1
0
mirror of https://github.com/hashcat/hashcat.git synced 2024-10-09 09:33:06 +00:00
hashcat/OpenCL/amp_a3.cl
jsteube 0bf4e3c34a - Dropped all vector code since new GPU's are all scalar, makes the code much easier
- Some performance on low-end GPU may drop because of that, but only for a few hash-modes
- Dropped scalar code (aka warp) since we do not have any vector datatypes anymore
- Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4
- Still need to fix kernels to new function names, needs to be done manually
- Temperature Management needs to be rewritten partially because of conflicting datatypes names
- Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id)
- Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD)
- First tests working for -m 0, for example
- Great performance increases in general for NV so far
- Tested amp_* and markov_* kernel
- Migrated special NV optimizations for rule processor
2015-12-15 12:04:22 +01:00

61 lines
1.5 KiB
Common Lisp

/**
* Author......: Jens Steube <jens.steube@gmail.com>
* License.....: MIT
*/
#include "include/constants.h"
#include "include/kernel_vendor.h"
#include "types_ocl.c"
__kernel void __attribute__((reqd_work_group_size (64, 1, 1))) amp (__global pw_t *pws, __global pw_t *pws_amp, __global gpu_rule_t *rules_buf, __global comb_t *combs_buf, __global bf_t *bfs_buf, const u32 combs_mode, const u32 gid_max)
{
const u32 gid = get_global_id (0);
if (gid >= gid_max) return;
const u32 pw_len = pws[gid].pw_len;
u32 w0[4];
u32 w1[4];
u32 w2[4];
u32 w3[4];
w0[0] = pws[gid].i[ 0];
w0[1] = pws[gid].i[ 1];
w0[2] = pws[gid].i[ 2];
w0[3] = pws[gid].i[ 3];
w1[0] = pws[gid].i[ 4];
w1[1] = pws[gid].i[ 5];
w1[2] = pws[gid].i[ 6];
w1[3] = pws[gid].i[ 7];
w2[0] = pws[gid].i[ 8];
w2[1] = pws[gid].i[ 9];
w2[2] = pws[gid].i[10];
w2[3] = pws[gid].i[11];
w3[0] = pws[gid].i[12];
w3[1] = pws[gid].i[13];
w3[2] = pws[gid].i[14];
w3[3] = pws[gid].i[15];
const u32 w0r = bfs_buf[0].i;
pws_amp[gid].i[ 0] = w0[0] | w0r;
pws_amp[gid].i[ 1] = w0[1];
pws_amp[gid].i[ 2] = w0[2];
pws_amp[gid].i[ 3] = w0[3];
pws_amp[gid].i[ 4] = w1[0];
pws_amp[gid].i[ 5] = w1[1];
pws_amp[gid].i[ 6] = w1[2];
pws_amp[gid].i[ 7] = w1[3];
pws_amp[gid].i[ 8] = w2[0];
pws_amp[gid].i[ 9] = w2[1];
pws_amp[gid].i[10] = w2[2];
pws_amp[gid].i[11] = w2[3];
pws_amp[gid].i[12] = w3[0];
pws_amp[gid].i[13] = w3[1];
pws_amp[gid].i[14] = w3[2];
pws_amp[gid].i[15] = w3[3];
pws_amp[gid].pw_len = pw_len;
}