To: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Subject: Re: posix-timers overrun broken?
MIME-Version: 1.0
Message-ID: <OFB812AC35.7E514449-ONC125727C.0029AA85-C125727C.002B65DA@ra.rockwell.com>
From: Milan Svoboda <msvoboda@ra.rockwell.com>
Date: Thu, 8 Feb 2007 08:55:01 +0100
Content-Type: text/plain; charset="US-ASCII"
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1340
Lines: 40

> On Wed, 2007-02-07 at 14:36 +0100, Milan Svoboda wrote:
> > But if the next interrupt arrives before function collect_signal is 
called 
> > to actually deliver the
> > siginfo_t to userspace, the si.overrun is cleared in posix_timer_event 
and 
> > we have just forgotten
> > one overrun...
> > 
> > Am I wrong?
> 
> Yes. posix_timer_event() is only called when the timer expires the first
> time. When the signal has been queued, the timer is not rearmed and the
> overrun is calculated in the actual signal delivery path, which calls
> do_schedule_next_timer().

Ok understand.

So it seems to me that drivers/char/mmtimer.c is wrong. It is
driver to additional hardware timers and mmtimer exports them as a posix 
timer.
When interrupt happens, they call posix_timer_event and set hardware 
interrupt
to the next tick. I use the same practise in my driver and found
that overruns are reported wrongly.

How to plug "external" source of ticks to the posix timers?

As far as I know current HRT doesn't allow us to use more sources of
ticks together, only one is selected, correct?

Milan


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/