Mozilla Bugzilla

Buffer overflow in LHA allows remote attackers to execute arbitrary code via long pathnames in LHarc format 2 headers for a .LHZ archive, as originally demonstrated using the "x" option but also exploitable through "l" and "v", and fixed in header.c, a different issue than CVE-2004-0771.

SQL injection vulnerability in Bugzilla 2.16.3 and earlier, and 2.17.1 through 2.17.4, allows remote authenticated users with editkeywords privileges to execute arbitrary SQL via the id parameter to editkeywords.cgi.

SQL injection vulnerability in collectstats.pl for Bugzilla 2.16.3 and earlier allows remote authenticated users with editproducts privileges to execute arbitrary SQL via the product name.

CGI.pl in Bugzilla before 2.14.1, when using LDAP, allows remote attackers to obtain an anonymous bind to the LDAP server via a request that does not include a password, which causes a null password to be sent to the LDAP server.

A third party website can access information available to a user with access to a restricted bug entry using the image generation in report.cgi in all Bugzilla versions prior to 4.4.

Util.pm in Bugzilla 2.x, 3.x, and 4.x before 4.2.15, 4.3.x and 4.4.x before 4.4.10, and 5.x before 5.0.1 mishandles long e-mail addresses during account registration, which allows remote attackers to obtain the default privileges for an arbitrary domain name by placing that name in a substring of an address, as demonstrated by truncation of an @mozilla.com.example.com address to an @mozilla.com address.

Bugzilla 2.10 allows remote attackers to execute arbitrary commands via shell metacharacters in a username that is then processed by (1) the Bugzilla_login cookie in post_bug.cgi, or (2) the who parameter in process_bug.cgi.

Bugzilla 2.14 through 2.22.7; 3.0.x, 3.1.x, and 3.2.x before 3.2.10; 3.4.x before 3.4.10; 3.6.x before 3.6.4; and 4.0.x before 4.0rc2 does not properly generate random values for cookies and tokens, which allows remote attackers to obtain access to arbitrary accounts via unspecified vectors, related to an insufficient number of calls to the srand function.

bugzilla_email_append.pl in Bugzilla 2.14.x before 2.14.4, and 2.16.x before 2.16.1, allows remote attackers to execute arbitrary code via shell metacharacters in a system call to processmail.

Bugzilla before 2.14.1 allows remote attackers to inject arbitrary SQL code and create files or gain privileges via (1) the sql parameter in buglist.cgi, (2) invalid field names from the "boolean chart" query in buglist.cgi, (3) the mybugslink parameter in userprefs.cgi, (4) a malformed bug ID in the buglist parameter in long_list.cgi, and (5) the value parameter in editusers.cgi, which allows groupset privileges to be modified by attackers with blessgroupset privileges.

The default .htaccess scripts for Bugzilla 2.14.x before 2.14.5, 2.16.x before 2.16.2, and 2.17.x before 2.17.3 do not include filenames for backup copies of the localconfig file that are made from editors such as vi and Emacs, which could allow remote attackers to obtain a database password by directly accessing the backup file.

Bugzilla 2.10 allows remote attackers to access sensitive information, including the database username and password, via an HTTP request for the globals.pl file, which is normally returned by the web server without being executed.

The offer_account_by_email function in User.pm in the WebService for Bugzilla before 3.0.2, and 3.1.x before 3.1.2, does not check the value of the createemailregexp parameter, which allows remote attackers to bypass intended restrictions on account creation.

Bugzilla before 2.14 does not properly escape untrusted parameters, which could allow remote attackers to conduct unauthorized activities via cross-site scripting (CSS) and possibly SQL injection attacks on (1) the product or output form variables for reports.cgi, (2) the voteon, bug_id, and user variables for showvotes.cgi, (3) an invalid email address in createaccount.cgi, (4) an invalid ID in showdependencytree.cgi, (5) invalid usernames and other fields in process_bug.cgi, and (6) error messages in buglist.cgi.

Bugzilla before 2.14.1 allows remote attackers to (1) spoof a user comment via an HTTP request to process_bug.cgi using the "who" parameter, instead of the Bugzilla_login cookie, or (2) post a bug as another user by modifying the reporter parameter to enter_bug.cgi, which is passed to post_bug.cgi.

The process_bug.cgi script in Bugzilla allows remote attackers to execute arbitrary commands via shell metacharacters.

Bugzilla 2.14 before 2.14.2, and 2.16 before 2.16rc2, may allow remote attackers to cause a denial of service or execute certain queries via a SQL injection attack on the sort order parameter to buglist.cgi.

Bugzilla before 2.14 does not properly restrict access to confidential bugs, which could allow Bugzilla users to bypass viewing permissions via modified bug id parameters in (1) process_bug.cgi, (2) show_activity.cgi, (3) showvotes.cgi, (4) showdependencytree.cgi, (5) showdependencygraph.cgi, (6) showattachment.cgi, or (7) describecomponents.cgi.

editproducts.cgi in Bugzilla 2.14.x before 2.14.4, and 2.16.x before 2.16.1, when the "usebuggroups" feature is enabled and more than 47 groups are specified, does not properly calculate bit values for large numbers, which grants extra permissions to users via known features of Perl math that set multiple bits.

post_bug.cgi in Bugzilla 2.10 through 2.18, 2.19.1, and 2.19.2 allows remote authenticated users to "enter bugs into products that are closed for bug entry" by modifying the URL to specify the name of the product.

The shadow database feature (syncshadowdb) in Bugzilla 2.9 through 2.16.10 allows local users to overwrite arbitrary files via a symlink attack on temporary files.

Bugzilla 2.14 before 2.14.2, and 2.16 before 2.16rc2, when configured to perform reverse DNS lookups, allows remote attackers to bypass IP restrictions by connecting from a system with a spoofed reverse DNS hostname.

SQL injection vulnerability in the Bug.search WebService function in Bugzilla 3.3.2 through 3.4.1, and 3.5, allows remote attackers to execute arbitrary SQL commands via unspecified parameters.

SQL injection vulnerability in the Bug.create WebService function in Bugzilla 2.23.4 through 3.0.8, 3.1.1 through 3.2.4, and 3.3.1 through 3.4.1 allows remote attackers to execute arbitrary SQL commands via unspecified parameters.

The mod_perl initialization script in Bugzilla 2.23.3 does not set the Bugzilla Apache configuration to allow .htaccess permissions to override file permissions, which allows remote attackers to obtain the database username and password via a direct request for the localconfig file.

describecomponents.cgi in Bugzilla 2.17.3 and 2.17.4 does not properly verify group membership when bug entry groups are used, which allows remote attackers to list component descriptions for otherwise restricted products.

Cross-site scripting vulnerabilities in Bugzilla 2.14 before 2.14.2, and 2.16 before 2.16rc2, could allow remote attackers to execute script as other Bugzilla users via the full name (real name) field, which is not properly quoted by editusers.cgi.

Bugzilla 2.19.3 through 2.20 does not properly handle "//" sequences in URLs when redirecting a user from the login form, which could cause it to generate a partial URL in a form action that causes the user's browser to send the form data to another domain.

Bugzilla 2.16.10 does not properly handle certain characters in the (1) maxpatchsize and (2) maxattachmentsize parameters in attachment.cgi, which allows remote attackers to trigger a SQL error.

Bugzilla before 2.14 allows Bugzilla users to bypass group security checks by marking a bug as the duplicate of a restricted bug, which adds the user to the CC list of the restricted bug and allows the user to view the bug.

editproducts.cgi in Bugzilla 2.16.3 and earlier, when usebuggroups is enabled, does not properly remove group add privileges from a group that is being deleted, which allows users with those privileges to perform unauthorized additions to the next group that is assigned with the original group ID.

Bugzilla before 2.14 stores user passwords in plaintext and sends password requests in an email message, which could allow attackers to gain privileges.

Bugzilla before 2.14 includes the username and password in URLs, which could allow attackers to gain privileges by reading the information from the web server logs, or by "shoulder-surfing" and observing the web browser's location bar.

Unknown vulnerability in the administrative controls in Bugzilla 2.17.1 through 2.17.7 allows users with "grant membership" privileges to grant memberships to groups that the user does not control.

Bugzilla 2.14 before 2.14.2, and 2.16 before 2.16rc2, when performing a mass change, sets the groupset of all bugs to the groupset of the first bug, which could inadvertently cause insecure groupset permissions to be assigned to some bugs.

Bugzilla 2.14 before 2.14.2, and 2.16 before 2.16rc2, does not properly handle URL-encoded field names that are generated by some browsers, which could cause certain fields to appear to be unset, which has the effect of removing group permissions on bugs when buglist.cgi is provided with the encoded field names.

Bugzilla 2.16.x before 2.16.1 does not properly filter apostrophes from an email address during account creation, which allows remote attackers to execute arbitrary SQL via a SQL injection attack.

SQL injection vulnerability in editusers.cgi in Bugzilla 2.16.x before 2.16.6, and 2.18 before 2.18rc1, allows remote attackers with privileges to grant membership to any group to execute arbitrary SQL.

Bugzilla 3.2.1, 3.0.7, and 3.3.2, when running under mod_perl, calls the srand function at startup time, which causes Apache children to have the same seed and produce insufficiently random numbers for random tokens, which allows remote attackers to bypass cross-site request forgery (CSRF) protection mechanisms and conduct unauthorized activities as other users.

Directory traversal vulnerability in importxml.pl in Bugzilla before 2.22.5, and 3.x before 3.0.5, when --attach_path is enabled, allows remote attackers to read arbitrary files via an XML file with a .. (dot dot) in the data element.