CVE-2026-99081 is a vulnerability in the Linux kernel's bnxt_en driver affecting Broadcom BCM57412 NetXtreme-E NICs. The issue arises from incorrect DMA mapping length for small padded packets, causing the NIC DMA engine to read beyond the mapped buffer boundary. This leads to Intel IOMMU DMA read faults and transmit queue timeouts. The flaw was introduced after a commit that increased VLAN headroom, causing small packets to align near page boundaries. The fix ensures the DMA mapping length matches the padded packet size, preventing DMA overruns.
CVE-2026-98965 is a vulnerability in the Linux kernel's ICE network driver affecting version 6.18. It involves improper ordering of memory-mapped I/O (MMIO) writes for transmit doorbells on arm64 architectures, leading to a race condition that can cause kernel crashes (BUG_ON). The issue arises because writel_relaxed() does not guarantee ordering between normal memory writes and MMIO writes, potentially causing the NIC to transmit before queue accounting updates are visible. This has been fixed by restoring the use of writel() to ensure proper ordering.
CVE-2026-112112 is a low-severity vulnerability in the CZ.NIC BIRD Internet Routing Daemon affecting IPv6 FlowSpec validation. It allows an established BGP peer to bypass validation and inject FlowSpec rules that match destinations outside the authorized address space. This occurs because BIRD fails to reject a destination prefix component with a nonzero offset as required by RFC 8956 section 5, validating a reconstructed prefix with zero-filled skipped bits instead. Consequently, unauthorized traffic filtering or redirection may occur if export policies permit propagation of these rules to downstream peers.
CVE-2026-65200 is an integer underflow vulnerability in Huawei openBMC's device-management process. The flaw occurs in the parsing of PLDM-over-MCTP and NCSI-over-MCTP responses, where the code subtracts a fixed header size from a device-supplied response length without checking for lower bounds. This can cause the length to wrap around to a large value, leading to out-of-bounds reads. The vulnerability affects the base PLDM control class and Huawei NCSI driver, making it reachable by any PLDM endpoint or Huawei NIC on the internal management bus. The CVSS score is 3.5, indicating low severity. No known exploits are reported in the wild, and no patch information is provided.
This vulnerability relates to the iwlwifi driver component handling multi-link device (mld) asynchronous notifications upon a network interface card (NIC) error. The provided information is minimal and does not include technical details, impact, or exploitation data.
In the Linux kernel, the following vulnerability has been resolved: qede: sync udp_tunnel ports outside qede_lock in the recovery path A TX timeout on a qede NIC that has VXLAN/GENEVE tunnel ports configured wedges the rtnetlink control plane of the whole machine: NETDEV WATCHDOG: ens6f1 (qede): transmit queue 2 timed out 10226 ms [qede_tx_timeout:586(ens6f1)]TX timeout on queue 2! [qede_recovery_handler:2665(ens6f0)]Starting a recovery process The recovery path deadlocks on the driver's own mutex: qede_sp_task rtnl_lock() mutex_lock(&edev->qede_lock) <- taken qede_recovery_handler qede_load udp_tunnel_nic_reset_ntf __udp_tunnel_nic_device_sync info->sync_table == qede_udp_tunnel_sync mutex_lock(&edev->qede_lock) <- same task: deadlock The mutex is not recursive, so the kworker blocks on itself with rtnl_lock held, and neither lock is ever released. Every task that calls rtnl_lock() afterwards (ip, ovs-vswitchd, lldpad, IPv6 addrconf, sshd) blocks forever while the node still answers ping. In a vmcore from an affected production node rtnl_mutex.owner decodes to the very kworker blocked at the innermost mutex_lock() above. Re-sync the tunnel ports from qede_sp_task() after the internal lock is dropped, still under rtnl_lock as the udp_tunnel API requires. This mirrors qede_open(), which calls udp_tunnel_nic_reset_ntf() under rtnl without the internal lock. qede_recovery_handler() now returns whether it has successfully reloaded an open device, and the caller re-syncs the ports only in that case. This keeps the old gating exactly: a device that was down or a failed recovery returns false, as those paths never reached the udp_tunnel_nic_reset_ntf() call before either. This was the only user of the qede_lock()/qede_unlock() helpers, so remove them.
In the Linux kernel, the following vulnerability has been resolved: net: udp_tunnel: fix memory leak in udp_tunnel_nic_unregister() syzbot reported a memory leak [1] in the UDP tunnel NIC offload code. When device registration fails (e.g. in register_netdevice()), netdev core unwinds by sending a single NETDEV_UNREGISTER notification. If work was queued during NETDEV_REGISTER (utn->work_pending is set), udp_tunnel_nic_unregister() returns early: if (utn->work_pending) return; Because failed registrations do not enter netdev_wait_allrefs_any(), no subsequent NETDEV_UNREGISTER rebroadcast will ever occur. As a result, the struct udp_tunnel_nic allocated in udp_tunnel_nic_alloc() is leaked permanently. Fix this by removing the early return. Instead, synchronously cancel any pending work with cancel_delayed_work_sync() before freeing @utn. To be able to call cancel_delayed_work_sync() while holding RTNL (the work also needs RTNL), switch udp_tunnel_nic_device_sync_work() to rtnl_trylock(). If RTNL is contended, requeue the work with a 1 jiffy delay (via queue_delayed_work()) to prevent high CPU contention while waiting for RTNL lock. The utn->work_pending bookkeeping is no longer needed and is removed, as the workqueue core already tracks the pending/running state of the work. [1] BUG: memory leak unreferenced object 0xffff888127d5f840 (size 96): comm "syz-executor", pid 5806, jiffies 4294942188 backtrace (crc 99fdb6c8): __kmalloc_noprof+0x3bf/0x550 udp_tunnel_nic_alloc net/ipv4/udp_tunnel_nic.c:756 [inline] udp_tunnel_nic_register net/ipv4/udp_tunnel_nic.c:833 [inline] udp_tunnel_nic_netdevice_event+0x804/0xab0 net/ipv4/udp_tunnel_nic.c:931 notifier_call_chain+0x59/0x160 kernel/notifier.c:85 call_netdevice_notifiers_info+0x7d/0xb0 net/core/dev.c:2250 register_netdevice+0xc10/0xeb0 net/core/dev.c:11478
A race condition in the Linux kernel's mlx5e network driver caused a potential NULL pointer dereference on weakly-ordered architectures such as ARM, PowerPC, and RISC-V. This occurred because the publication of new channel statistics entries and the increment of the stats counter lacked proper memory barriers, allowing readers to observe inconsistent state. The issue was observed on BlueField-3 DPUs during NIC bringup, leading to kernel crashes. The vulnerability has been fixed by adding memory barrier helpers to ensure proper ordering of updates and visibility of channel statistics entries.
CVE-2026-72409 is a vulnerability in the Linux kernel's mvneta network driver on Marvell MPIC platforms. It occurs because the per-CPU interrupt is not re-enabled after system resume from suspend, leading to permanent masking of the interrupt. This causes the network interface card (NIC) to generate interrupts that the CPU never receives, resulting in a complete loss of network connectivity. The issue arises when suspend happens while a NAPI poll is pending, preventing the interrupt from being unmasked on resume. The vulnerability has been fixed by ensuring the per-CPU interrupt is unmasked unconditionally during resume.
CVE-2026-72065 is a critical vulnerability in the Linux kernel's mana network driver where the packet length reported by the NIC in the RX Completion Queue Entry (CQE) was not properly validated before processing. This flaw could allow an attacker to exploit incorrect packet length information supplied by the NIC, potentially leading to severe impacts on confidentiality, integrity, and availability. The vulnerability affects Linux kernel version 5.13 and has been fixed in subsequent commits.
In the Linux kernel, the following vulnerability has been resolved: usb: core: port: Deattach Type-C connector on component unbind connector_unbind() is the mirror of connector_bind(), but it is missing the symmetric call to typec_deattach() that connector_bind() makes via: if (port_dev->child) typec_attach(port_dev->connector, &port_dev->child->dev); When a Thunderbolt dock is unplugged, two teardown paths race: 1. The component framework calls connector_unbind() first, which sets port_dev->connector = NULL without calling typec_deattach(). This leaves port->usb2_dev/port->usb3_dev in struct typec_port pointing at the USB device that is about to be freed. 2. usb_disconnect() then calls typec_deattach(port_dev->connector, ...), but port_dev->connector is already NULL, so the call is a no-op and port->usb2_dev is never cleared. 3. Concurrently, UCSI detects a PD partner-disconnect event and calls typec_unregister_partner(), which reads port->usb2_dev (now a dangling pointer to freed memory) and passes it to typec_partner_unlink_device() -> sysfs_remove_link() -> dev_name() on the freed device, corrupting the typec/UCSI partner state. This corruption leaves the Thunderbolt tunnel in an inconsistent state on the next dock hot-plug. On affected hardware the dock's I225/igc NIC fails to enumerate: AER fires a slot reset while the igc driver is still initialising ("PCIe link lost"), and the subsequent igc_reset attempt hits igc_rd32 on an already-detached device: igc 0000:2e:00.0 eth0: PCIe link lost, device now detached igc: Failed to read reg 0x0! WARNING: CPU: 9 PID: 129 at drivers/net/ethernet/intel/igc/igc_main.c:7005 igc_rd32+0xa4/0xc0 [igc] Call Trace: igc_disable_pcie_master+0x16/0xa0 [igc] igc_reset_hw_base+0x14/0x170 [igc] igc_reset+0x63/0x110 [igc] igc_io_slot_reset+0x9e/0xd0 [igc] report_slot_reset+0x5d/0xc0 pcie_do_recovery+0x209/0x400 aer_isr_one_error_type+0x235/0x430 aer_isr+0x4e/0x80 irq_thread+0xf4/0x1f0 4. UCSI later handles the PD partner-disconnect and calls typec_unregister_partner(), which still sees the stale port->usb2_dev and tries to remove its sysfs link a second time: kernfs: can not remove 'typec', no directory WARNING: CPU: 6 PID: 55 at fs/kernfs/dir.c:1706 kernfs_remove_by_name_ns+0xe9/0xf0 Workqueue: events ucsi_handle_connector_change [typec_ucsi] Call Trace: sysfs_remove_link+0x19/0x50 typec_unregister_partner+0x6e/0x120 [typec] ucsi_unregister_partner+0x107/0x150 [typec_ucsi] ucsi_handle_connector_change+0x3ec/0x490 [typec_ucsi] process_one_work+0x18e/0x3e0 worker_thread+0x2e3/0x420 kthread+0x10a/0x230 ret_from_fork+0x121/0x140 ret_from_fork_asm+0x1a/0x30 With worse timing the same stale pointer is dereferenced after the backing memory is freed, turning the warning into a use-after-free. Fix the asymmetry: call typec_deattach() before clearing port_dev->connector, matching what connector_bind() does on the bind side. typec_partner_deattach() is already protected by port->partner_link_lock, so it serialises safely with the concurrent typec_unregister_partner() path.
A vulnerability in the Linux kernel's Airoha network driver was resolved by adding missing reset operations in the airoha_qdma_cleanup_tx_queue() function. This fix ensures that DMA TX descriptors and the TX_DMA_IDX index are properly reset to notify the NIC that the QDMA TX ring is empty. The issue affects multiple Linux kernel versions and has been addressed in official updates.