Merge branch 'kvm-updates/2.6.39' of git://git.kernel.org/pub/scm/virt/kvm/kvm

* 'kvm-updates/2.6.39' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (55 commits)
  KVM: unbreak userspace that does not sets tss address
  KVM: MMU: cleanup pte write path
  KVM: MMU: introduce a common function to get no-dirty-logged slot
  KVM: fix rcu usage in init_rmode_* functions
  KVM: fix kvmclock regression due to missing clock update
  KVM: emulator: Fix permission checking in io permission bitmap
  KVM: emulator: Fix io permission checking for 64bit guest
  KVM: SVM: Load %gs earlier if CONFIG_X86_32_LAZY_GS=n
  KVM: x86: Remove useless regs_page pointer from kvm_lapic
  KVM: improve comment on rcu use in irqfd_deassign
  KVM: MMU: remove unused macros
  KVM: MMU: cleanup page alloc and free
  KVM: MMU: do not record gfn in kvm_mmu_pte_write
  KVM: MMU: move mmu pages calculated out of mmu lock
  KVM: MMU: set spte accessed bit properly
  KVM: MMU: fix kvm_mmu_slot_remove_write_access dropping intermediate W bits
  KVM: Start lock documentation
  KVM: better readability of efer_reserved_bits
  KVM: Clear async page fault hash after switching to real mode
  KVM: VMX: Initialize vm86 TSS only once.
  ...

Documentation/kvm/locking.txt

+KVM Lock Overview
+=================
+
+1. Acquisition Orders
+---------------------
+
+(to be written)
+
+2. Reference
+------------
+
+Name:		kvm_lock
+Type:		raw_spinlock
+Arch:		any
+Protects:	- vm_list
+		- hardware virtualization enable/disable
+Comment:	'raw' because hardware enabling/disabling must be atomic /wrt
+		migration.
+
+Name:		kvm_arch::tsc_write_lock
+Type:		raw_spinlock
+Arch:		x86
+Protects:	- kvm_arch::{last_tsc_write,last_tsc_nsec,last_tsc_offset}
+		- tsc offset in vmcb
+Comment:	'raw' because updating the tsc offsets must not be preempted.

arch/alpha/include/asm/errno.h

@@ -122,4 +122,6 @@
 
 #define	ERFKILL		138	/* Operation not possible due to RF-kill */
 
+#define EHWPOISON	139	/* Memory page has hardware error */
+
 #endif

arch/ia64/kvm/kvm-ia64.c

@@ -662,6 +662,7 @@ static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		goto vcpu_run_fail;
 
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
+	vcpu->mode = IN_GUEST_MODE;
 	kvm_guest_enter();
 
 	/*
@@ -683,6 +684,7 @@ static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	 */
 	barrier();
 	kvm_guest_exit();
+	vcpu->mode = OUTSIDE_GUEST_MODE;
 	preempt_enable();
 
 	idx = srcu_read_lock(&vcpu->kvm->srcu);

arch/mips/include/asm/errno.h

@@ -121,6 +121,8 @@
 
 #define	ERFKILL		167	/* Operation not possible due to RF-kill */
 
+#define EHWPOISON	168	/* Memory page has hardware error */
+
 #define EDQUOT		1133	/* Quota exceeded */
 
 #ifdef __KERNEL__

arch/parisc/include/asm/errno.h

@@ -122,4 +122,6 @@
 
 #define	ERFKILL		256	/* Operation not possible due to RF-kill */
 
+#define EHWPOISON	257	/* Memory page has hardware error */
+
 #endif

arch/powerpc/kvm/book3s.c

@@ -1141,9 +1141,10 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->sprg1 = vcpu->arch.shared->sprg1;
 	regs->sprg2 = vcpu->arch.shared->sprg2;
 	regs->sprg3 = vcpu->arch.shared->sprg3;
-	regs->sprg5 = vcpu->arch.sprg4;
-	regs->sprg6 = vcpu->arch.sprg5;
-	regs->sprg7 = vcpu->arch.sprg6;
+	regs->sprg4 = vcpu->arch.sprg4;
+	regs->sprg5 = vcpu->arch.sprg5;
+	regs->sprg6 = vcpu->arch.sprg6;
+	regs->sprg7 = vcpu->arch.sprg7;
 
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
@@ -1167,9 +1168,10 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.shared->sprg1 = regs->sprg1;
 	vcpu->arch.shared->sprg2 = regs->sprg2;
 	vcpu->arch.shared->sprg3 = regs->sprg3;
-	vcpu->arch.sprg5 = regs->sprg4;
-	vcpu->arch.sprg6 = regs->sprg5;
-	vcpu->arch.sprg7 = regs->sprg6;
+	vcpu->arch.sprg4 = regs->sprg4;
+	vcpu->arch.sprg5 = regs->sprg5;
+	vcpu->arch.sprg6 = regs->sprg6;
+	vcpu->arch.sprg7 = regs->sprg7;
 
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);

arch/powerpc/kvm/booke.c

@@ -546,9 +546,10 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->sprg1 = vcpu->arch.shared->sprg1;
 	regs->sprg2 = vcpu->arch.shared->sprg2;
 	regs->sprg3 = vcpu->arch.shared->sprg3;
-	regs->sprg5 = vcpu->arch.sprg4;
-	regs->sprg6 = vcpu->arch.sprg5;
-	regs->sprg7 = vcpu->arch.sprg6;
+	regs->sprg4 = vcpu->arch.sprg4;
+	regs->sprg5 = vcpu->arch.sprg5;
+	regs->sprg6 = vcpu->arch.sprg6;
+	regs->sprg7 = vcpu->arch.sprg7;
 
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
@@ -572,9 +573,10 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.shared->sprg1 = regs->sprg1;
 	vcpu->arch.shared->sprg2 = regs->sprg2;
 	vcpu->arch.shared->sprg3 = regs->sprg3;
-	vcpu->arch.sprg5 = regs->sprg4;
-	vcpu->arch.sprg6 = regs->sprg5;
-	vcpu->arch.sprg7 = regs->sprg6;
+	vcpu->arch.sprg4 = regs->sprg4;
+	vcpu->arch.sprg5 = regs->sprg5;
+	vcpu->arch.sprg6 = regs->sprg6;
+	vcpu->arch.sprg7 = regs->sprg7;
 
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);

arch/sparc/include/asm/errno.h

@@ -112,4 +112,6 @@
 
 #define	ERFKILL		134	/* Operation not possible due to RF-kill */
 
+#define EHWPOISON	135	/* Memory page has hardware error */
+
 #endif

arch/x86/include/asm/kvm_emulate.h

@@ -142,9 +142,9 @@ struct x86_emulate_ops {
 	int (*pio_out_emulated)(int size, unsigned short port, const void *val,
 				unsigned int count, struct kvm_vcpu *vcpu);
 
-	bool (*get_cached_descriptor)(struct desc_struct *desc,
+	bool (*get_cached_descriptor)(struct desc_struct *desc, u32 *base3,
 				      int seg, struct kvm_vcpu *vcpu);
-	void (*set_cached_descriptor)(struct desc_struct *desc,
+	void (*set_cached_descriptor)(struct desc_struct *desc, u32 base3,
 				      int seg, struct kvm_vcpu *vcpu);
 	u16 (*get_segment_selector)(int seg, struct kvm_vcpu *vcpu);
 	void (*set_segment_selector)(u16 sel, int seg, struct kvm_vcpu *vcpu);
@@ -239,6 +239,7 @@ struct x86_emulate_ctxt {
 	int interruptibility;
 
 	bool perm_ok; /* do not check permissions if true */
+	bool only_vendor_specific_insn;
 
 	bool have_exception;
 	struct x86_exception exception;

arch/x86/include/asm/kvm_host.h

@@ -85,7 +85,7 @@
 
 #define ASYNC_PF_PER_VCPU 64
 
-extern spinlock_t kvm_lock;
+extern raw_spinlock_t kvm_lock;
 extern struct list_head vm_list;
 
 struct kvm_vcpu;
@@ -255,6 +255,8 @@ struct kvm_mmu {
 	int (*sync_page)(struct kvm_vcpu *vcpu,
 			 struct kvm_mmu_page *sp);
 	void (*invlpg)(struct kvm_vcpu *vcpu, gva_t gva);
+	void (*update_pte)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
+			u64 *spte, const void *pte, unsigned long mmu_seq);
 	hpa_t root_hpa;
 	int root_level;
 	int shadow_root_level;
@@ -335,12 +337,6 @@ struct kvm_vcpu_arch {
 	u64  *last_pte_updated;
 	gfn_t last_pte_gfn;
 
-	struct {
-		gfn_t gfn;	/* presumed gfn during guest pte update */
-		pfn_t pfn;	/* pfn corresponding to that gfn */
-		unsigned long mmu_seq;
-	} update_pte;
-
 	struct fpu guest_fpu;
 	u64 xcr0;
 
@@ -448,7 +444,7 @@ struct kvm_arch {
 
 	unsigned long irq_sources_bitmap;
 	s64 kvmclock_offset;
-	spinlock_t tsc_write_lock;
+	raw_spinlock_t tsc_write_lock;
 	u64 last_tsc_nsec;
 	u64 last_tsc_offset;
 	u64 last_tsc_write;

arch/x86/include/asm/msr-index.h

@@ -43,6 +43,7 @@
 
 #define MSR_MTRRcap			0x000000fe
 #define MSR_IA32_BBL_CR_CTL		0x00000119
+#define MSR_IA32_BBL_CR_CTL3		0x0000011e
 
 #define MSR_IA32_SYSENTER_CS		0x00000174
 #define MSR_IA32_SYSENTER_ESP		0x00000175

arch/x86/kernel/kvm.c

@@ -493,7 +493,7 @@ static void __init kvm_smp_prepare_boot_cpu(void)
 	native_smp_prepare_boot_cpu();
 }
 
-static void kvm_guest_cpu_online(void *dummy)
+static void __cpuinit kvm_guest_cpu_online(void *dummy)
 {
 	kvm_guest_cpu_init();
 }

arch/x86/kvm/emulate.c

@@ -76,6 +76,7 @@
 #define Group       (1<<14)     /* Bits 3:5 of modrm byte extend opcode */
 #define GroupDual   (1<<15)     /* Alternate decoding of mod == 3 */
 /* Misc flags */
+#define VendorSpecific (1<<22) /* Vendor specific instruction */
 #define NoAccess    (1<<23) /* Don't access memory (lea/invlpg/verr etc) */
 #define Op3264      (1<<24) /* Operand is 64b in long mode, 32b otherwise */
 #define Undefined   (1<<25) /* No Such Instruction */
@@ -877,7 +878,8 @@ static void get_descriptor_table_ptr(struct x86_emulate_ctxt *ctxt,
 	if (selector & 1 << 2) {
 		struct desc_struct desc;
 		memset (dt, 0, sizeof *dt);
-		if (!ops->get_cached_descriptor(&desc, VCPU_SREG_LDTR, ctxt->vcpu))
+		if (!ops->get_cached_descriptor(&desc, NULL, VCPU_SREG_LDTR,
+						ctxt->vcpu))
 			return;
 
 		dt->size = desc_limit_scaled(&desc); /* what if limit > 65535? */
@@ -929,6 +931,7 @@ static int write_segment_descriptor(struct x86_emulate_ctxt *ctxt,
 	return ret;
 }
 
+/* Does not support long mode */
 static int load_segment_descriptor(struct x86_emulate_ctxt *ctxt,
 				   struct x86_emulate_ops *ops,
 				   u16 selector, int seg)
@@ -1040,7 +1043,7 @@ static int load_segment_descriptor(struct x86_emulate_ctxt *ctxt,
 	}
 load:
 	ops->set_segment_selector(selector, seg, ctxt->vcpu);
-	ops->set_cached_descriptor(&seg_desc, seg, ctxt->vcpu);
+	ops->set_cached_descriptor(&seg_desc, 0, seg, ctxt->vcpu);
 	return X86EMUL_CONTINUE;
 exception:
 	emulate_exception(ctxt, err_vec, err_code, true);
@@ -1560,7 +1563,7 @@ setup_syscalls_segments(struct x86_emulate_ctxt *ctxt,
 			struct desc_struct *ss)
 {
 	memset(cs, 0, sizeof(struct desc_struct));
-	ops->get_cached_descriptor(cs, VCPU_SREG_CS, ctxt->vcpu);
+	ops->get_cached_descriptor(cs, NULL, VCPU_SREG_CS, ctxt->vcpu);
 	memset(ss, 0, sizeof(struct desc_struct));
 
 	cs->l = 0;		/* will be adjusted later */
@@ -1607,9 +1610,9 @@ emulate_syscall(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops)
 		cs.d = 0;
 		cs.l = 1;
 	}
-	ops->set_cached_descriptor(&cs, VCPU_SREG_CS, ctxt->vcpu);
+	ops->set_cached_descriptor(&cs, 0, VCPU_SREG_CS, ctxt->vcpu);
 	ops->set_segment_selector(cs_sel, VCPU_SREG_CS, ctxt->vcpu);
-	ops->set_cached_descriptor(&ss, VCPU_SREG_SS, ctxt->vcpu);
+	ops->set_cached_descriptor(&ss, 0, VCPU_SREG_SS, ctxt->vcpu);
 	ops->set_segment_selector(ss_sel, VCPU_SREG_SS, ctxt->vcpu);
 
 	c->regs[VCPU_REGS_RCX] = c->eip;
@@ -1679,9 +1682,9 @@ emulate_sysenter(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops)
 		cs.l = 1;
 	}
 
-	ops->set_cached_descriptor(&cs, VCPU_SREG_CS, ctxt->vcpu);
+	ops->set_cached_descriptor(&cs, 0, VCPU_SREG_CS, ctxt->vcpu);
 	ops->set_segment_selector(cs_sel, VCPU_SREG_CS, ctxt->vcpu);
-	ops->set_cached_descriptor(&ss, VCPU_SREG_SS, ctxt->vcpu);
+	ops->set_cached_descriptor(&ss, 0, VCPU_SREG_SS, ctxt->vcpu);
 	ops->set_segment_selector(ss_sel, VCPU_SREG_SS, ctxt->vcpu);
 
 	ops->get_msr(ctxt->vcpu, MSR_IA32_SYSENTER_EIP, &msr_data);
@@ -1736,9 +1739,9 @@ emulate_sysexit(struct x86_emulate_ctxt *ctxt, struct x86_emulate_ops *ops)
 	cs_sel |= SELECTOR_RPL_MASK;
 	ss_sel |= SELECTOR_RPL_MASK;
 
-	ops->set_cached_descriptor(&cs, VCPU_SREG_CS, ctxt->vcpu);
+	ops->set_cached_descriptor(&cs, 0, VCPU_SREG_CS, ctxt->vcpu);
 	ops->set_segment_selector(cs_sel, VCPU_SREG_CS, ctxt->vcpu);
-	ops->set_cached_descriptor(&ss, VCPU_SREG_SS, ctxt->vcpu);
+	ops->set_cached_descriptor(&ss, 0, VCPU_SREG_SS, ctxt->vcpu);
 	ops->set_segment_selector(ss_sel, VCPU_SREG_SS, ctxt->vcpu);
 
 	c->eip = c->regs[VCPU_REGS_RDX];
@@ -1764,24 +1767,28 @@ static bool emulator_io_port_access_allowed(struct x86_emulate_ctxt *ctxt,
 					    u16 port, u16 len)
 {
 	struct desc_struct tr_seg;
+	u32 base3;
 	int r;
-	u16 io_bitmap_ptr;
-	u8 perm, bit_idx = port & 0x7;
+	u16 io_bitmap_ptr, perm, bit_idx = port & 0x7;
 	unsigned mask = (1 << len) - 1;
+	unsigned long base;
 
-	ops->get_cached_descriptor(&tr_seg, VCPU_SREG_TR, ctxt->vcpu);
+	ops->get_cached_descriptor(&tr_seg, &base3, VCPU_SREG_TR, ctxt->vcpu);
 	if (!tr_seg.p)
 		return false;
 	if (desc_limit_scaled(&tr_seg) < 103)
 		return false;
-	r = ops->read_std(get_desc_base(&tr_seg) + 102, &io_bitmap_ptr, 2,
-			  ctxt->vcpu, NULL);
+	base = get_desc_base(&tr_seg);
+#ifdef CONFIG_X86_64
+	base |= ((u64)base3) << 32;
+#endif
+	r = ops->read_std(base + 102, &io_bitmap_ptr, 2, ctxt->vcpu, NULL);
 	if (r != X86EMUL_CONTINUE)
 		return false;
 	if (io_bitmap_ptr + port/8 > desc_limit_scaled(&tr_seg))
 		return false;
-	r = ops->read_std(get_desc_base(&tr_seg) + io_bitmap_ptr + port/8,
-			  &perm, 1, ctxt->vcpu, NULL);
+	r = ops->read_std(base + io_bitmap_ptr + port/8, &perm, 2, ctxt->vcpu,
+			  NULL);
 	if (r != X86EMUL_CONTINUE)
 		return false;
 	if ((perm >> bit_idx) & mask)
@@ -2126,7 +2133,7 @@ static int emulator_do_task_switch(struct x86_emulate_ctxt *ctxt,
 	}
 
 	ops->set_cr(0,  ops->get_cr(0, ctxt->vcpu) | X86_CR0_TS, ctxt->vcpu);
-	ops->set_cached_descriptor(&next_tss_desc, VCPU_SREG_TR, ctxt->vcpu);
+	ops->set_cached_descriptor(&next_tss_desc, 0, VCPU_SREG_TR, ctxt->vcpu);
 	ops->set_segment_selector(tss_selector, VCPU_SREG_TR, ctxt->vcpu);
 
 	if (has_error_code) {
@@ -2365,7 +2372,8 @@ static struct group_dual group7 = { {
 	D(SrcMem16 | ModRM | Mov | Priv),
 	D(SrcMem | ModRM | ByteOp | Priv | NoAccess),
 }, {
-	D(SrcNone | ModRM | Priv), N, N, D(SrcNone | ModRM | Priv),
+	D(SrcNone | ModRM | Priv | VendorSpecific), N,
+	N, D(SrcNone | ModRM | Priv | VendorSpecific),
 	D(SrcNone | ModRM | DstMem | Mov), N,
 	D(SrcMem16 | ModRM | Mov | Priv), N,
 } };
@@ -2489,7 +2497,7 @@ static struct opcode opcode_table[256] = {
 static struct opcode twobyte_table[256] = {
 	/* 0x00 - 0x0F */
 	N, GD(0, &group7), N, N,
-	N, D(ImplicitOps), D(ImplicitOps | Priv), N,
+	N, D(ImplicitOps | VendorSpecific), D(ImplicitOps | Priv), N,
 	D(ImplicitOps | Priv), D(ImplicitOps | Priv), N, N,
 	N, D(ImplicitOps | ModRM), N, N,
 	/* 0x10 - 0x1F */
@@ -2502,7 +2510,8 @@ static struct opcode twobyte_table[256] = {
 	/* 0x30 - 0x3F */
 	D(ImplicitOps | Priv), I(ImplicitOps, em_rdtsc),
 	D(ImplicitOps | Priv), N,
-	D(ImplicitOps), D(ImplicitOps | Priv), N, N,
+	D(ImplicitOps | VendorSpecific), D(ImplicitOps | Priv | VendorSpecific),
+	N, N,
 	N, N, N, N, N, N, N, N,
 	/* 0x40 - 0x4F */
 	X16(D(DstReg | SrcMem | ModRM | Mov)),
@@ -2741,6 +2750,9 @@ x86_decode_insn(struct x86_emulate_ctxt *ctxt, void *insn, int insn_len)
 	if (c->d == 0 || (c->d & Undefined))
 		return -1;
 
+	if (!(c->d & VendorSpecific) && ctxt->only_vendor_specific_insn)
+		return -1;
+
 	if (mode == X86EMUL_MODE_PROT64 && (c->d & Stack))
 		c->op_bytes = 8;
 

arch/x86/kvm/i8259.c

@@ -61,9 +61,6 @@ static void pic_unlock(struct kvm_pic *s)
 			}
 		}
 
-		if (!found)
-			found = s->kvm->bsp_vcpu;
-
 		if (!found)
 			return;
 
@@ -75,7 +72,6 @@ static void pic_unlock(struct kvm_pic *s)
 static void pic_clear_isr(struct kvm_kpic_state *s, int irq)
 {
 	s->isr &= ~(1 << irq);
-	s->isr_ack |= (1 << irq);
 	if (s != &s->pics_state->pics[0])
 		irq += 8;
 	/*
@@ -89,16 +85,6 @@ static void pic_clear_isr(struct kvm_kpic_state *s, int irq)
 	pic_lock(s->pics_state);
 }
 
-void kvm_pic_clear_isr_ack(struct kvm *kvm)
-{
-	struct kvm_pic *s = pic_irqchip(kvm);
-
-	pic_lock(s);
-	s->pics[0].isr_ack = 0xff;
-	s->pics[1].isr_ack = 0xff;
-	pic_unlock(s);
-}
-
 /*
  * set irq level. If an edge is detected, then the IRR is set to 1
  */
@@ -281,7 +267,6 @@ void kvm_pic_reset(struct kvm_kpic_state *s)
 	s->irr = 0;
 	s->imr = 0;
 	s->isr = 0;
-	s->isr_ack = 0xff;
 	s->priority_add = 0;
 	s->irq_base = 0;
 	s->read_reg_select = 0;
@@ -545,15 +530,11 @@ static int picdev_read(struct kvm_io_device *this,
  */
 static void pic_irq_request(struct kvm *kvm, int level)
 {
-	struct kvm_vcpu *vcpu = kvm->bsp_vcpu;
 	struct kvm_pic *s = pic_irqchip(kvm);
-	int irq = pic_get_irq(&s->pics[0]);
 
-	s->output = level;
-	if (vcpu && level && (s->pics[0].isr_ack & (1 << irq))) {
-		s->pics[0].isr_ack &= ~(1 << irq);
+	if (!s->output)
 		s->wakeup_needed = true;
-	}
+	s->output = level;
 }
 
 static const struct kvm_io_device_ops picdev_ops = {
@@ -575,8 +556,6 @@ struct kvm_pic *kvm_create_pic(struct kvm *kvm)
 	s->pics[1].elcr_mask = 0xde;
 	s->pics[0].pics_state = s;
 	s->pics[1].pics_state = s;
-	s->pics[0].isr_ack = 0xff;
-	s->pics[1].isr_ack = 0xff;
 
 	/*
 	 * Initialize PIO device

arch/x86/kvm/lapic.c

@@ -417,10 +417,6 @@ static int __apic_accept_irq(struct kvm_lapic *apic, int delivery_mode,
 	case APIC_DM_INIT:
 		if (level) {
 			result = 1;
-			if (vcpu->arch.mp_state == KVM_MP_STATE_RUNNABLE)
-				printk(KERN_DEBUG
-				       "INIT on a runnable vcpu %d\n",
-				       vcpu->vcpu_id);
 			vcpu->arch.mp_state = KVM_MP_STATE_INIT_RECEIVED;
 			kvm_make_request(KVM_REQ_EVENT, vcpu);
 			kvm_vcpu_kick(vcpu);
@@ -875,8 +871,8 @@ void kvm_free_lapic(struct kvm_vcpu *vcpu)
 
 	hrtimer_cancel(&vcpu->arch.apic->lapic_timer.timer);
 
-	if (vcpu->arch.apic->regs_page)
-		__free_page(vcpu->arch.apic->regs_page);
+	if (vcpu->arch.apic->regs)
+		free_page((unsigned long)vcpu->arch.apic->regs);
 
 	kfree(vcpu->arch.apic);
 }
@@ -1065,13 +1061,12 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu)
 
 	vcpu->arch.apic = apic;
 
-	apic->regs_page = alloc_page(GFP_KERNEL|__GFP_ZERO);
-	if (apic->regs_page == NULL) {
+	apic->regs = (void *)get_zeroed_page(GFP_KERNEL);
+	if (!apic->regs) {
 		printk(KERN_ERR "malloc apic regs error for vcpu %x\n",
 		       vcpu->vcpu_id);
 		goto nomem_free_apic;
 	}
-	apic->regs = page_address(apic->regs_page);
 	apic->vcpu = vcpu;
 
 	hrtimer_init(&apic->lapic_timer.timer, CLOCK_MONOTONIC,

arch/x86/kvm/lapic.h

@@ -13,7 +13,6 @@ struct kvm_lapic {
 	u32 divide_count;
 	struct kvm_vcpu *vcpu;
 	bool irr_pending;
-	struct page *regs_page;
 	void *regs;
 	gpa_t vapic_addr;
 	struct page *vapic_page;

arch/x86/kvm/mmu.c

@@ -111,9 +111,6 @@ module_param(oos_shadow, bool, 0644);
 #define PT64_LEVEL_SHIFT(level) \
 		(PAGE_SHIFT + (level - 1) * PT64_LEVEL_BITS)
 
-#define PT64_LEVEL_MASK(level) \
-		(((1ULL << PT64_LEVEL_BITS) - 1) << PT64_LEVEL_SHIFT(level))
-
 #define PT64_INDEX(address, level)\
 	(((address) >> PT64_LEVEL_SHIFT(level)) & ((1 << PT64_LEVEL_BITS) - 1))
 
@@ -123,8 +120,6 @@ module_param(oos_shadow, bool, 0644);
 #define PT32_LEVEL_SHIFT(level) \
 		(PAGE_SHIFT + (level - 1) * PT32_LEVEL_BITS)
 
-#define PT32_LEVEL_MASK(level) \
-		(((1ULL << PT32_LEVEL_BITS) - 1) << PT32_LEVEL_SHIFT(level))
 #define PT32_LVL_OFFSET_MASK(level) \
 	(PT32_BASE_ADDR_MASK & ((1ULL << (PAGE_SHIFT + (((level) - 1) \
 						* PT32_LEVEL_BITS))) - 1))
@@ -379,15 +374,15 @@ static void mmu_free_memory_cache(struct kvm_mmu_memory_cache *mc,
 static int mmu_topup_memory_cache_page(struct kvm_mmu_memory_cache *cache,
 				       int min)
 {
-	struct page *page;
+	void *page;
 
 	if (cache->nobjs >= min)
 		return 0;
 	while (cache->nobjs < ARRAY_SIZE(cache->objects)) {
-		page = alloc_page(GFP_KERNEL);
+		page = (void *)__get_free_page(GFP_KERNEL);
 		if (!page)
 			return -ENOMEM;
-		cache->objects[cache->nobjs++] = page_address(page);
+		cache->objects[cache->nobjs++] = page;
 	}
 	return 0;
 }
@@ -554,13 +549,23 @@ static int host_mapping_level(struct kvm *kvm, gfn_t gfn)
 	return ret;
 }
 
-static bool mapping_level_dirty_bitmap(struct kvm_vcpu *vcpu, gfn_t large_gfn)
+static struct kvm_memory_slot *
+gfn_to_memslot_dirty_bitmap(struct kvm_vcpu *vcpu, gfn_t gfn,
+			    bool no_dirty_log)
 {
 	struct kvm_memory_slot *slot;
-	slot = gfn_to_memslot(vcpu->kvm, large_gfn);
-	if (slot && slot->dirty_bitmap)
-		return true;
-	return false;
+
+	slot = gfn_to_memslot(vcpu->kvm, gfn);
+	if (!slot || slot->flags & KVM_MEMSLOT_INVALID ||
+	      (no_dirty_log && slot->dirty_bitmap))
+		slot = NULL;
+
+	return slot;
+}
+
+static bool mapping_level_dirty_bitmap(struct kvm_vcpu *vcpu, gfn_t large_gfn)
+{
+	return gfn_to_memslot_dirty_bitmap(vcpu, large_gfn, true);
 }
 
 static int mapping_level(struct kvm_vcpu *vcpu, gfn_t large_gfn)
@@ -1032,9 +1037,9 @@ static void kvm_mmu_free_page(struct kvm *kvm, struct kvm_mmu_page *sp)
 	ASSERT(is_empty_shadow_page(sp->spt));
 	hlist_del(&sp->hash_link);
 	list_del(&sp->link);
-	__free_page(virt_to_page(sp->spt));
+	free_page((unsigned long)sp->spt);
 	if (!sp->role.direct)
-		__free_page(virt_to_page(sp->gfns));
+		free_page((unsigned long)sp->gfns);
 	kmem_cache_free(mmu_page_header_cache, sp);
 	kvm_mod_used_mmu_pages(kvm, -1);
 }
@@ -1199,6 +1204,13 @@ static void nonpaging_invlpg(struct kvm_vcpu *vcpu, gva_t gva)
 {
 }
 
+static void nonpaging_update_pte(struct kvm_vcpu *vcpu,
+				 struct kvm_mmu_page *sp, u64 *spte,
+				 const void *pte, unsigned long mmu_seq)
+{
+	WARN_ON(1);
+}
+
 #define KVM_PAGE_ARRAY_NR 16
 
 struct kvm_mmu_pages {
@@ -2150,26 +2162,13 @@ static void nonpaging_new_cr3(struct kvm_vcpu *vcpu)
 {
 }
 
-static struct kvm_memory_slot *
-pte_prefetch_gfn_to_memslot(struct kvm_vcpu *vcpu, gfn_t gfn, bool no_dirty_log)
-{
-	struct kvm_memory_slot *slot;
-
-	slot = gfn_to_memslot(vcpu->kvm, gfn);
-	if (!slot || slot->flags & KVM_MEMSLOT_INVALID ||
-	      (no_dirty_log && slot->dirty_bitmap))
-		slot = NULL;
-
-	return slot;
-}
-
 static pfn_t pte_prefetch_gfn_to_pfn(struct kvm_vcpu *vcpu, gfn_t gfn,
 				     bool no_dirty_log)
 {
 	struct kvm_memory_slot *slot;
 	unsigned long hva;
 
-	slot = pte_prefetch_gfn_to_memslot(vcpu, gfn, no_dirty_log);
+	slot = gfn_to_memslot_dirty_bitmap(vcpu, gfn, no_dirty_log);
 	if (!slot) {
 		get_page(bad_page);
 		return page_to_pfn(bad_page);
@@ -2190,7 +2189,7 @@ static int direct_pte_prefetch_many(struct kvm_vcpu *vcpu,
 	gfn_t gfn;
 
 	gfn = kvm_mmu_page_get_gfn(sp, start - sp->spt);
-	if (!pte_prefetch_gfn_to_memslot(vcpu, gfn, access & ACC_WRITE_MASK))
+	if (!gfn_to_memslot_dirty_bitmap(vcpu, gfn, access & ACC_WRITE_MASK))
 		return -1;
 
 	ret = gfn_to_page_many_atomic(vcpu->kvm, gfn, pages, end - start);
@@ -2804,6 +2803,7 @@ static int nonpaging_init_context(struct kvm_vcpu *vcpu,
 	context->prefetch_page = nonpaging_prefetch_page;
 	context->sync_page = nonpaging_sync_page;
 	context->invlpg = nonpaging_invlpg;
+	context->update_pte = nonpaging_update_pte;
 	context->root_level = 0;
 	context->shadow_root_level = PT32E_ROOT_LEVEL;
 	context->root_hpa = INVALID_PAGE;
@@ -2933,6 +2933,7 @@ static int paging64_init_context_common(struct kvm_vcpu *vcpu,
 	context->prefetch_page = paging64_prefetch_page;
 	context->sync_page = paging64_sync_page;
 	context->invlpg = paging64_invlpg;
+	context->update_pte = paging64_update_pte;
 	context->free = paging_free;
 	context->root_level = level;
 	context->shadow_root_level = level;
@@ -2961,6 +2962,7 @@ static int paging32_init_context(struct kvm_vcpu *vcpu,
 	context->prefetch_page = paging32_prefetch_page;
 	context->sync_page = paging32_sync_page;
 	context->invlpg = paging32_invlpg;
+	context->update_pte = paging32_update_pte;
 	context->root_level = PT32_ROOT_LEVEL;
 	context->shadow_root_level = PT32E_ROOT_LEVEL;
 	context->root_hpa = INVALID_PAGE;
@@ -2985,6 +2987,7 @@ static int init_kvm_tdp_mmu(struct kvm_vcpu *vcpu)
 	context->prefetch_page = nonpaging_prefetch_page;
 	context->sync_page = nonpaging_sync_page;
 	context->invlpg = nonpaging_invlpg;
+	context->update_pte = nonpaging_update_pte;
 	context->shadow_root_level = kvm_x86_ops->get_tdp_level();
 	context->root_hpa = INVALID_PAGE;
 	context->direct_map = true;
@@ -3089,8 +3092,6 @@ static int init_kvm_nested_mmu(struct kvm_vcpu *vcpu)
 
 static int init_kvm_mmu(struct kvm_vcpu *vcpu)
 {
-	vcpu->arch.update_pte.pfn = bad_pfn;
-
 	if (mmu_is_nested(vcpu))
 		return init_kvm_nested_mmu(vcpu);
 	else if (tdp_enabled)
@@ -3164,7 +3165,7 @@ static void mmu_pte_write_zap_pte(struct kvm_vcpu *vcpu,
 static void mmu_pte_write_new_pte(struct kvm_vcpu *vcpu,
 				  struct kvm_mmu_page *sp,
 				  u64 *spte,
-				  const void *new)
+				  const void *new, unsigned long mmu_seq)
 {
 	if (sp->role.level != PT_PAGE_TABLE_LEVEL) {
 		++vcpu->kvm->stat.mmu_pde_zapped;
@@ -3172,10 +3173,7 @@ static void mmu_pte_write_new_pte(struct kvm_vcpu *vcpu,
         }
 
 	++vcpu->kvm->stat.mmu_pte_updated;
-	if (!sp->role.cr4_pae)
-		paging32_update_pte(vcpu, sp, spte, new);
-	else
-		paging64_update_pte(vcpu, sp, spte, new);
+	vcpu->arch.mmu.update_pte(vcpu, sp, spte, new, mmu_seq);
 }
 
 static bool need_remote_flush(u64 old, u64 new)
@@ -3210,28 +3208,6 @@ static bool last_updated_pte_accessed(struct kvm_vcpu *vcpu)
 	return !!(spte && (*spte & shadow_accessed_mask));
 }
 
-static void mmu_guess_page_from_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
-					  u64 gpte)
-{
-	gfn_t gfn;
-	pfn_t pfn;
-
-	if (!is_present_gpte(gpte))
-		return;
-	gfn = (gpte & PT64_BASE_ADDR_MASK) >> PAGE_SHIFT;
-
-	vcpu->arch.update_pte.mmu_seq = vcpu->kvm->mmu_notifier_seq;
-	smp_rmb();
-	pfn = gfn_to_pfn(vcpu->kvm, gfn);
-
-	if (is_error_pfn(pfn)) {
-		kvm_release_pfn_clean(pfn);
-		return;
-	}
-	vcpu->arch.update_pte.gfn = gfn;
-	vcpu->arch.update_pte.pfn = pfn;
-}
-
 static void kvm_mmu_access_page(struct kvm_vcpu *vcpu, gfn_t gfn)
 {
 	u64 *spte = vcpu->arch.last_pte_updated;
@@ -3253,21 +3229,14 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	struct kvm_mmu_page *sp;
 	struct hlist_node *node;
 	LIST_HEAD(invalid_list);
-	u64 entry, gentry;
-	u64 *spte;
-	unsigned offset = offset_in_page(gpa);
-	unsigned pte_size;
-	unsigned page_offset;
-	unsigned misaligned;
-	unsigned quadrant;
-	int level;
-	int flooded = 0;
-	int npte;
-	int r;
-	int invlpg_counter;
+	unsigned long mmu_seq;
+	u64 entry, gentry, *spte;
+	unsigned pte_size, page_offset, misaligned, quadrant, offset;
+	int level, npte, invlpg_counter, r, flooded = 0;
 	bool remote_flush, local_flush, zap_page;
 
 	zap_page = remote_flush = local_flush = false;
+	offset = offset_in_page(gpa);
 
 	pgprintk("%s: gpa %llx bytes %d\n", __func__, gpa, bytes);
 
@@ -3275,9 +3244,8 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 
 	/*
 	 * Assume that the pte write on a page table of the same type
-	 * as the current vcpu paging mode.  This is nearly always true
-	 * (might be false while changing modes).  Note it is verified later
-	 * by update_pte().
+	 * as the current vcpu paging mode since we update the sptes only
+	 * when they have the same mode.
 	 */
 	if ((is_pae(vcpu) && bytes == 4) || !new) {
 		/* Handle a 32-bit guest writing two halves of a 64-bit gpte */
@@ -3303,15 +3271,17 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 		break;
 	}
 
-	mmu_guess_page_from_pte_write(vcpu, gpa, gentry);
+	mmu_seq = vcpu->kvm->mmu_notifier_seq;
+	smp_rmb();
+
 	spin_lock(&vcpu->kvm->mmu_lock);
 	if (atomic_read(&vcpu->kvm->arch.invlpg_counter) != invlpg_counter)
 		gentry = 0;
-	kvm_mmu_access_page(vcpu, gfn);
 	kvm_mmu_free_some_pages(vcpu);
 	++vcpu->kvm->stat.mmu_pte_write;
 	trace_kvm_mmu_audit(vcpu, AUDIT_PRE_PTE_WRITE);
 	if (guest_initiated) {
+		kvm_mmu_access_page(vcpu, gfn);
 		if (gfn == vcpu->arch.last_pt_write_gfn
 		    && !last_updated_pte_accessed(vcpu)) {
 			++vcpu->arch.last_pt_write_count;
@@ -3375,7 +3345,8 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 			if (gentry &&
 			      !((sp->role.word ^ vcpu->arch.mmu.base_role.word)
 			      & mask.word))
-				mmu_pte_write_new_pte(vcpu, sp, spte, &gentry);
+				mmu_pte_write_new_pte(vcpu, sp, spte, &gentry,
+						      mmu_seq);
 			if (!remote_flush && need_remote_flush(entry, *spte))
 				remote_flush = true;
 			++spte;
@@ -3385,10 +3356,6 @@ void kvm_mmu_pte_write(struct kvm_vcpu *vcpu, gpa_t gpa,
 	kvm_mmu_commit_zap_page(vcpu->kvm, &invalid_list);
 	trace_kvm_mmu_audit(vcpu, AUDIT_POST_PTE_WRITE);
 	spin_unlock(&vcpu->kvm->mmu_lock);
-	if (!is_error_pfn(vcpu->arch.update_pte.pfn)) {
-		kvm_release_pfn_clean(vcpu->arch.update_pte.pfn);
-		vcpu->arch.update_pte.pfn = bad_pfn;
-	}
 }
 
 int kvm_mmu_unprotect_page_virt(struct kvm_vcpu *vcpu, gva_t gva)
@@ -3538,14 +3505,23 @@ void kvm_mmu_slot_remove_write_access(struct kvm *kvm, int slot)
 		if (!test_bit(slot, sp->slot_bitmap))
 			continue;
 
-		if (sp->role.level != PT_PAGE_TABLE_LEVEL)
-			continue;
-
 		pt = sp->spt;
-		for (i = 0; i < PT64_ENT_PER_PAGE; ++i)
+		for (i = 0; i < PT64_ENT_PER_PAGE; ++i) {
+			if (!is_shadow_present_pte(pt[i]) ||
+			      !is_last_spte(pt[i], sp->role.level))
+				continue;
+
+			if (is_large_pte(pt[i])) {
+				drop_spte(kvm, &pt[i],
+					  shadow_trap_nonpresent_pte);
+				--kvm->stat.lpages;
+				continue;
+			}
+
 			/* avoid RMW */
 			if (is_writable_pte(pt[i]))
 				update_spte(&pt[i], pt[i] & ~PT_WRITABLE_MASK);
+		}
 	}
 	kvm_flush_remote_tlbs(kvm);
 }
@@ -3583,7 +3559,7 @@ static int mmu_shrink(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask)
 	if (nr_to_scan == 0)
 		goto out;
 
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 
 	list_for_each_entry(kvm, &vm_list, vm_list) {
 		int idx, freed_pages;
@@ -3606,7 +3582,7 @@ static int mmu_shrink(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask)
 	if (kvm_freed)
 		list_move_tail(&kvm_freed->vm_list, &vm_list);
 
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 
 out:
 	return percpu_counter_read_positive(&kvm_total_used_mmu_pages);

arch/x86/kvm/paging_tmpl.h

@@ -31,7 +31,6 @@
 	#define PT_LVL_ADDR_MASK(lvl) PT64_LVL_ADDR_MASK(lvl)
 	#define PT_LVL_OFFSET_MASK(lvl) PT64_LVL_OFFSET_MASK(lvl)
 	#define PT_INDEX(addr, level) PT64_INDEX(addr, level)
-	#define PT_LEVEL_MASK(level) PT64_LEVEL_MASK(level)
 	#define PT_LEVEL_BITS PT64_LEVEL_BITS
 	#ifdef CONFIG_X86_64
 	#define PT_MAX_FULL_LEVELS 4
@@ -48,7 +47,6 @@
 	#define PT_LVL_ADDR_MASK(lvl) PT32_LVL_ADDR_MASK(lvl)
 	#define PT_LVL_OFFSET_MASK(lvl) PT32_LVL_OFFSET_MASK(lvl)
 	#define PT_INDEX(addr, level) PT32_INDEX(addr, level)
-	#define PT_LEVEL_MASK(level) PT32_LEVEL_MASK(level)
 	#define PT_LEVEL_BITS PT32_LEVEL_BITS
 	#define PT_MAX_FULL_LEVELS 2
 	#define CMPXCHG cmpxchg
@@ -327,7 +325,7 @@ static bool FNAME(prefetch_invalid_gpte)(struct kvm_vcpu *vcpu,
 }
 
 static void FNAME(update_pte)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
-			      u64 *spte, const void *pte)
+			      u64 *spte, const void *pte, unsigned long mmu_seq)
 {
 	pt_element_t gpte;
 	unsigned pte_access;
@@ -339,14 +337,14 @@ static void FNAME(update_pte)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp,
 
 	pgprintk("%s: gpte %llx spte %p\n", __func__, (u64)gpte, spte);
 	pte_access = sp->role.access & FNAME(gpte_access)(vcpu, gpte);
-	if (gpte_to_gfn(gpte) != vcpu->arch.update_pte.gfn)
+	pfn = gfn_to_pfn_atomic(vcpu->kvm, gpte_to_gfn(gpte));
+	if (is_error_pfn(pfn)) {
+		kvm_release_pfn_clean(pfn);
 		return;
-	pfn = vcpu->arch.update_pte.pfn;
-	if (is_error_pfn(pfn))
-		return;
-	if (mmu_notifier_retry(vcpu, vcpu->arch.update_pte.mmu_seq))
+	}
+	if (mmu_notifier_retry(vcpu, mmu_seq))
 		return;
-	kvm_get_pfn(pfn);
+
 	/*
 	 * we call mmu_set_spte() with host_writable = true beacuse that
 	 * vcpu->arch.update_pte.pfn was fetched from get_user_pages(write = 1).
@@ -829,7 +827,6 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 #undef FNAME
 #undef PT_BASE_ADDR_MASK
 #undef PT_INDEX
-#undef PT_LEVEL_MASK
 #undef PT_LVL_ADDR_MASK
 #undef PT_LVL_OFFSET_MASK
 #undef PT_LEVEL_BITS

arch/x86/kvm/svm.c

@@ -135,6 +135,8 @@ struct vcpu_svm {
 
 	u32 *msrpm;
 
+	ulong nmi_iret_rip;
+
 	struct nested_state nested;
 
 	bool nmi_singlestep;
@@ -1153,7 +1155,9 @@ static void svm_vcpu_put(struct kvm_vcpu *vcpu)
 	wrmsrl(MSR_KERNEL_GS_BASE, current->thread.gs);
 	load_gs_index(svm->host.gs);
 #else
+#ifdef CONFIG_X86_32_LAZY_GS
 	loadsegment(gs, svm->host.gs);
+#endif
 #endif
 	for (i = 0; i < NR_HOST_SAVE_USER_MSRS; i++)
 		wrmsrl(host_save_user_msrs[i], svm->host_user_msrs[i]);
@@ -2653,6 +2657,7 @@ static int iret_interception(struct vcpu_svm *svm)
 	++svm->vcpu.stat.nmi_window_exits;
 	clr_intercept(svm, INTERCEPT_IRET);
 	svm->vcpu.arch.hflags |= HF_IRET_MASK;
+	svm->nmi_iret_rip = kvm_rip_read(&svm->vcpu);
 	return 1;
 }
 
@@ -3474,7 +3479,12 @@ static void svm_complete_interrupts(struct vcpu_svm *svm)
 
 	svm->int3_injected = 0;
 
-	if (svm->vcpu.arch.hflags & HF_IRET_MASK) {
+	/*
+	 * If we've made progress since setting HF_IRET_MASK, we've
+	 * executed an IRET and can allow NMI injection.
+	 */
+	if ((svm->vcpu.arch.hflags & HF_IRET_MASK)
+	    && kvm_rip_read(&svm->vcpu) != svm->nmi_iret_rip) {
 		svm->vcpu.arch.hflags &= ~(HF_NMI_MASK | HF_IRET_MASK);
 		kvm_make_request(KVM_REQ_EVENT, &svm->vcpu);
 	}
@@ -3641,19 +3651,30 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu)
 	wrmsrl(MSR_GS_BASE, svm->host.gs_base);
 #else
 	loadsegment(fs, svm->host.fs);
+#ifndef CONFIG_X86_32_LAZY_GS
+	loadsegment(gs, svm->host.gs);
+#endif
 #endif
 
 	reload_tss(vcpu);
 
 	local_irq_disable();
 
-	stgi();
-
 	vcpu->arch.cr2 = svm->vmcb->save.cr2;
 	vcpu->arch.regs[VCPU_REGS_RAX] = svm->vmcb->save.rax;
 	vcpu->arch.regs[VCPU_REGS_RSP] = svm->vmcb->save.rsp;
 	vcpu->arch.regs[VCPU_REGS_RIP] = svm->vmcb->save.rip;
 
+	if (unlikely(svm->vmcb->control.exit_code == SVM_EXIT_NMI))
+		kvm_before_handle_nmi(&svm->vcpu);
+
+	stgi();
+
+	/* Any pending NMI will happen here */
+
+	if (unlikely(svm->vmcb->control.exit_code == SVM_EXIT_NMI))
+		kvm_after_handle_nmi(&svm->vcpu);
+
 	sync_cr8_to_lapic(vcpu);
 
 	svm->next_rip = 0;

arch/x86/kvm/vmx.c

@@ -93,14 +93,14 @@ module_param(yield_on_hlt, bool, S_IRUGO);
  * These 2 parameters are used to config the controls for Pause-Loop Exiting:
  * ple_gap:    upper bound on the amount of time between two successive
  *             executions of PAUSE in a loop. Also indicate if ple enabled.
- *             According to test, this time is usually small than 41 cycles.
+ *             According to test, this time is usually smaller than 128 cycles.
  * ple_window: upper bound on the amount of time a guest is allowed to execute
  *             in a PAUSE loop. Tests indicate that most spinlocks are held for
  *             less than 2^12 cycles
  * Time is measured based on a counter that runs at the same rate as the TSC,
  * refer SDM volume 3b section 21.6.13 & 22.1.3.
  */
-#define KVM_VMX_DEFAULT_PLE_GAP    41
+#define KVM_VMX_DEFAULT_PLE_GAP    128
 #define KVM_VMX_DEFAULT_PLE_WINDOW 4096
 static int ple_gap = KVM_VMX_DEFAULT_PLE_GAP;
 module_param(ple_gap, int, S_IRUGO);
@@ -176,11 +176,11 @@ static inline struct vcpu_vmx *to_vmx(struct kvm_vcpu *vcpu)
 	return container_of(vcpu, struct vcpu_vmx, vcpu);
 }
 
-static int init_rmode(struct kvm *kvm);
 static u64 construct_eptp(unsigned long root_hpa);
 static void kvm_cpu_vmxon(u64 addr);
 static void kvm_cpu_vmxoff(void);
 static void vmx_set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3);
+static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr);
 
 static DEFINE_PER_CPU(struct vmcs *, vmxarea);
 static DEFINE_PER_CPU(struct vmcs *, current_vmcs);
@@ -1333,19 +1333,25 @@ static __init int vmx_disabled_by_bios(void)
 
 	rdmsrl(MSR_IA32_FEATURE_CONTROL, msr);
 	if (msr & FEATURE_CONTROL_LOCKED) {
+		/* launched w/ TXT and VMX disabled */
 		if (!(msr & FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX)
 			&& tboot_enabled())
 			return 1;
+		/* launched w/o TXT and VMX only enabled w/ TXT */
 		if (!(msr & FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX)
+			&& (msr & FEATURE_CONTROL_VMXON_ENABLED_INSIDE_SMX)
 			&& !tboot_enabled()) {
 			printk(KERN_WARNING "kvm: disable TXT in the BIOS or "
-				" activate TXT before enabling KVM\n");
+				"activate TXT before enabling KVM\n");
 			return 1;
 		}
+		/* launched w/o TXT and VMX disabled */
+		if (!(msr & FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX)
+			&& !tboot_enabled())
+			return 1;
 	}
 
 	return 0;
-	/* locked but not enabled */
 }
 
 static void kvm_cpu_vmxon(u64 addr)
@@ -1683,6 +1689,7 @@ static void enter_pmode(struct kvm_vcpu *vcpu)
 	vmx->emulation_required = 1;
 	vmx->rmode.vm86_active = 0;
 
+	vmcs_write16(GUEST_TR_SELECTOR, vmx->rmode.tr.selector);
 	vmcs_writel(GUEST_TR_BASE, vmx->rmode.tr.base);
 	vmcs_write32(GUEST_TR_LIMIT, vmx->rmode.tr.limit);
 	vmcs_write32(GUEST_TR_AR_BYTES, vmx->rmode.tr.ar);
@@ -1756,6 +1763,19 @@ static void enter_rmode(struct kvm_vcpu *vcpu)
 	vmx->emulation_required = 1;
 	vmx->rmode.vm86_active = 1;
 
+	/*
+	 * Very old userspace does not call KVM_SET_TSS_ADDR before entering
+	 * vcpu. Call it here with phys address pointing 16M below 4G.
+	 */
+	if (!vcpu->kvm->arch.tss_addr) {
+		printk_once(KERN_WARNING "kvm: KVM_SET_TSS_ADDR need to be "
+			     "called before entering vcpu\n");
+		srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx);
+		vmx_set_tss_addr(vcpu->kvm, 0xfeffd000);
+		vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu);
+	}
+
+	vmx->rmode.tr.selector = vmcs_read16(GUEST_TR_SELECTOR);
 	vmx->rmode.tr.base = vmcs_readl(GUEST_TR_BASE);
 	vmcs_writel(GUEST_TR_BASE, rmode_tss_base(vcpu->kvm));
 
@@ -1794,7 +1814,6 @@ static void enter_rmode(struct kvm_vcpu *vcpu)
 
 continue_rmode:
 	kvm_mmu_reset_context(vcpu);
-	init_rmode(vcpu->kvm);
 }
 
 static void vmx_set_efer(struct kvm_vcpu *vcpu, u64 efer)
@@ -2030,23 +2049,40 @@ static void vmx_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
 	vmcs_writel(GUEST_CR4, hw_cr4);
 }
 
-static u64 vmx_get_segment_base(struct kvm_vcpu *vcpu, int seg)
-{
-	struct kvm_vmx_segment_field *sf = &kvm_vmx_segment_fields[seg];
-
-	return vmcs_readl(sf->base);
-}
-
 static void vmx_get_segment(struct kvm_vcpu *vcpu,
 			    struct kvm_segment *var, int seg)
 {
+	struct vcpu_vmx *vmx = to_vmx(vcpu);
 	struct kvm_vmx_segment_field *sf = &kvm_vmx_segment_fields[seg];
+	struct kvm_save_segment *save;
 	u32 ar;
 
+	if (vmx->rmode.vm86_active
+	    && (seg == VCPU_SREG_TR || seg == VCPU_SREG_ES
+		|| seg == VCPU_SREG_DS || seg == VCPU_SREG_FS
+		|| seg == VCPU_SREG_GS)
+	    && !emulate_invalid_guest_state) {
+		switch (seg) {
+		case VCPU_SREG_TR: save = &vmx->rmode.tr; break;
+		case VCPU_SREG_ES: save = &vmx->rmode.es; break;
+		case VCPU_SREG_DS: save = &vmx->rmode.ds; break;
+		case VCPU_SREG_FS: save = &vmx->rmode.fs; break;
+		case VCPU_SREG_GS: save = &vmx->rmode.gs; break;
+		default: BUG();
+		}
+		var->selector = save->selector;
+		var->base = save->base;
+		var->limit = save->limit;
+		ar = save->ar;
+		if (seg == VCPU_SREG_TR
+		    || var->selector == vmcs_read16(sf->selector))
+			goto use_saved_rmode_seg;
+	}
 	var->base = vmcs_readl(sf->base);
 	var->limit = vmcs_read32(sf->limit);
 	var->selector = vmcs_read16(sf->selector);
 	ar = vmcs_read32(sf->ar_bytes);
+use_saved_rmode_seg:
 	if ((ar & AR_UNUSABLE_MASK) && !emulate_invalid_guest_state)
 		ar = 0;
 	var->type = ar & 15;
@@ -2060,6 +2096,18 @@ static void vmx_get_segment(struct kvm_vcpu *vcpu,
 	var->unusable = (ar >> 16) & 1;
 }
 
+static u64 vmx_get_segment_base(struct kvm_vcpu *vcpu, int seg)
+{
+	struct kvm_vmx_segment_field *sf = &kvm_vmx_segment_fields[seg];
+	struct kvm_segment s;
+
+	if (to_vmx(vcpu)->rmode.vm86_active) {
+		vmx_get_segment(vcpu, &s, seg);
+		return s.base;
+	}
+	return vmcs_readl(sf->base);
+}
+
 static int vmx_get_cpl(struct kvm_vcpu *vcpu)
 {
 	if (!is_protmode(vcpu))
@@ -2101,6 +2149,7 @@ static void vmx_set_segment(struct kvm_vcpu *vcpu,
 	u32 ar;
 
 	if (vmx->rmode.vm86_active && seg == VCPU_SREG_TR) {
+		vmcs_write16(sf->selector, var->selector);
 		vmx->rmode.tr.selector = var->selector;
 		vmx->rmode.tr.base = var->base;
 		vmx->rmode.tr.limit = var->limit;
@@ -2361,11 +2410,12 @@ static bool guest_state_valid(struct kvm_vcpu *vcpu)
 
 static int init_rmode_tss(struct kvm *kvm)
 {
-	gfn_t fn = rmode_tss_base(kvm) >> PAGE_SHIFT;
+	gfn_t fn;
 	u16 data = 0;
-	int ret = 0;
-	int r;
+	int r, idx, ret = 0;
 
+	idx = srcu_read_lock(&kvm->srcu);
+	fn = rmode_tss_base(kvm) >> PAGE_SHIFT;
 	r = kvm_clear_guest_page(kvm, fn, 0, PAGE_SIZE);
 	if (r < 0)
 		goto out;
@@ -2389,12 +2439,13 @@ static int init_rmode_tss(struct kvm *kvm)
 
 	ret = 1;
 out:
+	srcu_read_unlock(&kvm->srcu, idx);
 	return ret;
 }
 
 static int init_rmode_identity_map(struct kvm *kvm)
 {
-	int i, r, ret;
+	int i, idx, r, ret;
 	pfn_t identity_map_pfn;
 	u32 tmp;
 
@@ -2409,6 +2460,7 @@ static int init_rmode_identity_map(struct kvm *kvm)
 		return 1;
 	ret = 0;
 	identity_map_pfn = kvm->arch.ept_identity_map_addr >> PAGE_SHIFT;
+	idx = srcu_read_lock(&kvm->srcu);
 	r = kvm_clear_guest_page(kvm, identity_map_pfn, 0, PAGE_SIZE);
 	if (r < 0)
 		goto out;
@@ -2424,6 +2476,7 @@ static int init_rmode_identity_map(struct kvm *kvm)
 	kvm->arch.ept_identity_pagetable_done = true;
 	ret = 1;
 out:
+	srcu_read_unlock(&kvm->srcu, idx);
 	return ret;
 }
 
@@ -2699,22 +2752,6 @@ static int vmx_vcpu_setup(struct vcpu_vmx *vmx)
 	return 0;
 }
 
-static int init_rmode(struct kvm *kvm)
-{
-	int idx, ret = 0;
-
-	idx = srcu_read_lock(&kvm->srcu);
-	if (!init_rmode_tss(kvm))
-		goto exit;
-	if (!init_rmode_identity_map(kvm))
-		goto exit;
-
-	ret = 1;
-exit:
-	srcu_read_unlock(&kvm->srcu, idx);
-	return ret;
-}
-
 static int vmx_vcpu_reset(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
@@ -2722,10 +2759,6 @@ static int vmx_vcpu_reset(struct kvm_vcpu *vcpu)
 	int ret;
 
 	vcpu->arch.regs_avail = ~((1 << VCPU_REGS_RIP) | (1 << VCPU_REGS_RSP));
-	if (!init_rmode(vmx->vcpu.kvm)) {
-		ret = -ENOMEM;
-		goto out;
-	}
 
 	vmx->rmode.vm86_active = 0;
 
@@ -2805,7 +2838,7 @@ static int vmx_vcpu_reset(struct kvm_vcpu *vcpu)
 		vmcs_write64(VIRTUAL_APIC_PAGE_ADDR, 0);
 		if (vm_need_tpr_shadow(vmx->vcpu.kvm))
 			vmcs_write64(VIRTUAL_APIC_PAGE_ADDR,
-				page_to_phys(vmx->vcpu.arch.apic->regs_page));
+				     __pa(vmx->vcpu.arch.apic->regs));
 		vmcs_write32(TPR_THRESHOLD, 0);
 	}
 
@@ -2971,6 +3004,9 @@ static int vmx_set_tss_addr(struct kvm *kvm, unsigned int addr)
 	if (ret)
 		return ret;
 	kvm->arch.tss_addr = addr;
+	if (!init_rmode_tss(kvm))
+		return  -ENOMEM;
+
 	return 0;
 }
 
@@ -3962,7 +3998,7 @@ static void vmx_cancel_injection(struct kvm_vcpu *vcpu)
 #define Q "l"
 #endif
 
-static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
+static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
 
@@ -3991,6 +4027,7 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
 	asm(
 		/* Store host registers */
 		"push %%"R"dx; push %%"R"bp;"
+		"push %%"R"cx \n\t" /* placeholder for guest rcx */
 		"push %%"R"cx \n\t"
 		"cmp %%"R"sp, %c[host_rsp](%0) \n\t"
 		"je 1f \n\t"
@@ -4032,10 +4069,11 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
 		".Llaunched: " __ex(ASM_VMX_VMRESUME) "\n\t"
 		".Lkvm_vmx_return: "
 		/* Save guest registers, load host registers, keep flags */
-		"xchg %0,     (%%"R"sp) \n\t"
+		"mov %0, %c[wordsize](%%"R"sp) \n\t"
+		"pop %0 \n\t"
 		"mov %%"R"ax, %c[rax](%0) \n\t"
 		"mov %%"R"bx, %c[rbx](%0) \n\t"
-		"push"Q" (%%"R"sp); pop"Q" %c[rcx](%0) \n\t"
+		"pop"Q" %c[rcx](%0) \n\t"
 		"mov %%"R"dx, %c[rdx](%0) \n\t"
 		"mov %%"R"si, %c[rsi](%0) \n\t"
 		"mov %%"R"di, %c[rdi](%0) \n\t"
@@ -4053,7 +4091,7 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
 		"mov %%cr2, %%"R"ax   \n\t"
 		"mov %%"R"ax, %c[cr2](%0) \n\t"
 
-		"pop  %%"R"bp; pop  %%"R"bp; pop  %%"R"dx \n\t"
+		"pop  %%"R"bp; pop  %%"R"dx \n\t"
 		"setbe %c[fail](%0) \n\t"
 	      : : "c"(vmx), "d"((unsigned long)HOST_RSP),
 		[launched]"i"(offsetof(struct vcpu_vmx, launched)),
@@ -4076,7 +4114,8 @@ static void vmx_vcpu_run(struct kvm_vcpu *vcpu)
 		[r14]"i"(offsetof(struct vcpu_vmx, vcpu.arch.regs[VCPU_REGS_R14])),
 		[r15]"i"(offsetof(struct vcpu_vmx, vcpu.arch.regs[VCPU_REGS_R15])),
 #endif
-		[cr2]"i"(offsetof(struct vcpu_vmx, vcpu.arch.cr2))
+		[cr2]"i"(offsetof(struct vcpu_vmx, vcpu.arch.cr2)),
+		[wordsize]"i"(sizeof(ulong))
 	      : "cc", "memory"
 		, R"ax", R"bx", R"di", R"si"
 #ifdef CONFIG_X86_64
@@ -4183,8 +4222,11 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id)
 		if (!kvm->arch.ept_identity_map_addr)
 			kvm->arch.ept_identity_map_addr =
 				VMX_EPT_IDENTITY_PAGETABLE_ADDR;
+		err = -ENOMEM;
 		if (alloc_identity_pagetable(kvm) != 0)
 			goto free_vmcs;
+		if (!init_rmode_identity_map(kvm))
+			goto free_vmcs;
 	}
 
 	return &vmx->vcpu;

arch/x86/kvm/x86.c

@@ -81,9 +81,10 @@
  * - enable LME and LMA per default on 64 bit KVM
  */
 #ifdef CONFIG_X86_64
-static u64 __read_mostly efer_reserved_bits = 0xfffffffffffffafeULL;
+static
+u64 __read_mostly efer_reserved_bits = ~((u64)(EFER_SCE | EFER_LME | EFER_LMA));
 #else
-static u64 __read_mostly efer_reserved_bits = 0xfffffffffffffffeULL;
+static u64 __read_mostly efer_reserved_bits = ~((u64)EFER_SCE);
 #endif
 
 #define VM_STAT(x) offsetof(struct kvm, stat.x), KVM_STAT_VM
@@ -360,8 +361,8 @@ void kvm_propagate_fault(struct kvm_vcpu *vcpu, struct x86_exception *fault)
 
 void kvm_inject_nmi(struct kvm_vcpu *vcpu)
 {
+	kvm_make_request(KVM_REQ_NMI, vcpu);
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
-	vcpu->arch.nmi_pending = 1;
 }
 EXPORT_SYMBOL_GPL(kvm_inject_nmi);
 
@@ -525,8 +526,10 @@ int kvm_set_cr0(struct kvm_vcpu *vcpu, unsigned long cr0)
 
 	kvm_x86_ops->set_cr0(vcpu, cr0);
 
-	if ((cr0 ^ old_cr0) & X86_CR0_PG)
+	if ((cr0 ^ old_cr0) & X86_CR0_PG) {
 		kvm_clear_async_pf_completion_queue(vcpu);
+		kvm_async_pf_hash_reset(vcpu);
+	}
 
 	if ((cr0 ^ old_cr0) & update_bits)
 		kvm_mmu_reset_context(vcpu);
@@ -1017,7 +1020,7 @@ void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
 	unsigned long flags;
 	s64 sdiff;
 
-	spin_lock_irqsave(&kvm->arch.tsc_write_lock, flags);
+	raw_spin_lock_irqsave(&kvm->arch.tsc_write_lock, flags);
 	offset = data - native_read_tsc();
 	ns = get_kernel_ns();
 	elapsed = ns - kvm->arch.last_tsc_nsec;
@@ -1050,7 +1053,7 @@ void kvm_write_tsc(struct kvm_vcpu *vcpu, u64 data)
 	kvm->arch.last_tsc_write = data;
 	kvm->arch.last_tsc_offset = offset;
 	kvm_x86_ops->write_tsc_offset(vcpu, offset);
-	spin_unlock_irqrestore(&kvm->arch.tsc_write_lock, flags);
+	raw_spin_unlock_irqrestore(&kvm->arch.tsc_write_lock, flags);
 
 	/* Reset of TSC must disable overshoot protection below */
 	vcpu->arch.hv_clock.tsc_timestamp = 0;
@@ -1453,6 +1456,14 @@ static int kvm_pv_enable_async_pf(struct kvm_vcpu *vcpu, u64 data)
 	return 0;
 }
 
+static void kvmclock_reset(struct kvm_vcpu *vcpu)
+{
+	if (vcpu->arch.time_page) {
+		kvm_release_page_dirty(vcpu->arch.time_page);
+		vcpu->arch.time_page = NULL;
+	}
+}
+
 int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 {
 	switch (msr) {
@@ -1510,10 +1521,7 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 		break;
 	case MSR_KVM_SYSTEM_TIME_NEW:
 	case MSR_KVM_SYSTEM_TIME: {
-		if (vcpu->arch.time_page) {
-			kvm_release_page_dirty(vcpu->arch.time_page);
-			vcpu->arch.time_page = NULL;
-		}
+		kvmclock_reset(vcpu);
 
 		vcpu->arch.time = data;
 		kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
@@ -1592,6 +1600,12 @@ int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
 		} else
 			return set_msr_hyperv(vcpu, msr, data);
 		break;
+	case MSR_IA32_BBL_CR_CTL3:
+		/* Drop writes to this legacy MSR -- see rdmsr
+		 * counterpart for further detail.
+		 */
+		pr_unimpl(vcpu, "ignored wrmsr: 0x%x data %llx\n", msr, data);
+		break;
 	default:
 		if (msr && (msr == vcpu->kvm->arch.xen_hvm_config.msr))
 			return xen_hvm_config(vcpu, data);
@@ -1846,6 +1860,19 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata)
 		} else
 			return get_msr_hyperv(vcpu, msr, pdata);
 		break;
+	case MSR_IA32_BBL_CR_CTL3:
+		/* This legacy MSR exists but isn't fully documented in current
+		 * silicon.  It is however accessed by winxp in very narrow
+		 * scenarios where it sets bit #19, itself documented as
+		 * a "reserved" bit.  Best effort attempt to source coherent
+		 * read data here should the balance of the register be
+		 * interpreted by the guest:
+		 *
+		 * L2 cache control register 3: 64GB range, 256KB size,
+		 * enabled, latency 0x1, configured
+		 */
+		data = 0xbe702111;
+		break;
 	default:
 		if (!ignore_msrs) {
 			pr_unimpl(vcpu, "unhandled rdmsr: 0x%x\n", msr);
@@ -2100,8 +2127,8 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 		if (check_tsc_unstable()) {
 			kvm_x86_ops->adjust_tsc_offset(vcpu, -tsc_delta);
 			vcpu->arch.tsc_catchup = 1;
-			kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
 		}
+		kvm_make_request(KVM_REQ_CLOCK_UPDATE, vcpu);
 		if (vcpu->cpu != cpu)
 			kvm_migrate_timers(vcpu);
 		vcpu->cpu = cpu;
@@ -2575,9 +2602,6 @@ static int kvm_vcpu_ioctl_x86_set_mce(struct kvm_vcpu *vcpu,
 	if (mce->status & MCI_STATUS_UC) {
 		if ((vcpu->arch.mcg_status & MCG_STATUS_MCIP) ||
 		    !kvm_read_cr4_bits(vcpu, X86_CR4_MCE)) {
-			printk(KERN_DEBUG "kvm: set_mce: "
-			       "injects mce exception while "
-			       "previous one is in progress!\n");
 			kvm_make_request(KVM_REQ_TRIPLE_FAULT, vcpu);
 			return 0;
 		}
@@ -2648,8 +2672,6 @@ static int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu,
 	vcpu->arch.interrupt.pending = events->interrupt.injected;
 	vcpu->arch.interrupt.nr = events->interrupt.nr;
 	vcpu->arch.interrupt.soft = events->interrupt.soft;
-	if (vcpu->arch.interrupt.pending && irqchip_in_kernel(vcpu->kvm))
-		kvm_pic_clear_isr_ack(vcpu->kvm);
 	if (events->flags & KVM_VCPUEVENT_VALID_SHADOW)
 		kvm_x86_ops->set_interrupt_shadow(vcpu,
 						  events->interrupt.shadow);
@@ -4140,8 +4162,8 @@ static unsigned long emulator_get_cached_segment_base(int seg,
 	return get_segment_base(vcpu, seg);
 }
 
-static bool emulator_get_cached_descriptor(struct desc_struct *desc, int seg,
-					   struct kvm_vcpu *vcpu)
+static bool emulator_get_cached_descriptor(struct desc_struct *desc, u32 *base3,
+					   int seg, struct kvm_vcpu *vcpu)
 {
 	struct kvm_segment var;
 
@@ -4154,6 +4176,10 @@ static bool emulator_get_cached_descriptor(struct desc_struct *desc, int seg,
 		var.limit >>= 12;
 	set_desc_limit(desc, var.limit);
 	set_desc_base(desc, (unsigned long)var.base);
+#ifdef CONFIG_X86_64
+	if (base3)
+		*base3 = var.base >> 32;
+#endif
 	desc->type = var.type;
 	desc->s = var.s;
 	desc->dpl = var.dpl;
@@ -4166,8 +4192,8 @@ static bool emulator_get_cached_descriptor(struct desc_struct *desc, int seg,
 	return true;
 }
 
-static void emulator_set_cached_descriptor(struct desc_struct *desc, int seg,
-					   struct kvm_vcpu *vcpu)
+static void emulator_set_cached_descriptor(struct desc_struct *desc, u32 base3,
+					   int seg, struct kvm_vcpu *vcpu)
 {
 	struct kvm_segment var;
 
@@ -4175,6 +4201,9 @@ static void emulator_set_cached_descriptor(struct desc_struct *desc, int seg,
 	kvm_get_segment(vcpu, &var, seg);
 
 	var.base = get_desc_base(desc);
+#ifdef CONFIG_X86_64
+	var.base |= ((u64)base3) << 32;
+#endif
 	var.limit = get_desc_limit(desc);
 	if (desc->g)
 		var.limit = (var.limit << 12) | 0xfff;
@@ -4390,41 +4419,16 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu,
 		vcpu->arch.emulate_ctxt.have_exception = false;
 		vcpu->arch.emulate_ctxt.perm_ok = false;
 
+		vcpu->arch.emulate_ctxt.only_vendor_specific_insn
+			= emulation_type & EMULTYPE_TRAP_UD;
+
 		r = x86_decode_insn(&vcpu->arch.emulate_ctxt, insn, insn_len);
-		if (r == X86EMUL_PROPAGATE_FAULT)
-			goto done;
 
 		trace_kvm_emulate_insn_start(vcpu);
-
-		/* Only allow emulation of specific instructions on #UD
-		 * (namely VMMCALL, sysenter, sysexit, syscall)*/
-		if (emulation_type & EMULTYPE_TRAP_UD) {
-			if (!c->twobyte)
-				return EMULATE_FAIL;
-			switch (c->b) {
-			case 0x01: /* VMMCALL */
-				if (c->modrm_mod != 3 || c->modrm_rm != 1)
-					return EMULATE_FAIL;
-				break;
-			case 0x34: /* sysenter */
-			case 0x35: /* sysexit */
-				if (c->modrm_mod != 0 || c->modrm_rm != 0)
-					return EMULATE_FAIL;
-				break;
-			case 0x05: /* syscall */
-				if (c->modrm_mod != 0 || c->modrm_rm != 0)
-					return EMULATE_FAIL;
-				break;
-			default:
-				return EMULATE_FAIL;
-			}
-
-			if (!(c->modrm_reg == 0 || c->modrm_reg == 3))
-				return EMULATE_FAIL;
-		}
-
 		++vcpu->stat.insn_emulation;
 		if (r)  {
+			if (emulation_type & EMULTYPE_TRAP_UD)
+				return EMULATE_FAIL;
 			if (reexecute_instruction(vcpu, cr2))
 				return EMULATE_DONE;
 			if (emulation_type & EMULTYPE_SKIP)
@@ -4452,7 +4456,6 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu,
 		return handle_emulation_failure(vcpu);
 	}
 
-done:
 	if (vcpu->arch.emulate_ctxt.have_exception) {
 		inject_emulated_exception(vcpu);
 		r = EMULATE_DONE;
@@ -4562,7 +4565,7 @@ static int kvmclock_cpufreq_notifier(struct notifier_block *nb, unsigned long va
 
 	smp_call_function_single(freq->cpu, tsc_khz_changed, freq, 1);
 
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	list_for_each_entry(kvm, &vm_list, vm_list) {
 		kvm_for_each_vcpu(i, vcpu, kvm) {
 			if (vcpu->cpu != freq->cpu)
@@ -4572,7 +4575,7 @@ static int kvmclock_cpufreq_notifier(struct notifier_block *nb, unsigned long va
 				send_ipi = 1;
 		}
 	}
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 
 	if (freq->old < freq->new && send_ipi) {
 		/*
@@ -5185,6 +5188,8 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 			r = 1;
 			goto out;
 		}
+		if (kvm_check_request(KVM_REQ_NMI, vcpu))
+			vcpu->arch.nmi_pending = true;
 	}
 
 	r = kvm_mmu_reload(vcpu);
@@ -5213,14 +5218,18 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 		kvm_load_guest_fpu(vcpu);
 	kvm_load_guest_xcr0(vcpu);
 
-	atomic_set(&vcpu->guest_mode, 1);
-	smp_wmb();
+	vcpu->mode = IN_GUEST_MODE;
+
+	/* We should set ->mode before check ->requests,
+	 * see the comment in make_all_cpus_request.
+	 */
+	smp_mb();
 
 	local_irq_disable();
 
-	if (!atomic_read(&vcpu->guest_mode) || vcpu->requests
+	if (vcpu->mode == EXITING_GUEST_MODE || vcpu->requests
 	    || need_resched() || signal_pending(current)) {
-		atomic_set(&vcpu->guest_mode, 0);
+		vcpu->mode = OUTSIDE_GUEST_MODE;
 		smp_wmb();
 		local_irq_enable();
 		preempt_enable();
@@ -5256,7 +5265,7 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 
 	kvm_get_msr(vcpu, MSR_IA32_TSC, &vcpu->arch.last_guest_tsc);
 
-	atomic_set(&vcpu->guest_mode, 0);
+	vcpu->mode = OUTSIDE_GUEST_MODE;
 	smp_wmb();
 	local_irq_enable();
 
@@ -5574,7 +5583,7 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
 	int mmu_reset_needed = 0;
-	int pending_vec, max_bits;
+	int pending_vec, max_bits, idx;
 	struct desc_ptr dt;
 
 	dt.size = sregs->idt.limit;
@@ -5603,10 +5612,13 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	kvm_x86_ops->set_cr4(vcpu, sregs->cr4);
 	if (sregs->cr4 & X86_CR4_OSXSAVE)
 		update_cpuid(vcpu);
+
+	idx = srcu_read_lock(&vcpu->kvm->srcu);
 	if (!is_long_mode(vcpu) && is_pae(vcpu)) {
 		load_pdptrs(vcpu, vcpu->arch.walk_mmu, kvm_read_cr3(vcpu));
 		mmu_reset_needed = 1;
 	}
+	srcu_read_unlock(&vcpu->kvm->srcu, idx);
 
 	if (mmu_reset_needed)
 		kvm_mmu_reset_context(vcpu);
@@ -5617,8 +5629,6 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	if (pending_vec < max_bits) {
 		kvm_queue_interrupt(vcpu, pending_vec, false);
 		pr_debug("Set back pending irq %d\n", pending_vec);
-		if (irqchip_in_kernel(vcpu->kvm))
-			kvm_pic_clear_isr_ack(vcpu->kvm);
 	}
 
 	kvm_set_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
@@ -5814,10 +5824,7 @@ void kvm_put_guest_fpu(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu)
 {
-	if (vcpu->arch.time_page) {
-		kvm_release_page_dirty(vcpu->arch.time_page);
-		vcpu->arch.time_page = NULL;
-	}
+	kvmclock_reset(vcpu);
 
 	free_cpumask_var(vcpu->arch.wbinvd_dirty_mask);
 	fx_free(vcpu);
@@ -5878,6 +5885,8 @@ int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu)
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 	vcpu->arch.apf.msr_val = 0;
 
+	kvmclock_reset(vcpu);
+
 	kvm_clear_async_pf_completion_queue(vcpu);
 	kvm_async_pf_hash_reset(vcpu);
 	vcpu->arch.apf.halted = false;
@@ -6005,7 +6014,7 @@ int kvm_arch_init_vm(struct kvm *kvm)
 	/* Reserve bit 0 of irq_sources_bitmap for userspace irq source */
 	set_bit(KVM_USERSPACE_IRQ_SOURCE_ID, &kvm->arch.irq_sources_bitmap);
 
-	spin_lock_init(&kvm->arch.tsc_write_lock);
+	raw_spin_lock_init(&kvm->arch.tsc_write_lock);
 
 	return 0;
 }
@@ -6103,7 +6112,7 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
 				int user_alloc)
 {
 
-	int npages = mem->memory_size >> PAGE_SHIFT;
+	int nr_mmu_pages = 0, npages = mem->memory_size >> PAGE_SHIFT;
 
 	if (!user_alloc && !old.user_alloc && old.rmap && !npages) {
 		int ret;
@@ -6118,12 +6127,12 @@ void kvm_arch_commit_memory_region(struct kvm *kvm,
 			       "failed to munmap memory\n");
 	}
 
+	if (!kvm->arch.n_requested_mmu_pages)
+		nr_mmu_pages = kvm_mmu_calculate_mmu_pages(kvm);
+
 	spin_lock(&kvm->mmu_lock);
-	if (!kvm->arch.n_requested_mmu_pages) {
-		unsigned int nr_mmu_pages = kvm_mmu_calculate_mmu_pages(kvm);
+	if (nr_mmu_pages)
 		kvm_mmu_change_mmu_pages(kvm, nr_mmu_pages);
-	}
-
 	kvm_mmu_slot_remove_write_access(kvm, mem->slot);
 	spin_unlock(&kvm->mmu_lock);
 }
@@ -6157,7 +6166,7 @@ void kvm_vcpu_kick(struct kvm_vcpu *vcpu)
 
 	me = get_cpu();
 	if (cpu != me && (unsigned)cpu < nr_cpu_ids && cpu_online(cpu))
-		if (atomic_xchg(&vcpu->guest_mode, 0))
+		if (kvm_vcpu_exiting_guest_mode(vcpu) == IN_GUEST_MODE)
 			smp_send_reschedule(cpu);
 	put_cpu();
 }

drivers/infiniband/hw/ipath/ipath_user_pages.c

@@ -53,8 +53,8 @@ static void __ipath_release_user_pages(struct page **p, size_t num_pages,
 }
 
 /* call with current->mm->mmap_sem held */
-static int __get_user_pages(unsigned long start_page, size_t num_pages,
-			struct page **p, struct vm_area_struct **vma)
+static int __ipath_get_user_pages(unsigned long start_page, size_t num_pages,
+				  struct page **p, struct vm_area_struct **vma)
 {
 	unsigned long lock_limit;
 	size_t got;
@@ -165,7 +165,7 @@ int ipath_get_user_pages(unsigned long start_page, size_t num_pages,
 
 	down_write(&current->mm->mmap_sem);
 
-	ret = __get_user_pages(start_page, num_pages, p, NULL);
+	ret = __ipath_get_user_pages(start_page, num_pages, p, NULL);
 
 	up_write(&current->mm->mmap_sem);
 

drivers/infiniband/hw/qib/qib_user_pages.c

@@ -51,8 +51,8 @@ static void __qib_release_user_pages(struct page **p, size_t num_pages,
 /*
  * Call with current->mm->mmap_sem held.
  */
-static int __get_user_pages(unsigned long start_page, size_t num_pages,
-			    struct page **p, struct vm_area_struct **vma)
+static int __qib_get_user_pages(unsigned long start_page, size_t num_pages,
+				struct page **p, struct vm_area_struct **vma)
 {
 	unsigned long lock_limit;
 	size_t got;
@@ -136,7 +136,7 @@ int qib_get_user_pages(unsigned long start_page, size_t num_pages,
 
 	down_write(&current->mm->mmap_sem);
 
-	ret = __get_user_pages(start_page, num_pages, p, NULL);
+	ret = __qib_get_user_pages(start_page, num_pages, p, NULL);
 
 	up_write(&current->mm->mmap_sem);
 

include/asm-generic/errno.h

@@ -108,4 +108,6 @@
 
 #define ERFKILL		132	/* Operation not possible due to RF-kill */
 
+#define EHWPOISON	133	/* Memory page has hardware error */
+
 #endif

include/linux/kvm_host.h

@@ -43,6 +43,7 @@
 #define KVM_REQ_DEACTIVATE_FPU    10
 #define KVM_REQ_EVENT             11
 #define KVM_REQ_APF_HALT          12
+#define KVM_REQ_NMI               13
 
 #define KVM_USERSPACE_IRQ_SOURCE_ID	0
 
@@ -98,23 +99,31 @@ int kvm_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn,
 int kvm_async_pf_wakeup_all(struct kvm_vcpu *vcpu);
 #endif
 
+enum {
+	OUTSIDE_GUEST_MODE,
+	IN_GUEST_MODE,
+	EXITING_GUEST_MODE
+};
+
 struct kvm_vcpu {
 	struct kvm *kvm;
 #ifdef CONFIG_PREEMPT_NOTIFIERS
 	struct preempt_notifier preempt_notifier;
 #endif
+	int cpu;
 	int vcpu_id;
-	struct mutex mutex;
-	int   cpu;
-	atomic_t guest_mode;
-	struct kvm_run *run;
+	int srcu_idx;
+	int mode;
 	unsigned long requests;
 	unsigned long guest_debug;
-	int srcu_idx;
+
+	struct mutex mutex;
+	struct kvm_run *run;
 
 	int fpu_active;
 	int guest_fpu_loaded, guest_xcr0_loaded;
 	wait_queue_head_t wq;
+	struct pid *pid;
 	int sigset_active;
 	sigset_t sigset;
 	struct kvm_vcpu_stat stat;
@@ -140,6 +149,11 @@ struct kvm_vcpu {
 	struct kvm_vcpu_arch arch;
 };
 
+static inline int kvm_vcpu_exiting_guest_mode(struct kvm_vcpu *vcpu)
+{
+	return cmpxchg(&vcpu->mode, IN_GUEST_MODE, EXITING_GUEST_MODE);
+}
+
 /*
  * Some of the bitops functions do not support too long bitmaps.
  * This number must be determined not to exceed such limits.
@@ -212,7 +226,6 @@ struct kvm_memslots {
 
 struct kvm {
 	spinlock_t mmu_lock;
-	raw_spinlock_t requests_lock;
 	struct mutex slots_lock;
 	struct mm_struct *mm; /* userspace tied to this vm */
 	struct kvm_memslots *memslots;
@@ -223,6 +236,7 @@ struct kvm {
 #endif
 	struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
 	atomic_t online_vcpus;
+	int last_boosted_vcpu;
 	struct list_head vm_list;
 	struct mutex lock;
 	struct kvm_io_bus *buses[KVM_NR_BUSES];
@@ -719,11 +733,6 @@ static inline void kvm_make_request(int req, struct kvm_vcpu *vcpu)
 	set_bit(req, &vcpu->requests);
 }
 
-static inline bool kvm_make_check_request(int req, struct kvm_vcpu *vcpu)
-{
-	return test_and_set_bit(req, &vcpu->requests);
-}
-
 static inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu)
 {
 	if (test_bit(req, &vcpu->requests)) {

include/linux/mm.h

@@ -972,6 +972,10 @@ static inline int handle_mm_fault(struct mm_struct *mm,
 extern int make_pages_present(unsigned long addr, unsigned long end);
 extern int access_process_vm(struct task_struct *tsk, unsigned long addr, void *buf, int len, int write);
 
+int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
+		     unsigned long start, int len, unsigned int foll_flags,
+		     struct page **pages, struct vm_area_struct **vmas,
+		     int *nonblocking);
 int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 			unsigned long start, int nr_pages, int write, int force,
 			struct page **pages, struct vm_area_struct **vmas);
@@ -1535,6 +1539,7 @@ struct page *follow_page(struct vm_area_struct *, unsigned long address,
 #define FOLL_FORCE	0x10	/* get_user_pages read/write w/o permission */
 #define FOLL_MLOCK	0x40	/* mark page as mlocked */
 #define FOLL_SPLIT	0x80	/* don't return transhuge pages, split them */
+#define FOLL_HWPOISON	0x100	/* check page is hwpoisoned */
 
 typedef int (*pte_fn_t)(pte_t *pte, pgtable_t token, unsigned long addr,
 			void *data);
@@ -1627,14 +1632,6 @@ extern int sysctl_memory_failure_recovery;
 extern void shake_page(struct page *p, int access);
 extern atomic_long_t mce_bad_pages;
 extern int soft_offline_page(struct page *page, int flags);
-#ifdef CONFIG_MEMORY_FAILURE
-int is_hwpoison_address(unsigned long addr);
-#else
-static inline int is_hwpoison_address(unsigned long addr)
-{
-	return 0;
-}
-#endif
 
 extern void dump_page(struct page *page);
 

kernel/fork.c

@@ -193,6 +193,7 @@ void __put_task_struct(struct task_struct *tsk)
 	if (!profile_handoff_task(tsk))
 		free_task(tsk);
 }
+EXPORT_SYMBOL_GPL(__put_task_struct);
 
 /*
  * macro override instead of weak attribute alias, to workaround

kernel/pid.c

@@ -435,6 +435,7 @@ struct pid *get_task_pid(struct task_struct *task, enum pid_type type)
 	rcu_read_unlock();
 	return pid;
 }
+EXPORT_SYMBOL_GPL(get_task_pid);
 
 struct task_struct *get_pid_task(struct pid *pid, enum pid_type type)
 {
@@ -446,6 +447,7 @@ struct task_struct *get_pid_task(struct pid *pid, enum pid_type type)
 	rcu_read_unlock();
 	return result;
 }
+EXPORT_SYMBOL_GPL(get_pid_task);
 
 struct pid *find_get_pid(pid_t nr)
 {

mm/internal.h

@@ -245,11 +245,6 @@ static inline void mminit_validate_memmodel_limits(unsigned long *start_pfn,
 }
 #endif /* CONFIG_SPARSEMEM */
 
-int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
-		     unsigned long start, int len, unsigned int foll_flags,
-		     struct page **pages, struct vm_area_struct **vmas,
-		     int *nonblocking);
-
 #define ZONE_RECLAIM_NOSCAN	-2
 #define ZONE_RECLAIM_FULL	-1
 #define ZONE_RECLAIM_SOME	0

mm/memory-failure.c

@@ -1487,35 +1487,3 @@ int soft_offline_page(struct page *page, int flags)
 	/* keep elevated page count for bad page */
 	return ret;
 }
-
-/*
- * The caller must hold current->mm->mmap_sem in read mode.
- */
-int is_hwpoison_address(unsigned long addr)
-{
-	pgd_t *pgdp;
-	pud_t pud, *pudp;
-	pmd_t pmd, *pmdp;
-	pte_t pte, *ptep;
-	swp_entry_t entry;
-
-	pgdp = pgd_offset(current->mm, addr);
-	if (!pgd_present(*pgdp))
-		return 0;
-	pudp = pud_offset(pgdp, addr);
-	pud = *pudp;
-	if (!pud_present(pud) || pud_large(pud))
-		return 0;
-	pmdp = pmd_offset(pudp, addr);
-	pmd = *pmdp;
-	if (!pmd_present(pmd) || pmd_large(pmd))
-		return 0;
-	ptep = pte_offset_map(pmdp, addr);
-	pte = *ptep;
-	pte_unmap(ptep);
-	if (!is_swap_pte(pte))
-		return 0;
-	entry = pte_to_swp_entry(pte);
-	return is_hwpoison_entry(entry);
-}
-EXPORT_SYMBOL_GPL(is_hwpoison_address);

mm/memory.c

@@ -1410,6 +1410,55 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
 	return page;
 }
 
+/**
+ * __get_user_pages() - pin user pages in memory
+ * @tsk:	task_struct of target task
+ * @mm:		mm_struct of target mm
+ * @start:	starting user address
+ * @nr_pages:	number of pages from start to pin
+ * @gup_flags:	flags modifying pin behaviour
+ * @pages:	array that receives pointers to the pages pinned.
+ *		Should be at least nr_pages long. Or NULL, if caller
+ *		only intends to ensure the pages are faulted in.
+ * @vmas:	array of pointers to vmas corresponding to each page.
+ *		Or NULL if the caller does not require them.
+ * @nonblocking: whether waiting for disk IO or mmap_sem contention
+ *
+ * Returns number of pages pinned. This may be fewer than the number
+ * requested. If nr_pages is 0 or negative, returns 0. If no pages
+ * were pinned, returns -errno. Each page returned must be released
+ * with a put_page() call when it is finished with. vmas will only
+ * remain valid while mmap_sem is held.
+ *
+ * Must be called with mmap_sem held for read or write.
+ *
+ * __get_user_pages walks a process's page tables and takes a reference to
+ * each struct page that each user address corresponds to at a given
+ * instant. That is, it takes the page that would be accessed if a user
+ * thread accesses the given user virtual address at that instant.
+ *
+ * This does not guarantee that the page exists in the user mappings when
+ * __get_user_pages returns, and there may even be a completely different
+ * page there in some cases (eg. if mmapped pagecache has been invalidated
+ * and subsequently re faulted). However it does guarantee that the page
+ * won't be freed completely. And mostly callers simply care that the page
+ * contains data that was valid *at some point in time*. Typically, an IO
+ * or similar operation cannot guarantee anything stronger anyway because
+ * locks can't be held over the syscall boundary.
+ *
+ * If @gup_flags & FOLL_WRITE == 0, the page must not be written to. If
+ * the page is written to, set_page_dirty (or set_page_dirty_lock, as
+ * appropriate) must be called after the page is finished with, and
+ * before put_page is called.
+ *
+ * If @nonblocking != NULL, __get_user_pages will not wait for disk IO
+ * or mmap_sem contention, and if waiting is needed to pin all pages,
+ * *@nonblocking will be set to 0.
+ *
+ * In most cases, get_user_pages or get_user_pages_fast should be used
+ * instead of __get_user_pages. __get_user_pages should be used only if
+ * you need some special @gup_flags.
+ */
 int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 		     unsigned long start, int nr_pages, unsigned int gup_flags,
 		     struct page **pages, struct vm_area_struct **vmas,
@@ -1527,9 +1576,16 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 				if (ret & VM_FAULT_ERROR) {
 					if (ret & VM_FAULT_OOM)
 						return i ? i : -ENOMEM;
-					if (ret &
-					    (VM_FAULT_HWPOISON|VM_FAULT_HWPOISON_LARGE|
-					     VM_FAULT_SIGBUS))
+					if (ret & (VM_FAULT_HWPOISON |
+						   VM_FAULT_HWPOISON_LARGE)) {
+						if (i)
+							return i;
+						else if (gup_flags & FOLL_HWPOISON)
+							return -EHWPOISON;
+						else
+							return -EFAULT;
+					}
+					if (ret & VM_FAULT_SIGBUS)
 						return i ? i : -EFAULT;
 					BUG();
 				}
@@ -1578,6 +1634,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 	} while (nr_pages);
 	return i;
 }
+EXPORT_SYMBOL(__get_user_pages);
 
 /**
  * get_user_pages() - pin user pages in memory

virt/kvm/eventfd.c

@@ -313,8 +313,9 @@ kvm_irqfd_deassign(struct kvm *kvm, int fd, int gsi)
 		if (irqfd->eventfd == eventfd && irqfd->gsi == gsi) {
 			/*
 			 * This rcu_assign_pointer is needed for when
-			 * another thread calls kvm_irqfd_update before
-			 * we flush workqueue below.
+			 * another thread calls kvm_irq_routing_update before
+			 * we flush workqueue below (we synchronize with
+			 * kvm_irq_routing_update using irqfds.lock).
 			 * It is paired with synchronize_rcu done by caller
 			 * of that function.
 			 */

virt/kvm/kvm_main.c

@@ -69,7 +69,7 @@ MODULE_LICENSE("GPL");
  * 		kvm->lock --> kvm->slots_lock --> kvm->irq_lock
  */
 
-DEFINE_SPINLOCK(kvm_lock);
+DEFINE_RAW_SPINLOCK(kvm_lock);
 LIST_HEAD(vm_list);
 
 static cpumask_var_t cpus_hardware_enabled;
@@ -137,6 +137,14 @@ void vcpu_load(struct kvm_vcpu *vcpu)
 	int cpu;
 
 	mutex_lock(&vcpu->mutex);
+	if (unlikely(vcpu->pid != current->pids[PIDTYPE_PID].pid)) {
+		/* The thread running this VCPU changed. */
+		struct pid *oldpid = vcpu->pid;
+		struct pid *newpid = get_task_pid(current, PIDTYPE_PID);
+		rcu_assign_pointer(vcpu->pid, newpid);
+		synchronize_rcu();
+		put_pid(oldpid);
+	}
 	cpu = get_cpu();
 	preempt_notifier_register(&vcpu->preempt_notifier);
 	kvm_arch_vcpu_load(vcpu, cpu);
@@ -165,13 +173,16 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req)
 
 	zalloc_cpumask_var(&cpus, GFP_ATOMIC);
 
-	raw_spin_lock(&kvm->requests_lock);
-	me = smp_processor_id();
+	me = get_cpu();
 	kvm_for_each_vcpu(i, vcpu, kvm) {
-		if (kvm_make_check_request(req, vcpu))
-			continue;
+		kvm_make_request(req, vcpu);
 		cpu = vcpu->cpu;
-		if (cpus != NULL && cpu != -1 && cpu != me)
+
+		/* Set ->requests bit before we read ->mode */
+		smp_mb();
+
+		if (cpus != NULL && cpu != -1 && cpu != me &&
+		      kvm_vcpu_exiting_guest_mode(vcpu) != OUTSIDE_GUEST_MODE)
 			cpumask_set_cpu(cpu, cpus);
 	}
 	if (unlikely(cpus == NULL))
@@ -180,7 +191,7 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req)
 		smp_call_function_many(cpus, ack_flush, NULL, 1);
 	else
 		called = false;
-	raw_spin_unlock(&kvm->requests_lock);
+	put_cpu();
 	free_cpumask_var(cpus);
 	return called;
 }
@@ -209,6 +220,7 @@ int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id)
 	vcpu->cpu = -1;
 	vcpu->kvm = kvm;
 	vcpu->vcpu_id = id;
+	vcpu->pid = NULL;
 	init_waitqueue_head(&vcpu->wq);
 	kvm_async_pf_vcpu_init(vcpu);
 
@@ -233,6 +245,7 @@ EXPORT_SYMBOL_GPL(kvm_vcpu_init);
 
 void kvm_vcpu_uninit(struct kvm_vcpu *vcpu)
 {
+	put_pid(vcpu->pid);
 	kvm_arch_vcpu_uninit(vcpu);
 	free_page((unsigned long)vcpu->run);
 }
@@ -463,15 +476,14 @@ static struct kvm *kvm_create_vm(void)
 	kvm->mm = current->mm;
 	atomic_inc(&kvm->mm->mm_count);
 	spin_lock_init(&kvm->mmu_lock);
-	raw_spin_lock_init(&kvm->requests_lock);
 	kvm_eventfd_init(kvm);
 	mutex_init(&kvm->lock);
 	mutex_init(&kvm->irq_lock);
 	mutex_init(&kvm->slots_lock);
 	atomic_set(&kvm->users_count, 1);
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	list_add(&kvm->vm_list, &vm_list);
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 
 	return kvm;
 
@@ -544,9 +556,9 @@ static void kvm_destroy_vm(struct kvm *kvm)
 	struct mm_struct *mm = kvm->mm;
 
 	kvm_arch_sync_events(kvm);
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	list_del(&kvm->vm_list);
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 	kvm_free_irq_routing(kvm);
 	for (i = 0; i < KVM_NR_BUSES; i++)
 		kvm_io_bus_destroy(kvm->buses[i]);
@@ -588,6 +600,7 @@ static int kvm_vm_release(struct inode *inode, struct file *filp)
 	return 0;
 }
 
+#ifndef CONFIG_S390
 /*
  * Allocation size is twice as large as the actual dirty bitmap size.
  * This makes it possible to do double buffering: see x86's
@@ -608,6 +621,7 @@ static int kvm_create_dirty_bitmap(struct kvm_memory_slot *memslot)
 	memslot->dirty_bitmap_head = memslot->dirty_bitmap;
 	return 0;
 }
+#endif /* !CONFIG_S390 */
 
 /*
  * Allocate some memory and give it an address in the guest physical address
@@ -621,7 +635,7 @@ int __kvm_set_memory_region(struct kvm *kvm,
 			    struct kvm_userspace_memory_region *mem,
 			    int user_alloc)
 {
-	int r, flush_shadow = 0;
+	int r;
 	gfn_t base_gfn;
 	unsigned long npages;
 	unsigned long i;
@@ -741,8 +755,6 @@ int __kvm_set_memory_region(struct kvm *kvm,
 		if (kvm_create_dirty_bitmap(&new) < 0)
 			goto out_free;
 		/* destroy any largepage mappings for dirty tracking */
-		if (old.npages)
-			flush_shadow = 1;
 	}
 #else  /* not defined CONFIG_S390 */
 	new.user_alloc = user_alloc;
@@ -813,9 +825,6 @@ int __kvm_set_memory_region(struct kvm *kvm,
 	kvm_free_physmem_slot(&old, &new);
 	kfree(old_memslots);
 
-	if (flush_shadow)
-		kvm_arch_flush_shadow(kvm);
-
 	return 0;
 
 out_free:
@@ -1029,6 +1038,15 @@ static pfn_t get_fault_pfn(void)
 	return fault_pfn;
 }
 
+static inline int check_user_page_hwpoison(unsigned long addr)
+{
+	int rc, flags = FOLL_TOUCH | FOLL_HWPOISON | FOLL_WRITE;
+
+	rc = __get_user_pages(current, current->mm, addr, 1,
+			      flags, NULL, NULL, NULL);
+	return rc == -EHWPOISON;
+}
+
 static pfn_t hva_to_pfn(struct kvm *kvm, unsigned long addr, bool atomic,
 			bool *async, bool write_fault, bool *writable)
 {
@@ -1076,7 +1094,7 @@ static pfn_t hva_to_pfn(struct kvm *kvm, unsigned long addr, bool atomic,
 			return get_fault_pfn();
 
 		down_read(&current->mm->mmap_sem);
-		if (is_hwpoison_address(addr)) {
+		if (check_user_page_hwpoison(addr)) {
 			up_read(&current->mm->mmap_sem);
 			get_page(hwpoison_page);
 			return page_to_pfn(hwpoison_page);
@@ -1466,18 +1484,55 @@ void kvm_resched(struct kvm_vcpu *vcpu)
 }
 EXPORT_SYMBOL_GPL(kvm_resched);
 
-void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu)
+void kvm_vcpu_on_spin(struct kvm_vcpu *me)
 {
-	ktime_t expires;
-	DEFINE_WAIT(wait);
-
-	prepare_to_wait(&vcpu->wq, &wait, TASK_INTERRUPTIBLE);
-
-	/* Sleep for 100 us, and hope lock-holder got scheduled */
-	expires = ktime_add_ns(ktime_get(), 100000UL);
-	schedule_hrtimeout(&expires, HRTIMER_MODE_ABS);
+	struct kvm *kvm = me->kvm;
+	struct kvm_vcpu *vcpu;
+	int last_boosted_vcpu = me->kvm->last_boosted_vcpu;
+	int yielded = 0;
+	int pass;
+	int i;
 
-	finish_wait(&vcpu->wq, &wait);
+	/*
+	 * We boost the priority of a VCPU that is runnable but not
+	 * currently running, because it got preempted by something
+	 * else and called schedule in __vcpu_run.  Hopefully that
+	 * VCPU is holding the lock that we need and will release it.
+	 * We approximate round-robin by starting at the last boosted VCPU.
+	 */
+	for (pass = 0; pass < 2 && !yielded; pass++) {
+		kvm_for_each_vcpu(i, vcpu, kvm) {
+			struct task_struct *task = NULL;
+			struct pid *pid;
+			if (!pass && i < last_boosted_vcpu) {
+				i = last_boosted_vcpu;
+				continue;
+			} else if (pass && i > last_boosted_vcpu)
+				break;
+			if (vcpu == me)
+				continue;
+			if (waitqueue_active(&vcpu->wq))
+				continue;
+			rcu_read_lock();
+			pid = rcu_dereference(vcpu->pid);
+			if (pid)
+				task = get_pid_task(vcpu->pid, PIDTYPE_PID);
+			rcu_read_unlock();
+			if (!task)
+				continue;
+			if (task->flags & PF_VCPU) {
+				put_task_struct(task);
+				continue;
+			}
+			if (yield_to(task, 1)) {
+				put_task_struct(task);
+				kvm->last_boosted_vcpu = i;
+				yielded = 1;
+				break;
+			}
+			put_task_struct(task);
+		}
+	}
 }
 EXPORT_SYMBOL_GPL(kvm_vcpu_on_spin);
 
@@ -2122,9 +2177,9 @@ static void hardware_enable_nolock(void *junk)
 
 static void hardware_enable(void *junk)
 {
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	hardware_enable_nolock(junk);
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 }
 
 static void hardware_disable_nolock(void *junk)
@@ -2139,9 +2194,9 @@ static void hardware_disable_nolock(void *junk)
 
 static void hardware_disable(void *junk)
 {
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	hardware_disable_nolock(junk);
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 }
 
 static void hardware_disable_all_nolock(void)
@@ -2155,16 +2210,16 @@ static void hardware_disable_all_nolock(void)
 
 static void hardware_disable_all(void)
 {
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	hardware_disable_all_nolock();
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 }
 
 static int hardware_enable_all(void)
 {
 	int r = 0;
 
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 
 	kvm_usage_count++;
 	if (kvm_usage_count == 1) {
@@ -2177,7 +2232,7 @@ static int hardware_enable_all(void)
 		}
 	}
 
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 
 	return r;
 }
@@ -2339,10 +2394,10 @@ static int vm_stat_get(void *_offset, u64 *val)
 	struct kvm *kvm;
 
 	*val = 0;
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	list_for_each_entry(kvm, &vm_list, vm_list)
 		*val += *(u32 *)((void *)kvm + offset);
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 	return 0;
 }
 
@@ -2356,12 +2411,12 @@ static int vcpu_stat_get(void *_offset, u64 *val)
 	int i;
 
 	*val = 0;
-	spin_lock(&kvm_lock);
+	raw_spin_lock(&kvm_lock);
 	list_for_each_entry(kvm, &vm_list, vm_list)
 		kvm_for_each_vcpu(i, vcpu, kvm)
 			*val += *(u32 *)((void *)vcpu + offset);
 
-	spin_unlock(&kvm_lock);
+	raw_spin_unlock(&kvm_lock);
 	return 0;
 }
 
@@ -2402,7 +2457,7 @@ static int kvm_suspend(struct sys_device *dev, pm_message_t state)
 static int kvm_resume(struct sys_device *dev)
 {
 	if (kvm_usage_count) {
-		WARN_ON(spin_is_locked(&kvm_lock));
+		WARN_ON(raw_spin_is_locked(&kvm_lock));
 		hardware_enable_nolock(NULL);
 	}
 	return 0;