Date: Sat, 17 Jul 2004 15:53:13 -0700 (PDT) From: Doug White <dwhite@gumbysoft.com> To: Kris Kennaway <kris@obsecurity.org> Cc: "current@freebsd.org" <current@freebsd.org> Subject: Re: panic: APIC: Previous IPI is stuck Message-ID: <20040717154952.U41727@carver.gumbysoft.com> In-Reply-To: <20040717195632.GA67466@xor.obsecurity.org> References: <20040717103509.N24726@alpha.siliconlandmark.com> <20040717181026.GA63963@xor.obsecurity.org> <20040717195632.GA67466@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 17 Jul 2004, Kris Kennaway wrote: > > Does it only happen with a constant $NCPU+ load average? I have never > > seen it here. > > Yeah, it's usually a pretty busy machine. It does a lot of > buildworlds, nfs, concurrent www serving and ssh dispatching. I triggered one of these last weekend, too. Peter mentioned something to me about it being caused by both CPUs IPIng at once and there no deadlock avoidance in that case. It died for me in installworld, though, which made for a messy fixup job. I guess our work on getting the kernel free-running is uncovering all sorts of fun issues. -- Doug White | FreeBSD: The Power to Serve dwhite@gumbysoft.com | www.FreeBSD.org
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040717154952.U41727>