Bug 861032 - Segfault on startup if corosync.conf is wrong
Summary: Segfault on startup if corosync.conf is wrong
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: corosync
Version: 6.3
Hardware: x86_64
OS: Linux
Target Milestone: rc
: ---
Assignee: Jan Friesse
QA Contact: Cluster QE
Depends On: 860711
TreeView+ depends on / blocked
Reported: 2012-09-27 11:03 UTC by Jan Friesse
Modified: 2017-02-06 15:04 UTC (History)
5 users (show)

Fixed In Version: corosync-1.4.1-11.el6
Doc Type: Bug Fix
Doc Text:
Cause: corosync.conf contains ringnumber > 1 Consequence: Segfault Fix: Added check in corosync.conf parser. If ringnumber is > 1, correct error message is printed (interface ring number X is bigger then allowed maximum 1). Result: Corosync doesn't segfault if ringnumber in config is > 1 and proper error message is printed.
Clone Of: 860711
Last Closed: 2013-02-21 07:50:52 UTC
Target Upstream Version:

Attachments (Terms of Use)
Proposed patch (1.28 KB, patch)
2012-09-27 11:03 UTC, Jan Friesse
no flags Details | Diff

System ID Private Priority Status Summary Last Updated
Red Hat Product Errata RHBA-2013:0497 0 normal SHIPPED_LIVE corosync bug fix and enhancement update 2013-02-20 21:18:24 UTC

Description Jan Friesse 2012-09-27 11:03:18 UTC
Created attachment 618000 [details]
Proposed patch

+++ This bug was initially created as a clone of Bug #860711 +++

Created attachment 617571 [details]

Description of problem:

Moved config files forward from a 6.2_el6 version of Corosync.  Made minor change to config to use different multicast address than an existing corosync on same network.

Version-Release number of selected component (if applicable):

How reproducible:
Have an invalid config file, start corosync.

Steps to Reproduce:
1. use the config file attached
2. start corosync
Actual results:
*** glibc detected *** corosync: corrupted double-linked list: 0x0000000001cca190 ***
Under gdb:
Program received signal SIGSEGV, Segmentation fault.
0x00007ffff6e581e3 in _int_malloc () from /lib64/libc.so.6

Expected results:
Corosync to tell me it has an invalid config and exit

Additional info:

--- Additional comment from jfriesse@redhat.com on 2012-09-27 06:32:03 EDT ---

Pretty nasty bug. Main problem is hidden in "ringnumber: 2" line. We have dynamically allocated array with size of 2 and we are using direct indexing of ringnumber for that array (so we are using item 3).

This bug affects ALL releases RHEL 6 and also upstream master!

Many thanks for pointing this bug out.

--- Additional comment from jfriesse@redhat.com on 2012-09-27 07:01:14 EDT ---

Created attachment 617999 [details]
Proposed patch

Don't access invalid mem in totemconfig interfaces

When ringnumber in config file was set to value bigger or equal to
INTERFACE_MAX, we are using this big value as index to totemconfig
interfaces array, resulting to access to invalid memory and segfault.

Instead of that, ringnumber is now checked and proper error message is
printed if value is too big.

Comment 5 Jaroslav Kortus 2012-11-16 14:22:31 UTC
Verified with corosync-1.4.1-11.el6.x86_64:

corosync[32508]:   [MAIN  ] parse error in config: interface ring number 10 is bigger then allowed maximum 1

config used:
totem {
	version: 2
	secauth: off
	threads: 0
	rrp_mode: passive
	interface {
		ringnumber: 0
		mcastport: 5405
		ttl: 1

	interface {
		ringnumber: 10
		mcastport: 5405
		ttl: 1


logging {
	fileline: off
	to_stderr: no
	to_logfile: yes
	to_syslog: yes
	logfile: /var/log/cluster/corosync.log
	debug: off
	timestamp: on
	logger_subsys {
		subsys: AMF
		debug: off

amf {
	mode: disabled

It starts properly with valid ring ids.

Comment 7 errata-xmlrpc 2013-02-21 07:50:52 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.


Note You need to log in before you can comment on or make changes to this bug.