X-MimeOLE: Produced By Microsoft Exchange V6.5
Received: by onstor-exch02.onstor.net 
	id <01C7B9CA.1F6F4210@onstor-exch02.onstor.net>; Thu, 28 Jun 2007 13:20:16 -0800
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Content-class: urn:content-classes:message
Subject: RE: compac flash LEDs
Date: Thu, 28 Jun 2007 13:20:14 -0800
Message-ID: <BB375AF679D4A34E9CA8DFA650E2B04E0451E0CD@onstor-exch02.onstor.net>
In-Reply-To: <20070627110642.38e445ab@ripper.onstor.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: compac flash LEDs
Thread-Index: Ace45ewp+xZ6l9IlQYq6DJDEN0wQ+wA4tdLg
References: <20070626144241.58dc4fcc@ripper.onstor.net> <BB375AF679D4A34E9CA8DFA650E2B04E0451D575@onstor-exch02.onstor.net> <20070626164941.0455efcc@ripper.onstor.net> <BB375AF679D4A34E9CA8DFA650E2B04E0451D6F2@onstor-exch02.onstor.net> <20070627110642.38e445ab@ripper.onstor.net>
From: "Brian Stark" <brian.stark@onstor.com>
To: "Andy Sharp" <andy.sharp@onstor.com>

Andy,

Since the interrupts are level sensitive, the RM9K thinks the interrupt
is always active until it's serviced.  This is probably why Linux
eventually disables that IRQ.  Also, we don't currently use CSC
interrupts in BSD and only use functional interrupts that come straight
from the CF and are then routed by the controller to the interrupt pin
to the RM9K.  Are you enabling CSC interrupts?

You could try disabling functional interrupts by writing to the ExCA
interrupt register and disabling interrupts.  For example, for socket A,
set ExCA 0x3 to 0x60 instead of 0x64.  This will turn functional
interrupts off, and you shouldn't see Linux disable that IRQ.  If you
do, then it's CSC interrupts that are causing the issue, and you could
then try turning those off.


Brian



> -----Original Message-----
> From: Andy Sharp=20
> Sent: Wednesday, June 27, 2007 11:07 AM
> To: Brian Stark
> Subject: Re: compac flash LEDs
>=20
> On Tue, 26 Jun 2007 20:28:35 -0700 "Brian Stark"
> <brian.stark@onstor.com> wrote:
>=20
> > My comments below...
> >=20
> > =20
> >=20
> > > -----Original Message-----
> > > From: Andy Sharp
> > > Sent: Tuesday, June 26, 2007 4:50 PM
> > > To: Warren Gale
> > > Cc: Brian Stark
> > > Subject: Re: compac flash LEDs
> > >=20
> > > On Tue, 26 Jun 2007 15:35:09 -0700 "Warren Gale"
> > > <warren.gale@onstor.com> wrote:
> > >=20
> > > > Andy,
> > > >    See inline  < >
> > > > Warren
> > > >=20
> > > > -----Original Message-----
> > > > From: Andy Sharp
> > > > Sent: Tuesday, June 26, 2007 2:43 PM
> > > > To: Warren Gale
> > > > Cc: Brian Stark
> > > > Subject: compac flash LEDs
> > > >=20
> > > > As opposed to compact flash LEDs.
> > > >=20
> > > > Q1:
> > > >=20
> > > > Anyway, just curious who/what handles the LEDs by the compact=20
> > > > flash. Does the PROM do that?
> > > >=20
> > > > <PROM does NOT do anything with these LED's>
> > > >=20
> > > >  Could the OS do that?=20
> > > >=20
> > > > <U bet, all yours>
> > >=20
> > > that's funny
> > > now tell me how to program them, you bastaaaaaaaaaaaaards!
> >=20
> > Ah, LEDs are my favorite topic.  Chassisd controls the CF LEDs, and=20
> > the definition is as follows:
>=20
>=20
> Say no more, you have made me a happy man.  Since chassisd=20
> has been ported already, the LEDs are "done" ~:^)
>=20
> > > >=20
> > > > Q2:
> > > >=20
> > > > Is the pd6729 interrupt(s) edge or level?  Are they any
> > > different from
> > > > the natsemi interrupts?  We don't use them for OpenBSD, so I'm=20
> > > > breaking new ground here.
> > > >=20
> > > > < U mean the Intel pd6729 and the TI1520?
> > > >   PROM does very little with this and I don't know off hand.
> > > > Sorry>
> > >=20
> > > No, I mean the Cirrus/Intel pd6729 and the Natsemi=20
> dp83816 ethernet=20
> > > controllers, that also have interrupts on them.
> > >
> >=20
> > With the exception of the NMI, all interrupts to the RM9K are level=20
> > sensitive.
> >=20
> > Also, I am almost positive that the Natsemi interrupts are used in=20
> > OpenBSD.  I specifically added these to the RM9K to INT6_L=20
> and INT7_L=20
> > on the Rev 2 Bobcats.  The Rev 1 Bobcats were retired because they=20
> > didn't have this interrupt support.
>=20
> I'm sure the natsemi ints are used in BSD, it's the Cirrus=20
> ints that aren't used.  I just wanted to know if the Cirrus=20
> interrupts were level interrupts like the natsemi.  Now I=20
> know so that's useful information.  But of course, I want to=20
> use interrupts with the CF so Linux doesn't have to poll.=20
> What happens is, I enable the interrupts, and it just starts=20
> interrupting continuously.  I can't seem to "shut them off."=20
> Reading the CSC register doesn't deactivate them. There are=20
> no events but it just continues to interrupt anyway. After a=20
> 1 million unhandled interrupts, the kernel deactivates irq 10=20
> and I'm stuck. Any thoughts?
>=20
