[En-Nut-Discussion] Watchdog/Idle thread issues again

[Alexander Baranov]

Hi, I'm afraid that I haven't understand your application but I had similar
hang-up problem. I use interrupt from Timer 3 to restart external wdt and to
increment some TIMEOUT long variable while main thread zeroes it. If the
TIMEOUT grows to some predefined value ( i.e. main thread fails to reset it)
I do not restart  wdt in interrupt routine. Thus the reset timeout may be
extended to dozens seconds.
Alexander.
----- Original Message ----- 
From: "Ulrich Hertlein" <ulrich.hertlein at artcom.de>
To: "en-nut-discussion" <en-nut-discussion at egnite.de>
Sent: Wednesday, December 07, 2005 4:55 AM
Subject: [En-Nut-Discussion] Watchdog/Idle thread issues again


> Hi people,
>
> I'm struggling with fundamental idle thread/watchdog issues again and
> hopefully someone can shed some light on the problem.
>
> My application uses two threads:
> - an idle thread that checks and decrements a counter and issues a reset
> when the counter reaches zero.
> - a worker thread that polls I/O pins and sends out UDP packets. The
> worker thread periodically resets the counter.
>
> I thought this would save me from hangups but obviously it doesn't it
> the OS hangs and the idle thread never is scheduled.
>
> What's the way out here? I can't just use the hardware watchdog since
> its max. timeout is ~2s which is not enough for the network stack.
>
> The application runs fine for a few hours (~16 hours last time) and then
> hangs. Memory usage is stable.
>
> Thanks in advance,
> /ulrich
>
> -- 
> Ulrich Hertlein | Software Development
>
> ART+COM AG
> Kleiststr. 23-26 | 10787 Berlin | Germany
>
> phone: +49.30.21001-433
> fax: +49.30.21001-555
> http://www.artcom.de
> _______________________________________________
> En-Nut-Discussion mailing list
> En-Nut-Discussion at egnite.de
> http://www.egnite.de/mailman/listinfo.cgi/en-nut-discussion
>
>