In Eclipse Mosquitto versions 2.1 through 2.1.2, a listener configured with `protocol websockets` mishandles empty WebSocket data frames. `net__read_ws()` in `lib/net_ws.c` returns the number of framing bytes consumed (the 4 masking-key bytes) for a data frame with a zero-length payload, without writing any bytes to the caller’s buffer. The MQTT packet reader in `packet__read_single()` (`lib/packet_mosq.c`) trusts that return value as a payload byte count. A remote client can therefore advance the read position over an uninitialised heap allocation, causing the broker to process and forward uninitialised heap memory (information disclosure, including heap and libc pointers), and can underflow the unsigned `to_process` counter so that subsequent attacker-controlled frame data is written past the end of the allocation (heap out-of-bounds write). The out-of-bounds write is reachable from the CONNECT packet before any authentication check, so an unauthenticated client can corrupt the heap and crash the broker even when `allow_anonymous false` is set. Combining the disclosure and the write allows remote code execution as the broker’s user by any client able to establish a session.
In Eclipse Mosquitto versions 2.1.0 to 2.1.2, a broker with a listener that is configured to use the PROXY protocol v2 is vulnerable to an untrusted proxy providing details that cause the broker to accept fake SSL information.
This can be mitigated by following the documented requirements for PROXY listeners: It is very important that only trusted proxies are allowed to connect to any listener using the PROXY protocol, and that clients are not allowed to connect to the listener at all.
In Eclipse versions 1.5-1.6.13, 2.0.0-2.0.22 and 2.1.0-2.1.2, the client library libmosquitto does not reject Will messages with payloads greater than 65535 bytes when calling `mosquitto_will_set()` or `mosquitto_will_set_v5()`. This results in up to 65535 bytes of uninitialised heap memory being transmitted as the Will.
This can be mitigated by only passing payloads with a maximum length of 65535 bytes.
In the Apache Airflow Microsoft Azure provider 10.2.0 through 15.3.0, `S3ToAzureBlobStorageOperator` turned off TLS certificate verification for its connection to Amazon S3 by default (`s3_verify=False`), whatever the AWS connection's own verification settings were. An attacker positioned on the network between an Airflow worker and S3 could intercept that traffic, read the signed requests and alter the listed object keys and the copied file contents. Deployments are affected when a Dag uses `S3ToAzureBlobStorageOperator` without setting `s3_verify`. Users are advised to upgrade to apache-airflow-providers-microsoft-azure 15.3.1 or later, which uses the AWS connection's verification settings by default.
A bug in the Airlock Gateway (8.5.2 and earlier, and 8.6.0) allows attacker-controlled HTML in a backend response to exploit the Smart Form Protection feature, possibly leading to a Denial of Service (DoS) or Remote Code Execution (RCE) attack against the Airlock Gateway. The issue has been fixed in Airlock Gateway version 8.5.3 and 8.6.1.
For an attack to be executable, the Airlock Gateway needs to have 'Smart Form Protection' (or the corresponding expert settings) enabled on a mapping whose backend returns a response containing certain multi-byte HTML character references. The buffer overflow then happens during the parsing of that HTML response. Such a response can come either from a malicious or compromised backend, or from an otherwise trusted backend that returns attacker-supplied HTML, e.g. an application that stores and serves plain user-provided HTML files.
Successfully exploiting the vulnerability can enable an attacker to perform a DoS attack or, in the worst case, an RCE attack against the Airlock Gateway in front of the affected backend. Compromising the functionality of the Airlock Gateway could in this way also affect the availability and protection of other well-intentioned backends deployed behind the same Gateway.
Zohocorp Manageengine DDI Central builds below 6500 contain an authenticated command injection vulnerability in the Windows DDI Agent's clear_blocked_domains functionality. This flaw allows specially crafted blocked-domain values to execute arbitrary commands on affected Windows systems. The vulnerability has a high severity CVSS score of 9.1, indicating critical impact on confidentiality, integrity, and availability.
ManageEngine DDI Central builds below 6500 contain an authenticated PowerShell command injection vulnerability in the Telnet connectivity feature. This flaw allows attackers with valid credentials to execute arbitrary commands on managed Windows nodes by leveraging specially crafted IPv6 scope identifiers. The vulnerability has a high severity with a CVSS score of 9.1, indicating critical impact on confidentiality, integrity, and availability.
The WP Delicious – Recipe Plugin for Food Bloggers (formerly Delicious Recipes) plugin for WordPress is vulnerable to Stored Cross-Site Scripting via the 'recipeTitle' parameter in all versions up to, and including, 1.10.10 due to insufficient input sanitization and output escaping. This makes it possible for authenticated attackers, with contributor-level access and above, to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. WordPress core's block-attribute kses pass strips HTML tags but preserves bare double-quote characters, allowing a payload such as x" onerror="alert(1) to survive both the save pipeline and the render-time strip_tags() call, breaking out of the alt attribute of the rendered img element.
The WP Delicious – Recipe Plugin for Food Bloggers (formerly Delicious Recipes) plugin for WordPress is vulnerable to Stored Cross-Site Scripting via the 'details[].icon' parameter in all versions up to, and including, 1.10.11 due to insufficient input sanitization and output escaping. This makes it possible for authenticated attackers, with contributor-level access and above, to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. The payload bypasses wp_kses_post on save because Gutenberg serializes block attribute JSON with JSON_HEX_QUOT, encoding double-quotes as \u0022 so no raw quote or angle bracket appears in post_content; parse_blocks() restores the literal quote at render time, producing the HTML attribute breakout.
** UNSUPPORTED WHEN ASSIGNED **
Improper Authentication in Apache SeaTunnel Web 1.0.2 allows a user whose account has been disabled to authenticate with valid credentials and obtain a new login token. The login query does not check the account status, so disabling an account does not prevent new logins or access to protected APIs using the newly issued token. This issue requires valid credentials for the disabled account; it is not a credential-free authentication bypass.
Apache SeaTunnel Web is no longer maintained. The affected 1.0.2 release does not contain the account-status check added to the main branch in pull request #264. There is no maintained fixed release. Users should discontinue use of this unsupported component.
NOTE: This vulnerability only affects products that are no longer supported by the maintainer.
An incorrect implementation of authentication algorithm vulnerability in the SSO API in Synology DiskStation Manager (DSM) before 7.2.2-72806-5 and 7.3.1-86003-1 (7.2.1-69057 is not affected) allows remote attackers to read or write arbitrary files and conduct denial-of-service attacks with prior knowledge of the distinguished name (DN).
GNU Emacs 31.1 contains a code injection vulnerability in its handling of directory-local variables. A crafted .dir-locals.el can supply a (REGEXP FUNCTION t) entry in auto-mode-alist; when a matching file is opened, Emacs calls FUNCTION before set-auto-mode--dir-local-valid-p validates it, executing arbitrary Emacs Lisp with the privileges of the Emacs process. (Emacs 30.2 is unaffected.)
In OpenStack Cinder before 29.1.1, by making an os-initialize_connection request with an IQN that includes a path traversal as part of the connector's initiator element, an authenticated user with a 'member' role may create or overwrite files in the volume host filesystem. Only cinder-volume hosts using the LIO iSCSI target driver are affected. This target driver is typically used with the LVM volume driver.
A stack-based out-of-bounds write vulnerability exists in the GNU C Library (glibc) tdelete function during tree rebalancing. This occurs in versions from 2.1 up to but not including 2.39_2 when deleting nodes in a sufficiently deep binary tree. The flaw can cause a pointer to be written past the end of a stack-allocated array, potentially crashing the application. Exploitation requires a very large tree (depth 40 or multiples of 20) with at least a million nodes, making it difficult to trigger. No known affected applications in common distributions have been identified.
The OneLogin SAML SSO WordPress plugin versions before 3.7.1 contain a vulnerability where a user-supplied parameter is not validated before being used to redirect an authenticated user. This flaw allows unauthenticated attackers to craft links that redirect logged-in users to arbitrary external domains, potentially facilitating phishing attacks.
The ProfilePress WordPress plugin versions before 4.18.1 contain a flaw in processing membership plan changes. The plugin does not atomically consume the source subscription, which allows customers to perform concurrent plan-change checkouts on a single paid subscription. This can result in multiple active memberships being obtained for the price of one.