linux-kernel - [PATCH] KVM: VMX: switch to hrtimer for TSC deadline timer when L2 guest is running

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:	Wed,  6 Jul 2016 13:10:51 +0800
From:	Haozhong Zhang <haozhong.zhang@...el.com>
To:	kvm@...r.kernel.org
Cc:	Paolo Bonzini <pbonzini@...hat.com>, rkrcmar@...hat.com,
	Yunhong Jiang <yunhong.jiang@...el.com>,
	Wanpeng Li <wanpeng.li@...mail.com>,
	linux-kernel@...r.kernel.org,
	Haozhong Zhang <haozhong.zhang@...el.com>
Subject: [PATCH] KVM: VMX: switch to hrtimer for TSC deadline timer when L2 guest is running

A different VMCS is loaded when L2 guest is running, so it's incorrect
to use the VMX preemption timer for L1 TSC deadline timer. This patch
switches to hrtimer for L1 TSC deadline timer when entering L2 guest,
and switches back to VMX preemption timer when nested VMEXIT from L2 to
L1.

Signed-off-by: Haozhong Zhang <haozhong.zhang@...el.com>
---
 arch/x86/kvm/vmx.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 85e2f0a..cc29c2a 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -10203,6 +10203,9 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch)
 	if (!vmcs02)
 		return -ENOMEM;
 
+	if (kvm_lapic_hv_timer_in_use(vcpu))
+		kvm_lapic_switch_to_sw_timer(vcpu);
+
 	enter_guest_mode(vcpu);
 
 	vmx->nested.vmcs01_tsc_offset = vmcs_read64(TSC_OFFSET);
@@ -10227,6 +10230,7 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch)
 	if (msr_entry_idx) {
 		leave_guest_mode(vcpu);
 		vmx_load_vmcs01(vcpu);
+		kvm_lapic_switch_to_hv_timer(vcpu);
 		nested_vmx_entry_failure(vcpu, vmcs12,
 				EXIT_REASON_MSR_LOAD_FAIL, msr_entry_idx);
 		return 1;
@@ -10700,6 +10704,7 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason,
 		nested_vmx_abort(vcpu, VMX_ABORT_SAVE_GUEST_MSR_FAIL);
 
 	vmx_load_vmcs01(vcpu);
+	kvm_lapic_switch_to_hv_timer(vcpu);
 
 	if ((exit_reason == EXIT_REASON_EXTERNAL_INTERRUPT)
 	    && nested_exit_intr_ack_set(vcpu)) {
-- 
2.9.0