KVM: Halt vcpu if page it tries to access is swapped out
If a guest accesses swapped out memory do not swap it in from vcpu thread
context. Schedule work to do swapping and put vcpu into halted state
instead.
Interrupts will still be delivered to the guest and if interrupt will
cause reschedule guest will continue to run another task.
[avi: remove call to get_user_pages_noio(), nacked by Linus; this
makes everything synchrnous again]
Acked-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig
index ddc131f..50f6364 100644
--- a/arch/x86/kvm/Kconfig
+++ b/arch/x86/kvm/Kconfig
@@ -28,6 +28,7 @@
select HAVE_KVM_IRQCHIP
select HAVE_KVM_EVENTFD
select KVM_APIC_ARCHITECTURE
+ select KVM_ASYNC_PF
select USER_RETURN_NOTIFIER
select KVM_MMIO
---help---
diff --git a/arch/x86/kvm/Makefile b/arch/x86/kvm/Makefile
index 31a7035..c53bf19 100644
--- a/arch/x86/kvm/Makefile
+++ b/arch/x86/kvm/Makefile
@@ -9,6 +9,7 @@
coalesced_mmio.o irq_comm.o eventfd.o \
assigned-dev.o)
kvm-$(CONFIG_IOMMU_API) += $(addprefix ../../../virt/kvm/, iommu.o)
+kvm-$(CONFIG_KVM_ASYNC_PF) += $(addprefix ../../../virt/kvm/, async_pf.o)
kvm-y += x86.o mmu.o emulate.o i8259.o irq.o lapic.o \
i8254.o timer.o
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index fbb04ae..4ab04de 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -18,9 +18,11 @@
*
*/
+#include "irq.h"
#include "mmu.h"
#include "x86.h"
#include "kvm_cache_regs.h"
+#include "x86.h"
#include <linux/kvm_host.h>
#include <linux/types.h>
@@ -2587,6 +2589,50 @@
error_code & PFERR_WRITE_MASK, gfn);
}
+int kvm_arch_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn)
+{
+ struct kvm_arch_async_pf arch;
+ arch.gfn = gfn;
+
+ return kvm_setup_async_pf(vcpu, gva, gfn, &arch);
+}
+
+static bool can_do_async_pf(struct kvm_vcpu *vcpu)
+{
+ if (unlikely(!irqchip_in_kernel(vcpu->kvm) ||
+ kvm_event_needs_reinjection(vcpu)))
+ return false;
+
+ return kvm_x86_ops->interrupt_allowed(vcpu);
+}
+
+static bool try_async_pf(struct kvm_vcpu *vcpu, gfn_t gfn, gva_t gva,
+ pfn_t *pfn)
+{
+ bool async;
+
+ *pfn = gfn_to_pfn_async(vcpu->kvm, gfn, &async);
+
+ if (!async)
+ return false; /* *pfn has correct page already */
+
+ put_page(pfn_to_page(*pfn));
+
+ if (can_do_async_pf(vcpu)) {
+ trace_kvm_try_async_get_page(async, *pfn);
+ if (kvm_find_async_pf_gfn(vcpu, gfn)) {
+ trace_kvm_async_pf_doublefault(gva, gfn);
+ kvm_make_request(KVM_REQ_APF_HALT, vcpu);
+ return true;
+ } else if (kvm_arch_setup_async_pf(vcpu, gva, gfn))
+ return true;
+ }
+
+ *pfn = gfn_to_pfn(vcpu->kvm, gfn);
+
+ return false;
+}
+
static int tdp_page_fault(struct kvm_vcpu *vcpu, gva_t gpa,
u32 error_code)
{
@@ -2609,7 +2655,11 @@
mmu_seq = vcpu->kvm->mmu_notifier_seq;
smp_rmb();
- pfn = gfn_to_pfn(vcpu->kvm, gfn);
+
+ if (try_async_pf(vcpu, gfn, gpa, &pfn))
+ return 0;
+
+ /* mmio */
if (is_error_pfn(pfn))
return kvm_handle_bad_page(vcpu->kvm, gfn, pfn);
spin_lock(&vcpu->kvm->mmu_lock);
diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index cd7a833..c45376d 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -568,7 +568,9 @@
mmu_seq = vcpu->kvm->mmu_notifier_seq;
smp_rmb();
- pfn = gfn_to_pfn(vcpu->kvm, walker.gfn);
+
+ if (try_async_pf(vcpu, walker.gfn, addr, &pfn))
+ return 0;
/* mmio */
if (is_error_pfn(pfn))
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c05d477..3cd4d09 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -43,6 +43,7 @@
#include <linux/slab.h>
#include <linux/perf_event.h>
#include <linux/uaccess.h>
+#include <linux/hash.h>
#include <trace/events/kvm.h>
#define CREATE_TRACE_POINTS
@@ -155,6 +156,13 @@
u64 __read_mostly host_xcr0;
+static inline void kvm_async_pf_hash_reset(struct kvm_vcpu *vcpu)
+{
+ int i;
+ for (i = 0; i < roundup_pow_of_two(ASYNC_PF_PER_VCPU); i++)
+ vcpu->arch.apf.gfns[i] = ~0;
+}
+
static void kvm_on_user_return(struct user_return_notifier *urn)
{
unsigned slot;
@@ -5115,6 +5123,12 @@
vcpu->fpu_active = 0;
kvm_x86_ops->fpu_deactivate(vcpu);
}
+ if (kvm_check_request(KVM_REQ_APF_HALT, vcpu)) {
+ /* Page is swapped out. Do synthetic halt */
+ vcpu->arch.apf.halted = true;
+ r = 1;
+ goto out;
+ }
}
r = kvm_mmu_reload(vcpu);
@@ -5243,7 +5257,8 @@
r = 1;
while (r > 0) {
- if (vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE)
+ if (vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE &&
+ !vcpu->arch.apf.halted)
r = vcpu_enter_guest(vcpu);
else {
srcu_read_unlock(&kvm->srcu, vcpu->srcu_idx);
@@ -5256,6 +5271,7 @@
vcpu->arch.mp_state =
KVM_MP_STATE_RUNNABLE;
case KVM_MP_STATE_RUNNABLE:
+ vcpu->arch.apf.halted = false;
break;
case KVM_MP_STATE_SIPI_RECEIVED:
default:
@@ -5277,6 +5293,9 @@
vcpu->run->exit_reason = KVM_EXIT_INTR;
++vcpu->stat.request_irq_exits;
}
+
+ kvm_check_async_pf_completion(vcpu);
+
if (signal_pending(current)) {
r = -EINTR;
vcpu->run->exit_reason = KVM_EXIT_INTR;
@@ -5792,6 +5811,10 @@
kvm_make_request(KVM_REQ_EVENT, vcpu);
+ kvm_clear_async_pf_completion_queue(vcpu);
+ kvm_async_pf_hash_reset(vcpu);
+ vcpu->arch.apf.halted = false;
+
return kvm_x86_ops->vcpu_reset(vcpu);
}
@@ -5880,6 +5903,8 @@
if (!zalloc_cpumask_var(&vcpu->arch.wbinvd_dirty_mask, GFP_KERNEL))
goto fail_free_mce_banks;
+ kvm_async_pf_hash_reset(vcpu);
+
return 0;
fail_free_mce_banks:
kfree(vcpu->arch.mce_banks);
@@ -5938,8 +5963,10 @@
/*
* Unpin any mmu pages first.
*/
- kvm_for_each_vcpu(i, vcpu, kvm)
+ kvm_for_each_vcpu(i, vcpu, kvm) {
+ kvm_clear_async_pf_completion_queue(vcpu);
kvm_unload_vcpu_mmu(vcpu);
+ }
kvm_for_each_vcpu(i, vcpu, kvm)
kvm_arch_vcpu_free(vcpu);
@@ -6050,7 +6077,9 @@
int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu)
{
- return vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE
+ return (vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE &&
+ !vcpu->arch.apf.halted)
+ || !list_empty_careful(&vcpu->async_pf.done)
|| vcpu->arch.mp_state == KVM_MP_STATE_SIPI_RECEIVED
|| vcpu->arch.nmi_pending ||
(kvm_arch_interrupt_allowed(vcpu) &&
@@ -6109,6 +6138,83 @@
}
EXPORT_SYMBOL_GPL(kvm_set_rflags);
+static inline u32 kvm_async_pf_hash_fn(gfn_t gfn)
+{
+ return hash_32(gfn & 0xffffffff, order_base_2(ASYNC_PF_PER_VCPU));
+}
+
+static inline u32 kvm_async_pf_next_probe(u32 key)
+{
+ return (key + 1) & (roundup_pow_of_two(ASYNC_PF_PER_VCPU) - 1);
+}
+
+static void kvm_add_async_pf_gfn(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+ u32 key = kvm_async_pf_hash_fn(gfn);
+
+ while (vcpu->arch.apf.gfns[key] != ~0)
+ key = kvm_async_pf_next_probe(key);
+
+ vcpu->arch.apf.gfns[key] = gfn;
+}
+
+static u32 kvm_async_pf_gfn_slot(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+ int i;
+ u32 key = kvm_async_pf_hash_fn(gfn);
+
+ for (i = 0; i < roundup_pow_of_two(ASYNC_PF_PER_VCPU) &&
+ (vcpu->arch.apf.gfns[key] != gfn ||
+ vcpu->arch.apf.gfns[key] == ~0); i++)
+ key = kvm_async_pf_next_probe(key);
+
+ return key;
+}
+
+bool kvm_find_async_pf_gfn(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+ return vcpu->arch.apf.gfns[kvm_async_pf_gfn_slot(vcpu, gfn)] == gfn;
+}
+
+static void kvm_del_async_pf_gfn(struct kvm_vcpu *vcpu, gfn_t gfn)
+{
+ u32 i, j, k;
+
+ i = j = kvm_async_pf_gfn_slot(vcpu, gfn);
+ while (true) {
+ vcpu->arch.apf.gfns[i] = ~0;
+ do {
+ j = kvm_async_pf_next_probe(j);
+ if (vcpu->arch.apf.gfns[j] == ~0)
+ return;
+ k = kvm_async_pf_hash_fn(vcpu->arch.apf.gfns[j]);
+ /*
+ * k lies cyclically in ]i,j]
+ * | i.k.j |
+ * |....j i.k.| or |.k..j i...|
+ */
+ } while ((i <= j) ? (i < k && k <= j) : (i < k || k <= j));
+ vcpu->arch.apf.gfns[i] = vcpu->arch.apf.gfns[j];
+ i = j;
+ }
+}
+
+void kvm_arch_async_page_not_present(struct kvm_vcpu *vcpu,
+ struct kvm_async_pf *work)
+{
+ trace_kvm_async_pf_not_present(work->gva);
+
+ kvm_make_request(KVM_REQ_APF_HALT, vcpu);
+ kvm_add_async_pf_gfn(vcpu, work->arch.gfn);
+}
+
+void kvm_arch_async_page_present(struct kvm_vcpu *vcpu,
+ struct kvm_async_pf *work)
+{
+ trace_kvm_async_pf_ready(work->gva);
+ kvm_del_async_pf_gfn(vcpu, work->arch.gfn);
+}
+
EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_exit);
EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_inj_virq);
EXPORT_TRACEPOINT_SYMBOL_GPL(kvm_page_fault);