From owner-freebsd-bugs Mon Aug 25 14:00:04 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id OAA09444 for bugs-outgoing; Mon, 25 Aug 1997 14:00:04 -0700 (PDT) Received: (from gnats@localhost) by hub.freebsd.org (8.8.7/8.8.7) id OAA09409; Mon, 25 Aug 1997 14:00:01 -0700 (PDT) Date: Mon, 25 Aug 1997 14:00:01 -0700 (PDT) Message-Id: <199708252100.OAA09409@hub.freebsd.org> To: freebsd-bugs Cc: From: Steve Passe Subject: Re: kern/4382: CURRENT kernel has a "free vnode isn't" panic Reply-To: Steve Passe Sender: owner-freebsd-bugs@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk The following reply was made to PR kern/4382; it has been noted by GNATS. From: Steve Passe To: mark@grondar.za Cc: FreeBSD-gnats-submit@FreeBSD.ORG Subject: Re: kern/4382: CURRENT kernel has a "free vnode isn't" panic Date: Mon, 25 Aug 1997 14:51:14 -0600 Hi, this is a know problem, John is working on it and says: From: "John S. Dyson" Subject: Re: SMP seem to introduce some strange things... Date: Mon, 25 Aug 1997 05:35:05 -0500 (EST) I have been working on the problem (and alot of others regarding both UP and SMP in the VFS code.) Rather than hacking a fix, I finally decided to rework the code correctly. If I don't have a good fix by tonight (say, 21:00 EST) I'll back out my changes -- and we will have a proper fix within a few days. I don't want to hold up other's progress much longer. The management of the vnodes and VM interface, and associated reference counts is pretty tricky (and currently wrong.) PHK is working on the general vnode and namecache management (which is intricate in it's own right), but I am working the VM interface and making it fine grained SMP friendly. Currently, it isn't. -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD