 +148,8 @@ __kernel void m00020_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w2[3] = wordl2[3] | wordr2[3]; w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; - w3[2] = 0; - w3[3] = 0; + w3[2] = wordl3[2] | wordr3[2]; + w3[3] = wordl3[3] | wordr3[3]; /** * prepend salt @@ -281,17 +281,17 @@ __kernel void m00020_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -354,14 +354,14 @@ __kernel void m00020_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -405,8 +405,8 @@ __kernel void m00020_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w2[3] = wordl2[3] | wordr2[3]; w3[0] = wordl3[0] | wordr3[0]; w3[1] = wordl3[1] | wordr3[1]; - w3[2] = 0; - w3[3] = 0; + w3[2] = wordl3[2] | wordr3[2]; + w3[3] = wordl3[3] | wordr3[3]; /** * prepend salt diff --git a/OpenCL/m00020_a3.cl b/OpenCL/m00020_a3.cl index 742905c98..cd9e7ba43 100644 --- a/OpenCL/m00020_a3.cl +++ b/OpenCL/m00020_a3.cl @@ -81,8 +81,8 @@ static void m00020m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le w2[3] |= salt_buf2[3]; w3[0] |= salt_buf3[0]; w3[1] |= salt_buf3[1]; - w3[2] |= salt_buf3[2]; - w3[3] |= salt_buf3[3]; + w3[2] = pw_salt_len * 8; + w3[3] = 0; /** * loop @@ -113,8 +113,8 @@ static void m00020m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le w2_t[3] = w2[3]; w3_t[0] = w3[0]; w3_t[1] = w3[1]; - w3_t[2] = pw_salt_len * 8; - w3_t[3] = 0; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; overwrite_at_le_4x4 (w0_t, w1_t, w2_t, w3_t, w0lr, salt_len); @@ -272,8 +272,8 @@ static void m00020s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le w2[3] |= salt_buf2[3]; w3[0] |= salt_buf3[0]; w3[1] |= salt_buf3[1]; - w3[2] |= salt_buf3[2]; - w3[3] |= salt_buf3[3]; + w3[2] = pw_salt_len * 8; + w3[3] = 0; /** * loop @@ -304,8 +304,8 @@ static void m00020s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le w2_t[3] = w2[3]; w3_t[0] = w3[0]; w3_t[1] = w3[1]; - w3_t[2] = pw_salt_len * 8; - w3_t[3] = 0; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; overwrite_at_le_4x4 (w0_t, w1_t, w2_t, w3_t, w0lr, salt_len); diff --git a/OpenCL/m00030_a0.cl b/OpenCL/m00030_a0.cl index 7e043ea5b..93dc945d2 100644 --- a/OpenCL/m00030_a0.cl +++ b/OpenCL/m00030_a0.cl @@ -41,18 +41,16 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -97,66 +95,57 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1); + make_unicode (w1, w2, w3); + make_unicode (w0, w0, w1); + + const u32 out_len2 = out_len * 2; + /** * append salt */ u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; s0[2] = salt_buf0[2]; s0[3] = salt_buf0[3]; - - u32x s1[4]; - s1[0] = salt_buf1[0]; s1[1] = salt_buf1[1]; s1[2] = salt_buf1[2]; s1[3] = salt_buf1[3]; + s2[0] = salt_buf2[0]; + s2[1] = salt_buf2[1]; + s2[2] = salt_buf2[2]; + s2[3] = salt_buf2[3]; + s3[0] = salt_buf3[0]; + s3[1] = salt_buf3[1]; + s3[2] = salt_buf3[2]; + s3[3] = salt_buf3[3]; - u32x s2[4]; + switch_buffer_by_offset_le (s0, s1, s2, s3, out_len2); - s2[0] = 0; - s2[1] = 0; - s2[2] = 0; - s2[3] = 0; + const u32 pw_salt_len = out_len2 + salt_len; - u32x s3[4]; - - s3[0] = 0; - s3[1] = 0; - s3[2] = 0; - s3[3] = 0; - - switch_buffer_by_offset_le (s0, s1, s2, s3, (out_len * 2)); - - const u32 out_salt_len = (out_len * 2) + salt_len; - - u32x w0_t[4]; - u32x w1_t[4]; - u32x w2_t[4]; - u32x w3_t[4]; - - make_unicode (w0, w0_t, w1_t); - make_unicode (w1, w2_t, w3_t); - - w0_t[0] |= s0[0]; - w0_t[1] |= s0[1]; - w0_t[2] |= s0[2]; - w0_t[3] |= s0[3]; - w1_t[0] |= s1[0]; - w1_t[1] |= s1[1]; - w1_t[2] |= s1[2]; - w1_t[3] |= s1[3]; - w2_t[0] |= s2[0]; - w2_t[1] |= s2[1]; - w2_t[2] |= s2[2]; - w2_t[3] |= s2[3]; - w3_t[0] |= s3[0]; - w3_t[1] |= s3[1]; - w3_t[2] = out_salt_len * 8; - w3_t[3] = 0; + w0[0] |= s0[0]; + w0[1] |= s0[1]; + w0[2] |= s0[2]; + w0[3] |= s0[3]; + w1[0] |= s1[0]; + w1[1] |= s1[1]; + w1[2] |= s1[2]; + w1[3] |= s1[3]; + w2[0] |= s2[0]; + w2[1] |= s2[1]; + w2[2] |= s2[2]; + w2[3] |= s2[3]; + w3[0] |= s3[0]; + w3[1] |= s3[1]; + w3[2] = pw_salt_len * 8; + w3[3] = 0; /** * md5 @@ -167,73 +156,73 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x c = MD5M_C; u32x d = MD5M_D; - MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); COMPARE_M_SIMD (a, d, c, b); } @@ -264,18 +253,16 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -332,66 +319,57 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1); + make_unicode (w1, w2, w3); + make_unicode (w0, w0, w1); + + const u32 out_len2 = out_len * 2; + /** * append salt */ u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; s0[2] = salt_buf0[2]; s0[3] = salt_buf0[3]; - - u32x s1[4]; - s1[0] = salt_buf1[0]; s1[1] = salt_buf1[1]; s1[2] = salt_buf1[2]; s1[3] = salt_buf1[3]; + s2[0] = salt_buf2[0]; + s2[1] = salt_buf2[1]; + s2[2] = salt_buf2[2]; + s2[3] = salt_buf2[3]; + s3[0] = salt_buf3[0]; + s3[1] = salt_buf3[1]; + s3[2] = salt_buf3[2]; + s3[3] = salt_buf3[3]; - u32x s2[4]; + switch_buffer_by_offset_le (s0, s1, s2, s3, out_len2); - s2[0] = 0; - s2[1] = 0; - s2[2] = 0; - s2[3] = 0; + const u32 pw_salt_len = out_len2 + salt_len; - u32x s3[4]; - - s3[0] = 0; - s3[1] = 0; - s3[2] = 0; - s3[3] = 0; - - switch_buffer_by_offset_le (s0, s1, s2, s3, (out_len * 2)); - - const u32 out_salt_len = (out_len * 2) + salt_len; - - u32x w0_t[4]; - u32x w1_t[4]; - u32x w2_t[4]; - u32x w3_t[4]; - - make_unicode (w0, w0_t, w1_t); - make_unicode (w1, w2_t, w3_t); - - w0_t[0] |= s0[0]; - w0_t[1] |= s0[1]; - w0_t[2] |= s0[2]; - w0_t[3] |= s0[3]; - w1_t[0] |= s1[0]; - w1_t[1] |= s1[1]; - w1_t[2] |= s1[2]; - w1_t[3] |= s1[3]; - w2_t[0] |= s2[0]; - w2_t[1] |= s2[1]; - w2_t[2] |= s2[2]; - w2_t[3] |= s2[3]; - w3_t[0] |= s3[0]; - w3_t[1] |= s3[1]; - w3_t[2] = out_salt_len * 8; - w3_t[3] = 0; + w0[0] |= s0[0]; + w0[1] |= s0[1]; + w0[2] |= s0[2]; + w0[3] |= s0[3]; + w1[0] |= s1[0]; + w1[1] |= s1[1]; + w1[2] |= s1[2]; + w1[3] |= s1[3]; + w2[0] |= s2[0]; + w2[1] |= s2[1]; + w2[2] |= s2[2]; + w2[3] |= s2[3]; + w3[0] |= s3[0]; + w3[1] |= s3[1]; + w3[2] = pw_salt_len * 8; + w3[3] = 0; /** * md5 @@ -402,76 +380,76 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x c = MD5M_C; u32x d = MD5M_D; - MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); if (MATCHES_NONE_VS (a, search[0])) continue; - MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); + MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); COMPARE_S_SIMD (a, d, c, b); } diff --git a/OpenCL/m00030_a1.cl b/OpenCL/m00030_a1.cl index 2113be11b..d17d9fdf1 100644 --- a/OpenCL/m00030_a1.cl +++ b/OpenCL/m00030_a1.cl @@ -38,17 +38,17 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -99,14 +99,14 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -143,15 +143,15 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; - 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; + w1[3] = wordl1[3] | wordr1[3]; + w2[0] = wordl2[0] | wordr2[0]; + w2[1] = wordl2[1] | wordr2[1]; + w2[2] = wordl2[2] | wordr2[2]; + w2[3] = wordl2[3] | wordr2[3]; + w3[0] = wordl3[0] | wordr3[0]; + w3[1] = wordl3[1] | wordr3[1]; + w3[2] = wordl3[2] | wordr3[2]; + w3[3] = wordl3[3] | wordr3[3]; make_unicode (w1, w2, w3); make_unicode (w0, w0, w1); @@ -162,10 +162,10 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * append salt */ - u32x s0[4] = { 0 }; - u32x s1[4] = { 0 }; - u32x s2[4] = { 0 }; - u32x s3[4] = { 0 }; + u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; @@ -175,6 +175,14 @@ __kernel void m00030_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, s1[1] = salt_buf1[1]; s1[2] = salt_buf1[2]; s1[3] = salt_buf1[3]; + s2[0] = salt_buf2[0]; + s2[1] = salt_buf2[1]; + s2[2] = salt_buf2[2]; + s2[3] = salt_buf2[3]; + s3[0] = salt_buf3[0]; + s3[1] = salt_buf3[1]; + s3[2] = salt_buf3[2]; + s3[3] = salt_buf3[3]; switch_buffer_by_offset_le_VV (s0, s1, s2, s3, pw_len2); @@ -302,17 +310,17 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -375,14 +383,14 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -419,15 +427,15 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; - 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; + w1[3] = wordl1[3] | wordr1[3]; + w2[0] = wordl2[0] | wordr2[0]; + w2[1] = wordl2[1] | wordr2[1]; + w2[2] = wordl2[2] | wordr2[2]; + w2[3] = wordl2[3] | wordr2[3]; + w3[0] = wordl3[0] | wordr3[0]; + w3[1] = wordl3[1] | wordr3[1]; + w3[2] = wordl3[2] | wordr3[2]; + w3[3] = wordl3[3] | wordr3[3]; make_unicode (w1, w2, w3); make_unicode (w0, w0, w1); @@ -438,10 +446,10 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, * append salt */ - u32x s0[4] = { 0 }; - u32x s1[4] = { 0 }; - u32x s2[4] = { 0 }; - u32x s3[4] = { 0 }; + u32x s0[4]; + u32x s1[4]; + u32x s2[4]; + u32x s3[4]; s0[0] = salt_buf0[0]; s0[1] = salt_buf0[1]; @@ -451,6 +459,14 @@ __kernel void m00030_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, s1[1] = salt_buf1[1]; s1[2] = salt_buf1[2]; s1[3] = salt_buf1[3]; + s2[0] = salt_buf2[0]; + s2[1] = salt_buf2[1]; + s2[2] = salt_buf2[2]; + s2[3] = salt_buf2[3]; + s3[0] = salt_buf3[0]; + s3[1] = salt_buf3[1]; + s3[2] = salt_buf3[2]; + s3[3] = salt_buf3[3]; switch_buffer_by_offset_le_VV (s0, s1, s2, s3, pw_len2); diff --git a/OpenCL/m00030_a3.cl b/OpenCL/m00030_a3.cl index a79bfc392..82869bfb3 100644 --- a/OpenCL/m00030_a3.cl +++ b/OpenCL/m00030_a3.cl @@ -74,6 +74,8 @@ static void m00030m (u32 w[16], const u32 pw_len, __global pw_t *pws, __global k const u32 salt_len = salt_bufs[salt_pos].salt_len; + const u32 pw_salt_len = pw_len + salt_len; + switch_buffer_by_offset_le_S (salt_buf0, salt_buf1, salt_buf2, salt_buf3, pw_len); w[ 0] |= salt_buf0[0]; @@ -90,13 +92,8 @@ static void m00030m (u32 w[16], const u32 pw_len, __global pw_t *pws, __global k w[11] |= salt_buf2[3]; w[12] |= salt_buf3[0]; w[13] |= salt_buf3[1]; - w[14] |= salt_buf3[2]; - w[15] |= salt_buf3[3]; - - const u32 pw_salt_len = pw_len + salt_len; - - w[14] = pw_salt_len * 8; - w[15] = 0; + w[14] = pw_salt_len * 8; + w[15] = 0; /** * base diff --git a/OpenCL/m00040_a0.cl b/OpenCL/m00040_a0.cl index 6c41a85d8..d36ca1898 100644 --- a/OpenCL/m00040_a0.cl +++ b/OpenCL/m00040_a0.cl @@ -39,18 +39,16 @@ __kernel void m00040_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -95,43 +93,37 @@ __kernel void m00040_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1); + make_unicode (w1, w2, w3); + make_unicode (w0, w0, w1); + + const u32 out_len2 = out_len * 2; + /** * prepend salt */ - const u32 out_salt_len = (out_len * 2) + salt_len; + const u32 out_salt_len = out_len2 + salt_len; - u32x w0_t[4]; - u32x w1_t[4]; - u32x w2_t[4]; - u32x w3_t[4]; + switch_buffer_by_offset_le (w0, w1, w2, w3, salt_len); - make_unicode (w0, w0_t, w1_t); - make_unicode (w1, w2_t, w3_t); + 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] |= salt_buf2[2]; + w2[3] |= salt_buf2[3]; + w3[0] |= salt_buf3[0]; + w3[1] |= salt_buf3[1]; + w3[2] = out_salt_len * 8; + w3[3] = 0; - switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len); - - 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] |= salt_buf2[2]; - w2_t[3] |= salt_buf2[3]; - w3_t[0] |= salt_buf3[0]; - w3_t[1] |= salt_buf3[1]; - w3_t[2] |= salt_buf3[2]; - w3_t[3] |= salt_buf3[3]; - - append_0x80_4x4 (w0_t, w1_t, w2_t, w3_t, out_salt_len); - - w3_t[2] = out_salt_len * 8; - w3_t[3] = 0; + append_0x80_4x4 (w0, w1, w2, w3, out_salt_len); /** * md5 @@ -142,73 +134,73 @@ __kernel void m00040_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x c = MD5M_C; u32x d = MD5M_D; - MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); COMPARE_M_SIMD (a, d, c, b); } @@ -239,18 +231,16 @@ __kernel void m00040_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; 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; @@ -307,43 +297,37 @@ __kernel void m00040_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, const u32 out_len = apply_rules_vect (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1); + make_unicode (w1, w2, w3); + make_unicode (w0, w0, w1); + + const u32 out_len2 = out_len * 2; + /** * prepend salt */ - const u32 out_salt_len = (out_len * 2) + salt_len; + const u32 out_salt_len = out_len2 + salt_len; - u32x w0_t[4]; - u32x w1_t[4]; - u32x w2_t[4]; - u32x w3_t[4]; + switch_buffer_by_offset_le (w0, w1, w2, w3, salt_len); - make_unicode (w0, w0_t, w1_t); - make_unicode (w1, w2_t, w3_t); + 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] |= salt_buf2[2]; + w2[3] |= salt_buf2[3]; + w3[0] |= salt_buf3[0]; + w3[1] |= salt_buf3[1]; + w3[2] = out_salt_len * 8; + w3[3] = 0; - switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, salt_len); - - 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] |= salt_buf2[2]; - w2_t[3] |= salt_buf2[3]; - w3_t[0] |= salt_buf3[0]; - w3_t[1] |= salt_buf3[1]; - w3_t[2] |= salt_buf3[2]; - w3_t[3] |= salt_buf3[3]; - - append_0x80_4x4 (w0_t, w1_t, w2_t, w3_t, out_salt_len); - - w3_t[2] = out_salt_len * 8; - w3_t[3] = 0; + append_0x80_4x4 (w0, w1, w2, w3, out_salt_len); /** * md5 @@ -354,76 +338,76 @@ __kernel void m00040_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x c = MD5M_C; u32x d = MD5M_D; - MD5_STEP (MD5_Fo, a, b, c, d, w0_t[0], MD5C00, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w0_t[1], MD5C01, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w0_t[2], MD5C02, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w0_t[3], MD5C03, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w1_t[0], MD5C04, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w1_t[1], MD5C05, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w1_t[2], MD5C06, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w1_t[3], MD5C07, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w2_t[0], MD5C08, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w2_t[1], MD5C09, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w2_t[2], MD5C0a, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w2_t[3], MD5C0b, MD5S03); - MD5_STEP (MD5_Fo, a, b, c, d, w3_t[0], MD5C0c, MD5S00); - MD5_STEP (MD5_Fo, d, a, b, c, w3_t[1], MD5C0d, MD5S01); - MD5_STEP (MD5_Fo, c, d, a, b, w3_t[2], MD5C0e, MD5S02); - MD5_STEP (MD5_Fo, b, c, d, a, w3_t[3], MD5C0f, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w0[0], MD5C00, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w0[1], MD5C01, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w0[2], MD5C02, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w0[3], MD5C03, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w1[0], MD5C04, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w1[1], MD5C05, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w1[2], MD5C06, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w1[3], MD5C07, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w2[0], MD5C08, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w2[1], MD5C09, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w2[2], MD5C0a, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w2[3], MD5C0b, MD5S03); + MD5_STEP (MD5_Fo, a, b, c, d, w3[0], MD5C0c, MD5S00); + MD5_STEP (MD5_Fo, d, a, b, c, w3[1], MD5C0d, MD5S01); + MD5_STEP (MD5_Fo, c, d, a, b, w3[2], MD5C0e, MD5S02); + MD5_STEP (MD5_Fo, b, c, d, a, w3[3], MD5C0f, MD5S03); - MD5_STEP (MD5_Go, a, b, c, d, w0_t[1], MD5C10, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w1_t[2], MD5C11, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w2_t[3], MD5C12, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w0_t[0], MD5C13, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w1_t[1], MD5C14, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w2_t[2], MD5C15, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w3_t[3], MD5C16, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w1_t[0], MD5C17, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w2_t[1], MD5C18, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w3_t[2], MD5C19, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w0_t[3], MD5C1a, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w2_t[0], MD5C1b, MD5S13); - MD5_STEP (MD5_Go, a, b, c, d, w3_t[1], MD5C1c, MD5S10); - MD5_STEP (MD5_Go, d, a, b, c, w0_t[2], MD5C1d, MD5S11); - MD5_STEP (MD5_Go, c, d, a, b, w1_t[3], MD5C1e, MD5S12); - MD5_STEP (MD5_Go, b, c, d, a, w3_t[0], MD5C1f, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w0[1], MD5C10, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w1[2], MD5C11, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w2[3], MD5C12, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w0[0], MD5C13, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w1[1], MD5C14, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w2[2], MD5C15, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w3[3], MD5C16, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w1[0], MD5C17, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w2[1], MD5C18, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w3[2], MD5C19, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w0[3], MD5C1a, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w2[0], MD5C1b, MD5S13); + MD5_STEP (MD5_Go, a, b, c, d, w3[1], MD5C1c, MD5S10); + MD5_STEP (MD5_Go, d, a, b, c, w0[2], MD5C1d, MD5S11); + MD5_STEP (MD5_Go, c, d, a, b, w1[3], MD5C1e, MD5S12); + MD5_STEP (MD5_Go, b, c, d, a, w3[0], MD5C1f, MD5S13); - MD5_STEP (MD5_H , a, b, c, d, w1_t[1], MD5C20, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w2_t[0], MD5C21, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w2_t[3], MD5C22, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w3_t[2], MD5C23, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w0_t[1], MD5C24, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w1_t[0], MD5C25, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w1_t[3], MD5C26, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w2_t[2], MD5C27, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w3_t[1], MD5C28, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w0_t[0], MD5C29, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w0_t[3], MD5C2a, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w1_t[2], MD5C2b, MD5S23); - MD5_STEP (MD5_H , a, b, c, d, w2_t[1], MD5C2c, MD5S20); - MD5_STEP (MD5_H , d, a, b, c, w3_t[0], MD5C2d, MD5S21); - MD5_STEP (MD5_H , c, d, a, b, w3_t[3], MD5C2e, MD5S22); - MD5_STEP (MD5_H , b, c, d, a, w0_t[2], MD5C2f, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w1[1], MD5C20, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w2[0], MD5C21, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w2[3], MD5C22, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w3[2], MD5C23, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w0[1], MD5C24, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w1[0], MD5C25, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w1[3], MD5C26, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w2[2], MD5C27, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w3[1], MD5C28, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w0[0], MD5C29, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w0[3], MD5C2a, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w1[2], MD5C2b, MD5S23); + MD5_STEP (MD5_H , a, b, c, d, w2[1], MD5C2c, MD5S20); + MD5_STEP (MD5_H , d, a, b, c, w3[0], MD5C2d, MD5S21); + MD5_STEP (MD5_H , c, d, a, b, w3[3], MD5C2e, MD5S22); + MD5_STEP (MD5_H , b, c, d, a, w0[2], MD5C2f, MD5S23); - MD5_STEP (MD5_I , a, b, c, d, w0_t[0], MD5C30, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w1_t[3], MD5C31, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w3_t[2], MD5C32, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w1_t[1], MD5C33, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w3_t[0], MD5C34, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w0_t[3], MD5C35, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w2_t[2], MD5C36, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w0_t[1], MD5C37, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w2_t[0], MD5C38, MD5S30); - MD5_STEP (MD5_I , d, a, b, c, w3_t[3], MD5C39, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w1_t[2], MD5C3a, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w3_t[1], MD5C3b, MD5S33); - MD5_STEP (MD5_I , a, b, c, d, w1_t[0], MD5C3c, MD5S30); + MD5_STEP (MD5_I , a, b, c, d, w0[0], MD5C30, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w1[3], MD5C31, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w3[2], MD5C32, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w1[1], MD5C33, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w3[0], MD5C34, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w0[3], MD5C35, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w2[2], MD5C36, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w0[1], MD5C37, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w2[0], MD5C38, MD5S30); + MD5_STEP (MD5_I , d, a, b, c, w3[3], MD5C39, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w1[2], MD5C3a, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w3[1], MD5C3b, MD5S33); + MD5_STEP (MD5_I , a, b, c, d, w1[0], MD5C3c, MD5S30); if (MATCHES_NONE_VS (a, search[0])) continue; - MD5_STEP (MD5_I , d, a, b, c, w2_t[3], MD5C3d, MD5S31); - MD5_STEP (MD5_I , c, d, a, b, w0_t[2], MD5C3e, MD5S32); - MD5_STEP (MD5_I , b, c, d, a, w2_t[1], MD5C3f, MD5S33); + MD5_STEP (MD5_I , d, a, b, c, w2[3], MD5C3d, MD5S31); + MD5_STEP (MD5_I , c, d, a, b, w0[2], MD5C3e, MD5S32); + MD5_STEP (MD5_I , b, c, d, a, w2[1], MD5C3f, MD5S33); COMPARE_S_SIMD (a, d, c, b); } diff --git a/OpenCL/m00040_a1.cl b/OpenCL/m00040_a1.cl index 9fa9651da..fdff9a4cd 100644 --- a/OpenCL/m00040_a1.cl +++ b/OpenCL/m00040_a1.cl @@ -38,17 +38,17 @@ __kernel void m00040_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -99,14 +99,14 @@ __kernel void m00040_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -143,15 +143,15 @@ __kernel void m00040_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; - 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; + w1[3] = wordl1[3] | wordr1[3]; + w2[0] = wordl2[0] | wordr2[0]; + w2[1] = wordl2[1] | wordr2[1]; + w2[2] = wordl2[2] | wordr2[2]; + w2[3] = wordl2[3] | wordr2[3]; + w3[0] = wordl3[0] | wordr3[0]; + w3[1] = wordl3[1] | wordr3[1]; + w3[2] = wordl3[2] | wordr3[2]; + w3[3] = wordl3[3] | wordr3[3]; make_unicode (w1, w2, w3); make_unicode (w0, w0, w1); @@ -288,17 +288,17 @@ __kernel void m00040_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -361,14 +361,14 @@ __kernel void m00040_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -405,15 +405,15 @@ __kernel void m00040_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, w1[0] = wordl1[0] | wordr1[0]; w1[1] = wordl1[1] | wordr1[1]; w1[2] = wordl1[2] | wordr1[2]; - 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; + w1[3] = wordl1[3] | wordr1[3]; + w2[0] = wordl2[0] | wordr2[0]; + w2[1] = wordl2[1] | wordr2[1]; + w2[2] = wordl2[2] | wordr2[2]; + w2[3] = wordl2[3] | wordr2[3]; + w3[0] = wordl3[0] | wordr3[0]; + w3[1] = wordl3[1] | wordr3[1]; + w3[2] = wordl3[2] | wordr3[2]; + w3[3] = wordl3[3] | wordr3[3]; make_unicode (w1, w2, w3); make_unicode (w0, w0, w1); diff --git a/OpenCL/m00040_a3.cl b/OpenCL/m00040_a3.cl index fabe52b55..fd8449b0c 100644 --- a/OpenCL/m00040_a3.cl +++ b/OpenCL/m00040_a3.cl @@ -63,101 +63,60 @@ static void m00040m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le * prepend salt */ - u32 w0_t[4]; - u32 w1_t[4]; - u32 w2_t[4]; - u32 w3_t[4]; + const u32 w0l = w0[0]; - w0_t[0] = w0[0]; - w0_t[1] = w0[1]; - w0_t[2] = w0[2]; - w0_t[3] = w0[3]; - w1_t[0] = w1[0]; - w1_t[1] = w1[1]; - w1_t[2] = w1[2]; - w1_t[3] = w1[3]; - w2_t[0] = w2[0]; - w2_t[1] = w2[1]; - w2_t[2] = w2[2]; - w2_t[3] = w2[3]; - w3_t[0] = w3[0]; - w3_t[1] = w3[1]; - w3_t[2] = w3[2]; - w3_t[3] = w3[3]; + switch_buffer_by_offset_le_S (w0, w1, w2, w3, salt_len); - switch_buffer_by_offset_le_S (w0_t, w1_t, w2_t, w3_t, salt_len); - - 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] |= salt_buf2[2]; - w2_t[3] |= salt_buf2[3]; - w3_t[0] |= salt_buf3[0]; - w3_t[1] |= salt_buf3[1]; - w3_t[2] |= salt_buf3[2]; - w3_t[3] |= salt_buf3[3]; + 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] |= salt_buf2[2]; + w2[3] |= salt_buf2[3]; + w3[0] |= salt_buf3[0]; + w3[1] |= salt_buf3[1]; + w3[2] = pw_salt_len * 8; + w3[3] = 0; /** * loop */ - u32 w0l = w0[0]; - for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos += VECT_SIZE) { const u32x w0r = ix_create_bft (bfs_buf, il_pos); const u32x w0lr = w0l | w0r; - u32x wx[16]; - - wx[ 0] = w0_t[0]; - wx[ 1] = w0_t[1]; - wx[ 2] = w0_t[2]; - wx[ 3] = w0_t[3]; - wx[ 4] = w1_t[0]; - wx[ 5] = w1_t[1]; - wx[ 6] = w1_t[2]; - wx[ 7] = w1_t[3]; - wx[ 8] = w2_t[0]; - wx[ 9] = w2_t[1]; - wx[10] = w2_t[2]; - wx[11] = w2_t[3]; - wx[12] = w3_t[0]; - wx[13] = w3_t[1]; - wx[14] = w3_t[2]; - wx[15] = w3_t[3]; - - overwrite_at_le (wx, w0lr, salt_len); - u32x w0_t[4]; u32x w1_t[4]; u32x w2_t[4]; u32x w3_t[4]; - w0_t[0] = wx[ 0]; - w0_t[1] = wx[ 1]; - w0_t[2] = wx[ 2]; - w0_t[3] = wx[ 3]; - w1_t[0] = wx[ 4]; - w1_t[1] = wx[ 5]; - w1_t[2] = wx[ 6]; - w1_t[3] = wx[ 7]; - w2_t[0] = wx[ 8]; - w2_t[1] = wx[ 9]; - w2_t[2] = wx[10]; - w2_t[3] = wx[11]; - w3_t[0] = wx[12]; - w3_t[1] = wx[13]; - w3_t[2] = pw_salt_len * 8; - w3_t[3] = 0; + w0_t[0] = w0[0]; + w0_t[1] = w0[1]; + w0_t[2] = w0[2]; + w0_t[3] = w0[3]; + w1_t[0] = w1[0]; + w1_t[1] = w1[1]; + w1_t[2] = w1[2]; + w1_t[3] = w1[3]; + w2_t[0] = w2[0]; + w2_t[1] = w2[1]; + w2_t[2] = w2[2]; + w2_t[3] = w2[3]; + w3_t[0] = w3[0]; + w3_t[1] = w3[1]; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; + + overwrite_at_le_4x4 (w0_t, w1_t, w2_t, w3_t, w0lr, salt_len); /** * md5 @@ -295,101 +254,60 @@ static void m00040s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le * prepend salt */ - u32 w0_t[4]; - u32 w1_t[4]; - u32 w2_t[4]; - u32 w3_t[4]; + const u32 w0l = w0[0]; - w0_t[0] = w0[0]; - w0_t[1] = w0[1]; - w0_t[2] = w0[2]; - w0_t[3] = w0[3]; - w1_t[0] = w1[0]; - w1_t[1] = w1[1]; - w1_t[2] = w1[2]; - w1_t[3] = w1[3]; - w2_t[0] = w2[0]; - w2_t[1] = w2[1]; - w2_t[2] = w2[2]; - w2_t[3] = w2[3]; - w3_t[0] = w3[0]; - w3_t[1] = w3[1]; - w3_t[2] = w3[2]; - w3_t[3] = w3[3]; + switch_buffer_by_offset_le_S (w0, w1, w2, w3, salt_len); - switch_buffer_by_offset_le_S (w0_t, w1_t, w2_t, w3_t, salt_len); - - 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] |= salt_buf2[2]; - w2_t[3] |= salt_buf2[3]; - w3_t[0] |= salt_buf3[0]; - w3_t[1] |= salt_buf3[1]; - w3_t[2] |= salt_buf3[2]; - w3_t[3] |= salt_buf3[3]; + 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] |= salt_buf2[2]; + w2[3] |= salt_buf2[3]; + w3[0] |= salt_buf3[0]; + w3[1] |= salt_buf3[1]; + w3[2] = pw_salt_len * 8; + w3[3] = 0; /** * loop */ - u32 w0l = w0[0]; - for (u32 il_pos = 0; il_pos < bfs_cnt; il_pos += VECT_SIZE) { const u32x w0r = ix_create_bft (bfs_buf, il_pos); const u32x w0lr = w0l | w0r; - u32x wx[16]; - - wx[ 0] = w0_t[0]; - wx[ 1] = w0_t[1]; - wx[ 2] = w0_t[2]; - wx[ 3] = w0_t[3]; - wx[ 4] = w1_t[0]; - wx[ 5] = w1_t[1]; - wx[ 6] = w1_t[2]; - wx[ 7] = w1_t[3]; - wx[ 8] = w2_t[0]; - wx[ 9] = w2_t[1]; - wx[10] = w2_t[2]; - wx[11] = w2_t[3]; - wx[12] = w3_t[0]; - wx[13] = w3_t[1]; - wx[14] = w3_t[2]; - wx[15] = w3_t[3]; - - overwrite_at_le (wx, w0lr, salt_len); - u32x w0_t[4]; u32x w1_t[4]; u32x w2_t[4]; u32x w3_t[4]; - w0_t[0] = wx[ 0]; - w0_t[1] = wx[ 1]; - w0_t[2] = wx[ 2]; - w0_t[3] = wx[ 3]; - w1_t[0] = wx[ 4]; - w1_t[1] = wx[ 5]; - w1_t[2] = wx[ 6]; - w1_t[3] = wx[ 7]; - w2_t[0] = wx[ 8]; - w2_t[1] = wx[ 9]; - w2_t[2] = wx[10]; - w2_t[3] = wx[11]; - w3_t[0] = wx[12]; - w3_t[1] = wx[13]; - w3_t[2] = pw_salt_len * 8; - w3_t[3] = 0; + w0_t[0] = w0[0]; + w0_t[1] = w0[1]; + w0_t[2] = w0[2]; + w0_t[3] = w0[3]; + w1_t[0] = w1[0]; + w1_t[1] = w1[1]; + w1_t[2] = w1[2]; + w1_t[3] = w1[3]; + w2_t[0] = w2[0]; + w2_t[1] = w2[1]; + w2_t[2] = w2[2]; + w2_t[3] = w2[3]; + w3_t[0] = w3[0]; + w3_t[1] = w3[1]; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; + + overwrite_at_le_4x4 (w0_t, w1_t, w2_t, w3_t, w0lr, salt_len); /** * md5 diff --git a/OpenCL/m00050_a0.cl b/OpenCL/m00050_a0.cl index 09fda93ca..378673979 100644 --- a/OpenCL/m00050_a0.cl +++ b/OpenCL/m00050_a0.cl @@ -222,18 +222,16 @@ __kernel void m00050_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu if (gid >= gid_max) return; 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; @@ -282,59 +280,31 @@ __kernel void m00050_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu * pads */ - u32x w0_t[4]; - - w0_t[0] = w0[0]; - 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] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - - u32x w3_t[4]; - - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = 0; - u32x ipad[4]; u32x opad[4]; - hmac_md5_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad); + hmac_md5_pad (w0, w1, w2, w3, ipad, opad); - 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] = salt_buf2[2]; - w2_t[3] = salt_buf2[3]; - w3_t[0] = salt_buf3[0]; - w3_t[1] = salt_buf3[1]; - w3_t[2] = (64 + salt_len) * 8; - w3_t[3] = 0; + 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] = salt_buf2[2]; + w2[3] = salt_buf2[3]; + w3[0] = salt_buf3[0]; + w3[1] = salt_buf3[1]; + w3[2] = (64 + salt_len) * 8; + w3[3] = 0; u32x digest[4]; - hmac_md5_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest); + hmac_md5_run (w0, w1, w2, w3, ipad, opad, digest); COMPARE_M_SIMD (digest[0], digest[3], digest[2], digest[1]); } @@ -365,18 +335,16 @@ __kernel void m00050_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu if (gid >= gid_max) return; 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; @@ -437,59 +405,31 @@ __kernel void m00050_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu * pads */ - u32x w0_t[4]; - - w0_t[0] = w0[0]; - 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] = 0; - w2_t[1] = 0; - w2_t[2] = 0; - w2_t[3] = 0; - - u32x w3_t[4]; - - w3_t[0] = 0; - w3_t[1] = 0; - w3_t[2] = 0; - w3_t[3] = 0; - u32x ipad[4]; u32x opad[4]; - hmac_md5_pad (w0_t, w1_t, w2_t, w3_t, ipad, opad); + hmac_md5_pad (w0, w1, w2, w3, ipad, opad); - 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] = salt_buf2[2]; - w2_t[3] = salt_buf2[3]; - w3_t[0] = salt_buf3[0]; - w3_t[1] = salt_buf3[1]; - w3_t[2] = (64 + salt_len) * 8; - w3_t[3] = 0; + 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] = salt_buf2[2]; + w2[3] = salt_buf2[3]; + w3[0] = salt_buf3[0]; + w3[1] = salt_buf3[1]; + w3[2] = (64 + salt_len) * 8; + w3[3] = 0; u32x digest[4]; - hmac_md5_run (w0_t, w1_t, w2_t, w3_t, ipad, opad, digest); + hmac_md5_run (w0, w1, w2, w3, ipad, opad, digest); COMPARE_S_SIMD (digest[0], digest[3], digest[2], digest[1]); } diff --git a/OpenCL/m00050_a1.cl b/OpenCL/m00050_a1.cl index 83a04b527..18e10fdfb 100644 --- a/OpenCL/m00050_a1.cl +++ b/OpenCL/m00050_a1.cl @@ -219,17 +219,17 @@ __kernel void m00050_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -280,14 +280,14 @@ __kernel void m00050_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -392,17 +392,17 @@ __kernel void m00050_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -465,14 +465,14 @@ __kernel void m00050_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; diff --git a/OpenCL/m00050_a3.cl b/OpenCL/m00050_a3.cl index 54e5e6717..67a8e0862 100644 --- a/OpenCL/m00050_a3.cl +++ b/OpenCL/m00050_a3.cl @@ -257,32 +257,26 @@ static void m00050m (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le */ 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] = 0; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; u32x ipad[4]; u32x opad[4]; @@ -380,32 +374,26 @@ static void m00050s (u32 w0[4], u32 w1[4], u32 w2[4], u32 w3[4], const u32 pw_le */ 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] = 0; + w3_t[2] = w3[2]; + w3_t[3] = w3[3]; u32x ipad[4]; u32x opad[4]; diff --git a/OpenCL/m00060_a0.cl b/OpenCL/m00060_a0.cl index 558416cd1..1aff5e48e 100644 --- a/OpenCL/m00060_a0.cl +++ b/OpenCL/m00060_a0.cl @@ -222,18 +222,16 @@ __kernel void m00060_m04 (__global pw_t *pws, __global kernel_rule_t * rules_bu if (gid >= gid_max) return; 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; @@ -361,18 +359,16 @@ __kernel void m00060_s04 (__global pw_t *pws, __global kernel_rule_t * rules_bu if (gid >= gid_max) return; 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; diff --git a/OpenCL/m00060_a1.cl b/OpenCL/m00060_a1.cl index c1e32f57b..ca8f7c626 100644 --- a/OpenCL/m00060_a1.cl +++ b/OpenCL/m00060_a1.cl @@ -219,17 +219,17 @@ __kernel void m00060_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -311,14 +311,14 @@ __kernel void m00060_m04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 }; @@ -397,17 +397,17 @@ __kernel void m00060_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, if (gid >= gid_max) return; - u32 pws0[4] = { 0 }; - u32 pws1[4] = { 0 }; + u32 pw_buf0[4]; + u32 pw_buf1[4]; - pws0[0] = pws[gid].i[0]; - pws0[1] = pws[gid].i[1]; - pws0[2] = pws[gid].i[2]; - pws0[3] = pws[gid].i[3]; - pws1[0] = pws[gid].i[4]; - pws1[1] = pws[gid].i[5]; - pws1[2] = pws[gid].i[6]; - pws1[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_l_len = pws[gid].pw_len; @@ -501,14 +501,14 @@ __kernel void m00060_s04 (__global pw_t *pws, __global kernel_rule_t *rules_buf, u32x wordl2[4] = { 0 }; u32x wordl3[4] = { 0 }; - wordl0[0] = pws0[0]; - wordl0[1] = pws0[1]; - wordl0[2] = pws0[2]; - wordl0[3] = pws0[3]; - wordl1[0] = pws1[0]; - wordl1[1] = pws1[1]; - wordl1[2] = pws1[2]; - wordl1[3] = pws1[3]; + 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 };