If websocket data is sent with mixed text and binary in a single message, the binary data can be corrupted. This can result in an out-of-bounds read with the read memory sent to the originating server in response. This vulnerability affects Firefox < 60.

An out-of-bounds read in WebGL with a maliciously crafted "ImageInfo" object during WebGL operations. This vulnerability affects Firefox < 54, Firefox ESR < 52.2, and Thunderbird < 52.2.

An out of bounds read error occurs when parsing some HTTP digest authorization responses, resulting in information leakage through the reading of random memory containing matches to specifically set patterns. This vulnerability affects Firefox < 52 and Thunderbird < 52.

Insufficient bounds checking of data during inter-process communication might allow a compromised content process to be able to read memory from the parent process under certain conditions. This vulnerability affects Firefox < 66.

If a Sandbox content process is compromised, it can initiate an FTP download which will then use a child process to render the downloaded data. The downloaded data can then be passed to the Chrome process with an arbitrary file length supplied by an attacker, bypassing sandbox protections and allow for a potential memory read of adjacent data from the privileged Chrome process, which may include sensitive data. This vulnerability affects Firefox < 66.

The HTTP/2 protocol does not consider the role of the TCP congestion window in providing information about content length, which makes it easier for remote attackers to obtain cleartext data by leveraging a web-browser configuration in which third-party cookies are sent, aka a "HEIST" attack.

The HTTPS protocol does not consider the role of the TCP congestion window in providing information about content length, which makes it easier for remote attackers to obtain cleartext data by leveraging a web-browser configuration in which third-party cookies are sent, aka a "HEIST" attack.

A vulnerability exists in the Windows sandbox where an uninitialized value in memory can be leaked to a renderer from a broker when making a call to access an otherwise unavailable file. This results in the potential leaking of information stored at that memory location. *Note: this issue only occurs on Windows. Other operating systems are unaffected.*. This vulnerability affects Thunderbird < 60.7, Firefox < 67, and Firefox ESR < 60.7.

Mozilla Firefox before 48.0, Firefox ESR < 45.4 and Thunderbird < 45.4 allow remote attackers to obtain sensitive information about the previously retrieved page via Resource Timing API calls.

Mozilla Firefox 3.x before 3.0.5 allows remote attackers to bypass intended privacy restrictions by using the persist attribute in an XUL element to create and access data entities that are similar to cookies.

Mozilla Firefox 2.x before 2.0.0.18, Thunderbird 2.x before 2.0.0.18, and SeaMonkey 1.x before 1.1.13 do not properly change the source URI when processing a canvas element and an HTTP redirect, which allows remote attackers to bypass the same origin policy and access arbitrary images that are not directly accessible to the attacker. NOTE: this issue can be leveraged to enumerate software on the client by performing redirections related to moz-icon.

The XBM decoder in Mozilla Firefox before 2.0.0.17 and SeaMonkey before 1.1.12 allows remote attackers to read uninitialized memory, and possibly obtain sensitive information in opportunistic circumstances, via a crafted XBM image file.

Mozilla Firefox before 2.0.0.15 and SeaMonkey before 1.1.10 do not properly handle an invalid .properties file for an add-on, which allows remote attackers to read uninitialized memory, as demonstrated by use of ISO 8859 encoding instead of UTF-8 encoding in a French .properties file.

Firefox before 1.0.1 and Mozilla before 1.7.6 does not restrict xsl:include and xsl:import tags in XSLT stylesheets to the current domain, which allows remote attackers to determine the existence of files on the local system.

LiveConnect in Mozilla Firefox before 2.0.0.13 and SeaMonkey before 1.1.9 does not properly parse the content origin for jar: URIs before sending them to the Java plugin, which allows remote attackers to access arbitrary ports on the local machine. NOTE: this is closely related to CVE-2008-1195.

If a text string that happens to be a filename in the operating system's native format is dragged and dropped onto the addressbar the specified local file will be opened. This is contrary to policy and is what would happen if the string were the equivalent "file:" URL. This vulnerability affects Firefox < 60.

If an HTTP authentication prompt is triggered by a background network request from a page or extension, it is displayed over the currently loaded foreground page. Although the prompt contains the real domain making the request, this can result in user confusion about the originating site of the authentication request and may cause users to mistakenly send private credential information to a third party site. This vulnerability affects Firefox < 58.

A legacy extension's non-contentaccessible, defined resources can be loaded by an arbitrary web page through script. This script does this by using a maliciously crafted path string to reference the resources. Note: this vulnerability does not affect WebExtensions. This vulnerability affects Firefox < 59.

Mozilla Firefox before Firefox 2.0.0.13, and SeaMonkey before 1.1.9, can automatically install TLS client certificates with minimal user interaction, and automatically sends these certificates when requested, which makes it easier for remote web sites to track user activities across domains by requesting the TLS client certificates from other domains.

Using remote content in encrypted messages can lead to the disclosure of plaintext. This vulnerability affects Thunderbird ESR < 52.8 and Thunderbird < 52.8.

WebExtensions may use "view-source:" URLs to view local "file:" URL content, as well as content stored in "about:cache", bypassing restrictions that only allow WebExtensions to view specific content. This vulnerability affects Firefox < 59.

Same-origin protections for the PDF viewer can be bypassed, allowing a malicious site to intercept messages meant for the viewer. This could allow the site to retrieve PDF files restricted to viewing by an authenticated user on a third-party website. This vulnerability affects Firefox ESR < 52.8 and Firefox < 60.

Style editor traffic in the Developer Tools can be routed through a service worker hosted on a third party website if a user selects error links when these tools are open. This can allow style editor information used within Developer Tools to leak cross-origin. This vulnerability affects Firefox < 58.

Image for moz-icons can be accessed through the "moz-icon:" protocol through script in web content even when otherwise prohibited. This could allow for information leakage of which applications are associated with specific MIME types by a malicious page. This vulnerability affects Firefox < 59.

If a URL using the "file:" protocol is dragged and dropped onto an open tab that is running in a different child process the tab will open a local file corresponding to the dropped URL, contrary to policy. One way to make the target tab open more reliably in a separate process is to open it with the "noopener" keyword. This vulnerability affects Firefox < 60.

A shared worker created from a "data:" URL in one tab can be shared by another tab with a different origin, bypassing the same-origin policy. This vulnerability affects Firefox < 59.

Plaintext of decrypted emails can leak through the src attribute of remote images, or links. This vulnerability affects Thunderbird ESR < 52.8 and Thunderbird < 52.8.

WebExtensions can use request redirection and a "filterReponseData" filter to bypass host permission settings to redirect network traffic and access content from a host for which they do not have explicit user permission. This vulnerability affects Firefox < 60.

Development Tools panels of an extension are required to load URLs for the panels as relative URLs from the extension manifest file but this requirement was not enforced in all instances. This could allow the development tools panel for the extension to load a URL that it should not be able to access, including potentially privileged pages. This vulnerability affects Firefox < 58.

The "browser.identity.launchWebAuthFlow" function of WebExtensions is only allowed to load content over "https:" but this requirement was not properly enforced. This can potentially allow privileged pages to be loaded by the extension. This vulnerability affects Firefox < 58.

The WebService (XML-RPC) interface in Bugzilla 2.23.3 through 3.0.0 does not enforce permissions for the time-tracking fields of bugs, which allows remote attackers to obtain sensitive information via certain XML-RPC requests, as demonstrated by the (1) Deadline and (2) Estimated Time fields.

A previously installed malicious Android application with same signature-level permissions as Firefox can intercept AuthTokens meant for Firefox only. Note: This issue only affects Firefox for Android. Other versions and operating systems are unaffected. This vulnerability affects Firefox < 50.

The CheckLoadURI function in Mozilla Firefox 1.8 lists the about: URI as a ChromeProtocol and can be loaded via JavaScript, which allows remote attackers to obtain sensitive information by querying the browser's session history.

Mozilla Firefox 2.0, possibly only when running on Windows, allows remote attackers to bypass the Phishing Protection mechanism by representing an IP address in (1) dotted-hex, (2) dotted-octal, (3) single decimal integer, (4) single hex integer, or (5) single octal integer format, which is not captured by the blacklist filter.

The (1) Password Manager in Mozilla Firefox 2.0, and 1.5.0.8 and earlier; and the (2) Passcard Manager in Netscape 8.1.2 and possibly other versions, do not properly verify that an ACTION URL in a FORM element containing a password INPUT element matches the web site for which the user stored a password, which allows remote attackers to obtain passwords via a password INPUT element on a different web page located on the web site intended for this password.

Bugzilla 2.18.x before 2.18.6, 2.20.x before 2.20.3, 2.22.x before 2.22.1, and 2.23.x before 2.23.3 allow remote attackers to obtain (1) the description of arbitrary attachments by viewing the attachment in "diff" mode in attachment.cgi, and (2) the deadline field by viewing the XML format of the bug in show_bug.cgi.

Argument injection vulnerability in Mozilla Firefox 1.0.6 allows user-assisted remote attackers to modify command line arguments to an invoked mail client via " (double quote) characters in a mailto: scheme handler, as demonstrated by launching Microsoft Outlook with an arbitrary filename as an attachment. NOTE: it is not clear whether this issue is implementation-specific or a problem in the Microsoft API.

Mozilla Firefox 1.0.4 and 1.0.5 does not choose the challenge with the strongest authentication scheme available as required by RFC2617, which might cause credentials to be sent in plaintext even if an encrypted channel is available.

Bugzilla 2.19.1 through 2.20rc2 and 2.21, with user matching turned on in substring mode, allows attackers to list all users whose names match an arbitrary substring, even when the usevisibilitygroups parameter is set.

Bugzilla 2.17.1 through 2.18, 2.19.1, and 2.19.2, when a user is prompted to log in while attempting to view a chart, displays the password in the URL, which may allow local users to gain sensitive information from web logs or browser history.

Firefox before 1.0.5 and Mozilla before 1.7.9 allows a child frame to call top.focus and other methods in a parent frame, even when the parent is in a different domain, which violates the same origin policy and allows remote attackers to steal sensitive information such as cookies and passwords from web sites whose child frames do not verify that they are in the same domain as their parents.

Bugzilla 2.10 through 2.18, 2.19.1, and 2.19.2 displays a different error message depending on whether a product exists or not, which allows remote attackers to determine hidden products.

The find_replen function in jsstr.c in the Javascript engine for Mozilla Suite 1.7.6, Firefox 1.0.1 and 1.0.2, and Netscape 7.2 allows remote attackers to read portions of heap memory in a Javascript string via the lambda replace method.

The Form Fill feature in Firefox before 1.0.1 allows remote attackers to steal potentially sensitive information via an input control that monitors the values that are generated by the autocomplete capability.

Firefox before 1.0 and Mozilla before 1.7.5 allow remote attackers to obtain sensitive data from the clipboard via Javascript that generates a middle-click event on systems for which a middle-click performs a paste operation.

show_bug.cgi in Bugzilla 2.17.1 through 2.18rc2 and 2.19 from CVS, when using the insidergroup feature and exporting a bug to XML, shows comments and attachment summaries which are marked as private, which allows remote attackers to gain sensitive information.

Mozilla does not prevent cookies that are sent over an insecure channel (HTTP) from also being sent over a secure channel (HTTPS/SSL) in the same domain, which could allow remote attackers to steal cookies and conduct unauthorized activities, aka "Cross Security Boundary Cookie Injection."