Parallel Port Data Classification for Multi-Processor Deadlock Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In servers with multiple processors, the shared common parallel port can lead to deadlocks due to exclusive control issues, increasing production costs, especially when interruption request registers are required for each CPU, as seen in existing service request control methods.
Innovation Solution
An information processing system where each processor appends a processor identifier to transfer codes and writes them in a common parallel port without exclusive control, allowing a peripheral device with data classification and computing means to read and reconstruct the data, eliminating the need for shared interruption request registers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a common parallel port is shared by multiple processors, then production cost is reduced, but deadlock may occur due to exclusive control issues
Solution Approach 1:
The patent segments the data transfer process by assigning unique processor identifiers to each processor's data packets. This segmentation allows the peripheral device to distinguish and separately process data from different processors, eliminating the need for exclusive control mechanisms and preventing deadlocks while maintaining cost-effective shared port usage
Solution Approach 2:
The patent introduces processor identifiers as an intermediary element that mediates access to the common parallel port. These identifiers enable the peripheral device to act as an intelligent mediator that routes data from multiple processors without requiring traditional interrupt request registers or exclusive control protocols, thus avoiding deadlocks
2Ease of operation
If interruption request registers are provided for each processor, then data transfer control is improved, but production cost increases
Solution Approach 1:
The patent makes the common parallel port universal by enabling it to handle data from multiple processors simultaneously through identifier-based routing. The peripheral device's data classification unit serves multiple functions: receiving data, identifying the source processor, and routing to the appropriate processing unit, eliminating the need for dedicated interrupt request registers for each processor
Data Source
AI summary
An information processing system equipped with an information processing device, which includes multiple processors and a common parallel port (port), and a peripheral device, which includes a data classification means and multiple computing means. Each processor appends each processor identifier to a respective series of codes capable of expressing given data, thereby generating a writing unit capable of being written one time in the bus width of the port, and sequentially writes the writing unit to the port without performing exclusive control. A data classification means reads the writing units from the port sequentially and sequentially outputs the series of codes included in the writing unit to a computing means associated with the processor identifiers included in the writing unit. The computing means reconstructs the original data on the basis of the series of codes.