RESOURCE ACCESS PROTOCOLS - Hard Real-Time Computing Systems: Predictable Scheduling Algorithms and Applications - page 234

Let p ∗ be the highest priority among the ready tasks. If p exe <p ∗ , insert τ exe
in the ready queue and execute the ready task (other than τ exe ) with the highest
priority.

7.8 STACK RESOURCE POLICY

The Stack Resource Policy (SRP) is a technique proposed by Baker [Bak91] for ac-
cessing shared resources. It extends the Priority Ceiling Protocol (PCP) in three es-
sential points:

1. It allows the use of multi-unit resources.

2. It supports dynamic priority scheduling.

3. It allows the sharing of runtime stack-based resources.

From a scheduling point of view, the essential difference between the PCP and the
SRP is on the time at which a task is blocked. Whereas under the PCP a task is
blocked at the time it makes its first resource request, under the SRP a task is blocked
at the time it attempts to preempt. This early blocking slightly reduces concurrency
but saves unnecessary context switches, simplifies the implementation of the protocol,
and allows the sharing of runtime stack resources.

7.8.1 DEFINITIONS

Before presenting the formal description of the SRP we introduce the following defi-
nitions.

PRIORITY

Each task τ i is assigned a priority p i that indicates the importance (that is, the urgency)
of τ i with respect to the other tasks in the system. Priorities can be assigned to tasks
either statically or dynamically. At any time t , p a >p b means that the execution of τ a
is more important than that of τ b ; hence, τ b can be delayed in favor of τ a . For example,
priorities can be assigned to tasks based on Rate Monotonic (RM) or Earliest Deadline
First (EDF).