International Business Machines Corporation

Richard N Gustafson

Charles Franklin Webb

John Stephen Liptay

Bruce L McGilvray

Frederick O Flusche

Error detection and recovery is provided in a processor of small size and which can be integrated on a single chip by providing buffers for both data and processor status codes in order to contain errors until a subsequent check point preferably generated at the termination of each instruction is reached without detection of an error. Retry of an instruction can therefore be initiated using the status and data validated at the termination of the previous check point and without placing error correction processing in any critical path of the processor. Error detection is accomplished by comparing outputs of at least a pair of unchecked processors for both memory access requests and output data and status codes. Input to the processors is subjected to a parity check and parity check bits are generated for memory access requests. Error correcting codes are generated for data and status codes to allow correction of single bit errors during transmission within the processor or at a storage system. When an error is detected, all data which has not been validated, preferably by changing the logical value of a flag bit associated with each code, at the most recently generated check point is erased. Data codes in which the flag bit has been changed may be transferred to a storage system autonomously even after an error has occurred.

Data processor with enhanced error recovery

A multiprocessor (MP) system is described having central processors (CPs) in which each CP has a store-in-cache (SIC) with an associated processor directory (PD). Each PD has a plurality of line entries which define the content of corresponding line positions in the associated SIC. Each line entry has an associated data shareability control bit, designated EX, which may be set to a one or zero state to indicate, respectively, the exclusive (EX) or readonly (RO) state of the associated line. An exclusive line is not shareable, but a readonly line is shareable i.e. may exist validly in more than one SIC in the MP. Any CP in the MP can request data in an EX state from its SIC, which data may or may not be found in its SIC or in another CP's SIC. If a CP requests a line of storage data in EX state and the line is found in EX state in another CP's SIC, it may be allowed to remain in the other CP's SIC by being set to RO state in both CPU SICs for the situations in which: (1) the line is found unchanged in EX state in the other CP's SIC, or (2) the line is found in RO state in the other CP's SIC, in which case the line is received and set to RO state in the requesting SIC even though requested in EX state. But if the line is found to be changed in the other CP's SIC, its shareability designation in the requesting SIC will be EX and the line is invalidated in the other CP's SIC from where it is castout.

Growing community of inventors

Rye, NH, United States of America

Richard N Gustafson