1 files changed, 173 insertions, 54 deletions

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index d6464045d3b9..c6ae9c9e360d 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -69,6 +69,21 @@ DEFINE_SPINLOCK(hugetlb_lock);
 static int num_fault_mutexes;
 struct mutex *hugetlb_fault_mutex_table ____cacheline_aligned_in_smp;
 
+static inline bool PageHugeFreed(struct page *head)
+{
+	return page_private(head + 4) == -1UL;
+}
+
+static inline void SetPageHugeFreed(struct page *head)
+{
+	set_page_private(head + 4, -1UL);
+}
+
+static inline void ClearPageHugeFreed(struct page *head)
+{
+	set_page_private(head + 4, 0);
+}
+
 /* Forward declaration */
 static int hugetlb_acct_memory(struct hstate *h, long delta);
 
@@ -574,13 +589,20 @@ void hugetlb_fix_reserve_counts(struct inode *inode)
 {
 	struct hugepage_subpool *spool = subpool_inode(inode);
 	long rsv_adjust;
+	bool reserved = false;
 
 	rsv_adjust = hugepage_subpool_get_pages(spool, 1);
-	if (rsv_adjust) {
+	if (rsv_adjust > 0) {
 		struct hstate *h = hstate_inode(inode);
 
-		hugetlb_acct_memory(h, 1);
+		if (!hugetlb_acct_memory(h, 1))
+			reserved = true;
+	} else if (!rsv_adjust) {
+		reserved = true;
 	}
+
+	if (!reserved)
+		pr_warn("hugetlb: Huge Page Reserved count may go negative.\n");
 }
 
 /*
@@ -866,6 +888,7 @@ static void enqueue_huge_page(struct hstate *h, struct page *page)
 	list_move(&page->lru, &h->hugepage_freelists[nid]);
 	h->free_huge_pages++;
 	h->free_huge_pages_node[nid]++;
+	SetPageHugeFreed(page);
 }
 
 static struct page *dequeue_huge_page_node_exact(struct hstate *h, int nid)
@@ -883,6 +906,7 @@ static struct page *dequeue_huge_page_node_exact(struct hstate *h, int nid)
 		return NULL;
 	list_move(&page->lru, &h->hugepage_activelist);
 	set_page_refcounted(page);
+	ClearPageHugeFreed(page);
 	h->free_huge_pages--;
 	h->free_huge_pages_node[nid]--;
 	return page;
@@ -1191,14 +1215,16 @@ static inline int alloc_fresh_gigantic_page(struct hstate *h,
 static void update_and_free_page(struct hstate *h, struct page *page)
 {
 	int i;
+	struct page *subpage = page;
 
 	if (hstate_is_gigantic(h) && !gigantic_page_supported())
 		return;
 
 	h->nr_huge_pages--;
 	h->nr_huge_pages_node[page_to_nid(page)]--;
-	for (i = 0; i < pages_per_huge_page(h); i++) {
-		page[i].flags &= ~(1 << PG_locked | 1 << PG_error |
+	for (i = 0; i < pages_per_huge_page(h);
+	     i++, subpage = mem_map_next(subpage, page, i)) {
+		subpage->flags &= ~(1 << PG_locked | 1 << PG_error |
 				1 << PG_referenced | 1 << PG_dirty |
 				1 << PG_active | 1 << PG_private |
 				1 << PG_writeback);
@@ -1233,12 +1259,11 @@ struct hstate *size_to_hstate(unsigned long size)
  */
 bool page_huge_active(struct page *page)
 {
-	VM_BUG_ON_PAGE(!PageHuge(page), page);
-	return PageHead(page) && PagePrivate(&page[1]);
+	return PageHeadHuge(page) && PagePrivate(&page[1]);
 }
 
 /* never called for tail page */
-static void set_page_huge_active(struct page *page)
+void set_page_huge_active(struct page *page)
 {
 	VM_BUG_ON_PAGE(!PageHeadHuge(page), page);
 	SetPagePrivate(&page[1]);
@@ -1316,6 +1341,7 @@ static void prep_new_huge_page(struct hstate *h, struct page *page, int nid)
 	set_hugetlb_cgroup(page, NULL);
 	h->nr_huge_pages++;
 	h->nr_huge_pages_node[nid]++;
+	ClearPageHugeFreed(page);
 	spin_unlock(&hugetlb_lock);
 	put_page(page); /* free it into the hugepage allocator */
 }
@@ -1377,15 +1403,12 @@ int PageHeadHuge(struct page *page_head)
 	return get_compound_page_dtor(page_head) == free_huge_page;
 }
 
-pgoff_t __basepage_index(struct page *page)
+pgoff_t hugetlb_basepage_index(struct page *page)
 {
 	struct page *page_head = compound_head(page);
 	pgoff_t index = page_index(page_head);
 	unsigned long compound_idx;
 
-	if (!PageHuge(page_head))
-		return page_index(page);
-
 	if (compound_order(page_head) >= MAX_ORDER)
 		compound_idx = page_to_pfn(page) - page_to_pfn(page_head);
 	else
@@ -1479,6 +1502,7 @@ int dissolve_free_huge_page(struct page *page)
 {
 	int rc = 0;
 
+retry:
 	spin_lock(&hugetlb_lock);
 	if (PageHuge(page) && !page_count(page)) {
 		struct page *head = compound_head(page);
@@ -1488,6 +1512,26 @@ int dissolve_free_huge_page(struct page *page)
 			rc = -EBUSY;
 			goto out;
 		}
+
+		/*
+		 * We should make sure that the page is already on the free list
+		 * when it is dissolved.
+		 */
+		if (unlikely(!PageHugeFreed(head))) {
+			spin_unlock(&hugetlb_lock);
+			cond_resched();
+
+			/*
+			 * Theoretically, we should return -EBUSY when we
+			 * encounter this race. In fact, we have a chance
+			 * to successfully dissolve the page if we do a
+			 * retry. Because the race window is quite small.
+			 * If we seize this opportunity, it is an optimization
+			 * for increasing the success rate of dissolving page.
+			 */
+			goto retry;
+		}
+
 		/*
 		 * Move PageHWPoison flag from head page to the raw error page,
 		 * which makes any subpages rather than the error page reusable.
@@ -2604,8 +2648,10 @@ static int hugetlb_sysfs_add_hstate(struct hstate *h, struct kobject *parent,
 		return -ENOMEM;
 
 	retval = sysfs_create_group(hstate_kobjs[hi], hstate_attr_group);
-	if (retval)
+	if (retval) {
 		kobject_put(hstate_kobjs[hi]);
+		hstate_kobjs[hi] = NULL;
+	}
 
 	return retval;
 }
@@ -2911,6 +2957,22 @@ static unsigned int cpuset_mems_nr(unsigned int *array)
 }
 
 #ifdef CONFIG_SYSCTL
+static int proc_hugetlb_doulongvec_minmax(struct ctl_table *table, int write,
+					  void *buffer, size_t *length,
+					  loff_t *ppos, unsigned long *out)
+{
+	struct ctl_table dup_table;
+
+	/*
+	 * In order to avoid races with __do_proc_doulongvec_minmax(), we
+	 * can duplicate the @table and alter the duplicate of it.
+	 */
+	dup_table = *table;
+	dup_table.data = out;
+
+	return proc_doulongvec_minmax(&dup_table, write, buffer, length, ppos);
+}
+
 static int hugetlb_sysctl_handler_common(bool obey_mempolicy,
 			 struct ctl_table *table, int write,
 			 void __user *buffer, size_t *length, loff_t *ppos)
@@ -2922,9 +2984,8 @@ static int hugetlb_sysctl_handler_common(bool obey_mempolicy,
 	if (!hugepages_supported())
 		return -EOPNOTSUPP;
 
-	table->data = &tmp;
-	table->maxlen = sizeof(unsigned long);
-	ret = proc_doulongvec_minmax(table, write, buffer, length, ppos);
+	ret = proc_hugetlb_doulongvec_minmax(table, write, buffer, length, ppos,
+					     &tmp);
 	if (ret)
 		goto out;
 
@@ -2968,9 +3029,8 @@ int hugetlb_overcommit_handler(struct ctl_table *table, int write,
 	if (write && hstate_is_gigantic(h))
 		return -EINVAL;
 
-	table->data = &tmp;
-	table->maxlen = sizeof(unsigned long);
-	ret = proc_doulongvec_minmax(table, write, buffer, length, ppos);
+	ret = proc_hugetlb_doulongvec_minmax(table, write, buffer, length, ppos,
+					     &tmp);
 	if (ret)
 		goto out;
 
@@ -3324,8 +3384,9 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	struct page *page;
 	struct hstate *h = hstate_vma(vma);
 	unsigned long sz = huge_page_size(h);
-	const unsigned long mmun_start = start;	/* For mmu_notifiers */
-	const unsigned long mmun_end   = end;	/* For mmu_notifiers */
+	unsigned long mmun_start = start;	/* For mmu_notifiers */
+	unsigned long mmun_end   = end;		/* For mmu_notifiers */
+	bool force_flush = false;
 
 	WARN_ON(!is_vm_hugetlb_page(vma));
 	BUG_ON(start & ~huge_page_mask(h));
@@ -3337,6 +3398,11 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	 */
 	tlb_remove_check_page_size_change(tlb, sz);
 	tlb_start_vma(tlb, vma);
+
+	/*
+	 * If sharing possible, alert mmu notifiers of worst case.
+	 */
+	adjust_range_if_pmd_sharing_possible(vma, &mmun_start, &mmun_end);
 	mmu_notifier_invalidate_range_start(mm, mmun_start, mmun_end);
 	address = start;
 	for (; address < end; address += sz) {
@@ -3347,6 +3413,8 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 		ptl = huge_pte_lock(h, mm, ptep);
 		if (huge_pmd_unshare(mm, &address, ptep)) {
 			spin_unlock(ptl);
+			tlb_flush_pmd_range(tlb, address & PUD_MASK, PUD_SIZE);
+			force_flush = true;
 			continue;
 		}
 
@@ -3403,6 +3471,22 @@ void __unmap_hugepage_range(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	}
 	mmu_notifier_invalidate_range_end(mm, mmun_start, mmun_end);
 	tlb_end_vma(tlb, vma);
+
+	/*
+	 * If we unshared PMDs, the TLB flush was not recorded in mmu_gather. We
+	 * could defer the flush until now, since by holding i_mmap_rwsem we
+	 * guaranteed that the last refernece would not be dropped. But we must
+	 * do the flushing before we return, as otherwise i_mmap_rwsem will be
+	 * dropped and the last reference to the shared PMDs page might be
+	 * dropped as well.
+	 *
+	 * In theory we could defer the freeing of the PMD pages as well, but
+	 * huge_pmd_unshare() relies on the exact page_count for the PMD page to
+	 * detect sharing, so we cannot defer the release of the page either.
+	 * Instead, do flush now.
+	 */
+	if (force_flush)
+		tlb_flush_mmu(tlb);
 }
 
 void __unmap_hugepage_range_final(struct mmu_gather *tlb,
@@ -3429,12 +3513,23 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start,
 {
 	struct mm_struct *mm;
 	struct mmu_gather tlb;
+	unsigned long tlb_start = start;
+	unsigned long tlb_end = end;
+
+	/*
+	 * If shared PMDs were possibly used within this vma range, adjust
+	 * start/end for worst case tlb flushing.
+	 * Note that we can not be sure if PMDs are shared until we try to
+	 * unmap pages.  However, we want to make sure TLB flushing covers
+	 * the largest possible range.
+	 */
+	adjust_range_if_pmd_sharing_possible(vma, &tlb_start, &tlb_end);
 
 	mm = vma->vm_mm;
 
-	tlb_gather_mmu(&tlb, mm, start, end);
+	tlb_gather_mmu(&tlb, mm, tlb_start, tlb_end);
 	__unmap_hugepage_range(&tlb, vma, start, end, ref_page);
-	tlb_finish_mmu(&tlb, start, end);
+	tlb_finish_mmu(&tlb, tlb_start, tlb_end);
 }
 
 /*
@@ -3739,8 +3834,7 @@ retry:
 			 * handling userfault.  Reacquire after handling
 			 * fault to make calling code simpler.
 			 */
-			hash = hugetlb_fault_mutex_hash(h, mapping, idx,
-							address);
+			hash = hugetlb_fault_mutex_hash(h, mapping, idx);
 			mutex_unlock(&hugetlb_fault_mutex_table[hash]);
 			ret = handle_userfault(&vmf, VM_UFFD_MISSING);
 			mutex_lock(&hugetlb_fault_mutex_table[hash]);
@@ -3783,7 +3877,7 @@ retry:
 		 * So we need to block hugepage fault by PG_hwpoison bit check.
 		 */
 		if (unlikely(PageHWPoison(page))) {
-			ret = VM_FAULT_HWPOISON |
+			ret = VM_FAULT_HWPOISON_LARGE |
 				VM_FAULT_SET_HINDEX(hstate_index(h));
 			goto backout_unlocked;
 		}
@@ -3853,7 +3947,7 @@ backout_unlocked:
 
 #ifdef CONFIG_SMP
 u32 hugetlb_fault_mutex_hash(struct hstate *h, struct address_space *mapping,
-			    pgoff_t idx, unsigned long address)
+			    pgoff_t idx)
 {
 	unsigned long key[2];
 	u32 hash;
@@ -3861,7 +3955,7 @@ u32 hugetlb_fault_mutex_hash(struct hstate *h, struct address_space *mapping,
 	key[0] = (unsigned long) mapping;
 	key[1] = idx;
 
-	hash = jhash2((u32 *)&key, sizeof(key)/sizeof(u32), 0);
+	hash = jhash2((u32 *)&key, sizeof(key)/(sizeof(u32)), 0);
 
 	return hash & (num_fault_mutexes - 1);
 }
@@ -3871,7 +3965,7 @@ u32 hugetlb_fault_mutex_hash(struct hstate *h, struct address_space *mapping,
  * return 0 and avoid the hashing overhead.
  */
 u32 hugetlb_fault_mutex_hash(struct hstate *h, struct address_space *mapping,
-			    pgoff_t idx, unsigned long address)
+			    pgoff_t idx)
 {
 	return 0;
 }
@@ -3916,7 +4010,7 @@ int hugetlb_fault(struct mm_struct *mm, struct vm_area_struct *vma,
 	 * get spurious allocation failures if two CPUs race to instantiate
 	 * the same page in the page cache.
 	 */
-	hash = hugetlb_fault_mutex_hash(h, mapping, idx, address);
+	hash = hugetlb_fault_mutex_hash(h, mapping, idx);
 	mutex_lock(&hugetlb_fault_mutex_table[hash]);
 
 	entry = huge_ptep_get(ptep);
@@ -4037,10 +4131,20 @@ int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm,
 	struct page *page;
 
 	if (!*pagep) {
-		ret = -ENOMEM;
+		/* If a page already exists, then it's UFFDIO_COPY for
+		 * a non-missing case. Return -EEXIST.
+		 */
+		if (vm_shared &&
+		    hugetlbfs_pagecache_present(h, dst_vma, dst_addr)) {
+			ret = -EEXIST;
+			goto out;
+		}
+
 		page = alloc_huge_page(dst_vma, dst_addr, 0);
-		if (IS_ERR(page))
+		if (IS_ERR(page)) {
+			ret = -ENOMEM;
 			goto out;
+		}
 
 		ret = copy_huge_page_from_user(page,
 						(const void __user *) src_addr,
@@ -4320,11 +4424,21 @@ unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
 	pte_t pte;
 	struct hstate *h = hstate_vma(vma);
 	unsigned long pages = 0;
+	unsigned long f_start = start;
+	unsigned long f_end = end;
+	bool shared_pmd = false;
+
+	/*
+	 * In the case of shared PMDs, the area to flush could be beyond
+	 * start/end.  Set f_start/f_end to cover the maximum possible
+	 * range if PMD sharing is possible.
+	 */
+	adjust_range_if_pmd_sharing_possible(vma, &f_start, &f_end);
 
 	BUG_ON(address >= end);
-	flush_cache_range(vma, address, end);
+	flush_cache_range(vma, f_start, f_end);
 
-	mmu_notifier_invalidate_range_start(mm, start, end);
+	mmu_notifier_invalidate_range_start(mm, f_start, f_end);
 	i_mmap_lock_write(vma->vm_file->f_mapping);
 	for (; address < end; address += huge_page_size(h)) {
 		spinlock_t *ptl;
@@ -4335,6 +4449,7 @@ unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
 		if (huge_pmd_unshare(mm, &address, ptep)) {
 			pages++;
 			spin_unlock(ptl);
+			shared_pmd = true;
 			continue;
 		}
 		pte = huge_ptep_get(ptep);
@@ -4370,12 +4485,18 @@ unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
 	 * Must flush TLB before releasing i_mmap_rwsem: x86's huge_pmd_unshare
 	 * may have cleared our pud entry and done put_page on the page table:
 	 * once we release i_mmap_rwsem, another task can do the final put_page
-	 * and that page table be reused and filled with junk.
+	 * and that page table be reused and filled with junk.  If we actually
+	 * did unshare a page of pmds, flush the range corresponding to the pud.
 	 */
-	flush_hugetlb_tlb_range(vma, start, end);
-	mmu_notifier_invalidate_range(mm, start, end);
+	if (shared_pmd) {
+		flush_hugetlb_tlb_range(vma, f_start, f_end);
+		mmu_notifier_invalidate_range(mm, f_start, f_end);
+	} else {
+		flush_hugetlb_tlb_range(vma, start, end);
+		mmu_notifier_invalidate_range(mm, start, end);
+	}
 	i_mmap_unlock_write(vma->vm_file->f_mapping);
-	mmu_notifier_invalidate_range_end(mm, start, end);
+	mmu_notifier_invalidate_range_end(mm, f_start, f_end);
 
 	return pages << h->order;
 }
@@ -4575,25 +4696,23 @@ static bool vma_shareable(struct vm_area_struct *vma, unsigned long addr)
 void adjust_range_if_pmd_sharing_possible(struct vm_area_struct *vma,
 				unsigned long *start, unsigned long *end)
 {
-	unsigned long check_addr = *start;
+	unsigned long v_start = ALIGN(vma->vm_start, PUD_SIZE),
+		v_end = ALIGN_DOWN(vma->vm_end, PUD_SIZE);
 
-	if (!(vma->vm_flags & VM_MAYSHARE))
+	/*
+	 * vma need span at least one aligned PUD size and the start,end range
+	 * must at least partialy within it.
+	 */
+	if (!(vma->vm_flags & VM_MAYSHARE) || !(v_end > v_start) ||
+		(*end <= v_start) || (*start >= v_end))
 		return;
 
-	for (check_addr = *start; check_addr < *end; check_addr += PUD_SIZE) {
-		unsigned long a_start = check_addr & PUD_MASK;
-		unsigned long a_end = a_start + PUD_SIZE;
+	/* Extend the range to be PUD aligned for a worst case scenario */
+	if (*start > v_start)
+		*start = ALIGN_DOWN(*start, PUD_SIZE);
 
-		/*
-		 * If sharing is possible, adjust start/end if necessary.
-		 */
-		if (range_in_vma(vma, a_start, a_end)) {
-			if (a_start < *start)
-				*start = a_start;
-			if (a_end > *end)
-				*end = a_end;
-		}
-	}
+	if (*end < v_end)
+		*end = ALIGN(*end, PUD_SIZE);
 }
 
 /*
@@ -4855,9 +4974,9 @@ bool isolate_huge_page(struct page *page, struct list_head *list)
 {
 	bool ret = true;
 
-	VM_BUG_ON_PAGE(!PageHead(page), page);
 	spin_lock(&hugetlb_lock);
-	if (!page_huge_active(page) || !get_page_unless_zero(page)) {
+	if (!PageHeadHuge(page) || !page_huge_active(page) ||
+	    !get_page_unless_zero(page)) {
 		ret = false;
 		goto unlock;
 	}