1
0
mirror of https://github.com/hashcat/hashcat.git synced 2024-11-26 18:08:20 +00:00

Move autotune related stuff to autotune.c

This commit is contained in:
jsteube 2016-09-14 20:06:27 +02:00
parent 963cda3db6
commit e94ab8c113
9 changed files with 341 additions and 350 deletions

View File

@ -6,4 +6,6 @@
#ifndef _ATTACK_MODE_H #ifndef _ATTACK_MODE_H
#define _ATTACK_MODE_H #define _ATTACK_MODE_H
#define ATTACK_MODE 0
#endif // _ATTACK_MODE_H #endif // _ATTACK_MODE_H

View File

@ -6,4 +6,8 @@
#ifndef _AUTOTUNE_H #ifndef _AUTOTUNE_H
#define _AUTOTUNE_H #define _AUTOTUNE_H
#define OPENCL_VECTOR_WIDTH 0
int autotune (hc_device_param_t *device_param, hashconfig_t *hashconfig);
#endif // _AUTOTUNE_H #endif // _AUTOTUNE_H

View File

@ -6,6 +6,20 @@
#ifndef _DATA_H #ifndef _DATA_H
#define _DATA_H #define _DATA_H
#define global_free(attr) \
{ \
myfree ((void *) data.attr); \
\
data.attr = NULL; \
}
#define local_free(attr) \
{ \
myfree ((void *) attr); \
\
attr = NULL; \
}
typedef struct typedef struct
{ {
/** /**

View File

@ -6,4 +6,6 @@
#ifndef _POWERTUNE_H #ifndef _POWERTUNE_H
#define _POWERTUNE_H #define _POWERTUNE_H
#define POWERTUNE_ENABLE 0
#endif // _POWERTUNE_H #endif // _POWERTUNE_H

View File

@ -6,4 +6,7 @@
#ifndef _REMOVE_H #ifndef _REMOVE_H
#define _REMOVE_H #define _REMOVE_H
#define REMOVE 0
#define REMOVE_TIMER 60
#endif // _REMOVE_H #endif // _REMOVE_H

View File

@ -6,4 +6,6 @@
#ifndef _RUNTIME_H #ifndef _RUNTIME_H
#define _RUNTIME_H #define _RUNTIME_H
#define RUNTIME 0
#endif // _RUNTIME_H #endif // _RUNTIME_H

View File

@ -4,4 +4,308 @@
*/ */
#include "common.h" #include "common.h"
#include "types_int.h"
#include "types.h"
#include "logging.h"
#include "interface.h"
#include "timer.h"
#include "ext_OpenCL.h"
#include "ext_ADL.h"
#include "ext_nvapi.h"
#include "ext_nvml.h"
#include "ext_xnvctrl.h"
#include "hwmon.h"
#include "mpsp.h"
#include "rp_cpu.h"
#include "restore.h"
#include "opencl.h"
#include "outfile.h"
#include "potfile.h"
#include "debugfile.h"
#include "loopback.h"
#include "data.h"
#include "status.h"
#include "autotune.h" #include "autotune.h"
extern hc_global_data_t data;
static double TARGET_MS_PROFILE[4] = { 2, 12, 96, 480 };
static double try_run (hc_device_param_t *device_param, hashconfig_t *hashconfig, const u32 kernel_accel, const u32 kernel_loops)
{
const u32 kernel_power_try = device_param->device_processors * device_param->kernel_threads * kernel_accel;
device_param->kernel_params_buf32[28] = 0;
device_param->kernel_params_buf32[29] = kernel_loops; // not a bug, both need to be set
device_param->kernel_params_buf32[30] = kernel_loops; // because there's two variables for inner iters for slow and fast hashes
if (hashconfig->attack_exec == ATTACK_EXEC_INSIDE_KERNEL)
{
run_kernel (KERN_RUN_1, device_param, kernel_power_try, true, 0, hashconfig);
}
else
{
run_kernel (KERN_RUN_2, device_param, kernel_power_try, true, 0, hashconfig);
}
const double exec_ms_prev = get_avg_exec_time (device_param, 1);
return exec_ms_prev;
}
int autotune (hc_device_param_t *device_param, hashconfig_t *hashconfig)
{
const double target_ms = TARGET_MS_PROFILE[data.workload_profile - 1];
const u32 kernel_accel_min = device_param->kernel_accel_min;
const u32 kernel_accel_max = device_param->kernel_accel_max;
const u32 kernel_loops_min = device_param->kernel_loops_min;
const u32 kernel_loops_max = device_param->kernel_loops_max;
u32 kernel_accel = kernel_accel_min;
u32 kernel_loops = kernel_loops_min;
// in this case the user specified a fixed -u and -n on the commandline
// no way to tune anything
// but we need to run a few caching rounds
if ((kernel_loops_min == kernel_loops_max) && (kernel_accel_min == kernel_accel_max))
{
if (hashconfig->hash_mode != 2000)
{
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
}
device_param->kernel_accel = kernel_accel;
device_param->kernel_loops = kernel_loops;
const u32 kernel_power = device_param->device_processors * device_param->kernel_threads * device_param->kernel_accel;
device_param->kernel_power = kernel_power;
return 0;
}
// from here it's clear we are allowed to autotune
// so let's init some fake words
const u32 kernel_power_max = device_param->device_processors * device_param->kernel_threads * kernel_accel_max;
if (data.attack_kern == ATTACK_KERN_BF)
{
run_kernel_memset (device_param, device_param->d_pws_buf, 7, kernel_power_max * sizeof (pw_t));
}
else
{
for (u32 i = 0; i < kernel_power_max; i++)
{
device_param->pws_buf[i].i[0] = i;
device_param->pws_buf[i].i[1] = 0x01234567;
device_param->pws_buf[i].pw_len = 7 + (i & 7);
}
cl_int CL_err = hc_clEnqueueWriteBuffer (data.ocl, device_param->command_queue, device_param->d_pws_buf, CL_TRUE, 0, kernel_power_max * sizeof (pw_t), device_param->pws_buf, 0, NULL, NULL);
if (CL_err != CL_SUCCESS)
{
log_error ("ERROR: clEnqueueWriteBuffer(): %s\n", val2cstr_cl (CL_err));
return -1;
}
}
if (hashconfig->attack_exec == ATTACK_EXEC_INSIDE_KERNEL)
{
if (data.kernel_rules_cnt > 1)
{
cl_int CL_err = hc_clEnqueueCopyBuffer (data.ocl, device_param->command_queue, device_param->d_rules, device_param->d_rules_c, 0, 0, MIN (kernel_loops_max, KERNEL_RULES) * sizeof (kernel_rule_t), 0, NULL, NULL);
if (CL_err != CL_SUCCESS)
{
log_error ("ERROR: clEnqueueCopyBuffer(): %s\n", val2cstr_cl (CL_err));
return -1;
}
}
}
else
{
run_kernel_amp (device_param, kernel_power_max);
}
#define VERIFIER_CNT 1
// first find out highest kernel-loops that stays below target_ms
if (kernel_loops_min < kernel_loops_max)
{
for (kernel_loops = kernel_loops_max; kernel_loops > kernel_loops_min; kernel_loops >>= 1)
{
double exec_ms = try_run (device_param, hashconfig, kernel_accel_min, kernel_loops);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_v = try_run (device_param, hashconfig, kernel_accel_min, kernel_loops);
exec_ms = MIN (exec_ms, exec_ms_v);
}
if (exec_ms < target_ms) break;
}
}
// now the same for kernel-accel but with the new kernel-loops from previous loop set
#define STEPS_CNT 10
if (kernel_accel_min < kernel_accel_max)
{
for (int i = 0; i < STEPS_CNT; i++)
{
const u32 kernel_accel_try = 1u << i;
if (kernel_accel_try < kernel_accel_min) continue;
if (kernel_accel_try > kernel_accel_max) break;
double exec_ms = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_v = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops);
exec_ms = MIN (exec_ms, exec_ms_v);
}
if (exec_ms > target_ms) break;
kernel_accel = kernel_accel_try;
}
}
// at this point we want to know the actual runtime for the following reason:
// we need a reference for the balancing loop following up, and this
// the balancing loop can have an effect that the creates a new opportunity, for example:
// if the target is 95 ms and the current runtime is 48ms the above loop
// stopped the execution because the previous exec_ms was > 95ms
// due to the rebalance it's possible that the runtime reduces from 48ms to 47ms
// and this creates the possibility to double the workload -> 47 * 2 = 95ms, which is < 96ms
double exec_ms_pre_final = try_run (device_param, hashconfig, kernel_accel, kernel_loops);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_pre_final_v = try_run (device_param, hashconfig, kernel_accel, kernel_loops);
exec_ms_pre_final = MIN (exec_ms_pre_final, exec_ms_pre_final_v);
}
u32 diff = kernel_loops - kernel_accel;
if ((kernel_loops_min < kernel_loops_max) && (kernel_accel_min < kernel_accel_max))
{
u32 kernel_accel_orig = kernel_accel;
u32 kernel_loops_orig = kernel_loops;
for (u32 f = 1; f < 1024; f++)
{
const u32 kernel_accel_try = kernel_accel_orig * f;
const u32 kernel_loops_try = kernel_loops_orig / f;
if (kernel_accel_try > kernel_accel_max) break;
if (kernel_loops_try < kernel_loops_min) break;
u32 diff_new = kernel_loops_try - kernel_accel_try;
if (diff_new > diff) break;
diff_new = diff;
double exec_ms = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops_try);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_v = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops_try);
exec_ms = MIN (exec_ms, exec_ms_v);
}
if (exec_ms < exec_ms_pre_final)
{
exec_ms_pre_final = exec_ms;
kernel_accel = kernel_accel_try;
kernel_loops = kernel_loops_try;
}
}
}
const double exec_left = target_ms / exec_ms_pre_final;
const double accel_left = kernel_accel_max / kernel_accel;
const double exec_accel_min = MIN (exec_left, accel_left); // we want that to be int
if (exec_accel_min >= 1.0)
{
// this is safe to not overflow kernel_accel_max because of accel_left
kernel_accel *= (u32) exec_accel_min;
}
// reset them fake words
/*
memset (device_param->pws_buf, 0, kernel_power_max * sizeof (pw_t));
hc_clEnqueueWriteBuffer (data.ocl, device_param->command_queue, device_param->d_pws_buf, CL_TRUE, 0, kernel_power_max * sizeof (pw_t), device_param->pws_buf, 0, NULL, NULL);
hc_clEnqueueWriteBuffer (data.ocl, device_param->command_queue, device_param->d_pws_amp_buf, CL_TRUE, 0, kernel_power_max * sizeof (pw_t), device_param->pws_buf, 0, NULL, NULL);
*/
run_kernel_memset (device_param, device_param->d_pws_buf, 0, kernel_power_max * sizeof (pw_t));
if (hashconfig->attack_exec == ATTACK_EXEC_OUTSIDE_KERNEL)
{
run_kernel_memset (device_param, device_param->d_pws_amp_buf, 0, kernel_power_max * sizeof (pw_t));
}
// reset timer
device_param->exec_pos = 0;
memset (device_param->exec_ms, 0, EXEC_CACHE * sizeof (double));
memset (device_param->exec_us_prev1, 0, EXPECTED_ITERATIONS * sizeof (double));
memset (device_param->exec_us_prev2, 0, EXPECTED_ITERATIONS * sizeof (double));
memset (device_param->exec_us_prev3, 0, EXPECTED_ITERATIONS * sizeof (double));
// store
device_param->kernel_accel = kernel_accel;
device_param->kernel_loops = kernel_loops;
const u32 kernel_power = device_param->device_processors * device_param->kernel_threads * device_param->kernel_accel;
device_param->kernel_power = kernel_power;
#if defined (DEBUG)
if (data.quiet == 0)
{
clear_prompt ();
log_info ("- Device #%u: autotuned kernel-accel to %u\n"
"- Device #%u: autotuned kernel-loops to %u\n",
device_param->device_id + 1, kernel_accel,
device_param->device_id + 1, kernel_loops);
send_prompt ();
}
#endif
return 0;
}

View File

@ -91,6 +91,7 @@ extern int SUPPRESS_OUTPUT;
extern hc_thread_mutex_t mux_hwmon; extern hc_thread_mutex_t mux_hwmon;
extern hc_thread_mutex_t mux_display; extern hc_thread_mutex_t mux_display;
extern hc_thread_mutex_t mux_counter; extern hc_thread_mutex_t mux_counter;
extern hc_thread_mutex_t mux_dispatcher;
extern void (*get_next_word_func) (char *, u32, u32 *, u32 *); extern void (*get_next_word_func) (char *, u32, u32 *, u32 *);
@ -109,357 +110,8 @@ const int comptime = COMPTIME;
// remove
#define REMOVE 0
#define REMOVE_TIMER 60
// runtime
#define RUNTIME 0
// attack_mode
#define ATTACK_MODE 0
// powertune
#define POWERTUNE_ENABLE 0
// autotune
#define OPENCL_VECTOR_WIDTH 0
static double TARGET_MS_PROFILE[4] = { 2, 12, 96, 480 };
// thread
static hc_thread_mutex_t mux_dispatcher;
// data
#define global_free(attr) \
{ \
myfree ((void *) data.attr); \
\
data.attr = NULL; \
}
#define local_free(attr) \
{ \
myfree ((void *) attr); \
\
attr = NULL; \
}
static double try_run (hc_device_param_t *device_param, hashconfig_t *hashconfig, const u32 kernel_accel, const u32 kernel_loops)
{
const u32 kernel_power_try = device_param->device_processors * device_param->kernel_threads * kernel_accel;
device_param->kernel_params_buf32[28] = 0;
device_param->kernel_params_buf32[29] = kernel_loops; // not a bug, both need to be set
device_param->kernel_params_buf32[30] = kernel_loops; // because there's two variables for inner iters for slow and fast hashes
if (hashconfig->attack_exec == ATTACK_EXEC_INSIDE_KERNEL)
{
run_kernel (KERN_RUN_1, device_param, kernel_power_try, true, 0, hashconfig);
}
else
{
run_kernel (KERN_RUN_2, device_param, kernel_power_try, true, 0, hashconfig);
}
const double exec_ms_prev = get_avg_exec_time (device_param, 1);
return exec_ms_prev;
}
static int autotune (hc_device_param_t *device_param, hashconfig_t *hashconfig)
{
const double target_ms = TARGET_MS_PROFILE[data.workload_profile - 1];
const u32 kernel_accel_min = device_param->kernel_accel_min;
const u32 kernel_accel_max = device_param->kernel_accel_max;
const u32 kernel_loops_min = device_param->kernel_loops_min;
const u32 kernel_loops_max = device_param->kernel_loops_max;
u32 kernel_accel = kernel_accel_min;
u32 kernel_loops = kernel_loops_min;
// in this case the user specified a fixed -u and -n on the commandline
// no way to tune anything
// but we need to run a few caching rounds
if ((kernel_loops_min == kernel_loops_max) && (kernel_accel_min == kernel_accel_max))
{
if (hashconfig->hash_mode != 2000)
{
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
try_run (device_param, hashconfig, kernel_accel, kernel_loops);
}
device_param->kernel_accel = kernel_accel;
device_param->kernel_loops = kernel_loops;
const u32 kernel_power = device_param->device_processors * device_param->kernel_threads * device_param->kernel_accel;
device_param->kernel_power = kernel_power;
return 0;
}
// from here it's clear we are allowed to autotune
// so let's init some fake words
const u32 kernel_power_max = device_param->device_processors * device_param->kernel_threads * kernel_accel_max;
if (data.attack_kern == ATTACK_KERN_BF)
{
run_kernel_memset (device_param, device_param->d_pws_buf, 7, kernel_power_max * sizeof (pw_t));
}
else
{
for (u32 i = 0; i < kernel_power_max; i++)
{
device_param->pws_buf[i].i[0] = i;
device_param->pws_buf[i].i[1] = 0x01234567;
device_param->pws_buf[i].pw_len = 7 + (i & 7);
}
cl_int CL_err = hc_clEnqueueWriteBuffer (data.ocl, device_param->command_queue, device_param->d_pws_buf, CL_TRUE, 0, kernel_power_max * sizeof (pw_t), device_param->pws_buf, 0, NULL, NULL);
if (CL_err != CL_SUCCESS)
{
log_error ("ERROR: clEnqueueWriteBuffer(): %s\n", val2cstr_cl (CL_err));
return -1;
}
}
if (hashconfig->attack_exec == ATTACK_EXEC_INSIDE_KERNEL)
{
if (data.kernel_rules_cnt > 1)
{
cl_int CL_err = hc_clEnqueueCopyBuffer (data.ocl, device_param->command_queue, device_param->d_rules, device_param->d_rules_c, 0, 0, MIN (kernel_loops_max, KERNEL_RULES) * sizeof (kernel_rule_t), 0, NULL, NULL);
if (CL_err != CL_SUCCESS)
{
log_error ("ERROR: clEnqueueCopyBuffer(): %s\n", val2cstr_cl (CL_err));
return -1;
}
}
}
else
{
run_kernel_amp (device_param, kernel_power_max);
}
#define VERIFIER_CNT 1
// first find out highest kernel-loops that stays below target_ms
if (kernel_loops_min < kernel_loops_max)
{
for (kernel_loops = kernel_loops_max; kernel_loops > kernel_loops_min; kernel_loops >>= 1)
{
double exec_ms = try_run (device_param, hashconfig, kernel_accel_min, kernel_loops);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_v = try_run (device_param, hashconfig, kernel_accel_min, kernel_loops);
exec_ms = MIN (exec_ms, exec_ms_v);
}
if (exec_ms < target_ms) break;
}
}
// now the same for kernel-accel but with the new kernel-loops from previous loop set
#define STEPS_CNT 10
if (kernel_accel_min < kernel_accel_max)
{
for (int i = 0; i < STEPS_CNT; i++)
{
const u32 kernel_accel_try = 1u << i;
if (kernel_accel_try < kernel_accel_min) continue;
if (kernel_accel_try > kernel_accel_max) break;
double exec_ms = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_v = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops);
exec_ms = MIN (exec_ms, exec_ms_v);
}
if (exec_ms > target_ms) break;
kernel_accel = kernel_accel_try;
}
}
// at this point we want to know the actual runtime for the following reason:
// we need a reference for the balancing loop following up, and this
// the balancing loop can have an effect that the creates a new opportunity, for example:
// if the target is 95 ms and the current runtime is 48ms the above loop
// stopped the execution because the previous exec_ms was > 95ms
// due to the rebalance it's possible that the runtime reduces from 48ms to 47ms
// and this creates the possibility to double the workload -> 47 * 2 = 95ms, which is < 96ms
double exec_ms_pre_final = try_run (device_param, hashconfig, kernel_accel, kernel_loops);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_pre_final_v = try_run (device_param, hashconfig, kernel_accel, kernel_loops);
exec_ms_pre_final = MIN (exec_ms_pre_final, exec_ms_pre_final_v);
}
u32 diff = kernel_loops - kernel_accel;
if ((kernel_loops_min < kernel_loops_max) && (kernel_accel_min < kernel_accel_max))
{
u32 kernel_accel_orig = kernel_accel;
u32 kernel_loops_orig = kernel_loops;
for (u32 f = 1; f < 1024; f++)
{
const u32 kernel_accel_try = kernel_accel_orig * f;
const u32 kernel_loops_try = kernel_loops_orig / f;
if (kernel_accel_try > kernel_accel_max) break;
if (kernel_loops_try < kernel_loops_min) break;
u32 diff_new = kernel_loops_try - kernel_accel_try;
if (diff_new > diff) break;
diff_new = diff;
double exec_ms = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops_try);
for (int i = 0; i < VERIFIER_CNT; i++)
{
double exec_ms_v = try_run (device_param, hashconfig, kernel_accel_try, kernel_loops_try);
exec_ms = MIN (exec_ms, exec_ms_v);
}
if (exec_ms < exec_ms_pre_final)
{
exec_ms_pre_final = exec_ms;
kernel_accel = kernel_accel_try;
kernel_loops = kernel_loops_try;
}
}
}
const double exec_left = target_ms / exec_ms_pre_final;
const double accel_left = kernel_accel_max / kernel_accel;
const double exec_accel_min = MIN (exec_left, accel_left); // we want that to be int
if (exec_accel_min >= 1.0)
{
// this is safe to not overflow kernel_accel_max because of accel_left
kernel_accel *= (u32) exec_accel_min;
}
// reset them fake words
/*
memset (device_param->pws_buf, 0, kernel_power_max * sizeof (pw_t));
hc_clEnqueueWriteBuffer (data.ocl, device_param->command_queue, device_param->d_pws_buf, CL_TRUE, 0, kernel_power_max * sizeof (pw_t), device_param->pws_buf, 0, NULL, NULL);
hc_clEnqueueWriteBuffer (data.ocl, device_param->command_queue, device_param->d_pws_amp_buf, CL_TRUE, 0, kernel_power_max * sizeof (pw_t), device_param->pws_buf, 0, NULL, NULL);
*/
run_kernel_memset (device_param, device_param->d_pws_buf, 0, kernel_power_max * sizeof (pw_t));
if (hashconfig->attack_exec == ATTACK_EXEC_OUTSIDE_KERNEL)
{
run_kernel_memset (device_param, device_param->d_pws_amp_buf, 0, kernel_power_max * sizeof (pw_t));
}
// reset timer
device_param->exec_pos = 0;
memset (device_param->exec_ms, 0, EXEC_CACHE * sizeof (double));
memset (device_param->exec_us_prev1, 0, EXPECTED_ITERATIONS * sizeof (double));
memset (device_param->exec_us_prev2, 0, EXPECTED_ITERATIONS * sizeof (double));
memset (device_param->exec_us_prev3, 0, EXPECTED_ITERATIONS * sizeof (double));
// store
device_param->kernel_accel = kernel_accel;
device_param->kernel_loops = kernel_loops;
const u32 kernel_power = device_param->device_processors * device_param->kernel_threads * device_param->kernel_accel;
device_param->kernel_power = kernel_power;
#if defined (DEBUG)
if (data.quiet == 0)
{
clear_prompt ();
log_info ("- Device #%u: autotuned kernel-accel to %u\n"
"- Device #%u: autotuned kernel-loops to %u\n",
device_param->device_id + 1, kernel_accel,
device_param->device_id + 1, kernel_loops);
send_prompt ();
}
#endif
return 0;
}
@ -1094,7 +746,6 @@ static void *thread_outfile_remove (void *p)
return (p); return (p);
} }
static void set_kernel_power_final (const u64 kernel_power_final) static void set_kernel_power_final (const u64 kernel_power_final)
{ {
if (data.quiet == 0) if (data.quiet == 0)
@ -1744,6 +1395,13 @@ static void *thread_keypress (void *p)
return (p); return (p);
} }
int main (int argc, char **argv) int main (int argc, char **argv)
{ {
#if defined (_WIN) #if defined (_WIN)

View File

@ -30,6 +30,8 @@
extern hc_global_data_t data; extern hc_global_data_t data;
hc_thread_mutex_t mux_dispatcher;
#if defined (_WIN) #if defined (_WIN)
BOOL WINAPI sigHandler_default (DWORD sig) BOOL WINAPI sigHandler_default (DWORD sig)