nvidia libs have inconsistent ordering (#7473)

The runtime and management libraries may not always have
identical ordering, so use the device UUID to correlate instead of ID.
This commit is contained in:
Daniel Hiltgen
2024-11-02 16:35:41 -07:00
committed by GitHub
parent b8d5036e33
commit 29ab9fa7d7
3 changed files with 13 additions and 9 deletions

View File

@@ -25,7 +25,7 @@ typedef struct nvml_handle {
uint16_t verbose;
nvmlReturn_t (*nvmlInit_v2)(void);
nvmlReturn_t (*nvmlShutdown)(void);
nvmlReturn_t (*nvmlDeviceGetHandleByIndex)(unsigned int, nvmlDevice_t *);
nvmlReturn_t (*nvmlDeviceGetHandleByUUID)(const char *, nvmlDevice_t *);
nvmlReturn_t (*nvmlDeviceGetMemoryInfo)(nvmlDevice_t, nvmlMemory_t *);
} nvml_handle_t;
@@ -41,7 +41,7 @@ typedef struct nvml_compute_capability {
} nvml_compute_capability_t;
void nvml_init(char *nvml_lib_path, nvml_init_resp_t *resp);
void nvml_get_free(nvml_handle_t ch, int device_id, uint64_t *free, uint64_t *total, uint64_t *used);
void nvml_get_free(nvml_handle_t ch, char *uuid, uint64_t *free, uint64_t *total, uint64_t *used);
void nvml_release(nvml_handle_t ch);
#endif // __GPU_INFO_NVML_H__