Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Pull more KVM updates from Paolo Bonzini:
 "Second round of KVM changes for 4.6:

   - build fixes for PPC KVM
   - miscellaneous bugfixes for ARM KVM
   - cleanup of memory barrier and removal of redundant barriers
   - x86 fixes: page tracking oops, support for old buggy KVM nested on 4.5
   - support for protection keys in guests
   - lockdep fix
   - another conversion to simple wait queues and raw spinlocks,
     backported from PREEMPT_RT"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (27 commits)
  KVM: page_track: fix access to NULL slot
  KVM: PPC: do not compile in vfio.o unconditionally
  kvm, rt: change async pagefault code locking for PREEMPT_RT
  KVM/PPC: update the comment of memory barrier in the kvmppc_prepare_to_enter()
  KVM/x86: update the comment of memory barrier in the vcpu_enter_guest()
  KVM: Replace smp_mb() with smp_load_acquire() in the kvm_flush_remote_tlbs()
  KVM/x86: Call smp_wmb() before increasing tlbs_dirty
  KVM: Replace smp_mb() with smp_mb_after_atomic() in the kvm_make_all_cpus_request()
  KVM/x86: Replace smp_mb() with smp_store_mb/release() in the walk_shadow_page_lockless_begin/end()
  KVM: Remove redundant smp_mb() in the kvm_mmu_commit_zap_page()
  KVM, pkeys: expose CPUID/CR4 to guest
  KVM, pkeys: add pkeys support for permission_fault
  KVM, pkeys: introduce pkru_mask to cache conditions
  KVM, pkeys: save/restore PKRU when guest/host switches
  x86: pkey: introduce write_pkru() for KVM
  KVM, pkeys: add pkeys support for xsave state
  KVM, pkeys: disable pkeys for guests in non-paging mode
  KVM: x86: remove magic number with enum cpuid_leafs
  KVM: MMU: return page fault error code from permission_fault
  KVM: fix spin_lock_init order on x86
  ...

arch/arm/kvm/arm.c

@@ -373,7 +373,9 @@ static void exit_vm_noop(void *info)
 
 void force_vm_exit(const cpumask_t *mask)
 {
+	preempt_disable();
 	smp_call_function_many(mask, exit_vm_noop, NULL, true);
+	preempt_enable();
 }
 
 /**

arch/arm64/include/asm/kvm_asm.h

@@ -26,7 +26,13 @@
 #define KVM_ARM64_DEBUG_DIRTY_SHIFT	0
 #define KVM_ARM64_DEBUG_DIRTY		(1 << KVM_ARM64_DEBUG_DIRTY_SHIFT)
 
-#define kvm_ksym_ref(sym)		phys_to_virt((u64)&sym - kimage_voffset)
+#define kvm_ksym_ref(sym)						\
+	({								\
+		void *val = &sym;					\
+		if (!is_kernel_in_hyp_mode())				\
+			val = phys_to_virt((u64)&sym - kimage_voffset);	\
+		val;							\
+	 })
 
 #ifndef __ASSEMBLY__
 struct kvm;

arch/arm64/kvm/hyp/Makefile

@@ -16,3 +16,7 @@ obj-$(CONFIG_KVM_ARM_HOST) += fpsimd.o
 obj-$(CONFIG_KVM_ARM_HOST) += tlb.o
 obj-$(CONFIG_KVM_ARM_HOST) += hyp-entry.o
 obj-$(CONFIG_KVM_ARM_HOST) += s2-setup.o
+
+GCOV_PROFILE	:= n
+KASAN_SANITIZE	:= n
+UBSAN_SANITIZE	:= n

arch/powerpc/kvm/Makefile

@@ -8,7 +8,8 @@ ccflags-y := -Ivirt/kvm -Iarch/powerpc/kvm
 KVM := ../../../virt/kvm
 
 common-objs-y = $(KVM)/kvm_main.o $(KVM)/coalesced_mmio.o \
-		$(KVM)/eventfd.o $(KVM)/vfio.o
+		$(KVM)/eventfd.o
+common-objs-$(CONFIG_KVM_VFIO) += $(KVM)/vfio.o
 
 CFLAGS_e500_mmu.o := -I.
 CFLAGS_e500_mmu_host.o := -I.

arch/powerpc/kvm/book3s_64_vio.c

@@ -209,6 +209,32 @@ long kvm_vm_ioctl_create_spapr_tce(struct kvm *kvm,
 	return ret;
 }
 
+long kvmppc_h_put_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
+		      unsigned long ioba, unsigned long tce)
+{
+	struct kvmppc_spapr_tce_table *stt = kvmppc_find_table(vcpu, liobn);
+	long ret;
+
+	/* udbg_printf("H_PUT_TCE(): liobn=0x%lx ioba=0x%lx, tce=0x%lx\n", */
+	/* 	    liobn, ioba, tce); */
+
+	if (!stt)
+		return H_TOO_HARD;
+
+	ret = kvmppc_ioba_validate(stt, ioba, 1);
+	if (ret != H_SUCCESS)
+		return ret;
+
+	ret = kvmppc_tce_validate(stt, tce);
+	if (ret != H_SUCCESS)
+		return ret;
+
+	kvmppc_tce_put(stt, ioba >> stt->page_shift, tce);
+
+	return H_SUCCESS;
+}
+EXPORT_SYMBOL_GPL(kvmppc_h_put_tce);
+
 long kvmppc_h_put_tce_indirect(struct kvm_vcpu *vcpu,
 		unsigned long liobn, unsigned long ioba,
 		unsigned long tce_list, unsigned long npages)
@@ -264,3 +290,29 @@ long kvmppc_h_put_tce_indirect(struct kvm_vcpu *vcpu,
 	return ret;
 }
 EXPORT_SYMBOL_GPL(kvmppc_h_put_tce_indirect);
+
+long kvmppc_h_stuff_tce(struct kvm_vcpu *vcpu,
+		unsigned long liobn, unsigned long ioba,
+		unsigned long tce_value, unsigned long npages)
+{
+	struct kvmppc_spapr_tce_table *stt;
+	long i, ret;
+
+	stt = kvmppc_find_table(vcpu, liobn);
+	if (!stt)
+		return H_TOO_HARD;
+
+	ret = kvmppc_ioba_validate(stt, ioba, npages);
+	if (ret != H_SUCCESS)
+		return ret;
+
+	/* Check permission bits only to allow userspace poison TCE for debug */
+	if (tce_value & (TCE_PCI_WRITE | TCE_PCI_READ))
+		return H_PARAMETER;
+
+	for (i = 0; i < npages; ++i, ioba += (1ULL << stt->page_shift))
+		kvmppc_tce_put(stt, ioba >> stt->page_shift, tce_value);
+
+	return H_SUCCESS;
+}
+EXPORT_SYMBOL_GPL(kvmppc_h_stuff_tce);

arch/powerpc/kvm/book3s_64_vio_hv.c

@@ -180,7 +180,7 @@ long kvmppc_gpa_to_ua(struct kvm *kvm, unsigned long gpa,
 EXPORT_SYMBOL_GPL(kvmppc_gpa_to_ua);
 
 #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
-long kvmppc_h_put_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
+long kvmppc_rm_h_put_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
 		unsigned long ioba, unsigned long tce)
 {
 	struct kvmppc_spapr_tce_table *stt = kvmppc_find_table(vcpu, liobn);
@@ -204,7 +204,6 @@ long kvmppc_h_put_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
 
 	return H_SUCCESS;
 }
-EXPORT_SYMBOL_GPL(kvmppc_h_put_tce);
 
 static long kvmppc_rm_ua_to_hpa(struct kvm_vcpu *vcpu,
 		unsigned long ua, unsigned long *phpa)
@@ -296,7 +295,7 @@ long kvmppc_rm_h_put_tce_indirect(struct kvm_vcpu *vcpu,
 	return ret;
 }
 
-long kvmppc_h_stuff_tce(struct kvm_vcpu *vcpu,
+long kvmppc_rm_h_stuff_tce(struct kvm_vcpu *vcpu,
 		unsigned long liobn, unsigned long ioba,
 		unsigned long tce_value, unsigned long npages)
 {
@@ -320,7 +319,6 @@ long kvmppc_h_stuff_tce(struct kvm_vcpu *vcpu,
 
 	return H_SUCCESS;
 }
-EXPORT_SYMBOL_GPL(kvmppc_h_stuff_tce);
 
 long kvmppc_h_get_tce(struct kvm_vcpu *vcpu, unsigned long liobn,
 		      unsigned long ioba)

arch/powerpc/kvm/book3s_hv_rmhandlers.S

@@ -1942,7 +1942,7 @@ hcall_real_table:
 	.long	DOTSYM(kvmppc_h_clear_ref) - hcall_real_table
 	.long	DOTSYM(kvmppc_h_protect) - hcall_real_table
 	.long	DOTSYM(kvmppc_h_get_tce) - hcall_real_table
-	.long	DOTSYM(kvmppc_h_put_tce) - hcall_real_table
+	.long	DOTSYM(kvmppc_rm_h_put_tce) - hcall_real_table
 	.long	0		/* 0x24 - H_SET_SPRG0 */
 	.long	DOTSYM(kvmppc_h_set_dabr) - hcall_real_table
 	.long	0		/* 0x2c */
@@ -2020,7 +2020,7 @@ hcall_real_table:
 	.long	0		/* 0x12c */
 	.long	0		/* 0x130 */
 	.long	DOTSYM(kvmppc_h_set_xdabr) - hcall_real_table
-	.long	DOTSYM(kvmppc_h_stuff_tce) - hcall_real_table
+	.long	DOTSYM(kvmppc_rm_h_stuff_tce) - hcall_real_table
 	.long	DOTSYM(kvmppc_rm_h_put_tce_indirect) - hcall_real_table
 	.long	0		/* 0x140 */
 	.long	0		/* 0x144 */

arch/powerpc/kvm/powerpc.c

@@ -96,6 +96,9 @@ int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu)
 		 * so we don't miss a request because the requester sees
 		 * OUTSIDE_GUEST_MODE and assumes we'll be checking requests
 		 * before next entering the guest (and thus doesn't IPI).
+		 * This also orders the write to mode from any reads
+		 * to the page tables done while the VCPU is running.
+		 * Please see the comment in kvm_flush_remote_tlbs.
 		 */
 		smp_mb();
 

arch/x86/include/asm/kvm_host.h

@@ -84,7 +84,8 @@
 			  | X86_CR4_PSE | X86_CR4_PAE | X86_CR4_MCE     \
 			  | X86_CR4_PGE | X86_CR4_PCE | X86_CR4_OSFXSR | X86_CR4_PCIDE \
 			  | X86_CR4_OSXSAVE | X86_CR4_SMEP | X86_CR4_FSGSBASE \
-			  | X86_CR4_OSXMMEXCPT | X86_CR4_VMXE | X86_CR4_SMAP))
+			  | X86_CR4_OSXMMEXCPT | X86_CR4_VMXE | X86_CR4_SMAP \
+			  | X86_CR4_PKE))
 
 #define CR8_RESERVED_BITS (~(unsigned long)X86_CR8_TPR)
 
@@ -187,12 +188,14 @@ enum {
 #define PFERR_USER_BIT 2
 #define PFERR_RSVD_BIT 3
 #define PFERR_FETCH_BIT 4
+#define PFERR_PK_BIT 5
 
 #define PFERR_PRESENT_MASK (1U << PFERR_PRESENT_BIT)
 #define PFERR_WRITE_MASK (1U << PFERR_WRITE_BIT)
 #define PFERR_USER_MASK (1U << PFERR_USER_BIT)
 #define PFERR_RSVD_MASK (1U << PFERR_RSVD_BIT)
 #define PFERR_FETCH_MASK (1U << PFERR_FETCH_BIT)
+#define PFERR_PK_MASK (1U << PFERR_PK_BIT)
 
 /* apic attention bits */
 #define KVM_APIC_CHECK_VAPIC	0
@@ -335,6 +338,14 @@ struct kvm_mmu {
 	 */
 	u8 permissions[16];
 
+	/*
+	* The pkru_mask indicates if protection key checks are needed.  It
+	* consists of 16 domains indexed by page fault error code bits [4:1],
+	* with PFEC.RSVD replaced by ACC_USER_MASK from the page tables.
+	* Each domain has 2 bits which are ANDed with AD and WD from PKRU.
+	*/
+	u32 pkru_mask;
+
 	u64 *pae_root;
 	u64 *lm_root;
 
@@ -874,6 +885,7 @@ struct kvm_x86_ops {
 	void (*cache_reg)(struct kvm_vcpu *vcpu, enum kvm_reg reg);
 	unsigned long (*get_rflags)(struct kvm_vcpu *vcpu);
 	void (*set_rflags)(struct kvm_vcpu *vcpu, unsigned long rflags);
+	u32 (*get_pkru)(struct kvm_vcpu *vcpu);
 	void (*fpu_activate)(struct kvm_vcpu *vcpu);
 	void (*fpu_deactivate)(struct kvm_vcpu *vcpu);
 

arch/x86/include/asm/pgtable.h

@@ -107,6 +107,12 @@ static inline u32 read_pkru(void)
 	return 0;
 }
 
+static inline void write_pkru(u32 pkru)
+{
+	if (boot_cpu_has(X86_FEATURE_OSPKE))
+		__write_pkru(pkru);
+}
+
 static inline int pte_young(pte_t pte)
 {
 	return pte_flags(pte) & _PAGE_ACCESSED;

arch/x86/include/asm/special_insns.h

@@ -113,11 +113,27 @@ static inline u32 __read_pkru(void)
 		     : "c" (ecx));
 	return pkru;
 }
+
+static inline void __write_pkru(u32 pkru)
+{
+	u32 ecx = 0, edx = 0;
+
+	/*
+	 * "wrpkru" instruction.  Loads contents in EAX to PKRU,
+	 * requires that ecx = edx = 0.
+	 */
+	asm volatile(".byte 0x0f,0x01,0xef\n\t"
+		     : : "a" (pkru), "c"(ecx), "d"(edx));
+}
 #else
 static inline u32 __read_pkru(void)
 {
 	return 0;
 }
+
+static inline void __write_pkru(u32 pkru)
+{
+}
 #endif
 
 static inline void native_wbinvd(void)

arch/x86/kernel/kvm.c

@@ -36,6 +36,7 @@
 #include <linux/kprobes.h>
 #include <linux/debugfs.h>
 #include <linux/nmi.h>
+#include <linux/swait.h>
 #include <asm/timer.h>
 #include <asm/cpu.h>
 #include <asm/traps.h>
@@ -91,14 +92,14 @@ static void kvm_io_delay(void)
 
 struct kvm_task_sleep_node {
 	struct hlist_node link;
-	wait_queue_head_t wq;
+	struct swait_queue_head wq;
 	u32 token;
 	int cpu;
 	bool halted;
 };
 
 static struct kvm_task_sleep_head {
-	spinlock_t lock;
+	raw_spinlock_t lock;
 	struct hlist_head list;
 } async_pf_sleepers[KVM_TASK_SLEEP_HASHSIZE];
 
@@ -122,17 +123,17 @@ void kvm_async_pf_task_wait(u32 token)
 	u32 key = hash_32(token, KVM_TASK_SLEEP_HASHBITS);
 	struct kvm_task_sleep_head *b = &async_pf_sleepers[key];
 	struct kvm_task_sleep_node n, *e;
-	DEFINE_WAIT(wait);
+	DECLARE_SWAITQUEUE(wait);
 
 	rcu_irq_enter();
 
-	spin_lock(&b->lock);
+	raw_spin_lock(&b->lock);
 	e = _find_apf_task(b, token);
 	if (e) {
 		/* dummy entry exist -> wake up was delivered ahead of PF */
 		hlist_del(&e->link);
 		kfree(e);
-		spin_unlock(&b->lock);
+		raw_spin_unlock(&b->lock);
 
 		rcu_irq_exit();
 		return;
@@ -141,13 +142,13 @@ void kvm_async_pf_task_wait(u32 token)
 	n.token = token;
 	n.cpu = smp_processor_id();
 	n.halted = is_idle_task(current) || preempt_count() > 1;
-	init_waitqueue_head(&n.wq);
+	init_swait_queue_head(&n.wq);
 	hlist_add_head(&n.link, &b->list);
-	spin_unlock(&b->lock);
+	raw_spin_unlock(&b->lock);
 
 	for (;;) {
 		if (!n.halted)
-			prepare_to_wait(&n.wq, &wait, TASK_UNINTERRUPTIBLE);
+			prepare_to_swait(&n.wq, &wait, TASK_UNINTERRUPTIBLE);
 		if (hlist_unhashed(&n.link))
 			break;
 
@@ -166,7 +167,7 @@ void kvm_async_pf_task_wait(u32 token)
 		}
 	}
 	if (!n.halted)
-		finish_wait(&n.wq, &wait);
+		finish_swait(&n.wq, &wait);
 
 	rcu_irq_exit();
 	return;
@@ -178,8 +179,8 @@ static void apf_task_wake_one(struct kvm_task_sleep_node *n)
 	hlist_del_init(&n->link);
 	if (n->halted)
 		smp_send_reschedule(n->cpu);
-	else if (waitqueue_active(&n->wq))
-		wake_up(&n->wq);
+	else if (swait_active(&n->wq))
+		swake_up(&n->wq);
 }
 
 static void apf_task_wake_all(void)
@@ -189,14 +190,14 @@ static void apf_task_wake_all(void)
 	for (i = 0; i < KVM_TASK_SLEEP_HASHSIZE; i++) {
 		struct hlist_node *p, *next;
 		struct kvm_task_sleep_head *b = &async_pf_sleepers[i];
-		spin_lock(&b->lock);
+		raw_spin_lock(&b->lock);
 		hlist_for_each_safe(p, next, &b->list) {
 			struct kvm_task_sleep_node *n =
 				hlist_entry(p, typeof(*n), link);
 			if (n->cpu == smp_processor_id())
 				apf_task_wake_one(n);
 		}
-		spin_unlock(&b->lock);
+		raw_spin_unlock(&b->lock);
 	}
 }
 
@@ -212,7 +213,7 @@ void kvm_async_pf_task_wake(u32 token)
 	}
 
 again:
-	spin_lock(&b->lock);
+	raw_spin_lock(&b->lock);
 	n = _find_apf_task(b, token);
 	if (!n) {
 		/*
@@ -225,17 +226,17 @@ void kvm_async_pf_task_wake(u32 token)
 			 * Allocation failed! Busy wait while other cpu
 			 * handles async PF.
 			 */
-			spin_unlock(&b->lock);
+			raw_spin_unlock(&b->lock);
 			cpu_relax();
 			goto again;
 		}
 		n->token = token;
 		n->cpu = smp_processor_id();
-		init_waitqueue_head(&n->wq);
+		init_swait_queue_head(&n->wq);
 		hlist_add_head(&n->link, &b->list);
 	} else
 		apf_task_wake_one(n);
-	spin_unlock(&b->lock);
+	raw_spin_unlock(&b->lock);
 	return;
 }
 EXPORT_SYMBOL_GPL(kvm_async_pf_task_wake);
@@ -486,7 +487,7 @@ void __init kvm_guest_init(void)
 	paravirt_ops_setup();
 	register_reboot_notifier(&kvm_pv_reboot_nb);
 	for (i = 0; i < KVM_TASK_SLEEP_HASHSIZE; i++)
-		spin_lock_init(&async_pf_sleepers[i].lock);
+		raw_spin_lock_init(&async_pf_sleepers[i].lock);
 	if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF))
 		x86_init.irqs.trap_init = kvm_apf_trap_init;
 

arch/x86/kvm/cpuid.c

@@ -88,6 +88,16 @@ int kvm_update_cpuid(struct kvm_vcpu *vcpu)
 			apic->lapic_timer.timer_mode_mask = 1 << 17;
 	}
 
+	best = kvm_find_cpuid_entry(vcpu, 7, 0);
+	if (best) {
+		/* Update OSPKE bit */
+		if (boot_cpu_has(X86_FEATURE_PKU) && best->function == 0x7) {
+			best->ecx &= ~F(OSPKE);
+			if (kvm_read_cr4_bits(vcpu, X86_CR4_PKE))
+				best->ecx |= F(OSPKE);
+		}
+	}
+
 	best = kvm_find_cpuid_entry(vcpu, 0xD, 0);
 	if (!best) {
 		vcpu->arch.guest_supported_xcr0 = 0;
@@ -305,7 +315,7 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 	unsigned f_xsaves = kvm_x86_ops->xsaves_supported() ? F(XSAVES) : 0;
 
 	/* cpuid 1.edx */
-	const u32 kvm_supported_word0_x86_features =
+	const u32 kvm_cpuid_1_edx_x86_features =
 		F(FPU) | F(VME) | F(DE) | F(PSE) |
 		F(TSC) | F(MSR) | F(PAE) | F(MCE) |
 		F(CX8) | F(APIC) | 0 /* Reserved */ | F(SEP) |
@@ -315,7 +325,7 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		F(FXSR) | F(XMM) | F(XMM2) | F(SELFSNOOP) |
 		0 /* HTT, TM, Reserved, PBE */;
 	/* cpuid 0x80000001.edx */
-	const u32 kvm_supported_word1_x86_features =
+	const u32 kvm_cpuid_8000_0001_edx_x86_features =
 		F(FPU) | F(VME) | F(DE) | F(PSE) |
 		F(TSC) | F(MSR) | F(PAE) | F(MCE) |
 		F(CX8) | F(APIC) | 0 /* Reserved */ | F(SYSCALL) |
@@ -325,7 +335,7 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		F(FXSR) | F(FXSR_OPT) | f_gbpages | f_rdtscp |
 		0 /* Reserved */ | f_lm | F(3DNOWEXT) | F(3DNOW);
 	/* cpuid 1.ecx */
-	const u32 kvm_supported_word4_x86_features =
+	const u32 kvm_cpuid_1_ecx_x86_features =
 		/* NOTE: MONITOR (and MWAIT) are emulated as NOP,
 		 * but *not* advertised to guests via CPUID ! */
 		F(XMM3) | F(PCLMULQDQ) | 0 /* DTES64, MONITOR */ |
@@ -337,29 +347,32 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		0 /* Reserved*/ | F(AES) | F(XSAVE) | 0 /* OSXSAVE */ | F(AVX) |
 		F(F16C) | F(RDRAND);
 	/* cpuid 0x80000001.ecx */
-	const u32 kvm_supported_word6_x86_features =
+	const u32 kvm_cpuid_8000_0001_ecx_x86_features =
 		F(LAHF_LM) | F(CMP_LEGACY) | 0 /*SVM*/ | 0 /* ExtApicSpace */ |
 		F(CR8_LEGACY) | F(ABM) | F(SSE4A) | F(MISALIGNSSE) |
 		F(3DNOWPREFETCH) | F(OSVW) | 0 /* IBS */ | F(XOP) |
 		0 /* SKINIT, WDT, LWP */ | F(FMA4) | F(TBM);
 
 	/* cpuid 0xC0000001.edx */
-	const u32 kvm_supported_word5_x86_features =
+	const u32 kvm_cpuid_C000_0001_edx_x86_features =
 		F(XSTORE) | F(XSTORE_EN) | F(XCRYPT) | F(XCRYPT_EN) |
 		F(ACE2) | F(ACE2_EN) | F(PHE) | F(PHE_EN) |
 		F(PMM) | F(PMM_EN);
 
 	/* cpuid 7.0.ebx */
-	const u32 kvm_supported_word9_x86_features =
+	const u32 kvm_cpuid_7_0_ebx_x86_features =
 		F(FSGSBASE) | F(BMI1) | F(HLE) | F(AVX2) | F(SMEP) |
 		F(BMI2) | F(ERMS) | f_invpcid | F(RTM) | f_mpx | F(RDSEED) |
 		F(ADX) | F(SMAP) | F(AVX512F) | F(AVX512PF) | F(AVX512ER) |
 		F(AVX512CD) | F(CLFLUSHOPT) | F(CLWB) | F(PCOMMIT);
 
 	/* cpuid 0xD.1.eax */
-	const u32 kvm_supported_word10_x86_features =
+	const u32 kvm_cpuid_D_1_eax_x86_features =
 		F(XSAVEOPT) | F(XSAVEC) | F(XGETBV1) | f_xsaves;
 
+	/* cpuid 7.0.ecx*/
+	const u32 kvm_cpuid_7_0_ecx_x86_features = F(PKU) | 0 /*OSPKE*/;
+
 	/* all calls to cpuid_count() should be made on the same cpu */
 	get_cpu();
 
@@ -376,10 +389,10 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		entry->eax = min(entry->eax, (u32)0xd);
 		break;
 	case 1:
-		entry->edx &= kvm_supported_word0_x86_features;
-		cpuid_mask(&entry->edx, 0);
-		entry->ecx &= kvm_supported_word4_x86_features;
-		cpuid_mask(&entry->ecx, 4);
+		entry->edx &= kvm_cpuid_1_edx_x86_features;
+		cpuid_mask(&entry->edx, CPUID_1_EDX);
+		entry->ecx &= kvm_cpuid_1_ecx_x86_features;
+		cpuid_mask(&entry->ecx, CPUID_1_ECX);
 		/* we support x2apic emulation even if host does not support
 		 * it since we emulate x2apic in software */
 		entry->ecx |= F(X2APIC);
@@ -433,14 +446,20 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		entry->flags |= KVM_CPUID_FLAG_SIGNIFCANT_INDEX;
 		/* Mask ebx against host capability word 9 */
 		if (index == 0) {
-			entry->ebx &= kvm_supported_word9_x86_features;
-			cpuid_mask(&entry->ebx, 9);
+			entry->ebx &= kvm_cpuid_7_0_ebx_x86_features;
+			cpuid_mask(&entry->ebx, CPUID_7_0_EBX);
 			// TSC_ADJUST is emulated
 			entry->ebx |= F(TSC_ADJUST);
-		} else
+			entry->ecx &= kvm_cpuid_7_0_ecx_x86_features;
+			cpuid_mask(&entry->ecx, CPUID_7_ECX);
+			/* PKU is not yet implemented for shadow paging. */
+			if (!tdp_enabled)
+				entry->ecx &= ~F(PKU);
+		} else {
 			entry->ebx = 0;
-		entry->eax = 0;
 			entry->ecx = 0;
+		}
+		entry->eax = 0;
 		entry->edx = 0;
 		break;
 	}
@@ -514,7 +533,7 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 
 			do_cpuid_1_ent(&entry[i], function, idx);
 			if (idx == 1) {
-				entry[i].eax &= kvm_supported_word10_x86_features;
+				entry[i].eax &= kvm_cpuid_D_1_eax_x86_features;
 				entry[i].ebx = 0;
 				if (entry[i].eax & (F(XSAVES)|F(XSAVEC)))
 					entry[i].ebx =
@@ -564,10 +583,10 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		entry->eax = min(entry->eax, 0x8000001a);
 		break;
 	case 0x80000001:
-		entry->edx &= kvm_supported_word1_x86_features;
-		cpuid_mask(&entry->edx, 1);
-		entry->ecx &= kvm_supported_word6_x86_features;
-		cpuid_mask(&entry->ecx, 6);
+		entry->edx &= kvm_cpuid_8000_0001_edx_x86_features;
+		cpuid_mask(&entry->edx, CPUID_8000_0001_EDX);
+		entry->ecx &= kvm_cpuid_8000_0001_ecx_x86_features;
+		cpuid_mask(&entry->ecx, CPUID_8000_0001_ECX);
 		break;
 	case 0x80000007: /* Advanced power management */
 		/* invariant TSC is CPUID.80000007H:EDX[8] */
@@ -600,8 +619,8 @@ static inline int __do_cpuid_ent(struct kvm_cpuid_entry2 *entry, u32 function,
 		entry->eax = min(entry->eax, 0xC0000004);
 		break;
 	case 0xC0000001:
-		entry->edx &= kvm_supported_word5_x86_features;
-		cpuid_mask(&entry->edx, 5);
+		entry->edx &= kvm_cpuid_C000_0001_edx_x86_features;
+		cpuid_mask(&entry->edx, CPUID_C000_0001_EDX);
 		break;
 	case 3: /* Processor serial number */
 	case 5: /* MONITOR/MWAIT */

arch/x86/kvm/cpuid.h

@@ -80,6 +80,14 @@ static inline bool guest_cpuid_has_fsgsbase(struct kvm_vcpu *vcpu)
 	return best && (best->ebx & bit(X86_FEATURE_FSGSBASE));
 }
 
+static inline bool guest_cpuid_has_pku(struct kvm_vcpu *vcpu)
+{
+	struct kvm_cpuid_entry2 *best;
+
+	best = kvm_find_cpuid_entry(vcpu, 7, 0);
+	return best && (best->ecx & bit(X86_FEATURE_PKU));
+}
+
 static inline bool guest_cpuid_has_longmode(struct kvm_vcpu *vcpu)
 {
 	struct kvm_cpuid_entry2 *best;

arch/x86/kvm/kvm_cache_regs.h

@@ -84,6 +84,11 @@ static inline u64 kvm_read_edx_eax(struct kvm_vcpu *vcpu)
 		| ((u64)(kvm_register_read(vcpu, VCPU_REGS_RDX) & -1u) << 32);
 }
 
+static inline u32 kvm_read_pkru(struct kvm_vcpu *vcpu)
+{
+	return kvm_x86_ops->get_pkru(vcpu);
+}
+
 static inline void enter_guest_mode(struct kvm_vcpu *vcpu)
 {
 	vcpu->arch.hflags |= HF_GUEST_MASK;

arch/x86/kvm/mmu.c

@@ -632,12 +632,12 @@ static void walk_shadow_page_lockless_begin(struct kvm_vcpu *vcpu)
 	 * kvm_flush_remote_tlbs() IPI to all active vcpus.
 	 */
 	local_irq_disable();
-	vcpu->mode = READING_SHADOW_PAGE_TABLES;
+
 	/*
 	 * Make sure a following spte read is not reordered ahead of the write
 	 * to vcpu->mode.
 	 */
-	smp_mb();
+	smp_store_mb(vcpu->mode, READING_SHADOW_PAGE_TABLES);
 }
 
 static void walk_shadow_page_lockless_end(struct kvm_vcpu *vcpu)
@@ -647,8 +647,7 @@ static void walk_shadow_page_lockless_end(struct kvm_vcpu *vcpu)
 	 * reads to sptes.  If it does, kvm_commit_zap_page() can see us
 	 * OUTSIDE_GUEST_MODE and proceed to free the shadow page table.
 	 */
-	smp_mb();
-	vcpu->mode = OUTSIDE_GUEST_MODE;
+	smp_store_release(&vcpu->mode, OUTSIDE_GUEST_MODE);
 	local_irq_enable();
 }
 
@@ -2390,14 +2389,13 @@ static void kvm_mmu_commit_zap_page(struct kvm *kvm,
 		return;
 
 	/*
-	 * wmb: make sure everyone sees our modifications to the page tables
-	 * rmb: make sure we see changes to vcpu->mode
-	 */
-	smp_mb();
-
-	/*
-	 * Wait for all vcpus to exit guest mode and/or lockless shadow
-	 * page table walks.
+	 * We need to make sure everyone sees our modifications to
+	 * the page tables and see changes to vcpu->mode here. The barrier
+	 * in the kvm_flush_remote_tlbs() achieves this. This pairs
+	 * with vcpu_enter_guest and walk_shadow_page_lockless_begin/end.
+	 *
+	 * In addition, kvm_flush_remote_tlbs waits for all vcpus to exit
+	 * guest mode and/or lockless shadow page table walks.
 	 */
 	kvm_flush_remote_tlbs(kvm);
 
@@ -3923,6 +3921,81 @@ static void update_permission_bitmask(struct kvm_vcpu *vcpu,
 	}
 }
 
+/*
+* PKU is an additional mechanism by which the paging controls access to
+* user-mode addresses based on the value in the PKRU register.  Protection
+* key violations are reported through a bit in the page fault error code.
+* Unlike other bits of the error code, the PK bit is not known at the
+* call site of e.g. gva_to_gpa; it must be computed directly in
+* permission_fault based on two bits of PKRU, on some machine state (CR4,
+* CR0, EFER, CPL), and on other bits of the error code and the page tables.
+*
+* In particular the following conditions come from the error code, the
+* page tables and the machine state:
+* - PK is always zero unless CR4.PKE=1 and EFER.LMA=1
+* - PK is always zero if RSVD=1 (reserved bit set) or F=1 (instruction fetch)
+* - PK is always zero if U=0 in the page tables
+* - PKRU.WD is ignored if CR0.WP=0 and the access is a supervisor access.
+*
+* The PKRU bitmask caches the result of these four conditions.  The error
+* code (minus the P bit) and the page table's U bit form an index into the
+* PKRU bitmask.  Two bits of the PKRU bitmask are then extracted and ANDed
+* with the two bits of the PKRU register corresponding to the protection key.
+* For the first three conditions above the bits will be 00, thus masking
+* away both AD and WD.  For all reads or if the last condition holds, WD
+* only will be masked away.
+*/
+static void update_pkru_bitmask(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
+				bool ept)
+{
+	unsigned bit;
+	bool wp;
+
+	if (ept) {
+		mmu->pkru_mask = 0;
+		return;
+	}
+
+	/* PKEY is enabled only if CR4.PKE and EFER.LMA are both set. */
+	if (!kvm_read_cr4_bits(vcpu, X86_CR4_PKE) || !is_long_mode(vcpu)) {
+		mmu->pkru_mask = 0;
+		return;
+	}
+
+	wp = is_write_protection(vcpu);
+
+	for (bit = 0; bit < ARRAY_SIZE(mmu->permissions); ++bit) {
+		unsigned pfec, pkey_bits;
+		bool check_pkey, check_write, ff, uf, wf, pte_user;
+
+		pfec = bit << 1;
+		ff = pfec & PFERR_FETCH_MASK;
+		uf = pfec & PFERR_USER_MASK;
+		wf = pfec & PFERR_WRITE_MASK;
+
+		/* PFEC.RSVD is replaced by ACC_USER_MASK. */
+		pte_user = pfec & PFERR_RSVD_MASK;
+
+		/*
+		 * Only need to check the access which is not an
+		 * instruction fetch and is to a user page.
+		 */
+		check_pkey = (!ff && pte_user);
+		/*
+		 * write access is controlled by PKRU if it is a
+		 * user access or CR0.WP = 1.
+		 */
+		check_write = check_pkey && wf && (uf || wp);
+
+		/* PKRU.AD stops both read and write access. */
+		pkey_bits = !!check_pkey;
+		/* PKRU.WD stops write access. */
+		pkey_bits |= (!!check_write) << 1;
+
+		mmu->pkru_mask |= (pkey_bits & 3) << pfec;
+	}
+}
+
 static void update_last_nonleaf_level(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu)
 {
 	unsigned root_level = mmu->root_level;
@@ -3941,6 +4014,7 @@ static void paging64_init_context_common(struct kvm_vcpu *vcpu,
 
 	reset_rsvds_bits_mask(vcpu, context);
 	update_permission_bitmask(vcpu, context, false);
+	update_pkru_bitmask(vcpu, context, false);
 	update_last_nonleaf_level(vcpu, context);
 
 	MMU_WARN_ON(!is_pae(vcpu));
@@ -3968,6 +4042,7 @@ static void paging32_init_context(struct kvm_vcpu *vcpu,
 
 	reset_rsvds_bits_mask(vcpu, context);
 	update_permission_bitmask(vcpu, context, false);
+	update_pkru_bitmask(vcpu, context, false);
 	update_last_nonleaf_level(vcpu, context);
 
 	context->page_fault = paging32_page_fault;
@@ -4026,6 +4101,7 @@ static void init_kvm_tdp_mmu(struct kvm_vcpu *vcpu)
 	}
 
 	update_permission_bitmask(vcpu, context, false);
+	update_pkru_bitmask(vcpu, context, false);
 	update_last_nonleaf_level(vcpu, context);
 	reset_tdp_shadow_zero_bits_mask(vcpu, context);
 }
@@ -4078,6 +4154,7 @@ void kvm_init_shadow_ept_mmu(struct kvm_vcpu *vcpu, bool execonly)
 	context->direct_map = false;
 
 	update_permission_bitmask(vcpu, context, true);
+	update_pkru_bitmask(vcpu, context, true);
 	reset_rsvds_bits_mask_ept(vcpu, context, execonly);
 	reset_ept_shadow_zero_bits_mask(vcpu, context, execonly);
 }
@@ -4132,6 +4209,7 @@ static void init_kvm_nested_mmu(struct kvm_vcpu *vcpu)
 	}
 
 	update_permission_bitmask(vcpu, g_context, false);
+	update_pkru_bitmask(vcpu, g_context, false);
 	update_last_nonleaf_level(vcpu, g_context);
 }
 

arch/x86/kvm/mmu.h

@@ -10,10 +10,11 @@
 #define PT32_ENT_PER_PAGE (1 << PT32_PT_BITS)
 
 #define PT_WRITABLE_SHIFT 1
+#define PT_USER_SHIFT 2
 
 #define PT_PRESENT_MASK (1ULL << 0)
 #define PT_WRITABLE_MASK (1ULL << PT_WRITABLE_SHIFT)
-#define PT_USER_MASK (1ULL << 2)
+#define PT_USER_MASK (1ULL << PT_USER_SHIFT)
 #define PT_PWT_MASK (1ULL << 3)
 #define PT_PCD_MASK (1ULL << 4)
 #define PT_ACCESSED_SHIFT 5
@@ -141,11 +142,16 @@ static inline bool is_write_protection(struct kvm_vcpu *vcpu)
 }
 
 /*
- * Will a fault with a given page-fault error code (pfec) cause a permission
- * fault with the given access (in ACC_* format)?
+ * Check if a given access (described through the I/D, W/R and U/S bits of a
+ * page fault error code pfec) causes a permission fault with the given PTE
+ * access rights (in ACC_* format).
+ *
+ * Return zero if the access does not fault; return the page fault error code
+ * if the access faults.
  */
-static inline bool permission_fault(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
-				    unsigned pte_access, unsigned pfec)
+static inline u8 permission_fault(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
+				  unsigned pte_access, unsigned pte_pkey,
+				  unsigned pfec)
 {
 	int cpl = kvm_x86_ops->get_cpl(vcpu);
 	unsigned long rflags = kvm_x86_ops->get_rflags(vcpu);
@@ -166,10 +172,32 @@ static inline bool permission_fault(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu,
 	unsigned long smap = (cpl - 3) & (rflags & X86_EFLAGS_AC);
 	int index = (pfec >> 1) +
 		    (smap >> (X86_EFLAGS_AC_BIT - PFERR_RSVD_BIT + 1));
+	bool fault = (mmu->permissions[index] >> pte_access) & 1;
+
+	WARN_ON(pfec & (PFERR_PK_MASK | PFERR_RSVD_MASK));
+	pfec |= PFERR_PRESENT_MASK;
+
+	if (unlikely(mmu->pkru_mask)) {
+		u32 pkru_bits, offset;
+
+		/*
+		* PKRU defines 32 bits, there are 16 domains and 2
+		* attribute bits per domain in pkru.  pte_pkey is the
+		* index of the protection domain, so pte_pkey * 2 is
+		* is the index of the first bit for the domain.
+		*/
+		pkru_bits = (kvm_read_pkru(vcpu) >> (pte_pkey * 2)) & 3;
+
+		/* clear present bit, replace PFEC.RSVD with ACC_USER_MASK. */
+		offset = pfec - 1 +
+			((pte_access & PT_USER_MASK) << (PFERR_RSVD_BIT - PT_USER_SHIFT));
 
-	WARN_ON(pfec & PFERR_RSVD_MASK);
+		pkru_bits &= mmu->pkru_mask >> offset;
+		pfec |= -pkru_bits & PFERR_PK_MASK;
+		fault |= (pkru_bits != 0);
+	}
 
-	return (mmu->permissions[index] >> pte_access) & 1;
+	return -(uint32_t)fault & pfec;
 }
 
 void kvm_mmu_invalidate_zap_all_pages(struct kvm *kvm);

arch/x86/kvm/page_track.c

@@ -142,12 +142,17 @@ void kvm_slot_page_track_remove_page(struct kvm *kvm,
 bool kvm_page_track_is_active(struct kvm_vcpu *vcpu, gfn_t gfn,
 			      enum kvm_page_track_mode mode)
 {
-	struct kvm_memory_slot *slot = kvm_vcpu_gfn_to_memslot(vcpu, gfn);
-	int index = gfn_to_index(gfn, slot->base_gfn, PT_PAGE_TABLE_LEVEL);
+	struct kvm_memory_slot *slot;
+	int index;
 
 	if (WARN_ON(!page_track_mode_is_valid(mode)))
 		return false;
 
+	slot = kvm_vcpu_gfn_to_memslot(vcpu, gfn);
+	if (!slot)
+		return false;
+
+	index = gfn_to_index(gfn, slot->base_gfn, PT_PAGE_TABLE_LEVEL);
 	return !!ACCESS_ONCE(slot->arch.gfn_track[mode][index]);
 }
 

arch/x86/kvm/paging_tmpl.h

@@ -257,6 +257,17 @@ static int FNAME(update_accessed_dirty_bits)(struct kvm_vcpu *vcpu,
 	return 0;
 }
 
+static inline unsigned FNAME(gpte_pkeys)(struct kvm_vcpu *vcpu, u64 gpte)
+{
+	unsigned pkeys = 0;
+#if PTTYPE == 64
+	pte_t pte = {.pte = gpte};
+
+	pkeys = pte_flags_pkey(pte_flags(pte));
+#endif
+	return pkeys;
+}
+
 /*
  * Fetch a guest pte for a guest virtual address
  */
@@ -268,7 +279,7 @@ static int FNAME(walk_addr_generic)(struct guest_walker *walker,
 	pt_element_t pte;
 	pt_element_t __user *uninitialized_var(ptep_user);
 	gfn_t table_gfn;
-	unsigned index, pt_access, pte_access, accessed_dirty;
+	unsigned index, pt_access, pte_access, accessed_dirty, pte_pkey;
 	gpa_t pte_gpa;
 	int offset;
 	const int write_fault = access & PFERR_WRITE_MASK;
@@ -359,10 +370,10 @@ static int FNAME(walk_addr_generic)(struct guest_walker *walker,
 		walker->ptes[walker->level - 1] = pte;
 	} while (!is_last_gpte(mmu, walker->level, pte));
 
-	if (unlikely(permission_fault(vcpu, mmu, pte_access, access))) {
-		errcode |= PFERR_PRESENT_MASK;
+	pte_pkey = FNAME(gpte_pkeys)(vcpu, pte);
+	errcode = permission_fault(vcpu, mmu, pte_access, pte_pkey, access);
+	if (unlikely(errcode))
 		goto error;
-	}
 
 	gfn = gpte_to_gfn_lvl(pte, walker->level);
 	gfn += (addr & PT_LVL_OFFSET_MASK(walker->level)) >> PAGE_SHIFT;
@@ -949,6 +960,12 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 			return 0;
 
 		if (FNAME(prefetch_invalid_gpte)(vcpu, sp, &sp->spt[i], gpte)) {
+			/*
+			 * Update spte before increasing tlbs_dirty to make
+			 * sure no tlb flush is lost after spte is zapped; see
+			 * the comments in kvm_flush_remote_tlbs().
+			 */
+			smp_wmb();
 			vcpu->kvm->tlbs_dirty++;
 			continue;
 		}
@@ -964,6 +981,11 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 
 		if (gfn != sp->gfns[i]) {
 			drop_spte(vcpu->kvm, &sp->spt[i]);
+			/*
+			 * The same as above where we are doing
+			 * prefetch_invalid_gpte().
+			 */
+			smp_wmb();
 			vcpu->kvm->tlbs_dirty++;
 			continue;
 		}

arch/x86/kvm/svm.c

@@ -1280,6 +1280,11 @@ static void svm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags)
 	to_svm(vcpu)->vmcb->save.rflags = rflags;
 }
 
+static u32 svm_get_pkru(struct kvm_vcpu *vcpu)
+{
+	return 0;
+}
+
 static void svm_cache_reg(struct kvm_vcpu *vcpu, enum kvm_reg reg)
 {
 	switch (reg) {
@@ -4347,6 +4352,9 @@ static struct kvm_x86_ops svm_x86_ops = {
 	.cache_reg = svm_cache_reg,
 	.get_rflags = svm_get_rflags,
 	.set_rflags = svm_set_rflags,
+
+	.get_pkru = svm_get_pkru,
+
 	.fpu_activate = svm_fpu_activate,
 	.fpu_deactivate = svm_fpu_deactivate,
 

arch/x86/kvm/vmx.c

@@ -598,6 +598,10 @@ struct vcpu_vmx {
 	struct page *pml_pg;
 
 	u64 current_tsc_ratio;
+
+	bool guest_pkru_valid;
+	u32 guest_pkru;
+	u32 host_pkru;
 };
 
 enum segment_cache_field {
@@ -2107,6 +2111,7 @@ static void vmx_vcpu_pi_load(struct kvm_vcpu *vcpu, int cpu)
 	} while (cmpxchg(&pi_desc->control, old.control,
 			new.control) != old.control);
 }
+
 /*
  * Switches to specified vcpu, until a matching vcpu_put(), but assumes
  * vcpu mutex is already taken.
@@ -2167,6 +2172,7 @@ static void vmx_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
 	}
 
 	vmx_vcpu_pi_load(vcpu, cpu);
+	vmx->host_pkru = read_pkru();
 }
 
 static void vmx_vcpu_pi_put(struct kvm_vcpu *vcpu)
@@ -2286,6 +2292,11 @@ static void vmx_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags)
 	vmcs_writel(GUEST_RFLAGS, rflags);
 }
 
+static u32 vmx_get_pkru(struct kvm_vcpu *vcpu)
+{
+	return to_vmx(vcpu)->guest_pkru;
+}
+
 static u32 vmx_get_interrupt_shadow(struct kvm_vcpu *vcpu)
 {
 	u32 interruptibility = vmcs_read32(GUEST_INTERRUPTIBILITY_INFO);
@@ -2712,8 +2723,15 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx)
 	} else
 		vmx->nested.nested_vmx_ept_caps = 0;
 
+	/*
+	 * Old versions of KVM use the single-context version without
+	 * checking for support, so declare that it is supported even
+	 * though it is treated as global context.  The alternative is
+	 * not failing the single-context invvpid, and it is worse.
+	 */
 	if (enable_vpid)
 		vmx->nested.nested_vmx_vpid_caps = VMX_VPID_INVVPID_BIT |
+				VMX_VPID_EXTENT_SINGLE_CONTEXT_BIT |
 				VMX_VPID_EXTENT_GLOBAL_CONTEXT_BIT;
 	else
 		vmx->nested.nested_vmx_vpid_caps = 0;
@@ -3886,13 +3904,17 @@ static int vmx_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
 
 	if (!enable_unrestricted_guest && !is_paging(vcpu))
 		/*
-		 * SMEP/SMAP is disabled if CPU is in non-paging mode in
-		 * hardware.  However KVM always uses paging mode without
-		 * unrestricted guest.
-		 * To emulate this behavior, SMEP/SMAP needs to be manually
-		 * disabled when guest switches to non-paging mode.
+		 * SMEP/SMAP/PKU is disabled if CPU is in non-paging mode in
+		 * hardware.  To emulate this behavior, SMEP/SMAP/PKU needs
+		 * to be manually disabled when guest switches to non-paging
+		 * mode.
+		 *
+		 * If !enable_unrestricted_guest, the CPU is always running
+		 * with CR0.PG=1 and CR4 needs to be modified.
+		 * If enable_unrestricted_guest, the CPU automatically
+		 * disables SMEP/SMAP/PKU when the guest sets CR0.PG=0.
 		 */
-		hw_cr4 &= ~(X86_CR4_SMEP | X86_CR4_SMAP);
+		hw_cr4 &= ~(X86_CR4_SMEP | X86_CR4_SMAP | X86_CR4_PKE);
 
 	vmcs_writel(CR4_READ_SHADOW, cr4);
 	vmcs_writel(GUEST_CR4, hw_cr4);
@@ -7399,6 +7421,7 @@ static int handle_invept(struct kvm_vcpu *vcpu)
 	if (!(types & (1UL << type))) {
 		nested_vmx_failValid(vcpu,
 				VMXERR_INVALID_OPERAND_TO_INVEPT_INVVPID);
+		skip_emulated_instruction(vcpu);
 		return 1;
 	}
 
@@ -7457,6 +7480,7 @@ static int handle_invvpid(struct kvm_vcpu *vcpu)
 	if (!(types & (1UL << type))) {
 		nested_vmx_failValid(vcpu,
 			VMXERR_INVALID_OPERAND_TO_INVEPT_INVVPID);
+		skip_emulated_instruction(vcpu);
 		return 1;
 	}
 
@@ -7473,12 +7497,17 @@ static int handle_invvpid(struct kvm_vcpu *vcpu)
 	}
 
 	switch (type) {
+	case VMX_VPID_EXTENT_SINGLE_CONTEXT:
+		/*
+		 * Old versions of KVM use the single-context version so we
+		 * have to support it; just treat it the same as all-context.
+		 */
 	case VMX_VPID_EXTENT_ALL_CONTEXT:
 		__vmx_flush_tlb(vcpu, to_vmx(vcpu)->nested.vpid02);
 		nested_vmx_succeed(vcpu);
 		break;
 	default:
-		/* Trap single context invalidation invvpid calls */
+		/* Trap individual address invalidation invvpid calls */
 		BUG_ON(1);
 		break;
 	}
@@ -8621,6 +8650,9 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu)
 	if (vcpu->guest_debug & KVM_GUESTDBG_SINGLESTEP)
 		vmx_set_interrupt_shadow(vcpu, 0);
 
+	if (vmx->guest_pkru_valid)
+		__write_pkru(vmx->guest_pkru);
+
 	atomic_switch_perf_msrs(vmx);
 	debugctlmsr = get_debugctlmsr();
 
@@ -8760,6 +8792,20 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu)
 
 	vmx->exit_reason = vmcs_read32(VM_EXIT_REASON);
 
+	/*
+	 * eager fpu is enabled if PKEY is supported and CR4 is switched
+	 * back on host, so it is safe to read guest PKRU from current
+	 * XSAVE.
+	 */
+	if (boot_cpu_has(X86_FEATURE_OSPKE)) {
+		vmx->guest_pkru = __read_pkru();
+		if (vmx->guest_pkru != vmx->host_pkru) {
+			vmx->guest_pkru_valid = true;
+			__write_pkru(vmx->host_pkru);
+		} else
+			vmx->guest_pkru_valid = false;
+	}
+
 	/*
 	 * the KVM_REQ_EVENT optimization bit is only on for one entry, and if
 	 * we did not inject a still-pending event to L1 now because of
@@ -10884,6 +10930,9 @@ static struct kvm_x86_ops vmx_x86_ops = {
 	.cache_reg = vmx_cache_reg,
 	.get_rflags = vmx_get_rflags,
 	.set_rflags = vmx_set_rflags,
+
+	.get_pkru = vmx_get_pkru,
+
 	.fpu_activate = vmx_fpu_activate,
 	.fpu_deactivate = vmx_fpu_deactivate,
 

arch/x86/kvm/x86.c

@@ -723,7 +723,7 @@ int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
 {
 	unsigned long old_cr4 = kvm_read_cr4(vcpu);
 	unsigned long pdptr_bits = X86_CR4_PGE | X86_CR4_PSE | X86_CR4_PAE |
-				   X86_CR4_SMEP | X86_CR4_SMAP;
+				   X86_CR4_SMEP | X86_CR4_SMAP | X86_CR4_PKE;
 
 	if (cr4 & CR4_RESERVED_BITS)
 		return 1;
@@ -740,6 +740,9 @@ int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
 	if (!guest_cpuid_has_fsgsbase(vcpu) && (cr4 & X86_CR4_FSGSBASE))
 		return 1;
 
+	if (!guest_cpuid_has_pku(vcpu) && (cr4 & X86_CR4_PKE))
+		return 1;
+
 	if (is_long_mode(vcpu)) {
 		if (!(cr4 & X86_CR4_PAE))
 			return 1;
@@ -765,7 +768,7 @@ int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)
 	    (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE)))
 		kvm_mmu_reset_context(vcpu);
 
-	if ((cr4 ^ old_cr4) & X86_CR4_OSXSAVE)
+	if ((cr4 ^ old_cr4) & (X86_CR4_OSXSAVE | X86_CR4_PKE))
 		kvm_update_cpuid(vcpu);
 
 	return 0;
@@ -4326,9 +4329,14 @@ static int vcpu_mmio_gva_to_gpa(struct kvm_vcpu *vcpu, unsigned long gva,
 	u32 access = ((kvm_x86_ops->get_cpl(vcpu) == 3) ? PFERR_USER_MASK : 0)
 		| (write ? PFERR_WRITE_MASK : 0);
 
+	/*
+	 * currently PKRU is only applied to ept enabled guest so
+	 * there is no pkey in EPT page table for L1 guest or EPT
+	 * shadow page table for L2 guest.
+	 */
 	if (vcpu_match_mmio_gva(vcpu, gva)
 	    && !permission_fault(vcpu, vcpu->arch.walk_mmu,
-				 vcpu->arch.access, access)) {
+				 vcpu->arch.access, 0, access)) {
 		*gpa = vcpu->arch.mmio_gfn << PAGE_SHIFT |
 					(gva & (PAGE_SIZE - 1));
 		trace_vcpu_match_mmio(gva, *gpa, write, false);
@@ -6588,8 +6596,12 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 
 	srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx);
 
-	/* We should set ->mode before check ->requests,
-	 * see the comment in make_all_cpus_request.
+	/*
+	 * We should set ->mode before check ->requests,
+	 * Please see the comment in kvm_make_all_cpus_request.
+	 * This also orders the write to mode from any reads
+	 * to the page tables done while the VCPU is running.
+	 * Please see the comment in kvm_flush_remote_tlbs.
 	 */
 	smp_mb__after_srcu_read_unlock();
 
@@ -7123,7 +7135,7 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 
 	mmu_reset_needed |= kvm_read_cr4(vcpu) != sregs->cr4;
 	kvm_x86_ops->set_cr4(vcpu, sregs->cr4);
-	if (sregs->cr4 & X86_CR4_OSXSAVE)
+	if (sregs->cr4 & (X86_CR4_OSXSAVE | X86_CR4_PKE))
 		kvm_update_cpuid(vcpu);
 
 	idx = srcu_read_lock(&vcpu->kvm->srcu);

arch/x86/kvm/x86.h

@@ -183,7 +183,8 @@ bool kvm_vector_hashing_enabled(void);
 
 #define KVM_SUPPORTED_XCR0     (XFEATURE_MASK_FP | XFEATURE_MASK_SSE \
 				| XFEATURE_MASK_YMM | XFEATURE_MASK_BNDREGS \
-				| XFEATURE_MASK_BNDCSR | XFEATURE_MASK_AVX512)
+				| XFEATURE_MASK_BNDCSR | XFEATURE_MASK_AVX512 \
+				| XFEATURE_MASK_PKRU)
 extern u64 host_xcr0;
 
 extern u64 kvm_supported_xcr0(void);

virt/kvm/kvm_main.c

@@ -170,8 +170,8 @@ bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req)
 		kvm_make_request(req, vcpu);
 		cpu = vcpu->cpu;
 
-		/* Set ->requests bit before we read ->mode */
-		smp_mb();
+		/* Set ->requests bit before we read ->mode. */
+		smp_mb__after_atomic();
 
 		if (cpus != NULL && cpu != -1 && cpu != me &&
 		      kvm_vcpu_exiting_guest_mode(vcpu) != OUTSIDE_GUEST_MODE)
@@ -191,9 +191,23 @@ bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req)
 #ifndef CONFIG_HAVE_KVM_ARCH_TLB_FLUSH_ALL
 void kvm_flush_remote_tlbs(struct kvm *kvm)
 {
-	long dirty_count = kvm->tlbs_dirty;
+	/*
+	 * Read tlbs_dirty before setting KVM_REQ_TLB_FLUSH in
+	 * kvm_make_all_cpus_request.
+	 */
+	long dirty_count = smp_load_acquire(&kvm->tlbs_dirty);
 
-	smp_mb();
+	/*
+	 * We want to publish modifications to the page tables before reading
+	 * mode. Pairs with a memory barrier in arch-specific code.
+	 * - x86: smp_mb__after_srcu_read_unlock in vcpu_enter_guest
+	 * and smp_mb in walk_shadow_page_lockless_begin/end.
+	 * - powerpc: smp_mb in kvmppc_prepare_to_enter.
+	 *
+	 * There is already an smp_mb__after_atomic() before
+	 * kvm_make_all_cpus_request() reads vcpu->mode. We reuse that
+	 * barrier here.
+	 */
 	if (kvm_make_all_cpus_request(kvm, KVM_REQ_TLB_FLUSH))
 		++kvm->stat.remote_tlb_flush;
 	cmpxchg(&kvm->tlbs_dirty, dirty_count, 0);
@@ -536,6 +550,16 @@ static struct kvm *kvm_create_vm(unsigned long type)
 	if (!kvm)
 		return ERR_PTR(-ENOMEM);
 
+	spin_lock_init(&kvm->mmu_lock);
+	atomic_inc(&current->mm->mm_count);
+	kvm->mm = current->mm;
+	kvm_eventfd_init(kvm);
+	mutex_init(&kvm->lock);
+	mutex_init(&kvm->irq_lock);
+	mutex_init(&kvm->slots_lock);
+	atomic_set(&kvm->users_count, 1);
+	INIT_LIST_HEAD(&kvm->devices);
+
 	r = kvm_arch_init_vm(kvm, type);
 	if (r)
 		goto out_err_no_disable;
@@ -568,16 +592,6 @@ static struct kvm *kvm_create_vm(unsigned long type)
 			goto out_err;
 	}
 
-	spin_lock_init(&kvm->mmu_lock);
-	kvm->mm = current->mm;
-	atomic_inc(&kvm->mm->mm_count);
-	kvm_eventfd_init(kvm);
-	mutex_init(&kvm->lock);
-	mutex_init(&kvm->irq_lock);
-	mutex_init(&kvm->slots_lock);
-	atomic_set(&kvm->users_count, 1);
-	INIT_LIST_HEAD(&kvm->devices);
-
 	r = kvm_init_mmu_notifier(kvm);
 	if (r)
 		goto out_err;
@@ -602,6 +616,7 @@ static struct kvm *kvm_create_vm(unsigned long type)
 	for (i = 0; i < KVM_ADDRESS_SPACE_NUM; i++)
 		kvm_free_memslots(kvm, kvm->memslots[i]);
 	kvm_arch_free_vm(kvm);
+	mmdrop(current->mm);
 	return ERR_PTR(r);
 }