Glossary

acceptance

Acceptance is the acknowledgement by a certification authority that the ARINC 653 module, application, or system meets its defined requirements.

ACE

ACE: Agent for the Certified Environment.

AFDX

AFDX: Avionics Full Duplex Switched Ethernet. It is defined by the ARINC 664 specification, Part 7.

alarm

In the context of health monitoring, an alarm is an event. See also message.

AMIO

Applications multiplexed I/O (AMIO) allows you to provide input to and view output from multiple partitions over a single serial connection.

APEX

APEX: Application/Executive. The general-purpose interface between an OS and application software, specified by the ARINC 653 specification. The specification includes the list of services that lets the application control scheduling, communication, and status information of its internal processing elements.

APEX port

APEX port: see port.

API

API: application programming interface.

application

An application is a collection of software components that together perform a specific function in an embedded system. See also application partition.

application developer

An application developer develops one or more applications that will reside in a partition. This person or group may also be responsible for developing data binaries, which contain any databases used by the application. See also platform provider and system integrator.

application partition

An application partition is a partition that includes an application.

APPS

APPS: ARINC PPS. It is the module-wide scheduling scheme for partitions. This is a combination of ARINC 653 scheduling (TPS) and PPS scheduling in which the PPS scheme is used during idle time within the TPS scheme. The scheduling scheme applies to all PPS-enabled partitions in the module.

ARINC 653

ARINC 653 refers to ARINC Specification 653: the “Avionics Application Software Standard Interface.”

ARINC 653 scheduling

ARINC 653 scheduling is the scheduling that is specified by the ARINC 653 specification. It is time-preemptive scheduling (TPS). See also APPS scheduling and PPS scheduling.

ARINC PPS

ARINC PPS: see APPS scheduling.

black box

A black box is a set of configuration parameters that represent the memory requirements of an application, a shared library, or the core OS. The use of black boxes allows a VxWorks 653 module to be configured before all the applications and libraries are available. Applications, libraries, and the core OS must fit within the memory limits set by their black boxes.

board support package

BSP: board support package. It provides the libraries required to support a platform on a particular board. The BSP, along with the kernel and user-supplied extensions, makes up the core OS.

BSP

BSP: see board support package.

BSP developer

A BSP developer is a person or organization responsible for the development of a board support package.

BSS

BSS: block started by symbol. It is a data section in an ELF file that contains uninitialized global and static variables that are zeroed.

build spec

A build spec specifies compiler and linker options to produce particular output, such as cert, debug, and release.

callback routine

In the context of health monitoring, a callback routine is called when an event arrives at a partition health monitor task or module health monitor task. It is called before the handler for the given event is called.

CDF

CDF: component description file. It has the .cdf extension. It uses the component description language (CDL) to name and give values to the parameters of VxWorks 653 components.

cert

cert is the build spec that produces a certifiable image.

certifiable

An image that is certifiable can be certified to a specific level of the DO-178B avionics software standard.

certifiable subset

A certifiable subset is a subset of the core OS or a partition OS that can be certifiable to Level A of the DO-178B avionics software standard.

certification

Certification refers to certification to a specific level of the DO-178B avionics software standard.

channel

A channel defines a logical link between one source port and one or more destination ports. It also defines the message transfer mode and the characteristics of the messages. Channels are used for inter-partition communication, which can be between local partitions and/or pseudo-partitions. Channels conform to the ARINC 653 specification.

COIL

COIL: core OS interface library. A partition OS that provides a library of routines independent of the vThreads partition OS. The library supports the management of interrupts and exceptions, device I/O, interpartition messaging, and injection of health monitoring events.

COIL partition

A COIL partition is a partition whose partition OS is based on COIL. See also vThreads partition.

cold restart

A cold restart occurs when a module or partition is restarted and all data is reloaded. A cold restart takes longer than a warm restart.

configlette

A configlette is a component or part of a component that is distributed in source form, allowing compile time parameters to be set when the component is included in a build.

configuration parameter

A configuration parameter is used to change the configuration of VxWorks 653 component.

configuration record

A configuration record is a record of the information that makes up the configuration of a VxWorks 653 module or a part of it. Configuration records include both the system configuration record and user configuration records.

core OS

The core OS is the core operating system for a VxWorks 653 module. It provides fundamental operating system services and schedules partitions.

core OS interface library

Core OS interface library: see COIL.

CPU page size

The CPU page size is the smallest addressable unit of memory for the MMU. It is also called MMU page size. The page size depends on the CPU and is generally not configurable.

cross-development tools

Cross-development tools are programs that run on a host computer (running, for example, Windows or UNIX) and that are used to develop, debug, or control software running on an embedded processor, which is running a real-time operating system (for example, VxWorks 653). For VxWorks 653, the cross-development tools are based on Workbench. See also run-time software.

current partition

The current partition is the partition that is running. In an APPS scheduling environment, the current partition and the TPS partition may not be the same.

default schedule

The schedule that will be run when the module is booted.

destination port

A destination port is one of possibly many ports at the receiving end of a channel. See also source port.

direct-access port

A direct-access port is a type of pseudo-port which does not use software buffering. Buffering support is assumed to be provided by the communications hardware.

DO-178B

DO-178B: “Software Considerations in Airborne Systems and Equipment Certification.” The avionics software standard developed by RTCA.

domain

A domain is a software container. Each element of a VxWorks 653 module—the core OS (kernel), partitions (applications), shared libraries, system shared libraries, and shared data regions—exists in a domain.

dynamic memory allocation

Dynamic memory allocation refers to allocating memory from the heap at runtime.

EABI

EABI: Embedded Application Binary Interface.

ELF

ELF: Executable and Linking Format. It is an object module format used to encapsulate compiled software.

error handler process

See process health monitor.

event

In the context of health monitoring, an event is the base unit that is injected into the event handling framework. It could represent an alarm or a message, depending on the event code.

event code number

In the context of health monitoring, an event code number is the value of the event code, as defined in the HM_CODE enumeration type in hmTypes.h.

event queue

The module health monitor table and partition health monitor table each have an event queue. The module and partition health monitor event queues are sometimes called, simply, the module and partition health monitor queues. An event queue holds the events that have been dispatched to its associated health monitor for handling. Event queues are serviced before health monitor notification queues are serviced.

FAA

FAA: U.S. Federal Aviation Administration.

FIFO

FIFO: first-in, first-out queuing.

global file descriptor

Global file descriptors (standard in, standard out, and standard error) are available to all tasks in a partition. Their global assignment is controlled by the ioGlobalStdSet( ) and ioGlobalStdGet( ) routines, but may be overridden by the ioTaskStdSet( ) and ioTaskStdGet( ) routines.

GUI

GUI: graphical user interface.

health monitor

Health monitoring provides a framework to raise and handle events (which can be alarms or messages) in a VxWorks 653 module. Alarms are injected to represent faults, and handlers provide the opportunity to perform recovery actions. See module health monitor, partition health monitor, process health monitor, and system health monitor.

hosted function supplier

Hosted function supplier: see application developer.

IDE

IDE: integrated development environment.

injection

Injection is the act of creating a health monitor alarm event or message event.

interface subset

An interface subset defines part of the interface of a shared library. The use of interface subsets allows you to reuse parts of the interface definition among libraries that share some parts of their interface. For example, two different vThreads libraries containing different components would share the core vThreads interface.

interrupt level

Saying an event is injected at an interrupt level means the event is injected from an interrupt execution context.

ISR

ISR: interrupt service routine.

jitter

Jitter is a variation or deviation in the frequency of an expected occurrence. See also partition switch jitter.

kernel

Kernel is another term for the core OS.

kernel I/O region

A kernel I/O region is a region of target memory that corresponds to the address of an I/O device on the target and can be accessed only by the core OS.

Level A

Level A is the highest certification level for the DO-178B software standard.

loadable shared data region

A loadable shared data region is a data source, such as a database, that can be loaded into a shared data region as part of the module payload.

local partition

A local partition is a partition that is local to a VxWorks 653 module. Unless it might be confused with a pseudo-partition, it is called, simply, a partition.

local port

A local port is a port that is attached to a local partition. Unless it might be confused with a pseudo-port, it is called, simply, a port. See also null port.

log queue

The module health monitor and partition health monitor each have a log queue (sometimes called simply a log). Health monitor messages are always logged, whereas alarms are logged only if health monitor logging is enabled. If an event is injected from within a partition (HM_PROCESS_MODE or HM_PARTITION_MODE), the event is logged to the partition health monitor log. If the event is injected from outside the partition (HM_MODULE_MODE), the event is logged to the module health monitor log.

major frame

Each schedule consists of a major frame, which is divided into a series of variable-length minor frames.

message

In the context of health monitoring, a message is an event. See also alarm.

minor frame

Each schedule consists of a major frame, which is divided into a series of variable-length minor frames. Each minor frame defines the partition to run, its allowed duration, and whether or not the minor frame is a release point.

MMU

MMU: memory management unit.

module

A module is the “system” controlled by one RTOS, and in VxWorks 653, that RTOS is the core OS.

module health monitor

The module health monitor is present in parallel with all partitions in a VxWorks 653 module, and hence all partition health monitors in the module. The module health monitor is not part of any partition window and has priority over all partitions. The module health monitor resides in the core OS. It is associated with the module health monitor table, which among other things, defines notification queues, a log queue, and an event queue. See also system health monitor, partition health monitor, and process health monitor.

namespace

An XML namespace provides a unique identifier which can be associated with an XML element by means of a prefix. The namespace uniquely identifies the XML schema in which the element is defined.

NMI

NMI: non-maskable interrupt.

normal mode

Normal mode is the partition mode during which processes/threads are scheduled. (Other partition modes include idle, cold start, and warm start.)

notification

In the health monitoring context, notification is the act of informing another partition health monitor or the module health monitor of an event that has occurred in a given partition.

notification queue

The module health monitor table and partition health monitor table each have notification queues, one for each partition that wants to accept notification of events. Notification queues are serviced after health monitor event queues are serviced.

null port

A null port is a port that is created at system initialization time, but is not used. It is always considered to be empty when read from and have space when written to. A null port can be attached to a partition or the core OS of a VxWorks 653 module or to a pseudo-partition. See also local port and pseudo-port.

NVM

NVM: non-volatile memory.

online-loaded partition

With online-loaded partitions, the core OS does not install the partition code from flash or RAM into its final domain location in RAM as it does during the system initialization phase for regular partitions. Instead, an empty application domain is created for an online-loaded partition during the core OS initialization phase. The code of the online-loaded partition is made available to the core OS only at a later stage. In some cases this may not be until after all the regular partitions are already running.

OS

OS: operating system.

partition

A partition is a container for an application. An application running in a partition cannot interfere with applications in other partitions or with the core OS.

partition direct-access port

A partition direct-access port is a type of direct-access port residing in a partition. A partition direct access port can communicate only with a local port in the application resident in the partition.

partition health monitor

The partition health monitor is the health monitor that is present in parallel with vThreads to handle vThreads partition errors and events that may affect the operation of vThreads within the partition. The partition health monitor is scheduled as part of the partition window. It is associated with the partition health monitor table, which among other things, defines notification queues, a log queue, and an event queue. See also system health monitor, module health monitor, and process health monitor.

partition OS

A partition OS is a user-level software library running within a partition that provides operating system services to the partition. See also vThreads and COIL.

partition OS scheduler

The partition OS scheduler is the scheduler in a partition OS that allocates CPU time to threads in the partition. The partition OS scheduler in a vThreads partition is a priority-preemptive scheduler and is not related to the ARINC schedule.

partition port

Partition port: see local port.

partition scheduler

The partition scheduler is the scheduler in the core OS that allocates CPU time to partitions, allowing CPU time to become available to threads in those partitions. By default, the partition scheduler uses ARINC 653 (TPS) scheduling, but can optionally schedule designated partitions with APPS scheduling. See also partition OS scheduler.

partition switch jitter

Partition switch jitter is a variation or deviation in the configured partition switching schedule. For example, partition switch jitter might be caused by hardware latencies or when the core OS locks interrupts.

partition window

A partition window is the time in which a partition is allowed to run before being scheduled out.

payload

A payload is an image file (or files) that contains the code for a VxWorks 653 module in a form that is suitable for running on a target.

payload region

A payload region is the region of RAM or ROM where a payload is loaded.

periodic process

A periodic process is a process within a partition that is run on a schedule based on the passage of wall clock time (that is, the countdown to the next invocation of periodic process runs even when the partition itself is not scheduled).

PersistentBSS

A BSS section that is persistent across a warm restart.

platform

A platform is software on which applications can be built and from which a VxWorks 653 module can be developed.

platform provider

A platform provider is responsible for configuring the base system on which application developers will build their applications.

port

A port is one end of a channel, which is used for inter-partition communication. Ports have attributes, for example, direction (source or destination), mode (queuing or sampling), protocol (receiver discard, sender block, or none), and refresh rate. Ports conform to the ARINC 653 specification and its APEX interface and are also called APEX ports. See also pseudo-port.

POS

POS: See Partition Operating System.

POSIX

POSIX: Portable Operating Systems Interface. In this documentation, POSIX refers to the standard for real-time extensions (1003.1b), which specifies a set of interfaces to OS facilities. The POSIX API can be included in a vThreads partition if the APEX API is not included.

PPS

PPS: priority-preemptive scheduling. It allows for scheduling of partitions in a module-wide priority-preemptive scheme during the idle time within an ARINC 653 (TPS) schedule. See also APPS scheduling.

PPS-enabled

A PPS-enabled partition is a partition that is configured to indicate that it should be considered during APPS scheduling.

preemption locking

Preemption locking disables the scheduling of processes/threads/tasks, and only the current process/thread/task can be run until it decrements the lock level back to zero.

priority-preemptive scheduling

Priority-preemptive scheduling: see PPS.

process

Process is the APEX term for a thread. In the vThreads context, the term thread is preferred. See also task.

process health monitor

The process health monitor is the health monitor that is present within vThreads to handle process-related errors and events. It is also known as the error handler process. See also system health monitor, module health monitor, and partition health monitor.

pseudo-partition

A pseudo-partition is a communications object that is outside a VxWorks 653 module. See also local partition and pseudo-port.

pseudo-port

The term pseudo-port applies generally to any port that represents a data source or destination outside the current module. The term pseudo-port is also used in a more restrictive sense for a type of pseudo-port that uses software buffering. In this sense it is contrasted with direct-access port which is a type of pseudo-port that does not use software buffering. See also local port and null port.

queuing port

A queuing port is a port in queuing mode. In queuing ports, messages are queued. A protocol is required to manage the queues. See also sampling port.

RAM

RAM: random access memory.

RAM payload

A RAM payload is a payload that is designed to be downloaded into RAM on the target.

real-world time

Real-world time: see wall clock time.

receiver discard protocol

Receiver discard protocol is a port message protocol. If one of the channel’s destination ports is full, the source port discards the message for that port. Therefore, if all the destination ports are full, the message might be lost. When a message is so discarded, the port’s overflow flag is set to notify the application of the discarded (lost) message. See also sender block protocol.

refresh rate

The refresh rate (in seconds) indicates the maximum acceptable age of a valid message, from the time it was received by the port. It applies to destination sampling ports only.

release point

A release point is a way to synchronize a periodic process with the partition window of a partition. A periodic process spawned in a partition will be started only at the next release point.

ROM

ROM: read-only memory.

ROM payload

A ROM payload is a payload that is designed to be installed in ROM on the target.

root element

The root element is the element of an XML document that contains all the other elements in the document.

RTCA

RTCA: Radio Technical Commission for Aeronautics. The private, not-for-profit corporation that develops recommendations on communications, navigation, surveillance, and air-traffic management issues. RTCA developed the DO-178B avionics software standard.

RTOS

RTOS: real-time operating system.

run-time software

Run-time software is the operating system and application software that together run on a target. See also cross-development tools.

sampling port

A sampling port is a port in sampling mode. In sampling ports, messages are not queued. A message remains in the source port until it is sent or overwritten. Each new message overwrites the previous one when it reaches the destination port and remains there until it is overwritten itself. Sampling ports have refresh rates. See also queuing port.

SAP port

A service access point (SAP) is a special kind of queuing port. It is different from a normal queuing port because it allows access to addressing information when sending and receiving messages. The SAP services are similar to the ARINC 653 queuing port services but will have additional parameters to support address information. ARINC 653 Part 2, Supplement 2, defines two types of SAP services. Standard SAP services provide limited addressing capability to ensure that the source cannot alter its identity and that the destination is unambiguous. Extended SAP services provide complete accessibility to addressing. VxWorks 653 supports standard SAP ports, but not extended SAP ports.

schedule

Schedules define how the core OS schedules partitions. Each schedule consists of a major frame.

scheduler

See partition scheduler and partition OS scheduler.

select operation

The select operation refers to calling select( ) to pend on a set of file descriptors.

sender block protocol

Sender block protocol is a port message protocol. A queuing message is sent to all the channel’s destination ports. If any one is full, the message is queued in the source port in FIFO order. When the source port is full and if a timeout was specified, sender processes are blocked during the SEND_QUEUING_MESSAGE service. When a destination port is emptied, retransmission is attempted. Whether it succeeds depends on the state of the channel’s other destination ports. See also receiver discard protocol.

service access point

Service access point: see SAP port.

shared data region

A shared data region (sometimes called a shared data domain) is a data region that can be used by applications within partitions to share data. Outside a shared data region, applications have no access to the data of other applications. See also loadable shared data region.

shared I/O region

A shared I/O region is a region of target memory that corresponds to the address of an I/O device on the target and can be shared by partitions and the core OS.

shared library

A shared library is a library that contains code that can be shared by multiple applications. See also system shared library.

shared library region

A shared library region is the area of RAM that holds a shared library.

source port

A source port is the one port at the sending end of a channel. See also destination port.

standard port

Standard port: see local port.

static module

A static module file is a fully located object file that has been compiled and linked for use in a VxWorks 653 module. A static module file has a .sm file extension.

straight-line code

Straight-line code is code that does not use threads.

system call

A system call is a call from a partition to the core OS.

system clock

System clock refers to the system clock for a VxWorks 653 module.

system configuration record

The system configuration record is the record of all the configuration parameters in a VxWorks 653 module. During the configuration process, configuration information is expressed in the Module configuration document. The build process produces a binary version of this information in configRecord.reloc or configRecord.bin.

system health monitor

The system health monitor is the dispatcher for the health monitoring system. See also module health monitor, partition health monitor, and process health monitor.

system heap

System heap refers to the heap for the core OS.

system initialization

System initialization refers to the initialization of a VxWorks 653 module.

system integrator

A system integrator is responsible for integrating the applications created by the application developers with the platform created by the platform provider to create the final module.

system memory

System memory refers to memory controlled by the core OS.

system object

A system object is an object created by the core OS (or vThreads) for use by the core OS (or vThreads). An example is a semaphore.

system resource

A system resource is a resource allocated by the core OS for use by the core OS.

system restart

System restart refers to restarting a VxWorks 653 module.

system shared library

A system shared library is a special shared library that contains the code for a partition OS.

system start

System start refers to starting a VxWorks 653 module.

target

The target is the board for which you are developing an embedded system.

task

A task is an execution context. In VxWorks 653, it refers to a core OS object. See also thread.

TCB

TCB: task control block. The structure that contains critical runtime information for a single task.

thread

A thread is an execution context. It is the preferred term for what is sometimes called a process. A thread is a programming unit contained within a vThreads partition. It runs concurrently with other threads of the same partition. See also task and process.

time-preemptive scheduling

Time-preemptive scheduling: see TPS.

TLB

TLB: translation look-aside buffer. It is a specialized cache that holds a table of physical addresses as generated from the virtual addresses that program code uses.

TPS

TPS: time-preemptive scheduling. It is also called ARINC 653 scheduling. See also APPS scheduling and PPS scheduling.

TPS partition

A TPS partition is the partition that has been scheduled to be run by the ARINC 653 (TPS) scheduler. In an APPS scheduling environment, the current partition and the TPS partition may not be the same.

trusted partition

From the point of view of a given partition, a trusted partition is a partition from which it will allow the health monitor to accept health monitor notifications on its behalf. Since health monitor notifications are processed in the time slice of the partition on whose behalf they are received, limiting the number of partitions that a partition trusts limits the effect of health monitor notifications on the partition's time allotment.

user configuration record

A user configuration record is a collection of data that can be used for configuring user extensions to the core OS.

user memory region

The user memory region is that area of RAM that is needed for memory other than health monitor logs, core OS configuration records, core OS memory, core OS page pools, core pools, ports, and RAM payload.

user partition OS

A user partition OS is a partition OS that is based on COIL, augmented to perform other functions that are required by the application.

VAL

VAL: vThreads abstraction layer. It is a layer of the core OS. When a vThreads partition makes a system call, it communicates with this layer. It is a concept internal to VxWorks 653.

validation

In XML terms, validation is a process that ensures that an XML file is well formed according to the rules of XML and adheres to the structure specified in the appropriate XML schema. Validation is performed by an XML validator.

VME

VME: Versa Module Europa. VME is an open-ended bus system that makes use of the Eurocard standard. The VME bus was intended to be a flexible environment, supporting a variety of computing-intensive tasks, and has become a popular protocol in the computer industry. It is defined by the IEEE 1014-1987 standard.

vThreads

vThreads is the priority-preemptive OS that serves as a partition OS.

vThreads partition

A vThreads partition is a partition whose partition OS is based on vThreads. See also COIL partition.

vThreads scheduler

vThreads scheduler: see partition OS scheduler.

VxWorks 5.5

VxWorks 5.5 is the Wind River operating system on which the vThreads partition OS of VxWorks 653 is based.

VxWorks 653

VxWorks 653 is the Wind River operating system that supports the ARINC 653 specification.

W3C

W3C refers to the World Wide Web consortium at www.w3.org.

wall clock time

Wall clock time is time as measured in the real world by the clock on the wall. (As opposed, for instance, to the time elapsed in a particular application’s partition window.)

warm restart

A warm restart occurs when a module or partition is restarted but persistent data is retained, shortening the time required for the restart.

WDB

WDB refers to the Wind River debug agent.

Wind

Wind is the adjective applied to certain OS objects to distinguishes them from POSIX objects. For example, Wind semaphores to distinguishes from POSIX semaphores.

WindSh

WindSh is a host shell.

Workbench

Workbench is the Wind River Workbench development environment.

worker task

A worker task is a core OS task that is associated with a specific partition. Worker tasks perform blocking operations (typically blocking I/O) on behalf of the partition they are associated with.

write-protect

To write-protect is to guard an entity by a mechanism that prevents it from being changed or erased. For example, memory can be write-protected by using an MMU.

XInclude

XInclude is a W3C standard for including one XML file in another.

XML

XML: Extensible Markup Language. It is a standard for defining markup languages.

XML attribute

An XML attribute is an additional piece of information added to an XML element in the form of a key/value pair.

XML declaration

The XML declaration identifies a file as an XML document and contains information such as the version of XML used and the character encoding used in the file.

XML document

A document written using XML syntax.

XML document type

An XML document type is the grammar of a particular XML file as defined by the applicable XML schema.

XML editor

An XML editor is a program that provides support for editing XML files. This usually includes support for inserting tags and for validating the file against an XML schema.

XML element

An XML document consists of XML elements, each of which may contain data content and/or other elements. The elements allowed in a particular document type is determined by the applicable XML schema.

XML file

An XML file is an instantiation of an XML schema.

XML schema

An XML schema is a document that defines the structure of an XML document. In defines what elements are permitted in an XML document, the order and nesting of elements, and the types of data each element can contain.

XML schema file

An XML schema file is a file that contains all or part of the definition of an XML schema. An XML schema file can include other schema files by reference to construct a complete schema definition.

XPath

XPath is a W3C standard for expressing the location of an element or attribute in an XML file.




Generated on 2007-10-15 17:17:22.