The "atomic" statement is non-trivial. - Cory

On 11/15/06, Michael Schippling <[EMAIL PROTECTED]> wrote:
> If you look at system/LedsC.nc and you will see it does more than just
> call TOSH_xxx.
>
I just did that and the _more_ is setting a bit in a local variable
(which is actually unnecessary because one _could_ just read back the
value of the pin), but this should amount to about 3 instructions
plus the function call overhead. Assuming that SET and CLR are single
instructions I count maybe 6 @ 1us... I guess I should go look at the
generated ASM and see what's using the other 2/3 of the time...
MS



Cory Sharp wrote:
> On 11/15/06, *Chong Hui Kim* <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED] >> wrote:
>
>
>     Dear all,
>
>     Now, I have tried to measure the execution time of several modules.
>     To measure exeuction time, I used the logic analyzer.
>
>     But, I found some strange things.
>
>     1. When I called Leds interface sequentially as below, I found that
>     it took a time about 20us.
>                  call Leds.redOn();
>                  call Leds.redOff();
>         But I found that it took a time about 2us when I called TOSH_SET
>     and TOSh_CLR.
>                  TOSH_SET_RED_PIN();
>                  TOSH_CLR_RED_PIN();
>         In my thought, it took much time since Leds interface used
>     operating system.
>
>         Why does two methods task a different elapsed time?
>
>
> If you look at system/LedsC.nc and you will see it does more than just
> call TOSH_xxx.
>
>     2. When I call send ZigBee TX in OscilloscopeRF application, I found
>     that it took a time about 4-8ms between call Send and event sendDone.
>     There are two questions. 1) Why is the difference of transmission
>     time large? 2) I think that it takes a long time to transmit packet
>     about 4ms. If I want to synchronize two modules, I think that I must
>     send packet for synchronization with short transmission time. Why is
>     transmission time large?
>
>
> The delay is a combination of the backoff timer and clear channel
> assessment.
>
> Cory
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Tinyos-help mailing list
> [email protected]
> https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

_______________________________________________
Tinyos-help mailing list
[email protected]
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to