Skip to content

Display GPU architecture in json #494

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Mahsuba wants to merge 1 commit into
base: master
Choose a base branch
from
+155 −0
Open

Display GPU architecture in json #494

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions include/nvtop/extract_gpuinfo_common.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,7 @@
#define GPUINFO_STATIC_FIELD_VALID(structPtr, field) VALUE_IS_VALID(structPtr, field, gpuinfo_)
enum gpuinfo_static_info_valid {
gpuinfo_device_name_valid = 0,
gpuinfo_device_architecture_valid,
gpuinfo_max_pcie_gen_valid,
gpuinfo_max_pcie_link_width_valid,
gpuinfo_temperature_shutdown_threshold_valid,
Expand All @@ -68,6 +69,7 @@ enum gpuinfo_static_info_valid {

struct gpuinfo_static_info {
char device_name[MAX_DEVICE_NAME];
char device_architecture[MAX_DEVICE_NAME];
unsigned max_pcie_gen;
unsigned max_pcie_link_width;
unsigned temperature_shutdown_threshold;
Expand Down
98 changes: 98 additions & 0 deletions src/extract_gpuinfo_amdgpu.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -600,6 +600,104 @@ static void gpuinfo_amdgpu_populate_static_info(struct gpu_info *_gpu_info) {
}
}

if (info_query_success) {
const char *arch_name = NULL;
switch (info.family_id) {
#ifdef AMDGPU_FAMILY_SI
case AMDGPU_FAMILY_SI:
arch_name = "GCN 1";
break;
#endif
#ifdef AMDGPU_FAMILY_CI
case AMDGPU_FAMILY_CI:
arch_name = "GCN 2";
break;
#endif
#ifdef AMDGPU_FAMILY_KV
case AMDGPU_FAMILY_KV:
arch_name = "GCN 2";
break;
#endif
#ifdef AMDGPU_FAMILY_VI
case AMDGPU_FAMILY_VI:
arch_name = "GCN 3";
break;
#endif
#ifdef AMDGPU_FAMILY_CZ
case AMDGPU_FAMILY_CZ:
arch_name = "GCN 3";
break;
#endif
#ifdef AMDGPU_FAMILY_AI
case AMDGPU_FAMILY_AI:
arch_name = "GCN 5 (Vega)";
break;
#endif
#ifdef AMDGPU_FAMILY_RV
case AMDGPU_FAMILY_RV:
arch_name = "GCN 5 (Vega)";
break;
#endif
#ifdef AMDGPU_FAMILY_NV
case AMDGPU_FAMILY_NV:
arch_name = "RDNA";
break;
#endif
#ifdef AMDGPU_FAMILY_VGH
case AMDGPU_FAMILY_VGH:
arch_name = "RDNA 2";
break;
#endif
#ifdef AMDGPU_FAMILY_YC
case AMDGPU_FAMILY_YC:
arch_name = "RDNA 2";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_10_3_6
case AMDGPU_FAMILY_GC_10_3_6:
arch_name = "RDNA 2";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_10_3_7
case AMDGPU_FAMILY_GC_10_3_7:
arch_name = "RDNA 2";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_11_0_0
case AMDGPU_FAMILY_GC_11_0_0:
arch_name = "RDNA 3";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_11_0_1
case AMDGPU_FAMILY_GC_11_0_1:
arch_name = "RDNA 3";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_11_5_0
case AMDGPU_FAMILY_GC_11_5_0:
arch_name = "RDNA 3.5";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_12_0_0
case AMDGPU_FAMILY_GC_12_0_0:
arch_name = "RDNA 4";
break;
#endif
#ifdef AMDGPU_FAMILY_GC_12_0_1
case AMDGPU_FAMILY_GC_12_0_1:
arch_name = "RDNA 4";
break;
#endif
default:
break;
}
if (arch_name) {
strncpy(static_info->device_architecture, arch_name, MAX_DEVICE_NAME - 1);
static_info->device_architecture[MAX_DEVICE_NAME - 1] = '\0';
SET_VALID(gpuinfo_device_architecture_valid, static_info->valid);
}
}

// Retrieve infos from sysfs.

// 1) Fan
Expand Down
48 changes: 48 additions & 0 deletions src/extract_gpuinfo_nvidia.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,8 @@ static const char *(*nvmlErrorString)(nvmlReturn_t);

static nvmlReturn_t (*nvmlDeviceGetName)(nvmlDevice_t device, char *name, unsigned int length);

static nvmlReturn_t (*nvmlDeviceGetArchitecture)(nvmlDevice_t device, unsigned int *arch);

typedef struct {
char busIdLegacy[16];
unsigned int domain;
Expand Down Expand Up @@ -352,6 +354,9 @@ static bool gpuinfo_nvidia_init(void) {
if (!nvmlDeviceGetName)
goto init_error_clean_exit;

// Optional; not present in older drivers, absence is not an error
nvmlDeviceGetArchitecture = dlsym(libnvidia_ml_handle, "nvmlDeviceGetArchitecture");

nvmlDeviceGetPciInfo = dlsym(libnvidia_ml_handle, "nvmlDeviceGetPciInfo_v3");
if (!nvmlDeviceGetPciInfo)
nvmlDeviceGetPciInfo = dlsym(libnvidia_ml_handle, "nvmlDeviceGetPciInfo_v2");
Expand Down Expand Up @@ -560,6 +565,49 @@ static void gpuinfo_nvidia_populate_static_info(struct gpu_info *_gpu_info) {
if (last_nvml_return_status == NVML_SUCCESS)
SET_VALID(gpuinfo_device_name_valid, static_info->valid);

if (nvmlDeviceGetArchitecture) {
unsigned int arch = 0;
if (nvmlDeviceGetArchitecture(device, &arch) == NVML_SUCCESS) {
const char *arch_name = NULL;
switch (arch) {
case 2:
arch_name = "Kepler";
break;
case 3:
arch_name = "Maxwell";
break;
case 4:
arch_name = "Pascal";
break;
case 5:
arch_name = "Volta";
break;
case 6:
arch_name = "Turing";
break;
case 7:
arch_name = "Ampere";
break;
case 8:
arch_name = "Ada";
break;
case 9:
arch_name = "Hopper";
break;
case 10:
arch_name = "Blackwell";
break;
default:
break;
}
if (arch_name) {
strncpy(static_info->device_architecture, arch_name, MAX_DEVICE_NAME - 1);
static_info->device_architecture[MAX_DEVICE_NAME - 1] = '\0';
SET_VALID(gpuinfo_device_architecture_valid, static_info->valid);
}
}
}

last_nvml_return_status = nvmlDeviceGetMaxPcieLinkGeneration(device, &static_info->max_pcie_gen);
if (last_nvml_return_status == NVML_SUCCESS)
SET_VALID(gpuinfo_max_pcie_gen_valid, static_info->valid);
Expand Down
7 changes: 7 additions & 0 deletions src/interface.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2153,6 +2153,7 @@ void print_snapshot(struct list_head *devices, bool use_fahrenheit_option, bool
const char *indent_level_eight = " ";

const char *device_name_field = "device_name";
const char *device_architecture_field = "device_architecture";
const char *gpu_clock_field = "gpu_clock";
const char *mem_clock_field = "mem_clock";
const char *temp_field = "temp";
Expand All @@ -2172,6 +2173,12 @@ void print_snapshot(struct list_head *devices, bool use_fahrenheit_option, bool
else
printf("%s\"%s\": null,\n", indent_level_four, device_name_field);

// Device Architecture
if (GPUINFO_STATIC_FIELD_VALID(&device->static_info, device_architecture))
printf("%s\"%s\": \"%s\",\n", indent_level_four, device_architecture_field, device->static_info.device_architecture);
else
printf("%s\"%s\": null,\n", indent_level_four, device_architecture_field);

// GPU Clock Speed
if (GPUINFO_DYNAMIC_FIELD_VALID(&device->dynamic_info, gpu_clock_speed))
printf("%s\"%s\": \"%uMHz\",\n", indent_level_four, gpu_clock_field, device->dynamic_info.gpu_clock_speed);
Expand Down