rtems_iterate_over_all_threads

[Leon Pollak]

On Wednesday April 22 2009, Thomas Dörfler wrote:
> I strongly disagree. There are several reasons: On the one hand,
> changing the behaviour of the classic API may break several existing
> applications. 
This is correct. But I think that every designer making upgrade of RTEMS knows 
that he/she/ should check the application.
For example, switching to the new version of RTEMS changed a lot in the BSP 
API so, that mine stopped working at all.
This is the price of upgrading...


> On the other hand, I consider using nested suspend/resume
> to control the task scheduling as a bad practice which may lead to
> unexpected behaviour in bigger RTEMS systems. So I would not ecourage
> people to do so by making the API more feature-rich.
Thomas,
As I highly appreciate your qualification and experience, which are bigger and 
wider then mine, I am ready to accept your objection.
But why this option exists in other APIs?
And why do you think this is a bad practice?


Thanks.
Leon

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/users/attachments/20090422/b21979bf/attachment-0001.html>