URLs containing certain unicode glyphs for alternative hyphens and quotes do not properly trigger punycode display, allowing for domain name spoofing attacks in the location bar.
Event handlers on "marquee" elements were executed despite a strict Content Security Policy (CSP) that disallowed inline JavaScript. This vulnerability affects Firefox < 50.1, Firefox ESR < 45.6, and Thunderbird < 45.6.
On pages containing an iframe, the "data:" protocol can be used to create a modal alert that will render over arbitrary domains following page navigation, spoofing of the origin of the modal alert from the iframe content. This vulnerability affects Thunderbird < 52.3, Firefox ESR < 52.3, and Firefox < 55.
Using SVG filters that don't use the fixed point math implementation on a target iframe, a malicious page can extract pixel values from a targeted user. This can be used to extract history information and read text values across domains. This violates same-origin policy and leads to information disclosure.
Video files loaded video captions cross-origin without checking for the presence of CORS headers permitting such cross-origin use, leading to potential information disclosure for video captions.
Certain response codes in FTP connections can result in the use of uninitialized values for ports in FTP operations.
Last updated 24 July 2024
Last updated 24 July 2024
Last updated 24 July 2024
A malicious user or buggy application can inject code and trigger an infinite loop in inetdiagbcaudit()
Also make sure each instruction is aligned on 4 bytes boundary, to avoid unaligned accesses.
http://thread.gmane.org/gmane.linux.network/197206/focus=197386 http://patchwork.ozlabs.org/patch/100857/
Acknowledgements:
Red Hat would like to thank Dan Rosenberg for reporting this issue.
A signedness issue has been found in nextpidmap() function when the "last" parameter is negative as nextpidmap() just quietly accepted whatever "last" pid that was passed in, which is not all that safe when one of the users is /proc.
Setting fpos to negative value when accessing /proc via readdir()/getdents() resulted in sign extension of this value when map pointer was being constructed.
This later lead to #GP becasue the final pointer was not cannonical (x8664).
map = &pidns->pidmap[(last + 1)/BITSPERPAGE]; ffffffff810ac3b4: 48 63 f6 movslq %esi,%rsi <- sign extension (fpos+-) ffffffff810ac3b7: 81 e2 ff 7f 00 00 and $0x7fff,%edx ffffffff810ac3bd: 48 c1 ee 0f shr $0xf,%rsi ffffffff810ac3c1: 48 c1 e6 04 shl $0x4,%rsi ffffffff810ac3c5: 48 8d 5c 37 08 lea 0x8(%rdi,%rsi,1),%rbx ... if (unlikely(!map->page)) ffffffff810ac3cc: 48 8b 7b 08 mov 0x8(%rbx),%rdi <- #GP
Only x8664 architecture seems to be vulnerable to this particular issue (tested ppc64, s390x, ia64).
Reference: http://groups.google.com/group/fa.linux.kernel/browsethread/thread/93c1088451fd3522/4a28ecb7f755a88d?#4a28ecb7f755a88d
Upstream commit: http://git.kernel.org/linus/c78193e9 http://git.kernel.org/linus/d8bdc59f
Acknowledgements:
Red Hat would like to thank Robert Swiecki for reporting this issue.
Last updated 24 July 2024
Last updated 24 July 2024
Integer overflow in the agpgenericinsertmemory function in drivers/char/agp/generic.c in the Linux kernel before 2.6.38.5 allows local users to gain privileges or cause a denial of service (system crash) via a crafted AGPIOCBIND agpioctl ioctl call.
Created attachment 502027 [details] dmesg output with NFS/NLM/RPC debug
btw. I'm reporting this issue to akpm@ and to Trond Myklebust because other vendors and upstream linux kernel is affected too.
Description of problem: unprivileged user can use flock syscall for NFS share to trigger unlimited resource leak and this could cause node DoS.
Details: flock on NFS calls nlmclntlock() function, where nlmclntcall() executes synchronous RPC task NLMPROCLOCK. Via several subcalls (callbind) it successfully connects to NFS server and decodes its reply, however because server did not set up rport rpcbgetportdone() returns -EACCES, and following rpc subcall callbindstatus() in this case repeats callbind() after 3 sec timeout.
It isn't a big problem, and this cycle could be broken by signal, however error path of nlmclntlock() executes identical asynchronous RPC call to unlock taken lock. Waiting of result of this call is also canceled by signal, however in this case RPC call is asynchronous, it cycles too and never finished.
I would like to ask Trond to recheck my analyze and review my patch with fix of this issue or propose some better and more correct solution.
My idea is to break cycle described above, return -EOPNOTSUPP to nlmclntlock(), check it on error path and do not execute NLMPROCUNLOCK call because we already know that server does not support locking operations.
PS. one more separate issue: NLMPROCUNLOCK can be cycled in another place: nlmclntunlockcallback() restarts RPC task in case of any errors. Therefore we cannot execute NLMPROCUNLOCK on error path in case of server errors. I did not reproduce this issue, however I believe it's possible.
PPS. Also I've found similar cycles in other nlm callbacks, however I have no ideas is it possible to exploit it. However I would like to ask Trond to review this code too.
Version-Release number of selected component (if applicable): 2.6.32-131.0.15.el6.x8664 I've reproduced this issue on RHEL5, RHEL6, ubuntu 11.04 and upstream linux kernel 2.6.39.
How reproducible: 100%
Actual results: flock hangs, can be canceled by signal, as result one more rpc task leaks (FYI: sysctl -w sunrpc.rpcdebug=0 dumps status of active RPC taks)
Expected results: no hangs, no leaks
Additional info: dmesg output with NFS/NLM/RPC debug and concept of patch are attached. Because vendor-sec is no longer in use I'm preparing private report to akpm@ and to Trond Myklebust, probably they can propose some better solution
Last updated 24 July 2024
Mozilla Firefox before 15.0, Firefox ESR 10.x before 10.0.7, and SeaMonkey before 2.12 do not properly handle onLocationChange events during navigation between different https sites, which allows remote attackers to spoof the X.509 certificate information in the address bar via a crafted web page.
The format-number functionality in the XSLT implementation in Mozilla Firefox before 15.0, Firefox ESR 10.x before 10.0.7, Thunderbird before 15.0, Thunderbird ESR 10.x before 10.0.7, and SeaMonkey before 2.12 allows remote attackers to obtain sensitive information via unspecified vectors that trigger a heap-based buffer over-read.
The HZ-GB-2312 character-set implementation in Mozilla Firefox before 17.0, Firefox ESR 10.x before 10.0.11, Thunderbird before 17.0, Thunderbird ESR 10.x before 10.0.11, and SeaMonkey before 2.14 does not properly handle a ~ (tilde) character in proximity to a chunk delimiter, which allows remote attackers to conduct cross-site scripting (XSS) attacks via a crafted document.
The evalInSandbox implementation in Mozilla Firefox before 17.0, Firefox ESR 10.x before 10.0.11, Thunderbird before 17.0, Thunderbird ESR 10.x before 10.0.11, and SeaMonkey before 2.14 uses an incorrect context during the handling of JavaScript code that sets the location.href property, which allows remote attackers to conduct cross-site scripting (XSS) attacks or read arbitrary files by leveraging a sandboxed add-on.
Mozilla Firefox before 17.0, Firefox ESR 10.x before 10.0.11, Thunderbird before 17.0, Thunderbird ESR 10.x before 10.0.11, and SeaMonkey before 2.14 implement cross-origin wrappers with a filtering behavior that does not properly restrict write actions, which allows remote attackers to conduct cross-site scripting (XSS) attacks via a crafted web site.
Mozilla Firefox before 17.0, Firefox ESR 10.x before 10.0.11, Thunderbird before 17.0, Thunderbird ESR 10.x before 10.0.11, and SeaMonkey before 2.14 do not prevent use of a "top" frame name-attribute value to access the location property, which makes it easier for remote attackers to conduct cross-site scripting (XSS) attacks via vectors involving a binary plugin.
The nsLocation::CheckURL function in Mozilla Firefox before 16.0.2, Firefox ESR 10.x before 10.0.10, Thunderbird before 16.0.2, Thunderbird ESR 10.x before 10.0.10, and SeaMonkey before 2.13.2 does not properly determine the calling document and principal in its return value, which makes it easier for remote attackers to conduct cross-site scripting (XSS) attacks via a crafted web site, and makes it easier for remote attackers to execute arbitrary JavaScript code by leveraging certain add-on behavior.
The Chrome Object Wrapper (COW) implementation in Mozilla Firefox before 16.0, Firefox ESR 10.x before 10.0.8, Thunderbird before 16.0, Thunderbird ESR 10.x before 10.0.8, and SeaMonkey before 2.13 does not prevent access to properties of a prototype for a standard class, which allows remote attackers to execute arbitrary JavaScript code with chrome privileges via a crafted web site.
Mozilla Firefox before 16.0, Firefox ESR 10.x before 10.0.8, Thunderbird before 16.0, Thunderbird ESR 10.x before 10.0.8, and SeaMonkey before 2.13 do not properly restrict calls to DOMWindowUtils (aka nsDOMWindowUtils) methods, which allows remote attackers to bypass intended access restrictions via crafted JavaScript code.
Mozilla Firefox before 16.0, Firefox ESR 10.x before 10.0.8, Thunderbird before 16.0, Thunderbird ESR 10.x before 10.0.8, and SeaMonkey before 2.13 allow remote attackers to conduct cross-site scripting (XSS) attacks via a binary plugin that uses Object.defineProperty to shadow the top object, and leverages the relationship between top.location and the location property.
Mozilla Firefox before 16.0.1, Firefox ESR 10.x before 10.0.9, Thunderbird before 16.0.1, Thunderbird ESR 10.x before 10.0.9, and SeaMonkey before 2.13.1 omit a security check in the defaultValue function during the unwrapping of security wrappers, which allows remote attackers to bypass the Same Origin Policy and read the properties of a Location object, or execute arbitrary JavaScript code, via a crafted web site.
Mozilla Firefox before 16.0, Firefox ESR 10.x before 10.0.8, Thunderbird before 16.0, Thunderbird ESR 10.x before 10.0.8, and SeaMonkey before 2.13 do not properly manage history data, which allows remote attackers to conduct cross-site scripting (XSS) attacks or obtain sensitive POST content via vectors involving a location.hash write operation and history navigation that triggers the loading of a URL into the history object.
Mozilla Firefox before 18.0, Firefox ESR 10.x before 10.0.12 and 17.x before 17.0.2, Thunderbird before 17.0.2, Thunderbird ESR 10.x before 10.0.12 and 17.x before 17.0.2, and SeaMonkey before 2.15 allow remote attackers to spoof the address bar via vectors involving authentication information in the userinfo field of a URL, in conjunction with a 204 (aka No Content) HTTP status code.
The XBL.proto.toString implementation in Mozilla Firefox before 18.0, Firefox ESR 10.x before 10.0.12 and 17.x before 17.0.2, Thunderbird before 17.0.2, Thunderbird ESR 10.x before 10.0.12 and 17.x before 17.0.2, and SeaMonkey before 2.15 makes it easier for remote attackers to bypass the ASLR protection mechanism by calling the toString function of an XBL object.