Bug 880705 (CVE-2012-5373)

Summary:	CVE-2012-5373 java: Murmur hash function collisions (oCERT-2012-001)
Product:	[Other] Security Response	Reporter:	Jan Lieskovsky <jlieskov>
Component:	vulnerability	Assignee:	Red Hat Product Security <security-response-team>
Status:	CLOSED ERRATA	QA Contact:
Severity:	medium	Docs Contact:
Priority:	medium
Version:	unspecified	CC:	ahughes, dbhole, dpal, jerboaa, jvanek, lkundrak, mjw, mmatejov, omajid
Target Milestone:	---	Keywords:	Security
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2021-10-19 21:57:49 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:
Bug Depends On:	880713, 880714
Bug Blocks:	880720

Description Jan Lieskovsky 2012-11-27 16:08:09 UTC

A denial of service flaw was found in the Murmur hash function implementation, as being used by various Java implementations. A specially-crafted set of keys could trigger Murmur hash function collisions, which degrade hash table items insert performance by changing hash table operations complexity from an expected/average O(n) to the worst case O(n^2). Reporters were able to find colliding strings efficiently using equivalent substrings.

As various web application frameworks for Java automatically pre-fill certain arrays with data from the HTTP request (such as GET or POST parameters) for Java web applications, a remote attacker could use this flaw to make the Java virtual machine to use an excessive amount of CPU time by sending a POST request with a large number parameters which hash to the same value.

A different vulnerability than CVE-2012-2739.

References:
[1] http://www.openwall.com/lists/oss-security/2012/11/23/4
[2] http://www.ocert.org/advisories/ocert-2012-001.html
[3] http://2012.appsec-forum.ch/conferences/#c17
[4] https://www.131002.net/data/talks/appsec12_slides.pdf
[5] http://asfws12.files.wordpress.com/2012/11/asfws2012-jean_philippe_aumasson-martin_bosslet-hash_flooding_dos_reloaded.pdf

Comment 1 Jan Lieskovsky 2012-11-27 16:15:43 UTC

Ruby language upstream (which was also vulnerable to similar issue) in version ruby-1.9.3 patchlevel 327 has replaced the Murmur hash implementation with the SipHash-2-4 one (which is not vulnerable to this problem):
  http://www.ruby-lang.org/en/news/2012/11/09/ruby19-hashdos-cve-2012-5371/
  https://www.131002.net/siphash/

Comment 2 Jan Lieskovsky 2012-11-27 16:17:57 UTC

This issue affects the version of the java-1.6.0-openjdk package, as shipped with Fedora release of 16. Please schedule an update (once there is final upstream patch available).

--

This issue affects the versions of the java-1.7.0-openjdk packages, as shipped with Fedora release of 16 and 17. Please schedule an update (once there is final upstream patch available).

Comment 3 Jan Lieskovsky 2012-11-27 16:19:12 UTC

Created java-1.6.0-openjdk tracking bugs for this issue

Affects: fedora-16 [bug 880713]

Comment 4 Jan Lieskovsky 2012-11-27 16:21:09 UTC

Created java-1.7.0-openjdk tracking bugs for this issue

Affects: fedora-all [bug 880714]

Comment 5 Tomas Hoger 2012-11-27 16:49:40 UTC

Bug 750533 tracks the original HashDoS attack variant for Java.  Bug 750533, comment 15 points to a discussion of the change that introduced Murmur hash use to mitigate the original hash collisions problem.