@ -19,7 +19,7 @@
@@ -19,7 +19,7 @@
# include <stdio.h>
# include <stdlib.h>
# include <errno .h>
# include <string .h>
# ifndef _MSC_VER
# include <libgen.h>
# endif
@ -28,13 +28,21 @@
@@ -28,13 +28,21 @@
# include "cuda_runtime.h"
# include "nvml.h"
extern wrap_nvml_handle * hnvml ;
extern int device_map [ 8 ] ;
/*
* Wrappers to emulate dlopen ( ) on other systems like Windows
*/
# if defined(_MSC_VER) || defined(_WIN32) || defined(_WIN64)
# include <windows.h>
static void * wrap_dlopen ( const char * filename ) {
return ( void * ) LoadLibrary ( filename ) ;
HMODULE h = LoadLibrary ( filename ) ;
if ( ! h & & opt_debug ) {
applog ( LOG_DEBUG , " dlopen(%d): failed to load %s " ,
GetLastError ( ) , filename ) ;
}
return ( void * ) h ;
}
static void * wrap_dlsym ( void * h , const char * sym ) {
return ( void * ) GetProcAddress ( ( HINSTANCE ) h , sym ) ;
@ -46,9 +54,16 @@
@@ -46,9 +54,16 @@
# else
/* assume we can use dlopen itself... */
# include <dlfcn.h>
# include <errno.h>
static void * wrap_dlopen ( const char * filename ) {
return dlopen ( filename , RTLD_NOW ) ;
void * h = dlopen ( filename , RTLD_NOW ) ;
if ( h = = NULL & & opt_debug ) {
applog ( LOG_DEBUG , " dlopen(%d): failed to load %s " ,
errno , filename ) ;
}
return ( void * ) h ;
}
static void * wrap_dlsym ( void * h , const char * sym ) {
return dlsym ( h , sym ) ;
}
@ -57,7 +72,7 @@
@@ -57,7 +72,7 @@
}
# endif
# if defined(__cplusplus)
# ifdef __cplusplus
extern " C " {
# endif
@ -66,51 +81,27 @@ wrap_nvml_handle * wrap_nvml_create()
@@ -66,51 +81,27 @@ wrap_nvml_handle * wrap_nvml_create()
int i = 0 ;
wrap_nvml_handle * nvmlh = NULL ;
/*
* We use hard - coded library installation locations for the time being . . .
* No idea where or if libnvidia - ml . so is installed on MacOS X , a
* deep scouring of the filesystem on one of the Mac CUDA build boxes
* I used turned up nothing , so for now it ' s not going to work on OSX .
*/
# if defined(_WIN64)
/* 64-bit Windows */
# define libnvidia_ml "%PROGRAMFILES% / NVIDIA Corporation / NVSMI / nvml.dll"
# elif defined(_WIN32) || defined(_MSC_VER)
/* 32-bit Windows */
# define libnvidia_ml "%PROGRAMFILES% / NVIDIA Corporation / NVSMI / nvml.dll"
# elif defined(__linux) && (defined(__i386__) || defined(__ARM_ARCH_7A__))
/* 32-bit linux assumed */
# define libnvidia_ml " / usr / lib32 / libnvidia-ml.so"
# elif defined(__linux)
/* 64-bit linux assumed */
# define libnvidia_ml " / usr / lib / libnvidia-ml.so"
# if defined(WIN32)
/* Windows (do not use slashes, else ExpandEnvironmentStrings will mix them) */
# define libnvidia_ml "%PROGRAMFILES%\\NVIDIA Corporation\\NVSMI\\nvml.dll"
# else
# error "Unrecognized platform: need NVML DLL path for this platform..."
/* linux assumed */
# define libnvidia_ml "libnvidia-ml.so"
# endif
# if WIN32
char tmp [ 512 ] ;
ExpandEnvironmentStringsA ( libnvidia_ml , tmp , sizeof ( tmp ) ) ;
# ifdef WIN32
ExpandEnvironmentStrings ( libnvidia_ml , tmp , sizeof ( tmp ) ) ;
# else
char tmp [ 512 ] = libnvidia_ml ;
strcpy ( tmp , libnvidia_ml ) ;
# endif
void * nvml_dll = wrap_dlopen ( tmp ) ;
if ( nvml_dll = = NULL ) {
# ifdef WIN32
char lib [ ] = " nvml.dll " ;
# else
char lib [ 64 ] = { ' \0 ' } ;
snprintf ( lib , sizeof ( lib ) , " %s " , basename ( tmp ) ) ;
/* try dlopen without path, here /usr/lib/nvidia-340/libnvidia-ml.so */
nvml_dll = wrap_dlopen ( " nvml.dll " ) ;
if ( nvml_dll = = NULL )
# endif
nvml_dll = wrap_dlopen ( lib ) ;
if ( opt_debug )
applog ( LOG_DEBUG , " dlopen: %s=%p " , lib , nvml_dll ) ;
}
if ( nvml_dll = = NULL ) {
if ( opt_debug )
applog ( LOG_DEBUG , " dlopen(%d): failed to load %s " , errno , tmp ) ;
return NULL ;
}
@ -120,9 +111,10 @@ wrap_nvml_handle * wrap_nvml_create()
@@ -120,9 +111,10 @@ wrap_nvml_handle * wrap_nvml_create()
nvmlh - > nvmlInit = ( wrap_nvmlReturn_t ( * ) ( void ) )
wrap_dlsym ( nvmlh - > nvml_dll , " nvmlInit_v2 " ) ;
if ( ! nvmlh - > nvmlInit )
if ( ! nvmlh - > nvmlInit ) {
nvmlh - > nvmlInit = ( wrap_nvmlReturn_t ( * ) ( void ) )
wrap_dlsym ( nvmlh - > nvml_dll , " nvmlInit " ) ;
}
nvmlh - > nvmlDeviceGetCount = ( wrap_nvmlReturn_t ( * ) ( int * ) )
wrap_dlsym ( nvmlh - > nvml_dll , " nvmlDeviceGetCount_v2 " ) ;
nvmlh - > nvmlDeviceGetHandleByIndex = ( wrap_nvmlReturn_t ( * ) ( int , wrap_nvmlDevice_t * ) )
@ -153,11 +145,10 @@ wrap_nvml_handle * wrap_nvml_create()
@@ -153,11 +145,10 @@ wrap_nvml_handle * wrap_nvml_create()
nvmlh - > nvmlDeviceGetPciInfo = = NULL | |
nvmlh - > nvmlDeviceGetName = = NULL | |
nvmlh - > nvmlDeviceGetTemperature = = NULL | |
nvmlh - > nvmlDeviceGetFanSpeed = = NULL | |
nvmlh - > nvmlDeviceGetPowerUsage = = NULL )
nvmlh - > nvmlDeviceGetFanSpeed = = NULL )
{
if ( opt_debug )
applog ( LOG_DEBUG , " Failed to obtain all required NVML function pointers " ) ;
applog ( LOG_DEBUG , " Failed to obtain required NVML function pointers " ) ;
wrap_dlclose ( nvmlh - > nvml_dll ) ;
free ( nvmlh ) ;
return NULL ;
@ -342,14 +333,11 @@ int wrap_nvml_destroy(wrap_nvml_handle *nvmlh)
@@ -342,14 +333,11 @@ int wrap_nvml_destroy(wrap_nvml_handle *nvmlh)
/* api functions */
extern wrap_nvml_handle * nvmlh ;
extern int device_map [ 8 ] ;
unsigned int gpu_fanpercent ( struct cgpu_info * gpu )
{
unsigned int pct = 0 ;
if ( nvmlh ) {
wrap_nvml_get_fanpcnt ( nvmlh , device_map [ gpu - > thr_id ] , & pct ) ;
if ( h nvml) {
wrap_nvml_get_fanpcnt ( h nvml, device_map [ gpu - > thr_id ] , & pct ) ;
}
return pct ;
}
@ -357,9 +345,9 @@ unsigned int gpu_fanpercent(struct cgpu_info *gpu)
@@ -357,9 +345,9 @@ unsigned int gpu_fanpercent(struct cgpu_info *gpu)
double gpu_temp ( struct cgpu_info * gpu )
{
double tc = 0.0 ;
if ( nvmlh ) {
if ( h nvml) {
unsigned int tmp = 0 ;
wrap_nvml_get_tempC ( nvmlh , device_map [ gpu - > thr_id ] , & tmp ) ;
wrap_nvml_get_tempC ( h nvml, device_map [ gpu - > thr_id ] , & tmp ) ;
tc = ( double ) tmp ;
}
return tc ;
@ -368,8 +356,8 @@ double gpu_temp(struct cgpu_info *gpu)
@@ -368,8 +356,8 @@ double gpu_temp(struct cgpu_info *gpu)
unsigned int gpu_clock ( struct cgpu_info * gpu )
{
unsigned int freq = 0 ;
if ( nvmlh ) {
wrap_nvml_get_clock ( nvmlh , device_map [ gpu - > thr_id ] , NVML_CLOCK_SM , & freq ) ;
if ( h nvml) {
wrap_nvml_get_clock ( h nvml, device_map [ gpu - > thr_id ] , NVML_CLOCK_SM , & freq ) ;
}
return freq ;
}
@ -377,8 +365,8 @@ unsigned int gpu_clock(struct cgpu_info *gpu)
@@ -377,8 +365,8 @@ unsigned int gpu_clock(struct cgpu_info *gpu)
unsigned int gpu_power ( struct cgpu_info * gpu )
{
unsigned int mw = 0 ;
if ( nvmlh ) {
wrap_nvml_get_power_usage ( nvmlh , device_map [ gpu - > thr_id ] , & mw ) ;
if ( h nvml) {
wrap_nvml_get_power_usage ( h nvml, device_map [ gpu - > thr_id ] , & mw ) ;
}
return mw ;
}
@ -386,9 +374,8 @@ unsigned int gpu_power(struct cgpu_info *gpu)
@@ -386,9 +374,8 @@ unsigned int gpu_power(struct cgpu_info *gpu)
int gpu_pstate ( struct cgpu_info * gpu )
{
int pstate = 0 ;
if ( nvmlh ) {
wrap_nvml_get_pstate ( nvmlh , device_map [ gpu - > thr_id ] , & pstate ) ;
//gpu->gpu_pstate = pstate;
if ( hnvml ) {
wrap_nvml_get_pstate ( hnvml , device_map [ gpu - > thr_id ] , & pstate ) ;
}
return pstate ;
}
@ -433,7 +420,7 @@ int gpu_pstate(struct cgpu_info *gpu)
@@ -433,7 +420,7 @@ int gpu_pstate(struct cgpu_info *gpu)
* nvmlDeviceGetFanSpeed
nvmlDeviceGetGpuOperationMode
nvmlDeviceGetHandleByIndex
nvmlDeviceGetHandleByIndex_v2
* nvmlDeviceGetHandleByIndex_v2
nvmlDeviceGetHandleByPciBusId
nvmlDeviceGetHandleByPciBusId_v2
nvmlDeviceGetHandleBySerial
@ -450,7 +437,7 @@ int gpu_pstate(struct cgpu_info *gpu)
@@ -450,7 +437,7 @@ int gpu_pstate(struct cgpu_info *gpu)
nvmlDeviceGetMinorNumber
nvmlDeviceGetMultiGpuBoard
nvmlDeviceGetName
nvmlDeviceGetPciInfo
* nvmlDeviceGetPciInfo
nvmlDeviceGetPciInfo_v2
* nvmlDeviceGetPerformanceState
nvmlDeviceGetPersistenceMode