Merge pull request #2657 from matrix/hash-mode-10820

Added hash-mode 10820 - sha384(salt.pass)

OpenCL/m10820_a0-optimized.cl

@@ -0,0 +1,443 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+#define NEW_SIMD_CODE
+
+#ifdef KERNEL_STATIC
+#include "inc_vendor.h"
+#include "inc_types.h"
+#include "inc_platform.cl"
+#include "inc_common.cl"
+#include "inc_rp_optimized.h"
+#include "inc_rp_optimized.cl"
+#include "inc_simd.cl"
+#include "inc_hash_sha384.cl"
+#endif
+
+DECLSPEC void sha384_transform_intern (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u64x *digest)
+{
+  u64x w0_t = hl32_to_64 (w0[0], w0[1]);
+  u64x w1_t = hl32_to_64 (w0[2], w0[3]);
+  u64x w2_t = hl32_to_64 (w1[0], w1[1]);
+  u64x w3_t = hl32_to_64 (w1[2], w1[3]);
+  u64x w4_t = hl32_to_64 (w2[0], w2[1]);
+  u64x w5_t = hl32_to_64 (w2[2], w2[3]);
+  u64x w6_t = hl32_to_64 (w3[0], w3[1]);
+  u64x w7_t = 0;
+  u64x w8_t = 0;
+  u64x w9_t = 0;
+  u64x wa_t = 0;
+  u64x wb_t = 0;
+  u64x wc_t = 0;
+  u64x wd_t = 0;
+  u64x we_t = 0;
+  u64x wf_t = hl32_to_64 (w3[2], w3[3]);
+
+  u64x a = digest[0];
+  u64x b = digest[1];
+  u64x c = digest[2];
+  u64x d = digest[3];
+  u64x e = digest[4];
+  u64x f = digest[5];
+  u64x g = digest[6];
+  u64x h = digest[7];
+
+  #define ROUND_EXPAND()                            \
+  {                                                 \
+    w0_t = SHA384_EXPAND (we_t, w9_t, w1_t, w0_t);  \
+    w1_t = SHA384_EXPAND (wf_t, wa_t, w2_t, w1_t);  \
+    w2_t = SHA384_EXPAND (w0_t, wb_t, w3_t, w2_t);  \
+    w3_t = SHA384_EXPAND (w1_t, wc_t, w4_t, w3_t);  \
+    w4_t = SHA384_EXPAND (w2_t, wd_t, w5_t, w4_t);  \
+    w5_t = SHA384_EXPAND (w3_t, we_t, w6_t, w5_t);  \
+    w6_t = SHA384_EXPAND (w4_t, wf_t, w7_t, w6_t);  \
+    w7_t = SHA384_EXPAND (w5_t, w0_t, w8_t, w7_t);  \
+    w8_t = SHA384_EXPAND (w6_t, w1_t, w9_t, w8_t);  \
+    w9_t = SHA384_EXPAND (w7_t, w2_t, wa_t, w9_t);  \
+    wa_t = SHA384_EXPAND (w8_t, w3_t, wb_t, wa_t);  \
+    wb_t = SHA384_EXPAND (w9_t, w4_t, wc_t, wb_t);  \
+    wc_t = SHA384_EXPAND (wa_t, w5_t, wd_t, wc_t);  \
+    wd_t = SHA384_EXPAND (wb_t, w6_t, we_t, wd_t);  \
+    we_t = SHA384_EXPAND (wc_t, w7_t, wf_t, we_t);  \
+    wf_t = SHA384_EXPAND (wd_t, w8_t, w0_t, wf_t);  \
+  }
+
+  #define ROUND_STEP(i)                                                                   \
+  {                                                                                       \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha384[i +  0]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha384[i +  1]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha384[i +  2]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha384[i +  3]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha384[i +  4]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha384[i +  5]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha384[i +  6]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha384[i +  7]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha384[i +  8]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha384[i +  9]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha384[i + 10]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha384[i + 11]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha384[i + 12]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha384[i + 13]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, c, d, e, f, g, h, a, b, we_t, k_sha384[i + 14]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha384[i + 15]); \
+  }
+
+  ROUND_STEP (0);
+
+  #ifdef IS_CUDA
+  ROUND_EXPAND (); ROUND_STEP (16);
+  ROUND_EXPAND (); ROUND_STEP (32);
+  ROUND_EXPAND (); ROUND_STEP (48);
+  ROUND_EXPAND (); ROUND_STEP (64);
+  #else
+  #ifdef _unroll
+  #pragma unroll
+  #endif
+  for (int i = 16; i < 80; i += 16)
+  {
+    ROUND_EXPAND (); ROUND_STEP (i);
+  }
+  #endif
+
+  /* rev
+  digest[0] += a;
+  digest[1] += b;
+  digest[2] += c;
+  digest[3] += d;
+  digest[4] += e;
+  digest[5] += f;
+  digest[6] += g;
+  digest[7] += h;
+  */
+
+  digest[0] = a;
+  digest[1] = b;
+  digest[2] = c;
+  digest[3] = d;
+  digest[4] = e;
+  digest[5] = f;
+  digest[6] = 0;
+  digest[7] = 0;
+}
+
+KERNEL_FQ void m10820_m04 (KERN_ATTR_RULES ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 pw_buf0[4];
+  u32 pw_buf1[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];
+  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 & 63;
+
+  /**
+   * salt
+   */
+
+  u32 salt_buf0[4];
+  u32 salt_buf1[4];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = salt_bufs[SALT_POS].salt_buf[ 0];
+  salt_buf0[1] = salt_bufs[SALT_POS].salt_buf[ 1];
+  salt_buf0[2] = salt_bufs[SALT_POS].salt_buf[ 2];
+  salt_buf0[3] = salt_bufs[SALT_POS].salt_buf[ 3];
+  salt_buf1[0] = salt_bufs[SALT_POS].salt_buf[ 4];
+  salt_buf1[1] = salt_bufs[SALT_POS].salt_buf[ 5];
+  salt_buf1[2] = salt_bufs[SALT_POS].salt_buf[ 6];
+  salt_buf1[3] = salt_bufs[SALT_POS].salt_buf[ 7];
+  salt_buf2[0] = salt_bufs[SALT_POS].salt_buf[ 8];
+  salt_buf2[1] = salt_bufs[SALT_POS].salt_buf[ 9];
+  salt_buf2[2] = salt_bufs[SALT_POS].salt_buf[10];
+  salt_buf2[3] = salt_bufs[SALT_POS].salt_buf[11];
+  salt_buf3[0] = salt_bufs[SALT_POS].salt_buf[12];
+  salt_buf3[1] = salt_bufs[SALT_POS].salt_buf[13];
+  salt_buf3[2] = salt_bufs[SALT_POS].salt_buf[14];
+  salt_buf3[3] = salt_bufs[SALT_POS].salt_buf[15];
+
+  const u32 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
+
+    const u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
+
+    /**
+     * prepend salt
+     */
+
+    const u32x out_salt_len = out_len + salt_len;
+
+    switch_buffer_by_offset_le_VV (w0, w1, w2, w3, salt_len);
+
+    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] |= salt_buf3[2];
+    w3[3] |= salt_buf3[3];
+
+    append_0x80_4x4_VV (w0, w1, w2, w3, out_salt_len);
+
+    /**
+     * sha512
+     */
+
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = hc_swap32 (w0[0]);
+    w0_t[1] = hc_swap32 (w0[1]);
+    w0_t[2] = hc_swap32 (w0[2]);
+    w0_t[3] = hc_swap32 (w0[3]);
+    w1_t[0] = hc_swap32 (w1[0]);
+    w1_t[1] = hc_swap32 (w1[1]);
+    w1_t[2] = hc_swap32 (w1[2]);
+    w1_t[3] = hc_swap32 (w1[3]);
+    w2_t[0] = hc_swap32 (w2[0]);
+    w2_t[1] = hc_swap32 (w2[1]);
+    w2_t[2] = hc_swap32 (w2[2]);
+    w2_t[3] = hc_swap32 (w2[3]);
+    w3_t[0] = hc_swap32 (w3[0]);
+    w3_t[1] = hc_swap32 (w3[1]);
+    w3_t[2] = 0;
+    w3_t[3] = out_salt_len * 8;
+
+    u64x digest[8];
+
+    digest[0] = SHA384M_A;
+    digest[1] = SHA384M_B;
+    digest[2] = SHA384M_C;
+    digest[3] = SHA384M_D;
+    digest[4] = SHA384M_E;
+    digest[5] = SHA384M_F;
+    digest[6] = SHA384M_G;
+    digest[7] = SHA384M_H;
+
+    sha384_transform_intern (w0_t, w1_t, w2_t, w3_t, digest);
+
+    const u32x r0 = l32_from_64 (digest[3]);
+    const u32x r1 = h32_from_64 (digest[3]);
+    const u32x r2 = l32_from_64 (digest[2]);
+    const u32x r3 = h32_from_64 (digest[2]);
+
+    COMPARE_M_SIMD (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_m08 (KERN_ATTR_RULES ())
+{
+}
+
+KERNEL_FQ void m10820_m16 (KERN_ATTR_RULES ())
+{
+}
+
+KERNEL_FQ void m10820_s04 (KERN_ATTR_RULES ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 pw_buf0[4];
+  u32 pw_buf1[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];
+  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 & 63;
+
+  /**
+   * salt
+   */
+
+  u32 salt_buf0[4];
+  u32 salt_buf1[4];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = salt_bufs[SALT_POS].salt_buf[ 0];
+  salt_buf0[1] = salt_bufs[SALT_POS].salt_buf[ 1];
+  salt_buf0[2] = salt_bufs[SALT_POS].salt_buf[ 2];
+  salt_buf0[3] = salt_bufs[SALT_POS].salt_buf[ 3];
+  salt_buf1[0] = salt_bufs[SALT_POS].salt_buf[ 4];
+  salt_buf1[1] = salt_bufs[SALT_POS].salt_buf[ 5];
+  salt_buf1[2] = salt_bufs[SALT_POS].salt_buf[ 6];
+  salt_buf1[3] = salt_bufs[SALT_POS].salt_buf[ 7];
+  salt_buf2[0] = salt_bufs[SALT_POS].salt_buf[ 8];
+  salt_buf2[1] = salt_bufs[SALT_POS].salt_buf[ 9];
+  salt_buf2[2] = salt_bufs[SALT_POS].salt_buf[10];
+  salt_buf2[3] = salt_bufs[SALT_POS].salt_buf[11];
+  salt_buf3[0] = salt_bufs[SALT_POS].salt_buf[12];
+  salt_buf3[1] = salt_bufs[SALT_POS].salt_buf[13];
+  salt_buf3[2] = salt_bufs[SALT_POS].salt_buf[14];
+  salt_buf3[3] = salt_bufs[SALT_POS].salt_buf[15];
+
+  const u32 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  /**
+   * digest
+   */
+
+  const u32 search[4] =
+  {
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R0],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R1],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R2],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R3]
+  };
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    u32x w0[4] = { 0 };
+    u32x w1[4] = { 0 };
+    u32x w2[4] = { 0 };
+    u32x w3[4] = { 0 };
+
+    const u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
+
+    /**
+     * prepend salt
+     */
+
+    const u32x out_salt_len = out_len + salt_len;
+
+    switch_buffer_by_offset_le_VV (w0, w1, w2, w3, salt_len);
+
+    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] |= salt_buf3[2];
+    w3[3] |= salt_buf3[3];
+
+    append_0x80_4x4_VV (w0, w1, w2, w3, out_salt_len);
+
+    /**
+     * sha384
+     */
+
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = hc_swap32 (w0[0]);
+    w0_t[1] = hc_swap32 (w0[1]);
+    w0_t[2] = hc_swap32 (w0[2]);
+    w0_t[3] = hc_swap32 (w0[3]);
+    w1_t[0] = hc_swap32 (w1[0]);
+    w1_t[1] = hc_swap32 (w1[1]);
+    w1_t[2] = hc_swap32 (w1[2]);
+    w1_t[3] = hc_swap32 (w1[3]);
+    w2_t[0] = hc_swap32 (w2[0]);
+    w2_t[1] = hc_swap32 (w2[1]);
+    w2_t[2] = hc_swap32 (w2[2]);
+    w2_t[3] = hc_swap32 (w2[3]);
+    w3_t[0] = hc_swap32 (w3[0]);
+    w3_t[1] = hc_swap32 (w3[1]);
+    w3_t[2] = 0;
+    w3_t[3] = out_salt_len * 8;
+
+    u64x digest[8];
+
+    digest[0] = SHA384M_A;
+    digest[1] = SHA384M_B;
+    digest[2] = SHA384M_C;
+    digest[3] = SHA384M_D;
+    digest[4] = SHA384M_E;
+    digest[5] = SHA384M_F;
+    digest[6] = SHA384M_G;
+    digest[7] = SHA384M_H;
+
+    sha384_transform_intern (w0_t, w1_t, w2_t, w3_t, digest);
+
+    const u32x r0 = l32_from_64 (digest[3]);
+    const u32x r1 = h32_from_64 (digest[3]);
+    const u32x r2 = l32_from_64 (digest[2]);
+    const u32x r3 = h32_from_64 (digest[2]);
+
+    COMPARE_S_SIMD (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_s08 (KERN_ATTR_RULES ())
+{
+}
+
+KERNEL_FQ void m10820_s16 (KERN_ATTR_RULES ())
+{
+}

OpenCL/m10820_a0-pure.cl

@@ -0,0 +1,125 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+//#define NEW_SIMD_CODE
+
+#ifdef KERNEL_STATIC
+#include "inc_vendor.h"
+#include "inc_types.h"
+#include "inc_platform.cl"
+#include "inc_common.cl"
+#include "inc_rp.h"
+#include "inc_rp.cl"
+#include "inc_scalar.cl"
+#include "inc_hash_sha384.cl"
+#endif
+
+KERNEL_FQ void m10820_mxx (KERN_ATTR_RULES ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
+
+  COPY_PW (pws[gid]);
+
+  sha384_ctx_t ctx0;
+
+  sha384_init (&ctx0);
+
+  sha384_update_global_swap (&ctx0, salt_bufs[SALT_POS].salt_buf, salt_bufs[SALT_POS].salt_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    pw_t tmp = PASTE_PW;
+
+    tmp.pw_len = apply_rules (rules_buf[il_pos].cmds, tmp.i, tmp.pw_len);
+
+    sha384_ctx_t ctx = ctx0;
+
+    sha384_update_swap (&ctx, tmp.i, tmp.pw_len);
+
+    sha384_final (&ctx);
+
+    const u32 r0 = l32_from_64_S (ctx.h[3]);
+    const u32 r1 = h32_from_64_S (ctx.h[3]);
+    const u32 r2 = l32_from_64_S (ctx.h[2]);
+    const u32 r3 = h32_from_64_S (ctx.h[2]);
+
+    COMPARE_M_SCALAR (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_sxx (KERN_ATTR_RULES ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+  const u64 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
+   */
+
+  COPY_PW (pws[gid]);
+
+  sha384_ctx_t ctx0;
+
+  sha384_init (&ctx0);
+
+  sha384_update_global_swap (&ctx0, salt_bufs[SALT_POS].salt_buf, salt_bufs[SALT_POS].salt_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    pw_t tmp = PASTE_PW;
+
+    tmp.pw_len = apply_rules (rules_buf[il_pos].cmds, tmp.i, tmp.pw_len);
+
+    sha384_ctx_t ctx = ctx0;
+
+    sha384_update_swap (&ctx, tmp.i, tmp.pw_len);
+
+    sha384_final (&ctx);
+
+    const u32 r0 = l32_from_64_S (ctx.h[3]);
+    const u32 r1 = h32_from_64_S (ctx.h[3]);
+    const u32 r2 = l32_from_64_S (ctx.h[2]);
+    const u32 r3 = h32_from_64_S (ctx.h[2]);
+
+    COMPARE_S_SCALAR (r0, r1, r2, r3);
+  }
+}

OpenCL/m10820_a1-optimized.cl

@@ -0,0 +1,557 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+#define NEW_SIMD_CODE
+
+#ifdef KERNEL_STATIC
+#include "inc_vendor.h"
+#include "inc_types.h"
+#include "inc_platform.cl"
+#include "inc_common.cl"
+#include "inc_simd.cl"
+#include "inc_hash_sha384.cl"
+#endif
+
+DECLSPEC void sha384_transform_intern (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u64x *digest)
+{
+  u64x w0_t = hl32_to_64 (w0[0], w0[1]);
+  u64x w1_t = hl32_to_64 (w0[2], w0[3]);
+  u64x w2_t = hl32_to_64 (w1[0], w1[1]);
+  u64x w3_t = hl32_to_64 (w1[2], w1[3]);
+  u64x w4_t = hl32_to_64 (w2[0], w2[1]);
+  u64x w5_t = hl32_to_64 (w2[2], w2[3]);
+  u64x w6_t = hl32_to_64 (w3[0], w3[1]);
+  u64x w7_t = 0;
+  u64x w8_t = 0;
+  u64x w9_t = 0;
+  u64x wa_t = 0;
+  u64x wb_t = 0;
+  u64x wc_t = 0;
+  u64x wd_t = 0;
+  u64x we_t = 0;
+  u64x wf_t = hl32_to_64 (w3[2], w3[3]);
+
+  u64x a = digest[0];
+  u64x b = digest[1];
+  u64x c = digest[2];
+  u64x d = digest[3];
+  u64x e = digest[4];
+  u64x f = digest[5];
+  u64x g = digest[6];
+  u64x h = digest[7];
+
+  #define ROUND_EXPAND()                            \
+  {                                                 \
+    w0_t = SHA384_EXPAND (we_t, w9_t, w1_t, w0_t);  \
+    w1_t = SHA384_EXPAND (wf_t, wa_t, w2_t, w1_t);  \
+    w2_t = SHA384_EXPAND (w0_t, wb_t, w3_t, w2_t);  \
+    w3_t = SHA384_EXPAND (w1_t, wc_t, w4_t, w3_t);  \
+    w4_t = SHA384_EXPAND (w2_t, wd_t, w5_t, w4_t);  \
+    w5_t = SHA384_EXPAND (w3_t, we_t, w6_t, w5_t);  \
+    w6_t = SHA384_EXPAND (w4_t, wf_t, w7_t, w6_t);  \
+    w7_t = SHA384_EXPAND (w5_t, w0_t, w8_t, w7_t);  \
+    w8_t = SHA384_EXPAND (w6_t, w1_t, w9_t, w8_t);  \
+    w9_t = SHA384_EXPAND (w7_t, w2_t, wa_t, w9_t);  \
+    wa_t = SHA384_EXPAND (w8_t, w3_t, wb_t, wa_t);  \
+    wb_t = SHA384_EXPAND (w9_t, w4_t, wc_t, wb_t);  \
+    wc_t = SHA384_EXPAND (wa_t, w5_t, wd_t, wc_t);  \
+    wd_t = SHA384_EXPAND (wb_t, w6_t, we_t, wd_t);  \
+    we_t = SHA384_EXPAND (wc_t, w7_t, wf_t, we_t);  \
+    wf_t = SHA384_EXPAND (wd_t, w8_t, w0_t, wf_t);  \
+  }
+
+  #define ROUND_STEP(i)                                                                   \
+  {                                                                                       \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha384[i +  0]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha384[i +  1]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha384[i +  2]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha384[i +  3]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha384[i +  4]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha384[i +  5]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha384[i +  6]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha384[i +  7]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha384[i +  8]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha384[i +  9]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha384[i + 10]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha384[i + 11]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha384[i + 12]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha384[i + 13]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, c, d, e, f, g, h, a, b, we_t, k_sha384[i + 14]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha384[i + 15]); \
+  }
+
+  ROUND_STEP (0);
+
+  #ifdef IS_CUDA
+  ROUND_EXPAND (); ROUND_STEP (16);
+  ROUND_EXPAND (); ROUND_STEP (32);
+  ROUND_EXPAND (); ROUND_STEP (48);
+  ROUND_EXPAND (); ROUND_STEP (64);
+  #else
+  #ifdef _unroll
+  #pragma unroll
+  #endif
+  for (int i = 16; i < 80; i += 16)
+  {
+    ROUND_EXPAND (); ROUND_STEP (i);
+  }
+  #endif
+
+  /* rev
+  digest[0] += a;
+  digest[1] += b;
+  digest[2] += c;
+  digest[3] += d;
+  digest[4] += e;
+  digest[5] += f;
+  digest[6] += g;
+  digest[7] += h;
+  */
+
+  digest[0] = a;
+  digest[1] = b;
+  digest[2] = c;
+  digest[3] = d;
+  digest[4] = e;
+  digest[5] = f;
+  digest[6] = 0;
+  digest[7] = 0;
+}
+
+KERNEL_FQ void m10820_m04 (KERN_ATTR_BASIC ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 pw_buf0[4];
+  u32 pw_buf1[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];
+  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 & 63;
+
+  /**
+   * salt
+   */
+
+  u32 salt_buf0[4];
+  u32 salt_buf1[4];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = salt_bufs[SALT_POS].salt_buf[ 0];
+  salt_buf0[1] = salt_bufs[SALT_POS].salt_buf[ 1];
+  salt_buf0[2] = salt_bufs[SALT_POS].salt_buf[ 2];
+  salt_buf0[3] = salt_bufs[SALT_POS].salt_buf[ 3];
+  salt_buf1[0] = salt_bufs[SALT_POS].salt_buf[ 4];
+  salt_buf1[1] = salt_bufs[SALT_POS].salt_buf[ 5];
+  salt_buf1[2] = salt_bufs[SALT_POS].salt_buf[ 6];
+  salt_buf1[3] = salt_bufs[SALT_POS].salt_buf[ 7];
+  salt_buf2[0] = salt_bufs[SALT_POS].salt_buf[ 8];
+  salt_buf2[1] = salt_bufs[SALT_POS].salt_buf[ 9];
+  salt_buf2[2] = salt_bufs[SALT_POS].salt_buf[10];
+  salt_buf2[3] = salt_bufs[SALT_POS].salt_buf[11];
+  salt_buf3[0] = salt_bufs[SALT_POS].salt_buf[12];
+  salt_buf3[1] = salt_bufs[SALT_POS].salt_buf[13];
+  salt_buf3[2] = salt_bufs[SALT_POS].salt_buf[14];
+  salt_buf3[3] = salt_bufs[SALT_POS].salt_buf[15];
+
+  const u32 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos) & 63;
+
+    const u32x pw_len = (pw_l_len + pw_r_len) & 63;
+
+    /**
+     * concat password candidate
+     */
+
+    u32x wordl0[4] = { 0 };
+    u32x wordl1[4] = { 0 };
+    u32x wordl2[4] = { 0 };
+    u32x wordl3[4] = { 0 };
+
+    wordl0[0] = pw_buf0[0];
+    wordl0[1] = pw_buf0[1];
+    wordl0[2] = pw_buf0[2];
+    wordl0[3] = pw_buf0[3];
+    wordl1[0] = pw_buf1[0];
+    wordl1[1] = pw_buf1[1];
+    wordl1[2] = pw_buf1[2];
+    wordl1[3] = pw_buf1[3];
+
+    u32x wordr0[4] = { 0 };
+    u32x wordr1[4] = { 0 };
+    u32x wordr2[4] = { 0 };
+    u32x wordr3[4] = { 0 };
+
+    wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
+    wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
+    wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
+    wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
+    wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
+    wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
+    wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
+    wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
+
+    if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
+    {
+      switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
+    }
+    else
+    {
+      switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
+    }
+
+    u32x w0[4];
+    u32x w1[4];
+    u32x w2[4];
+    u32x w3[4];
+
+    w0[0] = wordl0[0] | wordr0[0];
+    w0[1] = wordl0[1] | wordr0[1];
+    w0[2] = wordl0[2] | wordr0[2];
+    w0[3] = wordl0[3] | wordr0[3];
+    w1[0] = wordl1[0] | wordr1[0];
+    w1[1] = wordl1[1] | wordr1[1];
+    w1[2] = wordl1[2] | wordr1[2];
+    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];
+
+    /**
+     * prepend salt
+     */
+
+    switch_buffer_by_offset_le (w0, w1, w2, w3, salt_len);
+
+    const u32x pw_salt_len = pw_len + salt_len;
+
+    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] |= salt_buf3[2];
+    w3[3] |= salt_buf3[3];
+
+    /**
+     * sha512
+     */
+
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = hc_swap32 (w0[0]);
+    w0_t[1] = hc_swap32 (w0[1]);
+    w0_t[2] = hc_swap32 (w0[2]);
+    w0_t[3] = hc_swap32 (w0[3]);
+    w1_t[0] = hc_swap32 (w1[0]);
+    w1_t[1] = hc_swap32 (w1[1]);
+    w1_t[2] = hc_swap32 (w1[2]);
+    w1_t[3] = hc_swap32 (w1[3]);
+    w2_t[0] = hc_swap32 (w2[0]);
+    w2_t[1] = hc_swap32 (w2[1]);
+    w2_t[2] = hc_swap32 (w2[2]);
+    w2_t[3] = hc_swap32 (w2[3]);
+    w3_t[0] = hc_swap32 (w3[0]);
+    w3_t[1] = hc_swap32 (w3[1]);
+    w3_t[2] = 0;
+    w3_t[3] = pw_salt_len * 8;
+
+    u64x digest[8];
+
+    digest[0] = SHA384M_A;
+    digest[1] = SHA384M_B;
+    digest[2] = SHA384M_C;
+    digest[3] = SHA384M_D;
+    digest[4] = SHA384M_E;
+    digest[5] = SHA384M_F;
+    digest[6] = SHA384M_G;
+    digest[7] = SHA384M_H;
+
+    sha384_transform_intern (w0_t, w1_t, w2_t, w3_t, digest);
+
+    const u32x r0 = l32_from_64 (digest[3]);
+    const u32x r1 = h32_from_64 (digest[3]);
+    const u32x r2 = l32_from_64 (digest[2]);
+    const u32x r3 = h32_from_64 (digest[2]);
+
+    COMPARE_M_SIMD (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_m08 (KERN_ATTR_BASIC ())
+{
+}
+
+KERNEL_FQ void m10820_m16 (KERN_ATTR_BASIC ())
+{
+}
+
+KERNEL_FQ void m10820_s04 (KERN_ATTR_BASIC ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 pw_buf0[4];
+  u32 pw_buf1[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];
+  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 & 63;
+
+  /**
+   * salt
+   */
+
+  u32 salt_buf0[4];
+  u32 salt_buf1[4];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = salt_bufs[SALT_POS].salt_buf[ 0];
+  salt_buf0[1] = salt_bufs[SALT_POS].salt_buf[ 1];
+  salt_buf0[2] = salt_bufs[SALT_POS].salt_buf[ 2];
+  salt_buf0[3] = salt_bufs[SALT_POS].salt_buf[ 3];
+  salt_buf1[0] = salt_bufs[SALT_POS].salt_buf[ 4];
+  salt_buf1[1] = salt_bufs[SALT_POS].salt_buf[ 5];
+  salt_buf1[2] = salt_bufs[SALT_POS].salt_buf[ 6];
+  salt_buf1[3] = salt_bufs[SALT_POS].salt_buf[ 7];
+  salt_buf2[0] = salt_bufs[SALT_POS].salt_buf[ 8];
+  salt_buf2[1] = salt_bufs[SALT_POS].salt_buf[ 9];
+  salt_buf2[2] = salt_bufs[SALT_POS].salt_buf[10];
+  salt_buf2[3] = salt_bufs[SALT_POS].salt_buf[11];
+  salt_buf3[0] = salt_bufs[SALT_POS].salt_buf[12];
+  salt_buf3[1] = salt_bufs[SALT_POS].salt_buf[13];
+  salt_buf3[2] = salt_bufs[SALT_POS].salt_buf[14];
+  salt_buf3[3] = salt_bufs[SALT_POS].salt_buf[15];
+
+  const u32 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  /**
+   * digest
+   */
+
+  const u32 search[4] =
+  {
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R0],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R1],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R2],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R3]
+  };
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    const u32x pw_r_len = pwlenx_create_combt (combs_buf, il_pos) & 63;
+
+    const u32x pw_len = (pw_l_len + pw_r_len) & 63;
+
+    /**
+     * concat password candidate
+     */
+
+    u32x wordl0[4] = { 0 };
+    u32x wordl1[4] = { 0 };
+    u32x wordl2[4] = { 0 };
+    u32x wordl3[4] = { 0 };
+
+    wordl0[0] = pw_buf0[0];
+    wordl0[1] = pw_buf0[1];
+    wordl0[2] = pw_buf0[2];
+    wordl0[3] = pw_buf0[3];
+    wordl1[0] = pw_buf1[0];
+    wordl1[1] = pw_buf1[1];
+    wordl1[2] = pw_buf1[2];
+    wordl1[3] = pw_buf1[3];
+
+    u32x wordr0[4] = { 0 };
+    u32x wordr1[4] = { 0 };
+    u32x wordr2[4] = { 0 };
+    u32x wordr3[4] = { 0 };
+
+    wordr0[0] = ix_create_combt (combs_buf, il_pos, 0);
+    wordr0[1] = ix_create_combt (combs_buf, il_pos, 1);
+    wordr0[2] = ix_create_combt (combs_buf, il_pos, 2);
+    wordr0[3] = ix_create_combt (combs_buf, il_pos, 3);
+    wordr1[0] = ix_create_combt (combs_buf, il_pos, 4);
+    wordr1[1] = ix_create_combt (combs_buf, il_pos, 5);
+    wordr1[2] = ix_create_combt (combs_buf, il_pos, 6);
+    wordr1[3] = ix_create_combt (combs_buf, il_pos, 7);
+
+    if (combs_mode == COMBINATOR_MODE_BASE_LEFT)
+    {
+      switch_buffer_by_offset_le_VV (wordr0, wordr1, wordr2, wordr3, pw_l_len);
+    }
+    else
+    {
+      switch_buffer_by_offset_le_VV (wordl0, wordl1, wordl2, wordl3, pw_r_len);
+    }
+
+    u32x w0[4];
+    u32x w1[4];
+    u32x w2[4];
+    u32x w3[4];
+
+    w0[0] = wordl0[0] | wordr0[0];
+    w0[1] = wordl0[1] | wordr0[1];
+    w0[2] = wordl0[2] | wordr0[2];
+    w0[3] = wordl0[3] | wordr0[3];
+    w1[0] = wordl1[0] | wordr1[0];
+    w1[1] = wordl1[1] | wordr1[1];
+    w1[2] = wordl1[2] | wordr1[2];
+    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];
+
+    /**
+     * prepend salt
+     */
+
+    switch_buffer_by_offset_le (w0, w1, w2, w3, salt_len);
+
+    const u32x pw_salt_len = pw_len + salt_len;
+
+    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] |= salt_buf3[2];
+    w3[3] |= salt_buf3[3];
+
+    /**
+     * sha512
+     */
+
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = hc_swap32 (w0[0]);
+    w0_t[1] = hc_swap32 (w0[1]);
+    w0_t[2] = hc_swap32 (w0[2]);
+    w0_t[3] = hc_swap32 (w0[3]);
+    w1_t[0] = hc_swap32 (w1[0]);
+    w1_t[1] = hc_swap32 (w1[1]);
+    w1_t[2] = hc_swap32 (w1[2]);
+    w1_t[3] = hc_swap32 (w1[3]);
+    w2_t[0] = hc_swap32 (w2[0]);
+    w2_t[1] = hc_swap32 (w2[1]);
+    w2_t[2] = hc_swap32 (w2[2]);
+    w2_t[3] = hc_swap32 (w2[3]);
+    w3_t[0] = hc_swap32 (w3[0]);
+    w3_t[1] = hc_swap32 (w3[1]);
+    w3_t[2] = 0;
+    w3_t[3] = pw_salt_len * 8;
+
+    u64x digest[8];
+
+    digest[0] = SHA384M_A;
+    digest[1] = SHA384M_B;
+    digest[2] = SHA384M_C;
+    digest[3] = SHA384M_D;
+    digest[4] = SHA384M_E;
+    digest[5] = SHA384M_F;
+    digest[6] = SHA384M_G;
+    digest[7] = SHA384M_H;
+
+    sha384_transform_intern (w0_t, w1_t, w2_t, w3_t, digest);
+
+    const u32x r0 = l32_from_64 (digest[3]);
+    const u32x r1 = h32_from_64 (digest[3]);
+    const u32x r2 = l32_from_64 (digest[2]);
+    const u32x r3 = h32_from_64 (digest[2]);
+
+    COMPARE_S_SIMD (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_s08 (KERN_ATTR_BASIC ())
+{
+}
+
+KERNEL_FQ void m10820_s16 (KERN_ATTR_BASIC ())
+{
+}

OpenCL/m10820_a1-pure.cl

@@ -0,0 +1,124 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+//#define NEW_SIMD_CODE
+
+#ifdef KERNEL_STATIC
+#include "inc_vendor.h"
+#include "inc_types.h"
+#include "inc_platform.cl"
+#include "inc_common.cl"
+#include "inc_scalar.cl"
+#include "inc_hash_sha384.cl"
+#endif
+
+KERNEL_FQ void m10820_mxx (KERN_ATTR_BASIC ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
+
+  sha384_ctx_t ctx0;
+
+  sha384_init (&ctx0);
+
+  sha384_update_global_swap (&ctx0, salt_bufs[SALT_POS].salt_buf, salt_bufs[SALT_POS].salt_len);
+
+  sha384_update_global_swap (&ctx0, pws[gid].i, pws[gid].pw_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    sha384_ctx_t ctx = ctx0;
+
+    sha384_update_global_swap (&ctx, combs_buf[il_pos].i, combs_buf[il_pos].pw_len);
+
+    sha384_final (&ctx);
+
+    const u32 r0 = l32_from_64_S (ctx.h[3]);
+    const u32 r1 = h32_from_64_S (ctx.h[3]);
+    const u32 r2 = l32_from_64_S (ctx.h[2]);
+    const u32 r3 = h32_from_64_S (ctx.h[2]);
+
+    COMPARE_M_SCALAR (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_sxx (KERN_ATTR_BASIC ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+  const u64 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 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  u32 s[64] = { 0 };
+
+  for (u32 i = 0, idx = 0; i < salt_len; i += 4, idx += 1)
+  {
+    s[idx] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[idx]);
+  }
+
+  sha384_ctx_t ctx0;
+
+  sha384_init (&ctx0);
+
+  sha384_update_global_swap (&ctx0, salt_bufs[SALT_POS].salt_buf, salt_bufs[SALT_POS].salt_len);
+
+  sha384_update_global_swap (&ctx0, pws[gid].i, pws[gid].pw_len);
+
+  /**
+   * loop
+   */
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos++)
+  {
+    sha384_ctx_t ctx = ctx0;
+
+    sha384_update_global_swap (&ctx, combs_buf[il_pos].i, combs_buf[il_pos].pw_len);
+
+    sha384_final (&ctx);
+
+    const u32 r0 = l32_from_64_S (ctx.h[3]);
+    const u32 r1 = h32_from_64_S (ctx.h[3]);
+    const u32 r2 = l32_from_64_S (ctx.h[2]);
+    const u32 r3 = h32_from_64_S (ctx.h[2]);
+
+    COMPARE_S_SCALAR (r0, r1, r2, r3);
+  }
+}

OpenCL/m10820_a3-optimized.cl

@@ -0,0 +1,597 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+#define NEW_SIMD_CODE
+
+#ifdef KERNEL_STATIC
+#include "inc_vendor.h"
+#include "inc_types.h"
+#include "inc_platform.cl"
+#include "inc_common.cl"
+#include "inc_simd.cl"
+#include "inc_hash_sha384.cl"
+#endif
+
+DECLSPEC void sha384_transform_intern (const u32x *w0, const u32x *w1, const u32x *w2, const u32x *w3, u64x *digest)
+{
+  u64x w0_t = hl32_to_64 (w0[0], w0[1]);
+  u64x w1_t = hl32_to_64 (w0[2], w0[3]);
+  u64x w2_t = hl32_to_64 (w1[0], w1[1]);
+  u64x w3_t = hl32_to_64 (w1[2], w1[3]);
+  u64x w4_t = hl32_to_64 (w2[0], w2[1]);
+  u64x w5_t = hl32_to_64 (w2[2], w2[3]);
+  u64x w6_t = hl32_to_64 (w3[0], w3[1]);
+  u64x w7_t = 0;
+  u64x w8_t = 0;
+  u64x w9_t = 0;
+  u64x wa_t = 0;
+  u64x wb_t = 0;
+  u64x wc_t = 0;
+  u64x wd_t = 0;
+  u64x we_t = 0;
+  u64x wf_t = hl32_to_64 (w3[2], w3[3]);
+
+  u64x a = digest[0];
+  u64x b = digest[1];
+  u64x c = digest[2];
+  u64x d = digest[3];
+  u64x e = digest[4];
+  u64x f = digest[5];
+  u64x g = digest[6];
+  u64x h = digest[7];
+
+  #define ROUND_EXPAND()                            \
+  {                                                 \
+    w0_t = SHA384_EXPAND (we_t, w9_t, w1_t, w0_t);  \
+    w1_t = SHA384_EXPAND (wf_t, wa_t, w2_t, w1_t);  \
+    w2_t = SHA384_EXPAND (w0_t, wb_t, w3_t, w2_t);  \
+    w3_t = SHA384_EXPAND (w1_t, wc_t, w4_t, w3_t);  \
+    w4_t = SHA384_EXPAND (w2_t, wd_t, w5_t, w4_t);  \
+    w5_t = SHA384_EXPAND (w3_t, we_t, w6_t, w5_t);  \
+    w6_t = SHA384_EXPAND (w4_t, wf_t, w7_t, w6_t);  \
+    w7_t = SHA384_EXPAND (w5_t, w0_t, w8_t, w7_t);  \
+    w8_t = SHA384_EXPAND (w6_t, w1_t, w9_t, w8_t);  \
+    w9_t = SHA384_EXPAND (w7_t, w2_t, wa_t, w9_t);  \
+    wa_t = SHA384_EXPAND (w8_t, w3_t, wb_t, wa_t);  \
+    wb_t = SHA384_EXPAND (w9_t, w4_t, wc_t, wb_t);  \
+    wc_t = SHA384_EXPAND (wa_t, w5_t, wd_t, wc_t);  \
+    wd_t = SHA384_EXPAND (wb_t, w6_t, we_t, wd_t);  \
+    we_t = SHA384_EXPAND (wc_t, w7_t, wf_t, we_t);  \
+    wf_t = SHA384_EXPAND (wd_t, w8_t, w0_t, wf_t);  \
+  }
+
+  #define ROUND_STEP(i)                                                                   \
+  {                                                                                       \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, a, b, c, d, e, f, g, h, w0_t, k_sha384[i +  0]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, h, a, b, c, d, e, f, g, w1_t, k_sha384[i +  1]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, g, h, a, b, c, d, e, f, w2_t, k_sha384[i +  2]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, f, g, h, a, b, c, d, e, w3_t, k_sha384[i +  3]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, e, f, g, h, a, b, c, d, w4_t, k_sha384[i +  4]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, d, e, f, g, h, a, b, c, w5_t, k_sha384[i +  5]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, c, d, e, f, g, h, a, b, w6_t, k_sha384[i +  6]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, b, c, d, e, f, g, h, a, w7_t, k_sha384[i +  7]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, a, b, c, d, e, f, g, h, w8_t, k_sha384[i +  8]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, h, a, b, c, d, e, f, g, w9_t, k_sha384[i +  9]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, g, h, a, b, c, d, e, f, wa_t, k_sha384[i + 10]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, f, g, h, a, b, c, d, e, wb_t, k_sha384[i + 11]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, e, f, g, h, a, b, c, d, wc_t, k_sha384[i + 12]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, d, e, f, g, h, a, b, c, wd_t, k_sha384[i + 13]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, c, d, e, f, g, h, a, b, we_t, k_sha384[i + 14]); \
+    SHA384_STEP (SHA384_F0o, SHA384_F1o, b, c, d, e, f, g, h, a, wf_t, k_sha384[i + 15]); \
+  }
+
+  ROUND_STEP (0);
+
+  #ifdef IS_CUDA
+  ROUND_EXPAND (); ROUND_STEP (16);
+  ROUND_EXPAND (); ROUND_STEP (32);
+  ROUND_EXPAND (); ROUND_STEP (48);
+  ROUND_EXPAND (); ROUND_STEP (64);
+  #else
+  #ifdef _unroll
+  #pragma unroll
+  #endif
+  for (int i = 16; i < 80; i += 16)
+  {
+    ROUND_EXPAND (); ROUND_STEP (i);
+  }
+  #endif
+
+  /* rev
+  digest[0] += a;
+  digest[1] += b;
+  digest[2] += c;
+  digest[3] += d;
+  digest[4] += e;
+  digest[5] += f;
+  digest[6] += g;
+  digest[7] += h;
+  */
+
+  digest[0] = a;
+  digest[1] = b;
+  digest[2] = c;
+  digest[3] = d;
+  digest[4] = e;
+  digest[5] = f;
+  digest[6] = 0;
+  digest[7] = 0;
+}
+
+DECLSPEC void m10820m (u32 *w, const u32 pw_len, KERN_ATTR_BASIC ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 gid = get_global_id (0);
+  const u64 lid = get_local_id (0);
+
+  /**
+   * salt
+   */
+
+  u32 salt_buf0[4];
+  u32 salt_buf1[4];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 0]);
+  salt_buf0[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 1]);
+  salt_buf0[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 2]);
+  salt_buf0[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 3]);
+  salt_buf1[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 4]);
+  salt_buf1[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 5]);
+  salt_buf1[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 6]);
+  salt_buf1[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 7]);
+  salt_buf2[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 8]);
+  salt_buf2[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 9]);
+  salt_buf2[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[10]);
+  salt_buf2[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[11]);
+  salt_buf3[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[12]);
+  salt_buf3[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[13]);
+  salt_buf3[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[14]);
+  salt_buf3[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[15]);
+
+  const u32 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  const u32 pw_salt_len = pw_len + salt_len;
+
+  /**
+   * loop
+   */
+
+  u32 w0l = w[0];
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    const u32x w0r = ix_create_bft (bfs_buf, il_pos);
+
+    const u32x w0lr = w0l | w0r;
+
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = w0lr;
+    w0_t[1] = w[ 1];
+    w0_t[2] = w[ 2];
+    w0_t[3] = w[ 3];
+    w1_t[0] = w[ 4];
+    w1_t[1] = w[ 5];
+    w1_t[2] = w[ 6];
+    w1_t[3] = w[ 7];
+    w2_t[0] = w[ 8];
+    w2_t[1] = w[ 9];
+    w2_t[2] = w[10];
+    w2_t[3] = w[11];
+    w3_t[0] = w[12];
+    w3_t[1] = w[13];
+    w3_t[2] = w[14];
+    w3_t[3] = w[15];
+
+    switch_buffer_by_offset_be (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]  = 0;
+    w3_t[3]  = pw_salt_len * 8;
+
+    /**
+     * sha512
+     */
+
+    u64x digest[8];
+
+    digest[0] = SHA384M_A;
+    digest[1] = SHA384M_B;
+    digest[2] = SHA384M_C;
+    digest[3] = SHA384M_D;
+    digest[4] = SHA384M_E;
+    digest[5] = SHA384M_F;
+    digest[6] = SHA384M_G;
+    digest[7] = SHA384M_H;
+
+    sha384_transform_intern (w0_t, w1_t, w2_t, w3_t, digest);
+
+    const u32x r0 = l32_from_64 (digest[3]);
+    const u32x r1 = h32_from_64 (digest[3]);
+    const u32x r2 = l32_from_64 (digest[2]);
+    const u32x r3 = h32_from_64 (digest[2]);
+
+    COMPARE_M_SIMD (r0, r1, r2, r3);
+  }
+}
+
+DECLSPEC void m10820s (u32 *w, const u32 pw_len, KERN_ATTR_BASIC ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 gid = get_global_id (0);
+  const u64 lid = get_local_id (0);
+
+  /**
+   * digest
+   */
+
+  const u32 search[4] =
+  {
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R0],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R1],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R2],
+    digests_buf[DIGESTS_OFFSET].digest_buf[DGST_R3]
+  };
+
+  /**
+   * salt
+   */
+
+  u32 salt_buf0[4];
+  u32 salt_buf1[4];
+  u32 salt_buf2[4];
+  u32 salt_buf3[4];
+
+  salt_buf0[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 0]);
+  salt_buf0[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 1]);
+  salt_buf0[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 2]);
+  salt_buf0[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 3]);
+  salt_buf1[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 4]);
+  salt_buf1[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 5]);
+  salt_buf1[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 6]);
+  salt_buf1[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 7]);
+  salt_buf2[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 8]);
+  salt_buf2[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[ 9]);
+  salt_buf2[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[10]);
+  salt_buf2[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[11]);
+  salt_buf3[0] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[12]);
+  salt_buf3[1] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[13]);
+  salt_buf3[2] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[14]);
+  salt_buf3[3] = hc_swap32_S (salt_bufs[SALT_POS].salt_buf[15]);
+
+  const u32 salt_len = salt_bufs[SALT_POS].salt_len;
+
+  const u32 pw_salt_len = pw_len + salt_len;
+
+  /**
+   * loop
+   */
+
+  u32 w0l = w[0];
+
+  for (u32 il_pos = 0; il_pos < il_cnt; il_pos += VECT_SIZE)
+  {
+    const u32x w0r = ix_create_bft (bfs_buf, il_pos);
+
+    const u32x w0lr = w0l | w0r;
+
+    u32x w0_t[4];
+    u32x w1_t[4];
+    u32x w2_t[4];
+    u32x w3_t[4];
+
+    w0_t[0] = w0lr;
+    w0_t[1] = w[ 1];
+    w0_t[2] = w[ 2];
+    w0_t[3] = w[ 3];
+    w1_t[0] = w[ 4];
+    w1_t[1] = w[ 5];
+    w1_t[2] = w[ 6];
+    w1_t[3] = w[ 7];
+    w2_t[0] = w[ 8];
+    w2_t[1] = w[ 9];
+    w2_t[2] = w[10];
+    w2_t[3] = w[11];
+    w3_t[0] = w[12];
+    w3_t[1] = w[13];
+    w3_t[2] = w[14];
+    w3_t[3] = w[15];
+
+    switch_buffer_by_offset_be (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]  = 0;
+    w3_t[3]  = pw_salt_len * 8;
+
+    /**
+     * sha512
+     */
+
+    u64x digest[8];
+
+    digest[0] = SHA384M_A;
+    digest[1] = SHA384M_B;
+    digest[2] = SHA384M_C;
+    digest[3] = SHA384M_D;
+    digest[4] = SHA384M_E;
+    digest[5] = SHA384M_F;
+    digest[6] = SHA384M_G;
+    digest[7] = SHA384M_H;
+
+    sha384_transform_intern (w0_t, w1_t, w2_t, w3_t, digest);
+
+    const u32x r0 = l32_from_64 (digest[3]);
+    const u32x r1 = h32_from_64 (digest[3]);
+    const u32x r2 = l32_from_64 (digest[2]);
+    const u32x r3 = h32_from_64 (digest[2]);
+
+    COMPARE_S_SIMD (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_m04 (KERN_ATTR_BASIC ())
+{
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 w[16];
+
+  w[ 0] = pws[gid].i[ 0];
+  w[ 1] = pws[gid].i[ 1];
+  w[ 2] = pws[gid].i[ 2];
+  w[ 3] = pws[gid].i[ 3];
+  w[ 4] = 0;
+  w[ 5] = 0;
+  w[ 6] = 0;
+  w[ 7] = 0;
+  w[ 8] = 0;
+  w[ 9] = 0;
+  w[10] = 0;
+  w[11] = 0;
+  w[12] = 0;
+  w[13] = 0;
+  w[14] = 0;
+  w[15] = pws[gid].i[15];
+
+  const u32 pw_len = pws[gid].pw_len & 63;
+
+  /**
+   * main
+   */
+
+  m10820m (w, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, SALT_POS, loop_pos, loop_cnt, il_cnt, digests_cnt, DIGESTS_OFFSET, combs_mode, pws_pos, gid_max);
+}
+
+KERNEL_FQ void m10820_m08 (KERN_ATTR_BASIC ())
+{
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 w[16];
+
+  w[ 0] = pws[gid].i[ 0];
+  w[ 1] = pws[gid].i[ 1];
+  w[ 2] = pws[gid].i[ 2];
+  w[ 3] = pws[gid].i[ 3];
+  w[ 4] = pws[gid].i[ 4];
+  w[ 5] = pws[gid].i[ 5];
+  w[ 6] = pws[gid].i[ 6];
+  w[ 7] = pws[gid].i[ 7];
+  w[ 8] = 0;
+  w[ 9] = 0;
+  w[10] = 0;
+  w[11] = 0;
+  w[12] = 0;
+  w[13] = 0;
+  w[14] = 0;
+  w[15] = pws[gid].i[15];
+
+  const u32 pw_len = pws[gid].pw_len & 63;
+
+  /**
+   * main
+   */
+
+  m10820m (w, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, SALT_POS, loop_pos, loop_cnt, il_cnt, digests_cnt, DIGESTS_OFFSET, combs_mode, pws_pos, gid_max);
+}
+
+KERNEL_FQ void m10820_m16 (KERN_ATTR_BASIC ())
+{
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 w[16];
+
+  w[ 0] = pws[gid].i[ 0];
+  w[ 1] = pws[gid].i[ 1];
+  w[ 2] = pws[gid].i[ 2];
+  w[ 3] = pws[gid].i[ 3];
+  w[ 4] = pws[gid].i[ 4];
+  w[ 5] = pws[gid].i[ 5];
+  w[ 6] = pws[gid].i[ 6];
+  w[ 7] = pws[gid].i[ 7];
+  w[ 8] = pws[gid].i[ 8];
+  w[ 9] = pws[gid].i[ 9];
+  w[10] = pws[gid].i[10];
+  w[11] = pws[gid].i[11];
+  w[12] = pws[gid].i[12];
+  w[13] = pws[gid].i[13];
+  w[14] = pws[gid].i[14];
+  w[15] = pws[gid].i[15];
+
+  const u32 pw_len = pws[gid].pw_len & 63;
+
+  /**
+   * main
+   */
+
+  m10820m (w, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, SALT_POS, loop_pos, loop_cnt, il_cnt, digests_cnt, DIGESTS_OFFSET, combs_mode, pws_pos, gid_max);
+}
+
+KERNEL_FQ void m10820_s04 (KERN_ATTR_BASIC ())
+{
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 w[16];
+
+  w[ 0] = pws[gid].i[ 0];
+  w[ 1] = pws[gid].i[ 1];
+  w[ 2] = pws[gid].i[ 2];
+  w[ 3] = pws[gid].i[ 3];
+  w[ 4] = 0;
+  w[ 5] = 0;
+  w[ 6] = 0;
+  w[ 7] = 0;
+  w[ 8] = 0;
+  w[ 9] = 0;
+  w[10] = 0;
+  w[11] = 0;
+  w[12] = 0;
+  w[13] = 0;
+  w[14] = 0;
+  w[15] = pws[gid].i[15];
+
+  const u32 pw_len = pws[gid].pw_len & 63;
+
+  /**
+   * main
+   */
+
+  m10820s (w, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, SALT_POS, loop_pos, loop_cnt, il_cnt, digests_cnt, DIGESTS_OFFSET, combs_mode, pws_pos, gid_max);
+}
+
+KERNEL_FQ void m10820_s08 (KERN_ATTR_BASIC ())
+{
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 w[16];
+
+  w[ 0] = pws[gid].i[ 0];
+  w[ 1] = pws[gid].i[ 1];
+  w[ 2] = pws[gid].i[ 2];
+  w[ 3] = pws[gid].i[ 3];
+  w[ 4] = pws[gid].i[ 4];
+  w[ 5] = pws[gid].i[ 5];
+  w[ 6] = pws[gid].i[ 6];
+  w[ 7] = pws[gid].i[ 7];
+  w[ 8] = 0;
+  w[ 9] = 0;
+  w[10] = 0;
+  w[11] = 0;
+  w[12] = 0;
+  w[13] = 0;
+  w[14] = 0;
+  w[15] = pws[gid].i[15];
+
+  const u32 pw_len = pws[gid].pw_len & 63;
+
+  /**
+   * main
+   */
+
+  m10820s (w, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, SALT_POS, loop_pos, loop_cnt, il_cnt, digests_cnt, DIGESTS_OFFSET, combs_mode, pws_pos, gid_max);
+}
+
+KERNEL_FQ void m10820_s16 (KERN_ATTR_BASIC ())
+{
+  /**
+   * base
+   */
+
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  u32 w[16];
+
+  w[ 0] = pws[gid].i[ 0];
+  w[ 1] = pws[gid].i[ 1];
+  w[ 2] = pws[gid].i[ 2];
+  w[ 3] = pws[gid].i[ 3];
+  w[ 4] = pws[gid].i[ 4];
+  w[ 5] = pws[gid].i[ 5];
+  w[ 6] = pws[gid].i[ 6];
+  w[ 7] = pws[gid].i[ 7];
+  w[ 8] = pws[gid].i[ 8];
+  w[ 9] = pws[gid].i[ 9];
+  w[10] = pws[gid].i[10];
+  w[11] = pws[gid].i[11];
+  w[12] = pws[gid].i[12];
+  w[13] = pws[gid].i[13];
+  w[14] = pws[gid].i[14];
+  w[15] = pws[gid].i[15];
+
+  const u32 pw_len = pws[gid].pw_len & 63;
+
+  /**
+   * main
+   */
+
+  m10820s (w, pw_len, pws, rules_buf, combs_buf, bfs_buf, tmps, hooks, bitmaps_buf_s1_a, bitmaps_buf_s1_b, bitmaps_buf_s1_c, bitmaps_buf_s1_d, bitmaps_buf_s2_a, bitmaps_buf_s2_b, bitmaps_buf_s2_c, bitmaps_buf_s2_d, plains_buf, digests_buf, hashes_shown, salt_bufs, esalt_bufs, d_return_buf, d_extra0_buf, d_extra1_buf, d_extra2_buf, d_extra3_buf, bitmap_mask, bitmap_shift1, bitmap_shift2, SALT_POS, loop_pos, loop_cnt, il_cnt, digests_cnt, DIGESTS_OFFSET, combs_mode, pws_pos, gid_max);
+}

OpenCL/m10820_a3-pure.cl

@@ -0,0 +1,149 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+#define NEW_SIMD_CODE
+
+#ifdef KERNEL_STATIC
+#include "inc_vendor.h"
+#include "inc_types.h"
+#include "inc_platform.cl"
+#include "inc_common.cl"
+#include "inc_simd.cl"
+#include "inc_hash_sha384.cl"
+#endif
+
+KERNEL_FQ void m10820_mxx (KERN_ATTR_VECTOR ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+  const u64 gid = get_global_id (0);
+
+  if (gid >= gid_max) return;
+
+  /**
+   * base
+   */
+
+  const u32 pw_len = pws[gid].pw_len;
+
+  u32x w[64] = { 0 };
+
+  for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1)
+  {
+    w[idx] = pws[gid].i[idx];
+  }
+
+  sha384_ctx_t ctx0;
+
+  sha384_init (&ctx0);
+
+  sha384_update_global_swap (&ctx0, salt_bufs[SALT_POS].salt_buf, salt_bufs[SALT_POS].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;
+
+    sha384_ctx_vector_t ctx;
+
+    sha384_init_vector_from_scalar (&ctx, &ctx0);
+
+    sha384_update_vector (&ctx, w, pw_len);
+
+    sha384_final_vector (&ctx);
+
+    const u32x r0 = l32_from_64 (ctx.h[3]);
+    const u32x r1 = h32_from_64 (ctx.h[3]);
+    const u32x r2 = l32_from_64 (ctx.h[2]);
+    const u32x r3 = h32_from_64 (ctx.h[2]);
+
+    COMPARE_M_SIMD (r0, r1, r2, r3);
+  }
+}
+
+KERNEL_FQ void m10820_sxx (KERN_ATTR_VECTOR ())
+{
+  /**
+   * modifier
+   */
+
+  const u64 lid = get_local_id (0);
+  const u64 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;
+
+  u32x w[64] = { 0 };
+
+  for (u32 i = 0, idx = 0; i < pw_len; i += 4, idx += 1)
+  {
+    w[idx] = pws[gid].i[idx];
+  }
+
+  sha384_ctx_t ctx0;
+
+  sha384_init (&ctx0);
+
+  sha384_update_global_swap (&ctx0, salt_bufs[SALT_POS].salt_buf, salt_bufs[SALT_POS].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;
+
+    sha384_ctx_vector_t ctx;
+
+    sha384_init_vector_from_scalar (&ctx, &ctx0);
+
+    sha384_update_vector (&ctx, w, pw_len);
+
+    sha384_final_vector (&ctx);
+
+    const u32x r0 = l32_from_64 (ctx.h[3]);
+    const u32x r1 = h32_from_64 (ctx.h[3]);
+    const u32x r2 = l32_from_64 (ctx.h[2]);
+    const u32x r3 = h32_from_64 (ctx.h[2]);
+
+    COMPARE_S_SIMD (r0, r1, r2, r3);
+  }
+}

docs/changes.txt

@@ -8,6 +8,8 @@
 - Added hash-mode: bcrypt(sha1($pass)) / bcryptsha1
 - Added hash-mode: Linux Kernel Crypto API (2.4)
 - Added hash-mode: MurmurHash
+- Added hash-mode: sha384($salt.$pass)
+- Added hash-mode: sha384($pass.$salt)
 
 ##
 ## Improvements
@@ -70,7 +72,6 @@
 - Added hash-mode: Umbraco HMAC-SHA1
 - Added hash-mode: sha1($salt.sha1($pass.$salt))
 - Added hash-mode: sha1(sha1($pass).$salt)
-- Added hash-mode: sha384($pass.$salt)
 
 ##
 ## Features

docs/readme.txt

@@ -109,6 +109,7 @@ NVIDIA GPUs require "NVIDIA Driver" (440.64 or later) and "CUDA Toolkit" (9.0 or
 - sha256(sha256($pass).$salt)
 - sha256(utf16le($pass).$salt)
 - sha384($pass.$salt)
+- sha384($salt.$pass)
 - sha512($pass.$salt)
 - sha512($salt.$pass)
 - sha512($salt.utf16le($pass))

src/modules/module_10820.c

@@ -0,0 +1,289 @@
+/**
+ * Author......: See docs/credits.txt
+ * License.....: MIT
+ */
+
+#include "common.h"
+#include "types.h"
+#include "modules.h"
+#include "bitops.h"
+#include "convert.h"
+#include "shared.h"
+
+static const u32   ATTACK_EXEC    = ATTACK_EXEC_INSIDE_KERNEL;
+static const u32   DGST_POS0      = 6;
+static const u32   DGST_POS1      = 7;
+static const u32   DGST_POS2      = 4;
+static const u32   DGST_POS3      = 5;
+static const u32   DGST_SIZE      = DGST_SIZE_8_8;
+static const u32   HASH_CATEGORY  = HASH_CATEGORY_RAW_HASH_SALTED;
+static const char *HASH_NAME      = "sha384($salt.$pass)";
+static const u64   KERN_TYPE      = 10820;
+static const u32   OPTI_TYPE      = OPTI_TYPE_ZERO_BYTE
+                                  | OPTI_TYPE_PRECOMPUTE_INIT
+                                  | OPTI_TYPE_EARLY_SKIP
+                                  | OPTI_TYPE_NOT_ITERATED
+                                  | OPTI_TYPE_PREPENDED_SALT
+                                  | OPTI_TYPE_USES_BITS_64
+                                  | OPTI_TYPE_RAW_HASH;
+static const u64   OPTS_TYPE      = OPTS_TYPE_PT_GENERATE_BE
+                                  | OPTS_TYPE_PT_ADD80
+                                  | OPTS_TYPE_PT_ADDBITS15;
+static const u32   SALT_TYPE      = SALT_TYPE_GENERIC;
+static const char *ST_PASS        = "hashcat";
+static const char *ST_HASH        = "63f63d7f82d4a4cb6b9ff37a6bc7c5ec39faaf9c9078551f5cbf7960e76ded87b643d37ac53c45bc544325e7ff83a1f2:93362";
+
+u32         module_attack_exec    (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ATTACK_EXEC;     }
+u32         module_dgst_pos0      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS0;       }
+u32         module_dgst_pos1      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS1;       }
+u32         module_dgst_pos2      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS2;       }
+u32         module_dgst_pos3      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_POS3;       }
+u32         module_dgst_size      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return DGST_SIZE;       }
+u32         module_hash_category  (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return HASH_CATEGORY;   }
+const char *module_hash_name      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return HASH_NAME;       }
+u64         module_kern_type      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return KERN_TYPE;       }
+u32         module_opti_type      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return OPTI_TYPE;       }
+u64         module_opts_type      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return OPTS_TYPE;       }
+u32         module_salt_type      (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return SALT_TYPE;       }
+const char *module_st_hash        (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ST_HASH;         }
+const char *module_st_pass        (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra) { return ST_PASS;         }
+
+bool module_unstable_warning (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra, MAYBE_UNUSED const hc_device_param_t *device_param)
+{
+  if (device_param->opencl_platform_vendor_id == VENDOR_ID_APPLE)
+  {
+    // self-test failed
+    if ((device_param->opencl_device_vendor_id == VENDOR_ID_AMD) && (device_param->opencl_device_type & CL_DEVICE_TYPE_GPU))
+    {
+      return true;
+    }
+  }
+
+  return false;
+}
+
+char *module_jit_build_options (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const user_options_t *user_options, MAYBE_UNUSED const user_options_extra_t *user_options_extra, MAYBE_UNUSED const hashes_t *hashes, MAYBE_UNUSED const hc_device_param_t *device_param)
+{
+  char *jit_build_options = NULL;
+
+  // Extra treatment for Apple systems
+  if (device_param->opencl_platform_vendor_id == VENDOR_ID_APPLE)
+  {
+    return jit_build_options;
+  }
+
+  // ROCM
+  if ((device_param->opencl_device_vendor_id == VENDOR_ID_AMD) && (device_param->has_vperm == true))
+  {
+    hc_asprintf (&jit_build_options, "-D _unroll");
+  }
+
+  return jit_build_options;
+}
+
+int module_hash_decode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED void *digest_buf, MAYBE_UNUSED salt_t *salt, MAYBE_UNUSED void *esalt_buf, MAYBE_UNUSED void *hook_salt_buf, MAYBE_UNUSED hashinfo_t *hash_info, const char *line_buf, MAYBE_UNUSED const int line_len)
+{
+  u64 *digest = (u64 *) digest_buf;
+
+  token_t token;
+
+  token.token_cnt  = 2;
+
+  token.sep[0]     = hashconfig->separator;
+  token.len_min[0] = 96;
+  token.len_max[0] = 96;
+  token.attr[0]    = TOKEN_ATTR_VERIFY_LENGTH
+                   | TOKEN_ATTR_VERIFY_HEX;
+
+  token.len_min[1] = SALT_MIN;
+  token.len_max[1] = SALT_MAX;
+  token.attr[1]    = TOKEN_ATTR_VERIFY_LENGTH;
+
+  if (hashconfig->opts_type & OPTS_TYPE_ST_HEX)
+  {
+    token.len_min[1] *= 2;
+    token.len_max[1] *= 2;
+
+    token.attr[1] |= TOKEN_ATTR_VERIFY_HEX;
+  }
+
+  const int rc_tokenizer = input_tokenizer ((const u8 *) line_buf, line_len, &token);
+
+  if (rc_tokenizer != PARSER_OK) return (rc_tokenizer);
+
+  const u8 *hash_pos = token.buf[0];
+
+  digest[0] = hex_to_u64 (hash_pos +  0);
+  digest[1] = hex_to_u64 (hash_pos + 16);
+  digest[2] = hex_to_u64 (hash_pos + 32);
+  digest[3] = hex_to_u64 (hash_pos + 48);
+  digest[4] = hex_to_u64 (hash_pos + 64);
+  digest[5] = hex_to_u64 (hash_pos + 80);
+  digest[6] = 0;
+  digest[7] = 0;
+
+  digest[0] = byte_swap_64 (digest[0]);
+  digest[1] = byte_swap_64 (digest[1]);
+  digest[2] = byte_swap_64 (digest[2]);
+  digest[3] = byte_swap_64 (digest[3]);
+  digest[4] = byte_swap_64 (digest[4]);
+  digest[5] = byte_swap_64 (digest[5]);
+  digest[6] = 0;
+  digest[7] = 0;
+
+  if (hashconfig->opti_type & OPTI_TYPE_OPTIMIZED_KERNEL)
+  {
+    digest[0] -= SHA384M_A;
+    digest[1] -= SHA384M_B;
+    digest[2] -= SHA384M_C;
+    digest[3] -= SHA384M_D;
+    digest[4] -= SHA384M_E;
+    digest[5] -= SHA384M_F;
+    digest[6] -= 0;
+    digest[7] -= 0;
+  }
+
+  const u8 *salt_pos = token.buf[1];
+  const int salt_len = token.len[1];
+
+  const bool parse_rc = generic_salt_decode (hashconfig, salt_pos, salt_len, (u8 *) salt->salt_buf, (int *) &salt->salt_len);
+
+  if (parse_rc == false) return (PARSER_SALT_LENGTH);
+
+  return (PARSER_OK);
+}
+
+int module_hash_encode (MAYBE_UNUSED const hashconfig_t *hashconfig, MAYBE_UNUSED const void *digest_buf, MAYBE_UNUSED const salt_t *salt, MAYBE_UNUSED const void *esalt_buf, MAYBE_UNUSED const void *hook_salt_buf, MAYBE_UNUSED const hashinfo_t *hash_info, char *line_buf, MAYBE_UNUSED const int line_size)
+{
+  const u64 *digest = (const u64 *) digest_buf;
+
+  // we can not change anything in the original buffer, otherwise destroying sorting
+  // therefore create some local buffer
+
+  u64 tmp[8];
+
+  tmp[0] = digest[0];
+  tmp[1] = digest[1];
+  tmp[2] = digest[2];
+  tmp[3] = digest[3];
+  tmp[4] = digest[4];
+  tmp[5] = digest[5];
+  tmp[6] = 0;
+  tmp[7] = 0;
+
+  if (hashconfig->opti_type & OPTI_TYPE_OPTIMIZED_KERNEL)
+  {
+    tmp[0] += SHA384M_A;
+    tmp[1] += SHA384M_B;
+    tmp[2] += SHA384M_C;
+    tmp[3] += SHA384M_D;
+    tmp[4] += SHA384M_E;
+    tmp[5] += SHA384M_F;
+    tmp[6] += 0;
+    tmp[7] += 0;
+  }
+
+  tmp[0] = byte_swap_64 (tmp[0]);
+  tmp[1] = byte_swap_64 (tmp[1]);
+  tmp[2] = byte_swap_64 (tmp[2]);
+  tmp[3] = byte_swap_64 (tmp[3]);
+  tmp[4] = byte_swap_64 (tmp[4]);
+  tmp[5] = byte_swap_64 (tmp[5]);
+  tmp[6] = 0;
+  tmp[7] = 0;
+
+  u8 *out_buf = (u8 *) line_buf;
+
+  int out_len = 0;
+
+  u64_to_hex (tmp[0], out_buf + out_len); out_len += 16;
+  u64_to_hex (tmp[1], out_buf + out_len); out_len += 16;
+  u64_to_hex (tmp[2], out_buf + out_len); out_len += 16;
+  u64_to_hex (tmp[3], out_buf + out_len); out_len += 16;
+  u64_to_hex (tmp[4], out_buf + out_len); out_len += 16;
+  u64_to_hex (tmp[5], out_buf + out_len); out_len += 16;
+
+  out_buf[out_len] = hashconfig->separator;
+
+  out_len += 1;
+
+  out_len += generic_salt_encode (hashconfig, (const u8 *) salt->salt_buf, (const int) salt->salt_len, out_buf + out_len);
+
+  return out_len;
+}
+
+void module_init (module_ctx_t *module_ctx)
+{
+  module_ctx->module_context_size             = MODULE_CONTEXT_SIZE_CURRENT;
+  module_ctx->module_interface_version        = MODULE_INTERFACE_VERSION_CURRENT;
+
+  module_ctx->module_attack_exec              = module_attack_exec;
+  module_ctx->module_benchmark_esalt          = MODULE_DEFAULT;
+  module_ctx->module_benchmark_hook_salt      = MODULE_DEFAULT;
+  module_ctx->module_benchmark_mask           = MODULE_DEFAULT;
+  module_ctx->module_benchmark_salt           = MODULE_DEFAULT;
+  module_ctx->module_build_plain_postprocess  = MODULE_DEFAULT;
+  module_ctx->module_deep_comp_kernel         = MODULE_DEFAULT;
+  module_ctx->module_dgst_pos0                = module_dgst_pos0;
+  module_ctx->module_dgst_pos1                = module_dgst_pos1;
+  module_ctx->module_dgst_pos2                = module_dgst_pos2;
+  module_ctx->module_dgst_pos3                = module_dgst_pos3;
+  module_ctx->module_dgst_size                = module_dgst_size;
+  module_ctx->module_dictstat_disable         = MODULE_DEFAULT;
+  module_ctx->module_esalt_size               = MODULE_DEFAULT;
+  module_ctx->module_extra_buffer_size        = MODULE_DEFAULT;
+  module_ctx->module_extra_tmp_size           = MODULE_DEFAULT;
+  module_ctx->module_forced_outfile_format    = MODULE_DEFAULT;
+  module_ctx->module_hash_binary_count        = MODULE_DEFAULT;
+  module_ctx->module_hash_binary_parse        = MODULE_DEFAULT;
+  module_ctx->module_hash_binary_save         = MODULE_DEFAULT;
+  module_ctx->module_hash_decode_potfile      = MODULE_DEFAULT;
+  module_ctx->module_hash_decode_zero_hash    = MODULE_DEFAULT;
+  module_ctx->module_hash_decode              = module_hash_decode;
+  module_ctx->module_hash_encode_status       = MODULE_DEFAULT;
+  module_ctx->module_hash_encode_potfile      = MODULE_DEFAULT;
+  module_ctx->module_hash_encode              = module_hash_encode;
+  module_ctx->module_hash_init_selftest       = MODULE_DEFAULT;
+  module_ctx->module_hash_mode                = MODULE_DEFAULT;
+  module_ctx->module_hash_category            = module_hash_category;
+  module_ctx->module_hash_name                = module_hash_name;
+  module_ctx->module_hashes_count_min         = MODULE_DEFAULT;
+  module_ctx->module_hashes_count_max         = MODULE_DEFAULT;
+  module_ctx->module_hlfmt_disable            = MODULE_DEFAULT;
+  module_ctx->module_hook_extra_param_size    = MODULE_DEFAULT;
+  module_ctx->module_hook_extra_param_init    = MODULE_DEFAULT;
+  module_ctx->module_hook_extra_param_term    = MODULE_DEFAULT;
+  module_ctx->module_hook12                   = MODULE_DEFAULT;
+  module_ctx->module_hook23                   = MODULE_DEFAULT;
+  module_ctx->module_hook_salt_size           = MODULE_DEFAULT;
+  module_ctx->module_hook_size                = MODULE_DEFAULT;
+  module_ctx->module_jit_build_options        = module_jit_build_options;
+  module_ctx->module_jit_cache_disable        = MODULE_DEFAULT;
+  module_ctx->module_kernel_accel_max         = MODULE_DEFAULT;
+  module_ctx->module_kernel_accel_min         = MODULE_DEFAULT;
+  module_ctx->module_kernel_loops_max         = MODULE_DEFAULT;
+  module_ctx->module_kernel_loops_min         = MODULE_DEFAULT;
+  module_ctx->module_kernel_threads_max       = MODULE_DEFAULT;
+  module_ctx->module_kernel_threads_min       = MODULE_DEFAULT;
+  module_ctx->module_kern_type                = module_kern_type;
+  module_ctx->module_kern_type_dynamic        = MODULE_DEFAULT;
+  module_ctx->module_opti_type                = module_opti_type;
+  module_ctx->module_opts_type                = module_opts_type;
+  module_ctx->module_outfile_check_disable    = MODULE_DEFAULT;
+  module_ctx->module_outfile_check_nocomp     = MODULE_DEFAULT;
+  module_ctx->module_potfile_custom_check     = MODULE_DEFAULT;
+  module_ctx->module_potfile_disable          = MODULE_DEFAULT;
+  module_ctx->module_potfile_keep_all_hashes  = MODULE_DEFAULT;
+  module_ctx->module_pwdump_column            = MODULE_DEFAULT;
+  module_ctx->module_pw_max                   = MODULE_DEFAULT;
+  module_ctx->module_pw_min                   = MODULE_DEFAULT;
+  module_ctx->module_salt_max                 = MODULE_DEFAULT;
+  module_ctx->module_salt_min                 = MODULE_DEFAULT;
+  module_ctx->module_salt_type                = module_salt_type;
+  module_ctx->module_separator                = MODULE_DEFAULT;
+  module_ctx->module_st_hash                  = module_st_hash;
+  module_ctx->module_st_pass                  = module_st_pass;
+  module_ctx->module_tmp_size                 = MODULE_DEFAULT;
+  module_ctx->module_unstable_warning         = module_unstable_warning;
+  module_ctx->module_warmup_disable           = MODULE_DEFAULT;
+}

tools/test_modules/m10820.pm

@@ -0,0 +1,44 @@
+#!/usr/bin/env perl
+
+##
+## Author......: See docs/credits.txt
+## License.....: MIT
+##
+
+use strict;
+use warnings;
+
+use Digest::SHA qw (sha384_hex);
+
+sub module_constraints { [[0, 256], [0, 256], [0, 55], [0, 55], [0, 55]] }
+
+sub module_generate_hash
+{
+  my $word = shift;
+  my $salt = shift;
+
+  my $digest = sha384_hex ($salt . $word);
+
+  my $hash = sprintf ("%s:%s", $digest, $salt);
+
+  return $hash;
+}
+
+sub module_verify_hash
+{
+  my $line = shift;
+
+  my ($hash, $salt, $word) = split (':', $line);
+
+  return unless defined $hash;
+  return unless defined $salt;
+  return unless defined $word;
+
+  my $word_packed = pack_if_HEX_notation ($word);
+
+  my $new_hash = module_generate_hash ($word_packed, $salt);
+
+  return ($new_hash, $word);
+}
+
+1;