From ${URL} : Description A weakness and two vulnerabilities have been reported in Asterisk, which can be exploited by malicious people to determine valid usernames, cause a DoS (Denial of Service), and compromise a vulnerable system. 1) A boundary error in the "h264_format_attr_sdp_parse()" function (res/res_format_attr_h264.c) when parsing attribute resource for H.264 videos can be exploited to cause a stack-based buffer overflow via a specially crafted SDP header. Successful exploitation of this vulnerability may allow execution of arbitrary code. 2) An error within the "ast_http_get_post_vars()" function (main/http.c) when parsing Content-Length values in HTTP POST requests can be exploited to exhaust available memory and disrupt the service. 3) The application sends different responses when using a valid or an invalid username in INVITE, SUBSCRIBE, and REGISTER messages. This can be exploited to determine valid usernames by sending specially crafted INVITE, SUBSCRIBE, and REGISTER messages. Successful exploitation of this vulnerability requires alwaysauthreject to be enabled, allowguest to be disable, and autocreatepeer to be disabled. The vulnerabilities are reported in versions prior to 11.2.2. Solution Update to version 11.2.2. Provided and/or discovered by 1) Ulf Härnhammar 2) Christoph Hebeisen, TELUS Security Labs 3) Walter Doekes, OSSO B.V. Original Advisory http://downloads.asterisk.org/pub/security/AST-2013-001.html http://downloads.asterisk.org/pub/security/AST-2013-002.html http://downloads.asterisk.org/pub/security/AST-2013-003.html https://issues.asterisk.org/jira/browse/ASTERISK-20901 https://issues.asterisk.org/jira/browse/ASTERISK-20967 https://issues.asterisk.org/jira/browse/ASTERISK-21013
+*asterisk-11.2.2 (28 Mar 2013) +*asterisk-1.8.20.2 (28 Mar 2013) + + 28 Mar 2013; Tony Vroon <chainsaw@gentoo.org> -asterisk-1.8.19.1.ebuild, + -asterisk-1.8.20.0.ebuild, +asterisk-1.8.20.2.ebuild, + -asterisk-11.1.2.ebuild, -asterisk-11.2.0.ebuild, -asterisk-11.2.1-r2.ebuild, + +asterisk-11.2.2.ebuild: + Security upgrade to address a boundary error in H264 video SDP handling, + naive Content-Length variable parsing in HTTP POST requests and an + information leak around account existence for the SIP channel driver. Arches, please test & mark stable: =net-misc/asterisk-1.8.2.20 =net-misc/asterisk-11.2.2 Target keywords: amd64 x86 Once installed with USE="samples" to gain default configuration files, please try to stop & start the daemon three times in sequence. Unless further test hardware is available, this is a suitable and complete test.
amd64 stable
x86 stable
Updated existing GLSA draft.
+ 30 Mar 2013; Tony Vroon <chainsaw@gentoo.org> -asterisk-1.8.20.1.ebuild, + -asterisk-11.2.1.ebuild: + Remove vulnerable ebuilds after stabling, for security bug #463622.
CVE-2013-2686 (http://nvd.nist.gov/nvd.cfm?cvename=CVE-2013-2686): main/http.c in the HTTP server in Asterisk Open Source 1.8.x before 1.8.20.2, 10.x before 10.12.2, and 11.x before 11.2.2; Certified Asterisk 1.8.15 before 1.8.15-cert2; and Asterisk Digiumphones 10.x-digiumphones before 10.12.2-digiumphones does not properly restrict Content-Length values, which allows remote attackers to conduct stack-consumption attacks and cause a denial of service (daemon crash) via a crafted HTTP POST request. NOTE: this vulnerability exists because of an incorrect fix for CVE-2012-5976. CVE-2013-2685 (http://nvd.nist.gov/nvd.cfm?cvename=CVE-2013-2685): Stack-based buffer overflow in res/res_format_attr_h264.c in Asterisk Open Source 11.x before 11.2.2 allows remote attackers to execute arbitrary code via a long sprop-parameter-sets H.264 media attribute in a SIP Session Description Protocol (SDP) header. CVE-2013-2264 (http://nvd.nist.gov/nvd.cfm?cvename=CVE-2013-2264): The SIP channel driver in Asterisk Open Source 1.8.x before 1.8.20.2, 10.x before 10.12.2, and 11.x before 11.2.2; Certified Asterisk 1.8.15 before 1.8.15-cert2; Asterisk Business Edition (BE) C.3.x before C.3.8.1; and Asterisk Digiumphones 10.x-digiumphones before 10.12.2-digiumphones exhibits different behavior for invalid INVITE, SUBSCRIBE, and REGISTER transactions depending on whether the user account exists, which allows remote attackers to enumerate account names by (1) reading HTTP status codes, (2) reading additional text in a 403 (aka Forbidden) response, or (3) observing whether certain retransmissions occur.
This issue was resolved and addressed in GLSA 201401-15 at http://security.gentoo.org/glsa/glsa-201401-15.xml by GLSA coordinator Sergey Popov (pinkbyte).