I call cyg_thread_delay() in my I2C driver at very beginning in the context
of main(), I think scheduler should switch to another thread(at least
"idle" thread) and switch back to main() when delay time expired.
But I found after calling cyg_thread_delay(), main() thread remains in
"SLEEPING" state.
When later other threads are resumed, if context switch happens again,
main() thread will never be switched back (because it is in "SLEEPING").
So I debugged it, I found when I call cyg_thread_delay(), main() is NOT
switched. So the cyg_thread_delay() returns immediately, and main() is not
switched back through normal wake() function, and so the thread's state is
not changed to "RUNNING".
Why doesn't the main() be switched to the "idle" thread?
I think when I call cyg_thread_delay() in driver, the scheduler is locked
in some early caller.
So is there a limitation that: the cyg_thread_delay() should not be called
when the scheduler is locked?
_________________________________________________________________
免费下载 MSN Explorer: http://explorer.msn.com/lccn/
--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss
