CVE-2026-64284

In the Linux kernel, the following vulnerability has been resolved: KVM: x86: Ensure vendor's exit handler runs before fastpath userspace exits Move the handling of fastpath userspace exits into vendor code to ensure KVM runs vendor specific operations that need to run before userspace gains control of the vCPU. E.g. for VMX (and soon to be for SVM as well), KVM needs to flush the PML buffer prior to exiting to userspace, otherwise any memory written by the final KVM_RUN might never be flagged as dirty. Note, waiting to snapshot CR0 and CR3 until svm_handle_exit() is flawed in general, as that risks consuming stale state in a fastpath handler. That will be addressed in a future change.

• git.kernel.orghttps://git.kernel.org/stable/c/0ffedf43910e44b76c2c1db4e9fbf12b268190c1
• git.kernel.orghttps://git.kernel.org/stable/c/4ad73ef0e7966ecfe67de0060537b4cb14d9acd4
• git.kernel.orghttps://git.kernel.org/stable/c/b3436d9b9b1affe1c3191ac9831308923f5f03c3
• git.kernel.orghttps://git.kernel.org/stable/c/f2ca2b5326211bd38490f0497eb583721ce0bbc0

No related CVEs by CWE or product.