• Journal of the Korea Society of Computer and Information
• /
• v.22 no.5
• /
• pp.11-19
• /
• 2017

In this paper, we propose a hypervisor for embedded systems based on ARM microprocessor. The proposed hypervisor makes it possible to run several real-time kernels concurrently on a single embedded system by virtualizing its microprocessor. With assistance of MMU, it supports virtual memory which enables each guest operating system has its own address space. Exploiting the fact that most embedded systems use memory-mapped I/O device, it provides a mechanism to distribute an external interrupt to virtual machines properly. It also achieves load balancing through live migration which moves a running virtual machine to other embedded system. Unlike other para-virtualization techniques, minor modifications are needed to run it on the hypervisor. Extensive performance measurement studies are conducted to show that the proposed hypervisor has enough potentiality of its real-world application.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.6 no.4
• /
• pp.231-241
• /
• 2011

Recently, system virtualization has been applied to consumer electronics (CE) such as smart mobile phones. Although multi-core processors have become a viable solution for complex applications of consumer electronics, the issue of utilizing multi-core resources in the virtualization layer has not been researched sufficiently. In this paper, we present a new hypervisor design and implementation for multi-core CE devices. We concretely describe virtualization methods for a multi-core processor and multi-core-related subsystems. We also analyze bottlenecks of network performance in a virtualization environment that supports multimedia applications and propose an efficient virtual interrupt distributor. Our new multi-core hypervisor improves network performance by 5.5 times as compared to a hypervisor without the virtual interrupt distributor.

• KIISE Transactions on Computing Practices
• /
• v.23 no.4
• /
• pp.199-209
• /
• 2017

Recently, there has been growing interest in workload consolidation via virtualization for real-time systems. Meanwhile, due to virtualization, additional overhead from intervention of hypervisor increases the execution time of applications running on virtual machine. The increase in execution time decreases the performance of workloads on virtual machines, thus satisfying real-time requirements are not easy. To resolve the problem, we designed and implemented a hypervisor (the K-Hypervisor) that allows programs on virtual machines to access the physical resources, without intervention of the hypervisor. Experimental results show that programs running on K-Hypervisor increase their execution time by about 3% on an average, compared to the native environment. Moreover, it is suitable for real-time workload execution because of its uniform performance degradation, regardless of the resources accessed from tasks, and the frequency it is accessed.

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.5
• /
• pp.103-112
• /
• 2007

In this paper, we implement a hypervisor that runs multiple uC/OS-II real-time kernels on one microprocessor. The hypervisor virtualizes microprocessor and memory that are main resources managed by uC/OS-II kernel. Microprocessor is virtualized by controlling interrupts that uC/OS-II real-time kernel handles and memory is virtualized by partitioning physical memory. The hypervisor consists of three components: interrupt control routines that virtualize timer interrupt and software interrupt, a startup code that initializes the hypervisor and uC/OS-II kernels, and an API that provides communication between two kernels. The original uC/OS-II kernel needs to be modified slightly in source-code level to run on the hypervisor. We performed a real-time test and an independent computation test on Jupiter 32-bit EISC microprocessor and showed that the virtualized kernels run without problem. The result of our research can reduce the hardware cost, the system space and weight, and system power consumption when the hypervisor is applied in embedded applications that require many embedded microprocessors.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.7 no.4
• /
• pp.193-199
• /
• 2012

Virtualization allows multiple commodity operating systems to share on a single physical machine. Resource allocation among virtual machines is a key to determine virtual machine performance. To satisfy time-sensitive task on a domain, hypervisor needs to observe the resource requirements and allocates proper amount of CPU resources in a timely manner. In this paper, we propose a realtime credit scheduler for latency sensitive application on virtual machines. The key idea is to register a time event in the Xen hypervisor. Experiment result shows that the proposed scheme is superior to Credit scheduler.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.10 no.4
• /
• pp.233-239
• /
• 2015

This paper describes a method that is integrated trace for real-time virtualization environment. This method has solved the problem that the performance trace may not be able to analyze integrated method between heterogeneous operating systems which is consists of real-time operating systems and general-purpose operating system. In order to solve this problem, we have attempted to reuse the performance analysis function in general-purpose operating system, thereby real-time operating systems can be analyzed along with general-operating system. Furthermore, we have implemented a prototype based on ARM Cortex-A15 dual-core processor. By using this integrated trace method, real-time system developers can be improved productivity and reliability of results on real-time virtualization environment.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.4
• /
• pp.177-187
• /
• 2020

Virtualization with hypervisors is one of emerging topics in multicore processors for space. Hypervisors are software layers to make several independent virtualized environments on one processor. Since all hardware resources are virtualized and distributed only by hypervisors, overall performance of processors can be improved by fully utilizing the resources. However at the same time, there are overheads for virtualizing and distributing hardware resources. Satellites are one of hard real time systems, and performance degradation with overheads should be analyzed thoroughly. Previous research on the overheads focused on single core systems. Even the overheads were analyzed in multicore systems, SMP environment was not fully included. This paper builds SMP environment with XtratuM, one of hypervisors for space missions, and analyzes performance degradation with overheads. Two boards of GR712RC with 2 LEON3FT CPUs and GR740 with 4 LEON4 CPUs are used in experiments. On each board, SMP benchmark functions are executed on SMP environment with XtratuM and on that without XtratuM respectively. Results are analyzed to find timing characteristics including overheads. Finally, applicability of the XtratuM to flight software in SMP is also reviewed.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06b
• /
• pp.381-384
• /
• 2011

오늘날의 높은 효율성을 가진 현대적인 가상화 솔루션의 가용성으로 인해, 서버와 클라우딩 컴퓨팅 분야와 같은 다양한 분야에서 주목받고 있다. Xen은 다양한 가상화 시스템 중 폭 넓게 사용되는 기술이기는 하나 스케줄링에 있어 약간의 약점이 노출되고 있다. 이는 Xen 하이퍼바이져 스케줄러가 유저 레벨의 쓰레드와 물리적 CPU 사이의 스케줄러들 가운데 가장 아래에 존재하기 때문에, 유저 레벌의 워크로드에 대한 정보를 얻는데 제약사항이 있기 때문이다. 이러한 특성은 시스템의 전체적인 처리량을 떨어뜨리고 리소스의 활용률을 저하시킬 수 있다. 본 연구는 게스트 운영체제에서 동작하는 동적인 워크로드에 대한 정보를 활용하는 유저 레벨의 스케줄링을 제안하고자 한다. 이를 위해 새로운 하이퍼콜과 모니터링 데몬을 가지는 유저레벨 스케줄링을 구현한다. 유저레벨 스케줄링이 동작하는 Xen 하이퍼바이져 기반에 Linux를 게스트 운영체제로 환경을 구축하여 다양한 유저레벨 워크로드를 동작시켜 시스템의 처리량 증가와 CPU 리소스의 활용률을 높일 수 있음을 보인다.

• Journal of KIISE:Computer Systems and Theory
• /
• v.37 no.1
• /
• pp.35-42
• /
• 2010

Guest operating systems running over the virtual machines share a variety of resources. Since CPU is allocated in a time division manner it consequently leads them to having the unknown physical time. It is not regarded as a serious problem in the server virtualization fields. However, it becomes critical in embedded systems because it prevents guest OS from executing real time tasks when it does not occupy CPU. In this paper we propose a hypercall to register a timer service to notify the timer request related real time. It enables hypervisor to schedule a virtual machine which has real time tasks to execute, and allows guest OS to take CPU on time to support real time. The following experiment shows its implementation on Xen-Arm and para-virtualized Linux. We also analyze the real time performance with response time of test application and frames per second of Mplayer.