Different vendors have different NVDIMM hardware, from 3D XPoint to NVDIMM-N and implement DSMs (Device Specific Method) functions in the BIOS/UEFI adhering to a variety of NVDIMM specs from the ones put out by Intel to Microsoft. As a result of this, if we want to get the health status of NVDIMM hardware, there does not exist a standard way of doing that. The solution to this problem is extending ndctl which is a user space utility library for managing NVDIMM hardware. This presentation will go through the ndctl code base and the function call sequence while executing a specific command, say 'ndctl list'. One will also learn how to add options for the 'ndctl list' command as well as how to add new commands such as 'ndctl dell-dsm'. Hope this presentation will help motivate different vendors (Dell, HPE, Lenovo, etc) to submit code to upstream nvdimm community to support their DSM.
