Date: Mon, 17 Sep 2012 10:08:38 +0200 From: Borja Marcos <borjam@sarenet.es> To: Mike A <mikea@mikea.ath.cx> Cc: freebsd-scsi@freebsd.org Subject: Re: Bug Report: IBM x3650M4 (32GB, 2x4-core Xeon E5-2600, IBM ServeRaid M5110e): fails in install with NMI Message-ID: <A70FA423-E4E8-41E4-8C5D-C092EC2F56BE@sarenet.es> In-Reply-To: <20120914164603.GA34637@mikea.ath.cx> References: <20120827203817.GB44988@mikea.ath.cx> <201208281238.48041.jhb@freebsd.org> <20120828210618.GD69985@mikea.ath.cx> <201208290818.20990.jhb@freebsd.org> <20120914164603.GA34637@mikea.ath.cx>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sep 14, 2012, at 6:46 PM, Mike A wrote: > All the stars finally moved to the right places, and I was able to try booting > from a 9.1 amd64 memstick made from FreeBSD-9.1-RC1-amd64-memstick.img. That > failed, both without and with boot loader hints. > > I had a movie camera running to catch the console message traffic. The last > normal messages on the screen have to do with device mfi0. Here is the last > screenfull: I had a similar issue on a Dell R720 with 9.0-RELEASE. Turns out it was a problem with the built-in RAID card. Currently there seems to be some sort of overlap or confusion between the mpt, mps and in your case mfi driver? In my case, the card was assigned the mpt driver on 9.0-RELEASE, only to crash with a NMI. With 9.1RC, however, the system assigned the mfi driver and it worked. If I recall _correctly_, the mps driver seems to be better for some disk controllers that were handled by mfi. I would try to compile a system without the mfi driver and see if either mpt or mps take over and work. It's a long shot, I know, but... Borja.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?A70FA423-E4E8-41E4-8C5D-C092EC2F56BE>