Add example -L kernel for algorithms using HMAC where the password is the data

OpenCL/m00050_a1-pure.cl

@@ -62,18 +62,21 @@ __kernel void m00050_mxx (__global pw_t *pws, __global const kernel_rule_t *rule
   {
     const u32 comb_len = combs_buf[il_pos].pw_len;
 
-    const u32 comb_lenv = ceil ((float) comb_len / 4);
+    u32 c[64];
 
-    u32 c[64] = { 0 };
-
-    for (int idx = 0; idx < comb_lenv; idx++)
+    #ifdef _unroll
+    #pragma unroll
+    #endif
+    for (int idx = 0; idx < 64; idx++)
     {
       c[idx] = combs_buf[il_pos].i[idx];
     }
 
     switch_buffer_by_offset_1x64_le_S (c, pw_len);
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 64; i++)
     {
       c[i] |= w[i];
@@ -157,18 +160,21 @@ __kernel void m00050_sxx (__global pw_t *pws, __global const kernel_rule_t *rule
   {
     const u32 comb_len = combs_buf[il_pos].pw_len;
 
-    const u32 comb_lenv = ceil ((float) comb_len / 4);
+    u32 c[64];
 
-    u32 c[64] = { 0 };
-
-    for (int idx = 0; idx < comb_lenv; idx++)
+    #ifdef _unroll
+    #pragma unroll
+    #endif
+    for (int idx = 0; idx < 64; idx++)
     {
       c[idx] = combs_buf[il_pos].i[idx];
     }
 
     switch_buffer_by_offset_1x64_le_S (c, pw_len);
 
+    #ifdef _unroll
     #pragma unroll
+    #endif
     for (int i = 0; i < 64; i++)
     {
       c[i] |= w[i];

OpenCL/m00060_a0-pure.cl

@@ -0,0 +1,164 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+//#define NEW_SIMD_CODE
+
+#include "inc_vendor.cl"
+#include "inc_hash_constants.h"
+#include "inc_hash_functions.cl"
+#include "inc_types.cl"
+#include "inc_common.cl"
+#include "inc_rp.h"
+#include "inc_rp.cl"
+#include "inc_scalar.cl"
+#include "inc_hash_md5.cl"
+
+__kernel void m00060_mxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+{
+  /**
+   * modifier
+   */
+
+  const u32 lid = get_local_id (0);
+  const u32 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  const u32 pw_lenv = ceil ((float) pw_len / 4);
+
+  u32 w[64] = { 0 };
+
+  for (int idx = 0; idx < pw_lenv; idx++)
+  {
+    w[idx] = pws[gid].i[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
+
+  const u32 salt_lenv = ceil ((float) salt_len / 4);
+
+  u32 s[64] = { 0 };
+
+  for (int idx = 0; idx < salt_lenv; idx++)
+  {
+    s[idx] = salt_bufs[salt_pos].salt_buf[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  md5_hmac_ctx_t ctx0;
+
+  md5_hmac_init (&ctx0, s, salt_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    // todo: add rules engine
+
+    md5_hmac_ctx_t ctx = ctx0;
+
+    md5_hmac_update (&ctx, w, pw_len);
+
+    md5_hmac_final (&ctx);
+
+    const u32x r0 = ctx.opad.h[DGST_R0];
+    const u32x r1 = ctx.opad.h[DGST_R1];
+    const u32x r2 = ctx.opad.h[DGST_R2];
+    const u32x r3 = ctx.opad.h[DGST_R3];
+
+    COMPARE_M_SCALAR (r0, r1, r2, r3);
+  }
+}
+
+__kernel void m00060_sxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+{
+  /**
+   * modifier
+   */
+
+  const u32 lid = get_local_id (0);
+  const u32 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * digest
+   */
+
+  const u32 search[4] =
+  {
+    digests_buf[digests_offset].digest_buf[DGST_R0],
+    digests_buf[digests_offset].digest_buf[DGST_R1],
+    digests_buf[digests_offset].digest_buf[DGST_R2],
+    digests_buf[digests_offset].digest_buf[DGST_R3]
+  };
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  const u32 pw_lenv = ceil ((float) pw_len / 4);
+
+  u32 w[64] = { 0 };
+
+  for (int idx = 0; idx < pw_lenv; idx++)
+  {
+    w[idx] = pws[gid].i[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
+
+  const u32 salt_lenv = ceil ((float) salt_len / 4);
+
+  u32 s[64] = { 0 };
+
+  for (int idx = 0; idx < salt_lenv; idx++)
+  {
+    s[idx] = salt_bufs[salt_pos].salt_buf[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  md5_hmac_ctx_t ctx0;
+
+  md5_hmac_init (&ctx0, s, salt_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    // todo: add rules engine
+
+    md5_hmac_ctx_t ctx = ctx0;
+
+    md5_hmac_update (&ctx, w, pw_len);
+
+    md5_hmac_final (&ctx);
+
+    const u32x r0 = ctx.opad.h[DGST_R0];
+    const u32x r1 = ctx.opad.h[DGST_R1];
+    const u32x r2 = ctx.opad.h[DGST_R2];
+    const u32x r3 = ctx.opad.h[DGST_R3];
+
+    COMPARE_S_SCALAR (r0, r1, r2, r3);
+  }
+}

OpenCL/m00060_a1-pure.cl

@@ -0,0 +1,202 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+//#define NEW_SIMD_CODE
+
+#include "inc_vendor.cl"
+#include "inc_hash_constants.h"
+#include "inc_hash_functions.cl"
+#include "inc_types.cl"
+#include "inc_common.cl"
+#include "inc_scalar.cl"
+#include "inc_hash_md5.cl"
+
+__kernel void m00060_mxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+{
+  /**
+   * modifier
+   */
+
+  const u32 lid = get_local_id (0);
+  const u32 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  const u32 pw_lenv = ceil ((float) pw_len / 4);
+
+  u32 w[64] = { 0 };
+
+  for (int idx = 0; idx < pw_lenv; idx++)
+  {
+    w[idx] = pws[gid].i[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
+
+  const u32 salt_lenv = ceil ((float) salt_len / 4);
+
+  u32 s[64] = { 0 };
+
+  for (int idx = 0; idx < salt_lenv; idx++)
+  {
+    s[idx] = salt_bufs[salt_pos].salt_buf[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  md5_hmac_ctx_t ctx0;
+
+  md5_hmac_init (&ctx0, s, salt_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    const u32 comb_len = combs_buf[il_pos].pw_len;
+
+    u32 c[64];
+
+    #ifdef _unroll
+    #pragma unroll
+    #endif
+    for (int idx = 0; idx < 64; idx++)
+    {
+      c[idx] = combs_buf[il_pos].i[idx];
+    }
+
+    switch_buffer_by_offset_1x64_le_S (c, pw_len);
+
+    #ifdef _unroll
+    #pragma unroll
+    #endif
+    for (int i = 0; i < 64; i++)
+    {
+      c[i] |= w[i];
+    }
+
+    md5_hmac_ctx_t ctx = ctx0;
+
+    md5_hmac_update (&ctx, c, pw_len + comb_len);
+
+    md5_hmac_final (&ctx);
+
+    const u32x r0 = ctx.opad.h[DGST_R0];
+    const u32x r1 = ctx.opad.h[DGST_R1];
+    const u32x r2 = ctx.opad.h[DGST_R2];
+    const u32x r3 = ctx.opad.h[DGST_R3];
+
+    COMPARE_M_SCALAR (r0, r1, r2, r3);
+  }
+}
+
+__kernel void m00060_sxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const bf_t *bfs_buf, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+{
+  /**
+   * modifier
+   */
+
+  const u32 lid = get_local_id (0);
+  const u32 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * digest
+   */
+
+  const u32 search[4] =
+  {
+    digests_buf[digests_offset].digest_buf[DGST_R0],
+    digests_buf[digests_offset].digest_buf[DGST_R1],
+    digests_buf[digests_offset].digest_buf[DGST_R2],
+    digests_buf[digests_offset].digest_buf[DGST_R3]
+  };
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  const u32 pw_lenv = ceil ((float) pw_len / 4);
+
+  u32 w[64] = { 0 };
+
+  for (int idx = 0; idx < pw_lenv; idx++)
+  {
+    w[idx] = pws[gid].i[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
+
+  const u32 salt_lenv = ceil ((float) salt_len / 4);
+
+  u32 s[64] = { 0 };
+
+  for (int idx = 0; idx < salt_lenv; idx++)
+  {
+    s[idx] = salt_bufs[salt_pos].salt_buf[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  md5_hmac_ctx_t ctx0;
+
+  md5_hmac_init (&ctx0, s, salt_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    const u32 comb_len = combs_buf[il_pos].pw_len;
+
+    u32 c[64];
+
+    #ifdef _unroll
+    #pragma unroll
+    #endif
+    for (int idx = 0; idx < 64; idx++)
+    {
+      c[idx] = combs_buf[il_pos].i[idx];
+    }
+
+    switch_buffer_by_offset_1x64_le_S (c, pw_len);
+
+    #ifdef _unroll
+    #pragma unroll
+    #endif
+    for (int i = 0; i < 64; i++)
+    {
+      c[i] |= w[i];
+    }
+
+    md5_hmac_ctx_t ctx = ctx0;
+
+    md5_hmac_update (&ctx, c, pw_len + comb_len);
+
+    md5_hmac_final (&ctx);
+
+    const u32x r0 = ctx.opad.h[DGST_R0];
+    const u32x r1 = ctx.opad.h[DGST_R1];
+    const u32x r2 = ctx.opad.h[DGST_R2];
+    const u32x r3 = ctx.opad.h[DGST_R3];
+
+    COMPARE_S_SCALAR (r0, r1, r2, r3);
+  }
+}

OpenCL/m00060_a3-pure.cl

@@ -0,0 +1,174 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+#define NEW_SIMD_CODE
+
+#include "inc_vendor.cl"
+#include "inc_hash_constants.h"
+#include "inc_hash_functions.cl"
+#include "inc_types.cl"
+#include "inc_common.cl"
+#include "inc_simd.cl"
+#include "inc_hash_md5.cl"
+
+__kernel void m00060_mxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const u32x *words_buf_r, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+{
+  /**
+   * modifier
+   */
+
+  const u32 lid = get_local_id (0);
+  const u32 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  const u32 pw_lenv = ceil ((float) pw_len / 4);
+
+  u32x w[64] = { 0 };
+
+  for (int idx = 0; idx < pw_lenv; idx++)
+  {
+    w[idx] = pws[gid].i[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
+
+  const u32 salt_lenv = ceil ((float) salt_len / 4);
+
+  u32x s[64] = { 0 };
+
+  for (int idx = 0; idx < salt_lenv; idx++)
+  {
+    s[idx] = salt_bufs[salt_pos].salt_buf[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  md5_hmac_ctx_vector_t ctx0;
+
+  md5_hmac_init_vector (&ctx0, s, salt_len);
+
+  /**
+   * loop
+   */
+
+  u32x w0l = w[0];
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    const u32x w0r = words_buf_r[il_pos / VECT_SIZE];
+
+    const u32x w0 = w0l | w0r;
+
+    w[0] = w0;
+
+    md5_hmac_ctx_vector_t ctx = ctx0;
+
+    md5_hmac_update_vector (&ctx, w, pw_len);
+
+    md5_hmac_final_vector (&ctx);
+
+    const u32x r0 = ctx.opad.h[DGST_R0];
+    const u32x r1 = ctx.opad.h[DGST_R1];
+    const u32x r2 = ctx.opad.h[DGST_R2];
+    const u32x r3 = ctx.opad.h[DGST_R3];
+
+    COMPARE_M_SIMD (r0, r1, r2, r3);
+  }
+}
+
+__kernel void m00060_sxx (__global pw_t *pws, __global const kernel_rule_t *rules_buf, __global const pw_t *combs_buf, __global const u32x *words_buf_r, __global void *tmps, __global void *hooks, __global const u32 *bitmaps_buf_s1_a, __global const u32 *bitmaps_buf_s1_b, __global const u32 *bitmaps_buf_s1_c, __global const u32 *bitmaps_buf_s1_d, __global const u32 *bitmaps_buf_s2_a, __global const u32 *bitmaps_buf_s2_b, __global const u32 *bitmaps_buf_s2_c, __global const u32 *bitmaps_buf_s2_d, __global plain_t *plains_buf, __global const digest_t *digests_buf, __global u32 *hashes_shown, __global const salt_t *salt_bufs, __global const void *esalt_bufs, __global u32 *d_return_buf, __global u32 *d_scryptV0_buf, __global u32 *d_scryptV1_buf, __global u32 *d_scryptV2_buf, __global u32 *d_scryptV3_buf, const u32 bitmap_mask, const u32 bitmap_shift1, const u32 bitmap_shift2, const u32 salt_pos, const u32 loop_pos, const u32 loop_cnt, const u32 il_cnt, const u32 digests_cnt, const u32 digests_offset, const u32 combs_mode, const u32 gid_max)
+{
+  /**
+   * modifier
+   */
+
+  const u32 lid = get_local_id (0);
+  const u32 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * digest
+   */
+
+  const u32 search[4] =
+  {
+    digests_buf[digests_offset].digest_buf[DGST_R0],
+    digests_buf[digests_offset].digest_buf[DGST_R1],
+    digests_buf[digests_offset].digest_buf[DGST_R2],
+    digests_buf[digests_offset].digest_buf[DGST_R3]
+  };
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  const u32 pw_lenv = ceil ((float) pw_len / 4);
+
+  u32x w[64] = { 0 };
+
+  for (int idx = 0; idx < pw_lenv; idx++)
+  {
+    w[idx] = pws[gid].i[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  const u32 salt_len = salt_bufs[salt_pos].salt_len;
+
+  const u32 salt_lenv = ceil ((float) salt_len / 4);
+
+  u32x s[64] = { 0 };
+
+  for (int idx = 0; idx < salt_lenv; idx++)
+  {
+    s[idx] = salt_bufs[salt_pos].salt_buf[idx];
+
+    barrier (CLK_GLOBAL_MEM_FENCE);
+  }
+
+  md5_hmac_ctx_vector_t ctx0;
+
+  md5_hmac_init_vector (&ctx0, s, salt_len);
+
+  /**
+   * loop
+   */
+
+  u32x w0l = w[0];
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    const u32x w0r = words_buf_r[il_pos / VECT_SIZE];
+
+    const u32x w0 = w0l | w0r;
+
+    w[0] = w0;
+
+    md5_hmac_ctx_vector_t ctx = ctx0;
+
+    md5_hmac_update_vector (&ctx, w, pw_len);
+
+    md5_hmac_final_vector (&ctx);
+
+    const u32x r0 = ctx.opad.h[DGST_R0];
+    const u32x r1 = ctx.opad.h[DGST_R1];
+    const u32x r2 = ctx.opad.h[DGST_R2];
+    const u32x r3 = ctx.opad.h[DGST_R3];
+
+    COMPARE_S_SIMD (r0, r1, r2, r3);
+  }
+}