CVE-2026-33298 is a high-severity heap-based buffer overflow vulnerability in the ggml-org llama.cpp library, affecting versions prior to b7824. The flaw arises from an integer overflow in the ggml_nbytes function when processing crafted GGUF files with malicious tensor dimensions, causing memory size miscalculations. This leads to a heap buffer overflow during tensor processing, enabling potential remote code execution via memory corruption. Exploitation requires local access and user interaction but no privileges. The vulnerability impacts confidentiality, integrity, and availability of affected systems. A fix is included in version b7824. No known exploits are currently reported in the wild. Organizations using llama.cpp for LLM inference should urgently update to the patched version to mitigate risk.
cgltf version 1.15 and prior contain an integer overflow vulnerability in the cgltf_validate() function when validating sparse accessors that allows attackers to trigger out-of-bounds reads by supplying crafted glTF/GLB input files with attacker-controlled size values. Attackers can exploit unchecked arithmetic operations in sparse accessor validation to cause heap buffer over-reads in cgltf_calc_index_bound(), resulting in denial of service crashes and potential memory disclosure.
libp2p-rust is the official rust language Implementation of the libp2p networking stack. In versions prior to 0.49.3, the Gossipsub implementation accepts attacker-controlled PRUNE backoff values and may perform unchecked time arithmetic when storing backoff state. A specially crafted PRUNE control message with an extremely large backoff (e.g.u64::MAX) can lead to Duration/Instant overflow during backoff update logic, triggering a panic in the networking state machine. This is remotely reachable over a normal libp2p connection and does not require authentication. Any application exposing a libp2p Gossipsub listener and using the affected backoff-handling path can be crashed by a network attacker that can reach the service port. The attack can be repeated by reconnecting and replaying the crafted control message. This issue has been fixed in version 0.49.3.
HTSlib is a library for reading and writing bioinformatics file formats. GZI files are used to index block-compressed GZIP [BGZF] files. In the GZI loading function, `bgzf_index_load_hfile()`, it was possible to trigger an integer overflow, leading to an under- or zero-sized buffer being allocated to store the index. Sixteen zero bytes would then be written to this buffer, and, depending on the result of the overflow the rest of the file may also be loaded into the buffer as well. If the function did attempt to load the data, it would eventually fail due to not reading the expected number of records, and then try to free the overflowed heap buffer. Exploiting this bug causes a heap buffer overflow. If a user opens a file crafted to exploit this issue, it could lead to the program crashing, or overwriting of data and heap structures in ways not expected by the program. It may be possible to use this to obtain arbitrary code execution. Versions 1.23.1, 1.22.2 and 1.21.1 include fixes for this issue. The easiest work-around is to discard any `.gzi` index files from untrusted sources, and use the `bgzip -r` option to recreate them.
CVE-2026-2809 is an integer overflow vulnerability in the Netskope Endpoint DLP Module for Windows clients. It can be triggered by a privileged user when the Endpoint DLP module is enabled, causing a Blue Screen of Death (BSOD) and resulting in a denial-of-service condition on the local machine. Exploitation does not require user interaction but does require high privileges. The vulnerability stems from improper handling of integer values in the DLL Injector component, leading to overflow or wraparound. No known exploits are currently reported in the wild. The CVSS 4.0 score is 6.7 (medium severity), reflecting the local attack vector and requirement for privileged access. Organizations using Netskope Endpoint DLP on Windows should prioritize patching or mitigating this issue to prevent potential service disruption. This vulnerability primarily impacts environments with the Endpoint DLP module enabled and Windows-based endpoints.
CVE-2025-15584 is an integer overflow vulnerability in the Netskope Endpoint DLP Module for Windows clients. An unprivileged local user can exploit this flaw by triggering an integer overflow in the filter communication port, causing a Blue Screen of Death (BSOD) and resulting in a denial-of-service condition. Exploitation requires the Endpoint DLP module to be enabled and does not require user interaction or elevated privileges beyond local user access. The vulnerability has a CVSS 4.0 base score of 6.8, indicating medium severity. There are no known exploits in the wild currently, and no patches have been published yet. Organizations using Netskope Endpoint DLP on Windows should be aware of this risk and monitor for updates. The impact is primarily local denial-of-service, affecting system availability but not confidentiality or integrity. Mitigation involves restricting local user access, disabling the Endpoint DLP module if not needed, and applying patches once available.
Yamux is a stream multiplexer over reliable, ordered connections such as TCP/IP. From 0.13.0 to before 0.13.9, a specially crafted WindowUpdate can cause arithmetic overflow in send-window accounting, which triggers a panic in the connection state machine. This is remotely reachable over a normal network connection and does not require authentication. This vulnerability is fixed in 0.13.9.
llama.cpp is an inference of several LLM models in C/C++. Prior to b8146, the gguf_init_from_file_impl() in gguf.cpp is vulnerable to an Integer overflow, leading to an undersized heap allocation. Using the subsequent fread() writes 528+ bytes of attacker-controlled data past the buffer boundary. This is a bypass of a similar bug in the same file - CVE-2025-53630, but the fix overlooked some areas. This vulnerability is fixed in b8146.
DNG SDK versions 1.7.1 2471 and earlier are affected by an Integer Overflow or Wraparound vulnerability that could lead to application denial-of-service. An attacker could exploit this vulnerability to cause the application to crash or become unresponsive. Exploitation of this issue requires user interaction in that a victim must open a malicious file.
Crypt::Sodium::XS versions through 0.001000 for Perl has potential integer overflows.
Combined aead encryption, combined signature creation, and bin2hex functions do not check that output size will be less than SIZE_MAX, which could lead to integer wraparound causing an undersized output buffer. This can cause a crash in bin2hex and encryption algorithms other than aes256gcm. For aes256gcm encryption and signatures, an undersized buffer could lead to buffer overflow.
Encountering this issue is unlikely as the message length would need to be very large.
For bin2hex the input size would have to be > SIZE_MAX / 2 For aegis encryption the input size would need to be > SIZE_MAX - 32U For other encryption the input size would need to be > SIZE_MAX - 16U For signatures the input size would need to be > SIZE_MAX - 64U
Crypt::NaCl::Sodium versions through 2.002 for Perl has potential integer overflows.bin2hex, encrypt, aes256gcm_encrypt_afternm and seal functions do not check that output size will be less than SIZE_MAX, which could lead to integer wraparound causing an undersized output buffer.
Encountering this issue is unlikely as the message length would need to be very large.
For bin2hex() the bin_len would have to be > SIZE_MAX / 2 For encrypt() the msg_len would need to be > SIZE_MAX - 16U For aes256gcm_encrypt_afternm() the msg_len would need to be > SIZE_MAX - 16U For seal() the enc_len would need to be > SIZE_MAX - 64U
TinyWeb is a web server (HTTP, HTTPS) written in Delphi for Win32. Prior to version 2.03, an integer overflow vulnerability in the string-to-integer conversion routine (_Val) allows an unauthenticated remote attacker to bypass Content-Length restrictions and perform HTTP Request Smuggling. This can lead to unauthorized access, security filter bypass, and potential cache poisoning. The impact is critical for servers using persistent connections (Keep-Alive). This issue has been patched in version 2.03.
WARNING:
Users of 6.x should upgrade to 6.2.4 or later as the fix was missed in previous 6.x releases.
See the following for more details:
https://activemq.apache.org/security-advisories.data/CVE-2026-40046-announcement.txt
https://www.cve.org/CVERecord?id=CVE-2026-40046
Original Report:
Apache ActiveMQ does not properly validate the remaining length field which may lead to an overflow during the decoding of malformed packets. When this integer overflow occurs, ActiveMQ may incorrectly compute the total Remaining Length and subsequently misinterpret the payload as multiple MQTT control packets which makes the broker susceptible to unexpected behavior when interacting with non-compliant clients. This behavior violates the MQTT v3.1.1 specification, which restricts Remaining Length to a maximum of 4 bytes. The scenario occurs on established connections after the authentication process. Brokers that are not enabling mqtt transport connectors are not impacted.
This issue affects Apache ActiveMQ: before 5.19.2, 6.0.0 to 6.1.8, and 6.2.0
Users are recommended to upgrade to version 5.19.2, 6.1.9, or 6.2.1, which fixes the issue.
pillow_heif is a Python library for working with HEIF images and plugin for Pillow. Prior to version 1.3.0, an integer overflow in the encode path buffer validation of `_pillow_heif.c` allows an attacker to bypass bounds checks by providing large image dimensions, resulting in a heap out-of-bounds read. This can lead to information disclosure (server heap memory leaking into encoded images) or denial of service (process crash). No special configuration is required — this triggers under default settings. Version 1.3.0 fixes the issue.
iccDEV provides a set of libraries and tools for working with ICC color management profiles. In versions up to and including 2.3.1.4, signed integer overflow in iccFromCube.cpp during multiplication triggers undefined behavior, potentially causing crashes or incorrect ICC profile generation when processing crafted/large cube inputs. Commit 43ae18dd69fc70190d3632a18a3af2f3da1e052a fixes the issue. No known workarounds are available.