Subject: kern/9285: Kernel hang in lockmgr()
To: None <gnats-bugs@gnats.netbsd.org>
From: None <Michael.Eriksson@era-t.ericsson.se>
List: netbsd-bugs
Date: 01/24/2000 00:18:38
>Number:         9285
>Category:       kern
>Synopsis:       Kernel hang in lockmgr()
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people (Kernel Bug People)
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Mon Jan 24 00:18:00 2000
>Last-Modified:
>Originator:     Michael Eriksson
>Organization:
Ericsson Radio Systems AB
>Release:        -current as of about January 17, 2000
>Environment:
System: NetBSD kafka 1.4P NetBSD 1.4P (KAFKA) #0: Fri Jan 21 16:11:07 CET 2000 eramore@kafka:/usr/src/sys/arch/i386/compile/KAFKA i386

>Description:

The entire machine will hang, not even answering pings. This has
happened a number of times on my new 700 MHz PIII Dell, I think mostly
(only?) during NFS use (client only).

This is a two-finger copy of ddb trace:

lockmgr(f02fa2c0, 400001,0) at lockmgr+0xc3
proclist_lock_read(f093e8c0,f093e864,fca2dd30,f0162566,22bc) at proclist_lock_read+0x28
pfind(22bc,f093e8c0,fca2dd44,f016e76c,f093e8c0) at pfind+0xc
selwakeup()
sowakeup()
udp4_sendup()
udp4_realinput()
udp_input()
ip_input()
ipintr()

This is the only trace I have, so I don't know if it is typical for
all hangings. I don't have any crash dumps, but I'll try to generate
one if/when this happens again.

Additional notes: When running make -j 4 or so on a NFS-mounted source
directory, I sometimes get some strange "file missing" errors.
Restarting the make always succeeds.

>How-To-Repeat:

Running make -j 8 on a NFS-mounted source directory will make the
machine hang after a while.

>Fix:

Not known :-(.
>Audit-Trail:
>Unformatted: