Procedures The Terminal Device Driver interfaces the HPC

on-board UART to the ISDN Software Device initialization

sequences service request tasks and accompanying inter-

rupt service routines are all defined in the IO Device Driver

section of this document

The Layer Protocol Tasks implement the ISDN DATA LINK

Layer 2 and the NETWORK Layer 3 requirements for the

HPC ISDN system These tasks are designed to be hard-

ware configuration and application independent The Layer

2 Task provides both the ‘‘USER SIDE’’ and the ‘‘NET-

WORK SIDE’’ implementation of the CCITT Specification

Q921 The Layer 3 Task provides the ‘‘USER SIDE’’ imple-

mentation of CCITT Specification Q931

The Layer 2 Task has been designed to use many of the

same routines to implement the link access procedures on

either the signaling D channel or the bearer B channel

(LAPD or LAPB) Design decisions have also been made to

facilitate the implementation of V120 the rate adaption pro-

tocol that processes LAPD frames on a bearer B channel

The Management Entity Task and the Call Control Task are

Application (Specific) Tasks that are closely coupled to the

specific system hardware configuration and the Central Of-

fice Network Entity Software These tasks are provided for

demonstration purposes to drive the ISDN layer entities Ap-

plication users must either replace or extensively rewrite

these tasks to match their particular ISDN Application envi-

ronment

The System Utilities include the power-up reset Main Task

the NMI handler the Timer interrupt handler and the

Watchdog Task

The Tracer utility provides the capability of on-line tracing of

intertask mail messages and task states Tracer is primarily

a passive task it displays messages that it receives from

other tasks Tracer also provides a user interface for Tele-

phone Simulation

The remainder of this document is devoted to defining each

of the software elements at the functional level Where ap-

plicable specific ISDN standard documents such as CCITT

Q921 Q931 and X25 will be referenced rather than dupli-

cating the information here

22 HPC EXECUTIVE

The HPC Executive provides a multitasking environment

within which the ISDN and applications tasks can run and it

provides various system services to those tasks The serv-

ices of the Executive are available to both tasks and inter-

rupt service routines

221 Tasks Priorities and the Ready Queue

A task is a subroutine which can be run (called) by the Exec-

utive Tasks are managed by the Executive as Task Control

Blocks (TCB’s) A task’s TCB contains all the parameters

needed by the Exeuctive to handle the task in particular

the task’s priority and its current starting address

Tasks which are not blocked waiting for a semaphore or for

mail are considered to be ready to run and their TCB’s are

queued on the Ready Queue in the order of the tasks’ prior-

ities The Task Scheduler runs the task at the head of the

Ready Queue ie the highest priority task that is ready to

run In this way the processor is always given to the highest

priority task that is ready to run

Once a task is started it continues to run until it does a

Semaphore Wait ReadMail or Return or until a higher pri-

ority task is put on the Ready Queue at which time the

scheduler has the opportunity to once again choose the

task at the head of prioritized Ready Queue and run that

task

A task may change the priority of any task including itself

The priority change takes place immediately to the extent

that the target task’s TCB is updated with the new priority

and the queue in which the target task’s TCB is waiting is

resorted to reflect the new priority

If the target task is in the Ready Queue and its new priority

is higher than the priority of the running task then the target

task will run once all protected sections are exited See

Section 223 below

222 Semaphores

A semaphore is a global variable accessed through the Ex-

ecutive which can be Signaled (incremented) by one task

and Waited on by another task A semaphore is typically

used to manage the sharing between tasks of some re-

source eg an IO device mail messages etc At any mo-

ment the value of a semaphore may be positve negative or

zero A positive value indicates the number of resources

available a negative value indicates the number of tasks

waiting for resources and a zero value indicates that there

are no resources available and no tasks waiting for them

When a task Waits on a semaphore if the semaphore has a

nonzero positive value the task will immediately go on the

Ready Queue and the semaphore value will be decrement-

ed by one On the other hand if the semaphore has a zero

or negative value the task will be queued on the semaphore

and the semaphore value will be decremented by one

When a task Signals a semaphore the semaphore’s value

is incremented by one and the highest priority task waiting

on the semaphore is put on the Ready Queue

A common use for a semaphore is the management of a

non-shareable resource such as an IO device When the

device is available the associated semaphore has the value

a

1 When a task wishes to obtain exclusive use of the de-

vice it Waits on the semaphore which is then decremented

to 0 with the task going immediately back on the Ready

Queue If another task then attempts to use the device its

Wait call will cause it to be placed on the Semaphore Queue

Other tasks may also Wait on the semaphore each decre-

menting its value by one The negative value of the sema-

phore indicates the number of tasks Waiting for the device

The waiting tasks are ordered in the semaphore queue ac-

cording to their priority When the first task is done with the

device it Signals the semaphore which moves the first wait-

ing task to the Ready Queue and increments the sema-

phore or if there are no waiting tasks returns the sema-

223 Preemptive Scheduling

Preemptive scheduling enables the executive to respond

quickly to high priority events If a task that is waiting on a

Semaphore Queue modes to the Ready Queue and if that

task is of higher priority than the currently running task

then as soon as the currently running task emerges from all

critical sections and non-preempt sections the currently

running task will stop running The task that was moved to

the Ready Queue will run The preempted task will be

placed on the Ready Queue in the normal manner

5