From owner-freebsd-current Thu Jul 22 14:55:14 1999 Delivered-To: freebsd-current@freebsd.org Received: from alcanet.com.au (border.alcanet.com.au [203.62.196.10]) by hub.freebsd.org (Postfix) with ESMTP id DCAF114F29 for ; Thu, 22 Jul 1999 14:54:59 -0700 (PDT) (envelope-from jeremyp@gsmx07.alcatel.com.au) Received: by border.alcanet.com.au id <40356>; Fri, 23 Jul 1999 07:35:10 +1000 Date: Fri, 23 Jul 1999 07:53:39 +1000 From: Peter Jeremy Subject: `stale FS getpages' messages To: current@FreeBSD.ORG Message-Id: <99Jul23.073510est.40356@border.alcanet.com.au> Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I've just received the following kernel messages (running a kernel from cvs-cur 5476 - about last Thursday): vnode_pager: *** WARNING *** stale FS getpages No strategy for buffer at 0xc1c03aa8 : 0xc6168380: type VREG, usecount 3, writecount 0, refcount 0, flags (VOBJBUF) tag VT_PROCFS, type 5, pid 36, mode 0x180, flags 0 : 0xc6168380: type VREG, usecount 3, writecount 0, refcount 0, flags (VOBJBUF) tag VT_PROCFS, type 5, pid 36, mode 0x180, flags 0 vnode_pager_getpages: I/O read error vm_fault: pager read error, pid 39554 (sylk1) pid 39554 (sylk1), uid 0: exited on signal 11 (core dumped) Looking at my core dump, it seems that sylk1 had mmap()d /proc/36/regs (FWIW, this is mfs) and caused the problem by trying to de-reference the first word of the mmap'd region. Accessing the data using a read(2) works correctly. I believe this behaviour is undesirable. If I can successfully mmap() a region (the actual parameters were PROT_READ, MAP_SHARED), I should be able to read the mmap'd region. Unfortunately, I don't understand the filesystem/vnode layering well enough to be able to easily identify what's missing. Peter To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message