KVM: x86: Switch KVM guest to using interrupts for page ready APF delivery

KVM now supports using interrupt for 'page ready' APF event delivery and
legacy mechanism was deprecated. Switch KVM guests to the new one.

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-Id: <20200525144125.143875-9-vkuznets@redhat.com>
[Use HYPERVISOR_CALLBACK_VECTOR instead of a separate vector. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

arch/x86/Kconfig

@@ -801,6 +801,7 @@ config KVM_GUEST
 	depends on PARAVIRT
 	select PARAVIRT_CLOCK
 	select ARCH_CPUIDLE_HALTPOLL
+	select X86_HV_CALLBACK_VECTOR
 	default y
 	---help---
 	  This option enables various optimizations for running under the KVM

arch/x86/entry/entry_32.S

@@ -1475,6 +1475,11 @@ BUILD_INTERRUPT3(hv_stimer0_callback_vector, HYPERV_STIMER0_VECTOR,
 
 #endif /* CONFIG_HYPERV */
 
+#ifdef CONFIG_KVM_GUEST
+BUILD_INTERRUPT3(kvm_async_pf_vector, HYPERVISOR_CALLBACK_VECTOR,
+		 kvm_async_pf_intr)
+#endif
+
 SYM_CODE_START(page_fault)
 	ASM_CLAC
 	pushl	$do_page_fault

arch/x86/entry/entry_64.S

@@ -1190,6 +1190,11 @@ apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \
 	acrn_hv_callback_vector acrn_hv_vector_handler
 #endif
 
+#ifdef CONFIG_KVM_GUEST
+apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \
+	kvm_async_pf_vector kvm_async_pf_intr
+#endif
+
 idtentry debug			do_debug		has_error_code=0	paranoid=1 shift_ist=IST_INDEX_DB ist_offset=DB_STACK_OFFSET
 idtentry int3			do_int3			has_error_code=0	create_gap=1
 idtentry stack_segment		do_stack_segment	has_error_code=1

arch/x86/include/asm/kvm_para.h

@@ -4,6 +4,7 @@
 
 #include <asm/processor.h>
 #include <asm/alternative.h>
+#include <linux/interrupt.h>
 #include <uapi/asm/kvm_para.h>
 
 extern void kvmclock_init(void);
@@ -104,6 +105,12 @@ static __always_inline bool kvm_handle_async_pf(struct pt_regs *regs, u32 token)
 		return false;
 }
 
+extern __visible void kvm_async_pf_vector(void);
+#ifdef CONFIG_TRACING
+#define trace_kvm_async_pf_vector kvm_async_pf_vector
+#endif
+__visible void __irq_entry kvm_async_pf_intr(struct pt_regs *regs);
+
 #ifdef CONFIG_PARAVIRT_SPINLOCKS
 void __init kvm_spinlock_init(void);
 #else /* !CONFIG_PARAVIRT_SPINLOCKS */

arch/x86/kernel/kvm.c

@@ -233,15 +233,10 @@ NOKPROBE_SYMBOL(kvm_read_and_reset_apf_flags);
 
 bool __kvm_handle_async_pf(struct pt_regs *regs, u32 token)
 {
-	u32 reason = kvm_read_and_reset_apf_flags();
+	u32 flags = kvm_read_and_reset_apf_flags();
 
-	switch (reason) {
+	if (!flags)
-	case KVM_PV_REASON_PAGE_NOT_PRESENT:
-	case KVM_PV_REASON_PAGE_READY:
-		break;
-	default:
 		return false;
-	}
 
 	/*
 	 * If the host managed to inject an async #PF into an interrupt
@@ -251,20 +246,39 @@ bool __kvm_handle_async_pf(struct pt_regs *regs, u32 token)
 	if (unlikely(!(regs->flags & X86_EFLAGS_IF)))
 		panic("Host injected async #PF in interrupt disabled region\n");
 
-	if (reason == KVM_PV_REASON_PAGE_NOT_PRESENT) {
+	if (flags & KVM_PV_REASON_PAGE_NOT_PRESENT) {
 		if (unlikely(!(user_mode(regs))))
 			panic("Host injected async #PF in kernel mode\n");
 		/* Page is swapped out by the host. */
 		kvm_async_pf_task_wait_schedule(token);
-	} else {
+		return true;
-		rcu_irq_enter();
-		kvm_async_pf_task_wake(token);
-		rcu_irq_exit();
 	}
+
+	WARN_ONCE(1, "Unexpected async PF flags: %x\n", flags);
 	return true;
 }
 NOKPROBE_SYMBOL(__kvm_handle_async_pf);
 
+__visible void __irq_entry kvm_async_pf_intr(struct pt_regs *regs)
+{
+	u32 token;
+
+	entering_ack_irq();
+
+	inc_irq_stat(irq_hv_callback_count);
+
+	if (__this_cpu_read(apf_reason.enabled)) {
+		token = __this_cpu_read(apf_reason.token);
+		rcu_irq_enter();
+		kvm_async_pf_task_wake(token);
+		rcu_irq_exit();
+		__this_cpu_write(apf_reason.token, 0);
+		wrmsrl(MSR_KVM_ASYNC_PF_ACK, 1);
+	}
+
+	exiting_irq();
+}
+
 static void __init paravirt_ops_setup(void)
 {
 	pv_info.name = "KVM";
@@ -308,17 +322,19 @@ static notrace void kvm_guest_apic_eoi_write(u32 reg, u32 val)
 
 static void kvm_guest_cpu_init(void)
 {
-	if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF) && kvmapf) {
+	if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF_INT) && kvmapf) {
-		u64 pa;
+		u64 pa = slow_virt_to_phys(this_cpu_ptr(&apf_reason));
 
 		WARN_ON_ONCE(!static_branch_likely(&kvm_async_pf_enabled));
 
 		pa = slow_virt_to_phys(this_cpu_ptr(&apf_reason));
-		pa |= KVM_ASYNC_PF_ENABLED;
+		pa |= KVM_ASYNC_PF_ENABLED | KVM_ASYNC_PF_DELIVERY_AS_INT;
 
 		if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF_VMEXIT))
 			pa |= KVM_ASYNC_PF_DELIVERY_AS_PF_VMEXIT;
 
+		wrmsrl(MSR_KVM_ASYNC_PF_INT, HYPERVISOR_CALLBACK_VECTOR);
+
 		wrmsrl(MSR_KVM_ASYNC_PF_EN, pa);
 		__this_cpu_write(apf_reason.enabled, 1);
 		pr_info("KVM setup async PF for cpu %d\n", smp_processor_id());
@@ -643,8 +659,10 @@ static void __init kvm_guest_init(void)
 	if (kvm_para_has_feature(KVM_FEATURE_PV_EOI))
 		apic_set_eoi_write(kvm_guest_apic_eoi_write);
 
-	if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF) && kvmapf)
+	if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF_INT) && kvmapf) {
 		static_branch_enable(&kvm_async_pf_enabled);
+		alloc_intr_gate(HYPERVISOR_CALLBACK_VECTOR, kvm_async_pf_vector);
+	}
 
 #ifdef CONFIG_SMP
 	smp_ops.smp_prepare_cpus = kvm_smp_prepare_cpus;