On Oct 22, 2009, at 2:54 PM, Wenjie Zeng wrote:
> Hi:
>
> I've come across a strange missing startDone() problem after I
> receive SUCCESS after calling
> CC2420ActiveMessageC.SplitControl.start() to start the radio. This
> contradicts the documentation which states that the system MUST
> signal a startDone() if it returns SUCCESS in start(). Then I looked
> into CC2420CsmaP.nc which is the implementation of CC2420 control
> and found the following code snippet:
>
> /***************** SplitControl Commands ****************/
> command error_t SplitControl.start() {
> if(call SplitControlState.requestState(S_STARTING) == SUCCESS) {
> call CC2420Power.startVReg();
> return SUCCESS;
>
> } else if(call SplitControlState.isState(S_STARTED)) {
> return EALREADY;
>
> } else if(call SplitControlState.isState(S_STARTING)) {
>
> return SUCCESS;
> }
>
> return EBUSY;
> }
>
> If the control flow falls into the first if statement, a
> startDone_task() will be posted later on. However, if the radio is
> somehow already in the state of S_STARTING, as in the third if
> statement, the system will return SUCCESS without posting the task
> startDone_task() which will signal the startDone() event. I guess
> the assumption here is that someone else's earlier call to start()
> must have triggered startDone_task() to be posted and as a result
> the callee will still receive startDone(). I guess this assumption
> is somehow INCORRECT.
That is the assumption. It would be good to know what causes it to be
incorrect.
Phil
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
