A method and apparatus for improving reliability of processor core

Abstract

The invention discloses a method and a device for improving the reliability of a processor core. The device for improving the reliability of a processor core includes error discovery and correction logic, correctable error processing logic, error recording and reporting logic, abnormal recording and reporting Logic and methods include error discovery and handling related to address translation, error discovery and handling related to the index fetch process except address translation, error discovery and handling related to Load operations except address translation, error discovery and handling related to Store operations except address translation and processing, Snoop or Evict operation-related error discovery and processing steps. The present invention can correct the errors in the storage body of the processor core as much as possible, and record and report these errors from multiple angles, thereby using multiple error handling means to jointly improve the reliability of the processor core, and try to ensure that the processor core and the system are It can still work normally after encountering errors.

Description

technical field [0001] The invention relates to the field of processor design, in particular to a method and device for improving the reliability of a processor core. Background technique [0002] In modern processor designs, SRAM is often used to cache portions of data in order to improve performance. Among them, the first-level instruction cache, the first-level data cache, and the address translation cache have all become mandatory components in high-performance processor cores. However, as the number of transistors integrated in processors has increased, the environment in which these memories operate has become more complex. Electromagnetic interference, erratic currents, etc. can often cause transient or permanent errors in memory banks in the processor core. These errors will directly manifest as abnormal changes in some data, which will cause failures in applications or systems using these data. For application scenarios such as servers and automobiles, once there...

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is to provide a method and device for improving the reliability of the processor core, aiming at the problem that the memory bank error during the operation of the processor core may cause the processor and the whole system to fail to work normally. The invention can correct the errors in the storage body of the processor core as much as possible, and record and report these errors from multiple angles, thereby using multiple error handling methods to jointly improve the reliability of the processor core, and try to ensure that the processor core and the system are It still works fine after the error

Images