Buffer Overflow

In the Linux kernel, the following vulnerability has been resolved: nfc: digital: clamp SENSF_RES length to the destination buffer digital_in_recv_sensf_res() memcpy()s resp->len bytes from a remote NFC-F device response into the NFC_SENSF_RES_MAXSIZE-byte target.sensf_res field without an upper-bound check. A nearby malicious NFC-F device can send an oversized SENSF_RES response to overflow the stack-local struct nfc_target. Clamp resp->len to NFC_SENSF_RES_MAXSIZE before the copy. Found by 0sec automated security-research tooling (https://0sec.ai).

In the Linux kernel, the following vulnerability has been resolved: nfc: fdp: bound the device-reported read length and fix an skb leak fdp_nci_i2c_read() takes the next packet length from two device-supplied bytes and never validates it. The value is a u16 used as the i2c_master_recv() count into a 261-byte on-stack buffer: a malicious, counterfeit or malfunctioning controller (or an i2c bus interposer) can drive it far past the buffer for a stack out-of-bounds write that clobbers the canary and return address, or below the minimum frame size (directly, or by truncating the computed sum) so the header/LRC strip and the next length read run past a short receive. Reject a length outside [FDP_NCI_I2C_MIN_PAYLOAD, FDP_NCI_I2C_MAX_PAYLOAD], as a corrupted packet already is, and force resynchronization. The same loop allocates one data skb per iteration and assumes a length packet followed by a data packet; a device that sends two data packets in one call leaks the first skb when the second allocation overwrites it. Free a previously allocated skb before allocating the next.

In the Linux kernel, the following vulnerability has been resolved: nfc: llcp: reject PDUs shorter than the LLCP header Every LLCP PDU begins with a two-byte header (DSAP/SSAP + PTYPE), but the receive path never checked that a frame is at least LLCP_HEADER_SIZE bytes before parsing it. nfc_llcp_rx_skb() reads the header via nfc_llcp_ptype()/nfc_llcp_dsap()/ nfc_llcp_ssap(), which dereference pdu->data[0] and pdu->data[1], and a CONNECT or CC PDU then computes tlv_array_len = skb->len - LLCP_HEADER_SIZE; as a size_t and hands it to the TLV walk. When the frame is shorter than the header the subtraction wraps to a huge value and the walk runs far past the buffer, an out-of-bounds read. A nearby NFC device can reach this without authentication; LLCP link activation happens automatically after NFC-DEP. Guard the common receive choke point __nfc_llcp_recv(), shared by both the target (nfc_llcp_data_received()) and initiator (nfc_llcp_recv()) paths, so a short skb is dropped before the rx_work worker parses it. Use pskb_may_pull() rather than a skb->len test so the two header bytes are guaranteed to sit in the skb linear area even for a non-linear skb, matching how the sibling NCI and HCI receive paths validate their headers. Reproduced with a KFENCE out-of-bounds read via /dev/virtual_nci on linux-next. Found by 0sec automated security-research tooling (https://0sec.ai).

In the Linux kernel, the following vulnerability has been resolved: nfc: nci: add data_len bound checks to activation parameter extractors nci_extract_activation_params_iso_dep() and nci_extract_activation_params_nfc_dep() read an inner length byte from the NCI RF_INTF_ACTIVATED_NTF payload and use it to memcpy() into fixed kernel buffers, but neither function receives the caller-validated activation_params_len. A crafted NCI notification with activation_params_len=1 and an inner length byte of up to 20 (NFC-A) or 50 (NFC-B) causes memcpy() to read that many bytes past the one valid byte in the activation params region -- a slab out-of-bounds read of kernel memory adjacent to the NCI skb. The sibling nci_extract_rf_params_*() family was given equivalent protection by commit 571dcbeb8e63 ("net: nfc: nci: Fix parameter validation for packet data"), but the two activation parameter extractors were not updated at that time. Add a data_len parameter to both functions, guard against an empty region before consuming the inner length byte, decrement the remaining count after consuming it, and clamp the copy length to what is actually available. Update both call sites to pass ntf.activation_params_len, which is already validated against the skb at ntf.c:801.

In the Linux kernel, the following vulnerability has been resolved: nfc: nci: fix out-of-bounds write in nci_target_auto_activated() nci_target_auto_activated() appends a target to the fixed-size array ndev->targets[NCI_MAX_DISCOVERED_TARGETS] and increments ndev->n_targets without first checking the array is full; unlike its sibling nci_add_new_target(), which bails out when n_targets already equals NCI_MAX_DISCOVERED_TARGETS. ndev->n_targets is only cleared by nci_clear_target_list(), so an NFCC that repeatedly re-runs discovery (RF_DISCOVER_RSP, which re-enters NCI_DISCOVERY without clearing the target list) and reports an auto-activated target (RF_INTF_ACTIVATED_NTF) drives n_targets past the limit. The append then writes a struct nfc_target past the end of the array (a slab out-of-bounds write), and nfc_targets_found() goes on to walk the array with the inflated count: BUG: KASAN: slab-out-of-bounds in nci_add_new_protocol+0x94/0x2ac [nci] Write of size 2 at addr ffff0000c7299a18 by task kworker/u8:0/12 Workqueue: nfc0_nci_rx_wq nci_rx_work [nci] Call trace: nci_add_new_protocol+0x94/0x2ac [nci] nci_ntf_packet+0xddc/0x11a0 [nci] nci_rx_work+0x15c/0x1e0 [nci] process_one_work+0x2dc/0x500 worker_thread+0x240/0x460 kthread+0x1c0/0x1d0 ret_from_fork+0x10/0x20 The buggy address belongs to the cache kmalloc-2k of size 2048 The buggy address is located 1024 bytes to the right of allocated 1560-byte region [ffff0000c7299000, ffff0000c7299618) Guard nci_target_auto_activated() with the same check used by nci_add_new_target().

In the Linux kernel, the following vulnerability has been resolved: HID: magicmouse: prevent unbounded recursion in magicmouse_raw_event() magicmouse_raw_event() handles DOUBLE_REPORT_ID (0xf7) packets, which pack two touch reports into one, by splitting the packet and calling itself on each half. The only guard against runaway recursion is a "size < 1" check, which stops zero-sized calls but does not bound the recursion depth. A malicious HID device that matches this driver can send a report starting with DOUBLE_REPORT_ID and filled with the sequence [0xf7, 0x00]. Each level consumes two bytes and recurses on the remainder, so an incoming report of up to HID_MAX_BUFFER_SIZE (16 KiB) drives roughly 8000 nested calls. That easily exhausts the 16 KiB kernel stack, leading to a stack overflow: a panic with CONFIG_VMAP_STACK, or memory corruption without it. A double report only ever wraps two normal reports; it is never legitimately nested. Refuse to re-enter the DOUBLE_REPORT_ID case from a recursive call so the recursion depth is bounded to two, while all valid packets keep being parsed exactly as before.

In the Linux kernel, the following vulnerability has been resolved: HID: core: fix OOB read of field->usage in hid_set_field() hid_set_field() hands field->usage + offset to hid_dump_input() before the guard that bounds offset: hid_dump_input(field->report->device, field->usage + offset, value); if (offset >= field->report_count) { hid_err(...); return -1; } Under CONFIG_DEBUG_FS hid_dump_input() dereferences that pointer, with buf = hid_resolv_usage(usage->hid, NULL). The usage[] array is allocated inline with the hid_field in hid_register_field() and holds field->maxusage entries, so an offset past it reads off the end of the kvzalloc()ed allocation and into a neighbouring object. Had the guard run first, offset < report_count <= maxusage would already have confined the pointer to the array. A caller supplies such an offset today. picolcd_fb_send_tile() validates only report->maxfield before issuing hid_set_field(report->field[0], 11 + i, ...) for i = 0..31, so its offsets are fixed at 11..42 and are never checked against the bound field. When the device registers that field with fewer usages, the framebuffer deferred-io work drives the read on every tile. KASAN reports a 4-byte slab-out-of-bounds read in hid_dump_input() below hid_set_field(), and the same boot logs "offset (1) exceeds report_count (1)" from the guard that runs only afterwards. Move the hid_dump_input() call below the guard. Because field->maxusage >= field->report_count, the guard then establishes that field->usage + offset lies inside the array before it is dereferenced, for every caller and without changing behaviour on the valid path. Discovered by XBOW, triaged by Baul Lee <baul.lee@xbow.com>

In the Linux kernel, the following vulnerability has been resolved: HID: pidff: fix OOB write when hid->inputs is empty hid_pidff_init_with_quirks() derives its input_dev from list_entry(hid->inputs.next, struct hid_input, list) without first checking that hid->inputs is non-empty. The list member of struct hid_input is at offset 0, so on an empty list list_entry() yields &hid->inputs itself and the following hidinput->input load reads an unrelated member of struct hid_device. dev is then a type-confused pointer, and force-feedback init writes through it: each set_bit(FF_*, dev->ffbit) stores 8 bytes at dev + 192, past the end of the object dev actually aliases, and input_ff_create() adds further writes of a heap pointer and two function pointers. Until hid-universal-pidff the only caller was hid_pidff_init() from usbhid, which runs under HID_CLAIMED_INPUT and therefore always has at least one hid_input. universal_pidff_probe() starts the device with HID_CONNECT_DEFAULT & ~HID_CONNECT_FF and then calls hid_pidff_init_with_quirks() directly whenever the descriptor carries a PID usage page, bypassing that gate. A report descriptor whose only application collection is on HID_UP_PID leaves hid->inputs empty while hid_connect() still succeeds through the hidraw claim, so probe reaches the unguarded list_entry(). The write happens in the USB probe path, on the hotplug workqueue, so plugging in a malicious device is enough to trigger it; no attacker software and no logged-in user are required. KASAN reports an 8-byte out-of-bounds write in hid_pidff_init_with_quirks() reached from universal_pidff_probe(). Check for an empty list before deriving dev and return -ENODEV, as the other HID force-feedback drivers already do. universal_pidff_probe() propagates the error and unwinds. Discovered by XBOW, triaged by Baul Lee <baul.lee@xbow.com>

In the Linux kernel, the following vulnerability has been resolved: HID: nintendo: fix out-of-bounds read in joycon_ctlr_read_handler() joycon_ctlr_read_handler() casts an incoming HID input report to struct joycon_input_report and parses it, guarding the cast only with a 12-byte length check: if (size >= 12) /* make sure it contains the input report */ joycon_parse_report(ctlr, (struct joycon_input_report *)data); struct joycon_input_report is 49 bytes: a 13-byte header followed by a union whose IMU arm is 36 bytes. For an IMU report joycon_parse_report() -> joycon_parse_imu_report() walks that union (struct offsets 13..48), so a report of exactly 12 bytes with data[0] == JC_INPUT_IMU_DATA passes the guard yet is read up to 37 bytes past its declared length. The over-read bytes are decoded into accelerometer/gyroscope values and forwarded to userspace through the "(IMU)" input device, leaking driver-internal memory. data[0] and size are fully controlled by a malicious or spoofed Joy-Con/Pro Controller. Receive buffers are sized to the maximum report length, so this is an over-read within the allocation rather than a slab OOB, but the decoded bytes still reach userspace. The sibling subcmd path in joycon_ctlr_handle_event() already bounds the same cast correctly: if (size < sizeof(struct joycon_input_report) || data[0] != JC_INPUT_SUBCMD_REPLY) break; Use the same sizeof(struct joycon_input_report) bound here.

Out-of-bounds read in libheif 1.23.1 persists because the patch shipped for GHSA-73p7-m7gg-w2jv did not fully remediate the underlying flaw. Processing a specially crafted HEIF or HEIC image file can trigger a read beyond allocated buffer bounds, potentially leaking heap memory contents or crashing the parsing application. Ubuntu has flagged this regression, meaning deployments that upgraded to 1.23.1 expecting full protection remain exposed to the original vulnerability class.

Out-of-bounds read in libheif 1.23.1 persists because the upstream patch for GHSA-73p7-m7gg-w2jv was incomplete, leaving the library still vulnerable in its latest release at time of disclosure. Any application or pipeline that uses libheif to parse HEIF/HEIC image files - including desktop viewers, photo management tools, and server-side media processors - is exposed when handling untrusted image input. The practical impact is at minimum information disclosure (heap memory leak) and potential application crash; no public exploit has been identified at time of analysis.

Stack corruption with potential code execution in MOOS core-moos through version 10.4.0 arises from an off-by-one error in CMOOSSerialPort::GetTelegram(), where a NUL terminator is written one byte past the end of a fixed-size stack buffer. An attacker who controls the serial communication line connected to a MOOS-enabled platform can send a full-length telegram to trigger the overwrite, corrupting adjacent stack memory and potentially redirecting execution. A vendor patch is available via GitHub PR #73 and commit befb04df; no public exploit has been identified at time of analysis.

Stack-based buffer overflow in OpenVPN's Windows service component (openvpnserv) affects versions 2.5.0 through 2.7.6, requiring high attack complexity and passive user interaction to trigger. The flaw produces limited integrity and availability impact on the vulnerable Windows system and marginally on downstream systems in the VPN path. Fixed in v2.7.7, released on GitHub and announced via the OpenVPN community security page, alongside several other Windows-specific CVEs addressed in the same release cycle. No public exploit or active exploitation has been identified at time of analysis.

Heap buffer out-of-bounds write in the MongoDB C Driver's BSON-to-JSON conversion helpers allows an unauthenticated party who supplies document content to corrupt a small number of bytes past the end of a heap-allocated buffer. The overflow is triggered specifically when a binary BSON field is being encoded and the caller has configured a maximum output length that causes the conversion to be cut short. Impact is limited to a low-integrity write of a few bytes adjacent to the buffer - no confidentiality loss and no confirmed availability impact - but heap corruption in native C applications can have unpredictable downstream effects depending on allocator layout. No public exploit code and no CISA KEV listing exist at time of analysis.

In the Linux kernel, the following vulnerability has been resolved: selinux: reject a class permission count below its inherited common security_get_permissions() maps an inherited common's permissions into an array sized by the class's own permissions.nprim, but class_read() takes that nprim verbatim from the policy image and never checks that it covers the common. A class that inherits a common of N permissions while declaring a smaller nprim is accepted, and on load the common's permissions are written past the class-sized array -- an out-of-bounds heap write. Reject a class whose permission count is below its inherited common's. Well-formed policies, where the class count already includes the inherited permissions, are unaffected.

In the Linux kernel, the following vulnerability has been resolved: selinux: reject a permission value exceeding the class permission count perm_read() bounds a permission value by SEL_VEC_MAX but never by the nprim of the owning class or common, which is taken verbatim from the policy image. security_get_permissions() then writes perms[value - 1] into an nprim-sized kcalloc() array, so a class declaring fewer permissions than its largest permission value drives an out-of-bounds heap write. The top-level symbol tables are validated this way; the nested per-class permission table is not. Reject a permission whose value exceeds nprim, which is already set when perm_read() runs. Well-formed policies are unaffected. [PM: tweak comment for line length]

Stack-based buffer overflow in slurmstepd, the SLURM job step daemon, can be triggered by an authenticated cluster user who submits a job containing an oversized SPANK (SLURM Plug-in Architecture for Node and job Control) option variable in the job environment. Because slurmstepd typically executes with elevated privileges on compute nodes, successful exploitation could allow a regular SLURM user to achieve privilege escalation or arbitrary code execution on cluster nodes. No public exploit or CISA KEV listing has been identified at time of analysis; this report originates from Ubuntu's vendor security tracking.

Out-of-bounds write in Samsung TizenFX's MediaBufferBase component allows a local low-privileged attacker to corrupt memory by supplying unchecked array indices through the Tizen.Multimedia media buffer API. The vulnerable indexer property in `Tizen.Multimedia.Common.Internal.MediaBufferBase` lacked all bounds validation prior to fixes introduced in upstream PRs #7814 and #7815, admitting negative or oversized index values. Under high-complexity conditions, successful exploitation can produce high integrity and availability impact - crashing the multimedia subsystem or enabling controlled memory corruption - on Tizen-based devices such as Samsung Smart TVs and Galaxy Watch wearables. No public exploit exists and no active exploitation has been confirmed.

Buffer overflow (CWE-119) in the Drupal Gammu SMS Daemon contributed module exposes high-privilege authenticated administrators to memory corruption that yields both confidentiality and integrity compromise. The vulnerability affects all released versions of the module (*.* wildcard CPE), with no patched release confirmed in the provided data. No public exploit code or CISA KEV listing has been identified at time of analysis, and the high attack complexity combined with the requirement for elevated Drupal privileges substantially limits opportunistic exploitation.

Buffer overflow in the Drupal Gammu SMS Daemon contributed module exposes sites to memory corruption with high confidentiality and integrity impact. The vulnerability (CWE-119) requires network-reachable access by a highly privileged user under specific conditions, limiting its practical reach but making it serious when those conditions are met. No public exploit code has been identified at time of analysis, and the Drupal security team has issued advisory SA-contrib-2026-100.

Buffer overflow (CWE-119) in the Drupal Gammu SMS Daemon contributed module allows a highly privileged, authenticated attacker to achieve high confidentiality and integrity impact against affected Drupal installations. The CVSS vector (PR:H, AC:H) constrains real-world exploitability significantly, requiring both administrative access and complex attack conditions. No active exploitation has been identified; all tracked versions of the module are reported as affected with no specific patched release confirmed in available data.

Stack overflow vulnerabilities in an HPE AOS-CX API endpoint allow an authenticated, high-privileged attacker to crash the affected system via a crafted request, resulting in a denial-of-service condition. The vulnerability is network-exploitable but requires prior administrative authentication, substantially limiting the attacker pool. This CVE is part of the same HPE advisory (hpesbnw05134) covering multiple AOS-CX weaknesses, including higher-severity unauthenticated and adjacent-network issues; no public exploit or CISA KEV listing has been identified at time of analysis.

Buffer overflow conditions in an underlying network service of HPE AOS-CX allow unauthenticated attackers on the adjacent network segment to crash the device by sending specially crafted packets, causing a denial-of-service disruption to normal OS operation. This vulnerability is one of several disclosed together under HPE advisory hpesbnw05134, affecting the CX series switch OS. No public exploit code or active exploitation has been identified at time of analysis.

Crash of the gvfsd-mtp daemon in gvfs is achievable by anyone who can connect a malicious USB MTP device to a target Linux system. The do_read() function in gvfsbackendmtp.c accepts the byte-count reported by the device at face value and passes it uncapped to memcpy(), causing an out-of-bounds read that triggers a segmentation fault when the device deliberately over-reports its response length. The vulnerability affects gvfs as packaged across Red Hat Enterprise Linux 6 through 10; no public exploit has been identified and exploitation is constrained by physical proximity.

Heap buffer overflow in the gvfs AFP (Apple Filing Protocol) backend allows a malicious AFP server to crash the gvfsd-afp process on connecting clients, resulting in denial of service. When a client mounts an AFP share, the DSI (Data Stream Interface) read path accepts a server-supplied length value without validating it against the pre-allocated reply buffer, enabling the server to write past buffer boundaries. This affects RHEL 6 through 10 and requires user interaction to mount a malicious share; no public exploit or CISA KEV listing has been identified at time of analysis.

Denial-of-service vulnerability in Rockwell Automation FactoryTalk® Historian Machine Edition allows a network-adjacent, high-privilege authenticated attacker to crash the device by sending crafted requests to its web interface, triggering a stack-based buffer overflow (CWE-121) that renders the historian unresponsive. All versions appear affected per the CPE wildcard, and the impact is limited to availability - no confidentiality or integrity compromise is possible. No public exploit code exists and the vulnerability is not listed in CISA KEV at time of analysis.

Remote code execution in Rockwell Automation FactoryTalk® Historian Machine Edition allows an adjacent-network attacker holding low-privilege credentials to fully compromise the historian host via a memory corruption flaw. The root cause is CWE-787 (Out-of-bounds Write), consistent with a buffer overflow in the historian's network-facing service, yielding full confidentiality, integrity, and availability impact on the affected device. No active exploitation has been confirmed in CISA KEV and no public proof-of-concept has been identified at the time of analysis.

Incorrect boundary conditions in Firefox's CSS Grid layout engine expose limited browser process memory to remote attackers who deliver crafted web content. All Firefox versions prior to 155 are affected; exploitation requires a victim to visit or render attacker-controlled content, producing low-level confidentiality disclosure with no integrity or availability impact. Mozilla issued the fix in Firefox 155 via MFSA2026-82; no public exploit code and no CISA KEV listing have been identified at time of analysis.

Heap metadata corruption in the popt command-line parsing library's `poptConfigFileToString` function can be triggered by a local, low-privileged attacker who supplies specially crafted configuration file content to an affected host. Successful exploitation causes the consuming process to crash, resulting in denial of service. CVSS 2.5 (Low) reflects the strictly local, high-complexity exploitation path with no confidentiality or integrity impact; no public exploit code or active exploitation (CISA KEV) is known at time of analysis.

Buffer overflow in FFmpeg v.7.0 and later within the libavformat/iamf_writer.c component allows an attacker to crash the process by supplying a crafted IAMF (Immersive Audio Model and Formats) file, resulting in a denial of service. The flaw, rooted in improper bounds checking during extradata padding operations (CWE-120), affects the IAMF audio encoder/writer path introduced alongside IAMF support in FFmpeg 7.0. A patch commit (8439e020) and upstream pull request #22988 are available, though a tagged release incorporating the fix has not been independently confirmed. No public exploit code has been identified and EPSS of 0.17% (7th percentile) reflects low exploitation probability at this time.

Off-by-one out-of-bounds write in Zephyr RTOS's LwM2M JSON formatter allows a LwM2M server peer to corrupt one byte of memory adjacent to a STRING resource buffer by sending a WRITE payload whose string length exactly equals the buffer size. Affected devices are Zephyr LwM2M clients with the JSON content formatter enabled (cpe:2.3:a:zephyrproject:zephyr:*:*:*:*:*:*:*:*). The overflow deterministically writes a NUL byte past the buffer boundary, enabling potential crash or silent data corruption depending on memory layout; no public exploit has been identified at time of analysis, but the fix commit and security advisory are publicly visible.

Heap-based buffer overflow in FLVMeta versions up to 1.2.2 allows remote attackers to trigger a crash via a crafted FLV file that manipulates the length argument passed to amf_string_new() during AMF string processing. A public proof-of-concept exists, but the CVSS 4.0 score of 2.1 reflects the strictly limited impact - low-severity availability disruption only, with no confidentiality or integrity consequences. The project maintainer has publicly disputed that the reported bugs are practically exploitable, further tempering the real-world risk assessment.

Race condition in Zephyr RTOS's I3C IBI workqueue subsystem allows a physically-present attacker controlling an I3C peripheral to corrupt kernel memory structures, most realistically causing denial of service. The unsynchronized free-list (sys_slist_t) shared between ISR context and the workqueue handler thread creates a window where a work node can be handed to two consumers simultaneously or a garbage pointer can be dereferenced in a memcpy, producing an out-of-bounds write. No public exploit code has been identified and no CISA KEV listing exists; the patch wrapping all free-list operations in a k_spinlock is confirmed in upstream commit e87e7e2.

Use-after-free in Valkey 9.1.0's blocked-on-keys subsystem allows a high-privilege authenticated attacker to crash the server by triggering a reentrant client teardown during ready-key iteration. When a key becomes ready and multiple clients are unblocked simultaneously, a reply callback for the first-served client can invoke CLIENT KILL on a successor client, freeing that client's memory while the iterator in `handleClientsBlockedOnKey` still holds a reference to it. A public proof-of-concept exists; the vulnerability is not in CISA KEV, and the high complexity and privilege requirements substantially limit practical exploitation.

Out-of-bounds read in Systerel S2OPC versions up to 1.7.3 allows an authenticated remote attacker to partially disrupt availability of an OPC UA server through crafted string array range write operations. The vulnerable function set_range_matrix_on_string_array in sopc_builtintypes.c fails to enforce memory boundaries during range matrix processing on string arrays. No vendor patch has been released; the project was notified via issue report but has not responded. No public exploit or CISA KEV listing exists at time of analysis.

Out-of-bounds read in Kamailio's IMS AVP handler exposes telecom infrastructure to remote memory disclosure and potential service disruption. The `get_4bytes()` function across five IMS modules - `ims_auth`, `ims_icscf`, `ims_registrar_scscf`, `ims_charging`, and `ims_ocs` - fails to validate that Diameter AVP data buffers contain at least 4 bytes before reading, allowing a remote attacker with low-privilege Diameter interface access to trigger the flaw via a crafted malformed message. Publicly available exploit code exists per GitHub issue #4816, and the vendor has confirmed that the 5.5.0 branch is end-of-life and will receive no further security updates, compounding risk for operators running that release.

Memory corruption in LatencyUtils 2.0.3 and earlier allows a local low-privileged user to cause availability disruption by manipulating input to the `LatencyStats.recordDetectedPause` function in the PauseDetector component. A public proof-of-concept has been disclosed via a GitHub issue, and the project maintainers have not responded to the disclosure, leaving the library without an upstream fix. The CVSS 4.0 score of 1.9 and availability-only impact (VA:L) reflect a tightly scoped, low-severity flaw with no confidentiality or integrity consequences and no confirmed active exploitation.

Out-of-bounds read in Exiv2's CrwMap::decodeBasic() function affects versions prior to 0.28.9, impacting applications that use the library to parse Canon CRW image metadata. Classified as low severity by the Exiv2 maintainers, this flaw has only been reproduced using a dedicated fuzz target - the standard exiv2 command-line tool cannot trigger it under normal operation. No public exploit exists and no active exploitation has been identified, making real-world impact minimal despite the memory safety class of defect.

Out-of-bounds read in Exiv2's RemoteIo::Impl::populateBlocks() function affects versions prior to 0.28.9, exploitable only when the tool is invoked against a remote URL rather than a local file. An attacker controlling a remote server can serve a crafted image to trigger the read when a user runs exiv2 against that URL, potentially causing information disclosure or a crash. No public exploit code or active exploitation has been identified; the vulnerability is rated low severity and fixed in the 0.28.9 release.

Heap-based buffer overflow in Open Asset Import Library (Assimp) through version 6.0.2 allows a local low-privilege attacker to corrupt heap memory by supplying a crafted MD5 mesh file. The root cause, confirmed by the patch diff, is that `MakeDataUnique()` pre-allocated only `faces × 3` vertex slots when the output can grow to `numOrigVerts + faces × 3` - meshes declaring more original vertices than face corners overflow the allocation. No public exploit exists and the vulnerability is absent from CISA KEV; impact is locally bounded with a CVSS 4.0 base score of 4.8.

Memory corruption in Open5GS versions up to 2.7.7 allows a low-privileged remote attacker to crash the Access and Mobility Management Function (AMF) by sending a crafted UE MM context transfer request with an oversized allowedNssai array. The vulnerable function amf_namf_comm_decode_ue_mm_context_list in src/amf/namf-handler.c lacked both null-pointer checks and array bounds enforcement on mmContextList[*].allowedNssai entries, relying instead on ogs_assert() which terminates the process rather than handling the error gracefully. No public exploit is confirmed in CISA KEV, but the CVSS 4.0 vector carries E:P indicating publicly disclosed exploit code exists.

Buffer overflow in NASA Core Flight System (cFS) up to version 7.0.1 allows adjacent-network attackers to corrupt memory through the SBN TCP Module's OS_read function. The MsgSz argument read from incoming TCP messages is not bounds-checked before use, enabling a crafted packet to overflow an internal buffer. No vendor patch exists - NASA did not respond to the coordinated disclosure - and no public exploit or KEV listing has been identified at time of analysis.

Stack-based buffer overflow in NASA Trick 19.6.0's JSONVariableServer component exposes the simulation framework to remote exploitation via crafted TCP requests, with no patch available and no vendor response to disclosure. The vulnerability resides in JSONVariableServerThread::parse_request, where user-supplied network data is processed through an unbounded sscanf call without bounds checking on a stack-allocated buffer. Unauthenticated remote attackers reaching the TCP socket can trigger this overflow, potentially crashing the simulation process or achieving code execution on the host.

Integer overflow in U-Boot's SquashFS filesystem parser (fs/squashfs/sqfs.c, function sqfs_concat_tokens) enables heap under-allocation followed by a heap-based buffer overflow via strcpy() when a crafted SquashFS image with manipulated token lists is processed during boot. All U-Boot releases through v2026.01-rc4 are affected. No public exploit code and no CISA KEV listing have been identified at time of analysis; a confirmed fix is available in v2026.04-rc1 (commit adccdb2).

Heap buffer overflow in U-Boot's do_mv shell command (fs/fs.c) affects all versions through v2026.01-rc4 and allows an attacker with U-Boot shell access to corrupt pre-boot heap memory and potentially execute arbitrary code before any OS-level security controls activate. The flaw is rooted in missing bounds checks during string length addition when constructing move-command path arguments: integer overflow causes a heap under-allocation, which is then overflowed by an unchecked strcpy() call. No public exploit code or active exploitation has been identified, and the upstream fix is confirmed in v2026.04-rc1.

Stack-based buffer overflow in MongoDB BI Connector ODBC Driver allows an authenticated user who can influence the numeric argument of a SQL LIMIT clause to crash the hosting application process or corrupt adjacent stack memory. The vulnerability is gated by a non-default driver configuration: the prefetch feature must be explicitly enabled, narrowing real-world exposure to that subset of deployments. No public exploit code has been identified and the issue is absent from CISA's Known Exploited Vulnerabilities catalog; however, where conditions are met, availability impact is rated High by the reporter.

Memory corruption in the TP-Link TL-WR841N v14 embedded HTTP service allows an unauthenticated attacker with adjacent network access to trigger a buffer overflow by submitting a crafted multipart/form-data request containing an oversized boundary parameter. The confirmed impact is limited to undefined application behavior resulting from process memory corruption; TP-Link's own advisory and the CVE description both explicitly state that arbitrary code execution, information disclosure, and denial-of-service have not been demonstrated, despite those labels appearing in the CVE tags. No public exploit code has been identified at time of analysis, and this CVE is not listed in CISA's Known Exploited Vulnerabilities catalog.

Heap out-of-bounds read and stack out-of-bounds access in GIMP's file-psd plugin allow an attacker to crash the application or disclose limited memory contents by convincing a user to open a specially crafted PSD image. The flaw affects GIMP as shipped in Red Hat Enterprise Linux 6 through 9 and stems from missing validation of the channel-count parameter during PSD parsing. No active exploitation has been identified; the attack requires local file delivery and deliberate user interaction, significantly constraining real-world risk.

Stack-based buffer overflow vulnerabilities in the PLANET GS-4210-16P2S managed PoE switch web management interface allow a remote authenticated attacker to crash the CGI process and disrupt web-based device administration. Three separate POST parameter handlers in /cgi-bin/dispatcher.cgi - processing usrPass, enbPass, usrName, and usrPass - copy user-supplied input into fixed-size stack buffers without bounds validation, triggering CWE-121 conditions. No public exploit code has been identified at time of analysis, and the vulnerability is not listed in the CISA KEV catalog; impact is confined to denial of service against the web management plane.

Stack-based buffer overflow and null pointer dereference vulnerabilities in the PLANET GS-4210-16P2S managed PoE switch firmware (all versions before 3.441b260626) allow a remote authenticated attacker to crash the CGI-based web management service by submitting crafted POST requests to RADIUS server configuration handlers. The `web_radiusSrv*_post` handler family in `/cgi-bin/dispatcher.cgi` copies user-controlled parameters into fixed-size stack buffers without bounds checking and dereferences `radName`/`radIp` pointers without confirming their presence. No public exploit code exists and the vulnerability is not listed in the CISA KEV catalog at time of analysis; a vendor-released firmware patch is available.

Stack-based buffer overflows in PLANET GS-4210-16P2S managed switch firmware before 3.441b260626 allow a remote administrator to crash the web management interface by sending crafted HTTP POST requests to /cgi-bin/dispatcher.cgi. More than 30 named CGI handler functions - spanning VLAN, SNMP v3, ACL, AAA, DHCP, STP, PoE, and cable diagnostics subsystems - copy attacker-supplied POST parameters into fixed-size stack buffers without length validation. The confirmed impact is denial of service of the web management service; the vulnerability is not listed in CISA KEV and no public exploit code has been identified at time of analysis, though a vendor patch is available.

Heap out-of-bounds read in GIMP's file-pvr plugin exposes users of Red Hat Enterprise Linux 6 through 9 to application crashes and limited heap memory disclosure when opening a crafted PVR image file. The VQ (Vector Quantization) compressed-texture decoder omits mandatory bounds validation, allowing a malicious image to read beyond allocated heap regions. Impact is primarily denial of service via crash, with a secondary low-severity information disclosure path; no public exploit code and no CISA KEV listing are associated with this issue at time of analysis.

Heap out-of-bounds read in GIMP's file-ico plugin allows a local attacker to crash GIMP or leak limited heap memory contents by delivering a specially crafted ICO image file to a victim user. Affected packages span Red Hat Enterprise Linux 6 through 9. No public exploit code or CISA KEV listing exists at time of analysis, but the user-interaction requirement (opening a malicious file) is the primary delivery mechanism and a realistic social-engineering target in image-processing workflows.

Out-of-bounds memory write in libsolv crashes RPM-based package manager tools (dnf, yum, zypper) when processing a corrupted or specially crafted .solv repository cache file. The root cause is an absent array-index bounds check on directory-id values read from compressed filelist data during .solv file rewriting - but because the out-of-bounds write stores only a fixed, non-attacker-controlled value, the vendor explicitly rules out arbitrary code execution, confining impact to denial of service. No public exploit has been identified and no active exploitation is confirmed; note that the 'RCE' metadata tag conflicts directly with the description and appears to be erroneous.

Heap out-of-bounds read in GIMP's file-iff plugin allows a local attacker to crash the application or disclose limited heap memory contents by delivering a specially crafted IFF/ILBM image file. The flaw stems from missing validation of HAM row size when the number of color planes (nPlanes) is zero, causing a size mismatch that bypasses memory bounds checking. No active exploitation has been confirmed (not in CISA KEV) and no public exploit code has been identified at time of analysis.

Stack buffer overflow in Ericsson CodeChecker through 6.28.2 results from an off-by-two error introduced during the CVE-2025-40843 remediation: the safe_strcpy() call at ldlogger-tool-gcc.c:129 receives destination pointer fullPath + 2 but the full PATH_MAX (4096) as its size bound, causing a deterministic 2-byte overwrite past the 4094 bytes that actually remain on every invocation of the GCC logger component. The overflow is unconditional and input-independent; a proof-of-concept is noted via the CVSS 4.0 E:P supplemental metric, though no active exploitation (CISA KEV) has been confirmed.

Authenticated availability disruption in OX Dovecot Pro and OX Dovecot CE allows any valid credential holder to crash the imap-hibernate service by opening many concurrent connections and sending malformed IMAP commands, triggering an intermittent out-of-bounds read (CWE-125). The crash terminates the imap-hibernate process, dropping hibernated IMAP sessions for co-hosted users and degrading service availability. No public exploit code is known and the vulnerability is not listed in CISA KEV; however, the low attack complexity means any authenticated user on a shared mail platform could attempt disruption.

Memory corruption via out-of-bounds write in OX Dovecot Pro and OX Dovecot CE's ManageSieve service allows authenticated network attackers to crash the ManageSieve process, denying all users the ability to remotely manage Sieve email filter scripts. The root cause is improper handling of extreme numeric literals during Sieve script compilation. The vendor notes that remote code execution may theoretically be achievable via the same memory corruption primitive, though no public exploit demonstrates this; no public exploits are known and the vulnerability is not listed in CISA KEV.

Heap buffer overflow in FFmpeg's TDSC screen-capture codec decoder exposes Red Hat Enterprise Linux AI 3 and OpenShift AI deployments to denial of service or potential arbitrary code execution. The tdsc_load_cursor() function performs erroneous stride-based pointer arithmetic in two locations when rendering cursor data from crafted TDSC video, writing past a heap-allocated buffer boundary. No public exploit or active exploitation has been confirmed at time of analysis, and the upstream fix is available as FFmpeg commit 242ff799c.

In the Linux kernel, the following vulnerability has been resolved: drm/vmwgfx: enforce cursor size limits for MOB cursors vmw_cursor_plane_atomic_check() bounds cursor width and height only on the legacy update path; the SVGA_CAP2_CURSOR_MOB path -- the default on modern hosts -- accepts any size. When the requested size exceeds SVGA_REG_CURSOR_MAX_DIMENSION or SVGA_REG_MOB_MAX_SIZE, vmw_cursor_mob_get() returns -EINVAL and leaves vps->cursor.mob NULL. Its return value is then discarded in vmw_cursor_plane_prepare_fb(), so the subsequent vmw_cursor_update_mob() calls vmw_bo_map_and_cache(NULL) and oopses inside vmw_bo_map_and_cache_size() on the tbo.base.size load. Reachable from any DRM master via DRM_IOCTL_MODE_CURSOR2 with a sufficiently large width or height (e.g. cursor_max_dim + 1). Reject oversized cursors in atomic_check for both MOB-backed cursor update types. The MOB byte-size limit only applies to the SVGA_CAP2_CURSOR_MOB path (vmw_cursor_mob_size() returns 0 for GB_ONLY); compute the required MOB size in 64-bit to avoid overflow when very large dimensions are requested. In prepare_fb only call vmw_cursor_mob_get()/_map() for VMW_CURSOR_UPDATE_MOB -- the GB_ONLY path uses bo->map.virtual directly and would otherwise be silently downgraded to NONE on hosts without SVGA_CAP2_CURSOR_MOB (where vmw_cursor_mob_get() always returns -EINVAL). Degrade the update to NONE if vmw_cursor_mob_get() or vmw_cursor_mob_map() fails so the update path does not run with a NULL backing MOB.

In the Linux kernel, the following vulnerability has been resolved: cxl/fwctl: Fix __fortify_panic Fix a runtime assertion in cxlctl_get_supported_features(). Fortify complains that it is potentially overflowing the entries array per __counted_by_le(num_entries). Quiet the false positive by initializing @num_entries earlier. memcpy: detected buffer overflow: 48 byte write of buffer size 0 WARNING: lib/string_helpers.c:1036 at __fortify_report+0x4d/0xa0, CPU#7: fwctl/1398 RIP: 0010:__fortify_report+0x50/0xa0 Call Trace: __fortify_panic+0xd/0xf cxlctl_get_supported_features.cold+0x23/0x35 [cxl_core]

In the Linux kernel, the following vulnerability has been resolved: cxl/test: Fix __fortify_panic Fix a runtime assertion in setup_xor_mapping(). Fortify complains that it is potentially overflowing the xormaps array per __counted_by(nr_maps). Quiet the false positive by initializing @nr_maps earlier. memcpy: detected buffer overflow: 32 byte write of buffer size 0 WARNING: lib/string_helpers.c:1036 at __fortify_report+0x4d/0xa0, CPU#8: modprobe/2728 Call Trace: __fortify_panic+0xd/0xf setup_xor_mapping+0x6c/0xa0 [cxl_translate] [ dj: Fixed up @nr_entries to @nr_maps in commit log. ]

In the Linux kernel, the following vulnerability has been resolved: octeontx2-af: npc: Fix size of entry2cntr_map KASAN prints below splat. This is caused by allocating counter for reserved mcam entry for cpt 2nd pass entry. But mcam->entry2cntr_map is not allocated for reserved entries. BUG: KASAN: slab-out-of-bounds in npc_map_mcam_entry_and_cntr+0xb0/0x1a0 Write of size 2 at addr ffff0001033e7ffe by task kworker/0:1/14 CPU: 0 PID: 14 Comm: kworker/0:1 Not tainted 6.1.67 #1 Hardware name: Marvell CN106XX board (DT) Workqueue: events work_for_cpu_fn Call trace: dump_backtrace.part.0+0xe4/0xf0 show_stack+0x18/0x30 dump_stack_lvl+0x88/0xb4 print_report+0x154/0x458 kasan_report+0xb8/0x194 __asan_store2+0x7c/0xa0 npc_map_mcam_entry_and_cntr+0xb0/0x1a0 rvu_mbox_handler_npc_mcam_write_entry+0x268/0x280 npc_install_flow+0x840/0xfe0 rvu_npc_install_cpt_pass2_entry+0x138/0x190 rvu_nix_init+0x148c/0x2880 rvu_probe+0x1800/0x30b0 local_pci_probe+0x78/0xe0 work_for_cpu_fn+0x30/0x50 process_one_work+0x4cc/0x97c worker_thread+0x360/0x630 kthread+0x1a0/0x1b0 ret_from_fork+0x10/0x20

In the Linux kernel, the following vulnerability has been resolved: MIPS: mm: Fix out-of-bounds write in maar_res_walk() maar_res_walk() uses wi->num_cfg as the index into the fixed-size wi->cfg array, but checks whether the array is full only after it has filled the selected entry. If walk_system_ram_range() reports more than 16 memory ranges, the overflow call writes one struct maar_config past the end of the array before WARN_ON() prevents num_cfg from advancing. Move the full-array check before taking the array slot and return non-zero when the scratch array is full, so walk_system_ram_range() terminates the walk instead of invoking the callback for further ranges.

In the Linux kernel, the following vulnerability has been resolved: Input: ims-pcu - add response length checks The driver processes response data from device buffers without verifying that the device actually sent enough data. This can lead to out-of-bounds reads or processing stale data. Add checks for the expected response length before accessing the buffers.

Heap-based buffer overflow in WatchGuard Fireware OS's iked process allows an authenticated administrator to crash the IKE daemon by saving a specially crafted configuration, resulting in denial of service and disruption of IPsec VPN services. The CVSS 4.0 score of 6.9 with PR:H reflects the high-privilege requirement, meaning exploitation is gated behind valid administrator credentials. No active exploitation has been identified (not in CISA KEV) and no public exploit code is known at time of analysis.

Heap buffer overflow in GNU C Library (glibc) 2.45 and earlier triggers when an effectively empty string is passed via the ,ccs= mode-argument extension to fopen(), enabling limited heap memory corruption in applications that expose this code path to externally controlled input. The vendor-assigned CVSS 3.1 score of 4.9 (Medium) with a local attack vector and high complexity accurately reflects the narrow preconditions required. No public exploit code has been identified at time of analysis, and the glibc advisory itself explicitly notes this vulnerable usage pattern is absent from mainstream GNU/Linux applications.

Out-of-bounds write in gdk-pixbuf's JPEG decoder affects any application on Red Hat Enterprise Linux 6 through 10 that renders JPEG images containing chunked ICC profile markers. When ICC profile parsing encounters an error, stale size metadata persists after the profile buffer is freed; a subsequent allocation within the same decode operation then writes beyond buffer boundaries, crashing the application. No active exploitation has been identified (no CISA KEV listing), and no public exploit code is known at time of analysis, but the availability-only impact and user-interaction requirement make this a realistic denial-of-service risk in desktop and server image-processing contexts.

Out-of-bounds read in Adobe DNG SDK 1.7.1 2502 and earlier exposes sensitive memory contents when a victim opens a maliciously crafted DNG file. The flaw resides in the SDK's file parser and affects any application built on top of the DNG SDK that processes untrusted DNG inputs. No public exploit has been identified at time of analysis, and the vulnerability is not listed in CISA KEV, limiting current urgency despite the High confidentiality impact.

Out-of-bounds write in Adobe DNG Software Development Kit versions 1.7.1 2502 and earlier allows memory corruption when a victim opens a maliciously crafted DNG file, causing the consuming application to crash or become unresponsive. The flaw (CWE-787) is constrained to a local, user-interaction-dependent attack path, limiting automated exploitation. No public exploit code has been identified at time of analysis and no CISA KEV listing exists.

Out-of-bounds heap read in the DHCPv4 packet capture code of wicked (all versions through 0.6.80) allows an unauthenticated attacker on the same Layer-2 network segment to send a crafted DHCP/UDP packet that causes wickedd-dhcp4 to read up to 68 bytes past the end of its 1500-byte receive buffer. Adjacent heap contents - including allocator metadata and pointer values - may be parsed as DHCP options and stored into lease fields, constituting limited information disclosure with a secondary low-integrity configuration impact. No public exploit code and no CISA KEV listing have been identified at time of analysis; an upstream fix exists in GitHub PR #1079, though a tagged release version has not been independently confirmed.

Integer underflow in the wicked DHCPv4 packet capture daemon (wickedd-dhcp4) on SUSE Linux systems allows an unauthenticated adjacent-network attacker to crash the daemon by sending a malformed DHCPv4 packet with an IP total length field smaller than the IP header length. The missing bounds check in ni_capture_inspect_udp_header() triggers an out-of-bounds read, causing a process crash depending on memory layout - resulting in a denial of service to network configuration. No information disclosure has been demonstrated, and no public exploit or CISA KEV listing exists at time of analysis.

Out-of-bounds read and write primitives in the hank-ai Darknet neural network framework (all versions through 6.0) are triggered by parsing a crafted configuration file, yielding a reliable crash and a single fixed-byte heap write at an attacker-controlled offset. The parser in src-lib/darknet_cfg.cpp consumes the `from` field of shortcut, scale_channels, and sam sections-and the `layers` field of route sections-as a direct array index into the layer heap allocation without validating it against the array's declared size. No public exploit has been identified at time of analysis, and the vulnerability is not listed in CISA KEV.

Out-of-bounds write in the SmiFlash SMM (System Management Mode) module affects ASUS FA507NU and FA507NV laptop BIOS firmware across all versions. A local administrator can dispatch a crafted Software SMI (SW SMI) request with an oversized length value, causing the SmiFlash handler to write beyond its allocated SMRAM buffer boundary, resulting in a system crash (BSOD) or irreversible BIOS flash corruption. The affected models are designated 'unsupported-when-assigned,' meaning these products were already outside active vendor support when this CVE was issued; no public exploit has been identified at time of analysis.

Out-of-bounds stack array access in the tdelete function affects the GNU C Library (glibc) as reported by Ubuntu. The tdelete function, part of POSIX binary tree search routines (tsearch family), contains a memory safety flaw where a stack-allocated array can be accessed beyond its bounds during tree node deletion. Exploitation depends on an application passing attacker-influenced input to tdelete, and no public exploit code or CISA KEV listing has been identified at time of analysis.

HCL BigFix Quantum Risk Analyzer's binary executable lacks multiple industry-standard binary hardening protections, leaving it susceptible to a stack-based buffer overflow (CWE-121). The vulnerability requires local access with high privileges and high attack complexity, yielding a CVSS score of 3.9 (Low). No public exploit code has been identified at the time of analysis, and the combination of local-only access, elevated privilege requirements, and high exploitation complexity significantly constrains real-world risk despite the theoretical severity of a stack overflow primitive.

Out-of-bounds read and in-object write in the Zephyr RTOS IEEE 1588 PTP stack affects all Zephyr versions with CONFIG_PTP enabled, exploitable by any unauthenticated attacker with Layer 2 adjacency. The PTP_MGMT_TIME management-id case in tlv_mgmt_post_recv() omits the length guard present for every sibling case, allowing a crafted short TLV to drive an 8-byte over-read and bounded write-back within the same struct ptp_msg allocation. No public exploit has been identified at time of analysis, and impact is constrained to minor adjacent-memory disclosure and corruption of the device's parsed management timestamp - no crash or past-allocation corruption is reachable.

Out-of-bounds read in Zephyr RTOS's LoRaWAN TS004 Fragmented Data Block Transport handler allows a party holding valid device session keys to trigger a bounded overread of up to 232 bytes beyond the 255-byte `MacCtx.RxPayload` static buffer, with overread data silently written into decoder buffers and the FUOTA flash image rather than returned to the attacker. The vulnerable function `frag_transport_package_callback()` in `subsys/lorawan/services/frag_transport.c` casts `rx_buf + rx_pos` to fixed-size structs and passes the pointer to the fragment decoder without first verifying that sufficient payload bytes remain. No public exploit is identified and this CVE is not listed in the CISA KEV catalog; CVSS 3.1 rates the vulnerability 3.1 (Low), reflecting the high-privilege access barrier and constrained, non-disclosing impact.

Out-of-bounds read in Zephyr RTOS's LoRaWAN clock synchronization service allows a malicious or compromised LoRaWAN application server to corrupt the target device's internal time offset by sending a crafted short AppTimeAns frame. The over-read is bounded by a 255-byte static backing buffer so no crash occurs, and the stray bytes are never transmitted back to the attacker, eliminating any information disclosure. Impact is confined to a minor, device-local clock integrity issue - the device's time estimate may silently diverge if a garbage time-correction value happens to match the pending token. No public exploit identified at time of analysis, and no CISA KEV listing.

CorvusSKK, a Windows-based SKK Japanese input method editor by Sasaki Nobuyuki, contains a signed 32-bit integer overflow (CWE-190) in its tar-format dictionary file size parsing, which can allow malicious data to be written to a dictionary file when a user opens a crafted input. All versions prior to 3.3.4 are affected; version 3.3.4 was released to address this flaw alongside a separate Lua-related vulnerability. No public exploit has been identified at time of analysis and the vulnerability is not listed in the CISA KEV catalog, indicating no confirmed active exploitation.

SMM buffer overflow in InsydeH2O's IHISI command handler for FMTS command 0x32 allows a physically-present, highly-privileged attacker to trigger out-of-bounds writes in System Management Mode firmware. The vulnerable function FMTSWriteUseIntelLib performs read and write operations without validating buffer sizes, enabling potential firmware-level memory corruption in Ring -2 execution context. No active exploitation has been identified and no public exploit exists; the combination of physical access, high privilege, high attack complexity, and required user interaction makes real-world risk extremely low despite the sensitive SMM execution environment.

Heap-based buffer overflow in Das U-Boot (the Denx universal embedded bootloader) before 2026.04 allows an attacker who controls an ext4 filesystem image to trigger an integer overflow in the ext4fs_get_bgdtable() block-group descriptor table size calculation, resulting in an undersized allocation that is later overrun by memcpy(). Successful exploitation can lead to arbitrary code execution in the highly privileged pre-OS boot context, denial of service, or memory corruption. No public exploit identified at time of analysis (a referenced GitHub gist may contain a proof-of-concept but is unverified), EPSS is low at 0.18% (7th percentile), and it is not listed in CISA KEV.

Stack-buffer-underflow in aria2's IOFile::getLine() function through version 1.37.0 allows a crash when the function processes an empty line, causing the expression buf[len - 1] to evaluate as buf[-1] and access memory before the buffer's start. The impact is limited to a local availability disruption - no confidentiality or integrity exposure has been identified - and the EPSS score of 0.15% (5th percentile) confirms negligible real-world exploitation interest. No public exploit code exists and this vulnerability is absent from CISA KEV, placing it firmly in the low-priority tier for most security teams.

Denial-of-service via heap buffer overflow in SQLite's optional SQLAR extension affects any application that compiles or loads ext/misc/sqlar.c and processes attacker-controlled archive entries. The root cause is an integer-width mismatch in sqlarUncompressFunc(): a 64-bit decompressed-size argument is silently truncated to 32 bits by sqlite3_value_int(), causing sqlite3_malloc() to allocate an undersized heap buffer before zlib's uncompress() writes the full payload into it (CWE-122). The vulnerability is absent from CISA KEV, EPSS stands at 0.28% (20th percentile), the upstream fix is confirmed at Git commit 169f68ed, and a third-party GitHub repository referencing the CVE suggests proof-of-concept code is publicly accessible.

Heap out-of-bounds read in GIMP's file-xwd plugin allows a crafted XWD image file to trigger an application crash or leak partial heap memory contents into the rendered image output. Users running GIMP on Red Hat Enterprise Linux 6 through 9 are affected when they open a malicious XWD file; the plugin fails to cross-validate image width and bytes-per-line parameters against the allocated buffer size, permitting an out-of-bounds read. No public exploit is identified at time of analysis and the vulnerability is not listed in CISA KEV, placing real-world urgency in the low-to-moderate range.

CPU exhaustion in libsoup's HTTP Range header processing affects any SoupServer deployment carrying the CVE-2025-32907 fix but not MR !550. Supplying a Range header with approximately 25,000 repeated satisfiable byte ranges triggers an O(N²) coalescing loop - each g_array_remove_index() call induces an O(N) memmove on GArray - blocking the server's event loop for roughly 90 ms per request from a single unauthenticated client. The description explicitly rules out memory corruption and information disclosure; this is a pure availability issue with no public exploit identified at time of analysis.

Out-of-bounds read in Chrome's Skia graphics library exposes sandbox memory to attackers who have already compromised the renderer process. Exploitation requires a prior renderer compromise plus user interaction with a crafted HTML page, making this a chained second-stage information disclosure rather than a standalone critical flaw. Chromium's own severity rating is Low; no public exploit or CISA KEV listing exists, and EPSS sits at just 0.17%.

Out-of-bounds memory read in Chrome's Tint graphics library on Android exposes sandbox-contained memory to remote attackers. Google Chrome for Android prior to 152.0.7977.65 is affected; an attacker who lures a user to a crafted HTML page can trigger the flaw and read arbitrary memory within the Chrome sandbox. No active exploitation has been identified - Chromium's own security team rates this as Low severity, consistent with the 0.17% EPSS score and the absence of a CISA KEV listing.

Out-of-bounds memory read in Google Chrome's FileSystem component (versions prior to 152.0.7977.65) allows a remote unauthenticated attacker to leak memory contents from outside the browser's renderer sandbox by directing a victim to a specially crafted HTML page via social engineering. Exploitation is contingent on user interaction, limiting opportunistic or automated mass exploitation. No public exploit code exists and EPSS sits at 0.17% (7th percentile), though the Confidentiality impact is rated High, meaning successful exploitation could expose sensitive data from browser memory. Vendor-released patch is available in Chrome 152.0.7977.65.

Out-of-bounds read in Google Chrome's Media component prior to version 152.0.7977.65 enables a remote attacker who has already compromised the renderer process to leak memory contents across the Chrome sandbox boundary via a crafted HTML page. The attack is a chained, second-stage capability - it requires prior renderer code execution and user interaction - not a standalone entry point. No public exploit has been identified at time of analysis, and the EPSS score of 0.29% (21st percentile) reflects low observed exploitation likelihood.