From owner-freebsd-hardware  Wed Feb  5 11:33:58 1997
Return-Path: <owner-hardware>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id LAA04530
          for hardware-outgoing; Wed, 5 Feb 1997 11:33:58 -0800 (PST)
Received: from Sisyphos.MI.Uni-Koeln.DE (Sisyphos.MI.Uni-Koeln.DE [134.95.212.10])
          by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id LAA04515;
          Wed, 5 Feb 1997 11:33:47 -0800 (PST)
Received: from x14.mi.uni-koeln.de (annexr3-8.slip.Uni-Koeln.DE) by Sisyphos.MI.Uni-Koeln.DE with SMTP id AA11238
  (5.67b/IDA-1.5); Wed, 5 Feb 1997 20:33:32 +0100
Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.5/8.6.9) id UAA05581; Wed, 5 Feb 1997 20:33:24 +0100 (CET)
Message-Id: <19970205203256.YE44873@x14.mi.uni-koeln.de>
Date: Wed, 5 Feb 1997 20:32:56 +0100
From: se@freebsd.org (Stefan Esser)
To: msmith@atrad.adelaide.edu.au (Michael Smith)
Cc: Brett_Glass@infoworld.com, se@freebsd.org, Andre.Albsmeier@mchp.siemens.de,
        freebsd-hardware@freebsd.org
Subject: Re: 5 devices in 4 PCI Slots ?
References: <9701038549.AA854983543@ccgate.infoworld.com> <199702040003.KAA07826@genesis.atrad.adelaide.edu.au>
X-Mailer: Mutt 0.60-PL0
Mime-Version: 1.0
In-Reply-To: <199702040003.KAA07826@genesis.atrad.adelaide.edu.au>; from Michael Smith on Feb 4, 1997 10:33:24 +1030
Sender: owner-hardware@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

On Feb 4, msmith@atrad.adelaide.edu.au (Michael Smith) wrote:
> Brett_Glass@infoworld.com stands accused of saying:
> > Interesting. Is interrupt sharing implemented for other drivers as well
> > (for instance, for PCMCIA multifunction cards that share an IRQ between an
> > Ethernet NIC and a modem)?
> 
> Interrupt sharing is a bus, not driver, issue.  We've just done to death

Well, it is a BUS + DRIVER issue, actually ...

The BUS must physically support it (some kind of 
wired OR, or dedicated interrupt lines per slot 
fed into the PIC chip).

The inputs must be level sensitive (it is possible
to share edge triggered interrupts, but this does
result in higher per interrupt overhead).

And finally, the device driver must also support 
interrupt sharing. Its first action in the interrupt
handler must be a test, whether THIS device was the
cause of the interrupt, and an immediate return from
the handler, if it was not ...

Regards, STefan