Memory diagnostics in a heterogeneous computing platform

A method for memory diagnostics by an Information Handling System (IHS), the method comprising:

• loading a plurality of memory device drivers for accessing one or more replaceable memory modules, each memory device driver supporting a respective diagnostic Application Program Interface (API);
• detecting, by a Unified Extensible Firmware Interface (UEFI) memory diagnostics module, a memory error resulting from an attempt to access the one or more replaceable memory modules, based at least in part on a callback registered with the UEFI memory diagnostics module to receive memory error notifications from a respective one of the plurality of memory device drivers and return a handle to the respective diagnostic API supported by the respective one of the plurality of memory device drivers;
• determining a memory address associated with the memory error;
• identifying, based on the memory address, a first memory device driver of the plurality of memory device drivers for accessing the memory address, and retrieving the handle to the respective diagnostic API supported by the first memory device driver;
• utilizing, via the handle, the respective diagnostic API supported by the first memory device driver to identify a first of the one or more replaceable memory modules corresponding to the memory address as a source of the memory error;
• issuing, via Operating System (OS) telemetry, a notification specifying a failure of the first of the one or more replaceable memory modules;
• and
• in response to the failure, repairing the first of the one or more replaceable memory modules, using a diagnostic tool.

A non-transitory storage device that comprises instructions stored thereon, wherein execution of the instructions by one or more processors of an Information Handling System (IHS) causes the one or more processors to:

• load a plurality of memory device drivers for access to one or more replaceable memory modules, wherein each memory device driver supports a respective diagnostic Application Program Interface (API);
• detect, by a Unified Extensible Firmware Interface (UEFI) memory diagnostics module, a memory error that resulted from an attempt to access the one or more replaceable memory modules, based at least in part on a callback registered with the UEFI memory diagnostics module to receive memory error notifications from a respective one of the plurality of memory device drivers and return a handle to the respective diagnostic API supported by the respective one of the plurality of memory device drivers;
• determine a memory address associated with the memory error;
• identify, based on the memory address, a first memory device driver of the plurality of memory device drivers for access to the memory address, and retrieve the handle to the respective diagnostic API supported by the first memory device driver;
• utilize, via the handle, the respective diagnostic API supported by the first memory device driver to identify a first of the one or more replaceable memory modules that correspond to the memory address as a source of the memory error;
• issue, via Operating System (OS) telemetry, a notification configured to specify a failure of the first of the one or more replaceable memory modules;
• and
• in response to the failure, repair the first of the one or more replaceable memory modules, based at least in part, on a diagnostic tool.