Linux

In the Linux kernel, the following vulnerability has been resolved: bpf: Fix unsound scalar forking in maybe_fork_scalars() for BPF_OR maybe_fork_scalars() is called for both BPF_AND and BPF_OR when the source operand is a constant. When dst has signed range [-1, 0], it forks the verifier state: the pushed path gets dst = 0, the current path gets dst = -1. For BPF_AND this is correct: 0 & K == 0. For BPF_OR this is wrong: 0 | K == K, not 0. The pushed path therefore tracks dst as 0 when the runtime value is K, producing an exploitable verifier/runtime divergence that allows out-of-bounds map access. Fix this by passing env->insn_idx (instead of env->insn_idx + 1) to push_stack(), so the pushed path re-executes the ALU instruction with dst = 0 and naturally computes the correct result for any opcode.

In the Linux kernel, the following vulnerability has been resolved: usb: gadget: f_mass_storage: Fix potential integer overflow in check_command_size_in_blocks() The `check_command_size_in_blocks()` function calculates the data size in bytes by left shifting `common->data_size_from_cmnd` by the block size (`common->curlun->blkbits`). However, it does not validate whether this shift operation will cause an integer overflow. Initially, the block size is set up in `fsg_lun_open()` , and the `common->data_size_from_cmnd` is set up in `do_scsi_command()`. During initialization, there is no integer overflow check for the interaction between two variables. So if a malicious USB host sends a SCSI READ or WRITE command requesting a large amount of data (`common->data_size_from_cmnd`), the left shift operation can wrap around. This results in a truncated data size, which can bypass boundary checks and potentially lead to memory corruption or out-of-bounds accesses. Fix this by using the check_shl_overflow() macro to safely perform the shift and catch any overflows.

Local privilege escalation in libcap allows unprivileged users to exploit a race condition in cap_set_file() by redirecting capability updates to attacker-controlled files via symlink manipulation in parent directories. An authenticated local attacker with write access to a parent directory and user interaction can inject or strip capabilities from arbitrary executables, achieving privilege escalation. No public exploit code or confirmed active exploitation has been identified at this time.

In the Linux kernel, the following vulnerability has been resolved: net: atm: fix crash due to unvalidated vcc pointer in sigd_send() Reproducer available at [1]. The ATM send path (sendmsg -> vcc_sendmsg -> sigd_send) reads the vcc pointer from msg->vcc and uses it directly without any validation. This pointer comes from userspace via sendmsg() and can be arbitrarily forged: int fd = socket(AF_ATMSVC, SOCK_DGRAM, 0); ioctl(fd, ATMSIGD_CTRL); // become ATM signaling daemon struct msghdr msg = { .msg_iov = &iov, ... }; *(unsigned long *)(buf + 4) = 0xdeadbeef; // fake vcc pointer sendmsg(fd, &msg, 0); // kernel dereferences 0xdeadbeef In normal operation, the kernel sends the vcc pointer to the signaling daemon via sigd_enq() when processing operations like connect(), bind(), or listen(). The daemon is expected to return the same pointer when responding. However, a malicious daemon can send arbitrary pointer values. Fix this by introducing find_get_vcc() which validates the pointer by searching through vcc_hash (similar to how sigd_close() iterates over all VCCs), and acquires a reference via sock_hold() if found. Since struct atm_vcc embeds struct sock as its first member, they share the same lifetime. Therefore using sock_hold/sock_put is sufficient to keep the vcc alive while it is being used. Note that there may be a race with sigd_close() which could mark the vcc with various flags (e.g., ATM_VF_RELEASED) after find_get_vcc() returns. However, sock_hold() guarantees the memory remains valid, so this race only affects the logical state, not memory safety. [1]: https://gist.github.com/mrpre/1ba5949c45529c511152e2f4c755b0f3

On Linux, if the target of Root.Chmod is replaced with a symlink while the chmod operation is in progress, Chmod can operate on the target of the symlink, even when the target lies outside the root. The Linux fchmodat syscall silently ignores the AT_SYMLINK_NOFOLLOW flag, which Root.Chmod uses to avoid symlink traversal. Root.Chmod checks its target before acting and returns an error if the target is a symlink lying outside the root, so the impact is limited to cases where the target is replaced with a symlink between the check and operation.

In the Linux kernel, the following vulnerability has been resolved: ksmbd: use volume UUID in FS_OBJECT_ID_INFORMATION Use sb->s_uuid for a proper volume identifier as the primary choice. For filesystems that do not provide a UUID, fall back to stfs.f_fsid obtained from vfs_statfs().

In the Linux kernel, the following vulnerability has been resolved: ksmbd: unset conn->binding on failed binding request When a multichannel SMB2_SESSION_SETUP request with SMB2_SESSION_REQ_FLAG_BINDING fails ksmbd sets conn->binding = true but never clears it on the error path. This leaves the connection in a binding state where all subsequent ksmbd_session_lookup_all() calls fall back to the global sessions table. This fix it by clearing conn->binding = false in the error path.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: SCO: Fix use-after-free in sco_recv_frame() due to missing sock_hold sco_recv_frame() reads conn->sk under sco_conn_lock() but immediately releases the lock without holding a reference to the socket. A concurrent close() can free the socket between the lock release and the subsequent sk->sk_state access, resulting in a use-after-free. Other functions in the same file (sco_sock_timeout(), sco_conn_del()) correctly use sco_sock_hold() to safely hold a reference under the lock. Fix by using sco_sock_hold() to take a reference before releasing the lock, and adding sock_put() on all exit paths.

In the Linux kernel, the following vulnerability has been resolved: netfilter: conntrack: add missing netlink policy validations Hyunwoo Kim reports out-of-bounds access in sctp and ctnetlink. These attributes are used by the kernel without any validation. Extend the netlink policies accordingly. Quoting the reporter: nlattr_to_sctp() assigns the user-supplied CTA_PROTOINFO_SCTP_STATE value directly to ct->proto.sctp.state without checking that it is within the valid range. [..] and: ... with exp->dir = 100, the access at ct->master->tuplehash[100] reads 5600 bytes past the start of a 320-byte nf_conn object, causing a slab-out-of-bounds read confirmed by UBSAN.

In the Linux kernel, the following vulnerability has been resolved: xfrm: Fix work re-schedule after cancel in xfrm_nat_keepalive_net_fini() After cancel_delayed_work_sync() is called from xfrm_nat_keepalive_net_fini(), xfrm_state_fini() flushes remaining states via __xfrm_state_delete(), which calls xfrm_nat_keepalive_state_updated() to re-schedule nat_keepalive_work. The following is a simple race scenario: cpu0 cpu1 cleanup_net() [Round 1] ops_undo_list() xfrm_net_exit() xfrm_nat_keepalive_net_fini() cancel_delayed_work_sync(nat_keepalive_work); xfrm_state_fini() xfrm_state_flush() xfrm_state_delete(x) __xfrm_state_delete(x) xfrm_nat_keepalive_state_updated(x) schedule_delayed_work(nat_keepalive_work); rcu_barrier(); net_complete_free(); net_passive_dec(net); llist_add(&net->defer_free_list, &defer_free_list); cleanup_net() [Round 2] rcu_barrier(); net_complete_free() kmem_cache_free(net_cachep, net); nat_keepalive_work() // on freed net To prevent this, cancel_delayed_work_sync() is replaced with disable_delayed_work_sync().

NULL pointer dereference in Linux kernel NFSD export cache cleanup allows local denial of service when RCU readers in e_show() and c_show() concurrently access export path and client name objects while cache_clean removes entries and drops the last reference prematurely. The vulnerability stems from path_put() and auth_domain_put() executing before the RCU grace period completes, freeing sub-objects still in use by readers. A fix has been merged upstream that defers these cleanup operations to a dedicated workqueue after the RCU grace period, ensuring safe resource release in process context where sleeping is permitted.

Use-after-free in Linux kernel NFSD /proc/fs/nfs/exports proc entry allows information disclosure when a network namespace is destroyed while an exports file descriptor remains open. The vulnerability occurs because exports_proc_open() captures a network namespace reference without holding a refcount, enabling nfsd_net_exit() to free the export cache while the fd is still active, leading to subsequent reads dereferencing freed memory. The fix holds a struct net reference for the lifetime of the open file descriptor, preventing namespace teardown while any exports fd is open.

Heap overflow in Linux kernel NFSv4.0 LOCK replay cache allows unauthenticated remote attackers to corrupt kernel memory by triggering a denial-of-service or potential code execution. The vulnerability exists in nfsd4_encode_operation() which copies encoded LOCK responses up to 1024 bytes into a fixed 112-byte inline buffer without bounds checking, resulting in up to 944 bytes of slab-out-of-bounds writes. Exploitation requires two cooperating NFSv4.0 clients but no special privileges; upstream fixes are available across multiple stable kernel branches.

Buffer overflow in Linux kernel HID-BPF subsystem allows arbitrary return values from dispatch_hid_bpf_raw_requests() to overflow the hid_hw_request buffer without validation. The vulnerability affects all Linux kernel versions with HID-BPF support; attackers with the ability to load or influence BPF programs targeting HID devices can trigger memory corruption. No CVSS score, EPSS data, or confirmed active exploitation has been assigned at time of analysis.

Linux kernel sunrpc subsystem fails to properly release cache_request objects when file descriptors are closed mid-read, resulting in memory leaks and potential information disclosure through stale cache entries. The vulnerability affects all Linux kernel versions with the affected sunrpc cache implementation, and requires no special privileges or network access to trigger since it occurs during normal file descriptor closure in the kernel's user-space cache management interface.

Use-after-free in Linux kernel nvdimm/bus asynchronous device registration allows local denial of service when device_add() fails during nd_async_device_register(). The vulnerability occurs because a device reference is dropped before the parent pointer is safely accessed, causing a kernel crash or memory corruption. No authenticated access required; only local access with ability to trigger device registration failures.

Linux kernel mm/rmap subsystem fails to correctly preserve page table entry attributes (writable and soft-dirty bits) when batching unmap operations on lazyfree folios, causing kernel panic via page table check violation when a large folio with mixed writable/non-writable PTEs is unmapped across multiple processes. The vulnerability affects all Linux kernel versions with the vulnerable folio_unmap_pte_batch() code path and can be triggered by local attackers through a specific sequence of memory management syscalls (MADV_DONTFORK, fork(), MADV_DOFORK, MADV_FREE, and memory reclaim), resulting in denial of service via kernel crash.

Linux kernel memory management allows installation of PMD entries pointing to non-existent physical memory or causes NULL pointer dereferences in move_pages_huge_pmd() when handling huge zero page migrations via UFFDIO_MOVE. The vulnerability occurs because the function incorrectly handles NULL folio pointers for huge zero pages, either producing bogus page frame numbers on SPARSEMEM_VMEMMAP systems or dereferencing NULL on other memory models. Additionally, destination PMD entries lose special mapping metadata (pmd_special flag), causing subsequent page reference counting corruption. No CVSS score is available; no active exploitation reported.

Use-after-free vulnerability in Linux kernel's Cadence MAC (macb) driver allows local attackers to read freed memory via ethtool get_ts_info calls on PTP-capable network interfaces. The PTP clock is registered when the interface opens and destroyed when it closes, but the ethtool handler can still access it after deallocation, causing a kernel memory access violation. No active exploitation confirmed; patch available in stable kernel releases.

Out-of-bounds memory access in the Linux kernel bnxt_en driver allows a malicious or compromised Broadcom NetXtreme network interface card to corrupt kernel heap memory or crash the system by supplying an unvalidated 16-bit type field in a debug buffer producer async event, affecting all Linux kernel versions using the vulnerable bnxt driver code path.

Null pointer dereference in Linux kernel mac80211 IEEE 802.11 wireless subsystem crashes AP_VLAN stations during channel bandwidth change operations. The ieee80211_chan_bw_change() function incorrectly accesses link data on VLAN interfaces (such as 4-address WDS clients) where the link structure is uninitialized, leading to kernel panic when dereferencing a NULL channel pointer. Any system with AP_VLAN wireless configurations and active channel state announcement (CSA) operations is vulnerable to local denial of service.

Out-of-bounds read in Linux kernel Bluetooth L2CAP layer allows remote attackers to read adjacent kernel memory via truncated L2CAP_INFO_RSP packets with insufficient payload length. The l2cap_information_rsp() function validates only the fixed 4-byte header but then unconditionally accesses variable-length payload fields (feat_mask at offset +4 and fixed_chan at offset +1) without verifying the payload is present, triggering kernel memory disclosure on specially crafted Bluetooth frames.

Linux kernel SMB client incorrectly reuses Kerberos authentication sessions across multiple mounts with different username options, allowing an attacker or misconfigured system to access shares using unintended credentials. The vulnerability affects CIFS/SMB mounting with Kerberos (sec=krb5) when the username mount option is specified; the kernel fails to validate that the username parameter matches the authenticated session, causing subsequent mounts to inherit the first mount's credentials rather than failing with ENOKEY when the requested principal is absent from the keytab. This is a session management flaw that enables credential confusion and potential unauthorized share access.

Memory leak in Linux kernel atmel-sha204a cryptographic driver allows local denial of service via resource exhaustion. The vulnerability occurs when memory allocation fails during tfm_count counter management; the counter is not properly decremented, causing subsequent read operations to block indefinitely. This affects all Linux kernel versions with the vulnerable atmel-sha204a implementation until patched.

Linux kernel xe (Intel GPU) driver leaks dynamically allocated virtual memory area (VMA) structures when argument validation fails in the xe_vm_madvise_ioctl handler, allowing local attackers to exhaust kernel memory and trigger denial of service. The vulnerability has been patched upstream in stable kernel branches with proper cleanup path addition.

Use-after-free vulnerability in Linux kernel SPI controller registration allows local attackers to trigger unclocked register accesses and potential information disclosure when per-CPU statistics allocation fails during controller initialization. The vulnerability affects all Linux kernel versions and is fixed via proper driver core deregistration on allocation failure; no CVSS score or active exploitation data available at time of analysis.

NULL-pointer dereference in Linux kernel SPI subsystem allows local denial of service via sysfs attribute access. The SPI controller's per-CPU statistics structure is not allocated until after the controller registers with the driver core, creating a race window where sysfs attribute reads can trigger a kernel panic. This affects all Linux kernel versions with the vulnerable SPI statistics implementation; exploitation requires local system access to read sysfs files.

Buffer overflow in Linux kernel's RedBoot partition table parser allows kernel panic during boot when CONFIG_FORTIFY_SOURCE is enabled with recent compilers. The MTD (Memory Technology Devices) subsystem reads beyond allocated buffer boundaries in partition name validation, triggering fortify-source detection and kernel crash (oops). This affects systems using RedBoot bootloader partitioning on embedded devices; exploitation is involuntary (denial of service via boot failure) rather than attacker-driven, with no public exploit code identified.

Linux kernel io_uring/poll multishot recv can hang indefinitely when a socket shutdown occurs concurrently with data reception, due to a race condition where accumulated poll wakeups are drained without consuming the persistent HUP event. The vulnerability affects all Linux kernel versions with io_uring poll support and requires a fix to explicitly check for HUP conditions and re-loop when multiple poll activations are pending.

Infinite loop in Linux kernel serial core driver handle_tx() affects systems using uninitialized PORT_UNKNOWN serial ports, where uart_write_room() and uart_write() behave inconsistently regarding null transmit buffers, causing denial of service through system hangs. The vulnerability impacts caif_serial and other drivers that rely on tty_write_room() to determine write capacity. Patch available in upstream kernel commits; no CVSS score assigned due to kernel-specific nature and relatively limited exposure scope.

Use-after-free in Linux kernel DRM subsystem when framebuffers and property blobs are dereferenced after drm_dev_unplug during device driver unload, causing kernel oops and general protection faults in drm_framebuffer_cleanup. Affects all Linux kernel versions with DRM enabled; upstream fix available via kernel commits referenced in stable tree.

Linux kernel drm/imagination driver deadlock in soft reset sequence allows local denial of service when the soft reset handler calls disable_irq() from within a threaded IRQ handler context, creating a self-deadlock condition. The fix replaces disable_irq() with disable_irq_nosync() to prevent the handler from waiting on itself. Affected systems running vulnerable kernel versions with imagination GPU drivers can experience system hangs during GPU reset operations; no public exploit code identified at time of analysis.

Linux kernel drm/imagination driver crashes when the GPU runtime PM suspend callback executes concurrently with an IRQ handler attempting to access GPU registers, causing kernel panics with SError interrupts on ARM64 platforms. The vulnerability affects the imagination GPU driver across Linux kernel versions and is triggered when power management suspend operations race with interrupt handling without proper synchronization. The fix adds synchronize_irq() calls to ensure IRQ handlers complete before GPU suspension and removes problematic runtime PM resume calls from the IRQ handler that could cause deadlocks.

Denial of service in Linux kernel amdgpu driver allows local attackers to exhaust system memory by passing an arbitrarily large BO (buffer object) list entry count via userspace, bypassing existing overflow checks but causing excessive allocation and processing delays; fixed by enforcing a 128k entry limit per BO list.

NULL pointer dereference in Linux kernel DRM i915 GPU driver allows local denial of service during system probe when DMC firmware initialization has not yet completed but hardware has DC6 power state enabled. The vulnerability occurs in intel_dmc_update_dc6_allowed_count() when called from gen9_set_dc_state() during intel_power_domains_init_hw(), which executes before DMC initialization, causing kernel oops if DC6 is unexpectedly enabled by BIOS firmware. No public exploit code identified; this is a kernel crash vulnerability requiring local system access triggered by atypical BIOS behavior.

Linux kernel DRM/xe driver fails to protect GPU memory (GGTT) MMIO access during failed driver load or asynchronous buffer object teardown, potentially enabling information disclosure or memory corruption. The vulnerability affects systems with Intel Xe graphics where the driver's hotplug-based protection mechanism does not activate if initialization fails, leaving GGTT memory accessible after the driver should have been disabled. CVSS and KEV status not available; patches have been released in upstream Linux stable branches.

Linux kernel btrfs filesystem fails to log new directory dentries when the parent directory of a conflicting inode is logged, causing new files and subdirectories to become inaccessible after power failure or system crash. The vulnerability affects all Linux kernel versions with btrfs; an attacker or system malfunction can trigger data loss through specific filesystem operation sequences involving deleted and recreated inodes with naming conflicts.

Memory leak in Linux kernel Microchip MPFS system controller driver (mpfs_sys_controller_probe) allows local attackers to exhaust kernel memory by repeatedly triggering the MTD device lookup failure path, eventually causing denial of service through memory exhaustion.

Race condition in Linux kernel QMan driver allows concurrent queue frame descriptor allocation and deallocation to corrupt internal state, causing WARN_ON triggers and potential information disclosure via stale fq_table entries. The vulnerability affects systems using Freescale/NXP QBMan queue management with dynamic FQID allocation enabled (QMAN_FQ_FLAG_DYNAMIC_FQID). No public exploit code or active exploitation confirmed; upstream fix merged via memory barrier enforcement to serialize table cleanup before FQID pool deallocation.

Use-after-free vulnerability in the Linux kernel's Bluetooth HIDP subsystem allows local attackers to trigger a kernel crash or potentially execute arbitrary code by failing to properly release L2CAP connection references when user callbacks are invoked. The flaw affects all Linux kernel versions in the CPE range and has been resolved through reference counting fixes in the L2CAP connection cleanup path; no public exploit code is currently identified, but the vulnerability requires local access to trigger via Bluetooth device manipulation.

Use-after-free in Linux kernel Bluetooth L2CAP layer allows local attackers to cause denial of service or potentially execute code via a race condition in l2cap_unregister_user(). The vulnerability arises because l2cap_register_user() and l2cap_unregister_user() access conn->users without proper locking (conn->lock), while l2cap_conn_del() protects the same structure with conn->lock, creating concurrent access to freed memory. All Linux kernel versions with Bluetooth L2CAP support are affected. Patch available via Linux stable kernel commits.

NULL pointer dereference in Linux kernel ROSE socket implementation allows local denial of service when rose_connect() is called twice during an active connection attempt. The vulnerability occurs because rose_connect() fails to validate TCP_SYN_SENT state, permitting rose->neighbour to be overwritten with NULL, which later causes a kernel crash when rose_transmit_link() dereferences the NULL pointer during socket closure. No active exploitation reported; fix available in upstream kernel commits.

Memory corruption and potential kernel freezes occur in the Linux kernel's IP tunnel implementation when VXLAN or GENEVE tunnels transmit packets, due to incorrect offset calculations in per-CPU statistics tracking on 32-bit systems. The vulnerability arises from iptunnel_xmit_stats() assuming all tunnels use NETDEV_PCPU_STAT_TSTATS, but VXLAN and GENEVE actually use NETDEV_PCPU_STAT_DSTATS with a different memory layout, causing syncp sequence counter overwrites that corrupt statistics or deadlock the kernel. Patch commits are available in the Linux kernel stable tree and address this by adapting the statistics handler and repositioning the pcpu_stat_type field to improve cache efficiency.

Use-after-free in Linux kernel netfilter ctnetlink module allows local attackers to read freed kernel memory by triggering multiple-round netlink dump operations on conntrack expectations, exploiting improper reference counting in ctnetlink_dump_exp_ct() that drops conntrack references before the dump callback completes. The vulnerability requires local network namespace access and CAP_NET_ADMIN capability but enables information disclosure of kernel heap contents via KASAN-detected slab-use-after-free on ct->ext dereference.

Integer truncation in Linux kernel netfilter SIP helper allows remote attackers to bypass Content-Length validation and cause information disclosure via malformed SIP messages. The sip_help_tcp() function stores SIP Content-Length header values (returned as unsigned long) into an unsigned int variable, causing values exceeding UINT_MAX (4,294,967,295) to truncate silently on 64-bit systems. This miscalculation causes the parser to misidentify message boundaries, treating trailing TCP segment data as additional SIP messages and passing them to the SDP parser, potentially leaking kernel memory or enabling further exploitation. Upstream patches are available across multiple stable kernel branches.

Out-of-bounds read in Linux kernel netfilter H.323/RAS packet decoding allows local or remote attackers to read 1-4 bytes beyond allocated buffer boundaries via malformed packets. The vulnerability exists in decode_int() within nf_conntrack_h323, where insufficient boundary validation before reading variable-length integer fields permits information disclosure or potential denial of service. No CVSS score or KEV status published; patch available across multiple stable kernel branches via upstream commits.

Out-of-bounds read in Linux kernel netfilter nf_conntrack_h323 DecodeQ931() function allows remote attackers to trigger a kernel memory disclosure or denial of service by sending a specially crafted H.323 packet with zero-length UserUserIE field, causing integer underflow when a 16-bit length value is decremented without validation. No public exploit code identified at time of analysis, and CVSS severity not quantified in available data.

Use-after-free in Linux kernel MANA hardware channel teardown (net/mana driver) allows concurrent interrupt handlers to dereference freed memory in mana_hwc_destroy_channel(), potentially causing NULL pointer dereference or memory corruption. The vulnerability stems from improper teardown ordering where hwc->caller_ctx is freed before CQ/EQ IRQ handlers are fully synchronized, affecting all Linux kernel versions with the MANA driver. Fixes are available across stable kernel branches via upstream commit reordering.

Memory leak in Linux kernel's TI ICSSG PRU Ethernet driver XDP_DROP path causes page pool exhaustion and out-of-memory conditions on systems using XDP packet dropping in non-zero-copy mode. The vulnerability affects all Linux kernel versions with the vulnerable icssg-prueth driver code; page recycling was incorrectly removed from the XDP_DROP handler to support AF_XDP zero-copy mode, but this created a resource leak in standard mode. No active exploitation identified; this is a kernel stability and denial-of-service issue affecting embedded and edge systems using TI PRU Ethernet hardware.

Linux kernel runtime PM subsystem contains a use-after-free race condition in pm_runtime_work() where the dev->parent pointer may be dereferenced after the parent device has been freed during device removal. This results in a KASAN-detectable memory safety violation that can trigger kernel panics or arbitrary memory access. The vulnerability affects all Linux kernel versions and is resolved by adding a flush_work() call to pm_runtime_remove() to serialize device removal with pending runtime PM work.

Infinite loop in Linux kernel bonding device header parsing allows local denial of service when two bonding devices are stacked. The bond_header_parse() function can recurse indefinitely because skb->dev always points to the top of the device hierarchy. The fix adds a bounded recursion parameter to the header_ops parse() method to ensure the leaf parse method is called and prevent endless loops. This vulnerability affects all Linux kernel versions with the vulnerable bonding code and requires local access to trigger.

NULL dereference and use-after-free in the Linux kernel's SMC (Shared Memory Communications) socket implementation occur when smc_tcp_syn_recv_sock() races with socket close operations, allowing a local attacker to trigger a kernel panic via concurrent manipulation of TCP SYN handling and SMC listen socket closure. The vulnerability affects the Linux kernel across multiple versions via the net/smc subsystem and is addressed through RCU-protected access and refcount validation rather than lock-based serialization.

Double-free memory corruption in the Linux kernel's TEQL (Trivial Link Equalizer) qdisc implementation allows local attackers to cause kernel crashes via denial of service. The vulnerability occurs when qdisc_reset is called without proper synchronization on lockless Qdisc root configurations, creating a race condition that results in use-after-free and double-free conditions in packet buffer management. This affects all Linux kernel versions with the vulnerable TEQL code path and requires local access to trigger via specially crafted packet scheduling operations.

Out-of-bounds memory read in Linux kernel USB CDC NCM driver allows local attackers to read kernel memory via malformed USB network devices. The cdc_ncm_rx_verify_ndp16() function fails to account for NDP header offset when validating DPE (Data Packet Element) array bounds, permitting buffer over-read when the NDP is positioned near the end of the network transfer block. No CVSS score, EPSS data, or active exploitation status currently available; patch available in stable kernel releases.

Buffer overflow in Linux kernel cdc_ncm driver allows out-of-bounds memory reads when processing malformed USB CDC NCM (Network Control Model) packets with NDP32 (Normal Data Packet) headers positioned near the end of the network transfer buffer. The vulnerability exists in cdc_ncm_rx_verify_ndp32() where bounds checking fails to account for the ndpoffset value when validating the DPE (Data Packet Element) array size, potentially enabling local denial-of-service or information disclosure on systems with affected USB CDC NCM network devices. No active exploitation or public proof-of-concept identified at time of analysis.

Linux kernel aqc111 USB driver deadlock in power management allows local denial of service via task hang during runtime suspend. The vulnerability occurs when aqc111_suspend() calls power-managed write operations during device suspension, triggering nested runtime PM calls that deadlock waiting for a state change that never occurs. This blocks the rtnl_lock and freezes the entire networking stack. Affected systems running vulnerable kernel versions with USB AQC111 network adapters are susceptible to local DoS that requires no authentication or user interaction beyond normal device suspension. No public exploit code identified; fix requires kernel upgrade or manual patch application.

Kernel page fault in Intel IGC network driver XDP TX timestamp handling allows local denial of service when an XDP application requesting TX timestamping shuts down while the interface link remains active. The vulnerability stems from stale xsk_meta pointers left in memory after TX ring shutdown, causing the IRQ handler to dereference invalid kernel addresses and trigger a kernel panic. This affects Linux kernel versions in the igc driver and requires no special privileges or network access, only the ability to run XDP programs on an affected system.

Memory leak in Linux kernel mac80211 subsystem's ieee80211_tx_prepare_skb() function fails to free socket buffers (skb) in one of three error paths, allowing local denial of service through memory exhaustion. The vulnerability affects all Linux kernel versions with the vulnerable code path in wireless MAC 802.11 handling; no active exploitation has been reported, but the fix addresses a resource leak that could be triggered by applications exercising error conditions in Wi-Fi frame preparation.

Use-after-free vulnerability in Linux kernel ACPI processor errata handling allows local attackers to cause denial of service or potentially execute code via device pointer dereference after reference dropping in acpi_processor_errata_piix4(). The vulnerability affects multiple Linux kernel versions and was introduced in a previous fix attempt (commit f132e089fe89); it has been resolved across stable kernel branches with no active public exploitation identified.

NULL pointer dereference in Linux kernel IPv6 SRv6 path processing allows local denial of service when __in6_dev_get() returns NULL due to missing IPv6 configuration or device unregistration. The vulnerability affects seg6_hmac_validate_skb() and ipv6_srh_rcv() functions which lacked NULL checks on the returned idev pointer, enabling a local attacker to crash the kernel by triggering these code paths on misconfigured or unregistering network devices.

Race condition in Linux kernel net/mlx5e IPSec offload driver allows concurrent access to a shared DMA-mapped ASO context, potentially causing information disclosure or incorrect IPSec processing results. The vulnerability affects systems using Mellanox MLX5 network adapters with IPSec offload functionality. An attacker with local access to initiate multiple IPSec operations in rapid succession can trigger the race condition, corrupting the shared context and causing subsequent operations to read invalid data, compromising confidentiality and integrity of IPSec-protected traffic.

Linux kernel net/mlx5e driver suffers a race condition during IPSec ESN (Extended Sequence Number) update handling that causes incorrect ESN high-order bit increments, leading to anti-replay failures and IPSec traffic halts. The vulnerability affects systems using Mellanox ConnectX adapters with IPSec full offload mode enabled. Attackers with local network access or the ability to trigger IPSec traffic patterns could exploit this to disrupt encrypted communications, though no public exploit code or active exploitation has been reported.

Linux kernel NULL pointer dereference in UDP tunnel socket creation when IPv6 is disabled causes denial of service. When CONFIG_IPV6=n, the udp_sock_create6() function incorrectly returns success (0) without creating a socket, leading callers such as fou_create() to dereference an uninitialized pointer. The vulnerability is triggered via netlink socket operations and requires privileged user access; no public exploit code or active exploitation has been identified at time of analysis.

Denial of service in Linux kernel mvpp2 network driver occurs when MTU changes or other operations trigger buffer pool switching on Marvell hardware lacking CM3 SRAM support, causing NULL pointer dereference in flow control register access. Affects systems running vulnerable kernel versions on Marvell Armada platforms where the CM3 SRAM device tree entry is absent; no authentication required. Upstream fix available via stable kernel commits.

Linux kernel net shaper module fails to validate netdev liveness during hierarchy read operations, allowing information disclosure through use-after-free conditions when a network device is unregistered while RCU-protected read operations are in progress. The vulnerability affects the netlink operation callbacks in the shaper subsystem, where references acquired during pre-callbacks are not validated before later lock/RCU acquisitions, creating a race condition that can expose kernel memory or cause denial of service. No public exploit code has been identified at time of analysis, and the issue requires local access to trigger netlink operations.

Linux kernel net shaper subsystem susceptible to race condition during hierarchy creation allows unauthenticated local attackers to leak kernel memory or trigger use-after-free conditions. The vulnerability arises when a netdev is unregistered between reference acquisition during Netlink operation preparation and subsequent lock acquisition, permitting hierarchy allocation after flush operations have completed. Fixed via upstream commits that consolidate locking to pre-callback phase, preventing concurrent write races with flush operations.

Linux kernel NULL pointer dereference in the x86 PMU NMI handler on AMD EPYC systems causes denial of service when perf event unthrottling races with PMU rescheduling. The vulnerability stems from commit 7e772a93eb61 moving event pointer initialization later in x86_pmu_enable(), allowing the unthrottle path to set active_mask bits without populating the corresponding events[] array entries, leading to NULL pointer dereference when subsequent PMC overflow interrupts fire. No public exploit code identified at time of analysis; patch fixes are available in upstream Linux kernel stable branches.

NAND flash device lock/unlock operations in the Linux kernel MTD subsystem can race with concurrent erase/write operations, causing cmd_pending conflicts on certain NAND controllers that use PIO-based SET_FEATURES. This race condition is resolved by serializing lock/unlock calls with the NAND device lock, preventing data corruption or system instability on affected controller implementations. The vulnerability affects all Linux kernel versions prior to the fix and is present in systems using raw NAND devices with specific controller hardware implementations.

Null pointer dereference in Linux kernel arm_mpam memory bandwidth monitoring causes kernel oops when an MSC supporting bandwidth monitoring transitions offline and back online. The mpam_restore_mbwu_state() function fails to initialize a value buffer before passing it to __ris_msmon_read() via IPI, triggering a crash in the bandwidth counter restoration routine. This affects ARM systems with MPAM (Memory Partitioning and Monitoring) support and results in denial of service through system instability when memory controllers are toggled.

A use-after-free vulnerability in the Linux kernel's mshv (Microsoft Hyper-V) driver allows local attackers to trigger a kernel panic by unmapping user memory after a failed mshv_map_user_memory() call. The error path incorrectly calls vfree() without unregistering the associated MMU notifier, leaving a dangling reference that fires when userspace performs subsequent memory operations. This is a memory safety issue affecting the Hyper-V virtualization subsystem in the Linux kernel.

Memory leak in the Linux kernel's Amlogic SPI controller driver (aml_spisg_probe) fails to release SPI controller resources in multiple error paths during probe, allowing local attackers to exhaust kernel memory through repeated driver load/unload cycles or failed probe attempts. The vulnerability has been resolved in the upstream kernel by converting to device-managed SPI allocation functions.

Memory leak in Linux kernel drm/vmwgfx driver caused by overwriting KMS surface dirty tracker without proper cleanup. The vulnerability affects the VMware graphics driver subsystem in the kernel, allowing local attackers to trigger memory exhaustion through repeated surface operations. No CVSS score, EPSS data, or KEV status available; fix commits exist in upstream stable kernel branches.

Linux kernel crashes in iommu_sva_unbind_device() when accessing a freed mm structure after iommu_domain_free() deallocates domain->mm->iommu_mm, causing denial of service on systems using IOMMU Shared Virtual Addressing (SVA). The fix reorders code to access the structure before the domain is freed. No CVSS score, EPSS, or KEV status available; no public exploit code identified.

Use-after-free in Linux kernel ksmbd SMB server allows local or remote attackers to read freed memory and potentially achieve denial of service or code execution via compound SMB2 requests that reuse a tree connection after it has been disconnected and its associated share_conf structure freed. The vulnerability exists because smb2_get_ksmbd_tcon() bypasses state validation checks when reusing connections in compound requests, enabling subsequent commands to dereference already-freed share_conf pointers. No CVE severity metrics are available, but KASAN confirms memory corruption is triggered in smb2_write operations during tree disconnect sequences.

Use-after-free in Linux kernel's ksmbd SMB server allows remote attackers to crash the kernel or potentially execute code via malicious SMB2 DURABLE_REQ_V2 replay operations. The vulnerability occurs when parse_durable_handle_context() unconditionally reassigns file handle connection pointers during replay operations, causing stale pointer dereferences when the reassigned connection is subsequently freed. A KASAN report confirms the use-after-free in spin_lock operations during file descriptor closure, triggered during SMB2 connection handling in the ksmbd-io workqueue. No public exploit code or active exploitation has been confirmed at time of analysis.

Linux kernel drm/logicvc driver fails to release a device node reference in logicvc_drm_config_parse(), causing a reference leak that can exhaust kernel memory resources over time. The vulnerability affects all Linux kernel versions with the logicvc DRM driver enabled; it requires local access to trigger repeated calls to the vulnerable code path. This is a low-severity resource exhaustion issue resolved via kernel patch implementing automatic cleanup attributes.

Linux kernel KVM ARM64 fails to properly initialize ID registers for non-protected pKVM guests, causing feature detection checks to incorrectly return zero and preventing proper save/restore of system registers like TCR2_EL1 during world switches, potentially leading to state corruption. The vulnerability affects the hypervisor's ability to detect CPU features in non-protected virtual machines despite the initialization flag being incorrectly set. This is a kernel-level logic error that impacts system register handling in ARM64 virtualization.

Linux kernel accel/amdxdna driver fails to validate command buffer payload count, allowing out-of-bounds reads in AMD XDNA accelerator command processing. The vulnerability affects the accel/amdxdna subsystem across unspecified Linux kernel versions and permits information disclosure through unvalidated payload size interpretation. No active exploitation, public proof-of-concept, or CVSS data currently available.

Linux kernel btrfs subsystem fails to free allocated pages in btrfs_uring_read_extent() when error conditions occur before asynchronous I/O completion, leading to memory leaks. The vulnerability affects all Linux kernel versions with the vulnerable btrfs implementation; while tagged as Information Disclosure, the primary impact is denial of service through memory exhaustion rather than data exposure. No public exploit code or active exploitation has been identified; this is a defensive fix addressing a code path that may never execute under normal conditions but represents a resource management defect.

Interrupt storm in Linux kernel dpaa2-switch driver occurs when a bounds check rejects an out-of-bounds if_id in the IRQ handler but fails to clear the interrupt status, causing repeated spurious interrupts. This denial-of-service condition affects the NXP DPAA2 Ethernet switch driver on systems running vulnerable Linux kernel versions. An attacker with the ability to trigger malformed frames or hardware state transitions could exhaust CPU resources through interrupt flooding.

Memory leak in Linux kernel DRM/XE configfs device release allows information disclosure through unfreed ctx_restore_mid_bb allocation. The xe_config_device_release() function fails to deallocate ctx_restore_mid_bb[0].cs memory that was previously allocated by wa_bb_store(), leaving sensitive kernel memory accessible when the configfs device is removed. Affected Linux kernel versions containing the vulnerable DRM/XE driver require patching to prevent potential information leakage.

Mutex lock-unlock mismatch in the Linux kernel's wlcore Wi-Fi driver allows potential information disclosure or system instability through improper synchronization. The vulnerability occurs when wl->mutex is unlocked without being locked first, detected by the Clang thread-safety analyzer. This affects all Linux kernel versions with the wlcore Wi-Fi driver. No active exploitation has been identified, but the bug creates a race condition that could be leveraged to access shared kernel state.

Linux kernel RDS TCP subsystem resolves circular locking dependency in rds_tcp_tune() function where socket lock contention with fs_reclaim memory allocation creates deadlock potential. The vulnerability affects all Linux kernel versions with the vulnerable code path in net/rds/tcp.c; the fix relocates sk_net_refcnt_upgrade() outside the socket lock critical section to eliminate the circular lock dependency without compromising synchronization semantics.

Memory leak in Linux kernel DRM/XE register save-restore (reg_sr) module fails to free allocated memory when xa_store() operation fails, potentially allowing local information disclosure or denial of service through repeated trigger of the error path. The vulnerability affects all Linux kernel versions containing the affected drm/xe/reg_sr code prior to the fix commits referenced. No CVSS score or exploit data provided; patch commits are available in upstream Linux repository.

Unblinded BPF immediate values in PROBE_MEM32 stores bypass constant hardening in the Linux kernel BPF JIT compiler when bpf_jit_harden >= 1, allowing user-controlled 32-bit immediates to leak into native code. The vulnerability affects Linux kernel versions where convert_ctx_accesses() rewrites arena pointer stores to BPF_ST|BPF_PROBE_MEM32 before constant blinding runs, but bpf_jit_blind_insn() only handles BPF_ST|BPF_MEM instructions. No public exploit code or active exploitation has been identified; the issue is a hardening bypass that could facilitate information disclosure or facilitate construction of more complex attacks against BPF programs.

Memory sealing (mseal) in the Linux kernel incorrectly tracks virtual memory area (VMA) boundaries during merge operations, causing curr_end to become stale and resulting in incorrect iteration state. This flaw in mm/mseal.c affects Linux kernel versions where the mseal feature is present, allowing local attackers to potentially bypass memory sealing protections or trigger information disclosure by manipulating VMA merge behavior during seal operations.

Use-after-free vulnerability in Linux kernel futex handling allows local attackers to read freed memory via race condition between futex_key_to_node_opt() and vma_replace_policy(). When mbind() concurrently replaces virtual memory area policies, __futex_key_to_node() may dereference a freed mempolicy structure, enabling information disclosure of kernel memory. The vulnerability requires local access and precise timing but poses memory safety risk in multi-threaded applications using futex operations alongside memory policy changes.

Linux kernel TLS subsystem leaks socket buffers (skbs) when asynchronous AEAD decryption operations fail during batch processing, allowing local attackers to exhaust kernel memory and potentially cause denial of service. The vulnerability exists in tls_decrypt_async_wait() and related functions that manage the async_hold queue, which pins encrypted input buffers for AEAD engine references; improper cleanup in failure paths leaves these buffers allocated indefinitely. This is a kernel memory leak affecting TLS decryption in the kernel's cryptographic stack, confirmed by multiple upstream patches across stable branches.

Use-after-free in Linux kernel clsact qdisc initialization and destruction rollback allows local denial of service or potential information disclosure when qdisc replacement fails midway during tcf_block_get_ext() operations. The vulnerability stems from asymmetric initialization and cleanup paths where egress_entry references from a previous clsact instance remain valid during failure scenarios, leading to double-free or use-after-free conditions. Affected Linux kernel versions across all distributions that include the clsact traffic control qdisc require patching.

Use-after-free in Linux kernel netfilter BPF hook memory management allows local attackers to read sensitive kernel memory via concurrent nfnetlink_hooks dumping operations. The vulnerability arises from premature memory release in hook structures before RCU readers complete their access, enabling information disclosure through netlink interface. No active exploitation confirmed, but the KASAN report demonstrates reliable reproducer availability.