PatchSiren cyber security CVE debrief
CVE-2026-46113 Linux CVE debrief
A use-after-free vulnerability in the Linux kernel's KVM x86 shadow MMU implementation allows privilege escalation and denial of service. The flaw occurs when guest page tables are modified between VM entries, causing KVM to create stale reverse map (rmap) entries that reference freed kvm_mmu_page structures. When a memslot is deleted, the rmap walk fails to locate entries outside the expected GFN range, leaving dangling pointers. Subsequent operations like dirty logging or MMU notifier invalidations dereference these freed structures. The vulnerability exists because KVM assumed GFNs would always match when walking shadow PTEs, an assumption that held until memory optimization changes in 2032a93d66fa removed separate GFN page allocation for direct MMU pages. The fix adds validation to detect GFN mismatches and zap existing SPTEs before installing new mappings.
- Vendor
- Linux
- Product
- Unknown
- CVSS
- HIGH 8.8
- CISA KEV
- Not listed in stored evidence
- Original CVE published
- 2026-05-28
- Original CVE updated
- 2026-05-30
- Advisory published
- 2026-05-28
- Advisory updated
- 2026-05-30
Who should care
Cloud providers running KVM-based virtualization, organizations hosting untrusted workloads on Linux hypervisors, security teams responsible for kernel patch management
Technical summary
The vulnerability stems from incorrect GFN range assumptions in KVM's shadow page table management. When a guest's 2MB PDE mapping is modified externally and the guest accesses another page in that region, KVM installs a leaf SPTE with a GFN outside the original kvm_mmu_page's [sp->gfn, sp->gfn + 511] range. The rmap entry is recorded at this incorrect offset. When the original memslot is deleted, rmap_remove() only searches the original range and misses the stale entry. Later rmap walks for dirty logging or MMU notifiers traverse this stale entry and dereference the freed kvm_mmu_page. The fix validates target GFNs against existing SPTEs and zaps mismatches before installing new mappings.
Defensive priority
high
Recommended defensive actions
- Apply kernel updates containing commits 06c19c967b845b63172601fe459667d973b7e6b7, 0cb2af2ea66ad8ff195c156ea690f11216285bdf, 14d1e55dfd2cf4711bff164a6aaaddb783552134, 488e386484ec8c0e558be6e156edf34ed9f4d5c8, or 738ec97b5
- Restart KVM virtual machines after kernel update to ensure shadow MMU state is reinitialized
- Audit systems for unauthorized VM memory access attempts in hypervisor logs
- Prioritize patching on multi-tenant virtualization hosts where untrusted guests execute
Evidence notes
Vulnerability description sourced from official CVE record published 2026-05-28. Fix commits identified in kernel.org stable tree. No CVSS score or severity assigned by NVD at time of disclosure.
Official resources
-
CVE-2026-46113 CVE record
CVE.org
-
CVE-2026-46113 NVD detail
NVD
-
Source item URL
nvd_modified
-
Source reference
416baaa9-dc9f-4396-8d5f-8c081fb06d67
-
Source reference
416baaa9-dc9f-4396-8d5f-8c081fb06d67
-
Source reference
416baaa9-dc9f-4396-8d5f-8c081fb06d67
-
Source reference
416baaa9-dc9f-4396-8d5f-8c081fb06d67
-
Source reference
416baaa9-dc9f-4396-8d5f-8c081fb06d67
2026-05-28