Date: Wed, 11 Aug 1999 00:04:20 +0200 From: Stefan Esser <se@zpr.uni-koeln.de> To: Matthew Jacob <mjacob@feral.com> Cc: scsi@FreeBSD.ORG, Stefan Esser <se@freebsd.org>, stefan.esser@o-tel-o.de Subject: Re: NCR/Multia Problems? Message-ID: <19990811000420.A587@dialup124.zpr.uni-koeln.de> In-Reply-To: <Pine.BSF.4.05.9908091341170.2325-100000@semuta.feral.com>; from Matthew Jacob on Mon, Aug 09, 1999 at 01:41:42PM -0700 References: <Pine.BSF.4.05.9908091341170.2325-100000@semuta.feral.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 1999-08-09 13:41 -0700, Matthew Jacob <mjacob@feral.com> wrote: > > nopilt.feral.com > ncr0:6: ERROR (a0:0) (0-0-0) (8/13) @ (mem > 4458e000:c0000004). You have DSTAT=0xa0, meaning DFE + BF. DFE is not an error condition (DMA fifo empty) but just an indicator of no ongoing DMA. But BF is Bus Fault, i.e. the DMA engine (or the instruction or indirect address fetch) aborted a transfer because of a bus error. I have no idea what can trigger that kind of failure in a multia. There are a number of conditions mentioned in the NCR data books. I can look them up tomorrow, if required. Anyway: This is an indication of a hardware problem. Either the chip-set doesn't deal correctly with multiple bus masters (and there was another bus-master active), or there was a problem with memory (invalid address touched, ...). The first problem existed with early PC chip-sets, but can be worked around in some cases (disable PCI buffers, PCI bursts, ...). But I never heard that the Multia had this fault. The second scenario may be caused by hardware or software. Either there is bad memory or access to memory is blocked for extended periods of time, or the address may have been wrong and the PCI transfer was not successful for that reason. > ncr0:6: ERROR (a0:0) (0-0-0) (8/13) @ (scripth 84:88080000). Same DMA status code (0xa0), but different SCRIPTS address. This points towards a hardware problem, IMHO. > ncr0:6: ERROR (a0:0) (0-0-0) (8/13) @ (scripth a0:c0000004). Again. I don't have time to investigate that further, right now. Please let me know, whether this problem was caused by a software upgrade or other configuration change, or whether you never had the NCR chip working on that particular box. Regards, STefan To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990811000420.A587>