Documentation available @: http://www.enea.com/templates/Extension____12926.aspx
The OSE RTOS was designed from the ground up to satisfy the requierments of today and tomorrow´s complex and mission-critical embedded systems.
Today’s real-time embedded systems are addressing requirements considered beyond the cutting edge
only a few years ago, such as supporting scalability across multiple CPUs, providing nonstop service
under all conditions, and running multiple communicating processes and applications in single-core,
multi-core and distributed processing environments. These complexities can increase costs, decrease
reliability, and lengthen time to market. A conventional Real-Time Operating System (RTOS) falls
short when asked to provide a foundation for the next generation of embedded applications. Designers
of complex embedded systems need a proven RTOS with the features to support the new generation of
The OSE RTOS was designed from the ground up to satisfy the needs of today’s complex and mission-critical
embedded systems. Originally developed for the telecommunications industry as a platform for delivering critical communication services, it has been used in millions of products worldwide for over 15 years. System designers focusing on reliability, scalability, and simplicity are increasingly making OSE their platform of choice. OSE is a solid technical market leader. The OSE RTOS was adopted by Ericsson in 1988, by Nokia in 1990, and made available throughout Europe in 1991 and in North America in the 1990`s. OSE quickly became the preferred operating system for high-availability applications and distributed systems, especially in telecommunication and wireless products. Today, OSE is at the core of millions of cellular telephones, many internet infrastructure systems and other devices.
A conventional RTOS involves application software in many routine chores. The OSE RTOS makes it possible to quickly and easily create applications that operate at a higher level of abstraction and execute reliably over their lifetime. The OSE RTOS is in a class above the others.
“We wanted an operating system that was fast, robust and had strong memory management capabilities. We already knew that OSE fulfi lled all of these criteria making the RTOS a natural fi rst choice for us.” Ericsson Licensing Technology
Messages are allocated from the OSE memory pool. Memory is conserved and fragmentation is avoided. OSE manages all of the details of buffer owner-ship as messages are passed from task to task, relieving the application of this responsibility.
The OSE Architecture
OSE is a powerful platform for the design of real-time embedded systems. OSE’s message-based architecture instantly and seamlessly achieves powerful simplicity in complex, distributed systems. OSE’s reliable task management and dynamic runtime confi guration enable faster, more reliable system deployment
and maintenance. OSE’s structured, multi-level facility for error detection enables more efficient code and more reliable and consistent exception handling in the system. OSE also includes built-in monitoring of critical tasks, alerting you before a software failure brings down your entire system. OSE makes it possible to develop highly reliable applications in far less time. Its architecture is designed specifically to meet the challenges of distributed and faulttolerant system designs.
At the heart of the OSE architecture is a direct message-passing model that provides fast, asynchronous intertask communication. Message buffers are allocated from the OSE memory pool. Memory is ecologically conserved and fragmentation is avoided. OSE manages all of the details of buffer ownership as
messages are passed from task to task, relieving the application of this responsibility. As a result, applications interoperate much more intuitively and avoid many of the program errors that result with other intertask communication models. This is the OSE direct messagepassing advatage.
OSE is the Leading RTOS for Use in High-Availability Designs
OSE was designed with built-in capabilities for creating service-critical, mission-critical, and safety-critical systems. This has always been a central design feature in OSE – not an after-thought. For example, OSE has always supported essential concepts such as user-supervisor modes and hardware-enabled memory protection. Rather than executing as a single task, or multiple tasks with no inherent memory protection, OSE enables an application to consist of a number of distributed components. Each task can be a separate entity with its own protected resources controlled entirely by the OSE kernel. An application written for OSE can easily be divided into logical parts, each independently designed and coded, with communications between them transparently managed as messages pass across protection domain boundaries. The result is high application performance, while maintaining stability and
reliability across the entire system. And OSE goes beyond supporting these essential concepts with architecture-
specific mechanisms that greatly enhance system availability. The deeper you look into OSE and compare it to other offerings on the market, the more you will understand why developers choose OSE when time to market and high availability are critical.
OSE TRANSPARENTLY EXTENDS ITS MESSAGE-PASSING ACROSS CPU BOUNDARIES
OSE can put critical tasks on a watch list and automatically notify the application, should a task of interest be
created or deleted. Applications can take action before a fault becomes a fatal error.
With OSE you design at a higher level, creating applications that are easier to write, understand, and maintain.
OSE HAS BUILT-IN SUPERVISION MECHANISMS FOR MONITORING CRITICAL TASKS
Enea´s unique OSE Link Handler technology manages the transparent connection of tasks to services regardless
of their location. If one task or an entire board fails, the software can automatically reconnect clients and their services without interruption to the application.
OSE IS UNMATCHED IN ITS ABILITY TO SUPPORT DYNAMIC SOFTWARE RECONFIGURATION
Live software replacement is a reality with OSE. New versions of programs can be loaded onto the system and
clients can be redirected to the new instance at any time, without rebooting. Stopping, recompiling and reloading
are no longer necessary.
DESIGN ON A HIGHER LEVEL WITH OSE AND PROJECTS ARE COMPLETED FASTER AND WITH FEWER ERRORS
Developers who use other commercial RTOS have had to dig through piles of manuals to learn the low-level
primitives found in most RTOS. OSE on the other hand provides a compact, well-designed set of services and simple elegant APIs for those services. OSE developers find that most applications can be managed with just eight powerful
kernel APIs. With OSE you design at a higher level, creating applications that are easier to write, understand,
and maintain. This has far-reaching benefits. Programmers learn and master OSE faster than other operating
systems, make fewer mistakes, and find errors more quickly. This shortens development time and produces
code that is more readily reused. Together with OSE’s automatic error detection and built-in application-level
debugging, these features greatly enhance productivity and quality.
OSE BUILDS IN AUTOMATIC ERROR DETECTION
OSE has an advanced, built-in error detection system. Should an error be detected, OSE automatically invokes
a user-defined error handler specific to its task, block, or system scope. This simple-to-use, but powerful feature
replaces the complex code and inconsistencies that often result when programmers handle errors differently
throughout their application tasks.
WITH OSE YOU CAN DEBUG AT THE APPLICATION LEVEL, AN EASIER PLACE TO WORK
OSE’s modern message-based paradigm provides the basis for powerful application-level debugging.Occurrences such as the passing of specific message types or context switching, can be followed step by step, or traced in real-time. Breakpoints can be placed on these occurrences to stop or monitor execution. During distributed and non-stop
system debugging, selected parts of the system can be stopped without halting the entire system. This type
of debugging can also be applied to simulated targets that run entirely on a host computer, through use of the OSE Soft Kernel.
OSE messages are sent directly from one task to another. A message contains the addresses of its sender and receiver, as well as a data payload.
“We wanted to move to a more modern distributed system – without having to develop it ourselves. We also wanted the operating
system to be founded on a message communication model rather than a shared memory approach.” Lucent Technologies
The OSE Platform
The OSE real-time kernel is complimented by a rich set of operating system add-on components to form a complete fault-tolerant platform for embedded computing.
OSE LINK HANDLER
The unique OSE Link Handler can connect nodes in a distributed system, and enable transparent communication
and supervision among applications running on the nodes. With built in support for redundant communication
paths across heterogeneous transports, OSE Link Handler provides a reliable means for connecting programs with services across any computing cluster.
OSE Gateway enables direct, task-to-task messagepassing between OSE and other operating systems. OSE
Gateway supports Solaris, Windows, Linux, VxWorks, Epoc and virtually any operating system.
EMBEDDED FILE SYSTEM
The OSE Embedded File System (EFS) can be used for storing program modules and for transferring data to and from many types of storage devices. Designed for distributed computing, OSE EFS is a fully distributed file system service that supports a variety of volume managers and media types including but not limited to RAM, flash, and FAT-based
OSE, in conjunction with best in class providers of advanced networking software, maximizes flexibility and
choice for its customers. OSE collaborates with providers of IPv4/IPv6 stacks, management (SNMP, web
based, etc), routing, security and utilities telecom protocols to deliver leading edge solutions to the market.
OSE PROGRAM HANDLER
The OSE Program Handler is a powerful utility for systems requiring high availability as it can load, modify, and
remove programs during run-time. While many embedded platforms support run-time loading of new programs,
most fall short when system designs require live software replacement. With the OSE Program Handler,
new versions of software can replace old ones without interrupting the application’s execution.
OSE BOARD SUPPORT PACKAGES
The OSE Board Support Packages (BSP) provide support for the latest standard boards from many suppliers.
BSP drivers conform to the OSE device-driver specification, allowing seamless migration from board to
OSE MEMORY MANAGEMENT SYSTEM
The OSE Memory Management System (MMS) isolates and protects functional software units from one another
and also protects the kernel from application software errors. The OSE MMS design is optimized for embedded
systems with a flexible and high performance implementation.
The complete OSE solution includes a target-based (on the right) operating system architecture and complete host-based (on the left) development environment for demanding, high-availability embedded applications.
The OSE real-time operating system enables customers to differentiate themselves from their competition by using the unique features and benefits found only in OSE products.
POLYHEDRA – HIGH PERFORMANCE DATA MANAGEMENT FOR EMBEDDED SYSTEMS
Today’s embedded systems have increasingly demanding requirements of their data management components.
Data management is now a core part of a system’s infrastructure, with requirements to store and preserve
mission-critical state and application data in a fault tolerant fashion. While traditional databases have
failed to address these requirements, OSE Polyhedra database offers high-performance active-relational
technology in a small footprint package designed specifically for this purpose. Data persistence and fault tolerance mechanisms are built into the Polyhedra database, ensuring survival of the database service across failure of part of the system, and robust data recovery in the event of a restart. Polyhedra’s unique active-relational technology provides the ability to encode application logic such as data integrity rules, directly into the database; and “Active
Queries”, enabling immediate notification to applications of data change. The Polyhedra database is integrated
with the OSE RTOS, using OSE messaging for communication between applications and the database server,
and operates seamlessly regardless of the physical location of the tasks. The database also connects to other embedded systems, servers, or workstations over TCP/IP – while automatically handling heterogeneity issues, making database location transparent to the developer.
Superior Tools for Rapid Development
Providing a modern and robust architecture for today’s complex embedded systems is only half of the story.
Developing reliable applications that take advantage of the features of the OSE RTOS is faster and less errorprone
than with a traditional RTOS, thanks to a powerful chain of OSE tools. This tool chain consists of industry-standard compilers and debuggers, powerful host-target tools, and accurate simulators, all supported by a comprehensive, integrated development and debugging environment.
OSE Illuminator is a host-based suite of software tools for debugging and analyzing OSE applications. OSE Illuminator includes a system browser, an event-action analyser, and profiling tools. It can also integrate a variety of plug-in
tools that provide additional debugging capability and system information. Illuminator also communicates with target boards through standard TCP/IP communication protocols and interconnects with source-level debuggers.
OSE Soft Kernel makes it possible to simulate your application on host computers. OSE Soft Kernel can simulate
part of a system, a whole system, or even a distributed system including a mix of actual hardware boards
and multiple OSE Soft Kernels, before moving all software to its final target.
The Polyhedra database for embedded applications provides data persistence and high performance on a single system or across a distributed system.
“Speed is one of our main criteria, and OSE perfomed best in specifi cation. OSE offered a more up-to-date, streamlined architecture
than the competition, which offered operating systems based on more traditionel technology.” ABB Environmental Systems
OSE SOFT ENVIRONMENT
The OSE Soft Environment complements the OSE Soft Kernel with tools and utilities for building a complete
OSE system in a host environment. This host environment can be used to test and debug an OSE application,
or even used as part of a distributed target system. The OSE Soft Environment provides users with exceptional
support for the simulation of an OSE system. Examples of run-time accessories supported in the OSE Soft Environment are:
• OSE Link Handler
• OSE Embedded File System
• Internet Protocols
• Internet Utilities
• Web Server
OSE – A Complete Solution
The OSE RTOS provides a complete framework for the implementation of ultra-reliable and ultra-efficient realtime
systems -from resource constrained single CPU systems to large, distributed systems. OSE includes a full suite of software development tools for building and deploying complex, mission-critical embedded systems. In addition, a full suite of simulation development components and tools allow an OSE system to run in a host environment -ideal for
parallel software-hardware development. The OSE host system may also be connected to a target system for debugging, or the host system may be a part of a deployed distributed system. It is the powerful, distributed
architecture of OSE combined with its higher level, dynamic design paradigm of messagepassing and its
built in mechanisms for fault tolerance and high availability that make OSE the clear platform choice for today’s
modern embedded designs.
Processor Families Supported
ARM consortium family:
ARM 7, ARM 9, ARM10, Xscale
MIPS 32 consortium family Motorola PowerPC family:
5xx, 6xx (603, 603e, 603ev, 604, 604e),7xx (740, 745, 750, 755), 74xx, MPC 8xx, MPC82xx (824x, 826x)
IBM PowerPC family:
4xx (403, 405GP, 440GP), 7xx (750, 750CX, 750FX), NPe405H, NPe405L, PowerNP4GS3)
Intel Strong ARM family:
StrongARM, Xscale incl. PXA25x
Intel Network Processor:
Priority-based preemptive scheduling
Memory Pools for dynamic memory allocation
Asynchronous direct messagepassing
Fault and error handling
Distributed system and multi-core support
Transparent inter processor communications (IPC)
“Attach” service to ‘watch over’ tasks
Dynamic application loading
Dynamic program management for persistent media
Fast response to real-time events
Non-fragmenting, efficient management of RAM buffers of differing sizes
Easy software design concept, for complex embedded systems
Reliable and robust service, supporting well-organized software architecture.
A single operating system and a single programmers’interface (“API”) for all processors.
Transparent message communication to tasks on other processors
Operating system monitors ‘liveness’ and accessibility of critical tasks, including tasks on other processors
“Hot swapping” for high-speed in-the-fifield software upgrades
Programs can be loaded to flash, execute in place, and survive restarts.
Leverages processor MMU memory protection and virtual memory translation facilities.