A buffer overflow occurs in pytorch v2.7.0 when a PyTorch model consists of torch.nn.Conv2d, torch.nn.functional.hardshrink, and torch.Tensor.view-torch.mv() and is compiled by Inductor, leading to a Denial of Service (DoS).

An issue in the component torch.linalg.lu of pytorch v2.8.0 allows attackers to cause a Denial of Service (DoS) when performing a slice operation.

An issue in pytorch v2.7.0 can lead to a Denial of Service (DoS) when a PyTorch model consists of torch.Tensor.to_sparse() and torch.Tensor.to_dense() and is compiled by Inductor.

In Indy Node 1.12.2, there is an Uncontrolled Resource Consumption vulnerability. Indy Node has a bug in TAA handling code. The current primary can be crashed with a malformed transaction from a client, which leads to a view change. Repeated rapid view changes have the potential of bringing down the network. This is fixed in version 1.12.3.

MaterialX is an open standard for the exchange of rich material and look-development content across applications and renderers. In version 1.39.2, nested imports of MaterialX files can lead to a crash via stack memory exhaustion, due to the lack of a limit on the "import chain" depth. When parsing file imports, recursion is used to process nested files; however, there is no limit imposed to the depth of files that can be parsed by the library. By building a sufficiently deep chain of MaterialX files one referencing the next, it is possible to crash the process using the MaterialX library via stack exhaustion. This is fixed in version 1.39.3.

opentelemetry-go-contrib is a collection of extensions for OpenTelemetry-Go. The v0.38.0 release of `go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp` uses the `httpconv.ServerRequest` function to annotate metric measurements for the `http.server.request_content_length`, `http.server.response_content_length`, and `http.server.duration` instruments. The `ServerRequest` function sets the `http.target` attribute value to be the whole request URI (including the query string)[^1]. The metric instruments do not "forget" previous measurement attributes when `cumulative` temporality is used, this means the cardinality of the measurements allocated is directly correlated with the unique URIs handled. If the query string is constantly random, this will result in a constant increase in memory allocation that can be used in a denial-of-service attack. This issue has been addressed in version 0.39.0. Users are advised to upgrade. There are no known workarounds for this issue.

KubeEdge is an open source system for extending native containerized application orchestration capabilities to hosts at Edge. Prior to versions 1.11.1, 1.10.2, and 1.9.4, the ServiceBus server on the edge side may be susceptible to a DoS attack if an HTTP request containing a very large Body is sent to it. It is possible for the node to be exhausted of memory. The consequence of the exhaustion is that other services on the node, e.g. other containers, will be unable to allocate memory and thus causing a denial of service. Malicious apps accidentally pulled by users on the host and have the access to send HTTP requests to localhost may make an attack. It will be affected only when users enable the `ServiceBus` module in the config file `edgecore.yaml`. This bug has been fixed in Kubeedge 1.11.1, 1.10.2, and 1.9.4. As a workaround, disable the `ServiceBus` module in the config file `edgecore.yaml`.

OpenTelemetry Java Instrumentation provides OpenTelemetry auto-instrumentation and instrumentation libraries for Java. In versions prior to 2.27.0, the RMI context propagation payload reader limits the number of context entries but does not limit the aggregate size of the strings read from the stream. An attacker who can reach an RMI endpoint on an instrumented JVM can send an oversized context propagation payload. This can cause excessive memory allocation while the JVM reads the payload, potentially leading to denial of service. The issue affects only deployments where RMI instrumentation is enabled and an RMI endpoint is network-reachable. This issue has been fixed in version 2.27.0.

indy-node is the server portion of Hyperledger Indy, a distributed ledger purpose-built for decentralized identity. In vulnerable versions of indy-node, an attacker can max out the number of client connections allowed by the ledger, leaving the ledger unable to be used for its intended purpose. However, the ledger content will not be impacted and the ledger will resume functioning after the attack. This attack exploits the trade-off between resilience and availability. Any protection against abusive client connections will also prevent the network being accessed by certain legitimate users. As a result, validator nodes must tune their firewall rules to ensure the right trade-off for their network's expected users. The guidance to network operators for the use of firewall rules in the deployment of Indy networks has been modified to better protect against denial of service attacks by increasing the cost and complexity in mounting such attacks. The mitigation for this vulnerability is not in the Hyperledger Indy code per se, but rather in the individual deployments of Indy. The mitigations should be applied to all deployments of Indy, and are not related to a particular release.

Missing access control restrictions in the Hypervisor component of the ACRN Project (v2.0 and v1.6.1) allow a malicious entity, with root access in the Service VM userspace, to abuse the PCIe assign/de-assign Hypercalls via crafted ioctls and payloads. This attack results in a corrupt state and Denial of Service (DoS) for previously assigned PCIe devices to the Service VM at runtime.

A Name Error occurs in pytorch v2.7.0 when a PyTorch model consists of torch.cummin and is compiled by Inductor, leading to a Denial of Service (DoS).

pytorch v2.8.0 was discovered to display unexpected behavior when the components torch.rot90 and torch.randn_like are used together.

An issue was discovered in Ceph through 13.2.9. A POST request with an invalid tagging XML can crash the RGW process by triggering a NULL pointer exception.

MaterialX is an open standard for the exchange of rich material and look-development content across applications and renderers. In version 1.39.2, when parsing shader nodes in a MTLX file, the MaterialXCore code accesses a potentially null pointer, which can lead to crashes with maliciously crafted files. An attacker could intentionally crash a target program that uses OpenEXR by sending a malicious MTLX file. This is fixed in version 1.39.3.

MaterialX is an open standard for the exchange of rich material and look-development content across applications and renderers. In versions 1.39.2 and below, when parsing an MTLX file with multiple nested nodegraph implementations, the MaterialX XML parsing logic can potentially crash due to stack exhaustion. An attacker could intentionally crash a target program that uses OpenEXR by sending a malicious MTLX file. This is fixed in version 1.39.3.

In wlan firmware, there is a possible firmware assertion due to improper input handling. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07932637; Issue ID: ALPS07932637.

MaterialX is an open standard for the exchange of rich material and look-development content across applications and renderers. In version 1.39.2, when parsing shader nodes in a MTLX file, the MaterialXCore code accesses a potentially null pointer, which can lead to crashes with maliciously crafted files. An attacker could intentionally crash a target program that uses MaterialX by sending a malicious MTLX file. This is fixed in version 1.39.3.

Rekor is an open source software supply chain transparency log. Rekor prior to version 1.1.1 may crash due to out of memory (OOM) conditions caused by reading archive metadata files into memory without checking their sizes first. Verification of a JAR file submitted to Rekor can cause an out of memory crash if files within the META-INF directory of the JAR are sufficiently large. Parsing of an APK file submitted to Rekor can cause an out of memory crash if the .SIGN or .PKGINFO files within the APK are sufficiently large. The OOM crash has been patched in Rekor version 1.1.1. There are no known workarounds.

In wlan firmware, there is possible system crash due to an integer overflow. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07664741; Issue ID: ALPS07664741.

In wlan firmware, there is possible system crash due to an integer overflow. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07664735; Issue ID: ALPS07664735.

In wlan firmware, there is possible system crash due to an uncaught exception. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07664711; Issue ID: ALPS07664711.

In wlan firmware, there is possible system crash due to an uncaught exception. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07664720; Issue ID: ALPS07664720.

containerd is an open-source container runtime. A bug was found in the containerd's CRI implementation where containerd, starting in version 2.0.1 and prior to version 2.0.5, doesn't put usernamespaced containers under the Kubernetes' cgroup hierarchy, therefore some Kubernetes limits are not honored. This may cause a denial of service of the Kubernetes node. This bug has been fixed in containerd 2.0.5+ and 2.1.0+. Users should update to these versions to resolve the issue. As a workaround, disable usernamespaced pods in Kubernetes temporarily.

An issue was discovered in Open Network Operating System (ONOS) 1.14. In the P4 tutorial application (org.onosproject.p4tutorial), the host event listener does not handle the following event types: HOST_MOVED, HOST_REMOVED, HOST_UPDATED. In combination with other applications, this could lead to the absence of intended code execution.

An issue was discovered in Open Network Operating System (ONOS) 1.14. In the virtual tenant network application (org.onosproject.vtn), the host event listener does not handle the following event types: HOST_MOVED. In combination with other applications, this could lead to the absence of intended code execution.

An issue was discovered in Open Network Operating System (ONOS) 1.14. In the access control application (org.onosproject.acl), the host event listener does not handle the following event types: HOST_REMOVED. In combination with other applications, this could lead to the absence of intended code execution.

An issue was discovered in Open Network Operating System (ONOS) 1.14. In the mobility application (org.onosproject.mobility), the host event listener does not handle the following event types: HOST_ADDED, HOST_REMOVED, HOST_UPDATED. In combination with other applications, this could lead to the absence of intended code execution.

LengthPrefixedMessageReader in gRPC Swift 1.1.0 and earlier allocates buffers of arbitrary length, which allows remote attackers to cause uncontrolled resource consumption and deny service.

In MPD before 0.23.8, as used on Automotive Grade Linux and other platforms, the PipeWire output plugin mishandles a Drain call in certain situations involving truncated files. Eventually there is an assertion failure in libmpdclient because libqtappfw passes in a NULL pointer.

An integer overflow in NATS Server before 2.0.2 allows a remote attacker to crash the server by sending a crafted request. If authentication is enabled, then the remote attacker must have first authenticated.

qubes-mirage-firewall (aka Mirage firewall for QubesOS) 0.8.x through 0.8.3 allows guest OS users to cause a denial of service (CPU consumption and loss of forwarding) via a crafted multicast UDP packet (IP address range of 224.0.0.0 through 239.255.255.255).

NATS Server is a high-performance server for NATS.io, the cloud and edge native messaging system. Prior to 2.12.8 and 2.11.17, an unauthenticated peer with network access to a leafnode listener with compression enabled could crash the server during the pre-authentication leafnode handshake by sending repeated leafnode INFO protocol messages before authentication and account setup completed. This issue is fixed in versions 2.12.8 and 2.11.17.

Magma versions <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) are susceptible to an assertion-based crash when an oversized NAS packet is received. An attacker may leverage this behavior to repeatedly crash the MME via either a compromised base station or via an unauthenticated cellphone within range of a base station managed by the MME, causing a denial of service.

In Wi-Fi driver, there is a possible way to disconnect Wi-Fi due to an improper resource release. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07030600; Issue ID: ALPS07030600.

In Wi-Fi, there is a possible low throughput due to misrepresentation of critical information. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: GN20220829014; Issue ID: GN20220829014.

This affects all versions of package github.com/nats-io/nats-server/server. Untrusted accounts are able to crash the server using configs that represent a service export/import cycles. Disclaimer from the maintainers: Running a NATS service which is exposed to untrusted users presents a heightened risk. Any remote execution flaw or equivalent seriousness, or denial-of-service by unauthenticated users, will lead to prompt releases by the NATS maintainers. Fixes for denial of service issues with no threat of remote execution, when limited to account holders, are likely to just be committed to the main development branch with no special attention. Those who are running such services are encouraged to build regularly from git.

NATS Server is a high-performance server for NATS.io, the cloud and edge native messaging system. Prior to 2.14.3 and 2.12.12, a WebSocket listener could route requests for the MQTT-over-WebSocket path into MQTT handling even when MQTT was not configured, allowing an unauthenticated client with access to the WebSocket listener to reach uninitialized MQTT state and crash the server process. This issue is fixed in versions 2.14.3 and 2.12.12.

grpc-swift is the Swift language implementation of gRPC, a remote procedure call (RPC) framework. Prior to version 1.7.2, a grpc-swift server is vulnerable to a denial of service attack via a reachable assertion. This is due to incorrect logic when handling GOAWAY frames. The attack is low-effort: it takes very little resources to construct and send the required sequence of frames. The impact on availability is high as the server will crash, dropping all in flight connections and requests. This issue is fixed in version 1.7.2. There are currently no known workarounds.

In wlan firmware, there is possible system crash due to an integer overflow. This could lead to remote denial of service with no additional execution privileges needed. User interaction is not needed for exploitation. Patch ID: ALPS07664731; Issue ID: ALPS07664731.

The JWT library in NATS nats-server before 2.1.9 allows a denial of service (a nil dereference in Go code).

AGL agl-service-can-low-level thru 17.1.12 contains a stack buffer overflow in the uds-c library. The send_diagnostic_request function in uds.c allocates a 6-byte stack buffer (MAX_DIAGNOSTIC_PAYLOAD_SIZE=6) but copies up to 7 bytes (MAX_UDS_REQUEST_PAYLOAD_LENGTH=7) via memcpy at an offset of 1+pid_length (2-3 bytes), resulting in 1-4 bytes of controlled stack overflow. The payload_length field (uint8_t) has no bounds check against the destination buffer. On 32-bit ARM automotive ECUs without stack canaries, this can lead to return address overwrite and RCE.

NATS-Server is a High-Performance server for NATS.io, a cloud and edge native messaging system. Prior to versions 2.11.15 and 2.12.6, a client which can connect to the leafnode port can crash the nats-server with a certain malformed message pre-authentication. Versions 2.11.15 and 2.12.6 contain a fix. As a workaround, disable leafnode support if not needed or restrict network connections to the leafnode port, if plausible without compromising the service offered.

NATS-Server is a High-Performance server for NATS.io, a cloud and edge native messaging system. Prior to versions 2.11.15 and 2.12.6, a malicious client which can connect to the WebSockets port can cause unbounded memory use in the nats-server before authentication; this requires sending a corresponding amount of data. This is a milder variant of CVE-2026-27571. That earlier issue was a compression bomb, this vulnerability is not. Attacks against this new issue thus require significant client bandwidth. Versions 2.11.15 and 2.12.6 contain a fix. As a workaround, disable websockets if not required for project deployment.

The Linux Foundation Magma <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) was discovered to contain a buffer overflow in the decode_protocol_configuration_options function at /3gpp/3gpp_24.008_sm_ies.c. This vulnerability allows attackers to cause a Denial of Service (DoS) via a crafted NAS packet.

The Linux Foundation Magma <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) was discovered to contain a buffer overflow in the decode_traffic_flow_template_packet_filter function at /3gpp/3gpp_24.008_sm_ies.c. This vulnerability allows attackers to cause a Denial of Service (DoS) via a crafted NAS packet.

NATS-Server is a High-Performance server for NATS.io, a cloud and edge native messaging system. Prior to versions 2.11.14 and 2.12.5, if the nats-server has the "leafnode" configuration enabled (not default), then anyone who can connect can crash the nats-server by triggering a panic. This happens pre-authentication and requires that compression be enabled (which it is, by default, when leafnodes are used). Versions 2.11.14 and 2.12.5 contain a fix. As a workaround, disable compression on the leafnode port.

The Linux Foundation Magma <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) was discovered to contain a buffer overflow in the decode_esm_message_container function at /nas/ies/EsmMessageContainer.cpp. This vulnerability allows attackers to cause a Denial of Service (DoS) via a crafted NAS packet.

A reachable assertion in the decode_linked_ti_ie function of Magma <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) allows attackers to cause a Denial of Service (DoS) via a crafted NAS packet.

The Linux Foundation Magma <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) was discovered to contain a buffer overflow in the decode_access_point_name_ie function at /3gpp/3gpp_24.008_sm_ies.c. This vulnerability allows attackers to cause a Denial of Service (DoS) via a crafted NAS packet.

The Linux Foundation Magma <= 1.8.0 (fixed in v1.9 commit 08472ba98b8321f802e95f5622fa90fec2dea486) was discovered to contain a buffer overflow in the decode_pdn_address function at /nas/ies/PdnAddress.cpp. This vulnerability allows attackers to cause a Denial of Service (DoS) via a crafted NAS packet.