Date: Wed, 22 Jul 1998 01:27:21 -0600 From: Steve Passe <smp@csn.net> To: "Jordan K. Hubbard" <jkh@time.cdrom.com> Cc: Bruce Evans <bde@zeta.org.au>, mike@smith.net.au, asami@FreeBSD.ORG, committers@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: Tree tagging put off by ~12 hours. Message-ID: <199807220727.BAA11941@Ilsa.StevesCafe.com> In-Reply-To: Your message of "Tue, 21 Jul 1998 22:24:11 PDT." <11820.901085051@time.cdrom.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi, > > > >Oh, BTW, freefall is taking NMI's now. No idea why except that it's a > > >fairly hot box (or was yesterday). It NMI'd while I was sitting at > > >the console yesterday and I simply continued it out of the debugger > > >again and then compiled a kernel with POWERFAIL_NMI so it wouldn't do > > >it again. > > > > Gack. FreeBSD mishandles NMIs (the NMI gate isn't an interrupt gate...). > > This normally doesn't cause any more problems than the correct response > > to an NMI (panic), but POWERFAIL_NMI prevents the correct response. Just my observation on NMIs in general, FreeBSD doesn't re-cock the PII chip correctly to take a subsequent NMI. I am using custom ISA hardware that generates an NMI to give us 'pseudo-realtime' response. To get it to work I had to add code to the NMI ISR to recock the PII chip to handle repeated NMIs. This is in the context of 2.2.6... -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199807220727.BAA11941>