Browse Source

nvapi: fix plimit argument on x64

now use the same percents as the x86

nvml watts (like linux) could work but requires admin rights
pull/2/head
Tanguy Pruvot 8 years ago
parent
commit
762c7f8656
  1. 2
      ccminer.cpp
  2. 3
      compat/nvapi/nvapi_ccminer.h
  3. 18
      nvml.cpp

2
ccminer.cpp

@ -321,7 +321,7 @@ Options:\n\
--max-diff=N Only mine if net difficulty is less than specified value\n\ --max-diff=N Only mine if net difficulty is less than specified value\n\
Can be tuned with --resume-diff=N to set a resume value\n\ Can be tuned with --resume-diff=N to set a resume value\n\
--max-log-rate Interval to reduce per gpu hashrate logs (default: 3)\n" --max-log-rate Interval to reduce per gpu hashrate logs (default: 3)\n"
#if defined(__linux) || defined(_WIN64) /* via nvml */ #if defined(__linux) /* via nvml */
"\ "\
--mem-clock=3505 Set the gpu memory max clock (346.72+ driver)\n\ --mem-clock=3505 Set the gpu memory max clock (346.72+ driver)\n\
--gpu-clock=1150 Set the gpu engine max clock (346.72+ driver)\n\ --gpu-clock=1150 Set the gpu engine max clock (346.72+ driver)\n\

3
compat/nvapi/nvapi_ccminer.h

@ -7,7 +7,8 @@ NvAPI_Status nvapi_dll_init();
typedef struct { typedef struct {
NvU32 version; NvU32 version;
NvU32 flags; NvU8 valid;
NvU8 count;
struct struct
{ {
NvU32 pstate; // Assumption NvU32 pstate; // Assumption

18
nvml.cpp

@ -44,6 +44,8 @@ uint32_t clock_prev[MAX_GPUS] = { 0 };
uint32_t clock_prev_mem[MAX_GPUS] = { 0 }; uint32_t clock_prev_mem[MAX_GPUS] = { 0 };
uint32_t limit_prev[MAX_GPUS] = { 0 }; uint32_t limit_prev[MAX_GPUS] = { 0 };
static bool nvml_plimit_set = false;
/* /*
* Wrappers to emulate dlopen() on other systems like Windows * Wrappers to emulate dlopen() on other systems like Windows
*/ */
@ -525,10 +527,13 @@ int nvml_set_plimit(nvml_handle *nvmlh, int dev_id)
plimit = max(plimit, pmin); plimit = max(plimit, pmin);
rc = nvmlh->nvmlDeviceSetPowerManagementLimit(nvmlh->devs[n], plimit); rc = nvmlh->nvmlDeviceSetPowerManagementLimit(nvmlh->devs[n], plimit);
if (rc != NVML_SUCCESS) { if (rc != NVML_SUCCESS) {
#ifndef WIN32
applog(LOG_WARNING, "GPU #%d: plimit %s", dev_id, nvmlh->nvmlErrorString(rc)); applog(LOG_WARNING, "GPU #%d: plimit %s", dev_id, nvmlh->nvmlErrorString(rc));
#endif
return -1; return -1;
} else { } else {
device_plimit[dev_id] = plimit / 1000; device_plimit[dev_id] = plimit / 1000;
nvml_plimit_set = true;
} }
if (!opt_quiet) { if (!opt_quiet) {
@ -1381,7 +1386,16 @@ int nvapi_pstateinfo(unsigned int devNum)
} }
uint32_t plim = nvapi_get_plimit(devNum); uint32_t plim = nvapi_get_plimit(devNum);
applog(LOG_RAW, " Power limit is set to %u%%", plim); double min_pw = 0, max_pw = 0; // percent
NVAPI_GPU_POWER_INFO nfo = { 0 };
nfo.version = NVAPI_GPU_POWER_INFO_VER;
ret = NvAPI_DLL_ClientPowerPoliciesGetInfo(phys[devNum], &nfo);
if (ret == NVAPI_OK && nfo.valid) {
min_pw = (double)nfo.entries[0].min_power / 1000;
max_pw = (double)nfo.entries[0].max_power / 1000;
}
applog(LOG_RAW, " Power limit is set to %u%%, range [%.0f-%.0f%%]", plim, min_pw, max_pw);
#if 0 #if 0
NVAPI_COOLER_SETTINGS *cooler; NVAPI_COOLER_SETTINGS *cooler;
@ -1812,7 +1826,7 @@ int nvapi_init_settings()
for (int n=0; n < opt_n_threads; n++) { for (int n=0; n < opt_n_threads; n++) {
int dev_id = device_map[n % MAX_GPUS]; int dev_id = device_map[n % MAX_GPUS];
if (device_plimit[dev_id] && !hnvml) { if (device_plimit[dev_id] && !nvml_plimit_set) {
if (nvapi_set_plimit(nvapi_dev_map[dev_id], device_plimit[dev_id]) == NVAPI_OK) { if (nvapi_set_plimit(nvapi_dev_map[dev_id], device_plimit[dev_id]) == NVAPI_OK) {
uint32_t res = nvapi_get_plimit(nvapi_dev_map[dev_id]); uint32_t res = nvapi_get_plimit(nvapi_dev_map[dev_id]);
gpulog(LOG_INFO, n, "Power limit is set to %u%%", res); gpulog(LOG_INFO, n, "Power limit is set to %u%%", res);

Loading…
Cancel
Save