|
|
|
@ -16,6 +16,113 @@
|
|
|
|
|
#include "inc_hash_sha1.cl"
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
const u32 replace_dots (u32 *w, const u32 idx, const u32 old_len, const u32 pw_len)
|
|
|
|
|
{
|
|
|
|
|
const u32 min_len = idx << 4; // 2 ^ 4 = 16 for each u32 w[4]
|
|
|
|
|
|
|
|
|
|
if (pw_len <= min_len) return 0;
|
|
|
|
|
|
|
|
|
|
const u32 max_len = pw_len - min_len - 1;
|
|
|
|
|
|
|
|
|
|
const u32 start_pos = (max_len < 15) ? max_len : 15;
|
|
|
|
|
|
|
|
|
|
u32 cur_len = old_len;
|
|
|
|
|
|
|
|
|
|
for (int pos = (int) start_pos; pos >= 0; pos--)
|
|
|
|
|
{
|
|
|
|
|
const u32 div = pos / 4;
|
|
|
|
|
const u32 mod = pos & 3;
|
|
|
|
|
const u32 sht = mod << 3;
|
|
|
|
|
|
|
|
|
|
if (((w[div] >> sht) & 0xff) == 0x2e) // '.'
|
|
|
|
|
{
|
|
|
|
|
w[div] += (cur_len - 0x2e) << sht;
|
|
|
|
|
|
|
|
|
|
cur_len = 0;
|
|
|
|
|
}
|
|
|
|
|
else
|
|
|
|
|
{
|
|
|
|
|
cur_len++;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return cur_len;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
const u32 replace_dot_by_len (u32 *w0, u32 *w1, u32 *w2, u32 *w3, const u32 pw_len)
|
|
|
|
|
{
|
|
|
|
|
u32 cur_len = 0;
|
|
|
|
|
|
|
|
|
|
// loop over w3...w0 (4 * 16 = 64 bytes):
|
|
|
|
|
|
|
|
|
|
cur_len = replace_dots (w3, 3, cur_len, pw_len);
|
|
|
|
|
cur_len = replace_dots (w2, 2, cur_len, pw_len);
|
|
|
|
|
cur_len = replace_dots (w1, 1, cur_len, pw_len);
|
|
|
|
|
cur_len = replace_dots (w0, 0, cur_len, pw_len);
|
|
|
|
|
|
|
|
|
|
return cur_len;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#define REPLACE_DOT_BY_LEN_VECT(n) \
|
|
|
|
|
if (out_len.s##n > 0) \
|
|
|
|
|
{ \
|
|
|
|
|
u32 tmp0[4]; \
|
|
|
|
|
\
|
|
|
|
|
tmp0[0] = w0_t[0].s##n; \
|
|
|
|
|
tmp0[1] = w0_t[1].s##n; \
|
|
|
|
|
tmp0[2] = w0_t[2].s##n; \
|
|
|
|
|
tmp0[3] = w0_t[3].s##n; \
|
|
|
|
|
\
|
|
|
|
|
u32 tmp1[4]; \
|
|
|
|
|
\
|
|
|
|
|
tmp1[0] = w1_t[0].s##n; \
|
|
|
|
|
tmp1[1] = w1_t[1].s##n; \
|
|
|
|
|
tmp1[2] = w1_t[2].s##n; \
|
|
|
|
|
tmp1[3] = w1_t[3].s##n; \
|
|
|
|
|
\
|
|
|
|
|
u32 tmp2[4]; \
|
|
|
|
|
\
|
|
|
|
|
tmp2[0] = w2_t[0].s##n; \
|
|
|
|
|
tmp2[1] = w2_t[1].s##n; \
|
|
|
|
|
tmp2[2] = w2_t[2].s##n; \
|
|
|
|
|
tmp2[3] = w2_t[3].s##n; \
|
|
|
|
|
\
|
|
|
|
|
u32 tmp3[4]; \
|
|
|
|
|
\
|
|
|
|
|
tmp3[0] = w3_t[0].s##n; \
|
|
|
|
|
tmp3[1] = w3_t[1].s##n; \
|
|
|
|
|
tmp3[2] = w3_t[2].s##n; \
|
|
|
|
|
tmp3[3] = w3_t[3].s##n; \
|
|
|
|
|
\
|
|
|
|
|
const u32 len = replace_dot_by_len (tmp0, tmp1, tmp2, tmp3, out_len.s##n); \
|
|
|
|
|
\
|
|
|
|
|
switch_buffer_by_offset_le_S (tmp0, tmp1, tmp2, tmp3, 1); \
|
|
|
|
|
\
|
|
|
|
|
tmp0[0] |= len & 0xff; \
|
|
|
|
|
\
|
|
|
|
|
w0_t[0].s##n = tmp0[0]; \
|
|
|
|
|
w0_t[1].s##n = tmp0[1]; \
|
|
|
|
|
w0_t[2].s##n = tmp0[2]; \
|
|
|
|
|
w0_t[3].s##n = tmp0[3]; \
|
|
|
|
|
\
|
|
|
|
|
w1_t[0].s##n = tmp1[0]; \
|
|
|
|
|
w1_t[1].s##n = tmp1[1]; \
|
|
|
|
|
w1_t[2].s##n = tmp1[2]; \
|
|
|
|
|
w1_t[3].s##n = tmp1[3]; \
|
|
|
|
|
\
|
|
|
|
|
w2_t[0].s##n = tmp2[0]; \
|
|
|
|
|
w2_t[1].s##n = tmp2[1]; \
|
|
|
|
|
w2_t[2].s##n = tmp2[2]; \
|
|
|
|
|
w2_t[3].s##n = tmp2[3]; \
|
|
|
|
|
\
|
|
|
|
|
w3_t[0].s##n = tmp3[0]; \
|
|
|
|
|
w3_t[1].s##n = tmp3[1]; \
|
|
|
|
|
w3_t[2].s##n = tmp3[2]; \
|
|
|
|
|
w3_t[3].s##n = tmp3[3]; \
|
|
|
|
|
\
|
|
|
|
|
out_len.s##n++; \
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
KERNEL_FQ void m08300_m04 (KERN_ATTR_RULES ())
|
|
|
|
|
{
|
|
|
|
|
/**
|
|
|
|
@ -91,7 +198,7 @@ KERNEL_FQ void m08300_m04 (KERN_ATTR_RULES ())
|
|
|
|
|
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);
|
|
|
|
|
u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* salt
|
|
|
|
@ -119,9 +226,44 @@ KERNEL_FQ void m08300_m04 (KERN_ATTR_RULES ())
|
|
|
|
|
w3_t[2] = w3[2];
|
|
|
|
|
w3_t[3] = w3[3];
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, 1);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= out_len & 0xff;
|
|
|
|
|
// replace "." with the length:
|
|
|
|
|
|
|
|
|
|
#if VECT_SIZE == 1
|
|
|
|
|
if (out_len > 0)
|
|
|
|
|
{
|
|
|
|
|
const u32 len = replace_dot_by_len (w0_t, w1_t, w2_t, w3_t, out_len);
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, 1);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= len & 0xff;
|
|
|
|
|
|
|
|
|
|
out_len++;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 2
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (0)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (1)
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 4
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (2)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (3)
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 8
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (4)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (5)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (6)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (7)
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 16
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (8)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (9)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (a)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (b)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (c)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (d)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (e)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (f)
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
u32x s0[4];
|
|
|
|
|
u32x s1[4];
|
|
|
|
@ -145,7 +287,7 @@ KERNEL_FQ void m08300_m04 (KERN_ATTR_RULES ())
|
|
|
|
|
s3[2] = 0;
|
|
|
|
|
s3[3] = 0;
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, 1 + out_len);
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, out_len);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= s0[0];
|
|
|
|
|
w0_t[1] |= s0[1];
|
|
|
|
@ -181,7 +323,7 @@ KERNEL_FQ void m08300_m04 (KERN_ATTR_RULES ())
|
|
|
|
|
s3[2] = 0;
|
|
|
|
|
s3[3] = 0;
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, 1 + out_len + domain_len + 1);
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, out_len + domain_len + 1);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= s0[0];
|
|
|
|
|
w0_t[1] |= s0[1];
|
|
|
|
@ -219,7 +361,7 @@ KERNEL_FQ void m08300_m04 (KERN_ATTR_RULES ())
|
|
|
|
|
w3_t[0] = hc_swap32 (w3_t[0]);
|
|
|
|
|
w3_t[1] = hc_swap32 (w3_t[1]);
|
|
|
|
|
w3_t[2] = 0;
|
|
|
|
|
w3_t[3] = (1 + out_len + domain_len + 1 + salt_len) * 8;
|
|
|
|
|
w3_t[3] = (out_len + domain_len + 1 + salt_len) * 8;
|
|
|
|
|
|
|
|
|
|
u32x digest[5];
|
|
|
|
|
|
|
|
|
@ -360,7 +502,7 @@ KERNEL_FQ void m08300_s04 (KERN_ATTR_RULES ())
|
|
|
|
|
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);
|
|
|
|
|
u32x out_len = apply_rules_vect_optimized (pw_buf0, pw_buf1, pw_len, rules_buf, il_pos, w0, w1);
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* salt
|
|
|
|
@ -388,9 +530,44 @@ KERNEL_FQ void m08300_s04 (KERN_ATTR_RULES ())
|
|
|
|
|
w3_t[2] = w3[2];
|
|
|
|
|
w3_t[3] = w3[3];
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, 1);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= out_len & 0xff;
|
|
|
|
|
// replace "." with the length:
|
|
|
|
|
|
|
|
|
|
#if VECT_SIZE == 1
|
|
|
|
|
if (out_len > 0)
|
|
|
|
|
{
|
|
|
|
|
const u32 len = replace_dot_by_len (w0_t, w1_t, w2_t, w3_t, out_len);
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le (w0_t, w1_t, w2_t, w3_t, 1);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= len & 0xff;
|
|
|
|
|
|
|
|
|
|
out_len++;
|
|
|
|
|
}
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 2
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (0)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (1)
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 4
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (2)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (3)
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 8
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (4)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (5)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (6)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (7)
|
|
|
|
|
#endif
|
|
|
|
|
#if VECT_SIZE >= 16
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (8)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (9)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (a)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (b)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (c)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (d)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (e)
|
|
|
|
|
REPLACE_DOT_BY_LEN_VECT (f)
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
u32x s0[4];
|
|
|
|
|
u32x s1[4];
|
|
|
|
@ -414,7 +591,7 @@ KERNEL_FQ void m08300_s04 (KERN_ATTR_RULES ())
|
|
|
|
|
s3[2] = 0;
|
|
|
|
|
s3[3] = 0;
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, 1 + out_len);
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, out_len);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= s0[0];
|
|
|
|
|
w0_t[1] |= s0[1];
|
|
|
|
@ -450,7 +627,7 @@ KERNEL_FQ void m08300_s04 (KERN_ATTR_RULES ())
|
|
|
|
|
s3[2] = 0;
|
|
|
|
|
s3[3] = 0;
|
|
|
|
|
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, 1 + out_len + domain_len + 1);
|
|
|
|
|
switch_buffer_by_offset_le_VV (s0, s1, s2, s3, out_len + domain_len + 1);
|
|
|
|
|
|
|
|
|
|
w0_t[0] |= s0[0];
|
|
|
|
|
w0_t[1] |= s0[1];
|
|
|
|
@ -488,7 +665,7 @@ KERNEL_FQ void m08300_s04 (KERN_ATTR_RULES ())
|
|
|
|
|
w3_t[0] = hc_swap32 (w3_t[0]);
|
|
|
|
|
w3_t[1] = hc_swap32 (w3_t[1]);
|
|
|
|
|
w3_t[2] = 0;
|
|
|
|
|
w3_t[3] = (1 + out_len + domain_len + 1 + salt_len) * 8;
|
|
|
|
|
w3_t[3] = (out_len + domain_len + 1 + salt_len) * 8;
|
|
|
|
|
|
|
|
|
|
u32x digest[5];
|
|
|
|
|
|
|
|
|
|