Bug 20167

Summary: Problem with RPC on jail(8) in 4.0-RELEASE and 4.0-STABLE
Product: Base System Reporter: zensyo <zensyo>
Component: kernAssignee: Poul-Henning Kamp <phk>
Status: Closed FIXED    
Severity: Affects Only Me    
Priority: Normal    
Version: 4.0-STABLE   
Hardware: Any   
OS: Any   

Description zensyo 2000-07-25 06:10:00 UTC
Some command that imprimented by RPC takes kernel panic. For example, When run following command on jail enviroment,
your kernel will panic.

rusers -l localhost


Fatal trap 12: page fault while in kernel mode
fault virtual address   = 0x0
fault code              = supervisor write, page not present
instruction pointer     = 0x8:0xc017827d
stack pointer           = 0x10:0xcfededf8
frame pointer           = 0x10:0xcfededfc
code segment            = base rx0, limit 0xfffff, type 0x1b
                        = DPL 0, pres 1, def32 1, gran 1
processor eflags        = interrupt enabled, resume, IOPL = 0
current process         = 2054 (rusers)
interrupt mask          =
trap number             = 12
panic: page fault

In another case, popper which use DRAC(for POP before SMTP)  takes same problem. DRAC patch for popper using RPC.
And, I found similer problem on GNATS database at misc/19839.

How-To-Repeat: Any time, please run command rusers -l localhost on jail enviroment.
May be any program that using RPC takes same probrem.
Comment 1 Sheldon Hearn freebsd_committer freebsd_triage 2000-07-25 09:52:14 UTC
Responsible Changed
From-To: freebsd-bugs->phk

pjk is the jail(8) warden, although he might defer this to 
one of the networking chaps.
Comment 2 Robert Watson freebsd_committer freebsd_triage 2000-07-26 21:59:53 UTC
I was unable to reproduce this bug on 5.0-CURRENT, but I'm wondering if it
doesn't have to do with IPv6.  ``localhost'' resolves to "::1" under
IPv6-enabled FreeBSD boxes, which is unfortunate, as the jail code in
4.0 doesn't really know what to do about that.  For that matter, the jail
code in 5.0-CURRENT doesn't really either.

I.e., since the introduction of IPv6, ``telnet localhost'' has telneted to
::1 instead of 127.0.0.1, causing operations on localhost that are
v6-enabled to go to the wrong host.  To solve this, we may have to make
jail more multi-protocol aware.

I'll have access to a 4.x box this evening and will see if I can reproduce
it there, also.

  Robert N M Watson 

robert@fledge.watson.org              http://www.watson.org/~robert/
PGP key fingerprint: AF B5 5F FF A6 4A 79 37  ED 5F 55 E9 58 04 6A B1
TIS Labs at Network Associates, Safeport Network Services
Comment 3 suzuki 2000-07-27 04:16:39 UTC
I tried it on only IPv4. And, the problem was reproduced using any IPv4
addresses.
Comment 4 Poul-Henning Kamp freebsd_committer freebsd_triage 2000-09-17 14:36:00 UTC
State Changed
From-To: open->closed

Thanks for the report, I found the problem and I belive it is fixed 
in current now.
Comment 5 Niels Bakker 2001-02-27 01:32:28 UTC
Has this bugfix been MFC'd?  I could still observe the behaviour with
4.2-BETA.


	-- Niels.