Please enable JavaScript.
Coggle requires JavaScript to display documents.
Real-Time Operating Systems, Aluno: Alexandre Nadolni Bonacim - Coggle…
Real-Time Operating Systems
RTOS Taxonomy and Architectures
Small, Fast, Proprietary Kernels
has a fast context switch
has a small size (minimal functionality)
quick response to external interrupts
minimizes intervals during which interrupts are disabled
fixed or variable sized partitions for memory management, can lock code and data in memory
provides special sequential files
multi-task support
priority-based preemptive scheduling mechanism
bounded execution time for most primitives
high-resolution real-time clock
special alarms and timeouts
real-time queuing (EDF) and primitives for jamming a message into the front of a queue
primitives to delay processing by a fixed amount of time and to suspend/resume execution
Real-Time Extensions to Commercial Operating Systems
compliant kernels
existing RTOS modified to emulate OS
dual kernels
thin real-time kernel below native OS
core kernel modifications
modify non real-time
The Resource kernel approach
kernel extended to provide resource reservations in addition to priority preemptive scheduling
Component-Based Kernels
OS made of components, modular
QoS-Based Kernels
Guaranteed resources for tasks and applications
Research Kernels
Topics
develop real-time process models
standard model both to program and at execution time
process model to program but different run-time model
real-time threads
real-time synchronization primitives
timing analysis
kernel and application predictability
retaining application semantics at run time
fault tolerance support
OOP approach
multiprocessor and distributed real-time systems
support for QoS
real-time micro-kernel
real-time programming languages
challenge basic assumptions and develop new paradigms
Examples
MARS
time-triggered model
a priori static analysis
distributed
automotive industry
SPRING
multiprocessor and distributed
dynamic planning-based scheduling
ARTS
distributed
static scheduling
real time objects and threads
HARTOS
hexagonal mesh network
uses pSOS
interprocess communication
RK
resource kernel
Paradigms
Hard and Soft Real-Time Guarantees
small and deterministic kernel for hard deadlines
larger, dynamic, probabilistic for soft
Admission Control
accepting new work to schedule
Resource Reservation
assigning resources to task
avoids resource conflicts via scheduling
Reflection
system meta data
Resource Kernels
resource reservations
Aluno: Alexandre Nadolni Bonacim