From: Like Xu <likexu@xxxxxxxxxxx> commit c6c937d673aaa1d603f62f134e1ca9c173eeeed3 upstream. Just like on the optional mmu_alloc_direct_roots() path, once shadow path reaches "r = -EIO" somewhere, the caller needs to know the actual state in order to enter error handling and avoid something worse. Fixes: 4a38162ee9f1 ("KVM: MMU: load PDPTRs outside mmu_lock") Signed-off-by: Like Xu <likexu@xxxxxxxxxxx> Reviewed-by: Sean Christopherson <seanjc@xxxxxxxxxx> Message-Id: <20220301124941.48412-1-likexu@xxxxxxxxxxx> Signed-off-by: Paolo Bonzini <pbonzini@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- arch/x86/kvm/mmu/mmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/arch/x86/kvm/mmu/mmu.c +++ b/arch/x86/kvm/mmu/mmu.c @@ -3579,7 +3579,7 @@ set_root_pgd: out_unlock: write_unlock(&vcpu->kvm->mmu_lock); - return 0; + return r; } static int mmu_alloc_special_roots(struct kvm_vcpu *vcpu) Patches currently in stable-queue which might be from likexu@xxxxxxxxxxx are queue-5.15/kvm-x86-mmu-passing-up-the-error-state-of-mmu_alloc_shadow_roots.patch