Multiple cross-site scripting (XSS) vulnerabilities in Horde Application Framework 3.0.0 through 3.0.10 and 3.1.0 through 3.1.1 allow remote attackers to inject arbitrary web script or HTML via a (1) javascript URI or an external (2) http, (3) https, or (4) ftp URI in the url parameter in services/go.php (aka the dereferrer), (5) a javascript URI in the module parameter in services/help (aka the help viewer), and (6) the name parameter in services/problem.php (aka the problem reporting screen).

Off-by-one buffer overflow in the Instance_options::complete_initialization function in instance_options.cc in the Instance Manager in MySQL before 5.0.23 and 5.1 before 5.1.12 might allow local users to cause a denial of service (application crash) via unspecified vectors, which triggers the overflow when the convert_dirname function is called. NOTE: the vendor has disputed this issue via e-mail to CVE, saying that it is only exploitable when the user has access to the configuration file or the Instance Manager daemon. Due to intended functionality, this level of access would already allow the user to disrupt program operation, so this does not cross security boundaries and is not a vulnerability

Format string vulnerability in time.cc in MySQL Server 4.1 before 4.1.21 and 5.0 before 1 April 2006 allows remote authenticated users to cause a denial of service (crash) via a format string instead of a date as the first parameter to the date_format function, which is later used in a formatted print call to display the error message.

Linux kernel 2.6.x, when using both NFS and EXT3, allows remote attackers to cause a denial of service (file system panic) via a crafted UDP packet with a V2 lookup procedure that specifies a bad file handle (inode number), which triggers an error and causes an exported directory to be remounted read-only.

Integer overflow in FreeType before 2.2 allows remote attackers to cause a denial of service (crash) and possibly execute arbitrary code via a crafted PCF file, as demonstrated by the Red Hat bad1.pcf test file, due to a partial fix of CVE-2006-1861.

Unspecified vulnerability in the custom tag support for the TIFF library (libtiff) before 3.8.2 allows remote attackers to cause a denial of service (instability or crash) and execute arbitrary code via unknown vectors.

TIFF library (libtiff) before 3.8.2 allows context-dependent attackers to pass numeric range checks and possibly execute code, and trigger assert errors, via large offset values in a TIFF directory that lead to an integer overflow and other unspecified vectors involving "unchecked arithmetic operations".

The EstimateStripByteCounts function in TIFF library (libtiff) before 3.8.2 uses a 16-bit unsigned short when iterating over an unsigned 32-bit value, which allows context-dependent attackers to cause a denial of service via a large td_nstrips value, which triggers an infinite loop.

Heap-based buffer overflow in the NeXT RLE decoder in the TIFF library (libtiff) before 3.8.2 might allow context-dependent attackers to execute arbitrary code via unknown vectors involving decoding large RLE images.

Heap-based buffer overflow in the PixarLog decoder in the TIFF library (libtiff) before 3.8.2 might allow context-dependent attackers to execute arbitrary code via unknown vectors.

Heap-based buffer overflow in the JPEG decoder in the TIFF library (libtiff) before 3.8.2 allows context-dependent attackers to cause a denial of service and possibly execute arbitrary code via an encoded JPEG stream that is longer than the scan line size (TiffScanLineSize).

Multiple stack-based buffer overflows in the TIFF library (libtiff) before 3.8.2, as used in Adobe Reader 9.3.0 and other products, allow context-dependent attackers to execute arbitrary code or cause a denial of service via unspecified vectors, including a large tdir_count value in the TIFFFetchShortPair function in tif_dirread.c.

Zope 2.7.0 to 2.7.8, 2.8.0 to 2.8.7, and 2.9.0 to 2.9.3 (Zope2) does not disable the "raw" command when providing untrusted users with restructured text (reStructuredText) functionality from docutils, which allows local users to read arbitrary files.

Tor before 0.1.1.20 uses OpenSSL pseudo-random bytes (RAND_pseudo_bytes) instead of cryptographically strong RAND_bytes, and seeds the entropy value at start-up with 160-bit chunks without reseeding, which makes it easier for attackers to conduct brute force guessing attacks.

Tor before 0.1.1.20 does not validate that a server descriptor's fingerprint line matches its identity key, which allows remote attackers to spoof the fingerprint line, which might be trusted by users or other applications.

Tor client before 0.1.1.20 prefers entry points based on is_fast or is_stable flags, which could allow remote attackers to be preferred over nodes that are identified as more trustworthy "entry guard" (is_guard) systems by directory authorities.

Tor before 0.1.1.20 uses improper logic to validate the "OR" destination, which allows remote attackers to perform a man-in-the-middle (MITM) attack via unspecified vectors.

Tor before 0.1.1.20 supports server descriptors that contain hostnames instead of IP addresses, which allows remote attackers to arbitrarily group users by providing preferential address resolution.

The privoxy configuration file in Tor before 0.1.1.20, when run on Apple OS X, logs all data via the "logfile", which allows attackers to obtain potentially sensitive information.

Tor before 0.1.1.20 does not sufficiently obey certain firewall options, which allows remote attackers to bypass intended access restrictions for dirservers, direct connections, or proxy servers.