update to 1.4.14 fixing CVE-2014-0480 CVE-2014-0481 CVE-2014-0482, CVE-2014-0483

Note that Tenable Network Security has extracted the preceding description block directly from the Fedora security advisory. Tenable has attempted to automatically clean and format it as much as possible without introducing additional issues.

Security fix for CVE-2014-0480, CVE-2014-0481, CVE-2014-0482, CVE-2014-0483

Note that Tenable Network Security has extracted the preceding description block directly from the Fedora security advisory. Tenable has attempted to automatically clean and format it as much as possible without introducing additional issues.

The remote host is affected by the vulnerability described in GLSA-201412-22 (Django: Multiple vulnerabilities)

    Multiple vulnerabilities have been discovered in Django. Please review       the CVE identifiers referenced below for details.
  Impact :

    A remote attacker may be able to create a Denial of Service condition,       obtain sensitive information, or hijack web sessions.
  Workaround :

    There is no known workaround at this time.

This update address an issue with reverse() generating external URLs;
a denial of service involving file uploads; a potential session hijacking issue in the remote-user middleware; and a data leak in the administrative interface.

http://www.freexian.com/services/debian-lts.html

CVE-2014-0480

Django includes the helper function django.core.urlresolvers.reverse, typically used to generate a URL from a reference to a view function or URL pattern name. However, when presented with input beginning with two forward-slash characters (//), reverse() could generate scheme-relative URLs to other hosts, allowing an attacker who is aware of unsafe use of reverse() (i.e., in a situation where an end user can control the target of a redirect, to take a common example) to generate links to sites of their choice, enabling phishing and other attacks.

To remedy this, URL reversing now ensures that no URL starts with two slashes (//), replacing the second slash with its URL encoded counterpart (%2F). This approach ensures that semantics stay the same, while making the URL relative to the domain and not to the scheme.

CVE-2014-0481

In the default configuration, when Django's file upload handling system is presented with a file that would have the same on-disk path and name as an existing file, it attempts to generate a new unique filename by appending an underscore and an integer to the end of the (as stored on disk) filename, incrementing the integer (i.e., _1, _2, etc.) until it has generated a name which does not conflict with any existing file.

An attacker with knowledge of this can exploit the sequential behavior of filename generation by uploading many tiny files which all share a filename; Django will, in processing them, generate ever-increasing numbers of os.stat() calls as it attempts to generate a unique filename. As a result, even a relatively small number of such uploads can significantly degrade performance.

To remedy this, Django's file-upload system will no longer use sequential integer names to avoid filename conflicts on disk; instead, a short random alphanumeric string will be appended, removing the ability to reliably generate many repeatedly-conflicting filenames.

CVE-2014-0482

Django provides a middleware -- django.contrib.auth.middleware.RemoteUserMiddleware -- and an authentication backend, django.contrib.auth.backends.RemoteUserBackend, which use the REMOTE_USER header for authentication purposes.

In some circumstances, use of this middleware and backend could result in one user receiving another user's session, if a change to the REMOTE_USER header occurred without corresponding logout/login actions.

To remedy this, the middleware will now ensure that a change to REMOTE_USER without an explicit logout will force a logout and subsequent login prior to accepting the new REMOTE_USER.

CVE-2014-0483

Django's administrative interface, django.contrib.admin, offers a feature whereby related objects can be displayed for selection in a popup window. The mechanism for this relies on placing values in the URL and querystring which specify the related model to display and the field through which the relationship is implemented. This mechanism does perform permission checks at the level of the model class as a whole.

This mechanism did not, however, verify that the specified field actually represents a relationship between models.
Thus a user with access to the admin interface, and with sufficient knowledge of model structure and the appropriate URLs, could construct popup views which would display the values of non-relationship fields, including fields the application developer had not intended to expose in such a fashion.

To remedy this, the admin interface will now, in addition to its normal permission checks, verify that the specified field does indeed represent a relationship, to a model registered with the admin, and will raise an exception if either condition is not true.

NOTE: Tenable Network Security has extracted the preceding description block directly from the DLA security advisory. Tenable has attempted to automatically clean and format it as much as possible without introducing additional issues.

ID	名稱	產品	系列	已發布	已更新	嚴重性
77591	Fedora 20 : python-django14-1.4.14-1.fc20 (2014-9788)	Nessus	Fedora Local Security Checks	2014/9/10	2021/1/11	medium
77614	Fedora 20 : python-django15-1.5.9-1.fc20 (2014-9866)	Nessus	Fedora Local Security Checks	2014/9/11	2021/1/11	medium
79975	GLSA-201412-22 : Django: Multiple vulnerabilities	Nessus	Gentoo Local Security Checks	2014/12/15	2021/1/6	medium
82210	Debian DLA-65-1 : python-django security update	Nessus	Debian Local Security Checks	2015/3/26	2021/1/11	medium

偵測

搜尋 Plugin

medium

medium

medium

medium