1
0
mirror of https://github.com/hashcat/hashcat.git synced 2024-11-15 20:39:17 +00:00
hashcat/OpenCL/m25200-pure.cl
2021-07-20 22:24:35 +02:00

219 lines
3.5 KiB
Common Lisp

/**
* 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_sha1.cl"
#endif
#define COMPARE_S "inc_comp_single.cl"
#define COMPARE_M "inc_comp_multi.cl"
#define SNMPV3_ENGINEID_MAX 32
#define SNMPV3_SALT_MAX 752
typedef struct hmac_sha1_tmp
{
u32 idx;
sha1_ctx_t ctx;
} hmac_sha1_tmp_t;
typedef struct snmpv3
{
u32 salt_buf[SNMPV3_SALT_MAX];
u32 salt_len;
u8 engineID_buf[SNMPV3_ENGINEID_MAX];
u32 engineID_len;
u8 packet_number[8+1];
} snmpv3_t;
KERNEL_FQ void m25200_init (KERN_ATTR_TMPS_ESALT (hmac_sha1_tmp_t, snmpv3_t))
{
/**
* modifier
*/
const u64 gid = get_global_id (0);
if (gid >= gid_max) return;
/**
* base
*/
const global u8 *pw_buf = (global u8 *) pws[gid].i;
const u32 pw_len = pws[gid].pw_len;
/**
* authkey
*/
u32 idx = 0;
u32 buf[32] = { 0 };
u8 *tmp_buf = (u8 *) buf;
sha1_ctx_t ctx;
sha1_init (&ctx);
for (int i = 0; i < 64; i++)
{
tmp_buf[i] = pw_buf[idx++];
if (idx >= pw_len) idx = 0;
}
sha1_update_swap (&ctx, buf, 64);
tmps[gid].idx = idx;
tmps[gid].ctx = ctx;
}
KERNEL_FQ void m25200_loop (KERN_ATTR_TMPS_ESALT (hmac_sha1_tmp_t, snmpv3_t))
{
/**
* base
*/
const u64 gid = get_global_id (0);
if (gid >= gid_max) return;
const global u8 *pw_buf = (global u8 *) pws[gid].i;
const u32 pw_len = pws[gid].pw_len;
u32 idx = tmps[gid].idx;
u32 buf[32] = { 0 };
u8 *tmp_buf = (u8 *) buf;
sha1_ctx_t ctx = tmps[gid].ctx;
for (u32 j = 0; j < loop_cnt; j++)
{
for (int i = 0; i < 64; i++)
{
tmp_buf[i] = pw_buf[idx++];
if (idx >= pw_len) idx = 0;
}
sha1_update_swap (&ctx, buf, 64);
}
tmps[gid].idx = idx;
tmps[gid].ctx = ctx;
}
KERNEL_FQ void m25200_comp (KERN_ATTR_TMPS_ESALT (hmac_sha1_tmp_t, snmpv3_t))
{
/**
* modifier
*/
const u64 gid = get_global_id (0);
if (gid >= gid_max) return;
const global u8 *engineID_buf = esalt_bufs[DIGESTS_OFFSET].engineID_buf;
u32 engineID_len = esalt_bufs[DIGESTS_OFFSET].engineID_len;
sha1_ctx_t ctx = tmps[gid].ctx;
sha1_final (&ctx);
const u32 h[5] = {
hc_swap32_S (ctx.h[0]),
hc_swap32_S (ctx.h[1]),
hc_swap32_S (ctx.h[2]),
hc_swap32_S (ctx.h[3]),
hc_swap32_S (ctx.h[4])
};
u32 tmp_buf[32] = { 0 };
tmp_buf[0] = h[0];
tmp_buf[1] = h[1];
tmp_buf[2] = h[2];
tmp_buf[3] = h[3];
tmp_buf[4] = h[4];
u8 *buf = (u8 *) (tmp_buf);
u32 i = 20;
u32 j;
for (j = 0; j < engineID_len; j++)
{
buf[i++] = engineID_buf[j];
}
for (j = 0; j < 20; j++)
{
buf[i++] = buf[j];
}
sha1_init (&ctx);
sha1_update_swap (&ctx, tmp_buf, i);
//sha1_update_swap (&ctx, (u32 *)buf, i);
sha1_final (&ctx);
u32 key[16] = { 0 };
key[0] = ctx.h[0];
key[1] = ctx.h[1];
key[2] = ctx.h[2];
key[3] = ctx.h[3];
key[4] = ctx.h[4];
u32 s[SNMPV3_SALT_MAX] = { 0 };
const u32 salt_len = esalt_bufs[DIGESTS_OFFSET].salt_len;
u32 idx;
for (i = 0, idx = 0; i < salt_len; i += 4, idx += 1)
{
s[idx] = esalt_bufs[DIGESTS_OFFSET].salt_buf[idx];
}
sha1_hmac_ctx_t hmac_ctx;
sha1_hmac_init (&hmac_ctx, key, 20);
sha1_hmac_update (&hmac_ctx, s, salt_len);
sha1_hmac_final (&hmac_ctx);
const u32 r0 = hmac_ctx.opad.h[DGST_R0];
const u32 r1 = hmac_ctx.opad.h[DGST_R1];
const u32 r2 = hmac_ctx.opad.h[DGST_R2];
const u32 r3 = 0;
#define il_pos 0
#ifdef KERNEL_STATIC
#include COMPARE_M
#endif
}