Security
Holes in the Linux random number generator?
Eye catching headlines are seen every day on the web, but one needs to be careful not to distort the contents of the article. A recent SecuriTeam article is headlined "Holes in the Linux Random Number Generator" but that title overstates the actual contents of the paper (PDF) it is announcing.
The three authors of the paper provide a nice detailed description of the Linux random number generator (RNG) and the algorithms that it uses, while also reporting a very theoretical attack. The basic attack is against the "forward security" of the RNG via a single compromise of the contents of the entropy pool. This value can be used to run the RNG algorithm in reverse and recover previous states of the entropy pool. Doing this enough times can recover keys that have been previously generated.
There are a number of reasons why this attack is considered to have little impact on real world systems. The most obvious is that if an attacker can access the state of the entropy pool, they have already broken the security of the system and can, as root, do any number of different things to the system. If recovering previously generated keys is the object of the attack, the paper outlines ways to do that, but the processing requirements are enormous as Ted Ts'o points out:
The paper also describes a well known feature of the Linux RNG implementation as if it were a newly discovered denial of service issue. The /dev/random device was specifically designed to block when the entropy pool had insufficient entropy to satisfy the request. The /dev/urandom device is provided as an alternative that generates very good random numbers and does not block (and is therefore not vulnerable to a denial of service). For any but the most sensitive applications (key generation being an obvious choice), /dev/urandom is the recommended source for random numbers. Alan Cox sums up the situation nicely:
The paper has sparked an interesting discussion on the linux kernel mailing list and has lead to some concrete suggestions for improving the algorithm, but it would be an exaggeration to conclude that the paper describes real world Linux security concerns. An administrator or security professional reading the SecuriTeam headline might easily be led astray.
New vulnerabilities
awstats: missing input sanitizing
| Package(s): | awstats | CVE #(s): | CVE-2006-2237 | ||||||||||||||||||||||||
| Created: | May 19, 2006 | Updated: | June 20, 2006 | ||||||||||||||||||||||||
| Description: | Hendrik Weimer discovered that specially crafted web requests can cause awstats, a powerful and featureful web server log analyzer, to execute arbitrary commands. | ||||||||||||||||||||||||||
| Alerts: |
| ||||||||||||||||||||||||||
cscope: buffer overflows
| Package(s): | cscope | CVE #(s): | CVE-2004-2541 | ||||||||||||||||||||||||
| Created: | May 22, 2006 | Updated: | June 19, 2009 | ||||||||||||||||||||||||
| Description: | A buffer overflow in Cscope 15.5, and possibly multiple overflows, allows remote attackers to execute arbitrary code via a C file with a long #include line that is later browsed by the target. | ||||||||||||||||||||||||||
| Alerts: |
| ||||||||||||||||||||||||||
dia: format string vulnerabilities
| Package(s): | dia | CVE #(s): | CVE-2006-2453 CVE-2006-2480 | ||||||||||||||||||||||||
| Created: | May 24, 2006 | Updated: | June 8, 2006 | ||||||||||||||||||||||||
| Description: | The dia drawing utility suffers from several format string vulnerabilities exploitable via a maliciously crafted dia file - or a file with a well-chosen name. | ||||||||||||||||||||||||||
| Alerts: |
| ||||||||||||||||||||||||||
hostapd: insufficient boundary checks
| Package(s): | hostapd | CVE #(s): | CVE-2006-2213 | ||||||||
| Created: | May 22, 2006 | Updated: | May 25, 2006 | ||||||||
| Description: | Matteo Rosi and Leonardo Maccari discovered that hostapd, a wifi network authenticator daemon, performs insufficient boundary checks on a key length value, which might be exploited to crash the service. | ||||||||||
| Alerts: |
| ||||||||||
kernel: denial of service
| Package(s): | kernel | CVE #(s): | CVE-2006-1859 CVE-2006-1860 | ||||||||||||||||
| Created: | May 19, 2006 | Updated: | May 24, 2006 | ||||||||||||||||
| Description: | Memory leak in __setlease in fs/locks.c in Linux kernel before 2.6.16.16
allows attackers to cause a denial of service (memory consumption) via
unspecified actions related to an "uninitialized return value," aka "slab
leak."
lease_init in fs/locks.c in Linux kernel before 2.6.16.16 allows attackers to cause a denial of service (fcntl_setlease lockup) via actions that cause lease_init to free a lock that might not have been allocated on the stack. | ||||||||||||||||||
| Alerts: |
| ||||||||||||||||||
kernel-patch-vserver: privilege escalation
| Package(s): | kernel-patch-vserver | CVE #(s): | CVE-2006-2110 | ||||
| Created: | May 22, 2006 | Updated: | May 24, 2006 | ||||
| Description: | Jan Rekorajski discovered that the kernel patch for virtual private servers does not limit context capabilities to the root user within the virtual server, which might lead to privilege escalation for some virtual server specific operations. | ||||||
| Alerts: |
| ||||||
kphone: insecure file creation
| Package(s): | kphone | CVE #(s): | CVE-2006-2442 | ||||||||
| Created: | May 22, 2006 | Updated: | May 25, 2006 | ||||||||
| Description: | Sven Dreyer discovered that KPhone, a Voice over IP client for KDE, creates a configuration file world-readable, which could leak sensitive information like SIP passwords. | ||||||||||
| Alerts: |
| ||||||||||
libextractor: heap-based buffer overflows
| Package(s): | libextractor | CVE #(s): | CVE-2006-2458 | ||||||||
| Created: | May 22, 2006 | Updated: | May 31, 2006 | ||||||||
| Description: | Luigi Auriemma has found two heap-based buffer overflows in libextractor 0.5.13 and earlier: one of them occurs in the asf_read_header function in the ASF plugin, and the other occurs in the parse_trak_atom function in the Qt plugin. | ||||||||||
| Alerts: |
| ||||||||||
mpg123: buffer overflows
| Package(s): | mpg123 | CVE #(s): | CVE-2006-1655 | ||||||||||||
| Created: | May 24, 2006 | Updated: | July 3, 2006 | ||||||||||||
| Description: | mpg123 does not properly validate MPEG 2.0 layer 3 files, leading to a number of buffer overflow vulnerabilities. | ||||||||||||||
| Alerts: |
| ||||||||||||||
OpenLDAP: boundary error
| Package(s): | openldap | CVE #(s): | |||||
| Created: | May 23, 2006 | Updated: | May 24, 2006 | ||||
| Description: | According to this Secunia advisory, a weakness exists in OpenLDAP which is caused due to a boundary error in slurpd within the handling of the status file. This can be exploited to cause a stack-based buffer overflow via an overly long hostname read from the status file. The weakness has been reported to be in OpenLDAP version 2.3.21 and earlier. | ||||||
| Alerts: |
| ||||||
phpbb2: missing input sanitizing
| Package(s): | phpbb2 | CVE #(s): | CVE-2006-1896 | ||||
| Created: | May 22, 2006 | Updated: | February 11, 2008 | ||||
| Description: | It was discovered that phpbb2, a web based bulletin board, insufficiently sanitizes values passed to the "Font Color 3" setting, which might lead to the execution of injected code by admin users. | ||||||
| Alerts: |
| ||||||
phpgroupware: missing input sanitizing
| Package(s): | phpgroupware | CVE #(s): | CVE-2005-2781 | ||||
| Created: | May 22, 2006 | Updated: | May 24, 2006 | ||||
| Description: | It was discovered that the Avatar upload feature of FUD Forum, a component of the web based groupware system phpgroupware, does not sufficiently validate uploaded files, which might lead to the execution of injected web script code. | ||||||
| Alerts: |
| ||||||
popfile: missing input sanitizing
| Package(s): | popfile | CVE #(s): | CVE-2006-0876 | ||||
| Created: | May 22, 2006 | Updated: | May 24, 2006 | ||||
| Description: | It has been discovered that popfile, a bayesian mail classifier, can be forced into a crash through malformed character sets within email messages, which allows denial of service. | ||||||
| Alerts: |
| ||||||
postgresql: SQL injection
| Package(s): | postgresql | CVE #(s): | CVE-2006-2313 CVE-2006-2314 | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| Created: | May 24, 2006 | Updated: | June 6, 2007 | ||||||||||||||||||||||||||||||||||||||||||||||||||||
| Description: | The PostgreSQL team has put out a set of "urgent updates" (in the form of the 7.3.15, 7.4.13, 8.0.8, and 8.1.4 releases) closing a newly-discovered set of SQL injection issues. Details about the problem can be found on the technical information page; in short: multi-byte encodings can be used to defeat normal string sanitizing techniques. The update fixes one problem related to invalid multi-byte characters, but punts on another by simply disallowing the old, unsafe technique of escaping single quotes with a backslash. | ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Alerts: |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
zoo: archive problem
| Package(s): | bin | CVE #(s): | |||||
| Created: | May 23, 2006 | Updated: | May 24, 2006 | ||||
| Description: | A security problem is zoo's fullpath() function could cause problems if zoo was run in an automated way, or if a user were to open a malicious zoo archive manually. | ||||||
| Alerts: |
| ||||||
Page editor: Jonathan Corbet
Next page:
Kernel development>>