rtems on m5282lite

Joel Sherrill <joel@OARcorp.com> joel.sherrill at OARcorp.com
Fri Apr 15 18:21:05 UTC 2005


g_YaneV wrote:
>   Hello again,
> 
>  Problem solved.
> 
> I insert this code in start.S just before the code that copy the new vector 
> table.
> 
>     lea.l __IPSBAR+0xc08,a0
>     move.l #0xfffffffe,(a0)
>     lea.l __IPSBAR+0xc0c,a0
>     move.l #0xffffffff,(a0)
>     lea.l __IPSBAR+0xd08,a0
>     move.l #0xffffffff,(a0)
>     lea.l __IPSBAR+0xd0c,a0
>     move.l #0xffffffff,(a0)
> 
> This way I am sure that all interrupts are unmasked by rtems code. Isn't this 
> a good practice in general(no one can know what is running befor calling 
> rtems). My problem was that dBUG allows one timer and rtems after that allows 
> another. 

Do you mean masked or unmasked?  All interrupt sources should be cleared 
and disabled before rtems_initialize_executive is executed.  RTEMS 
enables CPU interrupts when switching to the first task/thread so you 
don't want any spurious interrupt sources outstanding.

> thanks for showing me where the problem can be.


-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel at OARcorp.com                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
    Support Available             (256) 722-9985




More information about the users mailing list