OSDN Git Service

KVM: x86: Free wbinvd_dirty_mask if vCPU creation fails
authorSean Christopherson <sean.j.christopherson@intel.com>
Wed, 18 Dec 2019 21:54:48 +0000 (13:54 -0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 14 Feb 2020 21:29:58 +0000 (16:29 -0500)
commit 16be9ddea268ad841457a59109963fff8c9de38d upstream.

Free the vCPU's wbinvd_dirty_mask if vCPU creation fails after
kvm_arch_vcpu_init(), e.g. when installing the vCPU's file descriptor.
Do the freeing by calling kvm_arch_vcpu_free() instead of open coding
the freeing.  This adds a likely superfluous, but ultimately harmless,
call to kvmclock_reset(), which only clears vcpu->arch.pv_time_enabled.
Using kvm_arch_vcpu_free() allows for additional cleanup in the future.

Fixes: f5f48ee15c2ee ("KVM: VMX: Execute WBINVD to keep data consistency with assigned devices")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/x86/kvm/x86.c

index 36f8984..29e645c 100644 (file)
@@ -7498,7 +7498,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
        kvm_mmu_unload(vcpu);
        vcpu_put(vcpu);
 
-       kvm_x86_ops->vcpu_free(vcpu);
+       kvm_arch_vcpu_free(vcpu);
 }
 
 void kvm_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event)