arno
/
hashcat
mirror of https://github.com/hashcat/hashcat.git
1
0
Fork 0
Code Issues Releases Wiki Activity

speedup -m 5800 for NV

Browse Source
pull/33/head
jsteube 9 years ago
parent f1cdf15540
commit 8057d89333
1 changed files with 73 additions and 146 deletions
Show Stats Download Patch File Download Diff File

219
OpenCL/m05800.cl
Unescape View File

@ -1091,165 +1091,92 @@ static void append_word (u32 w0[4], u32 w1[4], const u32 append[4], const u32 of
static void append_salt (u32 w0[4], u32 w1[4], u32 w2[4], const u32 append[5], const u32 offset) static void append_salt (u32 w0[4], u32 w1[4], u32 w2[4], const u32 append[5], const u32 offset)
{ {
switch (offset) u32 tmp0;
{ u32 tmp1;
case 2: u32 tmp2;
w0[0] = w0[0] | append[0] << 16; u32 tmp3;
w0[1] = append[0] >> 16 | append[1] << 16; u32 tmp4;
w0[2] = append[1] >> 16 | append[2] << 16; u32 tmp5;
w0[3] = append[2] >> 16 | append[3] << 16;
w1[0] = append[3] >> 16 | append[4] << 16;
w1[1] = append[4] >> 16;
break;
case 3:
w0[0] = w0[0] | append[0] << 24;
w0[1] = append[0] >> 8 | append[1] << 24;
w0[2] = append[1] >> 8 | append[2] << 24;
w0[3] = append[2] >> 8 | append[3] << 24;
w1[0] = append[3] >> 8 | append[4] << 24;
w1[1] = append[4] >> 8;
break;
case 4:
w0[1] = append[0];
w0[2] = append[1];
w0[3] = append[2];
w1[0] = append[3];
w1[1] = append[4];
break;
case 5: #ifdef IS_AMD
w0[1] = w0[1] | append[0] << 8;
w0[2] = append[0] >> 24 | append[1] << 8;
w0[3] = append[1] >> 24 | append[2] << 8;
w1[0] = append[2] >> 24 | append[3] << 8;
w1[1] = append[3] >> 24 | append[4] << 8;
w1[2] = append[4] >> 24;
break;
case 6: const int offset_minus_4 = 4 - (offset & 3);
w0[1] = w0[1] | append[0] << 16;
w0[2] = append[0] >> 16 | append[1] << 16;
w0[3] = append[1] >> 16 | append[2] << 16;
w1[0] = append[2] >> 16 | append[3] << 16;
w1[1] = append[3] >> 16 | append[4] << 16;
w1[2] = append[4] >> 16;
break;
case 7: tmp0 = amd_bytealign (append[0], 0, offset_minus_4);
w0[1] = w0[1] | append[0] << 24; tmp1 = amd_bytealign (append[1], append[0], offset_minus_4);
w0[2] = append[0] >> 8 | append[1] << 24; tmp2 = amd_bytealign (append[2], append[1], offset_minus_4);
w0[3] = append[1] >> 8 | append[2] << 24; tmp3 = amd_bytealign (append[3], append[2], offset_minus_4);
w1[0] = append[2] >> 8 | append[3] << 24; tmp4 = amd_bytealign (append[4], append[3], offset_minus_4);
w1[1] = append[3] >> 8 | append[4] << 24; tmp5 = amd_bytealign ( 0, append[4], offset_minus_4);
w1[2] = append[4] >> 8;
break;
case 8: const u32 mod = offset & 3;
w0[2] = append[0];
w0[3] = append[1];
w1[0] = append[2];
w1[1] = append[3];
w1[2] = append[4];
break;
case 9: if (mod == 0)
w0[2] = w0[2] | append[0] << 8; {
w0[3] = append[0] >> 24 | append[1] << 8; tmp0 = tmp1;
w1[0] = append[1] >> 24 | append[2] << 8; tmp1 = tmp2;
w1[1] = append[2] >> 24 | append[3] << 8; tmp2 = tmp3;
w1[2] = append[3] >> 24 | append[4] << 8; tmp3 = tmp4;
w1[3] = append[4] >> 24; tmp4 = tmp5;
break; tmp5 = 0;
}
case 10:
w0[2] = w0[2] | append[0] << 16;
w0[3] = append[0] >> 16 | append[1] << 16;
w1[0] = append[1] >> 16 | append[2] << 16;
w1[1] = append[2] >> 16 | append[3] << 16;
w1[2] = append[3] >> 16 | append[4] << 16;
w1[3] = append[4] >> 16;
break;
case 11:
w0[2] = w0[2] | append[0] << 24;
w0[3] = append[0] >> 8 | append[1] << 24;
w1[0] = append[1] >> 8 | append[2] << 24;
w1[1] = append[2] >> 8 | append[3] << 24;
w1[2] = append[3] >> 8 | append[4] << 24;
w1[3] = append[4] >> 8;
break;
case 12: #endif
w0[3] = append[0];
w1[0] = append[1];
w1[1] = append[2];
w1[2] = append[3];
w1[3] = append[4];
break;
case 13: #ifdef IS_NV
w0[3] = w0[3] | append[0] << 8;
w1[0] = append[0] >> 24 | append[1] << 8;
w1[1] = append[1] >> 24 | append[2] << 8;
w1[2] = append[2] >> 24 | append[3] << 8;
w1[3] = append[3] >> 24 | append[4] << 8;
w2[0] = append[4] >> 24;
break;
case 14: const int offset_minus_4 = 4 - (offset & 3);
w0[3] = w0[3] | append[0] << 16;
w1[0] = append[0] >> 16 | append[1] << 16;
w1[1] = append[1] >> 16 | append[2] << 16;
w1[2] = append[2] >> 16 | append[3] << 16;
w1[3] = append[3] >> 16 | append[4] << 16;
w2[0] = append[4] >> 16;
break;
case 15: const int selector = (0x76543210 >> (offset_minus_4 * 4)) & 0xffff;
w0[3] = w0[3] | append[0] << 24;
w1[0] = append[0] >> 8 | append[1] << 24;
w1[1] = append[1] >> 8 | append[2] << 24;
w1[2] = append[2] >> 8 | append[3] << 24;
w1[3] = append[3] >> 8 | append[4] << 24;
w2[0] = append[4] >> 8;
break;
case 16: tmp0 = __byte_perm ( 0, append[0], selector);
w1[0] = append[0]; tmp1 = __byte_perm (append[0], append[1], selector);
w1[1] = append[1]; tmp2 = __byte_perm (append[1], append[2], selector);
w1[2] = append[2]; tmp3 = __byte_perm (append[2], append[3], selector);
w1[3] = append[3]; tmp4 = __byte_perm (append[3], append[4], selector);
w2[0] = append[4]; tmp5 = __byte_perm (append[4], 0, selector);
break;
case 17: #endif
w1[0] = w1[0] | append[0] << 8;
w1[1] = append[0] >> 24 | append[1] << 8;
w1[2] = append[1] >> 24 | append[2] << 8;
w1[3] = append[2] >> 24 | append[3] << 8;
w2[0] = append[3] >> 24 | append[4] << 8;
w2[1] = append[4] >> 24;
break;
case 18: const u32 div = offset / 4;
w1[0] = w1[0] | append[0] << 16;
w1[1] = append[0] >> 16 | append[1] << 16;
w1[2] = append[1] >> 16 | append[2] << 16;
w1[3] = append[2] >> 16 | append[3] << 16;
w2[0] = append[3] >> 16 | append[4] << 16;
w2[1] = append[4] >> 16;
break;
case 19: switch (div)
w1[0] = w1[0] | append[0] << 24; {
w1[1] = append[0] >> 8 | append[1] << 24; case 0: w0[0] |= tmp0;
w1[2] = append[1] >> 8 | append[2] << 24; w0[1] = tmp1;
w1[3] = append[2] >> 8 | append[3] << 24; w0[2] = tmp2;
w2[0] = append[3] >> 8 | append[4] << 24; w0[3] = tmp3;
w2[1] = append[4] >> 8; w1[0] = tmp4;
break; w1[1] = tmp5;
break;
case 1: w0[1] |= tmp0;
w0[2] = tmp1;
w0[3] = tmp2;
w1[0] = tmp3;
w1[1] = tmp4;
w1[2] = tmp5;
break;
case 2: w0[2] |= tmp0;
w0[3] = tmp1;
w1[0] = tmp2;
w1[1] = tmp3;
w1[2] = tmp4;
w1[3] = tmp5;
break;
case 3: w0[3] |= tmp0;
w1[0] = tmp1;
w1[1] = tmp2;
w1[2] = tmp3;
w1[3] = tmp4;
w2[0] = tmp5;
break;
case 4: w1[0] |= tmp0;
w1[1] = tmp1;
w1[2] = tmp2;
w1[3] = tmp3;
w2[0] = tmp4;
w2[1] = tmp5;
break;
} }
} }

Write Preview
Loading…
Cancel
Save