Mozilla Firefox 0.10

Mozilla Firefox before 3.5.9 and 3.6.x before 3.6.2, and SeaMonkey before 2.0.4, executes a mail application in situations where an IMG element has a SRC attribute that is a redirect to a mailto: URL, which allows remote attackers to cause a denial of service (excessive application launches) via an HTML document with many images.

The UITour::onPageEvent function in Mozilla Firefox before 36.0 does not ensure that an API call originates from a foreground tab, which allows remote attackers to conduct spoofing and clickjacking attacks by leveraging access to a UI Tour web site.

Mozilla Firefox before 27.0 does not properly restrict access to about:home buttons by script on other pages, which allows user-assisted remote attackers to cause a denial of service (session restore) via a crafted web site.

The nsGIFDecoder2::GifWrite function in decoders/gif/nsGIFDecoder2.cpp in libpr0n in Mozilla Firefox before 3.5.5 allows remote attackers to cause a denial of service (NULL pointer dereference and application crash) via an animated GIF file with a large image size, a different vulnerability than CVE-2009-3373.

The NS_SecurityCompareURIs function in netwerk/base/public/nsNetUtil.h in Mozilla Firefox before 3.5.16 and 3.6.x before 3.6.13, and SeaMonkey before 2.0.11, does not properly handle (1) about:neterror and (2) about:certerror pages, which allows remote attackers to spoof the location bar via a crafted web site.

Stack-based buffer underflow in the mozilla::MP3FrameParser::ParseBuffer function in Mozilla Firefox before 36.0 allows remote attackers to obtain sensitive information from process memory via a malformed MP3 file that improperly interacts with memory allocation during playback.

Cross-site scripting (XSS) vulnerability in the MozSearch plugin implementation in Mozilla Firefox before 3.0.9 allows user-assisted remote attackers to inject arbitrary web script or HTML via a javascript: URI in the SearchForm element.

Mozilla Firefox before 2.0.0.10 and SeaMonkey before 1.1.7 sets the Referer header to the window or frame in which script is running, instead of the address of the content that initiated the script, which allows remote attackers to spoof HTTP Referer headers and bypass Referer-based CSRF protection schemes by setting window.location and using a modal alert dialog that causes the wrong Referer to be sent.

Cross-site scripting (XSS) vulnerability in Mozilla Firefox before 3.6.24 and 4.x through 7.0 and Thunderbird before 3.1.6 and 5.0 through 7.0 allows remote attackers to inject arbitrary web script or HTML via crafted text with Shift JIS encoding.

Mozilla Gecko before 5.0, as used in Firefox before 5.0 and Thunderbird before 5.0, does not block use of a cross-domain image as a WebGL texture, which allows remote attackers to obtain approximate copies of arbitrary images via a timing attack involving a crafted WebGL fragment shader.

Mozilla before 1.7.6, and Firefox before 1.0.1, allows remote attackers to spoof arbitrary web sites by injecting content from one window into a target window whose name is known but resides in a different domain, as demonstrated using a pop-up window on a trusted web site, aka the "window injection" vulnerability.

Mozilla Firefox before 3.0.9, Thunderbird, and SeaMonkey do not properly implement the Same Origin Policy for (1) XMLHttpRequest, involving a mismatch for a document's principal, and (2) XPCNativeWrapper.toString, involving an incorrect __proto__ scope, which allows remote attackers to conduct cross-site scripting (XSS) attacks and possibly other attacks via a crafted document.

intl/uconv/util/nsUnicodeDecodeHelper.cpp in Mozilla Firefox before 3.6.7 and Thunderbird before 3.1.1 inserts a U+FFFD sequence into text in certain circumstances involving undefined positions, which might make it easier for remote attackers to conduct cross-site scripting (XSS) attacks via crafted 8-bit text.

The jar: URI implementation in Mozilla Firefox before 3.0.9, Thunderbird, and SeaMonkey does not follow the Content-Disposition header of the inner URI, which allows remote attackers to conduct cross-site scripting (XSS) attacks and possibly other attacks via an uploaded .jar file with a "Content-Disposition: attachment" designation.

The WebRTC subsystem in Mozilla Firefox before 36.0 recognizes turns: and stuns: URIs but accesses the TURN or STUN server without using TLS, which makes it easier for man-in-the-middle attackers to discover credentials by spoofing a server and completing a brute-force attack within a short time window.

The XMLDocument::load function in Mozilla Firefox before 3.5.9 and 3.6.x before 3.6.2, Thunderbird before 3.0.4, and SeaMonkey before 2.0.4 does not perform the expected nsIContentPolicy checks during loading of content by XML documents, which allows attackers to bypass intended access restrictions via crafted content.

Mozilla Firefox before 16.0 on Android assigns chrome privileges to Reader Mode pages, which allows user-assisted remote attackers to bypass intended access restrictions via a crafted web site.

Firefox before 1.0.5 and Mozilla before 1.7.9 does not clearly associate a Javascript dialog box with the web page that generated it, which allows remote attackers to spoof a dialog box from a trusted site and facilitates phishing attacks, aka the "Dialog Origin Spoofing Vulnerability."

Mozilla Firefox before 2.0.0.16, and 3.x before 3.0.1, interprets '|' (pipe) characters in a command-line URI as requests to open multiple tabs, which allows remote attackers to access chrome:i URIs, or read arbitrary local files via manipulations involving a series of URIs that is not entirely handled by a vector application, as exploited in conjunction with CVE-2008-2540. NOTE: this issue exists because of an insufficient fix for CVE-2005-2267.

Firefox before 1.0.2 allows remote attackers to execute arbitrary code by tricking a user into saving a page as a Firefox sidebar panel, then using the sidebar panel to inject Javascript into a privileged page.

Firefox before 1.0.1 allows remote attackers to spoof the (1) security and (2) download modal dialog boxes, which could be used to trick users into executing script or downloading and executing a file, aka "Firespoofing."

Firefox before 1.0.1 and Mozilla before 1.7.6 truncates long sub-domains or paths for display, which may allow remote malicious web sites to spoof legitimate sites and facilitate phishing attacks.

Mozilla Firefox before 36.0 does not properly restrict transitions of JavaScript objects from a non-extensible state to an extensible state, which allows remote attackers to bypass a Caja Compiler sandbox protection mechanism or a Secure EcmaScript sandbox protection mechanism via a crafted web site.

Firefox before 1.0.1 and Mozilla before 1.7.6 allows remote attackers to spoof the SSL "secure site" lock icon via (1) a web site that does not finish loading, which shows the lock of the previous site, (2) a non-HTTP server that uses SSL, which causes the lock to be displayed when the SSL handshake is completed, or (3) a URL that generates an HTTP 204 error, which updates the icon and location information but does not change the display of the original site.

Firefox before 1.0 does not properly distinguish between user-generated and synthetic click events, which allows remote attackers to use Javascript to bypass the file download prompt when the user uses the Alt-click feature.

Firefox before 1.0.1 and Mozilla before 1.7.6, when displaying the HTTP Authentication dialog, do not change the focus to the tab that generated the prompt, which could facilitate spoofing and phishing attacks.

Firefox before 1.0 and Mozilla before 1.7.5 display the SSL lock icon when an insecure page loads a binary file from a trusted site, which could facilitate phishing attacks.

Firefox before 1.0.1 and Mozilla before 1.7.6 allows remote malicious web sites to spoof the extensions of files to download via the Content-Disposition header, which could be used to trick users into downloading dangerous content.

Firefox before 1.0.1 and Mozilla Suite before 1.7.6 use a predictable filename for the plugin temporary directory, which allows local users to delete arbitrary files of other users via a symlink attack on the plugtmp directory.