Opera before 9.26 allows user-assisted remote attackers to execute arbitrary script via images that contain custom comments, which are treated as script when the user displays the image properties.

Opera before 9.26 allows remote attackers to "bypass sanitization filters" and conduct cross-site scripting (XSS) attacks via crafted attribute values in an XML document, which are not properly handled during DOM presentation.

Opera before 9.25 allows remote attackers to conduct cross-domain scripting attacks via unknown vectors related to plug-ins.

Unspecified vulnerability in Opera before 9.25 allows remote attackers to execute arbitrary code via crafted TLS certificates.

The rich text editing functionality in Opera before 9.25 allows remote attackers to conduct cross-domain scripting attacks by using designMode to modify contents of pages in other domains.

Algorithmic complexity vulnerability in Opera 9.50 beta and 9.x before 9.25 allows remote attackers to cause a denial of service (CPU consumption) via a crafted bitmap (BMP) file that triggers a large number of calculations and checks.

Opera before 9.25 allows remote attackers to obtain potentially sensitive memory contents via a crafted bitmap (BMP) file, as demonstrated using a CANVAS element and JavaScript in an HTML document for copying these contents from 9.50 beta, a related issue to CVE-2008-0420.

Unspecified vulnerability in Adobe Flash Player 9.0.47.0 and earlier, when running on Opera before 9.24 on Mac OS X, has unknown "Highly Severe" impact and unknown attack vectors.

Unspecified vulnerability in Opera before 9.24 allows remote attackers to overwrite functions on pages from other domains and bypass the same-origin policy via unknown vectors.

Unspecified vulnerability in Opera before 9.24, when using an "external" newsgroup or e-mail client, allows remote attackers to execute arbitrary commands via unknown vectors.

The canvas.createPattern function in Opera 9.x before 9.22 for Linux, FreeBSD, and Solaris does not clear memory before using it to process a new pattern, which allows remote attackers to obtain sensitive information (memory contents) via JavaScript.

Opera before 9.23 allows remote attackers to execute arbitrary code via crafted Javascript that triggers a "virtual function call on an invalid pointer."

Use-after-free vulnerability in the BitTorrent support in Opera before 9.22 allows user-assisted remote attackers to execute arbitrary code via a crafted header in a torrent file, which leaves a dangling pointer to an invalid object.

Buffer overflow in the transfer manager in Opera before 9.21 for Windows allows user-assisted remote attackers to execute arbitrary code via a crafted torrent file. NOTE: due to the lack of details, it is not clear if this is the same issue as CVE-2007-2274.

Adobe Macromedia Flash Player 7 and 9, when used with Opera before 9.20 or Konqueror before 20070613, allows remote attackers to obtain sensitive information (browser keystrokes), which are leaked to the Flash Player applet.

Opera 9.10 does not check URLs embedded in (1) object or (2) iframe HTML tags against the phishing site blacklist, which allows remote attackers to bypass phishing protection.

The FTP protocol implementation in Opera 9.10 allows remote attackers to allows remote servers to force the client to connect to other servers, perform a proxied port scan, or obtain sensitive information by specifying an alternate server address in an FTP PASV response.

The child frames in Opera 9 before 9.20 inherit the default charset from the parent window when a charset is not specified in an HTTP Content-Type header or META tag, which allows remote attackers to conduct cross-site scripting (XSS) attacks, as demonstrated using the UTF-7 character set.

Opera 9.10 Final allows remote attackers to bypass the Fraud Protection mechanism by adding certain characters to the end of a domain name, as demonstrated by the "." and "/" characters, which is not caught by the blacklist filter.

Mozilla Firefox 2.0.0.1 allows remote attackers to bypass the Phishing Protection mechanism by adding certain characters to the end of the domain name, as demonstrated by the "." and "/" characters, which is not caught by the Phishing List blacklist filter.