Date: Thu, 4 Mar 1999 12:20:02 -0800 (PST) From: "Kenneth D. Merry" <ken@plutotech.com> To: freebsd-bugs@FreeBSD.org Subject: Re: kern/10281: Crash of 3.1-STABLE system due to scsi error. Message-ID: <199903042020.MAA31910@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
The following reply was made to PR kern/10281; it has been noted by GNATS. From: "Kenneth D. Merry" <ken@plutotech.com> To: marijn@gewis.win.tue.nl (Marijn Meijles) Cc: mjacob@feral.com, freebsd-gnats-submit@freebsd.org Subject: Re: kern/10281: Crash of 3.1-STABLE system due to scsi error. Date: Thu, 4 Mar 1999 13:16:08 -0700 (MST) Marijn Meijles wrote... > You wrote: > > > > Umm, I know nothing about this. If there is a bug in the SCSI subsystem > > that causes crashes, why haven't I heard about it? > > > > And which machine at "cdrom.com" is crashing? > > > I think wcarchive, but it was during a party that I spoke with the guy, > so it was all a bit 'clouded'. I'll mail him and get back to you with > some more details. Yeah, I'd like to hear about it. If it is wcarchive, David Greenman has been uncharacteristically silent on it. > > > I think we have the same problem, do you know more about this? > > > The error message is: > > > devstat_end_transaction: HELP!! busy_count for da2 < 0 (-245774) > > > biodone: buffer already done > > > > > > the busy counts gets more and more negative. > > > > Now that is interesting. That indicates that, somehow or another, > > transactions are getting done more than once. > > > > Which version of sys/cam/scsi/scsi_da.c do you have? Can you mail me the > > file? > > > ok, I'll mail it seperately. It looks normal enough. I'm not sure what could be going on. I haven't seen this sort of problem before. > > If you're getting crashes, it would help immensely if you could provide a > > stack trace from the panic. > > > Yup, I know. I've instructed the guys at the university to write down > everything, but the crash of today just showed this and no panic. you > could only ping and it had to be rebooted. Hmm, sounds like it may just be a hang and not a panic. It's possible that it could be a bug in the aha driver or something. That driver hasn't been tested extensively. Some portion of the system is causing transactions to get completed more than once. Ken -- Kenneth Merry ken@plutotech.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199903042020.MAA31910>