[v12,23/46] KVM: x86: hyper-v: L2 TLB flush

Message ID 20221021153521.1216911-24-vkuznets@redhat.com
State New
Headers
Series KVM: x86: hyper-v: Fine-grained TLB flush + L2 TLB flush features |

Commit Message

Vitaly Kuznetsov Oct. 21, 2022, 3:34 p.m. UTC
  Handle L2 TLB flush requests by going through all vCPUs and checking
whether there are vCPUs running the same VM_ID with a VP_ID specified
in the requests. Perform synthetic exit to L2 upon finish.

Note, while checking VM_ID/VP_ID of running vCPUs seem to be a bit
racy, we count on the fact that KVM flushes the whole L2 VPID upon
transition. Also, KVM_REQ_HV_TLB_FLUSH request needs to be done upon
transition between L1 and L2 to make sure all pending requests are
always processed.

For the reference, Hyper-V TLFS refers to the feature as "Direct
Virtual Flush".

Note, nVMX/nSVM code does not handle VMCALL/VMMCALL from L2 yet.

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
---
 arch/x86/kvm/hyperv.c | 80 ++++++++++++++++++++++++++++++++++++-------
 arch/x86/kvm/hyperv.h |  3 --
 arch/x86/kvm/trace.h  | 21 +++++++-----
 3 files changed, 80 insertions(+), 24 deletions(-)
  

Patch

diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
index 68a2200d4a6f..ec3051699d2a 100644
--- a/arch/x86/kvm/hyperv.c
+++ b/arch/x86/kvm/hyperv.c
@@ -34,6 +34,7 @@ 
 #include <linux/eventfd.h>
 
 #include <asm/apicdef.h>
+#include <asm/mshyperv.h>
 #include <trace/events/kvm.h>
 
 #include "trace.h"
@@ -1832,18 +1833,16 @@  static int kvm_hv_get_tlb_flush_entries(struct kvm *kvm, struct kvm_hv_hcall *hc
 				  entries, consumed_xmm_halves, offset);
 }
 
-static void hv_tlb_flush_enqueue(struct kvm_vcpu *vcpu, u64 *entries, int count)
+static void hv_tlb_flush_enqueue(struct kvm_vcpu *vcpu,
+				 struct kvm_vcpu_hv_tlb_flush_fifo *tlb_flush_fifo,
+				 u64 *entries, int count)
 {
-	struct kvm_vcpu_hv_tlb_flush_fifo *tlb_flush_fifo;
 	struct kvm_vcpu_hv *hv_vcpu = to_hv_vcpu(vcpu);
 	u64 flush_all_entry = KVM_HV_TLB_FLUSHALL_ENTRY;
 
 	if (!hv_vcpu)
 		return;
 
-	/* kvm_hv_flush_tlb() is not ready to handle requests for L2s yet */
-	tlb_flush_fifo = &hv_vcpu->tlb_flush_fifo[HV_L1_TLB_FLUSH_FIFO];
-
 	spin_lock(&tlb_flush_fifo->write_lock);
 
 	/*
@@ -1912,6 +1911,7 @@  static u64 kvm_hv_flush_tlb(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
 	struct hv_tlb_flush_ex flush_ex;
 	struct hv_tlb_flush flush;
 	DECLARE_BITMAP(vcpu_mask, KVM_MAX_VCPUS);
+	struct kvm_vcpu_hv_tlb_flush_fifo *tlb_flush_fifo;
 	/*
 	 * Normally, there can be no more than 'KVM_HV_TLB_FLUSH_FIFO_SIZE'
 	 * entries on the TLB flush fifo. The last entry, however, needs to be
@@ -1955,7 +1955,8 @@  static u64 kvm_hv_flush_tlb(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
 		}
 
 		trace_kvm_hv_flush_tlb(flush.processor_mask,
-				       flush.address_space, flush.flags);
+				       flush.address_space, flush.flags,
+				       is_guest_mode(vcpu));
 
 		valid_bank_mask = BIT_ULL(0);
 		sparse_banks[0] = flush.processor_mask;
@@ -1986,7 +1987,7 @@  static u64 kvm_hv_flush_tlb(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
 		trace_kvm_hv_flush_tlb_ex(flush_ex.hv_vp_set.valid_bank_mask,
 					  flush_ex.hv_vp_set.format,
 					  flush_ex.address_space,
-					  flush_ex.flags);
+					  flush_ex.flags, is_guest_mode(vcpu));
 
 		valid_bank_mask = flush_ex.hv_vp_set.valid_bank_mask;
 		all_cpus = flush_ex.hv_vp_set.format !=
@@ -2030,19 +2031,57 @@  static u64 kvm_hv_flush_tlb(struct kvm_vcpu *vcpu, struct kvm_hv_hcall *hc)
 	 * vcpu->arch.cr3 may not be up-to-date for running vCPUs so we can't
 	 * analyze it here, flush TLB regardless of the specified address space.
 	 */
-	if (all_cpus) {
-		kvm_for_each_vcpu(i, v, kvm)
-			hv_tlb_flush_enqueue(v, tlb_flush_entries, hc->rep_cnt);
+	if (all_cpus && !is_guest_mode(vcpu)) {
+		kvm_for_each_vcpu(i, v, kvm) {
+			tlb_flush_fifo = kvm_hv_get_tlb_flush_fifo(v, false);
+			hv_tlb_flush_enqueue(v, tlb_flush_fifo,
+					     tlb_flush_entries, hc->rep_cnt);
+		}
 
 		kvm_make_all_cpus_request(kvm, KVM_REQ_HV_TLB_FLUSH);
-	} else {
+	} else if (!is_guest_mode(vcpu)) {
 		sparse_set_to_vcpu_mask(kvm, sparse_banks, valid_bank_mask, vcpu_mask);
 
 		for_each_set_bit(i, vcpu_mask, KVM_MAX_VCPUS) {
 			v = kvm_get_vcpu(kvm, i);
 			if (!v)
 				continue;
-			hv_tlb_flush_enqueue(v, tlb_flush_entries, hc->rep_cnt);
+			tlb_flush_fifo = kvm_hv_get_tlb_flush_fifo(v, false);
+			hv_tlb_flush_enqueue(v, tlb_flush_fifo,
+					     tlb_flush_entries, hc->rep_cnt);
+		}
+
+		kvm_make_vcpus_request_mask(kvm, KVM_REQ_HV_TLB_FLUSH, vcpu_mask);
+	} else {
+		struct kvm_vcpu_hv *hv_v;
+
+		bitmap_zero(vcpu_mask, KVM_MAX_VCPUS);
+
+		kvm_for_each_vcpu(i, v, kvm) {
+			hv_v = to_hv_vcpu(v);
+
+			/*
+			 * The following check races with nested vCPUs entering/exiting
+			 * and/or migrating between L1's vCPUs, however the only case when
+			 * KVM *must* flush the TLB is when the target L2 vCPU keeps
+			 * running on the same L1 vCPU from the moment of the request until
+			 * kvm_hv_flush_tlb() returns. TLB is fully flushed in all other
+			 * cases, e.g. when the target L2 vCPU migrates to a different L1
+			 * vCPU or when the corresponding L1 vCPU temporary switches to a
+			 * different L2 vCPU while the request is being processed.
+			 */
+			if (!hv_v || hv_v->nested.vm_id != hv_vcpu->nested.vm_id)
+				continue;
+
+			if (!all_cpus &&
+			    !hv_is_vp_in_sparse_set(hv_v->nested.vp_id, valid_bank_mask,
+						    sparse_banks))
+				continue;
+
+			__set_bit(i, vcpu_mask);
+			tlb_flush_fifo = kvm_hv_get_tlb_flush_fifo(v, true);
+			hv_tlb_flush_enqueue(v, tlb_flush_fifo,
+					     tlb_flush_entries, hc->rep_cnt);
 		}
 
 		kvm_make_vcpus_request_mask(kvm, KVM_REQ_HV_TLB_FLUSH, vcpu_mask);
@@ -2232,10 +2271,25 @@  static void kvm_hv_hypercall_set_result(struct kvm_vcpu *vcpu, u64 result)
 
 static int kvm_hv_hypercall_complete(struct kvm_vcpu *vcpu, u64 result)
 {
+	u32 tlb_lock_count = 0;
+	int ret;
+
+	if (hv_result_success(result) && is_guest_mode(vcpu) &&
+	    kvm_hv_is_tlb_flush_hcall(vcpu) &&
+	    kvm_read_guest(vcpu->kvm, to_hv_vcpu(vcpu)->nested.pa_page_gpa,
+			   &tlb_lock_count, sizeof(tlb_lock_count)))
+		result = HV_STATUS_INVALID_HYPERCALL_INPUT;
+
 	trace_kvm_hv_hypercall_done(result);
 	kvm_hv_hypercall_set_result(vcpu, result);
 	++vcpu->stat.hypercalls;
-	return kvm_skip_emulated_instruction(vcpu);
+
+	ret = kvm_skip_emulated_instruction(vcpu);
+
+	if (tlb_lock_count)
+		kvm_x86_ops.nested_ops->hv_inject_synthetic_vmexit_post_tlb_flush(vcpu);
+
+	return ret;
 }
 
 static int kvm_hv_hypercall_complete_userspace(struct kvm_vcpu *vcpu)
diff --git a/arch/x86/kvm/hyperv.h b/arch/x86/kvm/hyperv.h
index 5f9c76b45f46..7706e203ff43 100644
--- a/arch/x86/kvm/hyperv.h
+++ b/arch/x86/kvm/hyperv.h
@@ -159,9 +159,6 @@  static inline struct kvm_vcpu_hv_tlb_flush_fifo *kvm_hv_get_tlb_flush_fifo(struc
 	int i = is_guest_mode ? HV_L2_TLB_FLUSH_FIFO :
 				HV_L1_TLB_FLUSH_FIFO;
 
-	/* KVM does not handle L2 TLB flush requests yet */
-	WARN_ON_ONCE(i != HV_L1_TLB_FLUSH_FIFO);
-
 	return &hv_vcpu->tlb_flush_fifo[i];
 }
 
diff --git a/arch/x86/kvm/trace.h b/arch/x86/kvm/trace.h
index bc25589ad588..09f3392dd830 100644
--- a/arch/x86/kvm/trace.h
+++ b/arch/x86/kvm/trace.h
@@ -1547,38 +1547,41 @@  TRACE_EVENT(kvm_hv_timer_state,
  * Tracepoint for kvm_hv_flush_tlb.
  */
 TRACE_EVENT(kvm_hv_flush_tlb,
-	TP_PROTO(u64 processor_mask, u64 address_space, u64 flags),
-	TP_ARGS(processor_mask, address_space, flags),
+	TP_PROTO(u64 processor_mask, u64 address_space, u64 flags, bool guest_mode),
+	TP_ARGS(processor_mask, address_space, flags, guest_mode),
 
 	TP_STRUCT__entry(
 		__field(u64, processor_mask)
 		__field(u64, address_space)
 		__field(u64, flags)
+		__field(bool, guest_mode)
 	),
 
 	TP_fast_assign(
 		__entry->processor_mask = processor_mask;
 		__entry->address_space = address_space;
 		__entry->flags = flags;
+		__entry->guest_mode = guest_mode;
 	),
 
-	TP_printk("processor_mask 0x%llx address_space 0x%llx flags 0x%llx",
+	TP_printk("processor_mask 0x%llx address_space 0x%llx flags 0x%llx %s",
 		  __entry->processor_mask, __entry->address_space,
-		  __entry->flags)
+		  __entry->flags, __entry->guest_mode ? "(L2)" : "")
 );
 
 /*
  * Tracepoint for kvm_hv_flush_tlb_ex.
  */
 TRACE_EVENT(kvm_hv_flush_tlb_ex,
-	TP_PROTO(u64 valid_bank_mask, u64 format, u64 address_space, u64 flags),
-	TP_ARGS(valid_bank_mask, format, address_space, flags),
+	TP_PROTO(u64 valid_bank_mask, u64 format, u64 address_space, u64 flags, bool guest_mode),
+	TP_ARGS(valid_bank_mask, format, address_space, flags, guest_mode),
 
 	TP_STRUCT__entry(
 		__field(u64, valid_bank_mask)
 		__field(u64, format)
 		__field(u64, address_space)
 		__field(u64, flags)
+		__field(bool, guest_mode)
 	),
 
 	TP_fast_assign(
@@ -1586,12 +1589,14 @@  TRACE_EVENT(kvm_hv_flush_tlb_ex,
 		__entry->format = format;
 		__entry->address_space = address_space;
 		__entry->flags = flags;
+		__entry->guest_mode = guest_mode;
 	),
 
 	TP_printk("valid_bank_mask 0x%llx format 0x%llx "
-		  "address_space 0x%llx flags 0x%llx",
+		  "address_space 0x%llx flags 0x%llx %s",
 		  __entry->valid_bank_mask, __entry->format,
-		  __entry->address_space, __entry->flags)
+		  __entry->address_space, __entry->flags,
+		  __entry->guest_mode ? "(L2)" : "")
 );
 
 /*