CVE Feed

Buffer overflow in WebGL in Google Chrome prior to 153.0.8010.36 allowed a remote attacker to execute arbitrary code outside the sandbox via a crafted HTML page. (Chromium security severity: Critical)

Out of bounds write in WebGL in Google Chrome on on Android prior to 153.0.8010.36 allowed a remote attacker to execute arbitrary code outside the sandbox via a crafted HTML page. (Chromium security severity: Critical)

Use after free in WebGL in Google Chrome on on Android prior to 153.0.8010.36 allowed a remote attacker to execute arbitrary code outside the sandbox via a crafted HTML page. (Chromium security severity: Critical)

Use after free in WebGL in Google Chrome prior to 153.0.8010.36 allowed a remote attacker to execute arbitrary code outside the sandbox via a crafted HTML page. (Chromium security severity: Critical)

A flaw was found in the admin backend of gvfs. The privileged gvfsd-admin daemon changes the ownership of newly created private D-Bus sockets by calling the link-following chown() function on a pathname inside a user-controlled directory. A local attacker can exploit this via a Time-of-Check Time-of-Use (TOCTOU) race condition and exchange the socket pathname with a symbolic link pointing to an arbitrary root-owned file (such as /etc/pam.d/su). The daemon subsequently follows the symlink and changes the ownership of the targeted root-owned file to the attacker's user ID. This allows an authenticated local attacker to modify critical system files, leading to a full local privilege escalation to root.

An out-of-bounds read in the gf_dm_get_chunk_data function (src/utils/downloader.c) of GPAC v26.07.0 allows attackers to cause a Denial of Service (DoS) via sending a crafted HTTP request. Fixed in 2fd5a06ab226767900fd86edb5a1e8bfc1010640.

An issue in GPAC c2dee3aff638cd96f9617ac5b17dc2868cd90ef3 allows an attacker to cause a denial of service via the function gf_route_media_complete_object(). Fixed in 3c4e6c5b3e0c6fa9b16d55599701a08354538fab.

An issue in GPAC c2dee3aff638cd96f9617ac5b17dc2868cd90ef3 allows an attacker to execute arbitrary code via the src/media_tools/dvb_mpe.c, descriptorTime_slice_fec_identifier() and gf_m2ts_ipdatagram_reader() components. Fixed in 0e4093392e1f847c90d20e031e893cd942fef938.

The cupsUTF32ToUTF8() function in CUPS's cups/transcode.c lacks a source-length bound and can read past the end of the source buffer, resulting in a heap out-of-bounds read. This is reachable via SNMP supply-description parsing in backend/snmp-supplies.c with attacker-controlled content.

Two case-insensitive comparisons on request-derived usernames outside the main authorization path in CUPS's scheduler (printer ACL validation and private-attribute filtering) could allow bypass of username-based access controls in certain configurations.

SQL injection in the RelatedProduct block in Mageplaza Blog for Magento 2 (mageplaza/magento-2-blog-extension) through 4.3.2 allows remote unauthenticated attackers to execute arbitrary SQL commands and read arbitrary database contents via the id parameter to /mpblog/post/view.

Missing authorization in the Address Delete controller in Mageplaza GDPR for Magento 2 (mageplaza/module-gdpr) through 4.2.9 allows remote unauthenticated attackers to delete any customer's saved address, and to erase all stored addresses by iterating the address id, via a GET request to /customer/address/delete/id/{id}. The controller extends the legacy Action class instead of AbstractAccount, so no authentication, ownership or form key check is enforced.

zhitan-ems 1.0.0 is vulnerable to Cross Site Scripting (XSS) via SVG file upload through the /equipmentFile/upload endpoint.

A SQL Injection vulnerability in Siam Ordering (siam-server) 1.0.0 allows remote authenticated attackers to execute arbitrary SQL commands via the ${} string concatenation in AdminMapper.java and multiple other Mapper files (including MerchantWithdrawRecordMapper.java and MemberWithdrawRecordMapper.java).

A flaw was found in crun. After pivot_root, reopening /dev/null for stdio can follow a symlink and attach a host file to container stdio, then change that file's ownership. Affected versions are crun 1.29.1 and earlier. Default configurations that mount a fresh /dev are not exposed. No fixed release is available yet.

Information disclosure in the blogComments GraphQL query in Magefan Blog GraphQL for Magento 2 (magefan/module-blog-graph-ql) through 2.2.1 allows remote unauthenticated attackers to obtain blog commenter email addresses and internal customer and admin identifiers via a POST request to /graphql.

money-pos 1.0 contains a stored Cross-Site Scripting (XSS) vulnerability. When processing returns, the backend fails to filter or escape the goodsName parameter, directly concatenating it into the order log description; the frontend subsequently renders this content using v-html. An attacker with product creation privileges can inject a malicious JavaScript payload, causing unauthorized code execution when an administrator views the order logs.

An issue was discovered in Cyrus IMAP before 3.12.4. A JMAP email-header blob ID can reference an out-of-bounds index. An authenticated user could attempt to download a crafted JMAP blob ID of the form H<emailid>-<index>, which could read past the end of the internal blob_headers array during download, exposing adjacent heap memory.

An out-of-bounds read in the stbtt_GetGlyphShape component of nothings stb commit 31c1ad3 allows attackers to cause a Denial of Service (DoS) via sending a crafted TTF file.

An issue was discovered in Cyrus IMAP before 3.12.4. Mailbox/set let a sharee change a special-use role on shared mailboxes. An authenticated user with maySetKeywords on another user's mailbox could change that mailbox's specialuse annotation. This could allow the sharee to change the shared mailbox to perform the archived, snoozed, or other role, which might cause mail mail to be written to the shared mailbox, sharing more content than intended. (This is likely to be an unusual situation, made more unusual because if the target already has an non-shared mailbox with that role, role duplication suppression will prevent the update.)

A flaw was found in the memcached cache plugin of the community.general Ansible collection. Although its documentation states that records are stored in JSON format, the plugin performs no explicit serialization and relies on python-memcached, which pickles values on write and unpickles them on read. Because memcached is unauthenticated and cache keys are predictable, an attacker able to reach a network-exposed or shared memcached instance can write a crafted pickle payload that is deserialized and executed on the Ansible controller when the poisoned fact cache is next read, leading to remote code execution.

Authentication Bypass via Hardcoded Master Verification Code vulnerability in Siam Ordering (siam-server) 1.0.0 allows remote unauthenticated attackers to log in as any user, merchant, or administrator.

An out-of-bounds read in the stbsp_vsnprintf function (stb_sprintf.h) of nothings stb commit 31c1ad3 allows attackers to cause a Denial of Service (DoS) via sending a crafted input.

SQL injection vulnerability in PbootCMS versions 3.2.0 through 3.2.5 allows an authenticated user to modify arbitrary user account fields (including passwords and roles) via crafted parameters to the User/mod interface, enabling account takeover.

An out-of-bounds read in the gf_dm_data_received function (downloader.c) of GPAC v26.07.0 allows attackers to cause a Denial of Service (DoS) via sending a crafted HTTP request. Fixed in 2fd5a06ab226767900fd86edb5a1e8bfc1010640.

yshopmall <=3.3 is vulnerable to Cross Site Scripting (XSS). The file upload endpoint /api/upload of the system lacks file type validation. Attackers can upload files of any type, including HTML, JSP, and other executable files.

An arbitrary file upload and path traversal vulnerability exists in LZ-litchi 1.0.0. Unauthenticated remote attackers can upload arbitrary files and write them outside the intended storage directory via the directory parameter in POST /app-api/infra/file/upload.

In RuoYi-Cloud-Plus <= 2.6.2 in the ruoyi-workflow module, multiple core task APIs in FlwTaskController lack permission annotations, and the Service layer does not verify whether the current user is the task handler/related user. Authenticated low-privileged remote attackers can read sensitive workflow task details (/task/getTask/{taskId}) and trigger unauthorized workflow executions (/task/startWorkFlow).

Buffer Overflow vulnerability in GPAC c2dee3aff638cd96f9617ac5b17dc2868cd90ef3 allows an attacker to execute arbitrary code via the nhntdmx_process() function. Fixed in fac50e6a12ac27ffabdd5d3080b51afcc44ad8d6.

An issue was discovered in Cyrus IMAP before 3.12.4. There is a VPATCH BYPARAM double-free. An authenticated calendar user could crash a Cyrus CalDAV worker with a PATCH containing PATCH-ACTION="BYPARAM@..." against a resource with two or more properties of the matched kind. The memory holding the selector would be freed once on each iteration over the properties.

A use-after-free in the SocketDescriptor::tcpReadHandler1 function (liveMedia/RTPInterface.cpp) of LIVE555 Streaming Media (version 2026.02.26) allows attackers to cause a Denial of Service (DoS) via sending a series of crafted RTSP and HTTP requests to the server.

An issue in Actions Semiconductor Co. Ltd Tool- Media Player Utilities v.4.46 allows a physically proximate attacker execute arbitrary code via the Production.dll and RdiskUpgrade.exe components

A vulnerability in the Thesycon DPC Latency Checker driver (dpc.sys) thru 1.4.0 allows local unprivileged users to cause a denial-of-service (BSOD) condition on Windows systems. The driver exposes an IOCTL interface (0x81772008) that accepts user-controlled input without validating pointers before passing them to kernel APIs. Specifically, it dereferences a user-supplied pointer and uses the resulting value in a call to ExSetTimerResolution, leading to an arbitrary kernel memory access. Exploiting this flaw results in a system crash.

An issue was discovered in s-pms SPMS-Server through v1.0. The application contains a hardcoded default access token secret within its core configuration file, which is not overridden or removed in the production environment profile. A remote, unauthenticated attacker can locally forge valid administrative session tokens to completely bypass the authentication mechanism gaining full unauthorized access to protected backend APIs.

A divide-by-zero vulnerability in the gf_dash_get_timeline_duration function (src/media_tools/dash_client.c) of GPAC v26.07.0 allows attackers to cause a Denial of Service (DoS) via a crafted MPD SegmentTimeline. Fixed in 2fd5a06ab226767900fd86edb5a1e8bfc1010640.

An issue was discovered in Cyrus IMAP before 3.12.4. There is a Sieve mailbox existence oracle. An authenticated user could install a Sieve script that probed whether another user's private mailbox existed, or read the value of shared mailbox annotations, by observing which fileinto branch fired during LMTP delivery.

An issue in WIngs3D v.2.4.1 allows a local attacker to cause a denial of service via a crafted Wavefront OBJ file

An issue in SJRC F11 SJ-GPS-PRO firmware build 2019-09-17 allows a remote attacker to obtain sensitive information via the inetd service spawns /app/sh_for_telnet

A flaw was found in bubblewrap. During sandbox setup, creating files or directories under the new root can follow a parent symlink onto the host via /oldroot, writing attacker-chosen paths outside the sandbox as the launching user. This happens before the sandboxed process starts. This issue is GHSA-pxhw-h44j-8pfx. It is fixed in bubblewrap 0.12.0.

In the opam package before 2.5.2 for OCaml, the sandbox protection mechanism can be bypassed because symlinks are mishandled during use of .install files.

Buffer Overflow vulnerability in GPAC c2dee3aff638cd96f9617ac5b17dc2868cd90ef3 allows an attacker to execute arbitrary code via the j2kdec_process() function. Fixed in 9a253a07fd3f6b48022bba74302bf39388dda859.

A stored Cross-Site Scripting (XSS) vulnerability exists in the announcement preview component of REBUILD 4.4.3. Although the announcement content undergoes HTML escaping on the server side, the client-side preview code reverses the escaped entities using jQuery's `html().text()` method and subsequently injects the result into the DOM. An administrator or attacker capable of controlling the announcement content can exploit this vulnerability to execute arbitrary JavaScript code in the browsers of users viewing the affected pages "which may include the dashboard, activity feed, or login page, depending on the announcement's visibility settings.

An issue was discovered in Cyrus IMAP before 3.12.4. CalDAV/CardDAV multiget bypasses a per-href ACL. An authenticated DAV user with some shared access to another user's calendar or address book could read even unshared events or contacts by including the target hrefs in a calendar-multiget or addressbook-multiget REPORT.

An issue was discovered in Cyrus IMAP before 3.12.4. JMAP snooze bypasses the destination-mailbox ACL. An authenticated user with insert permissions on another user's snoozed mailbox could cause insertion of mail to that user's inbox, or any other of their mailboxes whose id was known to the user, despite having no insert permissions to the target mailbox.

An uncontrolled recursion vulnerability in the Windows SIPA event log parser of Google go-attestation versions up to and including 0.6.1 allows an attacker to cause a denial of service (DoS). The (*WinEvents).readELAMAggregation function recurses for every nested elamAggregation sub-event without enforcing a maximum recursion depth limit, while the size guard is bypassed on recursive execution paths. By submitting a crafted Windows event log containing deeply nested elamAggregation headers, an attacker can exhaust the goroutine call stack, triggering an unrecoverable fatal runtime error (stack overflow) that immediately crashes the verifier application.

OpenIDC/cjose is a C library implementing the Javascript Object Signing and Encryption (JOSE). In versions 0.6.1 through 0.6.2.5, when cjose encrypts a JWE using an AES-CBC-HMAC content-encryption algorithm (`A128CBC-HS256`, `A192CBC-HS384`, or `A256CBC-HS512`) together with any key-management algorithm that generates a fresh content-encryption key (CEK), the CEK is all zero bytes instead of being randomly generated. The resulting JWE is therefore encrypted and authenticated under a fixed, publicly known key, so anyone who obtains the JWE can recover the plaintext and forge or modify the content. This is fixed in version 0.6.2.6 by `_cjose_jwe_set_cek_aes_cbc()` generating the CEK from `RAND_bytes`. A regression test asserts that the `encrypted_key` differs across two encryptions for each AES-CBC-HMAC variant. Until upgrading, for data encrypted with cjose, three options are available. Use an AES-GCM `enc` (`A128GCM` / `A192GCM` / `A256GCM`) instead of an AES-CBC-HMAC `enc`, use `alg=dir` with a caller-supplied CEK, or avoid using cjose for JWE encryption with the affected algorithm pair. These are mitigations for new ciphertexts only; data already encrypted under the zero key remains compromised and should be re-encrypted (and any secrets it contained rotated).

OpenIDC/cjose is a C library implementing the Javascript Object Signing and Encryption (JOSE). Prior to version 0.6.2.5, cjose's JWE decryption path for the AES Key Wrap key-management algorithms (`alg` = `A128KW`, `A192KW`, `A256KW`) does not validate the length of the attacker-supplied `encrypted_key` (JWE Encrypted Key) before unwrapping it into a fixed-size, heap-allocated Content Encryption Key (CEK) buffer. A remote, unauthenticated attacker who can submit a crafted JWE to an application that decrypts it with an AES-KW symmetric key can trigger an out-of-bounds heap write, corrupting the heap. This leads at minimum to a crash (denial of service) and, depending on the heap layout and allocator, may be leverageable for further memory-corruption impact. `cjose_jwe_import()` / `cjose_jwe_decrypt()` are pre-authentication entry points: they parse and process fully attacker-controlled input. Upgrade to cjose 0.6.2.5 to receive a patch. If upgrading is not immediately possible, reject the AES Key Wrap algorithms (`A128KW`/`A192KW`/`A256KW`) for untrusted JWEs at the application layer.

MCP Kotlin SDK is the Kotlin Multiplatform software development kit for the Model Context Protocol. In versions 0.7.0 through 0.12.0, `ReadBuffer.append` in `kotlin-sdk-core/src/commonMain/kotlin/io/modelcontextprotocol/kotlin/sdk/shared/ReadBuffer.kt` writes every chunk of bytes received from the stdio transport into a `kotlinx.io.Buffer` with no size cap. Frames are extracted from that buffer only when a `\n` (0x0a) byte is observed. A peer that streams bytes without ever sending a newline causes the internal buffer to grow indefinitely until the JVM (or the surrounding host process) is OOM-killed. The leak is amplified by `StdioServerTransport` and `StdioClientTransport`, which both queue raw chunks through a `kotlinx.coroutines.channels.Channel<ByteArray>(Channel.UNLIMITED)` and then call `readBuffer.append(chunk)` without backpressure or size guard. This is a remote-pre-auth denial of service whenever an SDK stdio server's stdin is fed by an untrusted or attacker-controlled producer (for example: a host program that exec's the MCP server as a subprocess and pipes through bytes received from a network peer, or a sidecar wrapper that proxies bytes from an HTTP endpoint to the stdio transport). Version 0.13.0 fixes the issue.

The source-controller is a Kubernetes operator, specialised in artifacts acquisition from external sources such as Git, OCI, Helm repositories and S3-compatible buckets. In versions 0.0.17 through 1.8.4, an actor with the ability to influence the contents of a bucket referenced by a `Bucket` resource can cause source-controller to write fetched object data to paths outside the per-reconciliation working directory. The corruption surface is bounded by source-controller's own and downstream Flux controllers' digest verification: source-controller verifies stored artifact digests during reconciliation and rebuilds on divergence; consumers (kustomize-controller, helm-controller) verify the digest of fetched artifacts and reject mismatches. These checks prevent a manipulated artifact from reaching the cluster, but an attacker can still write files anywhere the source-controller pod has permission to write. Separately, a user with permission to create or update `GitRepository` resources can cause source-controller to test for the existence of paths outside the cloned repository. Because the result is exposed via the resource's status, this allows limited enumeration of file paths on the controller pod. This surface exists only on source-controller v1.6.0 and later, where the sparse-checkout feature was introduced. This vulnerability was fixed in source-controller v1.8.5. There is no in-product workaround. Users should upgrade to a patched version. As a defense-in-depth measure for the GitRepository sparse-checkout surface, a `ValidatingAdmissionPolicy` (or a third-party policy engine such as Kyverno or OPA Gatekeeper) can be deployed to reject `GitRepository` resources whose `.spec.sparseCheckout` entries contain `..` or absolute path segments.

A flaw was found in gdk-pixbuf. This vulnerability allows a remote attacker to cause a heap out-of-bounds read by providing a specially crafted Apple Icon Image (.icns) file. The uncompress() function, which handles RLE-encoded ICNS icon data, fails to validate the source buffer's boundaries during decompression. This can lead to a denial of service, where the application crashes, or to information disclosure, potentially revealing sensitive data from adjacent memory.