6.6-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sean Christopherson seanjc@google.com
[ Upstream commit 7b3d1bbf8d68d76fb21210932a5e8ed8ea80dbcc ]
Eat VMX treemption timer exits in the fastpath regardless of whether L1 or L2 is active. The VM-Exit is 100% KVM-induced, i.e. there is nothing directly related to the exit that KVM needs to do on behalf of the guest, thus there is no reason to wait until the slow path to do nothing.
Opportunistically add comments explaining why preemption timer exits for emulating the guest's APIC timer need to go down the slow path.
Link: https://lore.kernel.org/r/20240110012705.506918-6-seanjc@google.com Signed-off-by: Sean Christopherson seanjc@google.com Signed-off-by: Sasha Levin sashal@kernel.org --- arch/x86/kvm/vmx/vmx.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 4c991d514015..0ecc0e996386 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -6034,13 +6034,26 @@ static fastpath_t handle_fastpath_preemption_timer(struct kvm_vcpu *vcpu) if (vmx->req_immediate_exit) return EXIT_FASTPATH_EXIT_HANDLED;
+ /* + * If L2 is active, go down the slow path as emulating the guest timer + * expiration likely requires synthesizing a nested VM-Exit. + */ + if (is_guest_mode(vcpu)) + return EXIT_FASTPATH_NONE; + kvm_lapic_expired_hv_timer(vcpu); return EXIT_FASTPATH_REENTER_GUEST; }
static int handle_preemption_timer(struct kvm_vcpu *vcpu) { - handle_fastpath_preemption_timer(vcpu); + /* + * This non-fastpath handler is reached if and only if the preemption + * timer was being used to emulate a guest timer while L2 is active. + * All other scenarios are supposed to be handled in the fastpath. + */ + WARN_ON_ONCE(!is_guest_mode(vcpu)); + kvm_lapic_expired_hv_timer(vcpu); return 1; }
@@ -7258,7 +7271,12 @@ void noinstr vmx_spec_ctrl_restore_host(struct vcpu_vmx *vmx,
static fastpath_t vmx_exit_handlers_fastpath(struct kvm_vcpu *vcpu) { - if (is_guest_mode(vcpu)) + /* + * If L2 is active, some VMX preemption timer exits can be handled in + * the fastpath even, all other exits must use the slow path. + */ + if (is_guest_mode(vcpu) && + to_vmx(vcpu)->exit_reason.basic != EXIT_REASON_PREEMPTION_TIMER) return EXIT_FASTPATH_NONE;
switch (to_vmx(vcpu)->exit_reason.basic) {