Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 25 Dec 2007 14:30:05 GMT
From:      Robert Watson <rwatson@FreeBSD.org>
To:        freebsd-bugs@FreeBSD.org
Subject:   Re: kern/103432: panic: nfssvc_nfsd(): debug.mpsafenet=1 && Giant
Message-ID:  <200712251430.lBPEU5Gv013198@freefall.freebsd.org>

next in thread | raw e-mail | index | archive | help
The following reply was made to PR kern/103432; it has been noted by GNATS.

From: Robert Watson <rwatson@FreeBSD.org>
To: Daniel Rudy <dcrudy@pacbell.net>
Cc: freebsd-gnats-submit@FreeBSD.org, freebsd-bugs@FreeBSD.org
Subject: Re: kern/103432: panic: nfssvc_nfsd(): debug.mpsafenet=1 && Giant
Date: Tue, 25 Dec 2007 14:23:11 +0000 (GMT)

 On Wed, 20 Sep 2006, Daniel Rudy wrote:
 
 >> Number:         103432
 >> Category:       kern
 >> Synopsis:       panic: nfssvc_nfsd(): debug.mpsafenet=1 && Giant
 >> Confidential:   no
 >> Severity:       critical
 >> Priority:       high
 >> Responsible:    freebsd-bugs
 >> State:          open
 >> Quarter:
 >> Keywords:
 >> Date-Required:
 >> Class:          sw-bug
 >> Submitter-Id:   current-users
 >> Arrival-Date:   Wed Sep 20 15:40:17 GMT 2006
 >> Closed-Date:
 >> Last-Modified:
 >> Originator:     Daniel Rudy
 >> Release:        6.1-RELEASE-p7
 >> Organization:
 >> Environment:
 >> Description:
 > This computer is configured as a NFS server and it also has IPSEC enabled.  When a NFS client accesses this server, the server imediately panics with the following error "panic: nfssvc_nfsd(): debug.mpsafenet=1 && Giant" and reboots hanging the client.  On a release 6.0 machine, debug.mpsafe is forced to 0 because IPSEC requires Giant.
 >> How-To-Repeat:
 > On a 6.1 system, enable IPSEC and NFS then try to access the server remotely.
 >> Fix:
 > Can't.  The OID for debug.mpsafenet is readonly.
 
 It would be very helpful to have a stack trace on this one.
 
 There's a reasonable chance that this bug is gone in 7.x as we now have an 
 entirely MPSAFE IPSEC implementation, and the debug.mpssafenet compat code has 
 been entirely removed.  However, it would be very helpful to know.  You should 
 be able to compile and use a 7.x kernel with a 6.x userspace, which would give 
 you a lower-risk way to determine this.  Make sure to do a 7.x buildworld 
 before 7.x buildkernel, but only to a 7.x installkernel, not 7.x 
 installworld/mergemaster.
 
 Robert N M Watson
 Computer Laboratory
 University of Cambridge



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200712251430.lBPEU5Gv013198>