From owner-freebsd-alpha Wed Jul 26 15:33:57 2000 Delivered-To: freebsd-alpha@freebsd.org Received: from feral.com (feral.com [192.67.166.1]) by hub.freebsd.org (Postfix) with ESMTP id BD89837BF1A for ; Wed, 26 Jul 2000 15:33:53 -0700 (PDT) (envelope-from mjacob@feral.com) Received: from semuta.feral.com (semuta [192.67.166.70]) by feral.com (8.9.3/8.9.3) with ESMTP id PAA02663; Wed, 26 Jul 2000 15:33:33 -0700 Date: Wed, 26 Jul 2000 15:32:49 -0700 (PDT) From: Matthew Jacob Reply-To: mjacob@feral.com To: Bernd Walter Cc: David Greenman , Andrew Gallatin , "Koster, K.J." , "'FreeBSD Alpha mailing list'" Subject: Re: fxp0 hangs my AXPpci33 In-Reply-To: <20000727002844.A35107@cicely8.cicely.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Well, I have a Multia I haven't fired up lately... maybe it's time... :-).... -matt On Thu, 27 Jul 2000, Bernd Walter wrote: > On Wed, Jul 26, 2000 at 01:25:22PM -0700, Matthew Jacob wrote: > > > > > >I think that the SEEPROM reading might be very timing > > > >dependant... David? I'm worried that whatever the path in the chipset > > > > > > It is timing dependant, but the driver inserts more than enough delay > > > by using DELAY(1) for 1us. I think the requirement is much much less than > > > that (< .5us), so if timing really is the problem, then it leads me to wonder > > > if DELAY() might be somewhat broken on Alpha for small delays. A quick way > > > to test for that would be to change the DELAY(1)'s to DELAY(10) or more. > > > > It shouldn't be. It's using the rpcc register to do more precise timing. It > > *used* to be about 30% off, but now should be quite accurate- but it's > > certainly possible. I'll be interested in hearing this. > > > > I forget- is this with 4.0 or with -current/4.1? I seem to recall that this > > source base is a recent source base change even for 4.1- (July 4th). > > To add something to the problem I had. > My problem was after changing a 4.0-current system (short time before > 4.0-RELEASE) from an axppci33 board to a PC164. > After I got this problems I updated to a recent STABLE and asked on the > alpha list. > Even the original bug was there before the card worked fine with the > axppci33 board. I used it with and with cache installed. > The difference is that my card is older than the 559 - I asume it's a 557. > > -- > B.Walter COSMO-Project http://www.cosmo-project.de > ticso@cicely.de Usergroup info@cosmo-project.de > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message