Vmware, Inc.

Patrick Tullmann

Dilpreet S Bindra

Carl Alan Waldspurger

Osten Kit Colbert

Kinshuk Govil

Gagandeep Singh Arneja

Gregory Hutchins

Gregory T Harm

Swathi Koundinya

Gregory T Harmon

A method is provided for recovering from an uncorrected memory error located at a memory address as identified by a memory device. A stored hash value for a memory page corresponding to the identified memory address is used to determine the correct data. Because the memory device specifies the location of the corrupted data, and the size of the window where the corruption occurred, the stored hash can be used to verify memory page reconstruction. With the known good part of the data in hand, the hashes of the pages using possible values in place of the corrupted data are calculated. It is expected that there will be a match between the previously stored hash and one of the computed hashes. As long as there is one and only one match, then that value, used in the place of the corrupted data, is the correct value. The corrupt data, once replaced, allows operation of the memory device to continue without needing to interrupt or otherwise affect a system's operation.

Transparent recovery from hardware memory errors

In one approach, a method is described of generating an interface for an operating system kernel. The method calls for creating an input file, where the input file includes a node structure for the interface. A kernel component of the interface is generated from the input file. A user space component of the interface is also generated from the input file.

Static, configurable kernel interface

A software component is executed to carry out a task, the task including a subtask. An external function is called to perform the subtask, the external function executing in a separate thread or process. The component receives an observation recorded by the external function, the observation including an identifier of a possible error condition and instance data associated with the possible error condition. The possible error condition being a cause of the failure of the external function to carry out the subtask. If the task cannot be completed, then a new observation is recorded along with the received observation, the new observation being related to a possible error condition of the component, which is a cause of the failure of the component to carry out the task. When the task can be completed despite the failure of the external function, the observation recorded by the external function is cleared.

Error reporting through observation correlation

Growing community of inventors

Millbrae, CA, United States of America

Patrick Tullmann