ghsa

Blind XXE vulnerabilities in jackrabbit-spi-commons and jackrabbit-core in Apache Jackrabbit < 2.23.2 due to usage of an unsecured document build to load privileges. Users are recommended to upgrade to versions 2.20.17 (Java 8), 2.22.1 (Java 11) or 2.23.2 (Java 11, beta versions), which fix this issue. Earlier versions (up to 2.20.16) are not supported anymore, thus users should update to the respective supported version.

py-libp2p before 0.2.3 allows a peer to cause a denial of service (resource consumption) via a large RSA key.

In Roundup before 2.5.0, XSS can occur via interaction between URLs and issue tracker templates (devel and responsive).

The attacker can use the raft server protocol in an unauthenticated way. The attacker can see the server's resources, including directories and files. This issue affects Apache Zeppelin: from 0.10.1 up to 0.12.0. Users are recommended to upgrade to version 0.12.0, which fixes the issue by removing the Cluster Interpreter.

The affected function, `MemBump::new()`, would allocate memory without initializing it. Subsequently calling the created value's various `alloc`methods would then read and write the start of that memory as a `Cell` which isundefined behavior. Instead, it should zero initialize the start of the allocated memory. For instance, some values could violate the internal invariants of the type and cause an assertion failure. Nevertheless, no deterministic read is known tocause further uninitialized memory to be exposed. Affected downstream users that can not upgrade are advised to call `MemBump::reset` immediately after allocation to manually perform the missing write of the counter best-as-possible. The flaw was corrected in commit d8d6a7d096d3aaafd963b356a8f1bbd8d26fd967 by zeroing the Cell at the start of the allocated memory.

A heap-buffer-overflow vulnerability in the loading of ExecuTorch methods can cause the runtime to crash and potentially result in code execution or other undesirable effects. This issue affects ExecuTorch prior to commit 93b1a0c15f7eda49b2bc46b5b4c49557b4e9810f

### Summary Using a CDN that caches (`/**/*.png`, `/**/*.json`, `/**/*.css`, etc...) requests, a cache deception can emerge. This could lead to unauthorized access to user sessions and personal data when cached responses are served to other users. ### Details The vulnerability occurs in the request processing logic where path sanitization is insufficient. The library splits the path using `config.basePath` but doesn't properly validate the remaining path components. This allows specially crafted requests that appear to be static assets (like `/api/auth/get-session/api/auth/image.png` assuming `config.basePath`=`/api/auth`) to bypass typical CDN cache exclusion rules while actually returning sensitive data. The problematic code [here](https://github.com/Bekacru/better-call/blob/8b6f13e24fad7f4666a582601517bb3232d4f4af/src/router.ts#L124): ```js const processRequest = async (request: Request) => { const url = new URL(request.url); const path = config?.basePath ? url.pathname.spl...

gif_outputAsJpeg in phpThumb through 1.7.23 allows phpthumb.gif.php OS Command Injection via a crafted parameter value. This is fixed in 1.7.23-202506081709.

Uncontrolled Recursion vulnerability in Apache Commons Lang. This issue affects Apache Commons Lang: Starting with commons-lang:commons-lang 2.0 to 2.6, and, from org.apache.commons:commons-lang3 3.0 before 3.18.0. The methods ClassUtils.getClass(...) can throw StackOverflowError on very long inputs. Because an Error is usually not handled by applications and libraries, a StackOverflowError could cause an application to stop. Users are recommended to upgrade to version 3.18.0, which fixes the issue.

A Regular Expression Denial of Service (ReDoS) vulnerability was discovered in the Hugging Face Transformers library, specifically within the DonutProcessor class's `token2json()` method. This vulnerability affects versions 4.51.3 and earlier, and is fixed in version 4.52.1. The issue arises from the regex pattern `<s_(.*?)>` which can be exploited to cause excessive CPU consumption through crafted input strings due to catastrophic backtracking. This vulnerability can lead to service disruption, resource exhaustion, and potential API service vulnerabilities, impacting document processing tasks using the Donut model.