An uncontrolled recursion vulnerability exists in the LINE iOS app versions prior to 26.19.0. This flaw is due to inadequate validation of the structure of externally supplied Flex Message content, which can cause the app to crash when processing a crafted message. A server-side mitigation has been implemented to protect even unpatched clients.
FFmpeg versions 7.0 through 9.0.2 contain an out-of-bounds read vulnerability in the dvdvideo_menu_next_ps_block() function within libavformat/dvdvideodec.c. This flaw occurs when demuxing DVD menu domains, allowing crafted DVD images to bypass boundary checks and cause a process crash.
Dokploy versions from 0.28.0 up to but not including 0.31.1 contain a path traversal vulnerability in the patch.readRepoDirectories tRPC procedure. This flaw allows authenticated organization members with service read permission to supply a crafted repository path, enabling them to list tracked files of arbitrary git repositories, potentially including those of other organizations. The vulnerability arises because the crafted path is passed unconfined to the git ls-tree command, allowing unauthorized enumeration of source trees.
A vulnerability in Keycloak's handling of User-Managed Access (UMA) token requests allows an authenticated user to obtain valid Requesting Party Tokens (RPTs) for clients or resource servers that have been administratively disabled. This occurs because Keycloak does not verify whether the target client or resource server is enabled before issuing tokens. The flaw enables bypassing of administrative access controls and permission restrictions, but does not grant privileges beyond those already authorized for the user. The impact is limited in typical deployments where downstream services enforce online token introspection or verify client status.
A vulnerability in Keycloak allows an authenticated user to bypass token revocation when User-Managed Access (UMA) authorization is enabled. Specifically, Keycloak does not enforce the not-before revocation cutoff for UMA grant requests using ID token claims, enabling reuse of revoked ID tokens to mint fresh Requesting Party Tokens (RPTs). This flaw permits unauthorized access continuation but does not allow privilege escalation beyond the active session's permissions.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability that allows memory exhaustion. This occurs because the declared length of an ASN.1 element in its BER header is passed unchecked to sysread, causing Perl to allocate memory based on the declared length rather than the actual data size. A maliciously crafted 6-byte header can declare a length up to 4GB, leading to large memory allocation attempts. If the allocation fails, Perl exits with an "Out of memory" error that cannot be caught by eval.
Bar Assistant versions 3.0.0 up to but not including 6.11.0 contain an authorization bypass vulnerability in the BarController::join function. This flaw allows authenticated users to join bars without a valid invite code by sending POST requests omitting the invite_code parameter. Exploiting this, attackers can access data such as cocktails, ingredients, members, settings, and search tokens for bars that have invite codes disabled or set to NULL.
BeamNG.drive versions prior to 0.40.6 have an out-of-bounds write vulnerability in the JIT-compiled LuaJIT engine. This flaw arises because the JIT traces omit foreign function interface (FFI) array bounds checks, allowing malicious Lua code from untrusted mods to escape the sandbox. Exploitation can lead to arbitrary command execution with the privileges of the game user.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability that allows CPU exhaustion via unbounded arc length in OBJECT IDENTIFIER decoding and encoding. The flaw arises because the module processes arcs of an OID string with quadratic time complexity relative to the arc's length, without limiting that length. An attacker who can control the decoded data or the OID string passed to encode can cause excessive CPU usage with a single oversized arc.
The Canva Desktop App versions prior to 1.128.0 store encrypted application state data on disk using a hardcoded encryption key. This key is embedded as a plaintext string in the main.js file and is the same across all installations on Windows and macOS platforms. The vulnerability is classified as low severity with limited impact on confidentiality and no impact on integrity or availability.
A vulnerability in aoiflux mutant versions 2.3.0, 2.4.0, and 2.5.0 affects the Forensic Parser component in the file builtin/hive_builtins.go. The flaw involves manipulation of the argument Count, leading to uncontrolled memory allocation. This issue can be exploited locally. Upgrading to version 2.6.0 resolves the vulnerability.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability that allows memory exhaustion via unlimited recursion depth during decoding of constructed-form strings. The decoder recursively processes nested elements without bounding the recursion depth, leading to heap memory growth. An attacker controlling the decoded data can exploit this by crafting deeply nested constructed strings to exhaust system memory.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability where zero-length BIT STRING content can cause memory exhaustion. This occurs because a BIT STRING with zero content octets leads to a negative length parameter in a substring operation, causing Perl to read excessive input data repeatedly. An attacker controlling the decoded data can exploit this to exhaust system memory with relatively short input.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability where specially crafted exponent-notation strings in INTEGER encoding can cause CPU exhaustion. The encoder does not validate that large values are integers before processing, allowing an attacker to trigger a costly computation loop.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability in the asn_recv function that allows memory exhaustion. The function repeatedly peeks at socket data, doubling the buffer size each time without consuming data, due to a logic error in the guard condition and handling of incomplete length decoding. This can cause fatal memory exhaustion when processing a single truncated BER header.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability that allows CPU exhaustion via a quadratic bigint loop in the os2ip function. This occurs when decoding INTEGER or ENUMERATED values longer than four octets, leading to excessive CPU usage due to the algorithm's O(n^2) complexity. An attacker controlling the decoded data can exploit this by providing an oversized value, causing denial of service through resource exhaustion.
lwext4 versions through 1.0.0 have a heap buffer overflow vulnerability in the ext4_fs_init_inode_bitmap() function due to improper validation of the s_inodes_per_group field against the inode bitmap size. This flaw allows an attacker to craft a malicious ext4 image with an oversized s_inodes_per_group value, causing a memset operation to write beyond the allocated heap buffer, leading to memory corruption and application crashes during mounting.
lwext4 versions through 1.0.0 have an out-of-bounds write vulnerability in the ext4_fs_init_block_bitmap() function. This flaw allows attackers to corrupt heap memory by supplying a crafted ext4 image with oversized s_reserved_gdt_blocks and BLOCK_UNINIT groups. Exploitation can cause applications mounting the image to crash.
Convert::ASN1 versions before 0.37 for Perl contain a vulnerability in binary REAL decoding where the exponent length is not properly checked. This flaw allows an attacker controlling the decoded data to cause CPU exhaustion by crafting values that cause excessive looping during decoding.
CVE-2026-133480 is an authorization bypass vulnerability in hrshadhin CloudSchool versions 3.0.0 and 3.0.1. The flaw exists in the PublicController.checkSanity4Student function, allowing remote attackers to manipulate the student_id argument to bypass authorization checks. Exploit code has been published, but there is no vendor response or patch available yet. The vulnerability has a medium severity rating with a CVSS 4.0 score of 5.3.