How are concurrent task starts handled?

[ewan]

If there are three tasks and one is running, and the other two tasks are scheduled (using o.s. timers) to start before the first is complete, which task starts next, and when?

What happens during "idle" time?

I have looked at the literature and the various posts and still have trouble visualizing the scheduler process. I do understand one-shot timers.

Thanks,

Ewan.

[ewan]

Well, I have read some more. Is it true that the Tasker runs all the tasks in "round-robin" fashion except those which delay themselves or which are waiting for an event?

Regards,

Ewan.

[mark]

Ewan

It may be best to look at the following post for more background to this:
http://www.utasker.com/forum/index.php?topic=59.msg235#msg235
since it hopefully explains all and the reasons behind it.

Some other links which may be eye-opens are:
http://tech.groups.yahoo.com/group/lpc2000/message/31341
http://www.nilsenelektronikk.no/nenesos.html
http://tech.groups.yahoo.com/group/lpc2000/message/31262
http://www.embedded.com/2000/0009/0009feat4.htm

Regards

Mark

PS. As stated in the other uTasker thread, the main criteria for the chosen operation is the ability to support the use of the project. Simplicity is essential. (Users') Bugs which cause a failure each time the system is used are generally not a big deal - they are quite obvious and support can usually help solve these with little effort. Bugs which cause a system failure on a sporadic basis (user has not anticipated a context switch or correctly protected resources) can cause important projects to fail or result in huge additional costs - support can be very difficult.