From: Sean Christopherson Date: Fri, 5 Mar 2021 01:10:58 +0000 (-0800) Subject: KVM: x86: Defer the MMU unload to the normal path on an global INVPCID X-Git-Tag: v5.12.4~642 X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=bb2679a3a1c030d2f43200b575bb825f4ad3da99;p=thirdparty%2Fkernel%2Fstable.git KVM: x86: Defer the MMU unload to the normal path on an global INVPCID commit f66c53b3b94f658590e1012bf6d922f8b7e01bda upstream. Defer unloading the MMU after a INVPCID until the instruction emulation has completed, i.e. until after RIP has been updated. On VMX, this is a benign bug as VMX doesn't touch the MMU when skipping an emulated instruction. However, on SVM, if nrip is disabled, the emulator is used to skip an instruction, which would lead to fireworks if the emulator were invoked without a valid MMU. Fixes: eb4b248e152d ("kvm: vmx: Support INVPCID in shadow paging mode") Cc: stable@vger.kernel.org Signed-off-by: Sean Christopherson Message-Id: <20210305011101.3597423-15-seanjc@google.com> Signed-off-by: Paolo Bonzini Signed-off-by: Greg Kroah-Hartman --- diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index ee0dc58ac3a51..280e5492864c5 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -11539,7 +11539,7 @@ int kvm_handle_invpcid(struct kvm_vcpu *vcpu, unsigned long type, gva_t gva) fallthrough; case INVPCID_TYPE_ALL_INCL_GLOBAL: - kvm_mmu_unload(vcpu); + kvm_make_request(KVM_REQ_MMU_RELOAD, vcpu); return kvm_skip_emulated_instruction(vcpu); default: