SCIENTIFIC-LINUX-USERS Archives

February 2011

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Don Krause <[log in to unmask]>
Reply To:
Don Krause <[log in to unmask]>
Date:
Thu, 3 Feb 2011 15:52:06 -0800
Content-Type:
multipart/signed
Parts/Attachments:
text/plain (3665 bytes) , smime.p7s (4 kB)

On Feb 3, 2011, at 2:20 PM, Don Krause wrote:

> 
> On Feb 3, 2011, at 1:39 PM, Troy Dawson wrote:
> 
>> Don Krause wrote:
>>> Ok, this is definitely a bug. Well 2 actually.
>>> But It doesn't appear in the publicly accessible bugzilla at RH.
>>> Package ypbind actually depends on policycoreutils-python.
>> 
>> This is a problem for both SL and RedHat.  I just checked the dependencies.
>> 
>>> A fresh install of SL6 Beta1, using "Software Development Workstation", and selecting NIS under "Use Network Login", fails
>>> to install policycoreutils-python, which contains "sesetbool". "sesetbool" is called by /etc/init.d/ypbind to allow ypbind access.
>>> Installing as "Basic Server" at least includes policycoreutils-python.
>>> Unfortunately, bug number 2, is that "sesetbool allow_ypbind=1" doesn't work, since the default selinux policy doesn't have
>>> "allow_ypbind"
>> 
>> I'm trying to test this on a real RHEL6 system.
>> Aside from ypbind still not working, how can we tell "sesetbool allow_ypbind=1" doesn't work?
>> 
>> Troy
>> 
> 
> On both test boxes I tried, one installed as "Software Development" and one installed as "Basic Server",
> I just tried to start ypbind via "service ypbind start". It would actually start ypbind, but it wouldn't connect to
> the ypmaster.
> 
> In /var/log/audit/audit.log I'd get:
> 
> type=USER_AVC msg=audit(1296764305.009:32965): user pid=2503 uid=81 auid=4294967295 ses=4294967295 subj=system_u:system_r:system_dbusd_t:s0-s0:c0.c1023 msg='avc:  denied  { send_msg } for msgtype=method_call interface
> =org.freedesktop.NetworkManager member=state dest=org.freedesktop.NetworkManager spid=3718 tpid=3449 scontext=unconfined_u:system_r:ypbind_t:s0 tcontext=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 tclass=dbu
> s : exe="/bin/dbus-daemon" sauid=81 hostname=? addr=? terminal=?'
> 
> Then I'd run "sesetbool allow_ypbind=1" as it's found in /etc/init.d/ypbind, and attempt to restart ypbind. I got the same denial in audit.log
> 
> I ran the denial through audit2allow, which gave me:
> 
> module ypbind 1.0;
> 
> require {
> 	type unconfined_t;
> 	type ypbind_t;
> 	class dbus send_msg;
> }
> 
> #============= ypbind_t ==============
> allow ypbind_t unconfined_t:dbus send_msg;
> 
> Compile that to a module and install, then ypbind runs and connects as expected.
> 
> As Stephan recommend, I did a "getsebool -a | grep yp" which returned "allow_ypbind --> on", (this is AFTER I did "sesetbool allow_ypbind=1") but ypbind still wouldn't bind to the master.
> 
> On a completely fresh installation, "getsebool -a | grep yp" returns "allow_ypbind --> off".
> 
> Thanks for looking!


Now, because things aren't weird enough.. I've setup the PXE boot environment, copied the anaconda-ks from
the test VM that was installed as "Software Development Workstation".  You know, the same one where ypbind
refused to work until I compiled a pp file and installed it? Except, now ypbind binds as expected.

Even stranger, policycoreutils-python is NOT installed, so sesetbool as expected by /etc/init.d/ypbind doesn't exist, yet
there's no selinux denial when installed this way.

Now I'm just confused....

Is it possible that this is d-bus issue more than an selinux issue?

--
Don Krause                                                                   
Head Systems Geek, 
Waver of Deceased Chickens.
Optivus Proton Therapy, Inc.
P.O. Box 608
Loma Linda, California 92354
909.799.8327 Tel
909.799.8366 Fax
[log in to unmask]
www.optivus.com
"This message represents the official view of the voices in my head."








ATOM RSS1 RSS2