cParsimSynchronizer Class Reference
[Parallel simulation support]
Abstract base class for parallel simulation algorithms. More...
#include <cparsimsynchr.h>
Public Member Functions | |
| cParsimSynchronizer () | |
| virtual | ~cParsimSynchronizer () |
| virtual void | setContext (cSimulation *sim, cParsimPartition *seg, cParsimCommunications *co) |
| virtual void | startRun ()=0 |
| virtual void | endRun ()=0 |
| virtual cMessage * | getNextEvent ()=0 |
| virtual void | processOutgoingMessage (cMessage *msg, int procId, int moduleId, int gateId, void *data)=0 |
Detailed Description
Abstract base class for parallel simulation algorithms.
Subclasses implement the "null message algorithm" and others.
Task of this layer is to implement conservative synchronization, that is, block the execution at time points where it is not safe to proceed since events from other partitions will (may) arrive before the next local event.
Synchronizer integrates into the simulation kernel via the getNextEvent() and processOutgoingMessage() functions:
- getNextEvent() is the scheduler function (comes from cScheduler)
- processOutgoingMessage() is invoked when messages are sent out of the partition, from gates (see cProxyGate) of placeholder modules (cPlaceholderModule) via cParsimPartition.
Member Function Documentation
| virtual void cParsimSynchronizer::processOutgoingMessage | ( | cMessage * | msg, | |
| int | procId, | |||
| int | moduleId, | |||
| int | gateId, | |||
| void * | data | |||
| ) | [pure virtual] |
Hook, called when a cMessage is sent out of the partition.
It is provided here so that the synchronizer can potentially perform optimizations, such as piggybacking null messages (see null message algorithm) on outgoing messages.
Implemented in cISPEventLogger, cNullMessageProtocol, and cParsimProtocolBase.
The documentation for this class was generated from the following file:
