Re: how to generate external interrupts

["Frank Ch. Eigler" <fche at redhat dot com>]

Hi -

> [...]
> Kindly advise if my approach is correct. I edited the
> sid.conf as follows-
> new sid-sched-sim sensor-sched

This part is not necessary.  You could reuse the existing
sim-sched-sim instance - just use a new unused client number.
(If were to use a new scheduler instance, you would have to
connect its upstream "advance" pin to an appropriate source.)

> //then I set the num clients to 5...
> //then connected the sched event to the intrctrl...
> connect-pin sensor-sched 1-event -> intrctrl source-16
> //then I appropriately set the 'regular' and 'time'
> attributes.

All correct, I'm impressed!


> The problem is that I am unable to catch the
> interrupt.I am using eCos as the RTOS. [...]
> Also, when I checked the eCos header file
> (hal_intr.h), max interrupts were set to 8.If,possible
> could you kindly comment on this?

I am sorry, but I'm not familiar with that low-level aspect
of eCos.  I suspect choosing a lower interrupt number would
help, as might putting breakpoints into eCos functions for
setting interrupt masks and dispatching received interrupts.
If you suspect a sid (rather than eCos) problem, consider
monitoring the interrupt controller with tksm, and manually
sending interrupts from the pin view.


> P.S: my platform - ARM-PID.

This suggests that you should be using the an arm-cpu
configuration for sid that includes models for selected
peripherals of the pid board, like the sid/pregen/arm-pid.conf
file.  Sending a diff between that file and your modified
version would help diagnose problems in the future.


- FChE