Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!spool.mu.edu!sol.ctr.columbia.edu!math.ohio-state.edu!magnus.acs.ohio-state.edu!cis.ohio-state.edu!news.sei.cmu.edu!bb3.andrew.cmu.edu!andrew.cmu.edu!aw2t+ From: "Alex R.N. Wetmore" <aw2t+@andrew.cmu.edu> Newsgroups: comp.os.386bsd.bugs Subject: ed0 driver (was Re: NFS Crash) Date: Wed, 25 Aug 1993 07:55:30 -0400 Organization: Sophomore, Math/Computer Science, Carnegie Mellon, Pittsburgh, PA Lines: 27 Message-ID: <wgSpAmS00iV2016idL@andrew.cmu.edu> NNTP-Posting-Host: po2.andrew.cmu.edu In-Reply-To: <CCB1x9.43K@agora.rain.com> Excerpts from netnews.comp.os.386bsd.bugs: 25-Aug-93 Re: NFS crash (NetBSD 0.9) by David Greenman@agora.rai > The new driver does work with the 3c503, and the performance and > reliability with it exceed the 'ec' driver. As far as the over-run's that > you are seeing are concerned, the 'ec' driver experiances these as well - > it just doesn't do anything about it, which can lead to board lock-ups > and other not-so-nice behavior. And regarding the d0000 default for ed0, > nothing prevents you from setting this to anything you like. I use cc000 > myself so that I can switch between SMC 8013 boards and 3c503 boards > without rebuilding a different kernel. I had a lot of problems getting the ed0 driver setup with a 3c503 as well (using d0000, 250). I couldn't figure out how to configure the driver to use the aui port (instead of the thin port). It looked like a flags 0x01 in the config file would do it, but no such luck. Also, what sorts of crashes would I see from the ring buffer overflowing? A quick panic-reset or just having the machine die slowly (ie, things work, some things break, everything breaks, maybe over a 5 minute period). It did annoy me that the default NetBSD-0.9 kernel came with the ed0 driver installed in a way that the 3c503 didn't support, because I needed networking to get the kernel source. alex