GHSA-68JG-6V9H-662H
Vulnerability from github – Published: 2026-02-14 18:30 – Updated: 2026-02-14 18:30In the Linux kernel, the following vulnerability has been resolved:
KVM: Don't clobber irqfd routing type when deassigning irqfd
When deassigning a KVM_IRQFD, don't clobber the irqfd's copy of the IRQ's routing entry as doing so breaks kvm_arch_irq_bypass_del_producer() on x86 and arm64, which explicitly look for KVM_IRQ_ROUTING_MSI. Instead, to handle a concurrent routing update, verify that the irqfd is still active before consuming the routing information. As evidenced by the x86 and arm64 bugs, and another bug in kvm_arch_update_irqfd_routing() (see below), clobbering the entry type without notifying arch code is surprising and error prone.
As a bonus, checking that the irqfd is active provides a convenient location for documenting why KVM must not consume the routing entry for an irqfd that is in the process of being deassigned: once the irqfd is deleted from the list (which happens before the eventfd is detached), it will no longer receive updates via kvm_irq_routing_update(), and so KVM could deliver an event using stale routing information (relative to KVM_SET_GSI_ROUTING returning to userspace).
As an even better bonus, explicitly checking for the irqfd being active fixes a similar bug to the one the clobbering is trying to prevent: if an irqfd is deactivated, and then its routing is changed, kvm_irq_routing_update() won't invoke kvm_arch_update_irqfd_routing() (because the irqfd isn't in the list). And so if the irqfd is in bypass mode, IRQs will continue to be posted using the old routing information.
As for kvm_arch_irq_bypass_del_producer(), clobbering the routing type results in KVM incorrectly keeping the IRQ in bypass mode, which is especially problematic on AMD as KVM tracks IRQs that are being posted to a vCPU in a list whose lifetime is tied to the irqfd.
Without the help of KASAN to detect use-after-free, the most common sympton on AMD is a NULL pointer deref in amd_iommu_update_ga() due to the memory for irqfd structure being re-allocated and zeroed, resulting in irqfd->irq_bypass_data being NULL when read by avic_update_iommu_vcpu_affinity():
BUG: kernel NULL pointer dereference, address: 0000000000000018 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 40cf2b9067 P4D 40cf2b9067 PUD 408362a067 PMD 0 Oops: Oops: 0000 [#1] SMP CPU: 6 UID: 0 PID: 40383 Comm: vfio_irq_test Tainted: G U W O 6.19.0-smp--5dddc257e6b2-irqfd #31 NONE Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.78.2-0 09/05/2025 RIP: 0010:amd_iommu_update_ga+0x19/0xe0 Call Trace: avic_update_iommu_vcpu_affinity+0x3d/0x90 [kvm_amd] __avic_vcpu_load+0xf4/0x130 [kvm_amd] kvm_arch_vcpu_load+0x89/0x210 [kvm] vcpu_load+0x30/0x40 [kvm] kvm_arch_vcpu_ioctl_run+0x45/0x620 [kvm] kvm_vcpu_ioctl+0x571/0x6a0 [kvm] __se_sys_ioctl+0x6d/0xb0 do_syscall_64+0x6f/0x9d0 entry_SYSCALL_64_after_hwframe+0x4b/0x53 RIP: 0033:0x46893b ---[ end trace 0000000000000000 ]---
If AVIC is inhibited when the irfd is deassigned, the bug will manifest as list corruption, e.g. on the next irqfd assignment.
list_add corruption. next->prev should be prev (ffff8d474d5cd588), but was 0000000000000000. (next=ffff8d8658f86530). ------------[ cut here ]------------ kernel BUG at lib/list_debug.c:31! Oops: invalid opcode: 0000 [#1] SMP CPU: 128 UID: 0 PID: 80818 Comm: vfio_irq_test Tainted: G U W O 6.19.0-smp--f19dc4d680ba-irqfd #28 NONE Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.78.2-0 09/05/2025 RIP: 0010:__list_add_valid_or_report+0x97/0xc0 Call Trace: avic_pi_update_irte+0x28e/0x2b0 [kvm_amd] kvm_pi_update_irte+0xbf/0x190 [kvm] kvm_arch_irq_bypass_add_producer+0x72/0x90 [kvm] irq_bypass_register_consumer+0xcd/0x170 [irqbypa ---truncated---
{
"affected": [],
"aliases": [
"CVE-2026-23198"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-02-14T17:15:57Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nKVM: Don\u0027t clobber irqfd routing type when deassigning irqfd\n\nWhen deassigning a KVM_IRQFD, don\u0027t clobber the irqfd\u0027s copy of the IRQ\u0027s\nrouting entry as doing so breaks kvm_arch_irq_bypass_del_producer() on x86\nand arm64, which explicitly look for KVM_IRQ_ROUTING_MSI. Instead, to\nhandle a concurrent routing update, verify that the irqfd is still active\nbefore consuming the routing information. As evidenced by the x86 and\narm64 bugs, and another bug in kvm_arch_update_irqfd_routing() (see below),\nclobbering the entry type without notifying arch code is surprising and\nerror prone.\n\nAs a bonus, checking that the irqfd is active provides a convenient\nlocation for documenting _why_ KVM must not consume the routing entry for\nan irqfd that is in the process of being deassigned: once the irqfd is\ndeleted from the list (which happens *before* the eventfd is detached), it\nwill no longer receive updates via kvm_irq_routing_update(), and so KVM\ncould deliver an event using stale routing information (relative to\nKVM_SET_GSI_ROUTING returning to userspace).\n\nAs an even better bonus, explicitly checking for the irqfd being active\nfixes a similar bug to the one the clobbering is trying to prevent: if an\nirqfd is deactivated, and then its routing is changed,\nkvm_irq_routing_update() won\u0027t invoke kvm_arch_update_irqfd_routing()\n(because the irqfd isn\u0027t in the list). And so if the irqfd is in bypass\nmode, IRQs will continue to be posted using the old routing information.\n\nAs for kvm_arch_irq_bypass_del_producer(), clobbering the routing type\nresults in KVM incorrectly keeping the IRQ in bypass mode, which is\nespecially problematic on AMD as KVM tracks IRQs that are being posted to\na vCPU in a list whose lifetime is tied to the irqfd.\n\nWithout the help of KASAN to detect use-after-free, the most common\nsympton on AMD is a NULL pointer deref in amd_iommu_update_ga() due to\nthe memory for irqfd structure being re-allocated and zeroed, resulting\nin irqfd-\u003eirq_bypass_data being NULL when read by\navic_update_iommu_vcpu_affinity():\n\n BUG: kernel NULL pointer dereference, address: 0000000000000018\n #PF: supervisor read access in kernel mode\n #PF: error_code(0x0000) - not-present page\n PGD 40cf2b9067 P4D 40cf2b9067 PUD 408362a067 PMD 0\n Oops: Oops: 0000 [#1] SMP\n CPU: 6 UID: 0 PID: 40383 Comm: vfio_irq_test\n Tainted: G U W O 6.19.0-smp--5dddc257e6b2-irqfd #31 NONE\n Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE\n Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.78.2-0 09/05/2025\n RIP: 0010:amd_iommu_update_ga+0x19/0xe0\n Call Trace:\n \u003cTASK\u003e\n avic_update_iommu_vcpu_affinity+0x3d/0x90 [kvm_amd]\n __avic_vcpu_load+0xf4/0x130 [kvm_amd]\n kvm_arch_vcpu_load+0x89/0x210 [kvm]\n vcpu_load+0x30/0x40 [kvm]\n kvm_arch_vcpu_ioctl_run+0x45/0x620 [kvm]\n kvm_vcpu_ioctl+0x571/0x6a0 [kvm]\n __se_sys_ioctl+0x6d/0xb0\n do_syscall_64+0x6f/0x9d0\n entry_SYSCALL_64_after_hwframe+0x4b/0x53\n RIP: 0033:0x46893b\n \u003c/TASK\u003e\n ---[ end trace 0000000000000000 ]---\n\nIf AVIC is inhibited when the irfd is deassigned, the bug will manifest as\nlist corruption, e.g. on the next irqfd assignment.\n\n list_add corruption. next-\u003eprev should be prev (ffff8d474d5cd588),\n but was 0000000000000000. (next=ffff8d8658f86530).\n ------------[ cut here ]------------\n kernel BUG at lib/list_debug.c:31!\n Oops: invalid opcode: 0000 [#1] SMP\n CPU: 128 UID: 0 PID: 80818 Comm: vfio_irq_test\n Tainted: G U W O 6.19.0-smp--f19dc4d680ba-irqfd #28 NONE\n Tainted: [U]=USER, [W]=WARN, [O]=OOT_MODULE\n Hardware name: Google, Inc. Arcadia_IT_80/Arcadia_IT_80, BIOS 34.78.2-0 09/05/2025\n RIP: 0010:__list_add_valid_or_report+0x97/0xc0\n Call Trace:\n \u003cTASK\u003e\n avic_pi_update_irte+0x28e/0x2b0 [kvm_amd]\n kvm_pi_update_irte+0xbf/0x190 [kvm]\n kvm_arch_irq_bypass_add_producer+0x72/0x90 [kvm]\n irq_bypass_register_consumer+0xcd/0x170 [irqbypa\n---truncated---",
"id": "GHSA-68jg-6v9h-662h",
"modified": "2026-02-14T18:30:16Z",
"published": "2026-02-14T18:30:16Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-23198"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/2284bc168b148a17b5ca3b37b3d95c411f18a08d"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/4385b2f2843549bfb932e0dcf76bf4b065543a3c"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/6d14ba1e144e796b5fc81044f08cfba9024ca195"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/959a063e7f12524bc1871ad1f519787967bbcd45"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/b4d37cdb77a0015f51fee083598fa227cc07aaf1"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/b61f9b2fcf181451d0a319889478cc53c001123e"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/ff48c9312d042bfbe826ca675e98acc6c623211c"
}
],
"schema_version": "1.4.0",
"severity": []
}
Sightings
| Author | Source | Type | Date |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.