In the Linux kernel, the following vulnerability has been resolved: iommu/vt-d: Skip dev-iotlb flush for inaccessible PCIe device without scalable mode PCIe endpoints with ATS enabled and passed through to userspace (e.g., QEMU, DPDK) can hard-lock the host when their link drops, either by surprise removal or by a link fault. Commit 4fc82cd907ac ("iommu/vt-d: Don't issue ATS Invalidation request when device is disconnected") adds pci_dev_is_disconnected() to devtlb_invalidation_with_pasid() so ATS invalidation is skipped only when the device is being safely removed, but it applies only when Intel IOMMU scalable mode is enabled. With scalable mode disabled or unsupported, a system hard-lock occurs when a PCIe endpoint's link drops because the Intel IOMMU waits indefinitely for an ATS invalidation that cannot complete. Call Trace: qi_submit_sync qi_flush_dev_iotlb __context_flush_dev_iotlb.part.0 domain_context_clear_one_cb pci_for_each_dma_alias device_block_translation blocking_domain_attach_dev iommu_deinit_device __iommu_group_remove_device iommu_release_device iommu_bus_notifier blocking_notifier_call_chain bus_notify device_del pci_remove_bus_device pci_stop_and_remove_bus_device pciehp_unconfigure_device pciehp_disable_slot pciehp_handle_presence_or_link_change pciehp_ist Commit 81e921fd3216 ("iommu/vt-d: Fix NULL domain on device release") adds intel_pasid_teardown_sm_context() to intel_iommu_release_device(), which calls qi_flush_dev_iotlb() and can also hard-lock the system when a PCIe endpoint's link drops. Call Trace: qi_submit_sync qi_flush_dev_iotlb __context_flush_dev_iotlb.part.0 intel_context_flush_no_pasid device_pasid_table_teardown pci_pasid_table_teardown pci_for_each_dma_alias intel_pasid_teardown_sm_context intel_iommu_release_device iommu_deinit_device __iommu_group_remove_device iommu_release_device iommu_bus_notifier blocking_notifier_call_chain bus_notify device_del pci_remove_bus_device pci_stop_and_remove_bus_device pciehp_unconfigure_device pciehp_disable_slot pciehp_handle_presence_or_link_change pciehp_ist Sometimes the endpoint loses connection without a link-down event (e.g., due to a link fault); killing the process (virsh destroy) then hard-locks the host. Call Trace: qi_submit_sync qi_flush_dev_iotlb __context_flush_dev_iotlb.part.0 domain_context_clear_one_cb pci_for_each_dma_alias device_block_translation blocking_domain_attach_dev __iommu_attach_device __iommu_device_set_domain __iommu_group_set_domain_internal iommu_detach_group vfio_iommu_type1_detach_group vfio_group_detach_container vfio_group_fops_release __fput pci_dev_is_disconnected() only covers safe-removal paths; pci_device_is_present() tests accessibility by reading vendor/device IDs and internally calls pci_dev_is_disconnected(). On a ConnectX-5 (8 GT/s, x2) this costs ~70 µs. Since __context_flush_dev_iotlb() is only called on {attach,release}_dev paths (not hot), add pci_device_is_present() there to skip inaccessible devices and avoid the hard-lock.

The remote Oracle Linux 10 / 9 host has packages installed that are affected by multiple vulnerabilities as referenced in the ELSA-2026-50372 advisory.

    - net/sched: fix pedit partial COW leading to page cache corruption (Rajat Gupta)   {CVE-2026-46331}
    - eventpoll: fix ep_remove struct eventpoll / struct file UAF (Christian Brauner)  [Orabug: 39668743]     {CVE-2026-46242}
    - net: skbuff: fix missing zerocopy reference in pskb_carve helpers (Minh Nguyen)  [Orabug: 39656679]     {CVE-2026-52943}
    - arm64: errata: Mitigate TLBI errata on various Arm CPUs (Mark Rutland)  [Orabug: 39548792]     {CVE-2025-10263}
    - openvswitch: vport: fix self-deadlock on release of tunnel ports (Ilya Maximets) {CVE-2026-46165}
    - iommu/amd: serialize sequence allocation under concurrent TLB invalidations (Ankit Soni)     {CVE-2026-43220}
    - usb: ulpi: fix memory leak on ulpi_register() error paths (Felix Gu) {CVE-2026-46109}
    - smb: client: reject userspace cifs.spnego descriptions (Asim Viladi Oglu Manizada)  [Orabug: 39463671]     {CVE-2026-46243}
    - net: skbuff: propagate shared-frag marker through frag-transfer helpers (Hyunwoo Kim) [Orabug:
    39368826,39441323] {CVE-2026-43503,CVE-2026-46300}
    - net: skbuff: preserve shared-frag marker during coalescing (William Bowling) [Orabug: 39368826]     {CVE-2026-46300}
    - netfilter: reject zero shift in nft_bitwise (Kai Ma) [Orabug: 39452464] {CVE-2026-46101}
    - net: ipv6: fix NOREF dst use in seg6 and rpl lwtunnels (Andrea Mayer) [Orabug: 39452457]     {CVE-2026-46099}
    - ALSA: caiaq: fix usb_dev refcount leak on probe failure (Deepanshu Kartikey) [Orabug: 39452739]     {CVE-2026-46048}
    - ipv6: rpl: reserve mac_len headroom when recompressed SRH grows (Greg Kroah-Hartman) [Orabug: 39426003]     {CVE-2026-43501}
    - ALSA: caiaq: Fix potentially leftover ep1_in_urb at error path (Takashi Iwai) [Orabug: 39452746]     {CVE-2026-45992}
    - net: bonding: fix use-after-free in bond_xmit_broadcast() (Xiang Mei) [Orabug: 39205989]     {CVE-2026-31419}
    - crypto: authencesn - reject short ahash digests during instance creation (Yucheng Lu) [Orabug: 39452231]     {CVE-2026-46033}
    - spi: fix resource leaks on device setup failure (Johan Hovold) [Orabug: 39452400] {CVE-2026-46083}
    - net: qrtr: ns: Limit the total number of nodes (Manivannan Sadhasivam) [Orabug: 39452123]     {CVE-2026-46003}
    - net: bridge: use a stable FDB dst snapshot in RCU readers (Zhengchuan Liang) [Orabug: 39452411]     {CVE-2026-46086}
    - net: qrtr: ns: Limit the maximum number of lookups (Manivannan Sadhasivam) [Orabug: 39452207]     {CVE-2026-46026}
    - net: qrtr: ns: Limit the maximum server registration per node (Manivannan Sadhasivam) [Orabug: 39410851]     {CVE-2026-43491}
    - RDMA/mana_ib: Disable RX steering on RSS QP destroy (Long Li) [Orabug: 39452406] {CVE-2026-46084}
    - media: rc: igorplugusb: heed coherency rules (Oliver Neukum) [Orabug: 39452432] {CVE-2026-46091}
    - wifi: mwifiex: fix use-after-free in mwifiex_adapter_cleanup() (Daniel Hodges) [Orabug: 39452343]     {CVE-2026-46069}
    - thermal: core: Fix thermal zone governor cleanup issues (Rafael J. Wysocki) [Orabug: 39452186]     {CVE-2026-46021}
    - smb: client: validate the whole DACL before rewriting it in cifsacl (Michael Bommarito) [Orabug:
    39300611] {CVE-2026-31709}
    - scsi: sd: fix missing put_disk() when device_add(&disk_dev) fails (Yang Xiuwei) [Orabug: 39452100]     {CVE-2026-45997}
    - rtmutex: Use waiter::task instead of current in remove_waiter() (Keenan Dong) [Orabug: 39425998]     {CVE-2026-43499}
    - ceph: only d_add() negative dentries when they are unhashed (Max Kellermann) [Orabug: 39452291]     {CVE-2026-46052}
    - dm mirror: fix integer overflow in create_dirty_log() (Junrui Luo) [Orabug: 39452196] {CVE-2026-46023}
    - Bluetooth: hci_event: fix potential UAF in SSP passkey handlers (Shuvam Pandey) [Orabug: 39452304]     {CVE-2026-46056}
    - tcp: call sk_data_ready() after listener migration (Zhenzhong Wu) [Orabug: 39452159] {CVE-2026-46015}
    - inotify: fix watch count leak when fsnotify_add_inode_mark_locked() fails (Chia-Ming Chang) [Orabug:
    39452250] {CVE-2026-46040}
    - md/raid5: validate payload size before accessing journal metadata (Junrui Luo) [Orabug: 39452349]     {CVE-2026-46070}
    - md/raid5: fix soft lockup in retry_aligned_read() (Chia-Ming Chang) [Orabug: 39452287] {CVE-2026-46051}
    - ext4: fix missing brelse() in ext4_xattr_inode_dec_ref_all() (Sohei Koyama) [Orabug: 39452269]     {CVE-2026-46046}
    - ext4: fix bounds check in check_xattrs() to prevent out-of-bounds access (Deepanshu Kartikey) [Orabug:
    39452442] {CVE-2026-46094}
    - io_uring/poll: fix multishot recv missing EOF on wakeup race (Jens Axboe) {CVE-2026-23473}
    - KVM: nSVM: Raise #UD if unhandled VMMCALL isn't intercepted by L1 (Kevin Cheng) [Orabug: 39452372]     {CVE-2026-46076}
    - KVM: SVM: Inject #UD for INVLPGA if EFER.SVME=0 (Kevin Cheng) [Orabug: 39452394] {CVE-2026-46082}
    - KVM: nSVM: Sync interrupt shadow to cached vmcb12 after VMRUN of L2 (Yosry Ahmed) [Orabug: 39452061]     {CVE-2026-45987}
    - xfs: fix a resource leak in xfs_alloc_buftarg() (Haoxiang Li) [Orabug: 39452131] {CVE-2026-46005}
    - libceph: Prevent potential null-ptr-deref in ceph_handle_auth_reply() (Raphael Zimmer) [Orabug:
    39452201] {CVE-2026-46024}
    - ipv4: icmp: validate reply type before using icmp_pointers (Ruide Cao) [Orabug: 39452243]     {CVE-2026-46037}
    - RDMA/rxe: Validate pad and ICRC before payload_size() in rxe_rcv (Hkbinbin) [Orabug: 39452259]     {CVE-2026-46043}
    - net: qrtr: ns: Free the node during ctrl_cmd_bye() (Manivannan Sadhasivam) [Orabug: 39452246]     {CVE-2026-46038}
    - rxrpc: Fix rxkad crypto unalignment handling (David Howells) [Orabug: 39452728] {CVE-2026-46085}
    - ALSA: caiaq: Handle probe errors properly (Takashi Iwai) [Orabug: 39452126] {CVE-2026-46004}
    - rbd: fix null-ptr-deref when device_add_disk() fails (Dawei Feng) [Orabug: 39452385] {CVE-2026-46079}
    - md/raid10: fix deadlock with check operation and nowait requests (Josh Hunt) [Orabug: 39452284]     {CVE-2026-46050}
    - jbd2: fix deadlock in jbd2_journal_cancel_revoke() (Zhang Yi) [Orabug: 39452318] {CVE-2026-46061}
    - erofs: fix the out-of-bounds nameoff handling for trailing dirents (Gao Xiang) [Orabug: 39452380]     {CVE-2026-46078}
    - ALSA: ctxfi: Add fallback to default RSR for S/PDIF (Harin Lee) [Orabug: 39452280] {CVE-2026-46049}
    - ext2: reject inodes with zero i_nlink and valid mode in ext2_iget() (Vasiliy Kovalev) [Orabug: 39452119]     {CVE-2026-46002}
    - net: qrtr: ns: Fix use-after-free in driver remove() (Manivannan Sadhasivam) [Orabug: 39452274]     {CVE-2026-46047}
    - net: strparser: fix skb_head leak in strp_abort_strp() (Luxiao Xu) [Orabug: 39452468] {CVE-2026-46102}
    - ALSA: control: Validate buf_len before strnlen() in snd_ctl_elem_init_enum_names() (Ziqing Chen)     [Orabug: 39452416] {CVE-2026-46088}
    - crypto: pcrypt - Fix handling of MAY_BACKLOG requests (Herbert Xu) [Orabug: 39410863] {CVE-2026-43493}
    - wifi: rtw88: check for PCI upstream bridge existence (Fedor Pchelkin) [Orabug: 39452437]     {CVE-2026-46092}
    - zram: do not forget to endio for partial discard requests (Sergey Senozhatsky) [Orabug: 39452420]     {CVE-2026-46089}
    - ocfs2: split transactions in dio completion to avoid credit exhaustion (Heming Zhao) [Orabug: 39452388]     {CVE-2026-46080}
    - drm/amdgpu: Limit BO list entry count to prevent resource exhaustion (Jesse Zhang) [Orabug: 39167551]     {CVE-2026-23468}
    - padata: Fix pd UAF once and for all (Herbert Xu) [Orabug: 38335055] {CVE-2025-38584}
    - drm/nouveau: fix u32 overflow in pushbuf reloc bounds check (Greg Kroah-Hartman) [Orabug: 39452133]     {CVE-2026-46006}
    - ALSA: usb-audio: stop parsing UAC2 rates at MAX_NR_RATES (Cassio Gabriel) [Orabug: 39452170]     {CVE-2026-46018}
    - Buffer overflow in drivers/xen/sys-hypervisor.c (Juergen Gross) [Orabug: 39305898] {CVE-2026-31786}
    - xen/privcmd: fix double free via VMA splitting (Juergen Gross) [Orabug: 39305908] {CVE-2026-31787}
    - rxrpc: Fix missing validation of ticket length in non-XDR key preparsing (Anderson Nascimento) [Orabug:
    39300563] {CVE-2026-31696}
    - crypto: ccp: Don't attempt to copy ID to userspace if PSP command failed (Sean Christopherson)     {CVE-2026-31697}
    - crypto: ccp: Don't attempt to copy PDH cert to userspace if PSP command failed (Sean Christopherson)     {CVE-2026-31698}
    - crypto: ccp: Don't attempt to copy CSR to userspace if PSP command failed (Sean Christopherson)     {CVE-2026-31699}
    - net/packet: fix TOCTOU race on mmap'd vnet_hdr in tpacket_snd() (Bingquan Chen) {CVE-2026-31700}
    - ALSA: caiaq: take a reference on the USB device in create_card() (Berk Cem Goksel) [Orabug: 39300586]     {CVE-2026-31701}
    - smb: client: fix OOB read in smb2_ioctl_query_info QUERY_INFO path (Michael Bommarito) [Orabug:
    39300607] {CVE-2026-31708}
    - smb: client: require a full NFS mode SID before reading mode bits (Michael Bommarito) [Orabug: 39343707]     {CVE-2026-43350}
    - fuse: reject oversized dirents in page cache (Samuel Page) [Orabug: 39300556] {CVE-2026-31694}
    - mm/pagewalk: fix race between concurrent split and refault (Max Boone) [Orabug: 39250782]     {CVE-2026-31456}
    - wifi: mac80211: always free skb on ieee80211_tx_prepare_skb() failure (Felix Fietkau) [Orabug: 39167472]     {CVE-2026-23444}
    - mm/userfaultfd: fix hugetlb fault mutex hash calculation (Jianhui Zhou) [Orabug: 39273453]     {CVE-2026-31575}
    - ipv6: add NULL checks for idev in SRv6 paths (Heminhong) [Orabug: 39167467] {CVE-2026-23442}
    - ocfs2: fix out-of-bounds write in ocfs2_write_end_inline (Joseph Qi) [Orabug: 39331090] {CVE-2026-43075}
    - ocfs2: validate inline data i_size during inode read (Deepanshu Kartikey) [Orabug: 39331097]     {CVE-2026-43076}
    - media: as102: fix to not free memory after the device is registered in as102_usb_probe() (Jeongjun Park)     [Orabug: 39273467] {CVE-2026-31578}
    - bcache: fix cached_dev.sb_bio use-after-free and crash (Mingzhe Zou) [Orabug: 39273481] {CVE-2026-31580}
    - ALSA: 6fire: fix use-after-free on disconnect (Berk Cem Goksel) [Orabug: 39273486] {CVE-2026-31581}
    - media: em28xx: fix use-after-free in em28xx_v4l2_open() (Abhishek Kumar) [Orabug: 39273493]     {CVE-2026-31583}
    - mm: blk-cgroup: fix use-after-free in cgwb_release_workfn() (Breno Leitao) [Orabug: 39273507]     {CVE-2026-31586}
    - KVM: x86: Use scratch field in MMIO fragment to hold small write values (Sean Christopherson) [Orabug:
    39273521] {CVE-2026-31588}
    - x86-64: rename misleadingly named '__copy_user_nocache()' function (Linus Torvalds) [Orabug: 39323162]     {CVE-2026-43073}
    - kernel: be more careful about dup_mmap() failures and uprobe registering (Liam R. Howlett)     {CVE-2025-21709}
    - gpiolib: fix race condition for gdev->srcu (Pawel Narewski) [Orabug: 38887677] {CVE-2026-22986}
    - KVM: SEV: Drop WARN on large size for KVM_MEMORY_ENCRYPT_REG_REGION (Sean Christopherson) [Orabug:
    39273529] {CVE-2026-31590}
    - KVM: SEV: Reject attempts to sync VMSA of an already-launched/encrypted vCPU (Sean Christopherson)     [Orabug: 39273553] {CVE-2026-31593}
    - ocfs2: handle invalid dinode in ocfs2_group_extend (Zhengyuan Huang) [Orabug: 39273569] {CVE-2026-31596}
    - ocfs2: fix use-after-free in ocfs2_fault() when VM_FAULT_RETRY (Tejas Bharambe) [Orabug: 39273578]     {CVE-2026-31597}
    - ocfs2: fix possible deadlock between unlink and dio_end_io_write (Joseph Qi) [Orabug: 39273586]     {CVE-2026-31598}
    - dcache: Limit the minimal number of bucket to two (Zhihao Cheng) [Orabug: 39323133] {CVE-2026-43071}
    - ALSA: ctxfi: Limit PTP to a single page (Harin Lee) [Orabug: 39273608] {CVE-2026-31602}
    - wifi: rtw88: fix device leak on probe failure (Johan Hovold) [Orabug: 39273620] {CVE-2026-31604}

Tenable has extracted the preceding description block directly from the Oracle Linux security advisory.

Note that Nessus has not tested for these issues but has instead relied only on the application's self-reported version number.

The Linux/Unix host has one or more packages installed that are impacted by a vulnerability without a vendor supplied patch available.

  - iommu/vt-d: Skip dev-iotlb flush for inaccessible PCIe device without scalable mode PCIe endpoints with     ATS enabled and passed through to userspace (e.g., QEMU, DPDK) can hard-lock the host when their link     drops, either by surprise removal or by a link fault. Commit 4fc82cd907ac (iommu/vt-d: Don't issue ATS     Invalidation request when device is disconnected) adds pci_dev_is_disconnected() to     devtlb_invalidation_with_pasid() so ATS invalidation is skipped only when the device is being safely     removed, but it applies only when Intel IOMMU scalable mode is enabled. With scalable mode disabled or     unsupported, a system hard-lock occurs when a PCIe endpoint's link drops because the Intel IOMMU waits     indefinitely for an ATS invalidation that cannot complete. Call Trace: qi_submit_sync qi_flush_dev_iotlb
    __context_flush_dev_iotlb.part.0 domain_context_clear_one_cb pci_for_each_dma_alias     device_block_translation blocking_domain_attach_dev iommu_deinit_device __iommu_group_remove_device     iommu_release_device iommu_bus_notifier blocking_notifier_call_chain bus_notify device_del     pci_remove_bus_device pci_stop_and_remove_bus_device pciehp_unconfigure_device pciehp_disable_slot     pciehp_handle_presence_or_link_change pciehp_ist Commit 81e921fd3216 (iommu/vt-d: Fix NULL domain on     device release) adds intel_pasid_teardown_sm_context() to intel_iommu_release_device(), which calls     qi_flush_dev_iotlb() and can also hard-lock the system when a PCIe endpoint's link drops. Call Trace:
    qi_submit_sync qi_flush_dev_iotlb __context_flush_dev_iotlb.part.0 intel_context_flush_no_pasid     device_pasid_table_teardown pci_pasid_table_teardown pci_for_each_dma_alias     intel_pasid_teardown_sm_context intel_iommu_release_device iommu_deinit_device __iommu_group_remove_device     iommu_release_device iommu_bus_notifier blocking_notifier_call_chain bus_notify device_del     pci_remove_bus_device pci_stop_and_remove_bus_device pciehp_unconfigure_device pciehp_disable_slot     pciehp_handle_presence_or_link_change pciehp_ist Sometimes the endpoint loses connection without a link-     down event (e.g., due to a link fault); killing the process (virsh destroy) then hard-locks the host. Call     Trace: qi_submit_sync qi_flush_dev_iotlb __context_flush_dev_iotlb.part.0 domain_context_clear_one_cb     pci_for_each_dma_alias device_block_translation blocking_domain_attach_dev __iommu_attach_device
    __iommu_device_set_domain __iommu_group_set_domain_internal iommu_detach_group     vfio_iommu_type1_detach_group vfio_group_detach_container vfio_group_fops_release __fput     pci_dev_is_disconnected() only covers safe-removal paths; pci_device_is_present() tests accessibility by     reading vendor/device IDs and internally calls pci_dev_is_disconnected(). On a ConnectX-5 (8 GT/s, x2)     this costs ~70 s. Since __context_flush_dev_iotlb() is only called on {attach,release}_dev paths (not     hot), add pci_device_is_present() there to skip inaccessible devices and avoid the hard-lock.
    (CVE-2026-43161)

Note that Nessus relies on the presence of the package as reported by the vendor.

ID	Name	Product	Family	Severity
329305	openSUSE 16: cluster-md-kmp-64kb / cluster-md-kmp-azure / cluster-md-kmp-default / etc (openSUSE-SU-2026:21388-1)	Nessus	SuSE Local Security Checks	high
325292	Oracle Linux 10 / 9 : Unbreakable Enterprise kernel (ELSA-2026-50372)	Nessus	Oracle Linux Local Security Checks	high
312724	Linux Distros Unpatched Vulnerability : CVE-2026-43161	Nessus	Misc.	medium
311340	Amazon Linux 2023 : bpftool6.18, kernel6.18, kernel6.18-devel (ALAS2023-2026-1596)	Nessus	Amazon Linux Local Security Checks	high
305231	Amazon Linux 2023 : bpftool6.12, kernel6.12, kernel6.12-devel (ALAS2023-2026-1543)	Nessus	Amazon Linux Local Security Checks	high

Detections

Analytics

CVE-2026-43161

medium

Tenable Plugins

high

high

medium

high

high