Skip to contentSkip to conversation

KVM: x86: Free vCPUs before freeing VM state

PATCH Superseded Kevin Cheng 3 participants 6 messages Jul 25, 2025 – Jul 29, 2025 Stable branch

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 KVM: x86: Free vCPUs before freeing VM state

From: Sean Christopherson <[email protected]>

[ Upstream commit 17bcd714426386fda741a4bccd96a2870179344b ]

Free vCPUs before freeing any VM state, as both SVM and VMX may access VM state when "freeing" a vCPU that is currently "in" L2, i.e. that needs to be kicked out of nested guest mode.

Commit 6fcee03df6a1 ("KVM: x86: avoid loading a vCPU after .vm_destroy was called") partially fixed the issue, but for unknown reasons only moved the MMU unloading before VM destruction. Complete the change, and free all vCPU state prior to destroying VM state, as nVMX accesses even more state than nSVM.

In addition to the AVIC, KVM can hit a use-after-free on MSR filters:

  kvm_msr_allowed+0x4c/0xd0
  __kvm_set_msr+0x12d/0x1e0
  kvm_set_msr+0x19/0x40
  load_vmcs12_host_state+0x2d8/0x6e0 [kvm_intel]
  nested_vmx_vmexit+0x715/0xbd0 [kvm_intel]
  nested_vmx_free_vcpu+0x33/0x50 [kvm_intel]
  vmx_free_vcpu+0x54/0xc0 [kvm_intel]
  kvm_arch_vcpu_destroy+0x28/0xf0
  kvm_vcpu_destroy+0x12/0x50
  kvm_arch_destroy_vm+0x12c/0x1c0
  kvm_put_kvm+0x263/0x3c0
  kvm_vm_release+0x21/0x30

and an upcoming fix to process injectable interrupts on nested VM-Exit will access the PIC:

  BUG: kernel NULL pointer dereference, address: 0000000000000090
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  CPU: 23 UID: 1000 PID: 2658 Comm: kvm-nx-lpage-re
  RIP: 0010:kvm_cpu_has_extint+0x2f/0x60 [kvm]
  Call Trace:
   <TASK>
   kvm_cpu_has_injectable_intr+0xe/0x60 [kvm]
   nested_vmx_vmexit+0x2d7/0xdf0 [kvm_intel]
   nested_vmx_free_vcpu+0x40/0x50 [kvm_intel]
   vmx_vcpu_free+0x2d/0x80 [kvm_intel]
   kvm_arch_vcpu_destroy+0x2d/0x130 [kvm]
   kvm_destroy_vcpus+0x8a/0x100 [kvm]
   kvm_arch_destroy_vm+0xa7/0x1d0 [kvm]
   kvm_destroy_vm+0x172/0x300 [kvm]
   kvm_vcpu_release+0x31/0x50 [kvm]

Inarguably, both nSVM and nVMX need to be fixed, but punt on those cleanups for the moment. Conceptually, vCPUs should be freed before VM state. Assets like the I/O APIC and PIC _must_ be allocated before vCPUs are created, so it stands to reason that they must be freed _after_ vCPUs are destroyed.

---
 arch/x86/kvm/x86.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/x86.c +1 -1
index f378d479fea3f..7f91b11e6f0ec 100644
@@ -12888,11 +12888,11 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
 		mutex_unlock(&kvm->slots_lock);
 	}
 	kvm_unload_vcpu_mmus(kvm);
+	kvm_destroy_vcpus(kvm);
 	kvm_x86_call(vm_destroy)(kvm);
 	kvm_free_msr_filter(srcu_dereference_check(kvm->arch.msr_filter, &kvm->srcu, 1));
 	kvm_pic_destroy(kvm);
 	kvm_ioapic_destroy(kvm);
-	kvm_destroy_vcpus(kvm);
 	kvfree(rcu_dereference_check(kvm->arch.apic_map, 1));
 	kfree(srcu_dereference_check(kvm->arch.pmu_event_filter, &kvm->srcu, 1));
 	kvm_mmu_uninit_vm(kvm);
Signature · 1 line
2.50.1.470.g6ba607880d-goog
maintainer
#1 lore raw

[ Sasha's backport helper bot ]

Hi,

✅ All tests passed successfully. No issues detected. No action required from the submitter.

The upstream commit SHA1 provided is correct: 17bcd714426386fda741a4bccd96a2870179344b

WARNING: Author mismatch between patch and upstream commit:
Backport author: Kevin Cheng <[email protected]> Commit author: Sean Christopherson <[email protected]>

Status in newer kernel trees:
6.15.y | Present (exact SHA1)

Note: The patch differs from the upstream commit:
---
1: 17bcd7144263 ! 1: c2bfc381d88f KVM: x86: Free vCPUs before freeing VM state
@@ Metadata
## Commit message ##
KVM: x86: Free vCPUs before freeing VM state

+ [ Upstream commit 17bcd714426386fda741a4bccd96a2870179344b ]
+
Free vCPUs before freeing any VM state, as both SVM and VMX may access VM state when "freeing" a vCPU that is currently "in" L2, i.e. that needs to be kicked out of nested guest mode.
@@ Commit message

- Message-ID: <[email protected]>
- Signed-off-by: Paolo Bonzini <[email protected]>
+ Signed-off-by: Kevin Cheng <[email protected]>

## arch/x86/kvm/x86.c ##
@@ arch/x86/kvm/x86.c: void kvm_arch_destroy_vm(struct kvm *kvm)

---

Results of testing on various branches:

| Branch                    | Patch Apply | Build Test |
|---------------------------|-------------|------------|
| origin/linux-6.12.y       | Success     | Success    |
#2 lore raw

PATCH v2 1/1 KVM: x86: Free vCPUs before freeing VM state

From: Sean Christopherson <[email protected]>

[ Upstream commit 17bcd714426386fda741a4bccd96a2870179344b ]

Free vCPUs before freeing any VM state, as both SVM and VMX may access VM state when "freeing" a vCPU that is currently "in" L2, i.e. that needs to be kicked out of nested guest mode.

Commit 6fcee03df6a1 ("KVM: x86: avoid loading a vCPU after .vm_destroy was called") partially fixed the issue, but for unknown reasons only moved the MMU unloading before VM destruction. Complete the change, and free all vCPU state prior to destroying VM state, as nVMX accesses even more state than nSVM.

In addition to the AVIC, KVM can hit a use-after-free on MSR filters:

  kvm_msr_allowed+0x4c/0xd0
  __kvm_set_msr+0x12d/0x1e0
  kvm_set_msr+0x19/0x40
  load_vmcs12_host_state+0x2d8/0x6e0 [kvm_intel]
  nested_vmx_vmexit+0x715/0xbd0 [kvm_intel]
  nested_vmx_free_vcpu+0x33/0x50 [kvm_intel]
  vmx_free_vcpu+0x54/0xc0 [kvm_intel]
  kvm_arch_vcpu_destroy+0x28/0xf0
  kvm_vcpu_destroy+0x12/0x50
  kvm_arch_destroy_vm+0x12c/0x1c0
  kvm_put_kvm+0x263/0x3c0
  kvm_vm_release+0x21/0x30

and an upcoming fix to process injectable interrupts on nested VM-Exit will access the PIC:

  BUG: kernel NULL pointer dereference, address: 0000000000000090
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  CPU: 23 UID: 1000 PID: 2658 Comm: kvm-nx-lpage-re
  RIP: 0010:kvm_cpu_has_extint+0x2f/0x60 [kvm]
  Call Trace:
   <TASK>
   kvm_cpu_has_injectable_intr+0xe/0x60 [kvm]
   nested_vmx_vmexit+0x2d7/0xdf0 [kvm_intel]
   nested_vmx_free_vcpu+0x40/0x50 [kvm_intel]
   vmx_vcpu_free+0x2d/0x80 [kvm_intel]
   kvm_arch_vcpu_destroy+0x2d/0x130 [kvm]
   kvm_destroy_vcpus+0x8a/0x100 [kvm]
   kvm_arch_destroy_vm+0xa7/0x1d0 [kvm]
   kvm_destroy_vm+0x172/0x300 [kvm]
   kvm_vcpu_release+0x31/0x50 [kvm]

Inarguably, both nSVM and nVMX need to be fixed, but punt on those cleanups for the moment. Conceptually, vCPUs should be freed before VM state. Assets like the I/O APIC and PIC _must_ be allocated before vCPUs are created, so it stands to reason that they must be freed _after_ vCPUs are destroyed.

Message-ID: <[email protected]>

---
 arch/x86/kvm/x86.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/x86.c +1 -1
index f378d479fea3f..7f91b11e6f0ec 100644
@@ -12888,11 +12888,11 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
 		mutex_unlock(&kvm->slots_lock);
 	}
 	kvm_unload_vcpu_mmus(kvm);
+	kvm_destroy_vcpus(kvm);
 	kvm_x86_call(vm_destroy)(kvm);
 	kvm_free_msr_filter(srcu_dereference_check(kvm->arch.msr_filter, &kvm->srcu, 1));
 	kvm_pic_destroy(kvm);
 	kvm_ioapic_destroy(kvm);
-	kvm_destroy_vcpus(kvm);
 	kvfree(rcu_dereference_check(kvm->arch.apic_map, 1));
 	kfree(srcu_dereference_check(kvm->arch.pmu_event_filter, &kvm->srcu, 1));
 	kvm_mmu_uninit_vm(kvm);
Signature · 1 line
2.50.1.487.gc89ff58d15-goog
maintainer
#3 lore raw
Quoted 91 lines · Kevin Cheng

From: Sean Christopherson <[email protected]>

[ Upstream commit 17bcd714426386fda741a4bccd96a2870179344b ]

Free vCPUs before freeing any VM state, as both SVM and VMX may access VM state when "freeing" a vCPU that is currently "in" L2, i.e. that needs to be kicked out of nested guest mode.

Commit 6fcee03df6a1 ("KVM: x86: avoid loading a vCPU after .vm_destroy was called") partially fixed the issue, but for unknown reasons only moved the MMU unloading before VM destruction. Complete the change, and free all vCPU state prior to destroying VM state, as nVMX accesses even more state than nSVM.

In addition to the AVIC, KVM can hit a use-after-free on MSR filters:

  kvm_msr_allowed+0x4c/0xd0
  __kvm_set_msr+0x12d/0x1e0
  kvm_set_msr+0x19/0x40
  load_vmcs12_host_state+0x2d8/0x6e0 [kvm_intel]
  nested_vmx_vmexit+0x715/0xbd0 [kvm_intel]
  nested_vmx_free_vcpu+0x33/0x50 [kvm_intel]
  vmx_free_vcpu+0x54/0xc0 [kvm_intel]
  kvm_arch_vcpu_destroy+0x28/0xf0
  kvm_vcpu_destroy+0x12/0x50
  kvm_arch_destroy_vm+0x12c/0x1c0
  kvm_put_kvm+0x263/0x3c0
  kvm_vm_release+0x21/0x30

and an upcoming fix to process injectable interrupts on nested VM-Exit will access the PIC:

  BUG: kernel NULL pointer dereference, address: 0000000000000090
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  CPU: 23 UID: 1000 PID: 2658 Comm: kvm-nx-lpage-re
  RIP: 0010:kvm_cpu_has_extint+0x2f/0x60 [kvm]
  Call Trace:
   <TASK>
   kvm_cpu_has_injectable_intr+0xe/0x60 [kvm]
   nested_vmx_vmexit+0x2d7/0xdf0 [kvm_intel]
   nested_vmx_free_vcpu+0x40/0x50 [kvm_intel]
   vmx_vcpu_free+0x2d/0x80 [kvm_intel]
   kvm_arch_vcpu_destroy+0x2d/0x130 [kvm]
   kvm_destroy_vcpus+0x8a/0x100 [kvm]
   kvm_arch_destroy_vm+0xa7/0x1d0 [kvm]
   kvm_destroy_vm+0x172/0x300 [kvm]
   kvm_vcpu_release+0x31/0x50 [kvm]

Inarguably, both nSVM and nVMX need to be fixed, but punt on those cleanups for the moment. Conceptually, vCPUs should be freed before VM state. Assets like the I/O APIC and PIC _must_ be allocated before vCPUs are created, so it stands to reason that they must be freed _after_ vCPUs are destroyed.

Message-ID: <[email protected]>

---
 arch/x86/kvm/x86.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/x86.c +1 -1
index f378d479fea3f..7f91b11e6f0ec 100644
@@ -12888,11 +12888,11 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
 		mutex_unlock(&kvm->slots_lock);
 	}
 	kvm_unload_vcpu_mmus(kvm);
+	kvm_destroy_vcpus(kvm);
 	kvm_x86_call(vm_destroy)(kvm);
 	kvm_free_msr_filter(srcu_dereference_check(kvm->arch.msr_filter, &kvm->srcu, 1));
 	kvm_pic_destroy(kvm);
 	kvm_ioapic_destroy(kvm);
-	kvm_destroy_vcpus(kvm);
 	kvfree(rcu_dereference_check(kvm->arch.apic_map, 1));
 	kfree(srcu_dereference_check(kvm->arch.pmu_event_filter, &kvm->srcu, 1));
 	kvm_mmu_uninit_vm(kvm);
Signature · 1 line
2.50.1.487.gc89ff58d15-goog

What changed in this v2 version?

confused,

greg k-h

#4 lore raw

I left out

Message-ID: <[email protected]>

mistakingly.

Thank you,
Kevin

Quoted 99 lines · Greg KH
Quoted 88 lines · Kevin Cheng

From: Sean Christopherson <[email protected]>

[ Upstream commit 17bcd714426386fda741a4bccd96a2870179344b ]

Free vCPUs before freeing any VM state, as both SVM and VMX may access VM state when "freeing" a vCPU that is currently "in" L2, i.e. that needs to be kicked out of nested guest mode.

Commit 6fcee03df6a1 ("KVM: x86: avoid loading a vCPU after .vm_destroy was called") partially fixed the issue, but for unknown reasons only moved the MMU unloading before VM destruction. Complete the change, and free all vCPU state prior to destroying VM state, as nVMX accesses even more state than nSVM.

In addition to the AVIC, KVM can hit a use-after-free on MSR filters:

  kvm_msr_allowed+0x4c/0xd0
  __kvm_set_msr+0x12d/0x1e0
  kvm_set_msr+0x19/0x40
  load_vmcs12_host_state+0x2d8/0x6e0 [kvm_intel]
  nested_vmx_vmexit+0x715/0xbd0 [kvm_intel]
  nested_vmx_free_vcpu+0x33/0x50 [kvm_intel]
  vmx_free_vcpu+0x54/0xc0 [kvm_intel]
  kvm_arch_vcpu_destroy+0x28/0xf0
  kvm_vcpu_destroy+0x12/0x50
  kvm_arch_destroy_vm+0x12c/0x1c0
  kvm_put_kvm+0x263/0x3c0
  kvm_vm_release+0x21/0x30

and an upcoming fix to process injectable interrupts on nested VM-Exit will access the PIC:

  BUG: kernel NULL pointer dereference, address: 0000000000000090
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  CPU: 23 UID: 1000 PID: 2658 Comm: kvm-nx-lpage-re
  RIP: 0010:kvm_cpu_has_extint+0x2f/0x60 [kvm]
  Call Trace:
   <TASK>
   kvm_cpu_has_injectable_intr+0xe/0x60 [kvm]
   nested_vmx_vmexit+0x2d7/0xdf0 [kvm_intel]
   nested_vmx_free_vcpu+0x40/0x50 [kvm_intel]
   vmx_vcpu_free+0x2d/0x80 [kvm_intel]
   kvm_arch_vcpu_destroy+0x2d/0x130 [kvm]
   kvm_destroy_vcpus+0x8a/0x100 [kvm]
   kvm_arch_destroy_vm+0xa7/0x1d0 [kvm]
   kvm_destroy_vm+0x172/0x300 [kvm]
   kvm_vcpu_release+0x31/0x50 [kvm]

Inarguably, both nSVM and nVMX need to be fixed, but punt on those cleanups for the moment. Conceptually, vCPUs should be freed before VM state. Assets like the I/O APIC and PIC _must_ be allocated before vCPUs are created, so it stands to reason that they must be freed _after_ vCPUs are destroyed.

Message-ID: <[email protected]>

---
 arch/x86/kvm/x86.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/x86.c +1 -2
index f378d479fea3f..7f91b11e6f0ec 100644
@@ -12888,11 +12888,11 @@ void kvm_arch_destroy_vm(struct kvm *kvm)
              mutex_unlock(&kvm->slots_lock);
      }
      kvm_unload_vcpu_mmus(kvm);
+     kvm_destroy_vcpus(kvm);
      kvm_x86_call(vm_destroy)(kvm);
      kvm_free_msr_filter(srcu_dereference_check(kvm->arch.msr_filter, &kvm->srcu, 1));
      kvm_pic_destroy(kvm);
      kvm_ioapic_destroy(kvm);
-     kvm_destroy_vcpus(kvm);
      kvfree(rcu_dereference_check(kvm->arch.apic_map, 1));
      kfree(srcu_dereference_check(kvm->arch.pmu_event_filter, &kvm->srcu, 1));
      kvm_mmu_uninit_vm(kvm);
--
Quoted 3 lines

2.50.1.487.gc89ff58d15-goog

What changed in this v2 version?

confused,

greg k-h

maintainer
#5 lore raw
In reply to Kevin Cheng

[ Sasha's backport helper bot ]

Hi,

✅ All tests passed successfully. No issues detected. No action required from the submitter.

The upstream commit SHA1 provided is correct: 17bcd714426386fda741a4bccd96a2870179344b

WARNING: Author mismatch between patch and upstream commit:
Backport author: Kevin Cheng <[email protected]> Commit author: Sean Christopherson <[email protected]>

Status in newer kernel trees:
6.15.y | Present (exact SHA1)

Note: The patch differs from the upstream commit:
---
1: 17bcd7144263 ! 1: ee33d92e5a01 KVM: x86: Free vCPUs before freeing VM state
@@ Metadata
## Commit message ##
KVM: x86: Free vCPUs before freeing VM state

+ [ Upstream commit 17bcd714426386fda741a4bccd96a2870179344b ]
+
Free vCPUs before freeing any VM state, as both SVM and VMX may access VM state when "freeing" a vCPU that is currently "in" L2, i.e. that needs to be kicked out of nested guest mode.
@@ Commit message

Message-ID: <[email protected]>

+ Signed-off-by: Kevin Cheng <[email protected]>

## arch/x86/kvm/x86.c ##
@@ arch/x86/kvm/x86.c: void kvm_arch_destroy_vm(struct kvm *kvm)

---

Results of testing on various branches:

| Branch                    | Patch Apply | Build Test |
|---------------------------|-------------|------------|
| origin/linux-6.12.y       | Success     | Success    |