In this study, we describe obtaining hard real-time performance and rich Linux features together in a compounded real-time operating system (cRTOS). This system creates two realms with a partitioning hypervisor: a normal realm of Linux and a hard real-time realm of a swift RTOS (sRTOS). A rich real-time process running in the real-time realm can use not only the hard real-time performance of the RTOS but also the rich features of Linux through remote system calls. Unlike existing approaches for real-time Linux including the PREEMPT_RT patch and using interrupt-dispatching layers, this approach requires no modifications to Linux.

Centralized social networking services (SNSs) or online social networks (OSNs) inherently have privacy concerns. Ciphertext-policy attribute-based encryption (CP-ABE) is an effective and promising tool for protecting privacy in centralized and distributed systems. However, a large effort is required to develop new practical social applications using CP-ABE. Furthermore, such applications often have performance problems because of the substantial computation time needed for ABE. This paper describes the design and implementation of the ABnews system, a Usenet-like social messaging system using ABE and untrusted storage. ABnews is fast because it eliminates the heavy computation of ABE from the interactive access to messages. ABnews inherits application programs from Usenet, and allows Usenet's rich newsreaders to be utilized without any modification. Furthermore, ABnews provides overlay social applications, including a private blogging service, a presence service, and one-to-one direct messaging, on top of bulletin board systems. The ABnews system has been implemented using the cpabe toolkit and Google Drive, which holds encrypted messages. Experimental results show that the overhead of CP-ABE is negligible when implementing such a messaging system on a current cloud storage service.

As centralized social networking services (SNSs) or online social networks (OSNs) have inherent privacy concerns, researchers are working on decentralized SNSs as alternatives for addressing these problems. However, it is very difficult to decentralize all the functions of centralized SNSs. This paper describes a decentralized SNS using social virtual private networks (VPNs) that work together with centralized SNSs. In this decentralized SNS, members can share private information with friends through the social VPNs while hiding the information from centralized SNSs. SNS members run applications and servers on their PCs, allow friends to access the contents of these servers, and have complete control of their own data. Community leader members called hub users run VPN servers and maintain group member lists. Other regular members manipulate the VPN clients easily using Web browsers. The proposed decentralized SNS federates the user authentication of the VPN servers with that of centralized SNSs. Furthermore, the SNS provides a zero sign-on feature for social Web applications running on a closed VPN. For realizing such a social VPN, the authors have extended SoftEther VPN, their fast, open-source VPN software. Experimental results show that applications including file sharing, a text chat, a Wiki program, an interactive game, and Usenet-like messaging performed well using the social VPN over a current fast Internet access line.

Kernel scripting is a technique to run an extension code in a script language in an operating system kernel. Conventional kernel scripting has two limitations. First, it affects an entire system and only privileged users are allowed to install a new script. This prohibits developers from running their own application-specific code in the kernel. Second, its performance is not sufficient for some time-sensitive applications. In this paper, we address these problems. Our system call scripting allows developers to run their own application-specific code in the kernel without the root privilege. Our system call scripting runs with less overhead because we use a Just-In-Time (JIT) compiler. To evaluate our idea, we ported the LuaJIT compiler into the FreeBSD 10.1 x86 kernel. We modified Memcached to use system call scripting that processes multiple UDP GET requests at a time. With one worker thread and under a high-load condition, we achieved a 33% reduction in the average response time and a 44% improvement in the throughput when the response value size was small.

Modern authentication mechanisms, including Shibboleth and OAuth, provide user attributes such as affiliations and e-mail addresses. Conventional collaborative methods have problems using such attributes in egress access control for the Web. This paper proposes a new collaborative method using W

Most existing collaborative applications on the Web require centralized servers for storing shared data and relaying communication messages among browsers. This paper proposes building a distributed Web browser as a platform for Web-based collaborative applications to address these problems wit

The recent interest in runtime attestation requires modeling of a program’s runtime behavior to formulate its integrity properties. In this paper, we study the possibility of employing static source code analysis to derive integrity models of a commodity operating systems kernel. We develop a precis

This paper describes how users can collaborate through sharing personal protected Web resources. Personal protected Web resources are Web pages and services that are typically password-protected. One example is a personal page on an auction site. This paper introduces capability-based access control

This paper proposes an anti-spam scheme that uses capability-based access control. In this scheme, rights to bypass spam filters are represented as capabilities, and an email message containing a valid capability bypasses the spam filter and goes straight to the receiver's inbox. As a result, the false positive problem inherent in existing spam filters is eliminated. This scheme allows a user to delegate rights to another person and is compatible with existing email systems and applications. It was implemented in an email client, Mozilla Thunderbird, along with a tool, capability basket, that provides an API for email clients and a graphical user interface (GUI) for users. This tool makes it easy for users to browse and distribute their capabilities. Capabilities are safely distributed by using an overlay network by the instant messenger Skype.

Virtualized environments are important building blocks in consolidated data centers and cloud computing. Full virtualization (FV) allows unmodified guest OSes to run on virtualization-aware microprocessors. However, the significant overhead of device emulation in FV has caused high I/O overhead. Current implementations based on paravirtualization can only reduce such overhead partially. This paper describes the Linsock approach that applies the outsourcing method to speed up I/O in FV environments by combining different guest OS and host OS. Concretely, Linsock replaces the guest Windowspsila network processing with the host Linux kernel on the same machine. Linsock has been implemented on Linux Kernel-based Virtual Machine (KVM) as the host virtual machine (VM) environment. Our measurement results with Linsock show significant performance increase of more than 300% compared with device paravirtualization in a 10 Gbps Ethernet networking environment. In addition, Linsock also yields a fourfold increase in inter-VM communication performance.

Virtual machine monitors (VMMs), including hypervisors, are a popular platform for implementing various security functionalities. However, traditional VMMs require numerous components for providing virtual hardware devices and for sharing and protecting system resources among virtual machines (VMs), enlarging the code size of and reducing the reliability of the VMMs.This paper introduces a hypervisor architecture, called parapass-through, designed to minimize the code size of hypervisors by allowing most of the I/O access from the guest operating system (OS) to pass-through the hypervisor, while the minimum access necessary to implement security functionalities is completely mediated by the hypervisor. This architecture uses device drivers of the guest OS to handle devices, thereby reducing the size of components in the hypervisor to provide virtual devices. This architecture also allows to run only single VM on it, eliminating the components for sharing and protecting system resources among VMs.We implemented a hypervisor called BitVisor and a parapass-through driver for enforcing storage encryption of ATA devices based on the parapass-through architecture. The experimental result reveals that the hypervisor and ATA driver require approximately 20 kilo lines of code (KLOC) and 1.4 KLOC respectively.

Portable (standards-compliant) systems software is usually associated with unavoidable overhead from the standards-prescribed interface. For example, consider the POSIX Threads standard facility for using thread-specific data (TSD) to implement multithreaded code. The first TSD reference must be preceded by pthread/spl I.bar/getspecific( ), typically implemented as a function or macro with 40-50 instructions. This paper proposes a method that uses the runtime specialization'facility of the Tempo program specializer to convert such unavoidable source code into simple memory references of one or two instructions for execution. Consequently, the source code remains standard compliant and the executed code's performance is similar to direct global variable access. Measurements show significant performance gains over a range of code sizes. A random number generator (10 lines of C) shows a speedup of 4.8 times on a SPARC and 2.2 times on a Pentium. A time converter (2,800 lines) was sped up by 14 and 22 percent, respectively, and a parallel genetic algorithm system (14,000 lines) was sped up by 13 and 5 percent.

To install security modules or reference monitors into operating system kernels is a common and effective way for enhancing access control for networks. However, security modules in conventional kernel-level reference monitors are usually not portable to other kernels and require detailed knowledge about kernel internals. Furthermore, different security modules are often not composable and conflict with each other. This paper describes a reference monitor called SysGuard that addresses these problems. SysGuard uses modules called guards that are invoked before or after the execution of system calls. Unlike kernel-specific security modules, guards are attached to standard system calls that enhance their portability. The guard scoping on a per-process basis improves composability of individual guards, and it is implemented efficiently by using a per-process jump table of system calls. This paper describes the implementation of restricted execution environments for networks by composing simple and portable guards, and shows the advantages of the SysGuard security framework.

This paper presents several methods for harmonizing a distributed operating system with parallel and distributed applications. 'Harmonization' means that the system does not conflict with the parallel/distributed applications and enables them to control mapping and scheduling for speedup and reliability. In conventional distributed operating systems, one of the most important goals is to achieve network transparency. However, parallel/distributed applications have to bypass the facilities for network transparency because they require information on the locations and usage of system resources. Harmonization between the system and parallel/distributed applications are realized by putting some parts of the system into the same position as the applications.>