Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 30 Sep 1997 22:48:27 +0300 (EEST)
From:      Heikki Suonsivu <hsu@mail.clinet.fi>
To:        FreeBSD-gnats-submit@FreeBSD.ORG
Subject:   kern/4663: checkalias panic
Message-ID:  <199709301948.WAA24861@katiska.clinet.fi>
Resent-Message-ID: <199709302000.NAA18000@hub.freebsd.org>

next in thread | raw e-mail | index | archive | help

>Number:         4663
>Category:       kern
>Synopsis:       checkalias panic
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    freebsd-bugs
>State:          open
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Tue Sep 30 13:00:01 PDT 1997
>Last-Modified:
>Originator:     Heikki Suonsivu
>Organization:
Clinet, Espoo, Finland
>Release:        FreeBSD 2.2-STABLE i386
>Environment:

2.2-STABLE

>Description:

	We are getting very frequent, usually nfs-related panics ffrom
checkalias in vfs_subr.  Often these repeat and several machines go in a
row, which look a bit like it could be initiated from user level (>3000
users can access the machines, so it might be that someone found a
kill-freebsd-o-matic).  I think this happens much less frequently in closed
servers.

checkalias seem to refer lots of things without checking them for being
NULL.

We are also seeing nfs directories locking up and various instability with
nfs, which I pr'd some time ago.

fault virtual addr: 0x87654371
fault code: supervisor read, page not present
instruction pointer: 0x8:0xf0131dff
stack pointer: 0x10:0xcfbffd7c
frrame pointer: 0x10:0xcfbffd8c
code segment base 0x0 limit 0xfffff type 0x1b
	DPL 0 press 1 def 32 gran 1
cpu eflags interrupt enabled, resume IOPL = 0
current prcess 112 (nfsiod)
interrupt mask bio
panic page fault

(copied from handwriting, there may be some bit errors)

f01310c0 F vfs_subr.o
f01310c0 t ___gnu_compiled_c
f01310c0 t _sysctl___kern_maxvnodes
f01310c0 t gcc2_compiled.
f01310e8 t ___set_sysctl__kern_sym_sysctl___kern_maxvnodes
f01310ec T _vntblinit
f0131130 T _vfs_lock
f0131184 T _vfs_unlock
f01311c0 T _vfs_busy
f0131228 T _vfs_unbusy
f01312f4 T _vfs_unmountroot
f0131408 T _vfs_unmountall
f01314b4 T _getvfs
f01314ec T _getnewfsid
f0131574 T _vattr_null
f0131624 T _getnewvnode
f013178c T _insmntque
f01317f8 T _vwakeup
f0131878 T _vinvalbuf
f0131ab0 T _bgetvp
f0131b4c T _brelvp
f0131bdc T _pbgetvp
f0131c20 T _pbrelvp
f0131c5c T _reassignbuf
f0131d60 T _bdevvp
f0131dcc T _checkalias
f0131ef8 T _vget
f0131fdc T _vref
f0132024 T _vput
f0132078 T _vrele
f0132178 T _vflush
f013229c T _vclean
f01323fc T _vgoneall
f01324a8 T _vgone
f013265c T _vfinddev
f01326ac T _vcount
f01327c0 T _vprint

>How-To-Repeat:

Lots of users and nfs seem to be a good one.

>Fix:
	

>Audit-Trail:
>Unformatted:



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199709301948.WAA24861>