mirror of https://github.com/GOSTSec/sgminer
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
1442 lines
44 KiB
1442 lines
44 KiB
/* |
|
* Copyright 2011-2012 Con Kolivas |
|
* |
|
* This program is free software; you can redistribute it and/or modify it |
|
* under the terms of the GNU General Public License as published by the Free |
|
* Software Foundation; either version 3 of the License, or (at your option) |
|
* any later version. See COPYING for more details. |
|
*/ |
|
|
|
#include "config.h" |
|
|
|
#if defined(HAVE_ADL) && (defined(__linux) || defined (WIN32)) |
|
|
|
#include <stdio.h> |
|
#include <string.h> |
|
#include <math.h> |
|
|
|
#ifdef HAVE_CURSES |
|
#include <curses.h> |
|
#endif |
|
|
|
#include "miner.h" |
|
#include "ADL_SDK/adl_sdk.h" |
|
#include "compat.h" |
|
|
|
#if defined (__linux) |
|
#include <dlfcn.h> |
|
#include <stdlib.h> |
|
#include <unistd.h> |
|
#else /* WIN32 */ |
|
#include <windows.h> |
|
#include <tchar.h> |
|
#endif |
|
#include "adl_functions.h" |
|
|
|
#ifndef HAVE_CURSES |
|
#define wlogprint(...) applog(LOG_WARNING, __VA_ARGS__) |
|
#endif |
|
|
|
bool adl_active; |
|
bool opt_reorder = false; |
|
|
|
int opt_hysteresis = 3; |
|
const int opt_targettemp = 75; |
|
const int opt_overheattemp = 85; |
|
static pthread_mutex_t adl_lock; |
|
|
|
struct gpu_adapters { |
|
int iAdapterIndex; |
|
int iBusNumber; |
|
int virtual_gpu; |
|
int id; |
|
}; |
|
|
|
// Memory allocation function |
|
static void * __stdcall ADL_Main_Memory_Alloc(int iSize) |
|
{ |
|
void *lpBuffer = malloc(iSize); |
|
|
|
return lpBuffer; |
|
} |
|
|
|
// Optional Memory de-allocation function |
|
static void __stdcall ADL_Main_Memory_Free (void **lpBuffer) |
|
{ |
|
if (*lpBuffer) { |
|
free (*lpBuffer); |
|
*lpBuffer = NULL; |
|
} |
|
} |
|
|
|
#if defined (LINUX) |
|
// equivalent functions in linux |
|
static void *GetProcAddress(void *pLibrary, const char *name) |
|
{ |
|
return dlsym( pLibrary, name); |
|
} |
|
#endif |
|
|
|
static ADL_MAIN_CONTROL_CREATE ADL_Main_Control_Create; |
|
static ADL_MAIN_CONTROL_DESTROY ADL_Main_Control_Destroy; |
|
static ADL_ADAPTER_NUMBEROFADAPTERS_GET ADL_Adapter_NumberOfAdapters_Get; |
|
static ADL_ADAPTER_ADAPTERINFO_GET ADL_Adapter_AdapterInfo_Get; |
|
static ADL_ADAPTER_ID_GET ADL_Adapter_ID_Get; |
|
static ADL_OVERDRIVE5_TEMPERATURE_GET ADL_Overdrive5_Temperature_Get; |
|
static ADL_OVERDRIVE5_CURRENTACTIVITY_GET ADL_Overdrive5_CurrentActivity_Get; |
|
static ADL_OVERDRIVE5_ODPARAMETERS_GET ADL_Overdrive5_ODParameters_Get; |
|
static ADL_OVERDRIVE5_FANSPEEDINFO_GET ADL_Overdrive5_FanSpeedInfo_Get; |
|
static ADL_OVERDRIVE5_FANSPEED_GET ADL_Overdrive5_FanSpeed_Get; |
|
static ADL_OVERDRIVE5_FANSPEED_SET ADL_Overdrive5_FanSpeed_Set; |
|
static ADL_OVERDRIVE5_ODPERFORMANCELEVELS_GET ADL_Overdrive5_ODPerformanceLevels_Get; |
|
static ADL_OVERDRIVE5_ODPERFORMANCELEVELS_SET ADL_Overdrive5_ODPerformanceLevels_Set; |
|
static ADL_MAIN_CONTROL_REFRESH ADL_Main_Control_Refresh; |
|
static ADL_OVERDRIVE5_POWERCONTROL_GET ADL_Overdrive5_PowerControl_Get; |
|
static ADL_OVERDRIVE5_POWERCONTROL_SET ADL_Overdrive5_PowerControl_Set; |
|
static ADL_OVERDRIVE5_FANSPEEDTODEFAULT_SET ADL_Overdrive5_FanSpeedToDefault_Set; |
|
|
|
#if defined (LINUX) |
|
static void *hDLL; // Handle to .so library |
|
#else |
|
HINSTANCE hDLL; // Handle to DLL |
|
#endif |
|
static int iNumberAdapters; |
|
static LPAdapterInfo lpInfo = NULL; |
|
|
|
int set_fanspeed(int gpu, int iFanSpeed); |
|
static float __gpu_temp(struct gpu_adl *ga); |
|
|
|
static inline void lock_adl(void) |
|
{ |
|
mutex_lock(&adl_lock); |
|
} |
|
|
|
static inline void unlock_adl(void) |
|
{ |
|
mutex_unlock(&adl_lock); |
|
} |
|
|
|
/* This looks for the twin GPU that has the fanspeed control of a non fanspeed |
|
* control GPU on dual GPU cards */ |
|
static bool fanspeed_twin(struct gpu_adl *ga, struct gpu_adl *other_ga) |
|
{ |
|
if (!other_ga->has_fanspeed) |
|
return false; |
|
if (abs(ga->iBusNumber - other_ga->iBusNumber) != 1) |
|
return false; |
|
if (strcmp(ga->strAdapterName, other_ga->strAdapterName)) |
|
return false; |
|
return true; |
|
} |
|
|
|
static bool prepare_adl(void) |
|
{ |
|
int result; |
|
|
|
#if defined (LINUX) |
|
hDLL = dlopen( "libatiadlxx.so", RTLD_LAZY|RTLD_GLOBAL); |
|
#else |
|
hDLL = LoadLibrary("atiadlxx.dll"); |
|
if (hDLL == NULL) |
|
// A 32 bit calling application on 64 bit OS will fail to LoadLIbrary. |
|
// Try to load the 32 bit library (atiadlxy.dll) instead |
|
hDLL = LoadLibrary("atiadlxy.dll"); |
|
#endif |
|
if (hDLL == NULL) { |
|
applog(LOG_INFO, "Unable to load ati adl library"); |
|
return false; |
|
} |
|
ADL_Main_Control_Create = (ADL_MAIN_CONTROL_CREATE) GetProcAddress(hDLL,"ADL_Main_Control_Create"); |
|
ADL_Main_Control_Destroy = (ADL_MAIN_CONTROL_DESTROY) GetProcAddress(hDLL,"ADL_Main_Control_Destroy"); |
|
ADL_Adapter_NumberOfAdapters_Get = (ADL_ADAPTER_NUMBEROFADAPTERS_GET) GetProcAddress(hDLL,"ADL_Adapter_NumberOfAdapters_Get"); |
|
ADL_Adapter_AdapterInfo_Get = (ADL_ADAPTER_ADAPTERINFO_GET) GetProcAddress(hDLL,"ADL_Adapter_AdapterInfo_Get"); |
|
ADL_Adapter_ID_Get = (ADL_ADAPTER_ID_GET) GetProcAddress(hDLL,"ADL_Adapter_ID_Get"); |
|
ADL_Overdrive5_Temperature_Get = (ADL_OVERDRIVE5_TEMPERATURE_GET) GetProcAddress(hDLL,"ADL_Overdrive5_Temperature_Get"); |
|
ADL_Overdrive5_CurrentActivity_Get = (ADL_OVERDRIVE5_CURRENTACTIVITY_GET) GetProcAddress(hDLL, "ADL_Overdrive5_CurrentActivity_Get"); |
|
ADL_Overdrive5_ODParameters_Get = (ADL_OVERDRIVE5_ODPARAMETERS_GET) GetProcAddress(hDLL, "ADL_Overdrive5_ODParameters_Get"); |
|
ADL_Overdrive5_FanSpeedInfo_Get = (ADL_OVERDRIVE5_FANSPEEDINFO_GET) GetProcAddress(hDLL, "ADL_Overdrive5_FanSpeedInfo_Get"); |
|
ADL_Overdrive5_FanSpeed_Get = (ADL_OVERDRIVE5_FANSPEED_GET) GetProcAddress(hDLL, "ADL_Overdrive5_FanSpeed_Get"); |
|
ADL_Overdrive5_FanSpeed_Set = (ADL_OVERDRIVE5_FANSPEED_SET) GetProcAddress(hDLL, "ADL_Overdrive5_FanSpeed_Set"); |
|
ADL_Overdrive5_ODPerformanceLevels_Get = (ADL_OVERDRIVE5_ODPERFORMANCELEVELS_GET) GetProcAddress(hDLL, "ADL_Overdrive5_ODPerformanceLevels_Get"); |
|
ADL_Overdrive5_ODPerformanceLevels_Set = (ADL_OVERDRIVE5_ODPERFORMANCELEVELS_SET) GetProcAddress(hDLL, "ADL_Overdrive5_ODPerformanceLevels_Set"); |
|
ADL_Main_Control_Refresh = (ADL_MAIN_CONTROL_REFRESH) GetProcAddress(hDLL, "ADL_Main_Control_Refresh"); |
|
ADL_Overdrive5_PowerControl_Get = (ADL_OVERDRIVE5_POWERCONTROL_GET) GetProcAddress(hDLL, "ADL_Overdrive5_PowerControl_Get"); |
|
ADL_Overdrive5_PowerControl_Set = (ADL_OVERDRIVE5_POWERCONTROL_SET) GetProcAddress(hDLL, "ADL_Overdrive5_PowerControl_Set"); |
|
ADL_Overdrive5_FanSpeedToDefault_Set = (ADL_OVERDRIVE5_FANSPEEDTODEFAULT_SET) GetProcAddress(hDLL, "ADL_Overdrive5_FanSpeedToDefault_Set"); |
|
|
|
if (!ADL_Main_Control_Create || !ADL_Main_Control_Destroy || |
|
!ADL_Adapter_NumberOfAdapters_Get || !ADL_Adapter_AdapterInfo_Get || |
|
!ADL_Adapter_ID_Get || !ADL_Overdrive5_Temperature_Get || |
|
!ADL_Overdrive5_CurrentActivity_Get || |
|
!ADL_Overdrive5_ODParameters_Get || !ADL_Overdrive5_FanSpeedInfo_Get || |
|
!ADL_Overdrive5_FanSpeed_Get || !ADL_Overdrive5_FanSpeed_Set || |
|
!ADL_Overdrive5_ODPerformanceLevels_Get || !ADL_Overdrive5_ODPerformanceLevels_Set || |
|
!ADL_Main_Control_Refresh || !ADL_Overdrive5_PowerControl_Get || |
|
!ADL_Overdrive5_PowerControl_Set || !ADL_Overdrive5_FanSpeedToDefault_Set) { |
|
applog(LOG_WARNING, "ATI ADL's API is missing"); |
|
return false; |
|
} |
|
|
|
// Initialise ADL. The second parameter is 1, which means: |
|
// retrieve adapter information only for adapters that are physically present and enabled in the system |
|
result = ADL_Main_Control_Create (ADL_Main_Memory_Alloc, 1); |
|
if (result != ADL_OK) { |
|
applog(LOG_INFO, "ADL Initialisation Error! Error %d!", result); |
|
return false; |
|
} |
|
|
|
result = ADL_Main_Control_Refresh(); |
|
if (result != ADL_OK) { |
|
applog(LOG_INFO, "ADL Refresh Error! Error %d!", result); |
|
return false; |
|
} |
|
|
|
return true; |
|
} |
|
|
|
void init_adl(int nDevs) |
|
{ |
|
int result, i, j, devices = 0, last_adapter = -1, gpu = 0, dummy = 0; |
|
struct gpu_adapters adapters[MAX_GPUDEVICES], vadapters[MAX_GPUDEVICES]; |
|
bool devs_match = true; |
|
|
|
if (unlikely(pthread_mutex_init(&adl_lock, NULL))) { |
|
applog(LOG_ERR, "Failed to init adl_lock in init_adl"); |
|
return; |
|
} |
|
|
|
if (!prepare_adl()) |
|
return; |
|
|
|
// Obtain the number of adapters for the system |
|
result = ADL_Adapter_NumberOfAdapters_Get (&iNumberAdapters); |
|
if (result != ADL_OK) { |
|
applog(LOG_INFO, "Cannot get the number of adapters! Error %d!", result); |
|
return ; |
|
} |
|
|
|
if (iNumberAdapters > 0) { |
|
lpInfo = malloc ( sizeof (AdapterInfo) * iNumberAdapters ); |
|
memset ( lpInfo,'\0', sizeof (AdapterInfo) * iNumberAdapters ); |
|
|
|
lpInfo->iSize = sizeof(lpInfo); |
|
// Get the AdapterInfo structure for all adapters in the system |
|
result = ADL_Adapter_AdapterInfo_Get (lpInfo, sizeof (AdapterInfo) * iNumberAdapters); |
|
if (result != ADL_OK) { |
|
applog(LOG_INFO, "ADL_Adapter_AdapterInfo_Get Error! Error %d", result); |
|
return ; |
|
} |
|
} else { |
|
applog(LOG_INFO, "No adapters found"); |
|
return; |
|
} |
|
|
|
/* Iterate over iNumberAdapters and find the lpAdapterID of real devices */ |
|
for (i = 0; i < iNumberAdapters; i++) { |
|
int iAdapterIndex; |
|
int lpAdapterID; |
|
|
|
iAdapterIndex = lpInfo[i].iAdapterIndex; |
|
/* Get unique identifier of the adapter, 0 means not AMD */ |
|
result = ADL_Adapter_ID_Get(iAdapterIndex, &lpAdapterID); |
|
if (result != ADL_OK) { |
|
applog(LOG_INFO, "Failed to ADL_Adapter_ID_Get. Error %d", result); |
|
if (result == -10) |
|
applog(LOG_INFO, "This error says the device is not enabled"); |
|
continue; |
|
} |
|
|
|
/* Each adapter may have multiple entries */ |
|
if (lpAdapterID == last_adapter) |
|
continue; |
|
|
|
applog(LOG_DEBUG, "GPU %d " |
|
"iAdapterIndex %d " |
|
"strUDID %s " |
|
"iBusNumber %d " |
|
"iDeviceNumber %d " |
|
"iFunctionNumber %d " |
|
"iVendorID %d " |
|
"strAdapterName %s ", |
|
devices, |
|
iAdapterIndex, |
|
lpInfo[i].strUDID, |
|
lpInfo[i].iBusNumber, |
|
lpInfo[i].iDeviceNumber, |
|
lpInfo[i].iFunctionNumber, |
|
lpInfo[i].iVendorID, |
|
lpInfo[i].strAdapterName); |
|
|
|
adapters[devices].iAdapterIndex = iAdapterIndex; |
|
adapters[devices].iBusNumber = lpInfo[i].iBusNumber; |
|
adapters[devices].id = i; |
|
|
|
/* We found a truly new adapter instead of a logical |
|
* one. Now since there's no way of correlating the |
|
* opencl enumerated devices and the ADL enumerated |
|
* ones, we have to assume they're in the same order.*/ |
|
if (++devices > nDevs && devs_match) { |
|
applog(LOG_ERR, "ADL found more devices than opencl!"); |
|
applog(LOG_ERR, "There is possibly at least one GPU that doesn't support OpenCL"); |
|
applog(LOG_ERR, "Use the gpu map feature to reliably map OpenCL to ADL"); |
|
devs_match = false; |
|
} |
|
last_adapter = lpAdapterID; |
|
|
|
if (!lpAdapterID) { |
|
applog(LOG_INFO, "Adapter returns ID 0 meaning not AMD. Card order might be confused"); |
|
continue; |
|
} |
|
} |
|
|
|
if (devices < nDevs) { |
|
applog(LOG_ERR, "ADL found less devices than opencl!"); |
|
applog(LOG_ERR, "There is possibly more than one display attached to a GPU"); |
|
applog(LOG_ERR, "Use the gpu map feature to reliably map OpenCL to ADL"); |
|
devs_match = false; |
|
} |
|
|
|
for (i = 0; i < devices; i++) { |
|
vadapters[i].virtual_gpu = i; |
|
vadapters[i].id = adapters[i].id; |
|
} |
|
|
|
/* Apply manually provided OpenCL to ADL mapping, if any */ |
|
for (i = 0; i < nDevs; i++) { |
|
if (gpus[i].mapped) { |
|
vadapters[gpus[i].virtual_adl].virtual_gpu = i; |
|
applog(LOG_INFO, "Mapping OpenCL device %d to ADL device %d", i, gpus[i].virtual_adl); |
|
} else |
|
gpus[i].virtual_adl = i; |
|
} |
|
|
|
if (!devs_match) { |
|
applog(LOG_ERR, "WARNING: Number of OpenCL and ADL devices did not match!"); |
|
applog(LOG_ERR, "Hardware monitoring may NOT match up with devices!"); |
|
} else if (opt_reorder) { |
|
/* Windows has some kind of random ordering for bus number IDs and |
|
* ordering the GPUs according to ascending order fixes it. Linux |
|
* has usually sequential but decreasing order instead! */ |
|
for (i = 0; i < devices; i++) { |
|
int j, virtual_gpu; |
|
|
|
virtual_gpu = 0; |
|
for (j = 0; j < devices; j++) { |
|
if (i == j) |
|
continue; |
|
#ifdef WIN32 |
|
if (adapters[j].iBusNumber < adapters[i].iBusNumber) |
|
#else |
|
if (adapters[j].iBusNumber > adapters[i].iBusNumber) |
|
#endif |
|
virtual_gpu++; |
|
} |
|
if (virtual_gpu != i) { |
|
applog(LOG_INFO, "Mapping device %d to GPU %d according to Bus Number order", |
|
i, virtual_gpu); |
|
vadapters[virtual_gpu].virtual_gpu = i; |
|
vadapters[virtual_gpu].id = adapters[i].id; |
|
} |
|
} |
|
} |
|
|
|
if (devices > nDevs) |
|
devices = nDevs; |
|
|
|
for (gpu = 0; gpu < devices; gpu++) { |
|
struct gpu_adl *ga; |
|
int iAdapterIndex; |
|
int lpAdapterID; |
|
ADLODPerformanceLevels *lpOdPerformanceLevels; |
|
int lev, adlGpu; |
|
|
|
adlGpu = gpus[gpu].virtual_adl; |
|
i = vadapters[adlGpu].id; |
|
iAdapterIndex = lpInfo[i].iAdapterIndex; |
|
gpus[gpu].virtual_gpu = vadapters[adlGpu].virtual_gpu; |
|
|
|
/* Get unique identifier of the adapter, 0 means not AMD */ |
|
result = ADL_Adapter_ID_Get(iAdapterIndex, &lpAdapterID); |
|
if (result != ADL_OK) { |
|
applog(LOG_INFO, "Failed to ADL_Adapter_ID_Get. Error %d", result); |
|
continue; |
|
} |
|
|
|
if (gpus[gpu].deven == DEV_DISABLED) { |
|
gpus[gpu].gpu_engine = |
|
gpus[gpu].gpu_memclock = |
|
gpus[gpu].gpu_vddc = |
|
gpus[gpu].gpu_fan = |
|
gpus[gpu].gpu_powertune = 0; |
|
continue; |
|
} |
|
|
|
applog(LOG_INFO, "GPU %d %s hardware monitoring enabled", gpu, lpInfo[i].strAdapterName); |
|
if (gpus[gpu].name) |
|
free(gpus[gpu].name); |
|
gpus[gpu].name = lpInfo[i].strAdapterName; |
|
gpus[gpu].has_adl = true; |
|
/* Flag adl as active if any card is successfully activated */ |
|
adl_active = true; |
|
|
|
/* From here on we know this device is a discrete device and |
|
* should support ADL */ |
|
ga = &gpus[gpu].adl; |
|
ga->gpu = gpu; |
|
ga->iAdapterIndex = iAdapterIndex; |
|
ga->lpAdapterID = lpAdapterID; |
|
strcpy(ga->strAdapterName, lpInfo[i].strAdapterName); |
|
ga->DefPerfLev = NULL; |
|
ga->twin = NULL; |
|
|
|
ga->lpOdParameters.iSize = sizeof(ADLODParameters); |
|
if (ADL_Overdrive5_ODParameters_Get(iAdapterIndex, &ga->lpOdParameters) != ADL_OK) |
|
applog(LOG_INFO, "Failed to ADL_Overdrive5_ODParameters_Get"); |
|
|
|
lev = ga->lpOdParameters.iNumberOfPerformanceLevels - 1; |
|
/* We're only interested in the top performance level */ |
|
lpOdPerformanceLevels = malloc(sizeof(ADLODPerformanceLevels) + (lev * sizeof(ADLODPerformanceLevel))); |
|
lpOdPerformanceLevels->iSize = sizeof(ADLODPerformanceLevels) + sizeof(ADLODPerformanceLevel) * lev; |
|
|
|
/* Get default performance levels first */ |
|
if (ADL_Overdrive5_ODPerformanceLevels_Get(iAdapterIndex, 1, lpOdPerformanceLevels) != ADL_OK) |
|
applog(LOG_INFO, "Failed to ADL_Overdrive5_ODPerformanceLevels_Get"); |
|
/* Set the limits we'd use based on default gpu speeds */ |
|
ga->maxspeed = ga->minspeed = lpOdPerformanceLevels->aLevels[lev].iEngineClock; |
|
|
|
ga->lpTemperature.iSize = sizeof(ADLTemperature); |
|
ga->lpFanSpeedInfo.iSize = sizeof(ADLFanSpeedInfo); |
|
ga->lpFanSpeedValue.iSize = ga->DefFanSpeedValue.iSize = sizeof(ADLFanSpeedValue); |
|
/* Now get the current performance levels for any existing overclock */ |
|
ADL_Overdrive5_ODPerformanceLevels_Get(iAdapterIndex, 0, lpOdPerformanceLevels); |
|
/* Save these values as the defaults in case we wish to reset to defaults */ |
|
ga->DefPerfLev = lpOdPerformanceLevels; |
|
|
|
if (gpus[gpu].gpu_engine) { |
|
int setengine = gpus[gpu].gpu_engine * 100; |
|
|
|
/* Lower profiles can't have a higher setting */ |
|
for (j = 0; j < lev; j++) { |
|
if (lpOdPerformanceLevels->aLevels[j].iEngineClock > setengine) |
|
lpOdPerformanceLevels->aLevels[j].iEngineClock = setengine; |
|
} |
|
lpOdPerformanceLevels->aLevels[lev].iEngineClock = setengine; |
|
applog(LOG_INFO, "Setting GPU %d engine clock to %d", gpu, gpus[gpu].gpu_engine); |
|
ADL_Overdrive5_ODPerformanceLevels_Set(iAdapterIndex, lpOdPerformanceLevels); |
|
ga->maxspeed = setengine; |
|
if (gpus[gpu].min_engine) |
|
ga->minspeed = gpus[gpu].min_engine * 100; |
|
ga->managed = true; |
|
if (gpus[gpu].gpu_memdiff) |
|
set_memoryclock(gpu, gpus[gpu].gpu_engine + gpus[gpu].gpu_memdiff); |
|
} |
|
|
|
if (gpus[gpu].gpu_memclock) { |
|
int setmem = gpus[gpu].gpu_memclock * 100; |
|
|
|
for (j = 0; j < lev; j++) { |
|
if (lpOdPerformanceLevels->aLevels[j].iMemoryClock > setmem) |
|
lpOdPerformanceLevels->aLevels[j].iMemoryClock = setmem; |
|
} |
|
lpOdPerformanceLevels->aLevels[lev].iMemoryClock = setmem; |
|
applog(LOG_INFO, "Setting GPU %d memory clock to %d", gpu, gpus[gpu].gpu_memclock); |
|
ADL_Overdrive5_ODPerformanceLevels_Set(iAdapterIndex, lpOdPerformanceLevels); |
|
ga->managed = true; |
|
} |
|
|
|
if (gpus[gpu].gpu_vddc) { |
|
int setv = gpus[gpu].gpu_vddc * 1000; |
|
|
|
for (j = 0; j < lev; j++) { |
|
if (lpOdPerformanceLevels->aLevels[j].iVddc > setv) |
|
lpOdPerformanceLevels->aLevels[j].iVddc = setv; |
|
} |
|
lpOdPerformanceLevels->aLevels[lev].iVddc = setv; |
|
applog(LOG_INFO, "Setting GPU %d voltage to %.3f", gpu, gpus[gpu].gpu_vddc); |
|
ADL_Overdrive5_ODPerformanceLevels_Set(iAdapterIndex, lpOdPerformanceLevels); |
|
ga->managed = true; |
|
} |
|
|
|
ADL_Overdrive5_ODPerformanceLevels_Get(iAdapterIndex, 0, lpOdPerformanceLevels); |
|
ga->iEngineClock = lpOdPerformanceLevels->aLevels[lev].iEngineClock; |
|
ga->iMemoryClock = lpOdPerformanceLevels->aLevels[lev].iMemoryClock; |
|
ga->iVddc = lpOdPerformanceLevels->aLevels[lev].iVddc; |
|
ga->iBusNumber = lpInfo[i].iBusNumber; |
|
|
|
if (ADL_Overdrive5_FanSpeedInfo_Get(iAdapterIndex, 0, &ga->lpFanSpeedInfo) != ADL_OK) |
|
applog(LOG_INFO, "Failed to ADL_Overdrive5_FanSpeedInfo_Get"); |
|
else |
|
ga->has_fanspeed = true; |
|
|
|
/* Save the fanspeed values as defaults in case we reset later */ |
|
ADL_Overdrive5_FanSpeed_Get(ga->iAdapterIndex, 0, &ga->DefFanSpeedValue); |
|
if (gpus[gpu].gpu_fan) |
|
set_fanspeed(gpu, gpus[gpu].gpu_fan); |
|
else |
|
gpus[gpu].gpu_fan = 85; /* Set a nominal upper limit of 85% */ |
|
|
|
/* Not fatal if powercontrol get fails */ |
|
if (ADL_Overdrive5_PowerControl_Get(ga->iAdapterIndex, &ga->iPercentage, &dummy) != ADL_OK) |
|
applog(LOG_INFO, "Failed to ADL_Overdrive5_PowerControl_get"); |
|
|
|
if (gpus[gpu].gpu_powertune) { |
|
ADL_Overdrive5_PowerControl_Set(ga->iAdapterIndex, gpus[gpu].gpu_powertune); |
|
ADL_Overdrive5_PowerControl_Get(ga->iAdapterIndex, &ga->iPercentage, &dummy); |
|
ga->managed = true; |
|
} |
|
|
|
/* Set some default temperatures for autotune when enabled */ |
|
if (!ga->targettemp) |
|
ga->targettemp = opt_targettemp; |
|
if (!ga->overtemp) |
|
ga->overtemp = opt_overheattemp; |
|
if (!gpus[gpu].cutofftemp) |
|
gpus[gpu].cutofftemp = opt_cutofftemp; |
|
if (opt_autofan) { |
|
ga->autofan = true; |
|
/* Set a safe starting default if we're automanaging fan speeds */ |
|
set_fanspeed(gpu, 50); |
|
} |
|
if (opt_autoengine) { |
|
ga->autoengine = true; |
|
ga->managed = true; |
|
} |
|
ga->lasttemp = __gpu_temp(ga); |
|
} |
|
|
|
for (gpu = 0; gpu < devices; gpu++) { |
|
struct gpu_adl *ga = &gpus[gpu].adl; |
|
int j; |
|
|
|
for (j = 0; j < devices; j++) { |
|
struct gpu_adl *other_ga; |
|
|
|
if (j == gpu) |
|
continue; |
|
|
|
other_ga = &gpus[j].adl; |
|
|
|
/* Search for twin GPUs on a single card. They will be |
|
* separated by one bus id and one will have fanspeed |
|
* while the other won't. */ |
|
if (!ga->has_fanspeed) { |
|
if (fanspeed_twin(ga, other_ga)) { |
|
applog(LOG_INFO, "Dual GPUs detected: %d and %d", |
|
ga->gpu, other_ga->gpu); |
|
ga->twin = other_ga; |
|
other_ga->twin = ga; |
|
} |
|
} |
|
} |
|
} |
|
} |
|
|
|
static float __gpu_temp(struct gpu_adl *ga) |
|
{ |
|
if (ADL_Overdrive5_Temperature_Get(ga->iAdapterIndex, 0, &ga->lpTemperature) != ADL_OK) |
|
return -1; |
|
return (float)ga->lpTemperature.iTemperature / 1000; |
|
} |
|
|
|
float gpu_temp(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
float ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ret = __gpu_temp(ga); |
|
unlock_adl(); |
|
gpus[gpu].temp = ret; |
|
return ret; |
|
} |
|
|
|
static inline int __gpu_engineclock(struct gpu_adl *ga) |
|
{ |
|
return ga->lpActivity.iEngineClock / 100; |
|
} |
|
|
|
int gpu_engineclock(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
if (ADL_Overdrive5_CurrentActivity_Get(ga->iAdapterIndex, &ga->lpActivity) != ADL_OK) |
|
goto out; |
|
ret = __gpu_engineclock(ga); |
|
out: |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
static inline int __gpu_memclock(struct gpu_adl *ga) |
|
{ |
|
return ga->lpActivity.iMemoryClock / 100; |
|
} |
|
|
|
int gpu_memclock(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
if (ADL_Overdrive5_CurrentActivity_Get(ga->iAdapterIndex, &ga->lpActivity) != ADL_OK) |
|
goto out; |
|
ret = __gpu_memclock(ga); |
|
out: |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
static inline float __gpu_vddc(struct gpu_adl *ga) |
|
{ |
|
return (float)ga->lpActivity.iVddc / 1000; |
|
} |
|
|
|
float gpu_vddc(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
float ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
if (ADL_Overdrive5_CurrentActivity_Get(ga->iAdapterIndex, &ga->lpActivity) != ADL_OK) |
|
goto out; |
|
ret = __gpu_vddc(ga); |
|
out: |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
static inline int __gpu_activity(struct gpu_adl *ga) |
|
{ |
|
if (!ga->lpOdParameters.iActivityReportingSupported) |
|
return -1; |
|
return ga->lpActivity.iActivityPercent; |
|
} |
|
|
|
int gpu_activity(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ret = ADL_Overdrive5_CurrentActivity_Get(ga->iAdapterIndex, &ga->lpActivity); |
|
unlock_adl(); |
|
if (ret != ADL_OK) |
|
return ret; |
|
if (!ga->lpOdParameters.iActivityReportingSupported) |
|
return ret; |
|
return ga->lpActivity.iActivityPercent; |
|
} |
|
|
|
static inline int __gpu_fanspeed(struct gpu_adl *ga) |
|
{ |
|
if (!ga->has_fanspeed && ga->twin) |
|
return __gpu_fanspeed(ga->twin); |
|
|
|
if (!(ga->lpFanSpeedInfo.iFlags & ADL_DL_FANCTRL_SUPPORTS_RPM_READ)) |
|
return -1; |
|
ga->lpFanSpeedValue.iSpeedType = ADL_DL_FANCTRL_SPEED_TYPE_RPM; |
|
if (ADL_Overdrive5_FanSpeed_Get(ga->iAdapterIndex, 0, &ga->lpFanSpeedValue) != ADL_OK) |
|
return -1; |
|
return ga->lpFanSpeedValue.iFanSpeed; |
|
} |
|
|
|
int gpu_fanspeed(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ret = __gpu_fanspeed(ga); |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
static int __gpu_fanpercent(struct gpu_adl *ga) |
|
{ |
|
if (!ga->has_fanspeed && ga->twin) |
|
return __gpu_fanpercent(ga->twin); |
|
|
|
if (!(ga->lpFanSpeedInfo.iFlags & ADL_DL_FANCTRL_SUPPORTS_PERCENT_READ )) |
|
return -1; |
|
ga->lpFanSpeedValue.iSpeedType = ADL_DL_FANCTRL_SPEED_TYPE_PERCENT; |
|
if (ADL_Overdrive5_FanSpeed_Get(ga->iAdapterIndex, 0, &ga->lpFanSpeedValue) != ADL_OK) |
|
return -1; |
|
return ga->lpFanSpeedValue.iFanSpeed; |
|
} |
|
|
|
int gpu_fanpercent(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ret = __gpu_fanpercent(ga); |
|
unlock_adl(); |
|
if (unlikely(ga->has_fanspeed && ret == -1)) { |
|
#if 0 |
|
/* Recursive calling applog causes a hang, so disable messages */ |
|
applog(LOG_WARNING, "GPU %d stopped reporting fanspeed due to driver corruption", gpu); |
|
if (opt_restart) { |
|
applog(LOG_WARNING, "Restart enabled, will attempt to restart cgminer"); |
|
applog(LOG_WARNING, "You can disable this with the --no-restart option"); |
|
app_restart(); |
|
} |
|
applog(LOG_WARNING, "Disabling fanspeed monitoring on this device"); |
|
ga->has_fanspeed = false; |
|
if (ga->twin) { |
|
applog(LOG_WARNING, "Disabling fanspeed linking on GPU twins"); |
|
ga->twin->twin = NULL;; |
|
ga->twin = NULL; |
|
} |
|
#endif |
|
if (opt_restart) |
|
app_restart(); |
|
ga->has_fanspeed = false; |
|
if (ga->twin) { |
|
ga->twin->twin = NULL;; |
|
ga->twin = NULL; |
|
} |
|
} |
|
return ret; |
|
} |
|
|
|
static inline int __gpu_powertune(struct gpu_adl *ga) |
|
{ |
|
int dummy = 0; |
|
|
|
if (ADL_Overdrive5_PowerControl_Get(ga->iAdapterIndex, &ga->iPercentage, &dummy) != ADL_OK) |
|
return -1; |
|
return ga->iPercentage; |
|
} |
|
|
|
int gpu_powertune(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = -1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return ret; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ret = __gpu_powertune(ga); |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
bool gpu_stats(int gpu, float *temp, int *engineclock, int *memclock, float *vddc, |
|
int *activity, int *fanspeed, int *fanpercent, int *powertune) |
|
{ |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return false; |
|
|
|
ga = &gpus[gpu].adl; |
|
|
|
lock_adl(); |
|
*temp = __gpu_temp(ga); |
|
if (ADL_Overdrive5_CurrentActivity_Get(ga->iAdapterIndex, &ga->lpActivity) != ADL_OK) { |
|
*engineclock = 0; |
|
*memclock = 0; |
|
*vddc = 0; |
|
*activity = 0; |
|
} else { |
|
*engineclock = __gpu_engineclock(ga); |
|
*memclock = __gpu_memclock(ga); |
|
*vddc = __gpu_vddc(ga); |
|
*activity = __gpu_activity(ga); |
|
} |
|
*fanspeed = __gpu_fanspeed(ga); |
|
*fanpercent = __gpu_fanpercent(ga); |
|
*powertune = __gpu_powertune(ga); |
|
unlock_adl(); |
|
|
|
return true; |
|
} |
|
|
|
#ifdef HAVE_CURSES |
|
static void get_enginerange(int gpu, int *imin, int *imax) |
|
{ |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Get enginerange not supported\n"); |
|
return; |
|
} |
|
ga = &gpus[gpu].adl; |
|
*imin = ga->lpOdParameters.sEngineClock.iMin / 100; |
|
*imax = ga->lpOdParameters.sEngineClock.iMax / 100; |
|
} |
|
#endif |
|
|
|
int set_engineclock(int gpu, int iEngineClock) |
|
{ |
|
ADLODPerformanceLevels *lpOdPerformanceLevels; |
|
struct cgpu_info *cgpu; |
|
int i, lev, ret = 1; |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Set engineclock not supported\n"); |
|
return ret; |
|
} |
|
|
|
iEngineClock *= 100; |
|
ga = &gpus[gpu].adl; |
|
|
|
/* Keep track of intended engine clock in case the device changes |
|
* profile and drops while idle, not taking the new engine clock */ |
|
ga->lastengine = iEngineClock; |
|
|
|
lev = ga->lpOdParameters.iNumberOfPerformanceLevels - 1; |
|
lpOdPerformanceLevels = alloca(sizeof(ADLODPerformanceLevels) + (lev * sizeof(ADLODPerformanceLevel))); |
|
lpOdPerformanceLevels->iSize = sizeof(ADLODPerformanceLevels) + sizeof(ADLODPerformanceLevel) * lev; |
|
|
|
lock_adl(); |
|
if (ADL_Overdrive5_ODPerformanceLevels_Get(ga->iAdapterIndex, 0, lpOdPerformanceLevels) != ADL_OK) |
|
goto out; |
|
for (i = 0; i < lev; i++) { |
|
if (lpOdPerformanceLevels->aLevels[i].iEngineClock > iEngineClock) |
|
lpOdPerformanceLevels->aLevels[i].iEngineClock = iEngineClock; |
|
} |
|
lpOdPerformanceLevels->aLevels[lev].iEngineClock = iEngineClock; |
|
ADL_Overdrive5_ODPerformanceLevels_Set(ga->iAdapterIndex, lpOdPerformanceLevels); |
|
ADL_Overdrive5_ODPerformanceLevels_Get(ga->iAdapterIndex, 0, lpOdPerformanceLevels); |
|
if (lpOdPerformanceLevels->aLevels[lev].iEngineClock == iEngineClock) |
|
ret = 0; |
|
ga->iEngineClock = lpOdPerformanceLevels->aLevels[lev].iEngineClock; |
|
if (ga->iEngineClock > ga->maxspeed) |
|
ga->maxspeed = ga->iEngineClock; |
|
if (ga->iEngineClock < ga->minspeed) |
|
ga->minspeed = ga->iEngineClock; |
|
ga->iMemoryClock = lpOdPerformanceLevels->aLevels[lev].iMemoryClock; |
|
ga->iVddc = lpOdPerformanceLevels->aLevels[lev].iVddc; |
|
ga->managed = true; |
|
out: |
|
unlock_adl(); |
|
|
|
cgpu = &gpus[gpu]; |
|
if (cgpu->gpu_memdiff) |
|
set_memoryclock(gpu, iEngineClock / 100 + cgpu->gpu_memdiff); |
|
|
|
return ret; |
|
} |
|
|
|
#ifdef HAVE_CURSES |
|
static void get_memoryrange(int gpu, int *imin, int *imax) |
|
{ |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Get memoryrange not supported\n"); |
|
return; |
|
} |
|
ga = &gpus[gpu].adl; |
|
*imin = ga->lpOdParameters.sMemoryClock.iMin / 100; |
|
*imax = ga->lpOdParameters.sMemoryClock.iMax / 100; |
|
} |
|
#endif |
|
|
|
int set_memoryclock(int gpu, int iMemoryClock) |
|
{ |
|
ADLODPerformanceLevels *lpOdPerformanceLevels; |
|
int i, lev, ret = 1; |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Set memoryclock not supported\n"); |
|
return ret; |
|
} |
|
|
|
iMemoryClock *= 100; |
|
ga = &gpus[gpu].adl; |
|
|
|
lev = ga->lpOdParameters.iNumberOfPerformanceLevels - 1; |
|
lpOdPerformanceLevels = alloca(sizeof(ADLODPerformanceLevels) + (lev * sizeof(ADLODPerformanceLevel))); |
|
lpOdPerformanceLevels->iSize = sizeof(ADLODPerformanceLevels) + sizeof(ADLODPerformanceLevel) * lev; |
|
|
|
lock_adl(); |
|
if (ADL_Overdrive5_ODPerformanceLevels_Get(ga->iAdapterIndex, 0, lpOdPerformanceLevels) != ADL_OK) |
|
goto out; |
|
lpOdPerformanceLevels->aLevels[lev].iMemoryClock = iMemoryClock; |
|
for (i = 0; i < lev; i++) { |
|
if (lpOdPerformanceLevels->aLevels[i].iMemoryClock > iMemoryClock) |
|
lpOdPerformanceLevels->aLevels[i].iMemoryClock = iMemoryClock; |
|
} |
|
ADL_Overdrive5_ODPerformanceLevels_Set(ga->iAdapterIndex, lpOdPerformanceLevels); |
|
ADL_Overdrive5_ODPerformanceLevels_Get(ga->iAdapterIndex, 0, lpOdPerformanceLevels); |
|
if (lpOdPerformanceLevels->aLevels[lev].iMemoryClock == iMemoryClock) |
|
ret = 0; |
|
ga->iEngineClock = lpOdPerformanceLevels->aLevels[lev].iEngineClock; |
|
ga->iMemoryClock = lpOdPerformanceLevels->aLevels[lev].iMemoryClock; |
|
ga->iVddc = lpOdPerformanceLevels->aLevels[lev].iVddc; |
|
ga->managed = true; |
|
out: |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
#ifdef HAVE_CURSES |
|
static void get_vddcrange(int gpu, float *imin, float *imax) |
|
{ |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Get vddcrange not supported\n"); |
|
return; |
|
} |
|
ga = &gpus[gpu].adl; |
|
*imin = (float)ga->lpOdParameters.sVddc.iMin / 1000; |
|
*imax = (float)ga->lpOdParameters.sVddc.iMax / 1000; |
|
} |
|
|
|
static float curses_float(const char *query) |
|
{ |
|
float ret; |
|
char *cvar; |
|
|
|
cvar = curses_input(query); |
|
ret = atof(cvar); |
|
free(cvar); |
|
return ret; |
|
} |
|
#endif |
|
|
|
int set_vddc(int gpu, float fVddc) |
|
{ |
|
ADLODPerformanceLevels *lpOdPerformanceLevels; |
|
int i, iVddc, lev, ret = 1; |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Set vddc not supported\n"); |
|
return ret; |
|
} |
|
|
|
iVddc = 1000 * fVddc; |
|
ga = &gpus[gpu].adl; |
|
|
|
lev = ga->lpOdParameters.iNumberOfPerformanceLevels - 1; |
|
lpOdPerformanceLevels = alloca(sizeof(ADLODPerformanceLevels) + (lev * sizeof(ADLODPerformanceLevel))); |
|
lpOdPerformanceLevels->iSize = sizeof(ADLODPerformanceLevels) + sizeof(ADLODPerformanceLevel) * lev; |
|
|
|
lock_adl(); |
|
if (ADL_Overdrive5_ODPerformanceLevels_Get(ga->iAdapterIndex, 0, lpOdPerformanceLevels) != ADL_OK) |
|
goto out; |
|
for (i = 0; i < lev; i++) { |
|
if (lpOdPerformanceLevels->aLevels[i].iVddc > iVddc) |
|
lpOdPerformanceLevels->aLevels[i].iVddc = iVddc; |
|
} |
|
lpOdPerformanceLevels->aLevels[lev].iVddc = iVddc; |
|
ADL_Overdrive5_ODPerformanceLevels_Set(ga->iAdapterIndex, lpOdPerformanceLevels); |
|
ADL_Overdrive5_ODPerformanceLevels_Get(ga->iAdapterIndex, 0, lpOdPerformanceLevels); |
|
if (lpOdPerformanceLevels->aLevels[lev].iVddc == iVddc) |
|
ret = 0; |
|
ga->iEngineClock = lpOdPerformanceLevels->aLevels[lev].iEngineClock; |
|
ga->iMemoryClock = lpOdPerformanceLevels->aLevels[lev].iMemoryClock; |
|
ga->iVddc = lpOdPerformanceLevels->aLevels[lev].iVddc; |
|
ga->managed = true; |
|
out: |
|
unlock_adl(); |
|
return ret; |
|
} |
|
|
|
static void get_fanrange(int gpu, int *imin, int *imax) |
|
{ |
|
struct gpu_adl *ga; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Get fanrange not supported\n"); |
|
return; |
|
} |
|
ga = &gpus[gpu].adl; |
|
*imin = ga->lpFanSpeedInfo.iMinPercent; |
|
*imax = ga->lpFanSpeedInfo.iMaxPercent; |
|
} |
|
|
|
int set_fanspeed(int gpu, int iFanSpeed) |
|
{ |
|
struct gpu_adl *ga; |
|
int ret = 1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Set fanspeed not supported\n"); |
|
return ret; |
|
} |
|
|
|
ga = &gpus[gpu].adl; |
|
if (!(ga->lpFanSpeedInfo.iFlags & (ADL_DL_FANCTRL_SUPPORTS_RPM_WRITE | ADL_DL_FANCTRL_SUPPORTS_PERCENT_WRITE ))) { |
|
applog(LOG_DEBUG, "GPU %d doesn't support rpm or percent write", gpu); |
|
return ret; |
|
} |
|
|
|
/* Store what fanspeed we're actually aiming for for re-entrant changes |
|
* in case this device does not support fine setting changes */ |
|
ga->targetfan = iFanSpeed; |
|
|
|
lock_adl(); |
|
if (ADL_Overdrive5_FanSpeed_Get(ga->iAdapterIndex, 0, &ga->lpFanSpeedValue) != ADL_OK) { |
|
applog(LOG_DEBUG, "GPU %d call to fanspeed get failed", gpu); |
|
} |
|
if (!(ga->lpFanSpeedInfo.iFlags & ADL_DL_FANCTRL_SUPPORTS_PERCENT_WRITE)) { |
|
/* Must convert speed to an RPM */ |
|
iFanSpeed = ga->lpFanSpeedInfo.iMaxRPM * iFanSpeed / 100; |
|
ga->lpFanSpeedValue.iSpeedType = ADL_DL_FANCTRL_SPEED_TYPE_RPM; |
|
} else |
|
ga->lpFanSpeedValue.iSpeedType = ADL_DL_FANCTRL_SPEED_TYPE_PERCENT; |
|
if (!(ga->lpFanSpeedValue.iFlags & ADL_DL_FANCTRL_FLAG_USER_DEFINED_SPEED)) { |
|
/* If user defined is not already specified, set it first */ |
|
ga->lpFanSpeedValue.iFlags = ADL_DL_FANCTRL_FLAG_USER_DEFINED_SPEED; |
|
ADL_Overdrive5_FanSpeed_Set(ga->iAdapterIndex, 0, &ga->lpFanSpeedValue); |
|
} |
|
ga->lpFanSpeedValue.iFanSpeed = iFanSpeed; |
|
ret = ADL_Overdrive5_FanSpeed_Set(ga->iAdapterIndex, 0, &ga->lpFanSpeedValue); |
|
ga->managed = true; |
|
unlock_adl(); |
|
|
|
return ret; |
|
} |
|
|
|
#ifdef HAVE_CURSES |
|
static int set_powertune(int gpu, int iPercentage) |
|
{ |
|
struct gpu_adl *ga; |
|
int dummy, ret = 1; |
|
|
|
if (!gpus[gpu].has_adl || !adl_active) { |
|
wlogprint("Set powertune not supported\n"); |
|
return ret; |
|
} |
|
|
|
ga = &gpus[gpu].adl; |
|
|
|
lock_adl(); |
|
ADL_Overdrive5_PowerControl_Set(ga->iAdapterIndex, iPercentage); |
|
ADL_Overdrive5_PowerControl_Get(ga->iAdapterIndex, &ga->iPercentage, &dummy); |
|
if (ga->iPercentage == iPercentage) |
|
ret = 0; |
|
ga->managed = true; |
|
unlock_adl(); |
|
return ret; |
|
} |
|
#endif |
|
|
|
/* Returns whether the fanspeed is optimal already or not. The fan_window bool |
|
* tells us whether the current fanspeed is in the target range for fanspeeds. |
|
*/ |
|
static bool fan_autotune(int gpu, int temp, int fanpercent, int lasttemp, bool *fan_window) |
|
{ |
|
struct cgpu_info *cgpu = &gpus[gpu]; |
|
int tdiff = round(temp - lasttemp); |
|
struct gpu_adl *ga = &cgpu->adl; |
|
int top = gpus[gpu].gpu_fan; |
|
int bot = gpus[gpu].min_fan; |
|
int newpercent = fanpercent; |
|
int iMin = 0, iMax = 100; |
|
|
|
get_fanrange(gpu, &iMin, &iMax); |
|
if (temp > ga->overtemp && fanpercent < iMax) { |
|
applog(LOG_WARNING, "Overheat detected on GPU %d, increasing fan to 100%", gpu); |
|
newpercent = iMax; |
|
|
|
cgpu->device_last_not_well = time(NULL); |
|
cgpu->device_not_well_reason = REASON_DEV_OVER_HEAT; |
|
cgpu->dev_over_heat_count++; |
|
} else if (temp > ga->targettemp && fanpercent < top && tdiff >= 0) { |
|
applog(LOG_DEBUG, "Temperature over target, increasing fanspeed"); |
|
if (temp > ga->targettemp + opt_hysteresis) |
|
newpercent = ga->targetfan + 10; |
|
else |
|
newpercent = ga->targetfan + 5; |
|
if (newpercent > top) |
|
newpercent = top; |
|
} else if (fanpercent > bot && temp < ga->targettemp - opt_hysteresis) { |
|
/* Detect large swings of 5 degrees or more and change fan by |
|
* a proportion more */ |
|
if (tdiff <= 0) { |
|
applog(LOG_DEBUG, "Temperature %d degrees below target, decreasing fanspeed", opt_hysteresis); |
|
newpercent = ga->targetfan - 1 + tdiff / 5; |
|
} else if (tdiff >= 5) { |
|
applog(LOG_DEBUG, "Temperature climbed %d while below target, increasing fanspeed", tdiff); |
|
newpercent = ga->targetfan + tdiff / 5; |
|
} |
|
} else { |
|
|
|
/* We're in the optimal range, make minor adjustments if the |
|
* temp is still drifting */ |
|
if (fanpercent > bot && tdiff < 0 && lasttemp < ga->targettemp) { |
|
applog(LOG_DEBUG, "Temperature dropping while in target range, decreasing fanspeed"); |
|
newpercent = ga->targetfan + tdiff; |
|
} else if (fanpercent < top && tdiff > 0 && temp > ga->targettemp - opt_hysteresis) { |
|
applog(LOG_DEBUG, "Temperature rising while in target range, increasing fanspeed"); |
|
newpercent = ga->targetfan + tdiff; |
|
} |
|
} |
|
|
|
if (newpercent > iMax) |
|
newpercent = iMax; |
|
else if (newpercent < iMin) |
|
newpercent = iMin; |
|
|
|
if (newpercent <= top) |
|
*fan_window = true; |
|
else |
|
*fan_window = false; |
|
|
|
if (newpercent != fanpercent) { |
|
applog(LOG_INFO, "Setting GPU %d fan percentage to %d", gpu, newpercent); |
|
set_fanspeed(gpu, newpercent); |
|
/* If the fanspeed is going down and we're below the top speed, |
|
* consider the fan optimal to prevent minute changes in |
|
* fanspeed delaying GPU engine speed changes */ |
|
if (newpercent < fanpercent && *fan_window) |
|
return true; |
|
return false; |
|
} |
|
return true; |
|
} |
|
|
|
void gpu_autotune(int gpu, enum dev_enable *denable) |
|
{ |
|
int temp, fanpercent, engine, newengine, twintemp = 0; |
|
bool fan_optimal = true, fan_window = true; |
|
struct cgpu_info *cgpu; |
|
struct gpu_adl *ga; |
|
|
|
cgpu = &gpus[gpu]; |
|
ga = &cgpu->adl; |
|
|
|
lock_adl(); |
|
ADL_Overdrive5_CurrentActivity_Get(ga->iAdapterIndex, &ga->lpActivity); |
|
temp = __gpu_temp(ga); |
|
if (ga->twin) |
|
twintemp = __gpu_temp(ga->twin); |
|
fanpercent = __gpu_fanpercent(ga); |
|
unlock_adl(); |
|
|
|
newengine = engine = gpu_engineclock(gpu) * 100; |
|
|
|
if (temp && fanpercent >= 0 && ga->autofan) { |
|
if (!ga->twin) |
|
fan_optimal = fan_autotune(gpu, temp, fanpercent, ga->lasttemp, &fan_window); |
|
else if (ga->autofan && (ga->has_fanspeed || !ga->twin->autofan)) { |
|
/* On linked GPUs, we autotune the fan only once, based |
|
* on the highest temperature from either GPUs */ |
|
int hightemp, fan_gpu; |
|
int lasttemp; |
|
|
|
if (twintemp > temp) { |
|
lasttemp = ga->twin->lasttemp; |
|
hightemp = twintemp; |
|
} else { |
|
lasttemp = ga->lasttemp; |
|
hightemp = temp; |
|
} |
|
if (ga->has_fanspeed) |
|
fan_gpu = gpu; |
|
else |
|
fan_gpu = ga->twin->gpu; |
|
fan_optimal = fan_autotune(fan_gpu, hightemp, fanpercent, lasttemp, &fan_window); |
|
} |
|
} |
|
|
|
if (engine && ga->autoengine) { |
|
if (temp > cgpu->cutofftemp) { |
|
applog(LOG_WARNING, "Hit thermal cutoff limit on GPU %d, disabling!", gpu); |
|
*denable = DEV_RECOVER; |
|
newengine = ga->minspeed; |
|
|
|
cgpu->device_last_not_well = time(NULL); |
|
cgpu->device_not_well_reason = REASON_DEV_THERMAL_CUTOFF; |
|
cgpu->dev_thermal_cutoff_count++; |
|
} else if (temp > ga->overtemp && engine > ga->minspeed) { |
|
applog(LOG_WARNING, "Overheat detected, decreasing GPU %d clock speed", gpu); |
|
newengine = ga->minspeed; |
|
|
|
cgpu->device_last_not_well = time(NULL); |
|
cgpu->device_not_well_reason = REASON_DEV_OVER_HEAT; |
|
cgpu->dev_over_heat_count++; |
|
} else if (temp > ga->targettemp + opt_hysteresis && engine > ga->minspeed && fan_optimal) { |
|
applog(LOG_DEBUG, "Temperature %d degrees over target, decreasing clock speed", opt_hysteresis); |
|
newengine = engine - ga->lpOdParameters.sEngineClock.iStep; |
|
/* Only try to tune engine speed up if this GPU is not disabled */ |
|
} else if (temp < ga->targettemp && engine < ga->maxspeed && fan_window && *denable == DEV_ENABLED) { |
|
int iStep = ga->lpOdParameters.sEngineClock.iStep; |
|
|
|
applog(LOG_DEBUG, "Temperature below target, increasing clock speed"); |
|
if (temp < ga->targettemp - opt_hysteresis) |
|
iStep *= 2; |
|
newengine = engine + iStep; |
|
} else if (temp < ga->targettemp && *denable == DEV_RECOVER && opt_restart) { |
|
applog(LOG_NOTICE, "Device recovered to temperature below target, re-enabling"); |
|
*denable = DEV_ENABLED; |
|
} |
|
|
|
if (newengine > ga->maxspeed) |
|
newengine = ga->maxspeed; |
|
else if (newengine < ga->minspeed) |
|
newengine = ga->minspeed; |
|
|
|
/* Adjust engine clock speed if it's lower, or if it's higher |
|
* but higher than the last intended value as well as the |
|
* current speed, to avoid setting the engine clock speed to |
|
* a speed relateive to a lower profile during idle periods. */ |
|
if (newengine < engine || (newengine > engine && newengine > ga->lastengine)) { |
|
newengine /= 100; |
|
applog(LOG_INFO, "Setting GPU %d engine clock to %d", gpu, newengine); |
|
set_engineclock(gpu, newengine); |
|
} |
|
} |
|
ga->lasttemp = temp; |
|
} |
|
|
|
void set_defaultfan(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ADL_Overdrive5_FanSpeed_Set(ga->iAdapterIndex, 0, &ga->DefFanSpeedValue); |
|
unlock_adl(); |
|
} |
|
|
|
void set_defaultengine(int gpu) |
|
{ |
|
struct gpu_adl *ga; |
|
if (!gpus[gpu].has_adl || !adl_active) |
|
return; |
|
|
|
ga = &gpus[gpu].adl; |
|
lock_adl(); |
|
ADL_Overdrive5_ODPerformanceLevels_Set(ga->iAdapterIndex, ga->DefPerfLev); |
|
unlock_adl(); |
|
} |
|
|
|
#ifdef HAVE_CURSES |
|
void change_autosettings(int gpu) |
|
{ |
|
struct gpu_adl *ga = &gpus[gpu].adl; |
|
char input; |
|
int val; |
|
|
|
wlogprint("Target temperature: %d\n", ga->targettemp); |
|
wlogprint("Overheat temperature: %d\n", ga->overtemp); |
|
wlogprint("Cutoff temperature: %d\n", gpus[gpu].cutofftemp); |
|
wlogprint("Toggle [F]an auto [G]PU auto\nChange [T]arget [O]verheat [C]utoff\n"); |
|
wlogprint("Or press any other key to continue\n"); |
|
input = getch(); |
|
if (!strncasecmp(&input, "f", 1)) { |
|
ga->autofan ^= true; |
|
wlogprint("Fan autotune is now %s\n", ga->autofan ? "enabled" : "disabled"); |
|
if (!ga->autofan) { |
|
wlogprint("Resetting fan to startup settings\n"); |
|
set_defaultfan(gpu); |
|
} |
|
} else if (!strncasecmp(&input, "g", 1)) { |
|
ga->autoengine ^= true; |
|
wlogprint("GPU engine clock autotune is now %s\n", ga->autoengine ? "enabled" : "disabled"); |
|
if (!ga->autoengine) { |
|
wlogprint("Resetting GPU engine clock to startup settings\n"); |
|
set_defaultengine(gpu); |
|
} |
|
} else if (!strncasecmp(&input, "t", 1)) { |
|
val = curses_int("Enter target temperature for this GPU in C (0-200)"); |
|
if (val < 0 || val > 200) |
|
wlogprint("Invalid temperature"); |
|
else |
|
ga->targettemp = val; |
|
} else if (!strncasecmp(&input, "o", 1)) { |
|
wlogprint("Enter overheat temperature for this GPU in C (%d+)", ga->targettemp); |
|
val = curses_int(""); |
|
if (val <= ga->targettemp || val > 200) |
|
wlogprint("Invalid temperature"); |
|
else |
|
ga->overtemp = val; |
|
} else if (!strncasecmp(&input, "c", 1)) { |
|
wlogprint("Enter cutoff temperature for this GPU in C (%d+)", ga->overtemp); |
|
val = curses_int(""); |
|
if (val <= ga->overtemp || val > 200) |
|
wlogprint("Invalid temperature"); |
|
else |
|
gpus[gpu].cutofftemp = val; |
|
} |
|
} |
|
|
|
void change_gpusettings(int gpu) |
|
{ |
|
struct gpu_adl *ga = &gpus[gpu].adl; |
|
float fval, fmin = 0, fmax = 0; |
|
int val, imin = 0, imax = 0; |
|
char input; |
|
int engineclock = 0, memclock = 0, activity = 0, fanspeed = 0, fanpercent = 0, powertune = 0; |
|
float temp = 0, vddc = 0; |
|
|
|
updated: |
|
if (gpu_stats(gpu, &temp, &engineclock, &memclock, &vddc, &activity, &fanspeed, &fanpercent, &powertune)) |
|
wlogprint("Temp: %.1f C\n", temp); |
|
if (fanpercent >= 0 || fanspeed >= 0) { |
|
wlogprint("Fan Speed: "); |
|
if (fanpercent >= 0) |
|
wlogprint("%d%% ", fanpercent); |
|
if (fanspeed >= 0) |
|
wlogprint("(%d RPM)", fanspeed); |
|
wlogprint("\n"); |
|
} |
|
wlogprint("Engine Clock: %d MHz\nMemory Clock: %d Mhz\nVddc: %.3f V\nActivity: %d%%\nPowertune: %d%%\n", |
|
engineclock, memclock, vddc, activity, powertune); |
|
wlogprint("Fan autotune is %s (%d-%d)\n", ga->autofan ? "enabled" : "disabled", |
|
gpus[gpu].min_fan, gpus[gpu].gpu_fan); |
|
wlogprint("GPU engine clock autotune is %s (%d-%d)\n", ga->autoengine ? "enabled" : "disabled", |
|
ga->minspeed / 100, ga->maxspeed / 100); |
|
wlogprint("Change [A]utomatic [E]ngine [F]an [M]emory [V]oltage [P]owertune\n"); |
|
wlogprint("Or press any other key to continue\n"); |
|
input = getch(); |
|
|
|
if (!strncasecmp(&input, "a", 1)) { |
|
change_autosettings(gpu); |
|
} else if (!strncasecmp(&input, "e", 1)) { |
|
get_enginerange(gpu, &imin, &imax); |
|
wlogprint("Enter GPU engine clock speed (%d - %d Mhz)", imin, imax); |
|
val = curses_int(""); |
|
if (val < imin || val > imax) { |
|
wlogprint("Value is outside safe range, are you sure?\n"); |
|
input = getch(); |
|
if (strncasecmp(&input, "y", 1)) |
|
return; |
|
} |
|
if (!set_engineclock(gpu, val)) |
|
wlogprint("Driver reports success but check values below\n"); |
|
else |
|
wlogprint("Failed to modify engine clock speed\n"); |
|
} else if (!strncasecmp(&input, "f", 1)) { |
|
get_fanrange(gpu, &imin, &imax); |
|
wlogprint("Enter fan percentage (%d - %d %)", imin, imax); |
|
val = curses_int(""); |
|
if (val < imin || val > imax) { |
|
wlogprint("Value is outside safe range, are you sure?\n"); |
|
input = getch(); |
|
if (strncasecmp(&input, "y", 1)) |
|
return; |
|
} |
|
if (!set_fanspeed(gpu, val)) |
|
wlogprint("Driver reports success but check values below\n"); |
|
else |
|
wlogprint("Failed to modify fan speed\n"); |
|
} else if (!strncasecmp(&input, "m", 1)) { |
|
get_memoryrange(gpu, &imin, &imax); |
|
wlogprint("Enter GPU memory clock speed (%d - %d Mhz)", imin, imax); |
|
val = curses_int(""); |
|
if (val < imin || val > imax) { |
|
wlogprint("Value is outside safe range, are you sure?\n"); |
|
input = getch(); |
|
if (strncasecmp(&input, "y", 1)) |
|
return; |
|
} |
|
if (!set_memoryclock(gpu, val)) |
|
wlogprint("Driver reports success but check values below\n"); |
|
else |
|
wlogprint("Failed to modify memory clock speed\n"); |
|
} else if (!strncasecmp(&input, "v", 1)) { |
|
get_vddcrange(gpu, &fmin, &fmax); |
|
wlogprint("Enter GPU voltage (%.3f - %.3f V)", fmin, fmax); |
|
fval = curses_float(""); |
|
if (fval < fmin || fval > fmax) { |
|
wlogprint("Value is outside safe range, are you sure?\n"); |
|
input = getch(); |
|
if (strncasecmp(&input, "y", 1)) |
|
return; |
|
} |
|
if (!set_vddc(gpu, fval)) |
|
wlogprint("Driver reports success but check values below\n"); |
|
else |
|
wlogprint("Failed to modify voltage\n"); |
|
} else if (!strncasecmp(&input, "p", 1)) { |
|
val = curses_int("Enter powertune value (-20 - 20)"); |
|
if (val < -20 || val > 20) { |
|
wlogprint("Value is outside safe range, are you sure?\n"); |
|
input = getch(); |
|
if (strncasecmp(&input, "y", 1)) |
|
return; |
|
} |
|
if (!set_powertune(gpu, val)) |
|
wlogprint("Driver reports success but check values below\n"); |
|
else |
|
wlogprint("Failed to modify powertune value\n"); |
|
} else { |
|
clear_logwin(); |
|
return; |
|
} |
|
sleep(1); |
|
goto updated; |
|
} |
|
#endif |
|
|
|
static void free_adl(void) |
|
{ |
|
ADL_Main_Memory_Free ((void **)&lpInfo); |
|
ADL_Main_Control_Destroy (); |
|
#if defined (LINUX) |
|
dlclose(hDLL); |
|
#else |
|
FreeLibrary(hDLL); |
|
#endif |
|
} |
|
|
|
void clear_adl(int nDevs) |
|
{ |
|
struct gpu_adl *ga; |
|
int i; |
|
|
|
if (!adl_active) |
|
return; |
|
|
|
lock_adl(); |
|
/* Try to reset values to their defaults */ |
|
for (i = 0; i < nDevs; i++) { |
|
ga = &gpus[i].adl; |
|
/* Only reset the values if we've changed them at any time */ |
|
if (!gpus[i].has_adl || !ga->managed) |
|
continue; |
|
ADL_Overdrive5_ODPerformanceLevels_Set(ga->iAdapterIndex, ga->DefPerfLev); |
|
free(ga->DefPerfLev); |
|
ADL_Overdrive5_FanSpeed_Set(ga->iAdapterIndex, 0, &ga->DefFanSpeedValue); |
|
ADL_Overdrive5_FanSpeedToDefault_Set(ga->iAdapterIndex, 0); |
|
} |
|
adl_active = false; |
|
unlock_adl(); |
|
free_adl(); |
|
} |
|
#endif /* HAVE_ADL */
|
|
|