• What is Cheddar ?
  
• News
  
• Download binaries and source code
  
• Install Cheddar
  
• Last patch
  
• Work in progress
• Project examples
• Mailing List
• Related documents and publications
• Educational material
• Authors and contributors
• Bugs report and new tested platforms
  
• Feedback from users
  

What is Cheddar ?

Cheddar is a free real time scheduling tool. Cheddar is designed for checking task temporal constraints of a real time application/system. It can help you for quick prototyping of real time schedulers. It can also be used for educational purpose. Cheddar is developed and maintened by the LISYC Team, University of Brest.

Our real time scheduling simulator is composed of two independent parts : a graphical editor used to describe a real time application/system, and a framework which includes most of classical real time scheduling/feasibility algotithms/tests.

Cheddar is written in Ada. The graphical editor is made with GtkAda. Cheddar runs on Solaris, Linux and win32 boxes and should run on every GNAT/GtkAda supported platforms ( see AdaCore web site for details).

Cheddar provides two kind of features : a simulation engine and feasisibility tests. Feasibility tests allow a user to study a real time application/system without computing a scheduling. In the contrary, the simulation engine can be used firstly to computing a scheduling and secondly, to automatically looking for task constraint properties in the computed scheduling. Most of the time, feasibility tests are less complex tools but they are available only for few schedulers and tasks models. To solve this problem, the Cheddar simulation engine provides tools to design specific schedulers and task models : the system is then analyzed according to a given scheduling.

For each analysis result, Cheddar gives the name of the applied equation and the related publication. See the publications topic of this web page to get these publications.

The main features provided by Cheddar are then :

• Features to do scheduling simulation with classical real time schedulers :
  • Rate Monotonic analysis (sometimes called RMA, RM or RMS ).
  • Earliest Deadline First (or EDF).
  • Deadline Monotonic (or DM, Inverse Deadline).
  • Least Laxity First (or LLF).
  • POSIX 1003b scheduling and queueing policies, (SCHED_FIFO, SCHED_RR, SCHED_OTHERS).
  • From a scheduling simulation, much information on the studied real time system is computed :
    • Worst/best/Average blocking time,
    • Worst/best/Average response time,
    • Number of preemptions, of context switches,
    • Deadlocks, priority inversions,
    • Missed deadlines,
    • ...
  • ...
• Features to apply many feasibility tests in the preemptive case and in the non preemptive case :
  • Bound on Response times (with EDF, LL, DM and RM).
  • Bound on Blocking times (with PCP and PIP),
  • Feasibility tests based on the processor utilization factor (with EDF, LLF, RM and DM).
• Support for cyclic and aperiodic tasks. Cyclic tasks can be periodic (with jitter and deadline > period) or can be activated with a random poisson process distribution (useful to model random aperiodic arrivals pattern).
• Shared resource support (scheduling with shared resources and blocking time analysis). Supported protocols : PIP, PCP.
• Tools to express precedencies and do scheduliing simulations and feasibility tests :
  • Tindell end to end response time computation (feasibility test).
  • Chetto and Blazewicz task parameter modification algorithms (feasibility test and scheduling simulation).
  • Scheduling simulation with task precedencies hosted by different processors.
• Partitionning tools to assign tasks on muli-processors systems.
• Tools to study task sharing buffers :
  • Feasibility test to compute bounds on buffers.
  • Scheduling simulation to buffer utilization factor analysis.
• Scheduling simulation of applications distributed on several processor and sharing messages.
• Support to design your own schedulers, task pattern arrivals and event analyzers (event analyzers provide a way to express your own properties that you want to check from a scheduling simulation).
• ...

The picture below is a screenshot of Cheddar : on the top of the picture, scheduling is drawn according to the scheduler of each processor. On the bottom, some task information is provided (response times, missed deadlines, processor utilization level, ...).

News

• 12/07/2005 : provides solaris binaries
• 11/17/2005 : fix the install Makefile and the INSTALL file to the linux distribution. The current linux archive contains these updated files.
• 11/11/2005 : the gtkada compiled libraries are now include in the linux binaries distribution.
• 11/01/2005 : annual Cheddar's release. Current release is now 1.3p5
• 09/11/2004 : new patch on 1.3 (see Last patch section). Current release is now 1.3p4
• 09/08/2004 : a new Cheddar's release is distributed (1.3p3)
• 09/25/2003 : new version of the User's guide.
• 09/09/2003 : new patch on 1.3 (see Last patch section). Current release is now 1.3p2
• 08/28/2003 : new patch on 1.3 (see Last patch section). Current release is now 1.3p1
• 08/26/2003 : a new Cheddar's release is distributed (1.3p0)
• 07/03/2003 : a remote cvs access is now available. If you plan to write code with Cheddar, contact us to get an account on this cvs repository.
• 05/31/2003 : create a Cheddar users and developpers mailing list and provide dayly built cvs snapshot.
• 11/21/2002 : release of a new patch, see BUGS file for list of fixed bugs. Current version is now Cheddar-1.2p1.
• 10/21/2002 : improve installation Makefile and online User's guide.

Download Cheddar

• Linux binaries : Cheddar-1.3p5-linux-bin.tar.gz (this file also includes all required libraries).
• Windows binaries : Cheddar-1.3p5-win32-bin.zip (this file also includes all required DLL).
• Solaris binaries : Cheddar-1.3p5-solaris-bin.tar.gz (this file also includes all required libraries).

Cheddar is a free software under the GNU General Public License.
You can get the current stable Cheddar source code here .

Previous releases can be found here . See ChangesLog to have history of modifications.

You can also access to a dayly built cvs snapshot of Cheddar. Source code and binaries you get from there are not stable : use them with care. Finally, a remote cvs access is also available. If you plan to write code with Cheddar, contact us to get an account on this cvs repository.

Install Cheddar

1) Tested platforms :

Cheddar runs on :

• Linux Redhat (6.1, 6.2, 7.1 and 9.1).
• Sparc/Solaris 2.7, 2.8 and 2.9.
• Windows Me, 95, 98, 2000, NT4 and XP.
• Cheddar should work on all GNAT/GtkAda supported platforms.

2) What do you need ?

On windows boxes, just unzip the zip file. To install Cheddar on Unix boxes, you need the GNU make command (gmake).

All necessary software used to compile or install Cheddar can be found here.

3) Basic Unix Installation process :

For unix, the simpliest way to install this program is:

1. Edit the Makefile to give the name of the directory where cheddar must be installed. To do this, update the CHEDDAR_ROOT variable. If you do not change CHEDDAR_ROOT, Cheddar will be installed in /usr/local
2. Type `gmake' to install the package.
3. Update your PATH and LD_LIBRARY_PATH variables (see the INSTALL file).
4. Play and have fun.

Last patch

Work in progress

1. Update user's guide according to features related to buffer analysis and task precedencies tools.
2. Update the graphical editor in order to switch from GtkAda 1.2 towards GtkAda 2.x
3. Hierarchical schedulers support.
4. Provide a global memory analysis (task stack, text/data segments ...)
5. End to end response time with message scheduling.
6. User's defined message scheduler.
7. Improvement of message scheduling with :
   • Allowing message sending at any time of a task capacity
   • Providing a way to user-defined message delay communication by specification of user-defined message scheduling (as user-defined scheduler)
8. Fixing a buggy service which should detect deadlock from simulation.

Mailing List

Related documents and publications

• Cheddar : a Flexible Real Time Scheduling Framework. F. Singhoff, J. Legrand, L. Nana, L. Marcé. ACM Ada Letters journal, 24(4):1-8, ACM Press. Also published in the proceedings of the ACM SIGADA International Conference, Atlanta, 15-18 November, 2004.
• Scheduling and Memory requirement analysis with AADL. F. Singhoff, J. Legrand, L. Nana, and L. Marcé. ACM Ada Letters journal, 25(4):1-10, ACM Press. Also published in the proceedings of the ACM SIGADA International Conference, Atlanta, 14-17 November, 2005.
  
  
• Cheddar User's Guide. Frank Singhoff, Jérôme Legrand. EA 2215 Technical report number singhoff-01-2003, September 2003.
• The Cheddar FAQ.
• Framework User's guide. Jérôme Legrand, Frank Singhoff. EA 2215 Technical report number legrand-01-2004, July 2004.
  
  
• All the publications used by Cheddar to compute and analyse a real time system can be downloaded here.
• All the publications related to Cheddar can be downloaded here.

Educational Material

• The Monash University (C. Kopp). Cached documents . In english.
• Centro Universitario de Mérida (H. S. Santamaria). Cached documents . In spanish.
• University of Rhode Island (V. Fay-Wolfe). Cached documents . In english.
• Universitat Politécnica de Catalunya (R. Costa-Castelló). Cached documents . In English and Catalan.
• University of Brest (F. Singhoff). Cached documents . In french.

Authors and contributors

Bugs report and new tested platforms

Please, report bugs and new tested platforms to cheddar@listes.univ-brest.fr (Warning : to post, you have to subscribe to the list).

Feedback from users

You are using Cheddar ? If so, please, let us know why you are using it (study of real time applications, educational purposes, prototyping of schedulers, ...) and what kind of features you would like to see in this program.


Users can contact us with cheddar@listes.univ-brest.fr (Warning : to post, you have to subscribe to the list).

Thanks for all people who sent us comments, bugs and advices.