摘要:
In a lock-step synchronism fault-tolerant computer system including a plurality of computing modules having a processor and a memory in which each computing module processes the same instruction string in synchronization with each other. When detecting disagreement in a state of access to an external bus among the respective processors in each computing module, if no fault is detected in the system including each computing module, processing of resuming operation in synchronization is executed with respect to each computing module after generating an interruption to all the processors to execute delay adjustment for making a state of instruction execution among computing modules be coincident.
摘要:
An information processing apparatus of the present invention includes first and second computer elements which execute the same instructions substantially simultaneously and which are substantially synchronized with each other. The first computer element includes first and second memory elements, which are written by the first and second computer elements, respectively, during a first state. The information processing apparatus has a control element which makes the first computer element read from the second memory element during a second state. Another information processing apparatus has the first and second computer elements, and first and second memory areas which are provided in the first computer element. The first and second memory areas are written by the first computer element and the second computer element, respectively, during a first state. A control element makes the first computer element read from the second memory area during a second state.
摘要:
A lock-step synchronism fault-tolerant computer system includes a plurality of computing modules having a processor and a memory in which each computing module processes the same instruction string in synchronization with each other. When disagreement in a state of access to an external bus among the respective processors in each computing module is detected, if no fault is detected in the system including the respective computing modules, an interruption is notified to all of said processors. Synchronization among each computing module is recovered by adjusting timing of a response to an access which each processor executes by an interruption.
摘要:
Cache control protocols can be switched during running without changing an architecture for a segment descriptor or page descriptor for indicating an attribute of an area to be accessed. A plurality of processors each including a cache memory constitute a multiprocessor system which shares a main memory via a system bus. Each processor has module detecting means for detecting execution of a module which accesses a shared memory area on the main memory, by comparing the virtual space number and the instruction segment number concerning the accessing module with those numbers concerning the software modules preset which may access the shared memory area. Memory access executed in a module detected by the module detecting means is controlled in a cache control protocol of a store-through scheme which updates a main memory simultaneously with update of a cache memory. Memory access executed in other modules is controlled in a cache control protocol of a store-in scheme which does not update a main memory at update of a cache memory.