Data Processor Dedicated Peripheral Control Bypassing Hypervisor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The overhead associated with hypervisors in data processors is significant due to the extra processing time required for peripheral device control, as the hypervisor intervenes in access arbitration between operating systems and hardware resources, leading to inefficiencies.
Innovation Solution
Classifying peripheral devices into dedicated and shared categories, where dedicated devices are controlled solely by a specific operating system without hypervisor intervention, and shared devices are managed through the hypervisor, using registers to set control modes and allocate resources efficiently, thereby reducing hypervisor overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the hypervisor intervenes in access arbitration between operating systems and hardware resources, then the reliability and security of each operating system are maintained, but the processing time increases due to extra hypervisor intervention steps
Solution Approach 1:
The patent segments peripheral devices into two categories: shared peripheral devices (requiring hypervisor arbitration) and dedicated peripheral devices (accessible directly by a specific OS). This segmentation allows the system to maintain security for shared devices while eliminating unnecessary hypervisor intervention for dedicated devices, thus reducing processing time without compromising reliability.
Solution Approach 2:
The patent applies different control qualities to different peripheral devices based on their access patterns. For dedicated peripheral devices, the control quality is optimized for direct OS access without hypervisor intervention. For shared peripheral devices, the control quality maintains hypervisor-mediated arbitration. This local differentiation resolves the contradiction by applying the appropriate level of intervention based on device characteristics.
2Adaptability or versatility
If the hypervisor performs arbitration for all peripheral device accesses, then resource sharing between operating systems is ensured, but the productivity decreases due to repeated hypervisor involvement
Solution Approach 1:
The patent divides peripheral devices into shared and dedicated categories, allowing the system to maintain resource sharing adaptability for shared devices while improving productivity for dedicated devices by eliminating redundant hypervisor arbitration steps.
Solution Approach 2:
The patent applies hypervisor arbitration only partially - specifically for shared peripheral devices where it is necessary. For dedicated peripheral devices, the hypervisor arbitration is omitted entirely, avoiding excessive action that would reduce productivity without providing additional adaptability benefits.
3Reliability
If the processor generates an exception and transfers control to the hypervisor for every peripheral device access, then secure arbitration is achieved, but the number of processing steps increases
Solution Approach 1:
The patent segments the exception handling mechanism based on device type. For shared peripheral devices, the full exception and hypervisor arbitration process is maintained to ensure security. For dedicated peripheral devices, the exception generation and hypervisor transfer steps are eliminated, reducing device complexity while maintaining security where needed.
Solution Approach 2:
The patent extracts the exception generation and hypervisor transfer steps from the peripheral device access path for dedicated devices. By removing these unnecessary steps for dedicated devices, the system reduces processing complexity while maintaining the security mechanism for shared devices where it remains essential.
Data Source
AI summary
This invention is intended to reduce the hypervisor overhead. In the data processor disclosed herein, when a device driver calls for access to a control register to activate a process of a dedicated controlled peripheral device, the access is handled directly without intervention of processing by the hypervisor. When an interrupt is generated from a dedicated controlled peripheral device, a process is directly initiated by the device driver of the operating system governing the peripheral device without intervention of processing by the hypervisor. By implementing this manner of control in the processor, it becomes possible to carry out peripheral device control without intervention of processing by the hypervisor. Thereby, the hypervisor overhead is alleviated.