interrupts...

[Eric Valette]

Gregory.D.Menke.1 at gsfc.nasa.gov wrote:

> 1.  Are the high level C rtems_interrupt_* routines sufficient to set
>     up an interrupt handler on one of the Raven timers?  If not, I
>     assume a couple low level calls are needed to set up the hardware,
>     then a couple high level ones to get the isr running.  Is this
>     correct?

Provided they are routed through the openpic (raven) yes the
rtems_interrupt_* routines code is sufficient.

> 
> 2.  What is the relationship between an interrupt level, vector and
>     irq on this bsp- (or for RTEMS in general if relevant)?  On a
>     related note, what degree of correspondence exists between the bsp
>     vector and the RTEMS vector?  These issues are only tangentially
>     introduced in the docs and none of the example code even begins to
>     present the topic.  I'm missing some basic knowledge of RTEMS
>     architecture, which makes wading thru source extremely
>     inefficient.

The level is required for a software implementation of interrupt
priorities for programmable controller that do not have this feature
build-in...

I think the raven does manage priorities but then you have to deal also
with PC legacy interrupt priorities that are routed on a songle raven
interrupt source...

Then you should probably use a commercial OS and pay huge royalties fees
to have a more high level user oriented docs :-)
 
> 3.  After I get our applications together, I would be extremely
>     interested in doing a basic RTEMS Survival Guide targeted to those
>     who find getting started with RTEMS hard going.  If people think

-- eric