NetBSD Problem Report #45355

From apb@cequrux.com  Sat Sep 10 16:26:19 2011
Return-Path: <apb@cequrux.com>
Received: from mail.netbsd.org (mail.netbsd.org [204.152.190.11])
	by www.NetBSD.org (Postfix) with ESMTP id EF94363BB48
	for <gnats-bugs@gnats.NetBSD.org>; Sat, 10 Sep 2011 16:26:18 +0000 (UTC)
Message-Id: <20110910162150.8F40C2C5E1D3@apb-laptoy.apb.alt.za>
Date: Sat, 10 Sep 2011 16:21:50 +0000 (UTC)
From: apb@cequrux.com
To: gnats-bugs@gnats.NetBSD.org
Subject: Reader/writer lock error:  rw_vector_enter: locking against myself
X-Send-Pr-Version: 3.95

>Number:         45355
>Category:       kern
>Synopsis:       Reader/writer lock error:  rw_vector_enter: locking against myself
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sat Sep 10 16:30:00 +0000 2011
>Closed-Date:    Sun Nov 13 18:05:13 +0000 2011
>Last-Modified:  Sun Nov 13 18:05:13 +0000 2011
>Originator:     Alan Barrett
>Release:        NetBSD 5.99.55
>Organization:
Not much
>Environment:
System: NetBSD i386 5.99.55 (sources from 2011-09-02 14:00 UTC)
Architecture: i386
Machine: i386
>Description:
The system crashes frequently (more than once per day).  I am usually
not able to get any debug information, because after a crash pressing
alt-control-F1 does not switch the display from graphics to text mode.
The most recent crash was while I was using a text console, so I was
able to get the following information.  (This is transcribed from blurry
photos, and it's likely that there's confusion between the digits
0 and 8.)

Reader / writer lock error: rw_vector_enter: locking against myself

lock address : 0x00000000d483e0b0
current cpu  :                  0
current lwp  : 0x00000000d8eac2a0
owner/count  : 0x00000000d8eac2a0 flags    = 0x0000000000000004

panic: lock error
fatal breakpoint trap in supervisor mode
trap type 1 code 0 eip c0251254 cs 8 eflags 206 cr2 ce881000 ilevel 0
stopped in pid 15882.1 (vim) at netbsd:breakpoint+0x4: popl %ebp

db{0}> bt
breakpoint ...
panic ...
lockdebug_abort ...
rw_abort ...
rw_vector_enter ...
genfs_lock ...
layer_bypass ...
VOP_LOCK ...
vclean ...
getcleanvnode ...
getnewvnode ...
ffs_vget ...
ffs_valloc ...
ufs_makeinode ...
ufs_create ...
VOP_CREATE ...
vn_open ...
sys_open ...
syscall ...

>How-To-Repeat:
The crash happened as I was attempting to save a file from the
"vim" editor (installed from pkgsrc).  The file system is
ffs+wapbl on cgd.  The file system also happens to be the
backing layer for a read-only nullfs mount, and it's possible that
a process was attempting to access the same file through the
read-only nullfs mount at the same time as the editor was
attempting to save the file to the writabe backing layer.

>Fix:

>Release-Note:

>Audit-Trail:
From: "Juergen Hannken-Illjes" <hannken@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/45355 CVS commit: src/sys/kern
Date: Sun, 2 Oct 2011 13:00:08 +0000

 Module Name:	src
 Committed By:	hannken
 Date:		Sun Oct  2 13:00:07 UTC 2011

 Modified Files:
 	src/sys/kern: vfs_vnode.c

 Log Message:
 The path getnewvnode()->getcleanvnode()->vclean()->VOP_LOCK() will panic
 if the vnode we want to clean is a layered vnode and the caller already
 locked its lower vnode.

 Change getnewvnode() to always allocate a fresh vnode and add a helper
 thread (vdrain) to keep the number of allocated vnodes within desiredvnodes.

 Rename getcleanvnode() to cleanvnode() and let it take a vnode from the
 lists, clean and free it.

 Reviewed by: David Holland <dholland@netbsd.org>

 Should fix:
 PR #19110 (nullfs mounts over NFS cause lock manager problems)
 PR #34102 (ffs panic in NetBSD 3.0_STABLE)
 PR #45115 (lock error panic when build.sh*3 and daily script is running)
 PR #45355 (Reader/writer lock error:  rw_vector_enter: locking against myself)


 To generate a diff of this commit:
 cvs rdiff -u -r1.11 -r1.12 src/sys/kern/vfs_vnode.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

State-Changed-From-To: open->feedback
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Sat, 05 Nov 2011 16:38:30 +0000
State-Changed-Why:
is it working better now?


From: Alan Barrett <apb@cequrux.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: kern/45355 (Reader/writer lock error:  rw_vector_enter: locking
 against myself)
Date: Sun, 6 Nov 2011 20:23:57 +0200

 On Sat, 05 Nov 2011, dholland@NetBSD.org wrote:
 >is it working better now?

 I think this is fixed.  I am unable to trigger the panic using
 a kernel from 2011-10-16 (with src/sys/kern/kern_vnode.c revision 1.14).

 --apb (Alan Barrett)

State-Changed-From-To: feedback->closed
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Sun, 13 Nov 2011 18:05:13 +0000
State-Changed-Why:
Fixed, thanks.


>Unformatted:

Home	PR Database Search