Cloudanix VulnDB

CVE-2026-98715

None

CVSS

—

EPSS

0.2%

Published

Oct 18, 2026

CISA due date

—

Cloudanix analysis

Next patch cycle SSVC: Track

A Linux kernel bug could let one virtual machine read or change memory belonging to another virtual machine or the host because a needed memory translation update is sometimes skipped.

Exploitation
Local Privilege
Attacker needs
Low Privilege
Complexity
Moderate
Impact
Privilege Gain
Component
OS kernel
Class
Memory Corruption
Remediation
Upgrade
Effort
Medium
Patch risk
Low

What it actually does

The KVM SVM code incorrectly resets the VMCB TLB control field on a failed VMRUN, dropping a pending TLB flush that the hardware would otherwise have performed only on successful VMRUN; this leaves stale TLB entries that can be used by a subsequent guest execution on the same pCPU.

Why it matters to the business

Virtual machines may execute with stale address translations, increasing the chance of memory corruption, information disclosure between guests, or hypervisor escape in environments that rely on KVM for workload isolation.

How an attack unfolds

An attacker controlling a guest repeatedly triggers VMRUN failures while also causing ASID reuse or TLB flush requests; because the flush is silently dropped, the guest later runs with stale TLB mappings that allow access to host or other-guest memory.

How to fix it

Update to a Linux kernel containing the fix that explicitly tracks per-CPU ASID flush requirements and preserves TLB control across failed VMRUNs; no configuration workaround exists.

How to confirm the fix worked

Confirm the running kernel contains the commit that preserves TLB control on failed VMRUN and tracks flush_all_asids per pCPU; inspect dmesg or kernel logs for the presence of the new per-CPU tracking logic.

Typical attack path

  1. Guest triggers conditions that set TLB_CONTROL_FLUSH_ALL_ASID or FLUSH_ASID in the VMCB
  2. Guest forces a VMRUN failure path (e.g., via crafted state or timing)
  3. KVM clears the TLB control field, discarding the pending flush
  4. On the next successful VMRUN the guest executes with stale TLB entries
  5. Stale translations enable cross-guest or guest-to-host memory access

How to detect it

Examine hypervisor kernel logs and KVM tracepoints around VMRUN failures and TLB flush requests for unexpected reuse of ASIDs or missing flush operations; no specific user-space artifacts are produced.

What your security team should do

  • Prioritize kernel updates on all KVM hosts within the normal patch cycle given low EPSS and absence from KEV
  • Monitor upstream stable kernel releases for the referenced SVM TLB fix
  • Review hypervisor hardening guidance for environments with high guest density

The record

In the Linux kernel, the following vulnerability has been resolved: KVM: SVM: Preserve TLB control (i.e. pending TLB flush) on failed VMRUN Don't reset the VMCB's TLB control back to "do nothing" on a failed VMRUN, as empirical testing shows that the CPU performs the requested TLB flush if and only if VMRUN is successful, i.e. clearing TLB control on a failed VMRUN effectively drops a TLB flush. Explicitly track the need to flush all ASIDs on a per-CPU basis, as the ASID reuse condition is tied to the pCPU, not to the vCPU. As a bonus, this also obviates the need to avoid clobbering FLUSH_ALL_ASID with TLB_CONTROL_FLUSH_ASID, e.g. in svm_flush_tlb_asid(). Deliberately don't bother saving/restoring the "old" tlb_ctl on failure, in quotes because it's not exactly the old tlb_ctl, it's the tlb_ctl from after pre_svm_run(), but before updating tlb_ctl for flush_all_asids. If VMRUN fails and TLB_CONTROL_FLUSH_ALL_ASID is forced, then the next successful run of the VMCB *may* unnecessarily flush all ASIDs, which strictly speaking could result in noisy neighbor issues. However, the fact that new_asid() is already guest-triggerable, because of KVM's flawed behavior of clearing the ASID on emulated INIT, means that a guest can already trigger a flush of all ASIDs at roughly the same rate. And once KVM stops clobbering the ASID on emulated INIT, *or* assigns a static ASID to each vCPU, this flaw goes away.

Affected packages

Ecosystem Package Introduced Fixed