Hi Guilherme,
I disagree, even if in most cases this is not desirable, I think if
someone wants to do this and they know what the implications are, it
should not be forbidden.
After all, putting a scheduler into an interrupt handler would be
frowned upon by many purist, but this is what is making RTL/RTAI work
for us all right now.
Regards, Stuart
Guilherme Nelson F De Souza wrote:
>
> Adam,
>
> I think this has already been discussed a lot in this list,
> but it's never too much: It's not a good practice to "stuff" a
> interrupt handler with "complicated" code, e.g. floating point
> calculations. As a good thumb rule: If your program design re-
> quires floating point at the handler level, there is something
> wrong with your design or with you. :-)
>
> In *your* case, I believe it's the design. ;-)
>
>
> Take care.
>
> gnd
>
>
> >Date: Wed, 11 Oct 2000 05:00:59 +0200
> >From: Paolo Mantegazza <[EMAIL PROTECTED]>
> >Subject: Re: [rtl] Simple floating point question
> >
> >Adam Meyerowitz wrote:
> >>
> >> Hello All,
> >>
> >> I am using RTL 2.0 pre 2 with a 2.2.13 kernel (yes, I know I should upgrade
> >> but it suits me just
> >> fine at the time being). I have a very simple module that has a
> >> init_module, cleanup_module and
> >> an interrupt handler. I would like to do some floating point calculations
> >> in my interrupt handler.
> >
> >Go to the RTAI home site at: http://www.aero.polimi.it/projects/rtai/,
> >click your mouse on "Hardware FPU support in interrupt handler" and
> >you'll learn how to do it in a snap.
> >
> >Ciao, Paolo.
> >-- [rtl] ---
> >To unsubscribe:
> >echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
> >echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
> >---
> >For more information on Real-Time Linux see:
> >http://www.rtlinux.org/rtlinux/
> >
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
---
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
