Re: [PATCH] i386/kvm: fix a use-after-free when vcpu plug/unplug

Philippe Mathieu-Daudé <philmd@xxxxxxxxxx> · Tue, 12 May 2020 09:54:27 +0200

On 5/12/20 3:39 PM, Pan Nengyuan wrote:
When we hotplug vcpus, cpu_update_state is added to vm_change_state_head
in kvm_arch_init_vcpu(). But it forgot to delete in kvm_arch_destroy_vcpu() after
unplug. Then it will cause a use-after-free access. This patch delete it in
kvm_arch_destroy_vcpu() to fix that.

Reproducer:
     virsh setvcpus vm1 4 --live
     virsh setvcpus vm1 2 --live
     virsh suspend vm1
     virsh resume vm1

The UAF stack:
==qemu-system-x86_64==28233==ERROR: AddressSanitizer: heap-use-after-free on address 0x62e00002e798 at pc 0x5573c6917d9e bp 0x7fff07139e50 sp 0x7fff07139e40
WRITE of size 1 at 0x62e00002e798 thread T0
     #0 0x5573c6917d9d in cpu_update_state /mnt/sdb/qemu/target/i386/kvm.c:742
     #1 0x5573c699121a in vm_state_notify /mnt/sdb/qemu/vl.c:1290
     #2 0x5573c636287e in vm_prepare_start /mnt/sdb/qemu/cpus.c:2144
     #3 0x5573c6362927 in vm_start /mnt/sdb/qemu/cpus.c:2150
     #4 0x5573c71e8304 in qmp_cont /mnt/sdb/qemu/monitor/qmp-cmds.c:173
     #5 0x5573c727cb1e in qmp_marshal_cont qapi/qapi-commands-misc.c:835
     #6 0x5573c7694c7a in do_qmp_dispatch /mnt/sdb/qemu/qapi/qmp-dispatch.c:132
     #7 0x5573c7694c7a in qmp_dispatch /mnt/sdb/qemu/qapi/qmp-dispatch.c:175
     #8 0x5573c71d9110 in monitor_qmp_dispatch /mnt/sdb/qemu/monitor/qmp.c:145
     #9 0x5573c71dad4f in monitor_qmp_bh_dispatcher /mnt/sdb/qemu/monitor/qmp.c:234

Reported-by: Euler Robot <euler.robot@xxxxxxxxxx>
Signed-off-by: Pan Nengyuan <pannengyuan@xxxxxxxxxx>
---
  target/i386/cpu.h | 1 +
  target/i386/kvm.c | 5 ++++-
  2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index e818fc712a..afbd11b7a3 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1631,6 +1631,7 @@ struct X86CPU {
  
      CPUNegativeOffsetState neg;
      CPUX86State env;
+    VMChangeStateEntry *vmsentry;
  
      uint64_t ucode_rev;
  
diff --git a/target/i386/kvm.c b/target/i386/kvm.c
index 4901c6dd74..ff2848357e 100644
--- a/target/i386/kvm.c
+++ b/target/i386/kvm.c
@@ -1770,7 +1770,7 @@ int kvm_arch_init_vcpu(CPUState *cs)
          }
      }
  
-    qemu_add_vm_change_state_handler(cpu_update_state, env);
+    cpu->vmsentry = qemu_add_vm_change_state_handler(cpu_update_state, env);
  
      c = cpuid_find_entry(&cpuid_data.cpuid, 1, 0);
      if (c) {
@@ -1883,6 +1883,9 @@ int kvm_arch_destroy_vcpu(CPUState *cs)
          env->nested_state = NULL;
      }
  
+    qemu_del_vm_change_state_handler(cpu->vmsentry);
+    cpu->vmsentry = NULL;

Why set it to NULL? there is no non-NULL check.

Anyway if it matters to you, why not do it in 
qemu_del_vm_change_state_handler()?

Otherwise:
Reviewed-by: Philippe Mathieu-Daudé <philmd@xxxxxxxxxx>

+
      return 0;
  }
  








