Public Member Functions
	AlgorithmWcdo (TransactionSet gamma, Class<?extends AbstractAlgorithm > baseAlgoClass, boolean tight)
ResultList	calculateAllWcrt () throws AbortedException
void	checkAlgorithmCompatibility () throws IncompatibleDataException
void	doIterationStep () throws AbortedException
double	dyn_O (Task tau_ab)
Map< Task, Double >	getDynJitter ()
Map< Task, Double >	getDynWcrt ()
double	R (Task tauUA) throws AbortedException
double	R_b (Task tau_ab)
void	requestAbort (boolean abort)
void	updateJitter ()
Static Public Member Functions
static TransactionSet	createCompatibleGamma (TransactionSet ori)
static void	main (String...args) throws IOException, AbortedException
Protected Member Functions
AbstractAlgorithm	createInnerAlgorithm ()
double	dyn_J (Task tau_ab)
boolean	isConvergent ()
Task	previousTask (Task tau_ab)
Protected Attributes
Map< Task, Double >	dynJitter = new LinkedHashMap<Task, Double>()
Map< Task, Double >	dynWcrts = new LinkedHashMap<Task, Double>()
AbstractAlgorithm	innerAlgo
final Class<?extends AbstractAlgorithm >	innerAlgoClass
Map< Task, Double >	oldJitter
Map< Task, Double >	oldWcrts
TransactionSet	workingGamma
Private Member Functions
ResultList	toResultList ()
Static Private Attributes
static final boolean	USE_ORIGINAL_SEQUENCE_PREDECESSOR = false

Detailed Description

This class implements the base functionality of the WCDO algorithm. As a difference to the WCSO and exact algorithm, the WCDO algorithm allows as well the usage of precedence relations as the usage of multiple resources (e.g. processors).

The WCDO algorithm calculates temporary offsets according to the resource IDs and the precenence relations and uses a "lower" algorithm for the calculation of the WCRT. Then it calculates / adjusts a dynamic jitter and iterates the calculation until a fix point is reached.

As "lower" algorithm the WCSO or the exact algorithm can be used, and the WCSO algorithm can also be used in tight mode.

Author:: shanno

Constructor & Destructor Documentation

de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.AlgorithmWcdo	(	TransactionSet	gamma,
		Class<?extends AbstractAlgorithm >	baseAlgoClass,
		boolean	tight
	)

Parameters:

	gamma	The TransactionSet to use.
	baseAlgoClass	The class of the Algorithm to use in base calculation. Allowed base algorithms: AlgorithmWcso, AlgorithmExact and the cached variants of them.
	tight	When using AlgorithmWcso, this parameter determines whether a tight implementation is used.

Member Function Documentation

ResultList de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.calculateAllWcrt ( ) throws AbortedException

Reimplemented from de.fraunhofer.isst.axbench.timing.algorithms.AbstractAlgorithm.

void de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.checkAlgorithmCompatibility ( ) throws IncompatibleDataException [virtual]

Implements de.fraunhofer.isst.axbench.timing.algorithms.AbstractAlgorithm.

static TransactionSet de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.createCompatibleGamma ( TransactionSet ori ) [static]

Creates a compatible TransactionSet by creating predecessor relation according to task order. Using this method, you can use this WCDO algorithm with WCSO test data

AbstractAlgorithm de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.createInnerAlgorithm ( ) [protected]

Creates a instance of the given base algorithm (innerAlgoClass) via reflection. This is a dynamic version of the term "new AlgorithmXXX(gamma, isTight())".

Reimplemented in de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdopsPlusOuter, and de.fraunhofer.isst.axbench.timing.algorithms.cache.CachedAlgorithmWcdopsPlusOuter.

void de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.doIterationStep ( ) throws AbortedException

Iteration step of calculating all R values

double de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.dyn_J ( Task tau_ab ) [protected]

calculates the dynamic jitter, used by jitter update

double de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.dyn_O ( Task tau_ab )

Calculates the dynamic offset using the best case response times of the predecessors according to eq. 3 in WCDOPS+ paper

Reimplemented in de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdoSL.

Map<Task, Double> de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.getDynJitter ( )

Map<Task, Double> de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.getDynWcrt ( )

boolean de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.isConvergent ( ) [protected]

Reimplemented in de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdopsPlusOuter.

static void de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.main ( String... args ) throws IOException, AbortedException [static]

Task de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.previousTask ( Task tau_ab ) [protected]

double de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.R ( Task tauUA ) throws AbortedException [virtual]

Implements de.fraunhofer.isst.axbench.timing.algorithms.AbstractAlgorithm.

double de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.R_b ( Task tau_ab )

Best case response time, used for calculation of initial O and dynamic J

Reimplemented in de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdoSL.

void de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.requestAbort ( boolean abort )

Reimplemented from de.fraunhofer.isst.axbench.timing.algorithms.AbstractAlgorithm.

ResultList de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.toResultList ( ) [private]

void de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdo.updateJitter ( )

Creates a temporary transaction set for executing base algorithm. Dynamic offsets are calculated by the best case execution time of the previous task. The dynamic jitter is also filled.

Reimplemented in de.fraunhofer.isst.axbench.timing.algorithms.AlgorithmWcdopsPlusOuter.