An Essential Relationship between Real-time and Resource Partitioning
0 likes•265 views
The document discusses the essential relationship between real-time processing and resource partitioning in embedded systems, specifically focusing on using Linux for both real-time (RT) and general-purpose (GP) applications. It outlines the requirements for deterministic behavior, the need for hardware resource partitioning, and various implementation approaches, including hybrid and single operating system strategies. The document also addresses practical challenges such as interrupt latency and the management of cascade timers in real-time cores.
![23Embedded Linux Conference Europe 2013
0
20
40
60
80
100
120
140
160
180
200
0 20 40 60 80 100120140160180200220240260280300320340360380400420440460480
Numberofsamples
0
10
20
30
40
50
60
70
80
90
100
0 20 40 60 80 100120140160180200220240260280300320340360380400420440460480
NumberofsamplesEvaluation of interrupt latency with CPU affinity
Evaluation environment
Hardware: Pandaboard
Period: 300μs
Interrupt letency [us]
CPU Affinity
W/O CPU
Affinity](https://image.slidesharecdn.com/elce2013-kobayashi-anessentialrelationshipbetweenreal-timeandresourcepartitioning-170303135528/85/An-Essential-Relationship-between-Real-time-and-Resource-Partitioning-21-320.jpg)
An Essential Relationship between Real-time and Resource Partitioning
- 1. © 2013 Toshiba Corporation An Essential Relationship between Real-time and Resource Partitioning 2013/10/25 Yoshitake Kobayashi Advanced Software Technology Group Corporate Software Engineering Center TOSHIBA CORPORATION
- 2. 4Embedded Linux Conference Europe 2013 Overview Background Requirement Hardware resource partitioning Summary
- 3. 5Embedded Linux Conference Europe 2013 Background Hardware Multi-core CPU Larger memory Larger storage space Hardware assisted virtualization Software Operating system Linux Virtual Machine Monitor
- 4. 6Embedded Linux Conference Europe 2013 Background Hardware Multi-core CPU Larger memory Larger storage space Hardware assisted virtualization Software Operating system Linux Virtual Machine Monitor Issues on real-time systems Meet its required deadline ex. Control systems Performance requirement Interrupt latency Response time
- 5. 7Embedded Linux Conference Europe 2013 Requirement (1) All Real-time application should meet its real- time constrain Response time (Deadline): 100μs – 100ms Event response time (Interrupt latency): 10μs – 100μs Wakeup time Process time Deadline Period Event response time Wakeup time
- 6. 8Embedded Linux Conference Europe 2013 Requirement (2) A system needs to be able to run both real-time (RT) application and general purpose (GP) application at same time
- 7. 9Embedded Linux Conference Europe 2013 A sample of current implementation Prepare two hardware Implement RT application on a specific one Implement GP application on other one Connect each other by a bus or share memory CPU Devices Linux kernel Real time process General purpose processes RTOS BUS
- 8. 10Embedded Linux Conference Europe 2013 A system software able to control RT and GP System software: OS, VMM Requirement (3): Make a simple world CPU Devices System Software Real time process General purpose processes
- 9. 11Embedded Linux Conference Europe 2013 Hybrid OS vs. Single OS approach
- 10. 12Embedded Linux Conference Europe 2013 Hybrid OS vs. Single OS approach Hybrid OS Two or more operating systems runs on same hardware RT specific tasks run on RTOS (Real-Time Operating System) and the other tasks runs on GPOS (General Purpose Operating System) ex. uITRON for RTOS and Linux for GPOS Possible implementations By VMM Run GPOS as a task on RTOS RTOS and GPOS have different APIs Xenomai Single OS Just use one OS to run both RT and GP applications Same API can be used for all applications Possible implementations Kernel level RT process RT-Preempt patch
- 11. 13Embedded Linux Conference Europe 2013 Hybrid architecture (Xenomai) Xenomai Reference: http://www.xenomai.org Dual kernel approach based on Adeos/I-Pipe I-Pipe works to dispatch events (ex. Interrupts) Xenomai skins build on top of the Xenomai nucleus to provide RTOS APIs such as VxWoks, uITRON Adeos/I-Pipe Linux Xenomai nucleus devices Skins
- 12. 14Embedded Linux Conference Europe 2013 Hybrid architecture (TOPPERS SafeG) SafeG (Safety Gate) Reference: http://www.toppers.jp/en/safeg.html Dual-OS monitor Execute an RTOS (Real-Time Operating System) and a GPOS (General-Purpose Operating System) on the same hardware platform ARM TrustZone security extensions uses to introduce the concept of Trust and Non-Trust states On the other hand, code running under Non-Trust state, even in privileged mode, cannot access memory space (devices included) that was allocated for Trust state usage, nor can it execute certain instructions that are considered critical. SafeG LinuxRTOS (uITRON) IRQ devicesFIQ devices
- 13. 15Embedded Linux Conference Europe 2013 Hybrid OS vs. Single OS approach Hybrid OS More than one OS runs on same hardware RT specific tasks run on RTOS (Real-Time Operating System) and the other tasks runs on GPOS (General Purpose Operating System) ex. Linux and uITRON Possible implementations By VMM Run GPOS as a task on RTOS RTOS and GPOS have different APIs Xenomai Single OS Just use one OS to run both RT and GP applications Same API can be used for all applications Possible implementations Kernel level RT process RT-Preempt patch
- 14. 16Embedded Linux Conference Europe 2013 Actual requirement (3): Linux Linux runs both RT and GP applications CPU Devices Linux Real time process General purpose processes
- 15. 17Embedded Linux Conference Europe 2013 Summary of requirements 1. Run RT processes and GP processes on a hardware platform 2. Need to meet required deadlines One of the most important perspective for embedded systems 3. Use single OS approach Linux
- 16. 18Embedded Linux Conference Europe 2013 Actual requirement (3): Linux Linux runs both RT and GP applications This is not a good idea if you don’t care anything CPU Devices Linux Real time process General purpose processes
- 17. 19Embedded Linux Conference Europe 2013 Issues to run RT process and GP process Determinism RT process should have a deterministic behavior GP process doesn’t assume deterministic behaviour
- 18. 20Embedded Linux Conference Europe 2013 Real-time Preemption Patch Fully preemptive kernel Improvement for latency CPU affinity Prohibit process migration from one core to another Protect from GP process behaviour Maybe good for determinism How to improve real time performance? RT CPU core GP CPU core RT thread RTタスクGP thread GP thread Scheduler Scheduler RT thread RT thread Effects by workload
- 19. 21Embedded Linux Conference Europe 2013 Partition is a set of hardware resource CPU cores, Memory, Devices, .. Each partition must be isolated from the others No device sharing Definition of hardware resource partitioning CPU Devices Linux kernel Real time processes General purpose processes
- 20. 22Embedded Linux Conference Europe 2013 A use case of CPU affinity for RT process Run a set of process and thread on specific CPU core Advantage No process migration Process migration is not friendly with real time behviour Migration timing cannot be expected Just RT process runs on specific cores Isolate all GP process into the other cores CPU core 1 CPU core 2 RT process
- 21. 23Embedded Linux Conference Europe 2013 0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100120140160180200220240260280300320340360380400420440460480 Numberofsamples 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100120140160180200220240260280300320340360380400420440460480 NumberofsamplesEvaluation of interrupt latency with CPU affinity Evaluation environment Hardware: Pandaboard Period: 300μs Interrupt letency [us] CPU Affinity W/O CPU Affinity
- 22. 24Embedded Linux Conference Europe 2013 0 20 40 60 80 100 120 140 160 180 200 0 20 40 60 80 100120140160180200220240260280300320340360380400420440460480 Numberofsamples 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100120140160180200220240260280300320340360380400420440460480 NumberofsamplesEvaluation of interrupt latency with CPU affinity 0 10 20 30 40 50 60 70 80 90 100 0 100 200 300 400 500 600 700 800 900 Numberofsamples CPU Affinity (in some case) CPU Affinity W/O CPU Affinity
- 23. 25Embedded Linux Conference Europe 2013 Period ? What’s occurred? Timer cascade occurs here
- 24. 26Embedded Linux Conference Europe 2013 CPU core 0 Process CPU core specific kernel thread CPU core 1 Kernel thread Process Kernel thread CPU core specific kernel thread Limitation of CPU affinity Example for CPU core specific kernel thread Timer, High resolution timer Process migration Etc..
- 25. 27Embedded Linux Conference Europe 2013 List 1 List 2 List 3 List 4 List 5 Inside the each entry Timer 1 Timer 2 ・・・ cascade() Cascade timer list Cascade timer Register the next timer list to the end of current one Impact of cascade timer to interrupt latency Runs with interrupt disabled context No limits for the number of timers Timer process cost becomes higher when tickless kernel used
- 26. 28Embedded Linux Conference Europe 2013 Control cascade timers on RT CPU core Solution Keep the timer list empty on RT core to protect from cascade timers RT CPU core timer.c GP CPU core Timer timer.c Timer Timer list (Empty) Timer list Kernel thread
- 27. 29Embedded Linux Conference Europe 2013 RT CPU core Kernel thread __run_timers() GP CPU core timer __run_timers() Kernel thread Timer list Timer list Core specific kernel thread Timer Timer (Ⅲ) (Ⅱ) (Ⅰ) Three issues which cause cascade timers (Ⅰ) Registered by GP process before migration (Ⅱ) Registered by RT Core specific kernel thread (Ⅲ) Registered by RT Core specific kernel thread before RT task runs Expired timers causes cascade
- 28. 30Embedded Linux Conference Europe 2013 RT core Kernel thread timer.c GP core Timer timer.c Timer Kernel thread Timer list Timer list (イ) (A) (B) Solution for the issue (Ⅰ) Preparation Log all timer registration by kernel thread Solution (A) Migrate kernel threads or a GP processes to GP core (B) Migrate registered times to GP core refer the log （Timer migration）
- 29. 31Embedded Linux Conference Europe 2013 RT core timer.c GP core タイマ処理 Timer list Timer list Core specific kernel thread Timer Timer (Ⅱ) (Ⅲ) Solution for the issues (Ⅱ) and (Ⅲ) Restrict to resister new timers to the GP core only Case 1: Not enough time to migrate Wait for expiration Case 2: Enough time to migrate Migrate the timer to GP core (Ⅱ) Registered by RT Core specific kernel thread (Ⅲ) Registered by RT Core specific kernel thread before RT task runs
- 30. 32Embedded Linux Conference Europe 2013 ① ② Evaluation ① ② Cascade occurs here Before After
- 31. 33Embedded Linux Conference Europe 2013 Summary Requirement RT processes and GP processes on a same hardware platform Just use Linux for both processes Meet its required deadline for RT process Hardware resource partitioning Set of hardware resources which is isolated from the others Define CPU cores as RT core and GP core Issues to implement the resource partitioning Some kernel thread cannot be migrated Core specific kernel thread Need to care with CPU affinity feature Focused on cascade in timer.c Protect from cascade function on RT core Keep timer list empty Future plan Fixing issues SCHED_DEADLINE on RT core with fine granularity support
- 32. 34 Questions? The latest slide is available at the following URL: http://elinux.org/ELC_Europe_2013_Presentations