OS Switching for Real-Time Interrupt Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional technologies face difficulties in efficiently securing real-time processing when real-time tasks are distributed across multiple operating systems.
Innovation Solution
An information processing apparatus with a reception unit and switching unit that switches between operating systems based on priority, using a virtual machine monitor to manage interrupts and task processing, ensuring that real-time tasks are assigned to high-priority cores and maintaining efficient real-time processing without relying on hypervisor affinity information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If real-time tasks are distributed across multiple operating systems using conventional virtualization technologies, then operating system diversity and task distribution are improved, but real-time processing efficiency deteriorates
Solution Approach 1:
The patent segments the system into a host OS and multiple guest OSes, with the host OS containing a real-time kernel that directly manages CPU cores. This segmentation allows guest OSes to run on non-real-time cores while real-time tasks are isolated and executed on dedicated real-time cores with direct access to hardware, eliminating the need for real-time tasks to traverse the hypervisor layer and ensuring deterministic real-time processing regardless of OS distribution.
Solution Approach 2:
The patent introduces a real-time kernel in the host OS as an intermediary layer between the hardware and guest OSes. This real-time kernel directly manages real-time CPU cores and handles real-time task scheduling without requiring intervention from the hypervisor or guest OS schedulers, thereby ensuring that real-time processing efficiency is maintained even when tasks are distributed across multiple operating systems.
2Ease of operation
If hypervisor affinity information is used to manage task distribution, then task placement control is improved, but system complexity and configuration effort increase
Solution Approach 1:
The patent extracts the real-time task management functionality from the hypervisor and guest OS schedulers and places it directly in the host OS real-time kernel. This extraction eliminates the need for complex hypervisor affinity information and cross-OS task coordination mechanisms, simplifying the system architecture while maintaining precise control over real-time task placement on dedicated real-time cores.
Solution Approach 2:
The real-time kernel in the host OS performs self-service by directly scheduling and managing real-time tasks on real-time cores without requiring external coordination from hypervisors or guest OSes. This self-service approach simplifies the system by eliminating complex inter-OS communication and affinity management, reducing configuration effort while maintaining effective task placement control.
Data Source
AI summary
An information processing apparatus according to an embodiment includes a reception unit and switching unit. The reception unit receives an interrupt. The switching unit that switches a second operating system (OS) which is executing in a core to a first OS to which the interrupt for the first OS is input, when the reception unit receives an interrupt for the core in which the first OS is a priority OS and the second OS is not the priority OS.