CVEs of 2026.
Every CVE published in 2026 that we hold, ranked by exploit odds.
-
CVE-2026-63202
netty-incubator-codec-ohttp implements Oblivious HTTP (OHTTP) gateway and client functionality using Netty. Prior to 0.0.23.Final, BinaryHttpParser.readFieldSection in codec-bhttp/src/main/java/io/netty/incubator/codec/bhttp/BinaryHttpParser.java loops while fieldSectionLength is nonzero and relies on disabled-by-default assertions to require BinaryHttpParser.readFieldLine to return a value and consume bytes. A corrupt field line can consume more bytes than its understated declared section length and drive the counter negative, or a truncated field line can return without consuming data and leave the parser in the same state. An unauthenticated OHTTP peer can encrypt such Binary HTTP content under the gateway public configuration, reach OHttpRequestResponseContext.ContentDecoder.decodeChunk, and permanently pin a Netty event-loop thread at full CPU before application business logic. Multiple requests can exhaust the event-loop group and deny service until restart. This issue is fixed in version 0.0.23.Final.
-
CVE-2026-61827
netty-incubator-codec-ohttp implements Oblivious HTTP (OHTTP) gateway and client functionality using Netty. Prior to 0.0.23.Final, BinaryHttpParser does not enforce limits on encoded variable-length fields controlled by a remote peer. A remote peer can cause the decoder to buffer data indefinitely until memory is exhausted, causing denial of service. This issue is fixed in version 0.0.23.Final.
-
CVE-2026-133217
A weakness has been identified in OSGeo shapelib up to 1.6.3. This vulnerability affects the function DBFCloneEmpty of the file dbfopen.c. This manipulation causes null pointer dereference. It is possible to launch the attack on the local host. Patch name: f79050a900d2a05529501b2b714b8f89067db071. It is recommended to apply a patch to fix this issue.
-
CVE-2026-118233
Allocation of resources without limits or throttling vulnerability in Apache Tomcat during HTTP upgrade in some circumstances lead to possible DoS. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.27, from 10.1.0-M1 through 10.1.61, from 9.0.41 through 9.0.123. The following versions were EOL at the time the CVE was created but are known to be affected: from 8.5.61 through 8.5.100. Other unsupported versions may also be affected. Users are recommended to upgrade to version 11.0.29, 10.1.62 or 9.0.124, which fix the issue. Note: Tomcat 11.0.28 was not released
-
CVE-2026-118436
Allocation of resources without limits or throttling vulnerability in Apache Tomcat in the CGI Servlet when handling POST bodies can result in a DoS for some inputs. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.27, from 10.1.0-M1 through 10.1.61, from 9.0.0.M1 through 9.0.123. The following versions were EOL at the time the CVE was created but are known to be affected: from 8.5.0 through 8.5.100, from 7.0.0 through 7.0.109. Other unsupported versions may also be affected. Users are recommended to upgrade to version 11.0.29, 10.1.62 or 9.0.124, which fix the issue. Note: Tomcat 11.0.28 was not released
-
CVE-2026-116830
Missing release of resource after effective lifetime vulnerability in Apache Tomcat's HTTP/2 initial connection handling leads to DoS. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.27, from 10.1.0-M1 through 10.1.61, from 9.0.20 through 9.0.123. Users are recommended to upgrade to version 11.0.29, 10.1.62 or 9.0.124, which fix the issue. Note: Tomcat 11.0.28 was not released
-
CVE-2026-92906
IBM watsonx.data 2.4.0 through 2.4.0 patch 8 could allow a remote authenticated attacker to cause a denial of service due to client-side enforcement of business rules.
-
CVE-2026-103033
ALTCHA is a Go library for creating and verifying ALTCHA challenges. Prior to 2.1.0, parseServerSignature can leave ServerSignatureVerificationData.Extra uninitialized when url.ParseQuery returns partial values and an error for a malformed percent escape in VerificationData. A later write of an unknown parsed key to the nil map panics before VerifyServerSignature or VerifyServerSignatureSafe completes signature verification, so an unauthenticated client can abort a request. Standard net/http recovers the request goroutine, while application goroutines or frameworks without recovery can crash the process. This issue is fixed in version 2.1.0.
-
CVE-2026-131240
In Splunk SOAR versions below 8.8.0, an authenticated user could cause Splunk SOAR to store excessive data in the Ingestion Status history through the Representational State Transfer (REST) API. When another user with permission to view system settings retrieves the Ingestion Status history, Splunk SOAR could use additional resources and return a larger response, which could degrade the availability of administrative monitoring. The vulnerability is possible because Splunk SOAR does not limit the amount of request data that it records for failed container creation before it stores the data in the Ingestion Status history. For more information see REST Containers (https://help.splunk.com/en/splunk-soar/soar-on-premises/rest-api-reference/6.4.0/container-endpoints/rest-containers), REST Roles and Permissions (https://help.splunk.com/en/splunk-soar/soar-on-premises/rest-api-reference/6.4.0/role-management-endpoints/rest-roles-and-permissions), and View ingested container statistics using Ingestion Status (https://help.splunk.com/en/splunk-soar/soar-on-premises/administer-soar-on-premises/6.4.1/manage-your-splunk-soar-on-premises-system-health-and-performance/view-ingested-container-statistics-using-ingestion-status) in the Splunk documentation.
-
CVE-2026-131195
In Splunk Enterprise versions below 10.4.5, 10.2.9, 10.0.12, and 9.4.17, an unauthenticated user could send repeated HTTP PubSub channel subscription requests that cause a Denial of Service (DoS) by permanently increasing memory usage until Splunk Enterprise exhausts available memory. The vulnerability affects instances that turn on the [pubsubsvr-http] stanza in server.conf because the HTTP PubSub broker does not release memory associated with a channel after the client disconnects.
-
CVE-2026-104199
aMQTT is an open source MQTT broker and client. From 0.11.0 until 0.11.4, add_subscription() in amqtt/broker.py accepts malformed single-level + wildcard placement because it checks substrings instead of validating each slash-delimited level. The _matches() function converts each + into a greedy regular-expression fragment and evaluates it with re.fullmatch(), allowing a filter containing many adjacent + characters to trigger catastrophic backtracking when compared with a topic that partially matches but ultimately fails. Because matching runs synchronously in _run_broadcast(), one subscription followed by a publish can block the broker event loop, stopping message delivery and new connection processing until restart. Anonymous clients can trigger the issue with the default configuration. Any authenticated client permitted to create arbitrary subscriptions and publish messages may still exploit it when authentication is enabled. This issue is fixed in version 0.11.4.
-
CVE-2026-103025
bep/imagemeta is a Go library for reading EXIF, IPTC and XMP image meta data from JPEG, TIFF, PNG, and WebP files. Prior to 0.17.3, imagemeta.Decode could exhaust memory when parsing PNG zTXt chunks whose keyword is "Raw profile type iptc": the declared chunk length can drive an allocation far larger than the input file, and the embedded zlib stream is decompressed without an output limit. The advisory reports a 42-byte input causing a 256 MB allocation and a 127 KB input inflating to 460 MB. The oversized-allocation path's memory effect depends on platform overcommit and configured GOMEMLIMIT or cgroup limits, while the decompression path reliably creates resident memory. Applications that decode untrusted PNGs with IPTC metadata processing can therefore suffer denial of service through memory exhaustion. This issue is fixed in 0.17.3.
-
CVE-2026-130647
@fastify/http-proxy is a Fastify plugin that proxies HTTP and WebSocket traffic to an upstream server. In versions from 9.0.0 up to but not including 11.6.10, when the plugin is registered with websocket support enabled, a WebSocket upgrade that is pipelined behind another HTTP request on the same cleartext connection arrives while that request's response is still attached to the socket. The shared upgrade listener sees the socket is already claimed and returns early without taking the upgrade, but it does not attach an error listener to the raw socket, and Node.js has already removed its own error listener before emitting the upgrade event. A later socket error on that connection, such as a client connection reset, is therefore unhandled and terminates the Node.js process. A single unauthenticated connection that pipelines an upgrade behind a normal request and then resets is enough to crash the server, and it can be repeated to keep the process down. Upgrades over TLS are not affected. The issue is fixed in @fastify/http-proxy 11.6.10, and users should upgrade to 11.6.10 or later.
-
CVE-2026-129500
@fastify/http-proxy is a Fastify plugin that proxies HTTP and WebSocket traffic to an upstream server. In versions from 9.0.0 up to but not including 11.6.10, when the plugin is registered with websocket support enabled, it takes every upgraded socket and dispatches the request through the Fastify router before the WebSocket library takes ownership of the socket. Node.js removes its own error listener from the socket before emitting the upgrade event, and the plugin does not attach one of its own during routing, so a socket error such as a client connection reset that arrives before the WebSocket library takes over is unhandled and terminates the Node.js process. The error window is open whenever anything asynchronous runs between dispatch and the route handler, such as an async hook on the proxied route or any unrelated async route in the application. A single unauthenticated upgrade request followed by a connection reset is enough to crash the server, and it can be repeated to keep the process down. The issue is fixed in @fastify/http-proxy 11.6.10, and users should upgrade to 11.6.10 or later.
-
CVE-2026-129502
@fastify/http-proxy is a Fastify plugin that proxies HTTP and WebSocket traffic to an upstream server. In versions from 9.4.0 up to but not including 11.6.10, when the plugin is registered with websocket support enabled, a WebSocket upgrade request whose path carries a URL fragment is matched by the router and answered with a successful upgrade before the upstream link is created. Building the upstream target then throws, the error is only logged, and the accepted client socket is left open with no error listener, so the next frame the client sends becomes an unhandled error and terminates the Node.js process. A request target that is not a valid URL reference reaching a parameterized prefix triggers the same path. A single unauthenticated upgrade request followed by one frame is enough to crash the server, and it can be repeated to keep the process down. The issue is fixed in @fastify/http-proxy 11.6.10, and users should upgrade to 11.6.10 or later.
-
CVE-2026-129224
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, the CCG lexicon parser's matchBrackets and augParseCategory functions recursively process nested parenthesized categories without a depth bound. An attacker-controlled lexicon string with excessive nesting can raise RecursionError through nltk.ccg.lexicon.fromstring and terminate the parse operation or request worker. This issue is fixed in version 3.10.4.
-
CVE-2026-129228
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, ToolboxSettings.parse builds a deeply nested attacker-controlled settings tree iteratively, but the later public helpers to_settings_string, remove_blanks, add_default_fields, sort_fields, and add_blank_lines recursively traverse that tree without a depth bound. Invoking any of those helpers on input nested beyond the interpreter recursion limit can raise RecursionError and abort the operation or request worker. This issue is fixed in version 3.10.4.
-
CVE-2026-129231
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, SeekableUnicodeStreamReader.readline appends each new block to a growing character buffer and reruns splitlines over the entire buffer on every iteration. An attacker-controlled corpus file containing one very long unterminated line therefore causes quadratic CPU consumption in line-oriented corpus readers and can deny service. This issue is fixed in version 3.10.4.
-
CVE-2026-129230
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, WordNet Synset.max_depth, min_depth, hypernym_paths, and hypernym_distances recursively follow attacker-controlled corpus pointers without a visited-node set or depth cap. A cyclic or excessively deep hypernym graph can raise RecursionError and terminate traversal APIs and similarity operations that depend on them. This issue is fixed in version 3.10.4.
-
CVE-2026-129225
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. In 3.10.3 and earlier, DependencyGraph._tree and triples recurse once per dependency edge without a depth bound, and get_cycle_path also lacks a visited-set termination guard. A deeply chained or cyclic attacker-controlled CoNLL or Malt-TAB dependency structure can raise RecursionError and abort a parse worker. This issue is fixed in version 3.10.4.
-
CVE-2026-129229
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. From 3.10.0 until 3.10.4, TokenSearcher.findall and Text.findall accept caller-controlled regular expressions with an unbounded number of capturing groups throughout the affected range. Starting in 3.10.3, nltk.redos.compile and RegexpTokenizer expose the same compile-time vector. The third-party regex engine compiles group-dense patterns with roughly quadratic cost before the match-time timeout begins, allowing a small unauthenticated request to exhaust CPU and deny service. This issue is fixed in version 3.10.4.
-
CVE-2026-129222
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, DependencyGraph.tree and DependencyGraph.triples walk parsed dependency edges through unbounded Python recursion and do not use a working cycle or depth guard. A crafted CoNLL or Malt-TAB dependency file containing a sufficiently long chain can raise RecursionError when the documented traversal APIs are called, aborting the current worker. This issue is fixed in version 3.10.4.
-
CVE-2026-129226
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, the semantic-logic read_type parser recursively processes both halves of nested complex type strings without applying LogicParser.MAX_PARSE_DEPTH. An attacker-controlled type string or signature with excessive nesting can raise RecursionError through Type.fromstring or type-checking entry points and abort the current worker. This issue is fixed in version 3.10.4.
-
CVE-2026-129227
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, Tree.fromlist recursively converts nested Python list input into Tree objects without enforcing a maximum tree depth. An attacker-controlled list nested beyond the interpreter recursion limit can raise RecursionError and terminate the current parse operation or request worker. This issue is fixed in version 3.10.4.
-
CVE-2026-129223
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, WordNet Synset.tree and acyclic_tree traverse relation graphs through recursive utility helpers without a depth bound. Exploitation requires the host application to load an attacker-influenced custom or foreign WordNet-format lexical resource; no relation in NLTK's bundled WordNet 3.0 corpus reaches the crash threshold. An excessively deep relation graph in such a resource can exceed the Python recursion limit and terminate the traversal or request worker. This issue is fixed in version 3.10.4.
-
CVE-2026-128140
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. From 3.6.7 until 3.10.4, nltk.downloader follows collection membership without tracking already visited collection nodes. A self-referencing or mutually referencing collection index can therefore make the package-closure walk loop indefinitely while its queue and memory usage grow, denying service when a victim processes the attacker-influenced index. This issue is fixed in version 3.10.4.
-
CVE-2026-128139
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. From 3.6 to 3.10.3, XMLCorpusView._read_xml_fragment accumulates 1 KiB blocks and repeatedly applies whole-buffer XML validation while a comment or CDATA terminator remains ahead. A large comment in an attacker-controlled corpus file causes quadratic CPU consumption through documented XML corpus readers across the affected range. On 3.6 and 3.9.2, a large CDATA section incurs the quadratic scan before rejection, while on 3.10.3 both comment and CDATA documents are read successfully. This vulnerability is fixed in 3.10.4.
-
CVE-2026-77336
Aperi'Solve is an open-source steganalysis web platform. Prior to 3.4.2, an unauthenticated user can submit a PNG to POST /upload, which reaches aperisolve/utils/png.py through analyze_image, PCRTAnalyzer, PNG.repair(), and check_idat(). A malformed IDAT chunk can make attacker-controlled length and newline data drive an unbounded combinations search in PNG._fix_dos2unix(), causing repeated CRC checks that occupy a worker and delay or block image analysis for other users. Per-IP upload limits and eventual job termination bound the denial of service, and the issue has no confidentiality or integrity impact. This issue is fixed in version 3.4.2.
-
CVE-2026-133200
A flaw was found in quarkus-vertx-utils. When a client disconnects while receiving a streaming response, the server fails to detect the closed connection and continues processing the data indefinitely. An unauthenticated remote attacker can exploit this issue by repeatedly initiating streaming requests and disconnecting, exhausting worker threads and resulting in a Denial of Service (DoS).
-
CVE-2026-129232
NLTK (Natural Language Toolkit) is a suite of open source Python modules, data sets, and tutorials supporting research and development in Natural Language Processing. Prior to 3.10.4, XMLCorpusView.read_block rebuilds the full root-to-node context path with "/".join(context) for every start tag while searching for a matching element. Deeply nested attacker-influenced corpus XML therefore causes quadratic pure-Python work and can pin a processing worker, resulting in denial of service. This issue is fixed in version 3.10.4.
-
CVE-2026-129400
Joomla Extension - vi-solutions - Unauthenticated verification mail amplification in Visforms extension 1.0.0-5.6.3, 6.0.0-6.2.0 - The verification mail task accepted arbitrary form IDs and recipients without rate limiting, allowing unauthenticated attackers to trigger mass verification mails and database entries.
-
CVE-2026-133265
In JetBrains YouTrack before 2026.2.19840 the attachment upload size limit could be bypassed via the REST API
-
CVE-2026-133255
In JetBrains Ktor before 3.6.1 a DoS attack was possible via the CIO engine
-
CVE-2026-133245
parse5 7.0.0 before 8.1.1 contains an uncaught exception vulnerability in Preprocessor#_processSurrogate that allows remote attackers to crash applications by supplying two consecutive low surrogates. Attackers can submit input such as U+DC00 U+DC00, which decodes to out-of-range code point 0x110000 and throws a RangeError from parse, parseFragment, or streaming parsers, crashing Node.js processes.
-
CVE-2026-133234
Stoatchat before 0.16.5 contains an unbounded JXL image decoding vulnerability in revolt_files decode_image that allows unauthenticated attackers to exhaust memory via the january /proxy endpoint. Attackers can request a small hosted JXL file declaring 12000x12000 pixels through /proxy to trigger huge allocations, OOM-killing january and denying proxy and embed service.
-
CVE-2026-121592
Mattermost Plugins versions <=12.-1 11.0.11 11.11.7 11.2.10.0 fail to check for a nil result after decoding the HTTP request body which allows an authenticated user to crash the Zoom plugin process, causing denial of service, via sending a JSON null body to the plugin's action endpoints. Mattermost Advisory ID: MMSA-2026-00722
-
CVE-2025-71472
littlefs 2.8.0 through 2.11.3 contains a divide-by-zero vulnerability in lfs_mount_() that allows attackers to crash processes mounting crafted filesystem images when block_count is 0. Attackers can supply an image lacking a superblock entry or recording block_count 0, causing lfs->seed % lfs->block_count to trigger SIGFPE and denial of service.
-
CVE-2026-99560
In the Linux kernel, the following vulnerability has been resolved: wifi: mt76: mt7925: fix reset panic caused by the WM2 rx irq mask mt7925_irq_map sets wm2_complete_mask, but mt7925_dma_layout has no tx done ring, so the MT_RXQ_MCU_WA queue and its NAPI are never set up. On a chip reset, mt7925e_mac_reset() calls napi_disable() on that NAPI and crashes: BUG: kernel NULL pointer dereference, address: 0000000000000a40 Workqueue: mt76 mt7925_mac_reset_work [mt7925_common] RIP: 0010:mutex_lock+0x1c/0x50 Call Trace: napi_disable+0x1d/0x50 mt7925e_mac_reset+0x327/0x420 [mt7925e] mt7925_mac_reset_work+0x9a/0x1a0 [mt7925_common] Remove the WM2 mask from mt7925_irq_map and from MT7925_INT_RX_DONE_ALL.
-
CVE-2026-99555
In the Linux kernel, the following vulnerability has been resolved: media: usbtv: fix null-pointer dereference on disconnect A null-pointer dereference can occur in usb_make_path() when called from usbtv_querycap() (and other ioctl handlers) during device disconnection. Oops: general protection fault, probably for non-canonical address 0xdffffc000000000a: 0000 [#1] SMP KASAN NOPTI KASAN: null-ptr-deref in range [0x0000000000000050-0x0000000000000057] RIP: 0010:usb_make_path include/linux/usb.h:985 [inline] RIP: 0010:usbtv_querycap+0x9c/0x100 drivers/media/usb/usbtv/usbtv-video.c:612 ... Call Trace: <TASK> v4l_querycap+0x236/0x470 drivers/media/v4l2-core/v4l2-ioctl.c:1106 __video_do_ioctl+0x8af/0xc70 drivers/media/v4l2-core/v4l2-ioctl.c:3133 video_usercopy+0x860/0x1430 drivers/media/v4l2-core/v4l2-ioctl.c:3475 v4l2_ioctl+0x18d/0x1e0 drivers/media/v4l2-core/v4l2-dev.c:366 vfs_ioctl fs/ioctl.c:51 [inline] __do_sys_ioctl fs/ioctl.c:597 [inline] __se_sys_ioctl+0xfc/0x170 fs/ioctl.c:583 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0x174/0x580 arch/x86/entry/syscall_64.c:94 This happens due to a race condition between the USB device disconnect routine (usbtv_disconnect()) and V4L2 ioctls. When the device is disconnected, usbtv_disconnect() sets usbtv->udev = NULL without holding the usbtv->v4l2_lock mutex. Concurrently, an ioctl handler like usbtv_querycap() can be executing under the v4l2_lock and attempt to use usbtv->udev, leading to a crash. To fix this, move the usbtv->udev = NULL assignment into usbtv_video_free() and protect it with mutex_lock(&usbtv->v4l2_lock). This ensures that no non-queue ioctl handler can run concurrently with the pointer being nullified. If an ioctl is already running, the mutex will block the disconnect path until the ioctl finishes. If an ioctl is blocked waiting for the lock, it will acquire it after the disconnect routine releases it, but will then check video_is_registered() (which was cleared by vb2_video_unregister_device()) and safely return -ENODEV. Acquiring v4l2_lock after vb2_video_unregister_device() avoids reintroducing a historical deadlock issue, and usbtv_audio_free() remains safe as it is called before usbtv_video_free().
-
CVE-2026-99543
In the Linux kernel, the following vulnerability has been resolved: backlight: qcom-wled: Fix NULL pointer dereference in wled_remove() wled_probe() doesn't set the driver data for the platform device. As a result, platform_get_drvdata() in wled_remove() will return NULL, and the first dereference of it crashes on unbind. Set the platform device driver data in wled_probe().
-
CVE-2026-99420
In the Linux kernel, the following vulnerability has been resolved: ovl: check that the parents of volatile/dirty are dirs An unprivileged user crashes the kernel by mounting overlayfs with the volatile option. A dentry that serves as the parent of a lookup must be a directory. ovl_create_volatile_dirty() walks work/incompat/volatile/dirty by name, and ovl_lookup_or_create() returns an existing entry whatever its type. A file or a symlink that another task creates during the mount becomes the parent of the next lookup, which calls a NULL ->lookup. An unprivileged user mounted overlayfs with "volatile" in a user namespace while a second process created a symlink named work/incompat, and the mount task oopsed on a NULL ->lookup call. Fix this by returning -ENOTDIR from ovl_lookup_or_create() when the caller asked for a directory and the existing entry is a non-dir.
-
CVE-2026-99421
In the Linux kernel, the following vulnerability has been resolved: ovl: always encode the origin fh when index is enabled On an overlay with index=on,nfs_export=on, copy up of a directory crashes the kernel once an LSM has started to deny CAP_DAC_READ_SEARCH to the mounter, the task that mounted the overlay. The mount and the LSM policy change require root. After that, any user who can write to a directory that exists only in a lower layer triggers the oops by doing so. Before a file or directory that exists only in a lower layer is changed for the first time, overlayfs copies it to the writable upper layer. This is called copy up. On copy up, ovl_get_origin_fh() encodes the file handle of the lower inode, called the origin. It returns NULL if ovl_can_decode_fh() says that the lower filesystem cannot decode handles. Whether the callers can handle NULL depends on the index option: - Without an index, NULL is fine: the callers store an empty origin xattr. - With index enabled, overlayfs keeps entries for copied up inodes in an internal index directory. The name of an index entry is derived from the handle, so ovl_get_origin_fh() must return a handle or an error. The mount only enables the index if ovl_can_decode_fh() succeeds for all layers, but the result can change later. The function checks the export operations of the filesystem and capable(CAP_DAC_READ_SEARCH). Copy up runs with the mounter's credentials, and an LSM can start to deny the capability to those credentials after the mount. ovl_get_origin_fh() then returns NULL with index enabled. This goes wrong in two callers: - With nfs_export=on, copy up of a directory creates an index entry for the directory. ovl_create_index() passes the NULL handle to ovl_get_index_name_fh(), which dereferences it. - If the result changes while mount(2) is still running, ovl_get_indexdir() passes a NULL from ovl_get_origin_fh() to ovl_verify_set_fh() in the same way. Before the commit in the Fixes tag, these two callers encoded the handle themselves, without consulting ovl_can_decode_fh(). Reproduced with a BPF LSM program that started to deny the capability after the mount. An unprivileged touch in a directory that had not been copied up yet oopsed in ovl_get_index_name_fh(). With this patch the touch succeeds and the directory gets its index entry. Fix this by calling ovl_can_decode_fh() in ovl_get_origin_fh() only when index is disabled.
-
CVE-2026-99391
In the Linux kernel, the following vulnerability has been resolved: xfs: don't try to scrub self-referential dirent in metapath checker LOLLM observes that the _ilock_both helper will loop forever trying to double-lock the same inode if mpath->dp == sc->ip. Callers are not supposed to call us with parent == child because that makes no sense. Trigger an assertion, mark the scan incomplete, and return IO error so that this fails noisily.
-
CVE-2026-99353
In the Linux kernel, the following vulnerability has been resolved: SUNRPC: Reject a socket that already has an svc_sock attached Writing the same socket descriptor to /proc/fs/nfsd/portlist twice attaches a second svc_sock to one socket. svc_setup_socket() saves the socket's callbacks before installing its own, so the second attach records svc_write_space() as the old write_space callback. svc_udp_init() invokes that callback by way of svc_sock_setbufsize(), and svc_write_space() then calls itself until the kernel stack is exhausted: BUG: TASK stack guard page was hit at ffffc900037d7ff8 svc_write_space+0x90/0x2b0 net/sunrpc/svcsock.c:429 svc_write_space+0xe6/0x2b0 net/sunrpc/svcsock.c:430 ... 700 more ... svc_sock_setbufsize+0x18d/0x220 net/sunrpc/svcsock.c:386 svc_udp_init net/sunrpc/svcsock.c:854 [inline] svc_setup_socket+0xb2f/0x1090 net/sunrpc/svcsock.c:1498 svc_addsock+0x2fd/0x760 net/sunrpc/svcsock.c:1547 __write_ports_addfd fs/nfsd/nfsctl.c:742 [inline] write_ports+0xa5b/0xcc0 fs/nfsd/nfsctl.c:861 nfsctl_transaction_write+0x106/0x1a0 fs/nfsd/nfsctl.c:112 svc_data_ready() and svc_tcp_state_change() chain through their saved callbacks the same way, so a TCP descriptor added twice recurses on the next incoming segment instead. Reaching any of this takes a writer on portlist, and the nfsd filesystem sets no FS_USERNS_MOUNT, so the reproducer needs CAP_SYS_ADMIN in the initial user namespace. Reject a socket that already carries sk_user_data. svc_setup_socket() overwrites that field unconditionally, so a socket some other consumer has claimed is one NFSD would corrupt whether or not the callbacks recurse.
-
CVE-2026-99323
In the Linux kernel, the following vulnerability has been resolved: ksmbd: fix invalid pointer dereference when get_inode_acl() fails When get_inode_acl() fails, it returns an ERR_PTR(), but ksmbd_acls_fattr() stored that pointer in smb_fattr without checking for errors. Later callers then passed the error pointer to security descriptor helpers or posix_acl_release(), resulting in an invalid pointer dereference. This was reproduced on an XFS-backed share when xfs_get_acl() returned ERR_PTR(-ENOMEM). Querying the SMB ACL from a CIFS client triggered the crash: BUG: unable to handle page fault for address: fffffffffffffff8 RIP: smb_acl_sec_desc_scratch_len+0x58/0x90 [ksmbd] Call Trace: smb2_query_info+0x40d/0x17c0 [ksmbd] handle_ksmbd_work+0x1ca/0x770 [ksmbd] Fix this by checking the return values from get_inode_acl() in ksmbd_acls_fattr(). Treat -EOPNOTSUPP as indicating that no ACL is available, and propagate all other errors to the callers. Update smb2_open() and smb2_get_info_sec() to handle these errors.
-
CVE-2026-133171
Tiptap @tiptap/core 3.7.0 before 3.30.5 contains a regular expression denial of service vulnerability in parseAttributes and parseShortcodeAttributes that allows attackers to exhaust CPU through quadratic backtracking. Attackers can supply Markdown with long crafted attribute strings parsed by createBlockMarkdownSpec, createAtomBlockMarkdownSpec or createInlineMarkdownSpec to stall browser pages or server processes.
-
CVE-2026-133161
### Summary ui With `strip_ansi=True`, Pydantic AI Harness's `ToolOutputLimits` capability removes ANSI escape sequences from tool output with a pattern that takes time quadratic in the length of a line. A tool result of a few hundred kilobytes, crafted with many unterminated escape sequences, can occupy the event loop for tens of seconds to minutes, delaying every other request and agent run in the process. ### Details This affects applications that set `strip_ansi=True` on `ToolOutputLimits` (named `OverflowingToolOutput` before 0.13.0) and run tools whose output an attacker can influence, such as shell commands, file reads, fetched pages or MCP servers. Stripping runs on the event loop before the capability reduces the output, so its own size limits do not bound it. The assumed attacker can place text in content a tool returns. ### Mitigation Upgrade `pydantic-ai-harness` to 0.70.0 or later. Until then, leave `strip_ansi` at its default, `False`. On Pydantic AI v1, the harness cannot be upgraded past 0.4.0. Upgrade to Pydantic AI v2 and `pydantic-ai-harness` 0.70.0, or leave `strip_ansi` off.
-
CVE-2026-127960
Denial of service through folded HTTP response headers. The HTTP stream wrapper joins each continuation line of a folded response header to the accumulated value by allocating a new string and copying the whole value, so a response with many short continuation lines takes disproportionate CPU time to parse. An attacker who controls an HTTP or HTTPS response fetched through file_get_contents(), fopen() or other stream functions, by making the application request a chosen URL or by influencing a server it already contacts, can keep PHP workers busy. Repeated or concurrent requests can cause a denial of service. This issue affects PHP versions from 8.1.32 through 8.1.34, from 8.2.28 before 8.2.36, from 8.3.18 before 8.3.37, from 8.4.5 before 8.4.28 and 8.5.* before 8.5.13.
-
CVE-2026-121584
Mattermost versions 11.7.x <= 11.7.11, 11.10.x <= 11.10.2, 11.11.x <= 11.11.1 fail to bound the number of records a single GraphQL request may hydrate which allows an authenticated user to exhaust server memory and cause a denial of service via a small query that repeats the runs field under many aliases. Mattermost Advisory ID: MMSA-2026-00767
-
CVE-2026-126827
A maliciously crafted SBSAR file, when imported or loaded through the Substance in 3ds Max plugin in Autodesk 3ds Max, can trigger a Stack Overflow vulnerability. A malicious actor may leverage this vulnerability to cause an application crash.
CVE records published in 2026 (ids starting CVE-2026-), sorted by how likely they are to be exploited.