Suse
Monthly
OpenEXR versions 3.0.0-3.2.8, 3.3.0-3.3.10, and 3.4.0-3.4.10 suffer from unbounded shift operations in the readVariableLengthInteger() function when parsing variable-length integers from untrusted EXR files. Attackers can craft malicious EXR files with excessive continuation bytes to trigger left shifts exceeding 64 bits on a 64-bit integer, causing undefined behavior that may lead to information disclosure or denial of service. The vulnerability is remotely exploitable without authentication or user interaction against any application processing untrusted EXR input; no public exploit code has been identified at the time of analysis.
Policy rollback vulnerability in gittuf versions up to 0.13.1 allows attackers with push access to the Reference State Log (RSL) to downgrade repository policies to previously signed versions, bypassing security controls. An attacker cannot roll back to policies that would be unsigned by the current root keys, but can selectively choose any valid prior policy state. Vendor-released patch: gittuf v0.14.0 introduces monotonically increasing version numbers to all policy metadata to prevent rollback attacks.
ShellHub Community v0.24.1 and earlier allows authenticated API Key holders to enumerate any tenant namespace and retrieve sensitive membership data (user IDs, emails, roles), settings, and device counts via GET /api/namespaces/:tenant due to a bypassed authorization check. The vulnerability exploits API Key authentication flows that fail to set the user ID header, causing the membership verification to be skipped entirely. Publicly available proof-of-concept code demonstrates validated exploitation against v0.24.1, with complete disclosure of cross-tenant namespace configuration including member lists suitable for targeted phishing campaigns.
Kubetail Dashboard prior to version 0.14.0 fails to validate the Origin header on WebSocket connection upgrades, enabling Cross-Site WebSocket Hijacking (CSWSH) attacks. An authenticated user visiting a malicious web page can be exploited to stream their Kubernetes container logs-including credentials, tokens, and PII often present in logs-to an attacker-controlled server. The vulnerability affects both desktop deployments at localhost:7500 and cluster deployments behind HTTP basic auth, with browser ambient credentials automatically attached to the WebSocket handshake.
CSS Parser gem disables HTTPS certificate validation by setting OpenSSL::SSL::VERIFY_NONE, allowing man-in-the-middle attackers to inject or modify CSS content loaded via HTTPS. Any application using CSS Parser versions prior to 2.1.0 to fetch external stylesheets over HTTPS can be exploited by network-positioned attackers without authentication. A proof-of-concept using mitmproxy or Burp Suite demonstrates practical exploitation; CVSS 5.8 reflects the network attack vector and integrity impact, but real-world risk depends on whether the application loads stylesheets from untrusted or attacker-controllable URLs and whether the attacker can intercept network traffic.
Gotenberg versions 8.31.0 and earlier allow unauthenticated remote attackers to enumerate and read arbitrary files under /tmp/ via the /forms/chromium/convert/url and /forms/chromium/screenshot/url endpoints using file:// scheme URLs. An attacker can discover in-flight conversion request directories and exfiltrate source files (HTML, Markdown, Office documents, staged PDFs) from other users' concurrent conversion requests by timing attacks to coincide with long-running conversion operations. The vulnerability exploits a logic flaw where the URL routes fail to set per-request scope guards that HTML/Markdown routes correctly apply, causing file:// access control enforcement to silently skip for URL-based conversions.
Arbitrary PDF file read vulnerability in Gotenberg versions up to 8.31.0 allows unauthenticated remote attackers to extract PDF content via path traversal in stampExpression and watermarkExpression parameters on six conversion routes (pdfengines/merge, pdfengines/split, libreoffice/convert, chromium/convert/url, chromium/convert/html, chromium/convert/markdown). The vulnerability exists because these routes accept user-controlled file paths without validation when stamp or watermark source is set to PDF, unlike the dedicated stamp/watermark routes which enforce file upload requirements. An attacker can read any PDF accessible to the Gotenberg process by specifying its filesystem path, gaining access to potentially sensitive documents in containerized deployments or systems with mounted directories.
DNS rebinding vulnerability in Gotenberg allows unauthenticated remote attackers to bypass SSRF protections and access internal services via Chromium URL conversion routes. When a URL is submitted for PDF conversion, Gotenberg validates the resolved IP address against a deny-list but discards the pinned result. Chromium then performs independent DNS resolution multiple times, creating a race condition where an attacker controlling DNS can return a public IP during validation and a private IP during connection, allowing access to loopback services, cloud metadata endpoints, or internal networks. Exploitation succeeds approximately 10% per attempt with trivial automation.
CRLF injection in Netty's RedisEncoder allows remote command injection and response poisoning by injecting carriage return and line feed characters into InlineCommandRedisMessage, SimpleStringRedisMessage, and ErrorRedisMessage objects. Attackers can inject arbitrary Redis commands (such as CONFIG SET, FLUSHALL, or authentication bypass) or forge fake responses when user-controlled input is placed into these message types without sanitization. The vulnerability affects Netty 4.2.12.Final and all prior versions with the codec-redis module; no active exploitation has been reported in CISA KEV, but publicly available proof-of-concept code demonstrates the vulnerability.
HTTP request smuggling in Netty's HttpRequestDecoder allows remote unauthenticated attackers to inject arbitrary HTTP requests by sending malformed Transfer-Encoding headers (specifically 'Transfer-Encoding: chunked, identity'). When Netty is deployed behind a proxy that forwards such requests without rejection, an attacker can smuggle a second request inside the body of the first, bypassing security controls and accessing unintended resources. The vulnerability is confirmed by public proof-of-concept code demonstrating successful parsing of injected requests.
HTTP request smuggling in Netty's chunk size parser allows remote unauthenticated attackers to inject arbitrary HTTP requests by exploiting integer overflow in the hexadecimal chunk size parsing logic. The HttpObjectDecoder.getChunkSize method accumulates the chunk size without proper overflow validation, enabling an attacker to craft a malicious chunk size header that wraps around to a valid size, causing Netty to misinterpret the request boundary and parse injected requests as separate legitimate requests. Publicly available proof-of-concept demonstrates successful parsing of an injected GET request within a chunked POST body, with CVSS score 6.5 (network-accessible, low complexity, no authentication required).
Cross-site scripting (XSS) via improper Markdown attribute sanitization in Weblate allows authenticated users to inject malicious HTML attributes through crafted Markdown in user comments and other user-provided content. The vulnerability affects Weblate versions prior to 5.17.1 and can be exploited by any authenticated user with the ability to submit Markdown content; however, the application's strict Content Security Policy (CSP) significantly mitigates actual exploitation risk. Vendor-released patch version 5.17.1 is available.
Weblate versions before 5.17.1 allow authenticated users to enumerate translations in projects they cannot access via the screenshots, tasks, and component link API endpoints. An attacker with valid credentials but no project access can probe these APIs to discover the existence and metadata of private translations, leading to information disclosure of project structure and language coverage that should remain hidden. The vulnerability requires authentication but has a low attack complexity, affecting confidentiality only without enabling further compromise.
BSON operator injection in ShellHub Community v0.24.1 and earlier allows any authenticated user to crash the device list API endpoint by supplying MongoDB operators as filter field names or sort_by values without input validation. Two attack vectors exist in api/store/mongo/query-options.go: unvalidated filter param.Name values become raw BSON map keys, and the sort_by query parameter is passed directly into a $sort pipeline stage. A validated proof-of-concept exists against v0.24.1. No active exploitation is confirmed (not in CISA KEV), and EPSS sits at 0.06% (19th percentile), indicating low real-world exploitation probability despite a public PoC.
Cross-tenant IDOR in ShellHub Community v0.24.1 allows any authenticated user to read SSH session records belonging to other namespaces by supplying an arbitrary session UID to `GET /api/sessions/:uid`. The `GetSession` function in `api/services/session.go` omits the `InNamespace` tenant filter present elsewhere, so the `Authorize` middleware validates token presence but not resource ownership. Disclosed data includes SSH username, device UID, remote IP, geolocation, terminal type, authenticated flag, and session timestamps - enabling cross-tenant reconnaissance. No public exploitation is confirmed and EPSS is 0.03% (8th percentile), though the advisory includes a working proof-of-concept curl script.
Cross-tenant IDOR in ShellHub Community's device API exposes full device metadata across namespace boundaries to any authenticated user. An authenticated attacker who knows or can obtain a target device UID can issue a crafted GET request to /api/devices/:uid and receive the victim tenant's device object - including hostname, MAC address, OS fingerprint, public SSH key, namespace name, last-seen timestamp, and remote IP address. A validated proof-of-concept exists against v0.24.1; no public exploit identified at time of analysis and no CISA KEV listing, though EPSS is 0.03% (8th percentile), reflecting low but real probability of opportunistic exploitation in multi-tenant deployments.
Stored XSS in Kyverno policy-reporter-ui 2.5.1 and earlier allows an attacker with Kubernetes write access to PolicyReport resources to inject arbitrary HTML into the Vue 3 frontend, which executes in the browser of any authenticated UI user who views the affected namespace. The sink is the v-html directive in PropertyCard.vue, which is reached whenever a property value is not an http/https URL and exceeds 75 characters - a bypass condition trivially satisfied by padding any HTML payload. A public proof-of-concept demonstrating session cookie exfiltration via a crafted PolicyReport object is included in the GitHub advisory GHSA-q98m-7w8c-w388; no public exploit identified as confirmed actively exploited (not in CISA KEV), though EPSS of 0.03% reflects low opportunistic risk given the Kubernetes-specific prerequisite.
Cross-tenant information disclosure in Hatchet's `GET /api/v1/stable/dags/tasks` endpoint permits any authenticated user on a multi-tenant instance to read task metadata belonging to other tenants by supplying a victim's tenant UUID and DAG UUID in query parameters. The CVSS 5.3 (AV:N/AC:H/PR:L) reflects that an attacker must hold a valid account and obtain victim UUIDs through out-of-band channels before exploitation - but on Hatchet Cloud, where account registration is open by default, the account prerequisite is trivially satisfied. No public exploit code exists, no exploitation was observed prior to patching, and vendor-released fix v0.83.39 was deployed within hours of the April 23, 2026 report.
Denial of service in vLLM v0.18.0-v0.19.1 permanently crashes the EngineCore process when any API request includes a sampling penalty parameter against a deployment configured with `extract_hidden_states` speculative decoding. A refactor in PR #37013 (v0.18.0) removed the `.unsqueeze(-1)` call that normalized tensor shape, causing a broadcast shape mismatch (`(batch_size, 2)` vs. expected `(batch_size, 1)`) during penalty application and an unrecoverable RuntimeError. The crash is deterministic - a single request with `repetition_penalty: 1.1` suffices - but no public exploit code has been identified at time of analysis and EPSS is very low (0.04%, 12th percentile).
Hugo static site generator versions 0.43.0 through 0.160.x allow unrestricted file system access when building sites with Node-based asset pipelines (PostCSS, Babel, TailwindCSS), enabling arbitrary code execution through these tools to read or write files outside the project directory. The vulnerability affects only users who deploy Node asset pipelines; those building trusted sites or not using these pipelines are unaffected. A vendor-released patch in v0.161.0 enforces Node's permission model with strict filesystem defaults.
Integer overflow in Chrome's Network component prior to version 148.0.7778.96 allows remote attackers to bypass same-origin policy after compromising the renderer process via a crafted HTML page. The vulnerability requires renderer compromise and user interaction, limiting real-world risk despite network-accessible attack surface. No active exploitation has been confirmed; a vendor patch is available.
Insufficient input validation in the FileSystem API in Google Chrome prior to version 148.0.7778.96 allows a remote attacker with a compromised renderer process to perform arbitrary file read and write operations through a malicious HTML page. The vulnerability requires prior renderer compromise and user interaction with a crafted page, limiting its attack surface despite network-accessible vectors. Vendor has released a patched version.
Insufficient policy enforcement in DirectSockets API in Google Chrome prior to version 148.0.7778.96 allows remote attackers to perform arbitrary read and write operations through a crafted malicious Chrome extension. The vulnerability requires user interaction (extension installation) but affects the core security model of the browser's socket access control, exposing local network resources to unauthorized access by untrusted extensions.
Google Chrome prior to version 148.0.7778.96 contains insufficient validation of untrusted input in the Permissions system, allowing attackers on the local network segment to leak cross-origin data through malicious network traffic. The vulnerability requires network adjacency but no user interaction or authentication, affecting confidentiality with medium Chromium severity. Patch is available from Google.
Information disclosure in Google Chrome prior to 148.0.7778.96 allows remote attackers who have compromised the renderer process to extract potentially sensitive data from process memory through a race condition triggered by a crafted HTML page. The vulnerability requires renderer process compromise and user interaction but results in high confidentiality impact with no integrity or availability consequences. Chromium security team rates this as Medium severity; no active exploitation has been publicly confirmed.
Uncontrolled XSS vulnerability in Google Chrome's ServiceWorker implementation prior to version 148.0.7778.96 allows attackers to inject arbitrary scripts or HTML into web pages when users install a malicious Chrome extension, bypassing same-origin policy protections. The vulnerability requires user interaction (extension installation) but can be exploited remotely with low attack complexity. CVSS score of 5.4 reflects medium severity; no active exploitation has been publicly reported.
Uninitialized memory use in the GPU component of Google Chrome prior to version 148.0.7778.96 allows remote attackers who have compromised the renderer process to extract potentially sensitive information from process memory through a malicious HTML page. The vulnerability requires renderer process compromise as a precondition and user interaction to trigger, but once achieved, enables confidentiality breach with no code execution or denial of service impact. Vendor-released patch available in Chrome 148.0.7778.96.
Unvalidated Omnibox input in Google Chrome prior to version 148.0.7778.96 enables remote attackers to inject arbitrary scripts and HTML (universal XSS) via malicious network traffic, affecting users who click on crafted links. The vulnerability requires user interaction but crosses security boundaries due to its scope impact. No active exploitation has been publicly confirmed, though a patch is available from Google.
Insufficient policy enforcement in Chrome Extensions prior to version 148.0.7778.96 allows a remote attacker with a compromised renderer process to bypass discretionary access controls via a crafted HTML page, potentially leading to unauthorized information disclosure or modification. The vulnerability requires user interaction and a pre-existing renderer compromise, limiting its practical exploitation scope. A vendor-released patch is available in Chrome 148.0.7778.96 and later.
Out-of-bounds read and write in the GFX component of Google Chrome prior to version 148.0.7778.96 allows remote attackers to perform arbitrary memory read and write operations through malicious network traffic. The vulnerability requires user interaction (clicking a link or visiting a malicious page) but does not require authentication. Chromium rated the security severity as Medium; CVSS 5.4 reflects moderate impact (confidentiality and integrity compromise without availability loss). No active exploitation confirmed in CISA KEV at time of analysis.
UI spoofing via insufficient input validation in Google Chrome prior to 148.0.7778.96 allows a remote attacker who has already compromised the renderer process to craft malicious HTML pages that deceive users about the legitimate UI elements. The attack requires renderer process compromise as a prerequisite and user interaction with the crafted page, limiting real-world applicability to multi-stage attacks. Chromium assessed this as medium severity; no public exploit code or active exploitation has been identified.
Insufficient policy enforcement in Chrome's WebUI on Linux, Mac, Windows, and ChromeOS prior to 148.0.7778.96 allows a remote attacker with a compromised renderer process to bypass site isolation via a crafted HTML page, potentially exposing sensitive cross-site data. The vulnerability requires user interaction (UI:R) and prior renderer compromise, limiting its standalone exploitability. Vendor-released patch available in version 148.0.7778.96.
Insufficient input validation in ANGLE (Almost Native Graphics Layer Engine) in Google Chrome prior to version 148.0.7778.96 allows remote attackers who have compromised the renderer process to perform arbitrary memory read and write operations via a crafted HTML page. The vulnerability requires renderer process compromise as a precondition, user interaction, and high attack complexity, making it a post-exploitation vector rather than a primary attack surface. Patch is available from vendor.
Integer overflow in ANGLE graphics library in Google Chrome prior to version 148.0.7778.96 allows remote attackers to leak cross-origin data by delivering a crafted HTML page that triggers the vulnerability. The vulnerability requires user interaction (visiting a malicious webpage) but can bypass same-origin policy protections, exposing sensitive data from other domains. No public exploit code or active exploitation has been confirmed at analysis time.
Insufficient input validation in Google Chrome on Android prior to version 148.0.7778.96 allows local attackers to inject arbitrary scripts or HTML via a crafted Chrome Extension, leading to Universal Cross-Site Scripting (UXSS). The vulnerability requires user interaction and local system access but poses a medium risk due to its ability to bypass same-origin policy protections within the browser context. No public exploit has been identified at the time of analysis.
Google Chrome versions prior to 148.0.7778.96 contain an XSS vulnerability in the SanitizerAPI that allows remote attackers to inject arbitrary scripts or HTML through crafted HTML pages. The vulnerability requires user interaction (visiting a malicious page) but carries medium severity due to its ability to compromise confidentiality and integrity. No public exploit code or active exploitation in CISA KEV has been identified at the time of analysis.
Out-of-bounds memory read in Google Chrome's V8 JavaScript engine allows remote attackers to disclose sensitive information via a crafted HTML page. Affects Chrome versions prior to 148.0.7778.96. The vulnerability requires user interaction (opening a malicious page) but operates over the network with no authentication required. While classified as medium severity by Chromium, the impact is limited to information disclosure without code execution capability.
UI spoofing vulnerability in Google Chrome prior to 148.0.7778.96 allows remote attackers to deceive users through crafted HTML pages that manipulate the speech interface, potentially disclosing sensitive information or causing denial of service. The vulnerability requires user interaction (clicking or interacting with the malicious page) and affects the speech component's visual presentation. Chromium severity is rated Medium; no active exploitation has been publicly confirmed.
Insufficient input validation in Google Chrome's Popup Blocker prior to version 148.0.7778.96 enables a remote attacker with a compromised renderer process to bypass navigation restrictions and access restricted content via a crafted HTML page. The vulnerability requires renderer process compromise and user interaction, limiting real-world exploitability despite its authentication bypass classification. No public exploit code or active exploitation has been identified at the time of analysis.
Out-of-bounds read in WebCodecs video processing in Google Chrome versions prior to 148.0.7778.96 allows remote attackers to leak sensitive memory contents via a crafted video file. The vulnerability requires user interaction to open a malicious video, but affects all users regardless of authentication. Chrome 148.0.7778.96 and later versions patch this information disclosure vulnerability, which could expose cryptographic keys, session tokens, or other sensitive data resident in adjacent memory.
Insufficient policy enforcement in Chrome's Downloads feature prior to version 148.0.7778.96 allows local attackers with user interaction to bypass navigation restrictions and access sensitive download locations via a crafted HTML page, potentially leading to information disclosure or file manipulation. The vulnerability requires local access and user engagement with a malicious page, limiting its scope to targeted social engineering rather than remote mass exploitation.
UI spoofing in Google Chrome on iOS prior to version 148.0.7778.96 allows remote attackers to deceive users through crafted HTML pages that manipulate the browser interface. The vulnerability stems from insufficient validation of untrusted input and has a CVSS score of 5.4 (medium severity). Exploitation requires user interaction with a malicious webpage but can result in information disclosure and denial of service on affected iOS devices.
Uninitialized memory use in Dawn (GPU abstraction layer) in Google Chrome prior to version 148.0.7778.96 allows remote attackers to read potentially sensitive information from process memory by opening a crafted HTML page. The vulnerability requires user interaction (clicking/viewing the malicious page) but no authentication, and has a high confidentiality impact. Chromium security team classified this as high severity; no public exploit code or active exploitation has been confirmed at time of analysis.
Insufficient data validation in DevTools in Google Chrome on Android prior to version 148.0.7778.96 allows remote attackers to bypass navigation restrictions by sending a crafted HTML page, requiring user interaction to open the malicious page. The vulnerability has a low CVSS score (4.3) due to limited confidentiality impact and requirement for user click, but affects all Android users running vulnerable versions. A vendor-released patch is available.
Integer overflow in the GPU component of Google Chrome on Android prior to version 148.0.7778.96 enables a remote attacker whose renderer process has been compromised to execute arbitrary read and write operations via a malicious HTML page. The vulnerability requires prior compromise of the renderer process and user interaction, limiting its standalone exploitability but creating a significant secondary threat following renderer exploitation. Chromium security team rated this as high severity; no public exploit code or active exploitation has been identified at time of analysis.
Out-of-bounds memory read in Google Chrome's font handling allows remote attackers to leak sensitive information via a crafted HTML page when users visit a malicious website. Chrome versions prior to 148.0.7778.96 are affected. The vulnerability requires user interaction (clicking a link or visiting a page) but occurs over the network without authentication. Information disclosure risk is limited (CVSS C:L) with no impact on integrity or availability, but the Chromium security severity designation of High indicates concern about potential information leakage in real-world exploitation.
Velociraptor versions prior to 0.76.4 allow authenticated users with reader role in the root organization to bypass cross-organization authorization controls via HTTP API GET requests, enabling them to read arbitrary files from other organizations regardless of their permissions in those target organizations. The vulnerability requires high privilege context (authenticated reader in root org) but has high confidentiality impact across organization boundaries. This is a horizontal privilege escalation affecting multi-tenant deployments where organizational isolation is a critical security boundary.
Denial of service in Linux kernel KVM x86 nested virtualization allows local privileged attackers to crash virtual machines by manipulating vCPU state through userspace MP_STATE and injected event stuffing. When a vCPU is awakened from a blocking state in L2 (nested guest mode) with an already-injected event, the kernel generates spurious KVM_EXIT_UNKNOWN exits that typically terminate the VM. The vulnerability stems from insufficient validation of impossible vCPU states that userspace can artificially create, despite architectural safeguards that should prevent injected events during blocking. CVSS 5.5 (local, low complexity, high availability impact); EPSS 0.02% indicates minimal widespread exploitation likelihood.
Denial of service in Linux kernel media cx88 driver allows local authenticated attackers to exhaust system resources by triggering a missing DMA unmapping in the snd_cx88_hw_params() error path. The vulnerability causes resource leaks when audio hardware parameter initialization fails, potentially rendering the audio subsystem unavailable. CVSS 5.5 reflects local attack vector with low complexity; EPSS 0.02% indicates minimal real-world exploitation probability despite vendor-released patches across multiple kernel versions.
A use-after-free condition in the libertas USB wireless driver allows local attackers with user privileges to cause a denial of service by triggering a kernel warning and potential crash through rapid firmware loading or repeated usb_tx_block() calls while a USB request is still active. The vulnerability stems from insufficient synchronization of USB URB submissions, enabling concurrent requests on the same URB without enforcing completion of prior transmissions.
Denial of service via kernel warning in MPTCP path manager occurs when combining endpoint removal with fullmesh and flag-setting operations through netlink in the Linux kernel. A local attacker with low privileges can trigger a WARNING in net/mptcp/pm_kernel.c:1074 by sending a crafted sequence of netlink commands, causing the system to emit a kernel warning and potentially become unstable. No known public exploit code exists, but the low CVSS (5.5) and minimal EPSS (0.03%) indicate this is a local DoS with limited real-world impact.
Kernel panic via SError interrupt in Linux kernel media driver (chips-media wave5) when device enters autosuspend during video decoding operations. Local authenticated attackers can trigger a denial of service by queuing buffers to the wave5 video decoder while autosuspend timeout occurs, causing the CPU to access suspended hardware and generate an unrecoverable asynchronous SError, crashing the system. No privilege escalation or code execution; impact limited to availability.
Memory leak in the tw9906 media driver's probe function allows local authenticated attackers to cause denial of service through memory exhaustion. The vulnerability occurs in tw9906_probe() when an error path fails to free memory allocated by v4l2_ctrl_handler_init() and v4l2_ctrl_new_std(), potentially leading to kernel memory depletion on repeated device probe attempts. Vendor-released patches are available across multiple stable kernel branches.
Denial of service in Linux kernel KCM (kernel connection multiplexer) subsystem when processing messages with zero-fragment skbs in frag_list after partial sendmsg failures. A local authenticated attacker can trigger a kernel warning and potentially crash the system by sending a malformed message that fails during data copy, leaving an empty skb in the fragment list. The vulnerability requires local access and low-level socket manipulation, affecting systems running vulnerable Linux kernel versions prior to patching.
A denial of service vulnerability in the Linux kernel's AMD display driver (drm/amd/display) allows local authenticated users to crash the system by accessing link encoder functionality on DisplayPort over USB-C (DPIA) links without proper signal type validation. The vulnerability affects kernel versions before the patches released in stable branches 6.12.75, 6.18.16, 6.19.6, and 7.0. No public exploit code has been identified, and real-world exploitation probability is very low (EPSS 0.02%), suggesting this is primarily an edge-case denial of service affecting specific hardware configurations with DPIA displays.
Memory leak in the Linux kernel's TI K3 SoC info driver (soc/ti/k3-socinfo) fails to release an allocated mmio regmap, causing denial of service through resource exhaustion on probe failures and driver unbind. Local attackers with low privileges can trigger probe deferral or driver unbind to exhaust kernel memory, affecting systems running vulnerable Linux kernel versions. The vulnerability has a low EPSS score (0.02%) but enables practical local DoS against systems where low-privilege users can control driver lifecycle.
Memory leak in Linux kernel tw9903 media driver probe function allows local authenticated attackers to cause denial of service through repeated device initialization failures. The v4l2_ctrl_handler_init() and v4l2_ctrl_new_std() allocations are not freed in one error path of tw9903_probe(), enabling exhaustion of kernel memory. CVSS 5.5 (local, low complexity, requires low privileges). EPSS score of 0.02% indicates minimal exploitation probability.
Local denial of service in Linux kernel ring-buffer tracing code allows authenticated local users to crash the system by triggering invalid memory access through malformed event length fields. The vulnerability exists in rb_read_data_buffer(), which validates possibly corrupted ring buffers at boot but fails to verify event lengths are within acceptable ranges before calculating buffer offsets. EPSS exploitation probability is very low at 0.02%, and no public exploit code or active exploitation has been identified.
Denial of service in Linux kernel minix filesystem implementation allows local authenticated users to crash the system via crafted minix superblock structures due to missing validation of s_log_zone_size and other superblock fields in the minix_check_superblock() function.
Memory leak in Linux kernel fbdev vt8500lcdfb driver allows local authenticated users to cause denial of service via unfreed DMA-allocated buffer on error path. The vulnerability exists in the framebuffer initialization code where dma_alloc_coherent() allocation for fbi->fb.screen_buffer is not properly freed if an error occurs during probe, leading to memory exhaustion on repeated device initialization attempts. Local privilege required; no remote or unauthenticated attack vector.
Denial of service in Linux kernel APEI/GHES ARM processor error handling allows local authenticated attackers to trigger kernel oops by crafting malformed ARM Processor Error records with incomplete or oversized section data, causing out-of-bounds memory dereference. CVSS 5.5 (local, low complexity, authenticated, availability impact only) with EPSS 0.02% indicates low real-world exploitation probability despite public patch availability.
Denial of service in Linux kernel PCI endpoint configfs interface allows local attackers with low privileges to crash the kernel via swapped parameters in pci_primary_epc_epf_unlink() and pci_secondary_epc_epf_unlink() functions. When executing the unlink command in configfs, incorrect parameter ordering causes invalid memory access and kernel panic. CVSS 5.5 (local, low complexity, low privilege) with EPSS 0.02% suggests limited real-world exploitation despite confirmed availability of patches across multiple kernel branches.
Integer underflow in the Linux kernel's EFI/CPER firmware error logging function (cper_print_fw_err) allows local authenticated attackers to trigger denial of service via memory dump of unmapped regions, disclose kernel memory contents, or cause system crash when processing malformed EFI firmware error records with invalid offsets. The vulnerability stems from insufficient validation of error record length before subtracting an offset, causing integer wraparound that permits dumping of arbitrary kernel memory regions.
Kernel panic occurs in Linux kernel USB gadget driver (dwc3) when vbus_draw function executes PMIC power-supply APIs from atomic context, causing sleep operations in non-sleepable context. The vulnerability affects Linux kernel versions prior to 5.13 and various stable series (6.6.128, 6.12.75, 6.18.16, 6.19.6, 7.0) and is resolved by deferring vbus_draw operations to workqueue context. Local authenticated users with privilege level PR:L can trigger denial of service by invoking affected USB gadget functionality, with EPSS exploitation probability at 0.02% percentile indicating extremely low practical risk despite moderate CVSS severity.
Denial of service via kernel panic in the DRM buddy allocator when rounded allocation sizes exceed available memory. Local authenticated attackers can trigger a BUG_ON() crash by requesting contiguous or large-aligned GPU memory allocations that, after power-of-two or block-size rounding, exceed the total VRAM size. Example: a 9GB contiguous allocation request on 10GB VRAM rounds to 16GB, causing immediate kernel panic. No authentication bypass or privilege escalation; requires local access and appropriate user permissions to invoke DRM allocator.
Denial of service via incomplete xattr cleanup in the Linux kernel OCFS2 filesystem causes memory corruption when processing reflinked files with extended attributes. A local user with standard privileges can trigger this vulnerability through crafted xattr operations, resulting in system crashes or data integrity issues. The flaw affects Linux kernel versions from 2.6.32 through multiple recent releases (5.10, 5.15, 6.1, 6.6, 6.12, 6.18, 6.19), with vendor-released patches available for supported stable branches.
Resource leak in Linux kernel xfrm (IPsec) subsystem allows local authenticated users to exhaust kernel memory and trigger denial of service by preventing proper cleanup of IPsec hardware offload device references during network device unregistration. The vulnerability affects XFRM state management when IPsec hardware offloading is configured, particularly when the hardware offload capability (NETIF_F_HW_ESP) is disabled after state creation but before device removal.
Denial of service in Linux kernel dm-verity subsystem allows local authenticated users to crash the system by triggering dm_bufio_client_create() failure in verity_fec_ctr(), which incorrectly passes an error pointer to dm_bufio_client_destroy(). The vulnerability requires local access and authenticated privileges but no user interaction, affecting dm-verity configurations with forward error correction enabled.
Null pointer dereference in the Linux kernel AMD power management (drm/amd/pm) subsystem causes denial of service when SMU (System Management Unit) is disabled during RAS (Reliability, Availability, and Serviceability) initialization. Local authenticated attackers with low privileges can trigger this crash on affected systems, resulting in kernel panic and system unavailability. EPSS exploitation probability is very low (0.02%), indicating this requires specific configuration and local access.
Hard-lockup in Linux kernel IOMMU/VT-d subsystem when flushing device IOTLB during PCIe link-down events affects systems using scalable mode with PASID support. Local authenticated attackers can trigger denial of service by causing PCIe device disconnection followed by resource cleanup operations, such as releasing VFIO group file descriptors, which deadlock the system while attempting ATS invalidation on inaccessible devices. Patch available across multiple stable kernel series.
A circular locking dependency in the Linux kernel's ntfs3 filesystem driver allows local authenticated attackers with low privileges to cause a denial of service (system hang or crash) by triggering simultaneous operations that deadlock between the MFT run lock and bitmap read-write lock. The vulnerability affects kernel versions prior to 6.18.16, 6.19.6, and 7.0, and requires local access with user-level privileges to exploit.
Denial of service via null pointer dereference in Linux kernel's pstore persistent storage subsystem occurs when the vmap() function fails but the persistent_ram_vmap() function incorrectly returns success if a non-zero offset is present, allowing subsequent buffer access to dereference invalid memory and cause system crashes. Affects Linux kernel versions prior to 5.10.252, 5.15.202, 6.1.165, 6.6.128, 6.12.75, 6.18.16, 6.19.6, and 7.0. No public exploit identified at time of analysis, but vendor-released patches are available across multiple stable branches.
Data corruption in Linux kernel btrfs filesystem log replay allows local authenticated attackers to cause files to retain incorrect sizes after crash recovery. When a file is truncated to zero bytes, fssynced, and then a hardlink is created, the file incorrectly retains its pre-truncation size after a power failure and log replay, resulting in data integrity violation with availability impact.
A denial of service vulnerability in the Linux kernel's Qualcomm PD mapper service registry causes system crashes due to mismatched string element length validation in servreg_loc_pfr_req_ei. When a process daemon crashes and triggers a service registry location request, the QMI decoder rejects the reason field because its declared maximum length (65 bytes) is smaller than the actual field size (81 bytes), causing a decoding error and system halt. This affects all Linux kernel versions prior to the patch, triggered by local processes with standard user privileges.
Denial of service in Linux kernel xfrm (IPsec transform) subsystem allows local authenticated attackers to trigger a kernel panic via improper netlink message size calculation when handling XFRM_MSG_GETAE requests for states with interface ID set. The xfrm_aevent_msgsize() function fails to account for XFRMA_IF_ID attribute space, causing build_aevent() to exceed buffer bounds and hit a BUG_ON assertion, resulting in kernel crash. EPSS exploitation probability is very low at 0.02% despite the local attack vector, suggesting limited real-world impact.
Memory leak in the Airoha QDMA RX packet processing function allows local authenticated attackers to cause a denial of service through resource exhaustion. The vulnerability occurs when page pool fragments fail to properly return to the pool during error handling in airoha_qdma_rx_process(), allowing an attacker with local access and low privileges to exhaust kernel memory and crash the system. EPSS exploitation probability is extremely low at 0.02%, reflecting the local-only attack vector and privilege requirement.
Null pointer dereference in Linux kernel bridge VLAN filtering code allows local authenticated attackers to trigger a denial of service via a crafted RTM_NEWLINK netlink message with BR_BOOLOPT_FDB_LOCAL_VLAN_0 flag when CONFIG_BRIDGE_VLAN_FILTERING is disabled. The vulnerability occurs because br_fdb_delete_locals_per_vlan_port() and br_fdb_insert_locals_per_vlan_port() dereference a NULL vlan group pointer without validation, causing a kernel panic. No public exploit code identified at time of analysis.
Infinite vCPU fault loop in the Linux kernel's mshv (Microsoft Hypervisor) subsystem allows a local guest VM process to permanently spin a host vCPU thread, exhausting host CPU resources. The flaw exists in mshv_handle_gpa_intercept(), which unconditionally attempts page remaps on all movable-memory faults regardless of access permission - when a guest writes to a read-only Guest Physical Address region, the remap succeeds but the region retains its read-only designation, causing an immediate re-fault in a tight loop. Affected kernel versions run from commit b9a66cd5ccbb9fade15d0e427e19470d8ad35b75 through the fix commits; patched releases 6.19.14 and 7.0 are available. No public exploit has been identified and EPSS is 0.01%, consistent with the local, hypervisor-specific attack surface.
Linux kernel ASoC SDCA subsystem crashes on sound card teardown due to IRQ lifecycle mismanagement, causing a local denial of service. IRQ handlers registered via devm_request_threaded_irq() during component probe retain stale references to freed card and kcontrol structures after the sound card is torn down, resulting in null or dangling pointer dereferences and kernel panic. Exploitation requires local low-privilege access and SDCA-capable audio hardware; no public exploit exists and EPSS is extremely low at 0.02% (5th percentile).
The AF_XDP socket subsystem (xsk) in the Linux kernel fails to validate that a network device's MTU fits within the usable UMEM frame space at bind time, allowing a local low-privileged user to trigger a kernel denial of service. Usable frame space - chunk size minus headroom and tailroom - can fall below a standard 1500-byte MTU when 2k chunks are used, a gap that became exploitable once tailroom subtraction was introduced. The kernel also omits validation of hardware zero-copy capabilities via net_device::xdp_zc_max_segs. No public exploit has been identified and EPSS is 0.02% (5th percentile), indicating low immediate exploitation risk.
Reference count leak in the Linux kernel's xfrm IPsec subsystem allows a local low-privileged attacker to exhaust kernel memory, resulting in denial of service. The defect resides in xfrm_migrate_policy_find(), where xfrm_pol_hold_rcu() is called twice - once implicitly by the lookup path (which already returns a held reference) and once redundantly - creating a refcount imbalance that prevents memory reclamation. Discovered by the Linux Verification Center using Syzkaller fuzzing; no public exploit identified at time of analysis, and EPSS is very low at 0.02% (5th percentile), indicating minimal observed exploitation activity.
Kernel crash in the Linux pinctrl mcp23s08 driver triggers a NULL pointer dereference during device probe when an MCP23S08/MCP23008 GPIO expander retains non-zero interrupt-on-change state across a warm reboot. The crash occurs because the driver's IRQ handler fires against pins whose nested IRQ handlers have not yet been registered, causing an unhandled kernel oops and system denial of service. Exploitation is local and device-specific; no public exploit identified at time of analysis, and EPSS remains at 0.02% (5th percentile) consistent with no observed in-the-wild triggering.
NULL pointer dereference in the Linux kernel IPVS subsystem causes a kernel panic during IPVS service creation failure cleanup, resulting in a full host denial of service. Affected kernels from 6.2 onward contain a logic error in ip_vs_add_service() where a successful scheduler bind sets the local sched variable to NULL; if ip_vs_start_estimator() subsequently fails, the error path calls ip_vs_unbind_scheduler() with that NULL pointer, dereferencing offset 0x30 and triggering a general protection fault. Exploitation requires local authenticated access with CAP_NET_ADMIN privileges; no public exploit is identified and EPSS is 0.02%, indicating very low exploitation probability.
Denial-of-service via off-by-one allocation in the Linux kernel txgbe network driver allows a local low-privileged user to crash the kernel on systems hosting Wangxun 10GbE NICs. The driver allocates property_entry struct lists without reserving the mandatory null-terminator sentinel slot, meaning kernel subsystems iterating over the list read beyond allocated memory bounds. No active exploitation has been identified and EPSS is extremely low (0.02%, 5th percentile), but patches are available across multiple stable kernel branches including 6.6.136, 6.12.83, 6.18.24, 6.19.14, and 7.0.
Incorrect GENERIC_CMD register field masks in the Linux kernel's IPA (IP Accelerator) network driver for IPA v5.0+ hardware trigger a kernel WARN when a 'stop' command is sent to the MPSS (Modem Processor SubSystem) remoteproc while IPA is active. This availability-only vulnerability (CVSS C:N/I:N/A:H) affects authenticated local users on systems with Qualcomm IPA v5.0+ silicon. No public exploit exists and no KEV listing is present; with an EPSS of 0.02% this is a low-urgency stability fix rather than an active threat.
Out-of-bounds array write in the Linux kernel's Intel uncore PMU driver (`perf/x86/intel/uncore`) causes denial of service on affected Intel multi-die x86 systems. The driver fails to skip discovery table parsing for dies whose CPUs are all offline at boot, allowing the assignment `pmu->boxes[die] = box` in `uncore_pci_pmu_register()` to overflow the boxes array, triggering a kernel WARN or potential memory corruption. Exploitation requires local low-privilege access under a specific hardware and boot configuration; no public exploit exists and EPSS is 0.02%, indicating negligible real-world exploitation likelihood.
Local denial-of-service in the Linux kernel's AEAD crypto socket interface (`algif_aead`) allows a low-privileged local user to crash the kernel by submitting a decryption request where the minimum receive buffer size check fails to account for the authentication tag length. The CVSS vector (AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H) confirms this is a locally exploitable, high-availability-impact issue with no confidentiality or integrity risk. Patches have been released across multiple Linux LTS stable branches (5.10.254, 5.15.204, 6.1.170, 6.6.136, 6.12.83, 6.18.24, 6.19.14, 7.0) and Ubuntu has issued multiple USN advisories (USN-8277-1 through USN-8281-1). No public exploit code has been identified and EPSS is 0.02%, indicating no public exploitation activity.
Oracle OCI CLI version 3.77 allows local attackers with user interaction to place imported files outside the intended directory, compromising file integrity and enabling potential code execution or data exfiltration. The vulnerability requires local access and user interaction but carries high integrity impact through arbitrary file placement. No active exploitation or public exploit code has been identified at the time of analysis.
Kubewarden versions before 1.35.0 permit RBAC reconnaissance attacks when users with AdmissionPolicy or AdmissionPolicyGroup creation privileges craft policies using the unchecked `can_i` host capability. The vulnerability allows enumeration of any user or service account permissions across the cluster via SubjectAccessReview requests executed with policy-server privileges, despite the absence of context-aware resource grants. This information disclosure enables attackers to discover sensitive permission configurations without requiring cluster-wide policy creation rights, a capability not available by default but exploitable when granted.
PocketBase versions before 0.22.42 and 0.30.0-0.37.3 allow account pre-hijacking via OAuth2 autolinking, where an attacker knowing a victim's email can create an unverified account linked to one OAuth2 provider, then retain access when the victim authenticates with a different provider and the accounts are auto-merged, because previous OAuth2 links are not cleared during the upgrade from unverified to verified status. Publicly available exploit code exists; vendor recommends immediate upgrade to v0.37.4 or v0.22.42.
OpenEXR versions 3.0.0-3.2.8, 3.3.0-3.3.10, and 3.4.0-3.4.10 suffer from unbounded shift operations in the readVariableLengthInteger() function when parsing variable-length integers from untrusted EXR files. Attackers can craft malicious EXR files with excessive continuation bytes to trigger left shifts exceeding 64 bits on a 64-bit integer, causing undefined behavior that may lead to information disclosure or denial of service. The vulnerability is remotely exploitable without authentication or user interaction against any application processing untrusted EXR input; no public exploit code has been identified at the time of analysis.
Policy rollback vulnerability in gittuf versions up to 0.13.1 allows attackers with push access to the Reference State Log (RSL) to downgrade repository policies to previously signed versions, bypassing security controls. An attacker cannot roll back to policies that would be unsigned by the current root keys, but can selectively choose any valid prior policy state. Vendor-released patch: gittuf v0.14.0 introduces monotonically increasing version numbers to all policy metadata to prevent rollback attacks.
ShellHub Community v0.24.1 and earlier allows authenticated API Key holders to enumerate any tenant namespace and retrieve sensitive membership data (user IDs, emails, roles), settings, and device counts via GET /api/namespaces/:tenant due to a bypassed authorization check. The vulnerability exploits API Key authentication flows that fail to set the user ID header, causing the membership verification to be skipped entirely. Publicly available proof-of-concept code demonstrates validated exploitation against v0.24.1, with complete disclosure of cross-tenant namespace configuration including member lists suitable for targeted phishing campaigns.
Kubetail Dashboard prior to version 0.14.0 fails to validate the Origin header on WebSocket connection upgrades, enabling Cross-Site WebSocket Hijacking (CSWSH) attacks. An authenticated user visiting a malicious web page can be exploited to stream their Kubernetes container logs-including credentials, tokens, and PII often present in logs-to an attacker-controlled server. The vulnerability affects both desktop deployments at localhost:7500 and cluster deployments behind HTTP basic auth, with browser ambient credentials automatically attached to the WebSocket handshake.
CSS Parser gem disables HTTPS certificate validation by setting OpenSSL::SSL::VERIFY_NONE, allowing man-in-the-middle attackers to inject or modify CSS content loaded via HTTPS. Any application using CSS Parser versions prior to 2.1.0 to fetch external stylesheets over HTTPS can be exploited by network-positioned attackers without authentication. A proof-of-concept using mitmproxy or Burp Suite demonstrates practical exploitation; CVSS 5.8 reflects the network attack vector and integrity impact, but real-world risk depends on whether the application loads stylesheets from untrusted or attacker-controllable URLs and whether the attacker can intercept network traffic.
Gotenberg versions 8.31.0 and earlier allow unauthenticated remote attackers to enumerate and read arbitrary files under /tmp/ via the /forms/chromium/convert/url and /forms/chromium/screenshot/url endpoints using file:// scheme URLs. An attacker can discover in-flight conversion request directories and exfiltrate source files (HTML, Markdown, Office documents, staged PDFs) from other users' concurrent conversion requests by timing attacks to coincide with long-running conversion operations. The vulnerability exploits a logic flaw where the URL routes fail to set per-request scope guards that HTML/Markdown routes correctly apply, causing file:// access control enforcement to silently skip for URL-based conversions.
Arbitrary PDF file read vulnerability in Gotenberg versions up to 8.31.0 allows unauthenticated remote attackers to extract PDF content via path traversal in stampExpression and watermarkExpression parameters on six conversion routes (pdfengines/merge, pdfengines/split, libreoffice/convert, chromium/convert/url, chromium/convert/html, chromium/convert/markdown). The vulnerability exists because these routes accept user-controlled file paths without validation when stamp or watermark source is set to PDF, unlike the dedicated stamp/watermark routes which enforce file upload requirements. An attacker can read any PDF accessible to the Gotenberg process by specifying its filesystem path, gaining access to potentially sensitive documents in containerized deployments or systems with mounted directories.
DNS rebinding vulnerability in Gotenberg allows unauthenticated remote attackers to bypass SSRF protections and access internal services via Chromium URL conversion routes. When a URL is submitted for PDF conversion, Gotenberg validates the resolved IP address against a deny-list but discards the pinned result. Chromium then performs independent DNS resolution multiple times, creating a race condition where an attacker controlling DNS can return a public IP during validation and a private IP during connection, allowing access to loopback services, cloud metadata endpoints, or internal networks. Exploitation succeeds approximately 10% per attempt with trivial automation.
CRLF injection in Netty's RedisEncoder allows remote command injection and response poisoning by injecting carriage return and line feed characters into InlineCommandRedisMessage, SimpleStringRedisMessage, and ErrorRedisMessage objects. Attackers can inject arbitrary Redis commands (such as CONFIG SET, FLUSHALL, or authentication bypass) or forge fake responses when user-controlled input is placed into these message types without sanitization. The vulnerability affects Netty 4.2.12.Final and all prior versions with the codec-redis module; no active exploitation has been reported in CISA KEV, but publicly available proof-of-concept code demonstrates the vulnerability.
HTTP request smuggling in Netty's HttpRequestDecoder allows remote unauthenticated attackers to inject arbitrary HTTP requests by sending malformed Transfer-Encoding headers (specifically 'Transfer-Encoding: chunked, identity'). When Netty is deployed behind a proxy that forwards such requests without rejection, an attacker can smuggle a second request inside the body of the first, bypassing security controls and accessing unintended resources. The vulnerability is confirmed by public proof-of-concept code demonstrating successful parsing of injected requests.
HTTP request smuggling in Netty's chunk size parser allows remote unauthenticated attackers to inject arbitrary HTTP requests by exploiting integer overflow in the hexadecimal chunk size parsing logic. The HttpObjectDecoder.getChunkSize method accumulates the chunk size without proper overflow validation, enabling an attacker to craft a malicious chunk size header that wraps around to a valid size, causing Netty to misinterpret the request boundary and parse injected requests as separate legitimate requests. Publicly available proof-of-concept demonstrates successful parsing of an injected GET request within a chunked POST body, with CVSS score 6.5 (network-accessible, low complexity, no authentication required).
Cross-site scripting (XSS) via improper Markdown attribute sanitization in Weblate allows authenticated users to inject malicious HTML attributes through crafted Markdown in user comments and other user-provided content. The vulnerability affects Weblate versions prior to 5.17.1 and can be exploited by any authenticated user with the ability to submit Markdown content; however, the application's strict Content Security Policy (CSP) significantly mitigates actual exploitation risk. Vendor-released patch version 5.17.1 is available.
Weblate versions before 5.17.1 allow authenticated users to enumerate translations in projects they cannot access via the screenshots, tasks, and component link API endpoints. An attacker with valid credentials but no project access can probe these APIs to discover the existence and metadata of private translations, leading to information disclosure of project structure and language coverage that should remain hidden. The vulnerability requires authentication but has a low attack complexity, affecting confidentiality only without enabling further compromise.
BSON operator injection in ShellHub Community v0.24.1 and earlier allows any authenticated user to crash the device list API endpoint by supplying MongoDB operators as filter field names or sort_by values without input validation. Two attack vectors exist in api/store/mongo/query-options.go: unvalidated filter param.Name values become raw BSON map keys, and the sort_by query parameter is passed directly into a $sort pipeline stage. A validated proof-of-concept exists against v0.24.1. No active exploitation is confirmed (not in CISA KEV), and EPSS sits at 0.06% (19th percentile), indicating low real-world exploitation probability despite a public PoC.
Cross-tenant IDOR in ShellHub Community v0.24.1 allows any authenticated user to read SSH session records belonging to other namespaces by supplying an arbitrary session UID to `GET /api/sessions/:uid`. The `GetSession` function in `api/services/session.go` omits the `InNamespace` tenant filter present elsewhere, so the `Authorize` middleware validates token presence but not resource ownership. Disclosed data includes SSH username, device UID, remote IP, geolocation, terminal type, authenticated flag, and session timestamps - enabling cross-tenant reconnaissance. No public exploitation is confirmed and EPSS is 0.03% (8th percentile), though the advisory includes a working proof-of-concept curl script.
Cross-tenant IDOR in ShellHub Community's device API exposes full device metadata across namespace boundaries to any authenticated user. An authenticated attacker who knows or can obtain a target device UID can issue a crafted GET request to /api/devices/:uid and receive the victim tenant's device object - including hostname, MAC address, OS fingerprint, public SSH key, namespace name, last-seen timestamp, and remote IP address. A validated proof-of-concept exists against v0.24.1; no public exploit identified at time of analysis and no CISA KEV listing, though EPSS is 0.03% (8th percentile), reflecting low but real probability of opportunistic exploitation in multi-tenant deployments.
Stored XSS in Kyverno policy-reporter-ui 2.5.1 and earlier allows an attacker with Kubernetes write access to PolicyReport resources to inject arbitrary HTML into the Vue 3 frontend, which executes in the browser of any authenticated UI user who views the affected namespace. The sink is the v-html directive in PropertyCard.vue, which is reached whenever a property value is not an http/https URL and exceeds 75 characters - a bypass condition trivially satisfied by padding any HTML payload. A public proof-of-concept demonstrating session cookie exfiltration via a crafted PolicyReport object is included in the GitHub advisory GHSA-q98m-7w8c-w388; no public exploit identified as confirmed actively exploited (not in CISA KEV), though EPSS of 0.03% reflects low opportunistic risk given the Kubernetes-specific prerequisite.
Cross-tenant information disclosure in Hatchet's `GET /api/v1/stable/dags/tasks` endpoint permits any authenticated user on a multi-tenant instance to read task metadata belonging to other tenants by supplying a victim's tenant UUID and DAG UUID in query parameters. The CVSS 5.3 (AV:N/AC:H/PR:L) reflects that an attacker must hold a valid account and obtain victim UUIDs through out-of-band channels before exploitation - but on Hatchet Cloud, where account registration is open by default, the account prerequisite is trivially satisfied. No public exploit code exists, no exploitation was observed prior to patching, and vendor-released fix v0.83.39 was deployed within hours of the April 23, 2026 report.
Denial of service in vLLM v0.18.0-v0.19.1 permanently crashes the EngineCore process when any API request includes a sampling penalty parameter against a deployment configured with `extract_hidden_states` speculative decoding. A refactor in PR #37013 (v0.18.0) removed the `.unsqueeze(-1)` call that normalized tensor shape, causing a broadcast shape mismatch (`(batch_size, 2)` vs. expected `(batch_size, 1)`) during penalty application and an unrecoverable RuntimeError. The crash is deterministic - a single request with `repetition_penalty: 1.1` suffices - but no public exploit code has been identified at time of analysis and EPSS is very low (0.04%, 12th percentile).
Hugo static site generator versions 0.43.0 through 0.160.x allow unrestricted file system access when building sites with Node-based asset pipelines (PostCSS, Babel, TailwindCSS), enabling arbitrary code execution through these tools to read or write files outside the project directory. The vulnerability affects only users who deploy Node asset pipelines; those building trusted sites or not using these pipelines are unaffected. A vendor-released patch in v0.161.0 enforces Node's permission model with strict filesystem defaults.
Integer overflow in Chrome's Network component prior to version 148.0.7778.96 allows remote attackers to bypass same-origin policy after compromising the renderer process via a crafted HTML page. The vulnerability requires renderer compromise and user interaction, limiting real-world risk despite network-accessible attack surface. No active exploitation has been confirmed; a vendor patch is available.
Insufficient input validation in the FileSystem API in Google Chrome prior to version 148.0.7778.96 allows a remote attacker with a compromised renderer process to perform arbitrary file read and write operations through a malicious HTML page. The vulnerability requires prior renderer compromise and user interaction with a crafted page, limiting its attack surface despite network-accessible vectors. Vendor has released a patched version.
Insufficient policy enforcement in DirectSockets API in Google Chrome prior to version 148.0.7778.96 allows remote attackers to perform arbitrary read and write operations through a crafted malicious Chrome extension. The vulnerability requires user interaction (extension installation) but affects the core security model of the browser's socket access control, exposing local network resources to unauthorized access by untrusted extensions.
Google Chrome prior to version 148.0.7778.96 contains insufficient validation of untrusted input in the Permissions system, allowing attackers on the local network segment to leak cross-origin data through malicious network traffic. The vulnerability requires network adjacency but no user interaction or authentication, affecting confidentiality with medium Chromium severity. Patch is available from Google.
Information disclosure in Google Chrome prior to 148.0.7778.96 allows remote attackers who have compromised the renderer process to extract potentially sensitive data from process memory through a race condition triggered by a crafted HTML page. The vulnerability requires renderer process compromise and user interaction but results in high confidentiality impact with no integrity or availability consequences. Chromium security team rates this as Medium severity; no active exploitation has been publicly confirmed.
Uncontrolled XSS vulnerability in Google Chrome's ServiceWorker implementation prior to version 148.0.7778.96 allows attackers to inject arbitrary scripts or HTML into web pages when users install a malicious Chrome extension, bypassing same-origin policy protections. The vulnerability requires user interaction (extension installation) but can be exploited remotely with low attack complexity. CVSS score of 5.4 reflects medium severity; no active exploitation has been publicly reported.
Uninitialized memory use in the GPU component of Google Chrome prior to version 148.0.7778.96 allows remote attackers who have compromised the renderer process to extract potentially sensitive information from process memory through a malicious HTML page. The vulnerability requires renderer process compromise as a precondition and user interaction to trigger, but once achieved, enables confidentiality breach with no code execution or denial of service impact. Vendor-released patch available in Chrome 148.0.7778.96.
Unvalidated Omnibox input in Google Chrome prior to version 148.0.7778.96 enables remote attackers to inject arbitrary scripts and HTML (universal XSS) via malicious network traffic, affecting users who click on crafted links. The vulnerability requires user interaction but crosses security boundaries due to its scope impact. No active exploitation has been publicly confirmed, though a patch is available from Google.
Insufficient policy enforcement in Chrome Extensions prior to version 148.0.7778.96 allows a remote attacker with a compromised renderer process to bypass discretionary access controls via a crafted HTML page, potentially leading to unauthorized information disclosure or modification. The vulnerability requires user interaction and a pre-existing renderer compromise, limiting its practical exploitation scope. A vendor-released patch is available in Chrome 148.0.7778.96 and later.
Out-of-bounds read and write in the GFX component of Google Chrome prior to version 148.0.7778.96 allows remote attackers to perform arbitrary memory read and write operations through malicious network traffic. The vulnerability requires user interaction (clicking a link or visiting a malicious page) but does not require authentication. Chromium rated the security severity as Medium; CVSS 5.4 reflects moderate impact (confidentiality and integrity compromise without availability loss). No active exploitation confirmed in CISA KEV at time of analysis.
UI spoofing via insufficient input validation in Google Chrome prior to 148.0.7778.96 allows a remote attacker who has already compromised the renderer process to craft malicious HTML pages that deceive users about the legitimate UI elements. The attack requires renderer process compromise as a prerequisite and user interaction with the crafted page, limiting real-world applicability to multi-stage attacks. Chromium assessed this as medium severity; no public exploit code or active exploitation has been identified.
Insufficient policy enforcement in Chrome's WebUI on Linux, Mac, Windows, and ChromeOS prior to 148.0.7778.96 allows a remote attacker with a compromised renderer process to bypass site isolation via a crafted HTML page, potentially exposing sensitive cross-site data. The vulnerability requires user interaction (UI:R) and prior renderer compromise, limiting its standalone exploitability. Vendor-released patch available in version 148.0.7778.96.
Insufficient input validation in ANGLE (Almost Native Graphics Layer Engine) in Google Chrome prior to version 148.0.7778.96 allows remote attackers who have compromised the renderer process to perform arbitrary memory read and write operations via a crafted HTML page. The vulnerability requires renderer process compromise as a precondition, user interaction, and high attack complexity, making it a post-exploitation vector rather than a primary attack surface. Patch is available from vendor.
Integer overflow in ANGLE graphics library in Google Chrome prior to version 148.0.7778.96 allows remote attackers to leak cross-origin data by delivering a crafted HTML page that triggers the vulnerability. The vulnerability requires user interaction (visiting a malicious webpage) but can bypass same-origin policy protections, exposing sensitive data from other domains. No public exploit code or active exploitation has been confirmed at analysis time.
Insufficient input validation in Google Chrome on Android prior to version 148.0.7778.96 allows local attackers to inject arbitrary scripts or HTML via a crafted Chrome Extension, leading to Universal Cross-Site Scripting (UXSS). The vulnerability requires user interaction and local system access but poses a medium risk due to its ability to bypass same-origin policy protections within the browser context. No public exploit has been identified at the time of analysis.
Google Chrome versions prior to 148.0.7778.96 contain an XSS vulnerability in the SanitizerAPI that allows remote attackers to inject arbitrary scripts or HTML through crafted HTML pages. The vulnerability requires user interaction (visiting a malicious page) but carries medium severity due to its ability to compromise confidentiality and integrity. No public exploit code or active exploitation in CISA KEV has been identified at the time of analysis.
Out-of-bounds memory read in Google Chrome's V8 JavaScript engine allows remote attackers to disclose sensitive information via a crafted HTML page. Affects Chrome versions prior to 148.0.7778.96. The vulnerability requires user interaction (opening a malicious page) but operates over the network with no authentication required. While classified as medium severity by Chromium, the impact is limited to information disclosure without code execution capability.
UI spoofing vulnerability in Google Chrome prior to 148.0.7778.96 allows remote attackers to deceive users through crafted HTML pages that manipulate the speech interface, potentially disclosing sensitive information or causing denial of service. The vulnerability requires user interaction (clicking or interacting with the malicious page) and affects the speech component's visual presentation. Chromium severity is rated Medium; no active exploitation has been publicly confirmed.
Insufficient input validation in Google Chrome's Popup Blocker prior to version 148.0.7778.96 enables a remote attacker with a compromised renderer process to bypass navigation restrictions and access restricted content via a crafted HTML page. The vulnerability requires renderer process compromise and user interaction, limiting real-world exploitability despite its authentication bypass classification. No public exploit code or active exploitation has been identified at the time of analysis.
Out-of-bounds read in WebCodecs video processing in Google Chrome versions prior to 148.0.7778.96 allows remote attackers to leak sensitive memory contents via a crafted video file. The vulnerability requires user interaction to open a malicious video, but affects all users regardless of authentication. Chrome 148.0.7778.96 and later versions patch this information disclosure vulnerability, which could expose cryptographic keys, session tokens, or other sensitive data resident in adjacent memory.
Insufficient policy enforcement in Chrome's Downloads feature prior to version 148.0.7778.96 allows local attackers with user interaction to bypass navigation restrictions and access sensitive download locations via a crafted HTML page, potentially leading to information disclosure or file manipulation. The vulnerability requires local access and user engagement with a malicious page, limiting its scope to targeted social engineering rather than remote mass exploitation.
UI spoofing in Google Chrome on iOS prior to version 148.0.7778.96 allows remote attackers to deceive users through crafted HTML pages that manipulate the browser interface. The vulnerability stems from insufficient validation of untrusted input and has a CVSS score of 5.4 (medium severity). Exploitation requires user interaction with a malicious webpage but can result in information disclosure and denial of service on affected iOS devices.
Uninitialized memory use in Dawn (GPU abstraction layer) in Google Chrome prior to version 148.0.7778.96 allows remote attackers to read potentially sensitive information from process memory by opening a crafted HTML page. The vulnerability requires user interaction (clicking/viewing the malicious page) but no authentication, and has a high confidentiality impact. Chromium security team classified this as high severity; no public exploit code or active exploitation has been confirmed at time of analysis.
Insufficient data validation in DevTools in Google Chrome on Android prior to version 148.0.7778.96 allows remote attackers to bypass navigation restrictions by sending a crafted HTML page, requiring user interaction to open the malicious page. The vulnerability has a low CVSS score (4.3) due to limited confidentiality impact and requirement for user click, but affects all Android users running vulnerable versions. A vendor-released patch is available.
Integer overflow in the GPU component of Google Chrome on Android prior to version 148.0.7778.96 enables a remote attacker whose renderer process has been compromised to execute arbitrary read and write operations via a malicious HTML page. The vulnerability requires prior compromise of the renderer process and user interaction, limiting its standalone exploitability but creating a significant secondary threat following renderer exploitation. Chromium security team rated this as high severity; no public exploit code or active exploitation has been identified at time of analysis.
Out-of-bounds memory read in Google Chrome's font handling allows remote attackers to leak sensitive information via a crafted HTML page when users visit a malicious website. Chrome versions prior to 148.0.7778.96 are affected. The vulnerability requires user interaction (clicking a link or visiting a page) but occurs over the network without authentication. Information disclosure risk is limited (CVSS C:L) with no impact on integrity or availability, but the Chromium security severity designation of High indicates concern about potential information leakage in real-world exploitation.
Velociraptor versions prior to 0.76.4 allow authenticated users with reader role in the root organization to bypass cross-organization authorization controls via HTTP API GET requests, enabling them to read arbitrary files from other organizations regardless of their permissions in those target organizations. The vulnerability requires high privilege context (authenticated reader in root org) but has high confidentiality impact across organization boundaries. This is a horizontal privilege escalation affecting multi-tenant deployments where organizational isolation is a critical security boundary.
Denial of service in Linux kernel KVM x86 nested virtualization allows local privileged attackers to crash virtual machines by manipulating vCPU state through userspace MP_STATE and injected event stuffing. When a vCPU is awakened from a blocking state in L2 (nested guest mode) with an already-injected event, the kernel generates spurious KVM_EXIT_UNKNOWN exits that typically terminate the VM. The vulnerability stems from insufficient validation of impossible vCPU states that userspace can artificially create, despite architectural safeguards that should prevent injected events during blocking. CVSS 5.5 (local, low complexity, high availability impact); EPSS 0.02% indicates minimal widespread exploitation likelihood.
Denial of service in Linux kernel media cx88 driver allows local authenticated attackers to exhaust system resources by triggering a missing DMA unmapping in the snd_cx88_hw_params() error path. The vulnerability causes resource leaks when audio hardware parameter initialization fails, potentially rendering the audio subsystem unavailable. CVSS 5.5 reflects local attack vector with low complexity; EPSS 0.02% indicates minimal real-world exploitation probability despite vendor-released patches across multiple kernel versions.
A use-after-free condition in the libertas USB wireless driver allows local attackers with user privileges to cause a denial of service by triggering a kernel warning and potential crash through rapid firmware loading or repeated usb_tx_block() calls while a USB request is still active. The vulnerability stems from insufficient synchronization of USB URB submissions, enabling concurrent requests on the same URB without enforcing completion of prior transmissions.
Denial of service via kernel warning in MPTCP path manager occurs when combining endpoint removal with fullmesh and flag-setting operations through netlink in the Linux kernel. A local attacker with low privileges can trigger a WARNING in net/mptcp/pm_kernel.c:1074 by sending a crafted sequence of netlink commands, causing the system to emit a kernel warning and potentially become unstable. No known public exploit code exists, but the low CVSS (5.5) and minimal EPSS (0.03%) indicate this is a local DoS with limited real-world impact.
Kernel panic via SError interrupt in Linux kernel media driver (chips-media wave5) when device enters autosuspend during video decoding operations. Local authenticated attackers can trigger a denial of service by queuing buffers to the wave5 video decoder while autosuspend timeout occurs, causing the CPU to access suspended hardware and generate an unrecoverable asynchronous SError, crashing the system. No privilege escalation or code execution; impact limited to availability.
Memory leak in the tw9906 media driver's probe function allows local authenticated attackers to cause denial of service through memory exhaustion. The vulnerability occurs in tw9906_probe() when an error path fails to free memory allocated by v4l2_ctrl_handler_init() and v4l2_ctrl_new_std(), potentially leading to kernel memory depletion on repeated device probe attempts. Vendor-released patches are available across multiple stable kernel branches.
Denial of service in Linux kernel KCM (kernel connection multiplexer) subsystem when processing messages with zero-fragment skbs in frag_list after partial sendmsg failures. A local authenticated attacker can trigger a kernel warning and potentially crash the system by sending a malformed message that fails during data copy, leaving an empty skb in the fragment list. The vulnerability requires local access and low-level socket manipulation, affecting systems running vulnerable Linux kernel versions prior to patching.
A denial of service vulnerability in the Linux kernel's AMD display driver (drm/amd/display) allows local authenticated users to crash the system by accessing link encoder functionality on DisplayPort over USB-C (DPIA) links without proper signal type validation. The vulnerability affects kernel versions before the patches released in stable branches 6.12.75, 6.18.16, 6.19.6, and 7.0. No public exploit code has been identified, and real-world exploitation probability is very low (EPSS 0.02%), suggesting this is primarily an edge-case denial of service affecting specific hardware configurations with DPIA displays.
Memory leak in the Linux kernel's TI K3 SoC info driver (soc/ti/k3-socinfo) fails to release an allocated mmio regmap, causing denial of service through resource exhaustion on probe failures and driver unbind. Local attackers with low privileges can trigger probe deferral or driver unbind to exhaust kernel memory, affecting systems running vulnerable Linux kernel versions. The vulnerability has a low EPSS score (0.02%) but enables practical local DoS against systems where low-privilege users can control driver lifecycle.
Memory leak in Linux kernel tw9903 media driver probe function allows local authenticated attackers to cause denial of service through repeated device initialization failures. The v4l2_ctrl_handler_init() and v4l2_ctrl_new_std() allocations are not freed in one error path of tw9903_probe(), enabling exhaustion of kernel memory. CVSS 5.5 (local, low complexity, requires low privileges). EPSS score of 0.02% indicates minimal exploitation probability.
Local denial of service in Linux kernel ring-buffer tracing code allows authenticated local users to crash the system by triggering invalid memory access through malformed event length fields. The vulnerability exists in rb_read_data_buffer(), which validates possibly corrupted ring buffers at boot but fails to verify event lengths are within acceptable ranges before calculating buffer offsets. EPSS exploitation probability is very low at 0.02%, and no public exploit code or active exploitation has been identified.
Denial of service in Linux kernel minix filesystem implementation allows local authenticated users to crash the system via crafted minix superblock structures due to missing validation of s_log_zone_size and other superblock fields in the minix_check_superblock() function.
Memory leak in Linux kernel fbdev vt8500lcdfb driver allows local authenticated users to cause denial of service via unfreed DMA-allocated buffer on error path. The vulnerability exists in the framebuffer initialization code where dma_alloc_coherent() allocation for fbi->fb.screen_buffer is not properly freed if an error occurs during probe, leading to memory exhaustion on repeated device initialization attempts. Local privilege required; no remote or unauthenticated attack vector.
Denial of service in Linux kernel APEI/GHES ARM processor error handling allows local authenticated attackers to trigger kernel oops by crafting malformed ARM Processor Error records with incomplete or oversized section data, causing out-of-bounds memory dereference. CVSS 5.5 (local, low complexity, authenticated, availability impact only) with EPSS 0.02% indicates low real-world exploitation probability despite public patch availability.
Denial of service in Linux kernel PCI endpoint configfs interface allows local attackers with low privileges to crash the kernel via swapped parameters in pci_primary_epc_epf_unlink() and pci_secondary_epc_epf_unlink() functions. When executing the unlink command in configfs, incorrect parameter ordering causes invalid memory access and kernel panic. CVSS 5.5 (local, low complexity, low privilege) with EPSS 0.02% suggests limited real-world exploitation despite confirmed availability of patches across multiple kernel branches.
Integer underflow in the Linux kernel's EFI/CPER firmware error logging function (cper_print_fw_err) allows local authenticated attackers to trigger denial of service via memory dump of unmapped regions, disclose kernel memory contents, or cause system crash when processing malformed EFI firmware error records with invalid offsets. The vulnerability stems from insufficient validation of error record length before subtracting an offset, causing integer wraparound that permits dumping of arbitrary kernel memory regions.
Kernel panic occurs in Linux kernel USB gadget driver (dwc3) when vbus_draw function executes PMIC power-supply APIs from atomic context, causing sleep operations in non-sleepable context. The vulnerability affects Linux kernel versions prior to 5.13 and various stable series (6.6.128, 6.12.75, 6.18.16, 6.19.6, 7.0) and is resolved by deferring vbus_draw operations to workqueue context. Local authenticated users with privilege level PR:L can trigger denial of service by invoking affected USB gadget functionality, with EPSS exploitation probability at 0.02% percentile indicating extremely low practical risk despite moderate CVSS severity.
Denial of service via kernel panic in the DRM buddy allocator when rounded allocation sizes exceed available memory. Local authenticated attackers can trigger a BUG_ON() crash by requesting contiguous or large-aligned GPU memory allocations that, after power-of-two or block-size rounding, exceed the total VRAM size. Example: a 9GB contiguous allocation request on 10GB VRAM rounds to 16GB, causing immediate kernel panic. No authentication bypass or privilege escalation; requires local access and appropriate user permissions to invoke DRM allocator.
Denial of service via incomplete xattr cleanup in the Linux kernel OCFS2 filesystem causes memory corruption when processing reflinked files with extended attributes. A local user with standard privileges can trigger this vulnerability through crafted xattr operations, resulting in system crashes or data integrity issues. The flaw affects Linux kernel versions from 2.6.32 through multiple recent releases (5.10, 5.15, 6.1, 6.6, 6.12, 6.18, 6.19), with vendor-released patches available for supported stable branches.
Resource leak in Linux kernel xfrm (IPsec) subsystem allows local authenticated users to exhaust kernel memory and trigger denial of service by preventing proper cleanup of IPsec hardware offload device references during network device unregistration. The vulnerability affects XFRM state management when IPsec hardware offloading is configured, particularly when the hardware offload capability (NETIF_F_HW_ESP) is disabled after state creation but before device removal.
Denial of service in Linux kernel dm-verity subsystem allows local authenticated users to crash the system by triggering dm_bufio_client_create() failure in verity_fec_ctr(), which incorrectly passes an error pointer to dm_bufio_client_destroy(). The vulnerability requires local access and authenticated privileges but no user interaction, affecting dm-verity configurations with forward error correction enabled.
Null pointer dereference in the Linux kernel AMD power management (drm/amd/pm) subsystem causes denial of service when SMU (System Management Unit) is disabled during RAS (Reliability, Availability, and Serviceability) initialization. Local authenticated attackers with low privileges can trigger this crash on affected systems, resulting in kernel panic and system unavailability. EPSS exploitation probability is very low (0.02%), indicating this requires specific configuration and local access.
Hard-lockup in Linux kernel IOMMU/VT-d subsystem when flushing device IOTLB during PCIe link-down events affects systems using scalable mode with PASID support. Local authenticated attackers can trigger denial of service by causing PCIe device disconnection followed by resource cleanup operations, such as releasing VFIO group file descriptors, which deadlock the system while attempting ATS invalidation on inaccessible devices. Patch available across multiple stable kernel series.
A circular locking dependency in the Linux kernel's ntfs3 filesystem driver allows local authenticated attackers with low privileges to cause a denial of service (system hang or crash) by triggering simultaneous operations that deadlock between the MFT run lock and bitmap read-write lock. The vulnerability affects kernel versions prior to 6.18.16, 6.19.6, and 7.0, and requires local access with user-level privileges to exploit.
Denial of service via null pointer dereference in Linux kernel's pstore persistent storage subsystem occurs when the vmap() function fails but the persistent_ram_vmap() function incorrectly returns success if a non-zero offset is present, allowing subsequent buffer access to dereference invalid memory and cause system crashes. Affects Linux kernel versions prior to 5.10.252, 5.15.202, 6.1.165, 6.6.128, 6.12.75, 6.18.16, 6.19.6, and 7.0. No public exploit identified at time of analysis, but vendor-released patches are available across multiple stable branches.
Data corruption in Linux kernel btrfs filesystem log replay allows local authenticated attackers to cause files to retain incorrect sizes after crash recovery. When a file is truncated to zero bytes, fssynced, and then a hardlink is created, the file incorrectly retains its pre-truncation size after a power failure and log replay, resulting in data integrity violation with availability impact.
A denial of service vulnerability in the Linux kernel's Qualcomm PD mapper service registry causes system crashes due to mismatched string element length validation in servreg_loc_pfr_req_ei. When a process daemon crashes and triggers a service registry location request, the QMI decoder rejects the reason field because its declared maximum length (65 bytes) is smaller than the actual field size (81 bytes), causing a decoding error and system halt. This affects all Linux kernel versions prior to the patch, triggered by local processes with standard user privileges.
Denial of service in Linux kernel xfrm (IPsec transform) subsystem allows local authenticated attackers to trigger a kernel panic via improper netlink message size calculation when handling XFRM_MSG_GETAE requests for states with interface ID set. The xfrm_aevent_msgsize() function fails to account for XFRMA_IF_ID attribute space, causing build_aevent() to exceed buffer bounds and hit a BUG_ON assertion, resulting in kernel crash. EPSS exploitation probability is very low at 0.02% despite the local attack vector, suggesting limited real-world impact.
Memory leak in the Airoha QDMA RX packet processing function allows local authenticated attackers to cause a denial of service through resource exhaustion. The vulnerability occurs when page pool fragments fail to properly return to the pool during error handling in airoha_qdma_rx_process(), allowing an attacker with local access and low privileges to exhaust kernel memory and crash the system. EPSS exploitation probability is extremely low at 0.02%, reflecting the local-only attack vector and privilege requirement.
Null pointer dereference in Linux kernel bridge VLAN filtering code allows local authenticated attackers to trigger a denial of service via a crafted RTM_NEWLINK netlink message with BR_BOOLOPT_FDB_LOCAL_VLAN_0 flag when CONFIG_BRIDGE_VLAN_FILTERING is disabled. The vulnerability occurs because br_fdb_delete_locals_per_vlan_port() and br_fdb_insert_locals_per_vlan_port() dereference a NULL vlan group pointer without validation, causing a kernel panic. No public exploit code identified at time of analysis.
Infinite vCPU fault loop in the Linux kernel's mshv (Microsoft Hypervisor) subsystem allows a local guest VM process to permanently spin a host vCPU thread, exhausting host CPU resources. The flaw exists in mshv_handle_gpa_intercept(), which unconditionally attempts page remaps on all movable-memory faults regardless of access permission - when a guest writes to a read-only Guest Physical Address region, the remap succeeds but the region retains its read-only designation, causing an immediate re-fault in a tight loop. Affected kernel versions run from commit b9a66cd5ccbb9fade15d0e427e19470d8ad35b75 through the fix commits; patched releases 6.19.14 and 7.0 are available. No public exploit has been identified and EPSS is 0.01%, consistent with the local, hypervisor-specific attack surface.
Linux kernel ASoC SDCA subsystem crashes on sound card teardown due to IRQ lifecycle mismanagement, causing a local denial of service. IRQ handlers registered via devm_request_threaded_irq() during component probe retain stale references to freed card and kcontrol structures after the sound card is torn down, resulting in null or dangling pointer dereferences and kernel panic. Exploitation requires local low-privilege access and SDCA-capable audio hardware; no public exploit exists and EPSS is extremely low at 0.02% (5th percentile).
The AF_XDP socket subsystem (xsk) in the Linux kernel fails to validate that a network device's MTU fits within the usable UMEM frame space at bind time, allowing a local low-privileged user to trigger a kernel denial of service. Usable frame space - chunk size minus headroom and tailroom - can fall below a standard 1500-byte MTU when 2k chunks are used, a gap that became exploitable once tailroom subtraction was introduced. The kernel also omits validation of hardware zero-copy capabilities via net_device::xdp_zc_max_segs. No public exploit has been identified and EPSS is 0.02% (5th percentile), indicating low immediate exploitation risk.
Reference count leak in the Linux kernel's xfrm IPsec subsystem allows a local low-privileged attacker to exhaust kernel memory, resulting in denial of service. The defect resides in xfrm_migrate_policy_find(), where xfrm_pol_hold_rcu() is called twice - once implicitly by the lookup path (which already returns a held reference) and once redundantly - creating a refcount imbalance that prevents memory reclamation. Discovered by the Linux Verification Center using Syzkaller fuzzing; no public exploit identified at time of analysis, and EPSS is very low at 0.02% (5th percentile), indicating minimal observed exploitation activity.
Kernel crash in the Linux pinctrl mcp23s08 driver triggers a NULL pointer dereference during device probe when an MCP23S08/MCP23008 GPIO expander retains non-zero interrupt-on-change state across a warm reboot. The crash occurs because the driver's IRQ handler fires against pins whose nested IRQ handlers have not yet been registered, causing an unhandled kernel oops and system denial of service. Exploitation is local and device-specific; no public exploit identified at time of analysis, and EPSS remains at 0.02% (5th percentile) consistent with no observed in-the-wild triggering.
NULL pointer dereference in the Linux kernel IPVS subsystem causes a kernel panic during IPVS service creation failure cleanup, resulting in a full host denial of service. Affected kernels from 6.2 onward contain a logic error in ip_vs_add_service() where a successful scheduler bind sets the local sched variable to NULL; if ip_vs_start_estimator() subsequently fails, the error path calls ip_vs_unbind_scheduler() with that NULL pointer, dereferencing offset 0x30 and triggering a general protection fault. Exploitation requires local authenticated access with CAP_NET_ADMIN privileges; no public exploit is identified and EPSS is 0.02%, indicating very low exploitation probability.
Denial-of-service via off-by-one allocation in the Linux kernel txgbe network driver allows a local low-privileged user to crash the kernel on systems hosting Wangxun 10GbE NICs. The driver allocates property_entry struct lists without reserving the mandatory null-terminator sentinel slot, meaning kernel subsystems iterating over the list read beyond allocated memory bounds. No active exploitation has been identified and EPSS is extremely low (0.02%, 5th percentile), but patches are available across multiple stable kernel branches including 6.6.136, 6.12.83, 6.18.24, 6.19.14, and 7.0.
Incorrect GENERIC_CMD register field masks in the Linux kernel's IPA (IP Accelerator) network driver for IPA v5.0+ hardware trigger a kernel WARN when a 'stop' command is sent to the MPSS (Modem Processor SubSystem) remoteproc while IPA is active. This availability-only vulnerability (CVSS C:N/I:N/A:H) affects authenticated local users on systems with Qualcomm IPA v5.0+ silicon. No public exploit exists and no KEV listing is present; with an EPSS of 0.02% this is a low-urgency stability fix rather than an active threat.
Out-of-bounds array write in the Linux kernel's Intel uncore PMU driver (`perf/x86/intel/uncore`) causes denial of service on affected Intel multi-die x86 systems. The driver fails to skip discovery table parsing for dies whose CPUs are all offline at boot, allowing the assignment `pmu->boxes[die] = box` in `uncore_pci_pmu_register()` to overflow the boxes array, triggering a kernel WARN or potential memory corruption. Exploitation requires local low-privilege access under a specific hardware and boot configuration; no public exploit exists and EPSS is 0.02%, indicating negligible real-world exploitation likelihood.
Local denial-of-service in the Linux kernel's AEAD crypto socket interface (`algif_aead`) allows a low-privileged local user to crash the kernel by submitting a decryption request where the minimum receive buffer size check fails to account for the authentication tag length. The CVSS vector (AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H) confirms this is a locally exploitable, high-availability-impact issue with no confidentiality or integrity risk. Patches have been released across multiple Linux LTS stable branches (5.10.254, 5.15.204, 6.1.170, 6.6.136, 6.12.83, 6.18.24, 6.19.14, 7.0) and Ubuntu has issued multiple USN advisories (USN-8277-1 through USN-8281-1). No public exploit code has been identified and EPSS is 0.02%, indicating no public exploitation activity.
Oracle OCI CLI version 3.77 allows local attackers with user interaction to place imported files outside the intended directory, compromising file integrity and enabling potential code execution or data exfiltration. The vulnerability requires local access and user interaction but carries high integrity impact through arbitrary file placement. No active exploitation or public exploit code has been identified at the time of analysis.
Kubewarden versions before 1.35.0 permit RBAC reconnaissance attacks when users with AdmissionPolicy or AdmissionPolicyGroup creation privileges craft policies using the unchecked `can_i` host capability. The vulnerability allows enumeration of any user or service account permissions across the cluster via SubjectAccessReview requests executed with policy-server privileges, despite the absence of context-aware resource grants. This information disclosure enables attackers to discover sensitive permission configurations without requiring cluster-wide policy creation rights, a capability not available by default but exploitable when granted.
PocketBase versions before 0.22.42 and 0.30.0-0.37.3 allow account pre-hijacking via OAuth2 autolinking, where an attacker knowing a victim's email can create an unverified account linked to one OAuth2 provider, then retain access when the victim authenticates with a different provider and the accounts are auto-merged, because previous OAuth2 links are not cleared during the upgrade from unverified to verified status. Publicly available exploit code exists; vendor recommends immediate upgrade to v0.37.4 or v0.22.42.