| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| Software installed and run as a non-privileged user may conduct improper GPU system calls to subvert GPU HW to write to arbitrary physical memory pages.
Under certain circumstances this exploit could be used to corrupt data pages not allocated by the GPU driver but memory pages in use by the kernel and drivers running on the platform altering their behaviour. |
| conda-forge-metadata provides programatic access to conda-forge's metadata. conda-forge-metadata uses an optional dependency - "conda-oci-mirror" which was neither present on the PyPi repository nor registered by any entity. If conda-oci-mirror is taken over by a threat actor, it can result in remote code execution. |
| Access of Uninitialized Pointer vulnerability in TP-Link WR940N and WR941ND allows local unauthenticated attackers the ability to execute DoS attack
and potentially arbitrary code execution
under the context of the ‘root’ user.This issue affects WR940N and WR941ND: ≤ WR940N v5 3.20.1 Build 200316,
≤
WR941ND v6 3.16.9 Build 151203. |
| To trigger the issue, three configuration parameters must have specific settings: "hostname-char-set" must be left at the default setting, which is "[^A-Za-z0-9.-]"; "hostname-char-replacement" must be empty (the default); and "ddns-qualifying-suffix" must *NOT* be empty (the default is empty). DDNS updates do not need to be enabled for this issue to manifest. A client that sends certain option content would then cause kea-dhcp4 to exit unexpectedly.
This issue affects Kea versions 3.0.1 through 3.0.1 and 3.1.1 through 3.1.2. |
| The install() function of ProviderInstaller.java in Magisk App before canary version 27007 does not verify the GMS app before loading it, which allows a local untrusted app with no additional privileges to silently execute arbitrary code in the Magisk app and escalate privileges to root via a crafted package, aka Bug #8279. User interaction is not needed for exploitation. |
| A critical security vulnerability exists in remote cache extensions for common build systems utilizing bucket-based remote cache (such as those using Amazon S3, Google Cloud Storage, or similar object storage) that allows any contributor with pull request privileges to inject compromised artifacts from an untrusted environment into trusted production environments without detection.
The vulnerability exploits a fundamental design flaw in the "first-to-cache wins" principle, where artifacts built in untrusted environments (feature branches, pull requests) can poison the cache used by trusted environments (protected branches, production deployments).
This attack bypasses all traditional security measures including encryption, access controls, and checksum validation because the poisoning occurs during the artifact construction phase, before any security measures are applied. |
| An internal product security audit discovered a UEFI SMM (System Management Mode) callout vulnerability in some ThinkSystem servers that could allow a local attacker with elevated privileges to execute arbitrary code. |
| A vulnerability has been identified in Spectrum Power 4 (All versions < V4.70 SP12 Update 2). The affected application is vulnerable to run arbitrary commands via the user interface. This user interface can be used via the network and allows the execution of commands as administrative application user. |
| In the Linux kernel before 6.9, an untrusted hypervisor can inject virtual interrupt 29 (#VC) at any point in time and can trigger its handler. This affects AMD SEV-SNP and AMD SEV-ES. |
| Untrusted Pointer Dereference in I/O subsystem for some Intel(R) QAT software before version 2.0.5 may allow authenticated user to potentially enable information disclosure via local operating system access. |
| Untrusted data inclusion in pg_dump in PostgreSQL allows a malicious superuser of the origin server to inject arbitrary code for restore-time execution as the client operating system account running psql to restore the dump, via psql meta-commands. pg_dumpall is also affected. pg_restore is affected when used to generate a plain-format dump. This is similar to MySQL CVE-2024-21096. Versions before PostgreSQL 17.6, 16.10, 15.14, 14.19, and 13.22 are affected. |
| mtr through 0.95, in certain privileged contexts, mishandles execution of a program specified by the MTR_PACKET environment variable. NOTE: mtr on macOS may often have Sudo rules, as an indirect consequence of Homebrew not installing setuid binaries. |
| Fastjson before 1.2.48 mishandles autoType because, when an @type key is in a JSON document, and the value of that key is the name of a Java class, there may be calls to certain public methods of that class. Depending on the behavior of those methods, there may be JNDI injection with an attacker-supplied payload located elsewhere in that JSON document. This was exploited in the wild in 2023 through 2025. NOTE: this issue exists because of an incomplete fix for CVE-2017-18349. Also, a later bypass is covered by CVE-2022-25845. |
| theshit is a command-line utility that automatically detects and fixes common mistakes in shell commands. Prior to version 0.1.1, the application loads custom Python rules and configuration files from user-writable locations (e.g., `~/.config/theshit/`) without validating ownership or permissions when executed with elevated privileges. If the tool is invoked with `sudo` or otherwise runs with an effective UID of root, it continues to trust configuration files originating from the unprivileged user's environment. This allows a local attacker to inject arbitrary Python code via a malicious rule or configuration file, which is then executed with root privileges. Any system where this tool is executed with elevated privileges is affected. In environments where the tool is permitted to run via `sudo` without a password (`NOPASSWD`), a local unprivileged user can escalate privileges to root without additional interaction. The issue has been fixed in version 0.1.1. The patch introduces strict ownership and permission checks for all configuration files and custom rules. The application now enforces that rules are only loaded if they are owned by the effective user executing the tool. When executed with elevated privileges (`EUID=0`), the application refuses to load any files that are not owned by root or that are writable by non-root users. When executed as a non-root user, it similarly refuses to load rules owned by other users. This prevents both vertical and horizontal privilege escalation via execution of untrusted code. If upgrading is not possible, users should avoid executing the application with `sudo` or as the root user. As a temporary mitigation, ensure that directories containing custom rules and configuration files are owned by root and are not writable by non-root users. Administrators may also audit existing custom rules before running the tool with elevated privileges. |
| Inclusion of Functionality from Untrusted Control Sphere vulnerability in HCL MyXalytics. v6.6
allows Loading third-party scripts without integrity checks or validation can allow external code run in the application's context, risking data exposure. |
| Kernel software installed and running inside a Guest VM may exploit memory shared with the GPU Firmware to write data outside the Guest's virtualised GPU memory. |
| Software installed and run as a non-privileged user may conduct improper GPU system calls to trigger a crash of the FW running on the GPU freezing graphics output. |
| Improper syscall input validation in ASP (AMD Secure Processor) may force the kernel into reading syscall parameter values from its own memory space allowing an attacker to infer the contents of the kernel memory leading to potential information disclosure. |
| Kernel software installed and running inside a Guest VM may post improper commands to the GPU Firmware to read data outside the Guest's virtualised GPU memory. |
| Folo organizes feeds content into one timeline. Using pull_request_target on .github/workflows/auto-fix-lint-format-commit.yml can be exploited by attackers, since untrusted code can be executed having full access to secrets (from the base repo). By exploiting the vulnerability is possible to exfiltrate GITHUB_TOKEN which has high privileges. GITHUB_TOKEN can be used to completely overtake the repo since the token has content write privileges. This vulnerability is fixed in commit 585c6a591440cd39f92374230ac5d65d7dd23d6a. |
