Rate Monotonic Analysis Rob Oshana Southern Methodist University.

Slides:



Advertisements
Similar presentations
Real Time Scheduling.
Advertisements

Priority Inheritance and Priority Ceiling Protocols
Introduction to Embedded Systems Resource Management - III Lecture 19.
Priority INHERITANCE PROTOCOLS
CS5270 Lecture 31 Uppaal, and Scheduling, and Resource Access Protocols CS 5270 Lecture 3.
CprE 458/558: Real-Time Systems (G. Manimaran)1 CprE 458/558: Real-Time Systems Resource Access Control Protocols.
CSE 522 Real-Time Scheduling (4)
Mutual Exclusion.
Real-time concepts Lin Zhong ELEC424, Fall Real time Correctness – Logical correctness – Timing Hard vs. Soft – Hard: lateness is intolerable Pass/Fail.
Chapter 9 Uniprocessor Scheduling Operating Systems: Internals and Design Principles, 6/E William Stallings Dave Bremer Otago Polytechnic, N.Z. ©2008,
CS 3013 & CS 502 Summer 2006 Scheduling1 The art and science of allocating the CPU and other resources to processes.
Wk 2 – Scheduling 1 CS502 Spring 2006 Scheduling The art and science of allocating the CPU and other resources to processes.
By Group: Ghassan Abdo Rayyashi Anas to’meh Supervised by Dr. Lo’ai Tawalbeh.
Spring 2002Real-Time Systems (Shin) Rate Monotonic Analysis Assumptions – A1. No nonpreemptible parts in a task, and negligible preemption cost –
Real-Time Operating System Chapter – 8 Embedded System: An integrated approach.
Real-Time Software Design Yonsei University 2 nd Semester, 2014 Sanghyun Park.
Computer Architecture and Operating Systems CS 3230: Operating System Section Lecture OS-3 CPU Scheduling Department of Computer Science and Software Engineering.
OPERATING SYSTEMS CPU SCHEDULING.  Introduction to CPU scheduling Introduction to CPU scheduling  Dispatcher Dispatcher  Terms used in CPU scheduling.
Chapter 9 Uniprocessor Scheduling Spring, 2011 School of Computer Science & Engineering Chung-Ang University.
1 Previous lecture review n Out of basic scheduling techniques none is a clear winner: u FCFS - simple but unfair u RR - more overhead than FCFS may not.
Real-Time Scheduling CS4730 Fall 2010 Dr. José M. Garrido Department of Computer Science and Information Systems Kennesaw State University.
Scheduling policies for real- time embedded systems.
CPU Scheduling CSCI 444/544 Operating Systems Fall 2008.
Reference: Ian Sommerville, Chap 15  Systems which monitor and control their environment.  Sometimes associated with hardware devices ◦ Sensors: Collect.
Real-Time Scheduling CS4730 Fall 2010 Dr. José M. Garrido Department of Computer Science and Information Systems Kennesaw State University.
Real-Time Scheduling CS 3204 – Operating Systems Lecture 20 3/3/2006 Shahrooz Feizabadi.
Deadlocks Silberschatz Ch. 7 and Priority Inversion Problems.
2.5 Scheduling Given a multiprogramming system. Given a multiprogramming system. Many times when more than 1 process is waiting for the CPU (in the ready.
Deadlock Detection and Recovery
Real Time Systems Real-Time Schedulability Part I.
RTOS task scheduling models
- 1 -  P. Marwedel, Univ. Dortmund, Informatik 12, 2006 Universität Dortmund Periodic scheduling For periodic scheduling, the best that we can do is to.
Real-Time Scheduling CS 3204 – Operating Systems Lecture 13 10/3/2006 Shahrooz Feizabadi.
1 Real-Time Scheduling. 2Today Operating System task scheduling –Traditional (non-real-time) scheduling –Real-time scheduling.
CSCI1600: Embedded and Real Time Software Lecture 24: Real Time Scheduling II Steven Reiss, Fall 2015.
2.5 Scheduling. Given a multiprogramming system, there are many times when more than 1 process is waiting for the CPU (in the ready queue). Given a multiprogramming.
Introduction to Embedded Systems Rabie A. Ramadan 5.
A presentation for Brian Evans’ Embedded Software Class By Nate Forman Liaison Technology Inc. 3/30/2000 For Real-Time Scheduling.
CSCI1600: Embedded and Real Time Software Lecture 23: Real Time Scheduling I Steven Reiss, Fall 2015.
CS333 Intro to Operating Systems Jonathan Walpole.
For a good summary, visit:
Time Management.  Time management is concerned with OS facilities and services which measure real time.  These services include:  Keeping track of.
1.  System Characteristics  Features of Real-Time Systems  Implementing Real-Time Operating Systems  Real-Time CPU Scheduling  An Example: VxWorks5.x.
Undergraduate course on Real-time Systems Linköping University TDDD07 Real-time Systems Lecture 2: Scheduling II Simin Nadjm-Tehrani Real-time Systems.
CPU Scheduling CS Introduction to Operating Systems.
Distributed Process Scheduling- Real Time Scheduling Csc8320(Fall 2013)
Real-Time Operating Systems RTOS For Embedded systems.
CPU Scheduling Scheduling processes (or kernel-level threads) onto the cpu is one of the most important OS functions. The cpu is an expensive resource.
Embedded System Scheduling
REAL-TIME OPERATING SYSTEMS
Processes and threads.
RTOS Scheduling 2.0 Problems - Solutions
Scheduling and Resource Access Protocols: Basic Aspects
Wayne Wolf Dept. of EE Princeton University
EEE 6494 Embedded Systems Design
Chapter 2 Scheduling.
Uniprocessor Scheduling
Lecture 4 Schedulability and Tasks
Real-time Software Design
Chapter 6: CPU Scheduling
Rate Monotonic Analysis For Real-Time Scheduling A presentation for
Chapter 6: CPU Scheduling
CSCI1600: Embedded and Real Time Software
CPU SCHEDULING.
Lecture 2 Part 3 CPU Scheduling
CSCI1600: Embedded and Real Time Software
Processes and operating systems
Chapter 10 Multiprocessor and Real-Time Scheduling
Real-Time Process Scheduling Concepts, Design and Implementations
Real-Time Process Scheduling Concepts, Design and Implementations
Presentation transcript:

Rate Monotonic Analysis Rob Oshana Southern Methodist University

Scheduling policies for real- time systems

Scheduling Policies in RT systems Two general categories –fixed or static scheduling policies –dynamic scheduling policies Many commercial RTOSs today support fixed priority scheduling policies Fixed priority scheduling algorithms do not modify a job’s priority while the task is running

Scheduling Policies in RT systems –The task itself is allowed to modify its own priority for reasons –approach requires very little support code on the scheduler to implement this functionality –scheduler is fast and predictable with this approach –scheduling is mostly done off-line (before the system runs)

Scheduling Policies in RT systems Requires system designer to know the task set a-priori (ahead of time) –not suitable for tasks that are created dynamically during run time The priority of the task set must be determined beforehand and cannot change when the system runs unless the task itself changes its own priority

Scheduling Policies in RT systems Dynamic scheduling algorithms allow a scheduler to modify a jobs priority based on one of several scheduling algorithms or policies –more complicated approach and requires more code in the scheduler to implement

Scheduling Policies in RT systems –leads to more overhead in managing a task set –scheduler must spend more time dynamically sorting through the system task set and prioritize tasks for execution based on the scheduling policy –leads to non-determinism which is not favorable, especially for hard real-time systems

Scheduling Policies in RT systems Dynamic scheduling algorithms are on-line scheduling algorithms –scheduling policy is applied to the task set during the execution of the system –active task set changes dynamically as the system runs –priority of the tasks can also change dynamically

Examples Static scheduling policies –rate monotonic scheduling –deadline monotonic scheduling Dynamic scheduling policies –earliest deadline first –least slack scheduling

Rate Monotonic Scheduling Optimal fixed priority policy –the higher the frequency (1/period) of a task, the higher is its priority Approach can be implemented in any OS supporting the fixed priority preemptive scheme Rate monotonic scheduling assumes the deadline of a periodic task is the same as its period

Deadline Monotonic Scheduling Generalization of the Rate-Monotonic scheduling policy Deadline of a task is a fixed (relative) point in time from the beginning of the period The shorter this (fixed) deadline, the higher the priority

Earliest Deadline First Dynamic priority preemptive policy Deadline of a task instance is the absolute point in time by which the instance must complete Task deadline is computed when the instance is created OS scheduler picks the task with the earliest deadline to run A task with an earlier deadline preempts a task with a later deadline

Least Slack Dynamic priority preemptive policy Slack of a task instance is the absolute deadline minus the remaining execution time for the instance to complete OS scheduler picks the task with the shortest slack to run first Task with a smaller slack preempts a task with a larger slack This approach maximizes the minimum lateness of tasks

More on dynamic policies Priority of a task can change from instance to instance or within the execution of an instance Higher priority task preempts a lower priority task Very few commercial RTOS support such policies –systems that are hard to analyze for real-time and determinism properties

Periodic tasks Many systems are multi-rate systems –multiple tasks in the system running at different periodic rates Muli-rate systems can be managed using non-preemptive as well as preemptive scheduling techniques Non-preemptive techniques include using state machines as well as cyclic executives

Examples of periodic tasks  Audio sampling in hardware  Audio sample processing  Video capture and processing  Feedback control (sensing and processing)  Navigation  Temperature and speed monitoring

Scheduling periodic tasks Preemptive scheduling is an effective approach for scheduling real-time DSP systems –modularity simplifies the overall design Application can be viewed as a collection of independent tasks or jobs –complexity is reduced as the functionality becomes encapsulated into a set of well defined tasks

Scheduling periodic tasks Systems designed using preemptive scheduling are also more maintainable –issue of changes to one task in the system affecting other jobs in the system is removed –New functionality can easily be added by adding a new task

Scheduling periodic tasks Preemptive scheduling approach also makes the system more efficient –preemptive scheduling is more efficient at utilizing time slots that may not be fully utilized Scheduling algorithms –rate monotonic scheduling –deadline monotonic scheduling

cost of handling event C = 4 periodic arrivals. period T = System Utilization = C/T =.40 Periodic Arrivals with Fixed Cost of Processing System will be able to meet all deadlines. It can finish processing arrivals before the next arrival occurs.

1. periodic arrival, period T = 10 and C=4 2. periodic arrival, T=10 and C=3 ?? Can a second periodic event be accommodated?

1. periodic arrival, period T = 10 and C=4 2. periodic arrival, T=10 and C=3 ?? System Utilization C/T =.70 Can a second periodic event be accommodated?

1. periodic arrival, period T = 10 and C=4 2. periodic arrival, T=6 and C=3 ?? How about 2nd periodic event with T=6 and C=3?

1. periodic arrival, period T = 10 and C=4 2. periodic arrival, T=6 and C=3 ?? System Utilization C/T =.90 How about 2nd periodic event with T=6 and C=3?

If we process Event #1 before Event #2 then, 2nd event processing will not complete before the next comparable event occurs Can’t Meet Deadline! Event #1 Event #2 Task #1 Task #2

Event #1 Event #2 Try Event #2 before Event #1- We still cannot complete task 1 before the next task 2 event occurs at t=6 unless... Task #1 Task #2

Event #1 Event #2 Try Event #2 before Event #1- We still cannot complete task 1 before the next task 2 event occurs at t=6 unless…we Interrupt task 1 Task #1 Task #2

Event #1 Event #2 Try Event #2 before Event #1- We still cannot complete task 1 before the next task 2 event occurs at t=6 unless…we Interrupt task 1 Giving event #2 priority means that we can meet our deadline IF we preempt the processing of event #1 when event #2 occurs Task #1 Task #2

Rate Monotonic Analysis

Assume a set of “n” periodic tasks –period Ti –worst case execution time Ci Rate-monotonic priority assignment –task with a shorter period (higher rate) assigned a fixed higher priority

Rate Monotonic Analysis Rate Monotonic scheduling addresses how to determine whether a group of tasks, whose individual CPU utilization is known, will meet their deadlines –assumes a priority preemption scheduling algorithm –assumes independent tasks (no communication or synchronization)

Rate Monotonic Analysis –restriction of no communication or synchronization may appear to be unrealistic, but there are techniques for dealing with this –Each task is a periodic task which has a period T, which is the frequency with which it executes

Rate Monotonic Analysis  An execution time C, which is the CPU time required during the period  A utilization U, which is the ratio C/T A task is schedulable if all its deadlines are met (i.e., the task completes its execution before its period elapses.) –A group of tasks is considered to be schedulable if each task can meet its deadlines

Rate Monotonic Analysis RMA is a mathematical solution to the scheduling problem for periodic tasks with known cost –assumption is that the total utilization must always be less than or equal to 100% Any more and you are exceeding the capacity of the CPU Are you asking for more computing power than you have? IF so, forget it!

Rate Monotonic Analysis For a set of independent periodic tasks, the rate monotonic algorithm assigns each task a fixed priority based on its period, such that the shorter the period of a task, the higher the priority

Rate Monotonic Analysis For three tasks T1, T2, and T3 with periods of 5, 15 and 40 msec respectively the highest priority is given to the task, T1, as it has the shortest period, the medium priority to task T2, and the lowest priority to task T3 –priority assignment is independent of the applications “priority” i.e. how important meeting this deadline is to the functioning of the system or user concerns

Rate Monotonic Analysis A mathematical solution to the scheduling problem for Periodic Tasks with known Cost Tasks will have: – Cost (Time to complete a task) – Period (Time between events) – Utilization ( Cost/Period) Assumption – Total Utilization must always be <= 100%

3 levels of analysis using RMA Utilization bound test Completion time test Response time test

Utilization bound test If this simple rule is followed, then all tasks are guaranteed to meet their requirements if the following holds true;

Utilization bound test In this rule, the bound is 1.0 for harmonic task sets A task set is said to be harmonic if the periods of all its tasks are either integral multiples or sub-multiples of one another – On the average, for random Cs and Ts, this number will be about 0.88.

Utilization bound test Theory is a worst case approximation For a randomly chosen group of tasks, it has been shown that the likely upper bound is 88% –Harmonic periods can give even higher upper bounds –The algorithm is stable in conditions where there is a transient overload

Utilization bound test In this case, there is a subset of the total number of tasks, namely those with the highest priorities that will still meet their deadlines

Example of UB test  Task t1: C1=20; T1= 100; U1 =.2  Task t2: C2=30; T2= 150; U2 =.2  Task t3: C3=60; T3= 200; U3 =.3 –The total utilization for this task set is =.7. Since this is less than the utilization bound for this task set, all deadlines will be met.

Example Can these 4 tasks be scheduled? – Can the system run and meet all hard deadlines? TaskC i T i U i

Example Can these 4 tasks be scheduled? –Can the system run and meet all hard deadlines? –NO! The Total Utilization = 115% TaskC i T i U i

Example Can these tasks always meet their deadlines? Total Utilization = 80% It MAY be possible - Rate Monotonic Scheduling applies! TaskC i T i U i

Rate Monotonic Theorem For PERIODIC Tasks Most frequent task gets highest priority THEOREM (Simple Version) – IF the utilization of all tasks is less than or equal to 69%, then all tasks will ALWAYS meet their deadlines

Are These Tasks Schedulable? TaskC i T i U i

Are These Tasks Schedulable? TaskC i T i U i Yes. Total CPU Utilization is 65% < 69%

Are These Tasks Schedulable? TaskC i T i U i

Are These Tasks Schedulable? TaskC i T i U i Total CPU Utilization is 65% ???

Priority Inversion Task h Task med Task low Normal execution Execution in critical section Priority inversion

Unbounded Priority Inversion Task h Task med Task low Normal execution Execution in critical section Priority inversion Task med Uncontrolled priority inversion

Motor control example Single DSP will be used to control a motor DSP will also be responsible for interfacing to an operator using a keypad, updating a simple display device, and sending data out one of the DSP ports The operator uses the keypad to control the system

Motor control example The motor speed must be sampled at a 1 kHz rate A timer will be used to interrupt processing at this rate to allow the DSP to execute some simple motor control algorithms At each interrupt, the DSP will read the motor speed, run some simple calculations, and adjust the speed of the motor accordingly

Motor control example Diagnostic data is transmitted out the RS232 port when selected by the operator using the keypad

Simple motor control Motor Drive MotorTach RS232 Data Out

Motor control example The first step in developing a multitasking system is to architect the application into isolated independent execution threads –tools available to help the system designer during this phase This architecture definition will produce data flow diagrams, system block diagrams, or finite state machines

Motor control example Four independent threads in this design; –main motor control algorithm which is a periodic task, running at a 1kHz rate –keypad control thread which is an aperiodic task controlled by the operator –display update thread which is a periodic task executing at a 2 Hz rate

Motor control example –data output thread which runs as a background task and outputting data when there is no other processing required

Requirements Control motor speed (1 Khz sampling rate – dV/dT) Accept keyboard commands to control the motor, change the display, or send data out the RS232 port Drive a simple display and refresh 2 times per second Send data out the RS232 port when there is nothing else to do

Independent execution threads ADC DAC Motor Drive Control algorithm Motor Tach Motor control Keypad control Display control Remote output System Control algorithm Keypad control Display interface Display algorithm RS-232 Remote algorithm McBSP

Motor control example Relative priority of the threads must now be determined Since this motor control example is a real-time system (there are hard real- time deadlines for critical operations to complete), there must be a priority assigned to the thread execution

Motor control example One hard real-time thread –motor control algorithm –must execute at a 1 kHz rate Soft real-time tasks in the system as well –display update at a two hertz rate is a soft-real time task (this is a soft real-time task because although the display update is a requirement, the system will not fail if the display is not updated precisely twice per second.)

Motor control example –keypad control is also a soft real-time task but since it is the primary control input, it should have a higher priority than the display update –remote output thread is a background task that will run when no other task is running

Motor control example Motor control system will be designed to use a hardware interrupt to control the motor control thread –Interrupts have fast context switching times (faster than that of thread context switch) and can be generated from the timer on the DSP

Assignment of priorities TaskRatePriorityPeriodic or aperiodic Activation mechanism Motor control1 kHz1PeriodicHardware Interrupt Keypad control 5 hertz2AperiodicHardware Interrupt Display control 2 hertz3PeriodicSoftware interrupt Remote output l Background4AperiodicIdle loop (runs continuously in the background)

Motor control example Example of a rate monotonic priority assignment; –the shorter the period of the thread, the higher the priority Along with the priority, the activation mechanism is described The highest priority motor control thread will use a hardware interrupt to trigger execution

Motor control example Hardware interrupts are the highest priority scheduling mechanism in most real time operating systems The keypad control function is an interface to the environment (operator console) and will use a hardware interrupt to signal the keypad control actions –This priority will be at a lower priority than the motor control interrupt

Motor control example Display control thread will use a software interrupt to schedule a display update at a two hertz rate –Software interrupts operate similar to hardware interrupts but at a lower priority than hardware interrupts (but at a higher priority than threads) Lowest priority task, the remote output task, will be executed as an idle loop

Motor control example –idle loop will runs continuously in the background while there are no other higher priority threads to run The highest priority thread, the motor control thread, is a periodic thread –Like many DSP applications this thread processes data periodically, in this case at a 1 kHz rate

Motor control example This motor control example is actually a multi-rate system –multiple periodic operations in the system (motor control and display control, for example) –threads operate at different rates –DSP RTOSs allow for multiple periodic threads to run

Motor control example For each of these threads, the DSP system designer must determine the period the specific operation must run and the time required to complete the operation The DSP developer will then program the DSP timer functions in such a way to produce an interrupt or other scheduling approach to enable the thread to run at the desired rate

Motor control example Most DSP RTOSs have a standard clock manager and API function to perform this setup operation

Periodic threads DSP timer User Defined clocking Selectable source Clock source Periodic Function manager Motor control DSP function Display control DSP function Periodic clockPeriodic functions tick n ticks m ticks

What Happened on Mars ?

What happened on Mars ? Mars pathfinder “flawless” in early days of mission –unconventional landing with airbags –deployment of Sojourner rover –gathering and transmitting data back to earth A few days into the mission the Pathfinder began experiencing total system resets, each including losses of data

What happened on Mars ? Press reported these as “software glitches” VxWorks RTOS provides preemptive priority scheduling of tasks –tasks executed as threads –priorities assigned reflecting relative urgency of the tasks

What happened on Mars ? bus management task information bus mutex meteorological data gathering task communication task high priority - frequent execution low priority - infrequent execution medium priority

Combination worked fine most of the time Possible for interrupt to occur that caused the long running medium priority task to be scheduled during the short interval while the high priority task was blocked waiting on the semaphore that the low priority task had. What happened on Mars ?

Watchdog timer would go off, notice data bus task not in use for some time, conclude that something bad went wrong, and initiate a total system reset Classic case of priority inversionClassic case of priority inversion What happened on Mars ?

How was this debugged ? VxWorks can run in trace mode, recording interesting events. JPL engineers spent hours in lab trying to reproduce the problem on the ground. When finally reproduced, the trace data indicated the priority inversion problem

How was this problem corrected? Mutex object accepts boolean parameter indicating whether priority inheritance should be used Initialized with parameter off –if on, the low-pri thread would have inherited the pri of the high-pri thread –medium pri thread would never have been executed

How was this problem corrected? VxWorks has a C language interpreter that allows C commands to be executed on the fly JPL engineers left this in the software Changed global variables by uploading a short program to the spacecraft No more system resets occurred after re- programming!

Analysis and Lessons Diagnosing this problem as a black box would have been impossible –trace data was required Leaving debugging facilities in the system saved the day Time critical situations requires additional correctness measures even at the expense of some performance

Human nature, Deadline Pressures One or two system resets had occurred on the ground prior to launch Never reproducable or explainable “it was probably caused by a hardware glitch” Engineer focus caused part of the problem –extremely focused on ensuring quality and flawless operation of landing software –the occasional glitch was dismissed

Importance of good Theory/Algorithms Some of the heros were people from CMU who published a paper years ago on the priority inversion problem –“An Approach to Real-Time Synchronization” IEEE Transaction on Computers, Vol39, pp , September 1990