This: (fhpd) go (fhpd) Breakpoint 0 main The code like: actionpoint = bpManager.addFunctionBreakpoint(breakpt, die); actionpoint.addObserver(new CLIBreakpointObserver() { public void updateHit(final SourceBreakpoint bpt, Task task, long address) { // See comment in case above. Manager.eventLoop.add(new Event() { public void execute() { FunctionBreakpoint fbpt = (FunctionBreakpoint) bpt; outWriter.print("Breakpoint "); outWriter.print(fbpt.getId()); outWriter.print(" "); outWriter.println(fbpt.getName()); in BreakpointCommand, it should instead notify the stepping engine that a breakpoint has occured so that the stepping engine can stop it (and possibly other) threads; and then notify the cli that things have stopped due to a breakpoint.
-> tthomas@redhat.com
Synchronize fhpd, stepping engine and breakpoints. frysk-core/frysk/hpd/ChangeLog: 2008-06-06 Teresa Thomas <tthomas@redhat.com> * BreakpointCommand.java (*.updateHit): Use SteppingEngine.updateActionpointDoneEvent. * TestBreakpoints.java (testGoAfterSimultaneousBreaks): New. frysk-core/frysk/stepping/ChangeLog: 2008-06-06 Teresa Thomas <tthomas@redhat.com> * SteppingEngine.java (updateActionpointDoneEvent): New. (blockedByActionPoint): Use updateActionpointDoneEvent. (blockedByActionPoint(Task,TaskObserver)): Test if messages set empty.