BSP_install_rtems_irq_handler API

[Till Straumann]

VALETTE Eric wrote:

> Thomas Doerfler wrote:
>
> en libbsp and libcpu code/preprocessor statements.
> >
> > I am currently moving the helas403 BSP and PPC403/405 CPU code
> > to the new exception handling, and I really have a split
> > opinion on the new API. I think that some part are really
> > nice, like to "irq_on" and "irq_off" functions in the drivers,
> > but you are right, not passing the vector numbers to the irq
> > function is really one design flaw among others (IMHO):
>
> No. You can do int by adding your own vector that push the argument and
> then call a geberic function. In that case you pay the cost because you
> need it but not every user that do not need it. Cost in term of
> instruction is exactly the same.
>

Sorry, I don't quite agree - having every driver that supports more
than one device coding multiple wrappers or even multiple ISRs
(as is the case for libbsp/powerpc/shared/uart.c:) is IMHO highly
unelegant. On the other hand, passing at least the vector
number (besides: that's what i386/shared/irq.c already does)
hardly adds _any_ overhead because it's already sitting in a
register anyway.

>

-- snip --

> Starting the discussion is fine but do not try to reinvent the whell and
> read the mailling lists archive...

Sorry if I miss something but browsing the archive from 1/1999 until today
(search seems not to work), I only found one thread (MPC8xx interrupts,
10/2001) where you also state that this had already been discussed many
times.

I'd appreciate a pointer to the thread where the pro's and con's of this
issue (the "new" IRQ API) have been thoroughly discussed.

The reason why I bring this up is that I have to decide upon an API to
adopt for VME interrupts (a driver for the Universe II will be available
soon)
and for my SVGM5 BSP (PPC G4 board by Synergy microsystems).

-- Till.

>
>
> --
>     __
>    /  `                          Eric Valette - Canon CRF
>   /--   __  o _.                 Product Dev. Group Software Team Leader
> (___, / (_(_(__                 Rue de la touche lambert
>                                  35517 Cesson-Sevigne  Cedex
>                                  FRANCE
> Tel: +33 (0)2 99 87 68 91       Fax: +33 (0)2 99 84 11 30
> E-mail: valette at crf.canon.fr    http://www.crf.canon.fr