The Advanced Linux Sound Architecture (ALSA) library before 1.2.16.1 contains a double-free vulnerability in parse_def() in src/conf.c that allows attackers to corrupt memory by supplying maliciously crafted ALSA configuration text. When parsing nested compound or array configuration blocks, parse_def() fails to check return values before continuing, causing snd_config_delete() to be called twice on the same already-freed node, resulting in a NULL-pointer write or invalid memory read.

Angular is a development platform for building mobile and desktop web applications using TypeScript/JavaScript and other languages. Prior to 22.0.1, 21.2.17, and 20.3.25, Angular's HttpTransferCache caches HTTP requests made during Server-Side Rendering (SSR) so that they can be reused during client-side hydration. This avoids repeating the same HTTP requests on the client. The cached responses are stored in TransferState using a cache key generated by hashing request properties (method, response type, mapped URL, serialized body, and sorted query parameters). The cache keys are generated using a weak 32-bit DJB2-like polynomial rolling hash. The 32-bit hash space is extremely small, allowing attackers to find hash collisions. An attacker can easily find a query parameter string (e.g., q=aaCAZMMM for a search request) that produces the exact same 32-bit hash as a sensitive endpoint (e.g., /api/user/profile). When a victim visits a crafted link containing the colliding parameter, the SSR process executes both the search request and the profile request. Due to the hash collision, the search response overwrites the profile response in the TransferState cache. This vulnerability is fixed in 22.0.1, 21.2.17, and 20.3.25.

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.22.0, an assert-based security check in vLLM's activation function loading allows any unauthenticated attacker to achieve arbitrary code execution on the server by publishing a malicious HuggingFace model, when vLLM runs in Python optimized mode (python -O or PYTHONOPTIMIZE=1). This vulnerability is fixed in 0.22.0.

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, vLLM's /v1/audio/transcriptions endpoint limits compressed upload size but not decoded PCM output. A 25MB OPUS file expands to ~14.9GB of float32 PCM at decode time. This vulnerability is fixed in 0.23.1rc0.

Angular is a development platform for building mobile and desktop web applications using TypeScript/JavaScript and other languages. Prior to 22.0.1, 21.2.17, and 20.3.25, an information disclosure vulnerability exists in the @angular/service-worker package of the Angular framework. When the Service Worker fetches assets, it preserves metadata (such as headers) from the original request. However, on cross-origin redirects, the Service Worker fails to strip sensitive headers, violating the Fetch redirect algorithm. This allows a remote attacker to obtain sensitive credentials (e.g., Authorization tokens, Proxy-Authorization credentials, or session cookies) by triggering a cross-origin redirect to an untrusted external origin. This vulnerability is fixed in 22.0.1, 21.2.17, and 20.3.25.

vLLM is an inference and serving engine for large language models (LLMs). From 0.5.5 until 0.23.1rc0, integer truncation of tensor dimensions in vLLM's GGUF dequantize kernels (csrc/quantization/gguf/gguf_kernel.cu) causes partial tensor processing. The output tensor is allocated at full size via torch::empty (uninitialized memory), but the dequantize CUDA kernel processes only a truncated number of elements. The unfilled portion of the output tensor retains whatever was previously in GPU memory. In multi-tenant inference deployments, this residual GPU memory may contain tensor data from other users' inference requests, constituting information disclosure. This vulnerability is fixed in 0.23.1rc0.

Angular is a development platform for building mobile and desktop web applications using TypeScript/JavaScript and other languages. Prior to 22.0.0-rc.2, 21.2.15, 20.3.22, and 19.2.23, a vulnerability was discovered in @angular/common when Server-Side Rendering (SSR) and hydration are enabled. The HttpTransferCache utility optimizes hydration by caching outgoing HTTP requests performed during SSR and transferring the cached state to the client-side application via TransferState. However, the caching mechanism fails to inspect the withCredentials flag or the Cookie header of outgoing requests. As a result, credentialed, user-specific responses may be cached by default in the shared TransferState payload. When these responses are serialized into the HTML, any caching layer (such as a CDN, reverse proxy, or shared server cache) that caches the SSR-rendered HTML page could inadvertently cache and leak one user's private data to other users, leading to a high-severity information disclosure vulnerability. This vulnerability is fixed in 22.0.0-rc.2, 21.2.15, 20.3.22, and 19.2.23.

Angular is a development platform for building mobile and desktop web applications using TypeScript/JavaScript and other languages. Prior to 22.0.0-rc.2, 21.2.16, 20.3.24, and 19.2.25, a Cross-Site Scripting (XSS) vulnerability exists in @angular/platform-server's DOM emulation dependency (domino) when serializing the content of raw-text elements (such as <script>, <style>, and <iframe>). domino supports escaping raw-text elements during serialization to prevent closing-tag breakout. However, a Unicode index alignment bug existed in this escaping logic. In JavaScript, string lengths and character indices are calculated based on UTF-16 code units (where astral characters—such as emojis—occupy 2 code units / 4 bytes). If the bound dynamic text contained astral Unicode characters before the closing tag (e.g. </script>, </style>, or </iframe>), the index offset calculation in domino's replacement logic shifted. This misalignment caused domino to fail to replace or escape the closing tag, leaving it raw and unescaped in the output HTML. An attacker who controls the dynamic text can supply a payload containing both an astral Unicode character and a closing tag (e.g., 😀</iframe><script>alert(1)</script>). When serialized on the server during SSR, the browser parses the unescaped closing tag, exits the raw-text context early, and executes the subsequent <script> block, leading to same-origin Cross-Site Scripting (XSS). This vulnerability is fixed in 22.0.0-rc.2, 21.2.16, 20.3.24, and 19.2.25.

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.84.0, This vulnerability is fixed in 1.84.0.

WebOb provides objects for HTTP requests and responses. Prior to 1.8.10, the normalization of the HTTP Location header during a redirect is vulnerable to an open redirect: WebOb joins the redirect target to the request URI using Python's urljoin, and since Python 3.10 the underlying urlsplit strips ASCII tab, carriage return, and newline characters before parsing, so a redirect target containing such characters can be reinterpreted as a protocol-relative URL whose authority is an attacker-controlled host. This bypasses the CVE-2024-42353 fix that escaped a leading double slash, allowing an attacker who influences the redirect location to send users to an arbitrary external site instead of the intended one. This vulnerability is fixed in 1.8.10.