1 files changed, 149 insertions, 101 deletions

diff --git a/virt/kvm/arm/mmu.c b/virt/kvm/arm/mmu.c
index 1d90d79706bd..5eca48bdb1a6 100644
--- a/virt/kvm/arm/mmu.c
+++ b/virt/kvm/arm/mmu.c
@@ -45,7 +45,6 @@ static phys_addr_t hyp_idmap_vector;
 
 static unsigned long io_map_base;
 
-#define S2_PGD_SIZE	(PTRS_PER_S2_PGD * sizeof(pgd_t))
 #define hyp_pgd_order get_order(PTRS_PER_PGD * sizeof(pgd_t))
 
 #define KVM_S2PTE_FLAG_IS_IOMAP		(1UL << 0)
@@ -150,20 +149,20 @@ static void *mmu_memory_cache_alloc(struct kvm_mmu_memory_cache *mc)
 
 static void clear_stage2_pgd_entry(struct kvm *kvm, pgd_t *pgd, phys_addr_t addr)
 {
-	pud_t *pud_table __maybe_unused = stage2_pud_offset(pgd, 0UL);
-	stage2_pgd_clear(pgd);
+	pud_t *pud_table __maybe_unused = stage2_pud_offset(kvm, pgd, 0UL);
+	stage2_pgd_clear(kvm, pgd);
 	kvm_tlb_flush_vmid_ipa(kvm, addr);
-	stage2_pud_free(pud_table);
+	stage2_pud_free(kvm, pud_table);
 	put_page(virt_to_page(pgd));
 }
 
 static void clear_stage2_pud_entry(struct kvm *kvm, pud_t *pud, phys_addr_t addr)
 {
-	pmd_t *pmd_table __maybe_unused = stage2_pmd_offset(pud, 0);
-	VM_BUG_ON(stage2_pud_huge(*pud));
-	stage2_pud_clear(pud);
+	pmd_t *pmd_table __maybe_unused = stage2_pmd_offset(kvm, pud, 0);
+	VM_BUG_ON(stage2_pud_huge(kvm, *pud));
+	stage2_pud_clear(kvm, pud);
 	kvm_tlb_flush_vmid_ipa(kvm, addr);
-	stage2_pmd_free(pmd_table);
+	stage2_pmd_free(kvm, pmd_table);
 	put_page(virt_to_page(pud));
 }
 
@@ -177,6 +176,35 @@ static void clear_stage2_pmd_entry(struct kvm *kvm, pmd_t *pmd, phys_addr_t addr
 	put_page(virt_to_page(pmd));
 }
 
+static inline void kvm_set_pte(pte_t *ptep, pte_t new_pte)
+{
+	WRITE_ONCE(*ptep, new_pte);
+	dsb(ishst);
+}
+
+static inline void kvm_set_pmd(pmd_t *pmdp, pmd_t new_pmd)
+{
+	WRITE_ONCE(*pmdp, new_pmd);
+	dsb(ishst);
+}
+
+static inline void kvm_pmd_populate(pmd_t *pmdp, pte_t *ptep)
+{
+	kvm_set_pmd(pmdp, kvm_mk_pmd(ptep));
+}
+
+static inline void kvm_pud_populate(pud_t *pudp, pmd_t *pmdp)
+{
+	WRITE_ONCE(*pudp, kvm_mk_pud(pmdp));
+	dsb(ishst);
+}
+
+static inline void kvm_pgd_populate(pgd_t *pgdp, pud_t *pudp)
+{
+	WRITE_ONCE(*pgdp, kvm_mk_pgd(pudp));
+	dsb(ishst);
+}
+
 /*
  * Unmapping vs dcache management:
  *
@@ -196,6 +224,10 @@ static void clear_stage2_pmd_entry(struct kvm *kvm, pmd_t *pmd, phys_addr_t addr
  * This is why right after unmapping a page/section and invalidating
  * the corresponding TLBs, we call kvm_flush_dcache_p*() to make sure
  * the IO subsystem will never hit in the cache.
+ *
+ * This is all avoided on systems that have ARM64_HAS_STAGE2_FWB, as
+ * we then fully enforce cacheability of RAM, no matter what the guest
+ * does.
  */
 static void unmap_stage2_ptes(struct kvm *kvm, pmd_t *pmd,
 		       phys_addr_t addr, phys_addr_t end)
@@ -219,7 +251,7 @@ static void unmap_stage2_ptes(struct kvm *kvm, pmd_t *pmd,
 		}
 	} while (pte++, addr += PAGE_SIZE, addr != end);
 
-	if (stage2_pte_table_empty(start_pte))
+	if (stage2_pte_table_empty(kvm, start_pte))
 		clear_stage2_pmd_entry(kvm, pmd, start_addr);
 }
 
@@ -229,9 +261,9 @@ static void unmap_stage2_pmds(struct kvm *kvm, pud_t *pud,
 	phys_addr_t next, start_addr = addr;
 	pmd_t *pmd, *start_pmd;
 
-	start_pmd = pmd = stage2_pmd_offset(pud, addr);
+	start_pmd = pmd = stage2_pmd_offset(kvm, pud, addr);
 	do {
-		next = stage2_pmd_addr_end(addr, end);
+		next = stage2_pmd_addr_end(kvm, addr, end);
 		if (!pmd_none(*pmd)) {
 			if (pmd_thp_or_huge(*pmd)) {
 				pmd_t old_pmd = *pmd;
@@ -248,7 +280,7 @@ static void unmap_stage2_pmds(struct kvm *kvm, pud_t *pud,
 		}
 	} while (pmd++, addr = next, addr != end);
 
-	if (stage2_pmd_table_empty(start_pmd))
+	if (stage2_pmd_table_empty(kvm, start_pmd))
 		clear_stage2_pud_entry(kvm, pud, start_addr);
 }
 
@@ -258,14 +290,14 @@ static void unmap_stage2_puds(struct kvm *kvm, pgd_t *pgd,
 	phys_addr_t next, start_addr = addr;
 	pud_t *pud, *start_pud;
 
-	start_pud = pud = stage2_pud_offset(pgd, addr);
+	start_pud = pud = stage2_pud_offset(kvm, pgd, addr);
 	do {
-		next = stage2_pud_addr_end(addr, end);
-		if (!stage2_pud_none(*pud)) {
-			if (stage2_pud_huge(*pud)) {
+		next = stage2_pud_addr_end(kvm, addr, end);
+		if (!stage2_pud_none(kvm, *pud)) {
+			if (stage2_pud_huge(kvm, *pud)) {
 				pud_t old_pud = *pud;
 
-				stage2_pud_clear(pud);
+				stage2_pud_clear(kvm, pud);
 				kvm_tlb_flush_vmid_ipa(kvm, addr);
 				kvm_flush_dcache_pud(old_pud);
 				put_page(virt_to_page(pud));
@@ -275,7 +307,7 @@ static void unmap_stage2_puds(struct kvm *kvm, pgd_t *pgd,
 		}
 	} while (pud++, addr = next, addr != end);
 
-	if (stage2_pud_table_empty(start_pud))
+	if (stage2_pud_table_empty(kvm, start_pud))
 		clear_stage2_pgd_entry(kvm, pgd, start_addr);
 }
 
@@ -299,7 +331,7 @@ static void unmap_stage2_range(struct kvm *kvm, phys_addr_t start, u64 size)
 	assert_spin_locked(&kvm->mmu_lock);
 	WARN_ON(size & ~PAGE_MASK);
 
-	pgd = kvm->arch.pgd + stage2_pgd_index(addr);
+	pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
 	do {
 		/*
 		 * Make sure the page table is still active, as another thread
@@ -308,8 +340,8 @@ static void unmap_stage2_range(struct kvm *kvm, phys_addr_t start, u64 size)
 		 */
 		if (!READ_ONCE(kvm->arch.pgd))
 			break;
-		next = stage2_pgd_addr_end(addr, end);
-		if (!stage2_pgd_none(*pgd))
+		next = stage2_pgd_addr_end(kvm, addr, end);
+		if (!stage2_pgd_none(kvm, *pgd))
 			unmap_stage2_puds(kvm, pgd, addr, next);
 		/*
 		 * If the range is too large, release the kvm->mmu_lock
@@ -338,9 +370,9 @@ static void stage2_flush_pmds(struct kvm *kvm, pud_t *pud,
 	pmd_t *pmd;
 	phys_addr_t next;
 
-	pmd = stage2_pmd_offset(pud, addr);
+	pmd = stage2_pmd_offset(kvm, pud, addr);
 	do {
-		next = stage2_pmd_addr_end(addr, end);
+		next = stage2_pmd_addr_end(kvm, addr, end);
 		if (!pmd_none(*pmd)) {
 			if (pmd_thp_or_huge(*pmd))
 				kvm_flush_dcache_pmd(*pmd);
@@ -356,11 +388,11 @@ static void stage2_flush_puds(struct kvm *kvm, pgd_t *pgd,
 	pud_t *pud;
 	phys_addr_t next;
 
-	pud = stage2_pud_offset(pgd, addr);
+	pud = stage2_pud_offset(kvm, pgd, addr);
 	do {
-		next = stage2_pud_addr_end(addr, end);
-		if (!stage2_pud_none(*pud)) {
-			if (stage2_pud_huge(*pud))
+		next = stage2_pud_addr_end(kvm, addr, end);
+		if (!stage2_pud_none(kvm, *pud)) {
+			if (stage2_pud_huge(kvm, *pud))
 				kvm_flush_dcache_pud(*pud);
 			else
 				stage2_flush_pmds(kvm, pud, addr, next);
@@ -376,10 +408,11 @@ static void stage2_flush_memslot(struct kvm *kvm,
 	phys_addr_t next;
 	pgd_t *pgd;
 
-	pgd = kvm->arch.pgd + stage2_pgd_index(addr);
+	pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
 	do {
-		next = stage2_pgd_addr_end(addr, end);
-		stage2_flush_puds(kvm, pgd, addr, next);
+		next = stage2_pgd_addr_end(kvm, addr, end);
+		if (!stage2_pgd_none(kvm, *pgd))
+			stage2_flush_puds(kvm, pgd, addr, next);
 	} while (pgd++, addr = next, addr != end);
 }
 
@@ -576,7 +609,6 @@ static void create_hyp_pte_mappings(pmd_t *pmd, unsigned long start,
 		pte = pte_offset_kernel(pmd, addr);
 		kvm_set_pte(pte, pfn_pte(pfn, prot));
 		get_page(virt_to_page(pte));
-		kvm_flush_dcache_to_poc(pte, sizeof(*pte));
 		pfn++;
 	} while (addr += PAGE_SIZE, addr != end);
 }
@@ -601,9 +633,8 @@ static int create_hyp_pmd_mappings(pud_t *pud, unsigned long start,
 				kvm_err("Cannot allocate Hyp pte\n");
 				return -ENOMEM;
 			}
-			pmd_populate_kernel(NULL, pmd, pte);
+			kvm_pmd_populate(pmd, pte);
 			get_page(virt_to_page(pmd));
-			kvm_flush_dcache_to_poc(pmd, sizeof(*pmd));
 		}
 
 		next = pmd_addr_end(addr, end);
@@ -634,9 +665,8 @@ static int create_hyp_pud_mappings(pgd_t *pgd, unsigned long start,
 				kvm_err("Cannot allocate Hyp pmd\n");
 				return -ENOMEM;
 			}
-			pud_populate(NULL, pud, pmd);
+			kvm_pud_populate(pud, pmd);
 			get_page(virt_to_page(pud));
-			kvm_flush_dcache_to_poc(pud, sizeof(*pud));
 		}
 
 		next = pud_addr_end(addr, end);
@@ -671,9 +701,8 @@ static int __create_hyp_mappings(pgd_t *pgdp, unsigned long ptrs_per_pgd,
 				err = -ENOMEM;
 				goto out;
 			}
-			pgd_populate(NULL, pgd, pud);
+			kvm_pgd_populate(pgd, pud);
 			get_page(virt_to_page(pgd));
-			kvm_flush_dcache_to_poc(pgd, sizeof(*pgd));
 		}
 
 		next = pgd_addr_end(addr, end);
@@ -868,7 +897,7 @@ int kvm_alloc_stage2_pgd(struct kvm *kvm)
 	}
 
 	/* Allocate the HW PGD, making sure that each page gets its own refcount */
-	pgd = alloc_pages_exact(S2_PGD_SIZE, GFP_KERNEL | __GFP_ZERO);
+	pgd = alloc_pages_exact(stage2_pgd_size(kvm), GFP_KERNEL | __GFP_ZERO);
 	if (!pgd)
 		return -ENOMEM;
 
@@ -957,7 +986,7 @@ void kvm_free_stage2_pgd(struct kvm *kvm)
 
 	spin_lock(&kvm->mmu_lock);
 	if (kvm->arch.pgd) {
-		unmap_stage2_range(kvm, 0, KVM_PHYS_SIZE);
+		unmap_stage2_range(kvm, 0, kvm_phys_size(kvm));
 		pgd = READ_ONCE(kvm->arch.pgd);
 		kvm->arch.pgd = NULL;
 	}
@@ -965,7 +994,7 @@ void kvm_free_stage2_pgd(struct kvm *kvm)
 
 	/* Free the HW pgd, one page at a time */
 	if (pgd)
-		free_pages_exact(pgd, S2_PGD_SIZE);
+		free_pages_exact(pgd, stage2_pgd_size(kvm));
 }
 
 static pud_t *stage2_get_pud(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
@@ -974,16 +1003,16 @@ static pud_t *stage2_get_pud(struct kvm *kvm, struct kvm_mmu_memory_cache *cache
 	pgd_t *pgd;
 	pud_t *pud;
 
-	pgd = kvm->arch.pgd + stage2_pgd_index(addr);
-	if (WARN_ON(stage2_pgd_none(*pgd))) {
+	pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
+	if (stage2_pgd_none(kvm, *pgd)) {
 		if (!cache)
 			return NULL;
 		pud = mmu_memory_cache_alloc(cache);
-		stage2_pgd_populate(pgd, pud);
+		stage2_pgd_populate(kvm, pgd, pud);
 		get_page(virt_to_page(pgd));
 	}
 
-	return stage2_pud_offset(pgd, addr);
+	return stage2_pud_offset(kvm, pgd, addr);
 }
 
 static pmd_t *stage2_get_pmd(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
@@ -996,15 +1025,15 @@ static pmd_t *stage2_get_pmd(struct kvm *kvm, struct kvm_mmu_memory_cache *cache
 	if (!pud)
 		return NULL;
 
-	if (stage2_pud_none(*pud)) {
+	if (stage2_pud_none(kvm, *pud)) {
 		if (!cache)
 			return NULL;
 		pmd = mmu_memory_cache_alloc(cache);
-		stage2_pud_populate(pud, pmd);
+		stage2_pud_populate(kvm, pud, pmd);
 		get_page(virt_to_page(pud));
 	}
 
-	return stage2_pmd_offset(pud, addr);
+	return stage2_pmd_offset(kvm, pud, addr);
 }
 
 static int stage2_set_pmd_huge(struct kvm *kvm, struct kvm_mmu_memory_cache
@@ -1015,19 +1044,35 @@ static int stage2_set_pmd_huge(struct kvm *kvm, struct kvm_mmu_memory_cache
 	pmd = stage2_get_pmd(kvm, cache, addr);
 	VM_BUG_ON(!pmd);
 
-	/*
-	 * Mapping in huge pages should only happen through a fault.  If a
-	 * page is merged into a transparent huge page, the individual
-	 * subpages of that huge page should be unmapped through MMU
-	 * notifiers before we get here.
-	 *
-	 * Merging of CompoundPages is not supported; they should become
-	 * splitting first, unmapped, merged, and mapped back in on-demand.
-	 */
-	VM_BUG_ON(pmd_present(*pmd) && pmd_pfn(*pmd) != pmd_pfn(*new_pmd));
-
 	old_pmd = *pmd;
 	if (pmd_present(old_pmd)) {
+		/*
+		 * Multiple vcpus faulting on the same PMD entry, can
+		 * lead to them sequentially updating the PMD with the
+		 * same value. Following the break-before-make
+		 * (pmd_clear() followed by tlb_flush()) process can
+		 * hinder forward progress due to refaults generated
+		 * on missing translations.
+		 *
+		 * Skip updating the page table if the entry is
+		 * unchanged.
+		 */
+		if (pmd_val(old_pmd) == pmd_val(*new_pmd))
+			return 0;
+
+		/*
+		 * Mapping in huge pages should only happen through a
+		 * fault.  If a page is merged into a transparent huge
+		 * page, the individual subpages of that huge page
+		 * should be unmapped through MMU notifiers before we
+		 * get here.
+		 *
+		 * Merging of CompoundPages is not supported; they
+		 * should become splitting first, unmapped, merged,
+		 * and mapped back in on-demand.
+		 */
+		VM_BUG_ON(pmd_pfn(old_pmd) != pmd_pfn(*new_pmd));
+
 		pmd_clear(pmd);
 		kvm_tlb_flush_vmid_ipa(kvm, addr);
 	} else {
@@ -1090,7 +1135,7 @@ static int stage2_set_pte(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
 		if (!cache)
 			return 0; /* ignore calls from kvm_set_spte_hva */
 		pte = mmu_memory_cache_alloc(cache);
-		pmd_populate_kernel(NULL, pmd, pte);
+		kvm_pmd_populate(pmd, pte);
 		get_page(virt_to_page(pmd));
 	}
 
@@ -1102,6 +1147,10 @@ static int stage2_set_pte(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
 	/* Create 2nd stage page table mapping - Level 3 */
 	old_pte = *pte;
 	if (pte_present(old_pte)) {
+		/* Skip page table update if there is no change */
+		if (pte_val(old_pte) == pte_val(*new_pte))
+			return 0;
+
 		kvm_set_pte(pte, __pte(0));
 		kvm_tlb_flush_vmid_ipa(kvm, addr);
 	} else {
@@ -1158,8 +1207,9 @@ int kvm_phys_addr_ioremap(struct kvm *kvm, phys_addr_t guest_ipa,
 		if (writable)
 			pte = kvm_s2pte_mkwrite(pte);
 
-		ret = mmu_topup_memory_cache(&cache, KVM_MMU_CACHE_MIN_PAGES,
-						KVM_NR_MEM_OBJS);
+		ret = mmu_topup_memory_cache(&cache,
+					     kvm_mmu_cache_min_pages(kvm),
+					     KVM_NR_MEM_OBJS);
 		if (ret)
 			goto out;
 		spin_lock(&kvm->mmu_lock);
@@ -1181,8 +1231,14 @@ static bool transparent_hugepage_adjust(kvm_pfn_t *pfnp, phys_addr_t *ipap)
 {
 	kvm_pfn_t pfn = *pfnp;
 	gfn_t gfn = *ipap >> PAGE_SHIFT;
+	struct page *page = pfn_to_page(pfn);
 
-	if (PageTransCompoundMap(pfn_to_page(pfn))) {
+	/*
+	 * PageTransCompoungMap() returns true for THP and
+	 * hugetlbfs. Make sure the adjustment is done only for THP
+	 * pages.
+	 */
+	if (!PageHuge(page) && PageTransCompoundMap(page)) {
 		unsigned long mask;
 		/*
 		 * The address we faulted on is backed by a transparent huge
@@ -1247,19 +1303,21 @@ static void stage2_wp_ptes(pmd_t *pmd, phys_addr_t addr, phys_addr_t end)
 
 /**
  * stage2_wp_pmds - write protect PUD range
+ * kvm:		kvm instance for the VM
  * @pud:	pointer to pud entry
  * @addr:	range start address
  * @end:	range end address
  */
-static void stage2_wp_pmds(pud_t *pud, phys_addr_t addr, phys_addr_t end)
+static void stage2_wp_pmds(struct kvm *kvm, pud_t *pud,
+			   phys_addr_t addr, phys_addr_t end)
 {
 	pmd_t *pmd;
 	phys_addr_t next;
 
-	pmd = stage2_pmd_offset(pud, addr);
+	pmd = stage2_pmd_offset(kvm, pud, addr);
 
 	do {
-		next = stage2_pmd_addr_end(addr, end);
+		next = stage2_pmd_addr_end(kvm, addr, end);
 		if (!pmd_none(*pmd)) {
 			if (pmd_thp_or_huge(*pmd)) {
 				if (!kvm_s2pmd_readonly(pmd))
@@ -1279,18 +1337,19 @@ static void stage2_wp_pmds(pud_t *pud, phys_addr_t addr, phys_addr_t end)
   *
   * Process PUD entries, for a huge PUD we cause a panic.
   */
-static void  stage2_wp_puds(pgd_t *pgd, phys_addr_t addr, phys_addr_t end)
+static void  stage2_wp_puds(struct kvm *kvm, pgd_t *pgd,
+			    phys_addr_t addr, phys_addr_t end)
 {
 	pud_t *pud;
 	phys_addr_t next;
 
-	pud = stage2_pud_offset(pgd, addr);
+	pud = stage2_pud_offset(kvm, pgd, addr);
 	do {
-		next = stage2_pud_addr_end(addr, end);
-		if (!stage2_pud_none(*pud)) {
+		next = stage2_pud_addr_end(kvm, addr, end);
+		if (!stage2_pud_none(kvm, *pud)) {
 			/* TODO:PUD not supported, revisit later if supported */
-			BUG_ON(stage2_pud_huge(*pud));
-			stage2_wp_pmds(pud, addr, next);
+			BUG_ON(stage2_pud_huge(kvm, *pud));
+			stage2_wp_pmds(kvm, pud, addr, next);
 		}
 	} while (pud++, addr = next, addr != end);
 }
@@ -1306,7 +1365,7 @@ static void stage2_wp_range(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
 	pgd_t *pgd;
 	phys_addr_t next;
 
-	pgd = kvm->arch.pgd + stage2_pgd_index(addr);
+	pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
 	do {
 		/*
 		 * Release kvm_mmu_lock periodically if the memory region is
@@ -1320,9 +1379,9 @@ static void stage2_wp_range(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
 		cond_resched_lock(&kvm->mmu_lock);
 		if (!READ_ONCE(kvm->arch.pgd))
 			break;
-		next = stage2_pgd_addr_end(addr, end);
-		if (stage2_pgd_present(*pgd))
-			stage2_wp_puds(pgd, addr, next);
+		next = stage2_pgd_addr_end(kvm, addr, end);
+		if (stage2_pgd_present(kvm, *pgd))
+			stage2_wp_puds(kvm, pgd, addr, next);
 	} while (pgd++, addr = next, addr != end);
 }
 
@@ -1401,20 +1460,14 @@ static void invalidate_icache_guest_page(kvm_pfn_t pfn, unsigned long size)
 static void kvm_send_hwpoison_signal(unsigned long address,
 				     struct vm_area_struct *vma)
 {
-	siginfo_t info;
-
-	clear_siginfo(&info);
-	info.si_signo   = SIGBUS;
-	info.si_errno   = 0;
-	info.si_code    = BUS_MCEERR_AR;
-	info.si_addr    = (void __user *)address;
+	short lsb;
 
 	if (is_vm_hugetlb_page(vma))
-		info.si_addr_lsb = huge_page_shift(hstate_vma(vma));
+		lsb = huge_page_shift(hstate_vma(vma));
 	else
-		info.si_addr_lsb = PAGE_SHIFT;
+		lsb = PAGE_SHIFT;
 
-	send_sig_info(SIGBUS, &info, current);
+	send_sig_mceerr(BUS_MCEERR_AR, (void __user *)address, lsb, current);
 }
 
 static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
@@ -1471,7 +1524,7 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	up_read(&current->mm->mmap_sem);
 
 	/* We need minimum second+third level pages */
-	ret = mmu_topup_memory_cache(memcache, KVM_MMU_CACHE_MIN_PAGES,
+	ret = mmu_topup_memory_cache(memcache, kvm_mmu_cache_min_pages(kvm),
 				     KVM_NR_MEM_OBJS);
 	if (ret)
 		return ret;
@@ -1714,7 +1767,7 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	}
 
 	/* Userspace should not be able to register out-of-bounds IPAs */
-	VM_BUG_ON(fault_ipa >= KVM_PHYS_SIZE);
+	VM_BUG_ON(fault_ipa >= kvm_phys_size(vcpu->kvm));
 
 	if (fault_status == FSC_ACCESS) {
 		handle_access_fault(vcpu, fault_ipa);
@@ -1768,18 +1821,6 @@ static int kvm_unmap_hva_handler(struct kvm *kvm, gpa_t gpa, u64 size, void *dat
 	return 0;
 }
 
-int kvm_unmap_hva(struct kvm *kvm, unsigned long hva)
-{
-	unsigned long end = hva + PAGE_SIZE;
-
-	if (!kvm->arch.pgd)
-		return 0;
-
-	trace_kvm_unmap_hva(hva);
-	handle_hva_to_gpa(kvm, hva, end, &kvm_unmap_hva_handler, NULL);
-	return 0;
-}
-
 int kvm_unmap_hva_range(struct kvm *kvm,
 			unsigned long start, unsigned long end)
 {
@@ -1811,13 +1852,20 @@ static int kvm_set_spte_handler(struct kvm *kvm, gpa_t gpa, u64 size, void *data
 void kvm_set_spte_hva(struct kvm *kvm, unsigned long hva, pte_t pte)
 {
 	unsigned long end = hva + PAGE_SIZE;
+	kvm_pfn_t pfn = pte_pfn(pte);
 	pte_t stage2_pte;
 
 	if (!kvm->arch.pgd)
 		return;
 
 	trace_kvm_set_spte_hva(hva);
-	stage2_pte = pfn_pte(pte_pfn(pte), PAGE_S2);
+
+	/*
+	 * We've moved a page around, probably through CoW, so let's treat it
+	 * just like a translation fault and clean the cache to the PoC.
+	 */
+	clean_dcache_guest_page(pfn, PAGE_SIZE);
+	stage2_pte = pfn_pte(pfn, PAGE_S2);
 	handle_hva_to_gpa(kvm, hva, end, &kvm_set_spte_handler, &stage2_pte);
 }
 
@@ -2018,7 +2066,7 @@ int kvm_arch_prepare_memory_region(struct kvm *kvm,
 	 * space addressable by the KVM guest IPA space.
 	 */
 	if (memslot->base_gfn + memslot->npages >=
-	    (KVM_PHYS_SIZE >> PAGE_SHIFT))
+	    (kvm_phys_size(kvm) >> PAGE_SHIFT))
 		return -EFAULT;
 
 	down_read(&current->mm->mmap_sem);