Heap-based buffer overflow in the wordwrap function in PHP after 4.1.2 and before 4.3.0 may allow attackers to cause a denial of service or execute arbitrary code.

Internet Message (IM) 141-18 and earlier uses predictable file and directory names, which allows local users to (1) obtain unauthorized directory permissions via a temporary directory used by impwagent, and (2) overwrite and create arbitrary files via immknmz.

Apache Tomcat 4.0.5 and earlier, when using both the invoker servlet and the default servlet, allows remote attackers to read source code for server files or bypass certain protections, a variant of CAN-2002-1148.

Multiple vulnerabilities in KDE 2 and KDE 3.x through 3.0.5 do not quote certain parameters that are inserted into a shell command, which could allow remote attackers to execute arbitrary commands via (1) URLs, (2) filenames, or (3) e-mail addresses.

faxspool in mgetty before 1.1.29 uses a world-writable spool directory for outgoing faxes, which allows local users to modify fax transmission privileges.

Buffer overflow in cnd-program for mgetty before 1.1.29 allows remote attackers to cause a denial of service and possibly execute arbitrary code via a Caller ID string with a long CallerName argument.

The daemon for GeneWeb before 4.09 does not properly handle requested paths, which allows remote attackers to read arbitrary files via a crafted URL.

Buffer overflow in typespeed 0.4.2 and earlier allows local users to gain privileges via long input.

Cross-site scripting (XSS) vulnerability in MHonArc before 2.5.14 allows remote attackers to inject arbitrary HTML into web archive pages via HTML mail messages.

The spray mode in traceroute-nanog (aka traceroute-ng) may allow local users to overwrite arbitrary memory locations via an array index overflow using the nprobes (number of probes) argument.

Buffer overflow in traceroute-nanog (aka traceroute-ng) may allow local users to execute arbitrary code via a long hostname argument.

openwebmail_init in Open WebMail 1.81 and earlier allows local users to execute arbitrary code via .. (dot dot) sequences in a login name, such as the name provided in the sessionid parameter for openwebmail-abook.pl, which is used to find a configuration file that specifies additional code to be executed.

Integer overflow in pdftops, as used in Xpdf 2.01 and earlier, xpdf-i, and CUPS before 1.1.18, allows local users to execute arbitrary code via a ColorSpace entry with a large number of elements, as demonstrated by cups-pdf.

Multiple integer overflows in Common Unix Printing System (CUPS) 1.1.14 through 1.1.17 allow remote attackers to execute arbitrary code via (1) the CUPSd HTTP interface, as demonstrated by vanilla-coke, and (2) the image handling code in CUPS filters, as demonstrated by mksun.

Macromedia Flash Player before 6.0.65.0 allows remote attackers to execute arbitrary code via certain malformed data headers in Shockwave Flash file format (SWF) files, a different issue than CAN-2002-0846.

Format string vulnerability in daemon.c for Exim 4.x through 4.10, and 3.x through 3.36, allows exim administrative users to execute arbitrary code by modifying the pid_file_path value.

Linux kernel 2.2.x allows local users to cause a denial of service (crash) by using the mmap() function with a PROT_READ parameter to access non-readable memory pages through the /proc/pid/mem interface.

OpenLDAP2 (OpenLDAP 2) 2.2.0 and earlier allows remote or local attackers to execute arbitrary code when libldap reads the .ldaprc file within applications that are running with extra privileges.

Multiple buffer overflows in OpenLDAP2 (OpenLDAP 2) 2.2.0 and earlier allow remote attackers to execute arbitrary code via (1) long -t or -r parameters to slurpd, (2) a malicious ldapfilter.conf file that is not properly handled by getfilter functions, (3) a malicious ldaptemplates.conf that causes an overflow in libldap, (4) a certain access control list that causes an overflow in slapd, or (5) a long generated filename for logging rejected replication requests.

vim 6.0 and 6.1, and possibly other versions, allows attackers to execute arbitrary commands using the libcall feature in modelines, which are not sandboxed but may be executed when vim is used to edit a malicious file, as demonstrated using mutt.