Subject: kern/17107: repeatable kernel panic in nfs_reclaim
To: None <gnats-bugs@gnats.netbsd.org>
From: None <art@riverstonenet.com>
List: netbsd-bugs
Date: 05/30/2002 10:31:44
>Number:         17107
>Category:       kern
>Synopsis:       repeatable kernel panic in nfs_reclaim
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu May 30 10:32:00 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     Artem Belevich
>Release:        1.6_BETA1
>Organization:
>Environment:
NetBSD tyr 1.6_BETA1 NetBSD 1.6_BETA1 (GENERIC_DIAGNOSTIC) #0: Sat May 25 16:12:18 UTC 2002     autobuild@tgm.daemon.org:/autobuild/i386/OBJ/autobuild/Src-build-real/sys/arch/i386/compile/GENERIC_DIAGNOSTIC i386
>Description:
System uses amd to automount filesystems from NetAPP fileserver and from various SUN workstations. When daily cron job is run system panics in nfs_reclaim. The crash is repeatable. It also happens with the -current snapshot taken in mid-April. I have core dump for the DIAGNOSTICS kernel from the 5/25 daily build.

uvm_fault(0xe5bbfeb0, 0x0, 0, 1) -> e
kernel: page fault trap, code=0
Stopped in pid 782 (find) at    nfs_reclaim+0x54:       testb   $0x1,       0x1(%esi)
db> t
nfs_reclaim(e6e4abe4,8,0,c01c42eb,caa431d0) at nfs_reclaim+0x54
VOP_RECLAIM(e6b2014c,e5e2b570,200000,c01c3ae8) at VOP_RECLAIM+0x2e
vclean(e6b2014c,8,e5e2b570,c01c7321) at vclean+0x107
vgonel(e6b2014c,e5e2b570,0,c0189cff) at vgonel+0x46
getnewvnode(1,c15cd600,c13e7100,e6e4acdc,e08f10ec) at getnewvnode+0x210
ffs_vget(c15cd600,2796ff,e6e4ad68,e5e2b570,e65f15b8) at ffs_vget+0x4f
ufs_lookup(e6e4ada0,30002,e6e4adb0,c01ceb41,e6e4ae98) at ufs_lookup+0x9bd
VOP_LOOKUP(e65f15b8,e6e4aea8,e6e4aebc,e6e4adfc,e65f15b8) at VOP_LOOKUP+0x35
lookup(e6e4ae98,e5bcc400,400,e6e4aeb0,e6e4af80) at lookup+0x259
namei(e6e4ae98,805884c,60,1,e6e4af80) at namei+0x2f1
sys___lstat13(e5e2b570,e6e4af80,e6e4af78,c1579820) at sys___lstat13+0x4f
syscall_plain(1f,1f,1f,bfbf001f,8058900) at syscall_plain+0xa7

>How-To-Repeat:
It seems that the crash usually happens when /etc/daily runs "find \*.core" on nfs-mounted filesystems.
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: