[LinuxPPS] LinuxPPS 5.3.2 strange behaviour

[Rodolfo Giometti]

On Wed, Oct 01, 2008 at 07:04:07AM -0400, George Spelvin wrote:
> Rodolfo Giometti <giometti at enneenne.com> wrote:
> 
> > On Tue, Sep 30, 2008 at 06:25:43PM -0400, George Spelvin wrote:
> >>> On Tue, Sep 30, 2008 at 04:00:04PM -0400, George Spelvin wrote:
> >
> >>> The PPS_IRQ_EVENTS option is not so clear explained into the LinuxPPS
> >>> wiki... maybe me or someone else should write something about
> >>> it... any candidates? :)
> >> 
> >> I thought it just grabbed the timestamp early in the interrupt handler
> >> and passed it down.  That should WORK fine with a shared interrupt,
> >> it'll just add overhead when the other handler doesn't need the timestamp.
> >
> > No, the IRQ is grabbed so early that we cannot know which is the IRQ
> > source at all! That's why sharing the IRQ line is wrong.
> 
> Er... that's what I tried to say.  It's grabbed very early in the
> generic interrupt handler and passed to the device-specific ones.
> That adds overhead when the interrupt is for a device that doesn't
> need the timestamp, but it shouldn't stop it from functioning!

Ah, ok! Sorry, I misunderstood you. :)

Ciao,

Rodolfo

-- 

GNU/Linux Solutions                  e-mail: giometti at enneenne.com
Linux Device Driver                          giometti at linux.it
Embedded Systems                     phone:  +39 349 2432127
UNIX programming                     skype:  rodolfo.giometti