Mozilla developers fixed several memory safety bugs in the browser engine used in Firefox and other Mozilla-based products. Some of these bugs showed evidence of memory corruption under certain circumstances, and we presume that with enough effort at least some of these could be exploited to run arbitrary code.
Security researcher Nicolas Golubovic reported that a malicious page can overwrite files on the user's machine using Content Security Policy (CSP) violation reports. The file contents are restricted to the JSON format of the report. In many cases overwriting a local file may simply be destructive, breaking the functionality of that file. The CSP error reports can include HTML fragments which could be rendered by browsers. If a user has disabled add-on signing and has installed an "unpacked" add-on, a malicious page could overwrite one of the add-on resources. Depending on how this resource is used, this could lead to privilege escalation.
Security researcher Muneaki Nishimura (nishimunea) of Recruit Technologies Co.,Ltd. reported that Content Security Policy (CSP) violation reports contained full path information for cross-origin iframe navigations in violation of the CSP specification. This could result in information disclosure.
Security researcher Ucha Gobejishvili reported a denial of service (DOS) attack when doing certain WebGL operations in a canvas requiring an unusually large amount buffer to be allocated from video memory. This resulted in memory resource exhaustion with some Intel video cards, requiring the computer to be rebooted to return functionality. This was resolved by putting in additional checks on the amount of memory to be allocated during graphics processing.
Security researchers Jose Martinez and Romina Santillan reported a memory leak in the libstagefright library when array destruction occurs during MPEG4 video file processing.
Security researcher Abdulrahman Alqabandi reported an issue where an attacker can load an arbitrary web page but the addressbar's displayed URL will be blank or filled with page defined content. This can be used to obfuscate which page is currently loaded and allows for an attacker to spoof an existing page without the malicious page's address being displayed correctly.
Security researcher Looben Yang reported a mechanism where the Clients API in Service Workers can be used to trigger an out-of-bounds read in ServiceWorkerManager. This results in a potentially exploitable crash.
Security researcher ca0nguyen, working with HP's Zero Day Initiative, reported a use-after-free issue in the HTML5 string parser when parsing a particular set of table-related tags in a foreign fragment context such as SVG. This results in a potentially exploitable crash.
Security researcher lokihardt, working with HP's Zero Day Initiative, reported a use-after-free issue in the SetBody function of HTMLDocument. This results in a potentially exploitable crash.
Security researcher Dominique Hazaël-Massieux reported a use-after-free issue when using multiple WebRTC data channel connections. This causes a potentially exploitable crash when a data channel connection is freed from within a call through it.
Security researcher Holger Fuhrmannek and Mozilla security engineer Tyson Smith reported a number of security vulnerabilities in the Graphite 2 library affecting version 1.3.5. The issue reported by Holger Fuhrmannek is a mechanism to induce stack corruption with a malicious graphite font. This leads to a potentially exploitable crash when the font is loaded. Tyson Smith used the Address Sanitizer tool in concert with a custom software fuzzer to find a series of uninitialized memory, out-of-bounds read, and out-of-bounds write errors when working with fuzzed graphite fonts.
Security researcher Oriol reported memory corruption when local files are modified (by either the user or another program) at the same time being read using the FileReader API. This flaw requires that input be taken from a local file in order to be triggered and cannot be triggered by web content. This results in a potentially exploitable crash when triggered.
Security researcher Nicolas Grégoire used the Address Sanitizer to find a use-after-free during XML transformation operations. This results in a potentially exploitable crash triggerable by web content.
Security researcher Tsubasa Iinuma reported a mechanism where the displayed addressbar can be spoofed to users. This issue involves using history navigation in concert with the Location protocol property. After navigating from a malicious page to another, if the user navigates back to the initial page, the displayed URL will not reflect the reloaded page. This could be used to trick users into potentially treating the page as a different and trusted site.
The Communications Electronics Security Group (UK) of the GCHQ reported a dangling pointer dereference within the Netscape Plugin Application Programming Interface (NPAPI) that could lead to the NPAPI subsystem crashing. This issue requires a maliciously crafted NPAPI plugin in concert with scripted web content, resulting in a potentially exploitable crash when triggered.
Security researcher Jordi Chancel discovered a variant of Mozilla Foundation Security Advisory 2015-136 which was fixed in Firefox 43. In the original bug, it was possible to read cross-origin URLs following a redirect if performance.getEntries() was used along with an iframe to host a page. Navigating back in history through script, content was pulled from the browser cache for the redirected location instead of going to the original location. In the newly reported variant issue, it was found that if a browser session was restored, history navigation would still allow for the same attack as content was restored from the browser cache. This is a same-origin policy violation and could allow for data theft.
Security researcher Luke Li reported a pointer underflow bug in the Brotli library's decompression that leads to a buffer overflow. This results in a potentially exploitable crash when triggered.
Security researcher Ronald Crane reported five "moderate" rated vulnerabilities affecting released code that were found through code inspection. These included the following issues in WebRTC: an integer underflow, a missing status check, race condition, and a use of deleted pointers to create new object. A race condition in LibVPX was also identified. These do not all have clear mechanisms to be exploited through web content but are vulnerable if a mechanism can be found to trigger them.
Security researcher Ronald Crane reported a race condition in GetStaticInstance in WebRTC which results in a use-after-free. This could result in a potentially exploitable crash. This issue was found through code inspection and does not have clear mechanism to be exploited through web content but is vulnerable if a mechanism can be found to trigger it.
Security researcher Ronald Crane reported an out-of-bounds read following a failed allocation in the HTML parser while working with unicode strings. This can also affect the parsing of XML and SVG format data. This leads to a potentially exploitable crash.