From owner-freebsd-hackers Sun Oct 6 13:48:27 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id NAA10164 for hackers-outgoing; Sun, 6 Oct 1996 13:48:27 -0700 (PDT) Received: from mailhost.coppe.ufrj.br (root@cisigw.coppe.ufrj.br [146.164.2.31]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id NAA10157 for ; Sun, 6 Oct 1996 13:48:24 -0700 (PDT) Received: (from jonny@localhost) by mailhost.coppe.ufrj.br (8.7.5/8.7.3) id SAA07079; Sun, 6 Oct 1996 18:48:10 -0200 (EDT) From: Joao Carlos Mendes Luis Message-Id: <199610062048.SAA07079@mailhost.coppe.ufrj.br> Subject: Found a BUG and partial solution (was: nullfs: Panicing...) To: adam@veda.is (Adam David) Date: Sun, 6 Oct 1996 18:48:10 -0200 (EDT) Cc: jonny@mailhost.coppe.ufrj.br, freebsd-hackers@freebsd.org In-Reply-To: <199610062030.UAA14571@veda.is> from Adam David at "Oct 6, 96 08:30:07 pm" X-Mailer: ELM [version 2.4ME+ PL14 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk #define quoting(Adam David) // > // >PS: I'm now using NFS to localhost, as a workaround. // > // // > // I had to abandon this approach and use symlinks instead, once the directory // > // had outgrown a certain size. Perhaps there is no longer any such size limit, // > // but I don't have the stomach at the moment to risk it. // > // > I cannot symlink to anon ftp. It's chrooted, remember ? :) // // You can symlink in the other direction, but perhaps that is not appropriate // for your particular situation. It might at least be acceptable, though less // than convenient. Disk Space, the final frontier. These are the voyages of the StarShip BSD. :-) Analysing the sources I found the (first) point of failure: Seens to me that there's a line missing near the end of nullfs_mount() calling nullfs_statfs( mp, &mp->mnt_stat, p ); It's strange, since nullfs_statfs() has a check to verify if it has been called this way... The problem was caused by a division by zero at vnode_pager_haspage(), line 313: block = offset / vp->v_mount->mnt_stat.f_iosize; ^^^^^^^^ But yet, it still does not work. Now it's a problem of page fault while in kernel mode. My knowledge of BSD internals is not enough to find the problem anymore... Jonny PS: I'm talking about the sources of 2.1.5-RELEASE -- Joao Carlos Mendes Luis jonny@gta.ufrj.br +55 21 290-4698 ( Job ) jonny@cisi.coppe.ufrj.br Network Manager UFRJ/COPPE/CISI Universidade Federal do Rio de Janeiro