From: Lai Jiangshan <laijs@xxxxxxxxxxxxxxxxx> [ Upstream commit 552617382c197949ff965a3559da8952bf3c1fa5 ] X86_CR4_PCIDE doesn't participate in kvm_mmu_role, so the mmu context doesn't need to be reset. It is only required to flush all the guest tlb. Signed-off-by: Lai Jiangshan <laijs@xxxxxxxxxxxxxxxxx> Reviewed-by: Sean Christopherson <seanjc@xxxxxxxxxx> Message-Id: <20210919024246.89230-2-jiangshanlai@xxxxxxxxx> Signed-off-by: Paolo Bonzini <pbonzini@xxxxxxxxxx> Signed-off-by: Mathias Krause <minipli@xxxxxxxxxxxxxx> # backport to v5.10.x --- - no kvm_post_set_cr4() in this kernel yet, it's part of kvm_set_cr4() arch/x86/kvm/x86.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index d6bb2c300e16..952281f18987 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1042,9 +1042,10 @@ int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4) kvm_x86_ops.set_cr4(vcpu, cr4); - if (((cr4 ^ old_cr4) & KVM_MMU_CR4_ROLE_BITS) || - (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE))) + if ((cr4 ^ old_cr4) & KVM_MMU_CR4_ROLE_BITS) kvm_mmu_reset_context(vcpu); + else if (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE)) + kvm_make_request(KVM_REQ_TLB_FLUSH_GUEST, vcpu); if ((cr4 ^ old_cr4) & (X86_CR4_OSXSAVE | X86_CR4_PKE)) kvm_update_cpuid_runtime(vcpu); -- 2.39.2
