NetBSD Problem Report #43290
From bch@methodlogic.net Tue May 11 19:51:06 2010
Return-Path: <bch@methodlogic.net>
Received: from mail.netbsd.org (mail.netbsd.org [204.152.190.11])
by www.NetBSD.org (Postfix) with ESMTP id D333863BA59
for <gnats-bugs@gnats.NetBSD.org>; Tue, 11 May 2010 19:51:06 +0000 (UTC)
Message-Id: <20100511203144.GC3395@methodlogic.net>
Date: Tue, 11 May 2010 20:31:44 +0000
From: bch@methodlogic.net
Reply-To:
To: gnats-bugs@gnats.NetBSD.org
Subject: kern crash on socket read/write amd64 -current
X-Send-Pr-Version: 3.95
>Number: 43290
>Category: kern
>Synopsis: crash system networking on lo0
>Confidential: no
>Severity: serious
>Priority: high
>Responsible: kern-bug-people
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Tue May 11 19:55:00 +0000 2010
>Last-Modified: Sun Jun 12 23:20:02 +0000 2011
>Originator: bch@methodlogic.net
>Release: NetBSD 5.99.29
>Organization:
Method Logic Digital
http://methodlogic.net
>Environment:
System: NetBSD kamloops 5.99.29 NetBSD 5.99.29 (kamloops) #30: Mon May 10 22:55:08 PDT 2010 root@kamloops:/usr/obj/sys/arch/amd64/compile/kamloops amd64
Architecture: x86_64
Machine: amd64
>Description:
Transcribed by hand from ddb output:
uvm_fault(0xffff80004e7e8a30, 0x0, 1) -> e
fatal page fault in supervisor mode
trap type 6 code 0 rip ffffffff803ded94 cs 8 rflags 10286 cr2 40 cpl 0 rsp ffff8000437cd6f0
kernel: page fault trap, code 0
Stopped in pid 479.1 (netstat) at netbsd:kath_cred_getuid + 0x4: movl 0 x40 (%rdi), %eax
>How-To-Repeat:
Was observing no reponse on network socket (which should
have held data), then ran netstat. Seems to be a specific
test case which generates fault, which I can elaborate on
offline if required.
On the same host, I was listening on a socket, writing to it,
observing no response, then ran netstat and crashed the system.
>Fix:
>Audit-Trail:
From: David Holland <dholland-bugs@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc:
Subject: Re: kern/43290: kern crash on socket read/write amd64 -current
Date: Sun, 12 Jun 2011 23:19:06 +0000
On Tue, May 11, 2010 at 07:55:00PM +0000, bch@methodlogic.net wrote:
> kernel: page fault trap, code 0
> Stopped in pid 479.1 (netstat) at netbsd:kath_cred_getuid + 0x4:
> movl 0x40 (%rdi), %eax
>
>
> >How-To-Repeat:
> Was observing no reponse on network socket (which should
> have held data), then ran netstat. Seems to be a specific
> test case which generates fault, which I can elaborate on
> offline if required.
>
> On the same host, I was listening on a socket, writing to it,
> observing no response, then ran netstat and crashed the system.
Did anyone ever look into this? If you see it again, try to get a
backtrace; it will be a lot easier to figure if we know what's calling
kauth_cred_getuid.
--
David A. Holland
dholland@netbsd.org
(Contact us)
$NetBSD: query-full-pr,v 1.39 2013/11/01 18:47:49 spz Exp $
$NetBSD: gnats_config.sh,v 1.8 2006/05/07 09:23:38 tsutsui Exp $
Copyright © 1994-2007
The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.