Bug 470316

Summary:	semodule segfaults when loading a base policy with fewer categories than the currently loaded policy
Product:	Red Hat Enterprise Linux 5	Reporter:	Stuart Sears <ssears>
Component:	libsepol	Assignee:	Daniel Walsh <dwalsh>
Status:	CLOSED NEXTRELEASE	QA Contact:	BaseOS QE Security Team <qe-baseos-security>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	5.2	CC:	dwalsh, mmalik
Target Milestone:	rc
Target Release:	---
Hardware:	i386
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2009-12-21 14:36:41 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Stuart Sears 2008-11-06 17:12:00 UTC

Description of problem:
When building a new base policy module from the serefpolicy sources (in our source RPM), loading the new base policy segfaults if you have fewer than 1024 categories defined.

Version-Release number of selected component (if applicable):
policycoreutils-1.33.12-12.el5
(1.33.12-14 has the same issue)

How reproducible:
Every time

Steps to Reproduce:
1. unpack the source RPM (selinux-policy-2.4.6-106.el5_1.3)
rpmbuild -bp <specfile>
then in the serefpolicy directory...

2. edit build.conf:
DISTRO=redhat
TYPE=targeted-mcs
NAME=targeted
POLY=y
MONOLITHIC=n
QUIET=n
DIRECT_INITRC=y

3. make bare
4. make conf

5. cp %_topdir/SOURCES/modules,booleans .conf into the policy dir.

6. build the base policy
make base.pp

then try and load your new base policy module with 

7. semodule -b base.pp
  
Actual results:

Segmentation fault with no additional error messaged


Expected results:
new base policy loads (or fails with a comprehensible error message)

Additional info:
This appears to be related to the number of categories in the new base (256), when existing/loaded policy modules expect more (1024).
A more meaningful error message would be nice, rather than just a segfault :)

Comment 3 Daniel Walsh 2009-12-21 14:36:41 UTC

This has been fixed in upstream, I will work in RHEL6.  But I think we just need to close next release for now.  Not a problem many customers will face.