[PATCH 3.16 104/306] mm/hugetlb: check for reserved hugepages during memory offline

[Ben Hutchings]

3.16.40-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Gerald Schaefer <gerald.schaefer@xxxxxxxxxx>

commit 082d5b6b60e9f25e1511557fcfcb21eedd267446 upstream.

In dissolve_free_huge_pages(), free hugepages will be dissolved without
making sure that there are enough of them left to satisfy hugepage
reservations.

Fix this by adding a return value to dissolve_free_huge_pages() and
checking h->free_huge_pages vs.  h->resv_huge_pages.  Note that this may
lead to the situation where dissolve_free_huge_page() returns an error
and all free hugepages that were dissolved before that error are lost,
while the memory block still cannot be set offline.

Fixes: c8721bbb ("mm: memory-hotplug: enable memory hotplug to handle hugepage")
Link: http://lkml.kernel.org/r/20160926172811.94033-3-gerald.schaefer@xxxxxxxxxx
Signed-off-by: Gerald Schaefer <gerald.schaefer@xxxxxxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxx>
Acked-by: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Cc: "Kirill A . Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx>
Cc: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Mike Kravetz <mike.kravetz@xxxxxxxxxx>
Cc: "Aneesh Kumar K . V" <aneesh.kumar@xxxxxxxxxxxxxxxxxx>
Cc: Martin Schwidefsky <schwidefsky@xxxxxxxxxx>
Cc: Heiko Carstens <heiko.carstens@xxxxxxxxxx>
Cc: Rui Teng <rui.teng@xxxxxxxxxxxxxxxxxx>
Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
---
 include/linux/hugetlb.h |  6 +++---
 mm/hugetlb.c            | 26 +++++++++++++++++++++-----
 mm/memory_hotplug.c     |  4 +++-
 3 files changed, 27 insertions(+), 9 deletions(-)

--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -396,8 +396,8 @@ static inline pgoff_t basepage_index(str
 	return __basepage_index(page);
 }
 
-extern void dissolve_free_huge_pages(unsigned long start_pfn,
-				     unsigned long end_pfn);
+extern int dissolve_free_huge_pages(unsigned long start_pfn,
+				    unsigned long end_pfn);
 static inline int hugepage_migration_supported(struct hstate *h)
 {
 #ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION
@@ -452,7 +452,7 @@ static inline pgoff_t basepage_index(str
 {
 	return page->index;
 }
-#define dissolve_free_huge_pages(s, e)	do {} while (0)
+#define dissolve_free_huge_pages(s, e)	0
 #define hugepage_migration_supported(h)	0
 
 static inline spinlock_t *huge_pte_lockptr(struct hstate *h,
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1067,21 +1067,31 @@ static int free_pool_huge_page(struct hs
 
 /*
  * Dissolve a given free hugepage into free buddy pages. This function does
- * nothing for in-use (including surplus) hugepages.
+ * nothing for in-use (including surplus) hugepages. Returns -EBUSY if the
+ * number of free hugepages would be reduced below the number of reserved
+ * hugepages.
  */
-static void dissolve_free_huge_page(struct page *page)
+static int dissolve_free_huge_page(struct page *page)
 {
+	int rc = 0;
+
 	spin_lock(&hugetlb_lock);
 	if (PageHuge(page) && !page_count(page)) {
 		struct page *head = compound_head(page);
 		struct hstate *h = page_hstate(head);
 		int nid = page_to_nid(head);
+		if (h->free_huge_pages - h->resv_huge_pages == 0) {
+			rc = -EBUSY;
+			goto out;
+		}
 		list_del(&head->lru);
 		h->free_huge_pages--;
 		h->free_huge_pages_node[nid]--;
 		update_and_free_page(h, head);
 	}
+out:
 	spin_unlock(&hugetlb_lock);
+	return rc;
 }
 
 /*
@@ -1089,16 +1099,22 @@ static void dissolve_free_huge_page(stru
  * make specified memory blocks removable from the system.
  * Note that this will dissolve a free gigantic hugepage completely, if any
  * part of it lies within the given range.
+ * Also note that if dissolve_free_huge_page() returns with an error, all
+ * free hugepages that were dissolved before that error are lost.
  */
-void dissolve_free_huge_pages(unsigned long start_pfn, unsigned long end_pfn)
+int dissolve_free_huge_pages(unsigned long start_pfn, unsigned long end_pfn)
 {
 	unsigned long pfn;
+	int rc = 0;
 
 	if (!hugepages_supported())
-		return;
+		return rc;
 
 	for (pfn = start_pfn; pfn < end_pfn; pfn += 1 << minimum_order)
-		dissolve_free_huge_page(pfn_to_page(pfn));
+		if (rc = dissolve_free_huge_page(pfn_to_page(pfn)))
+			break;
+
+	return rc;
 }
 
 static struct page *alloc_buddy_huge_page(struct hstate *h, int nid)
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -1732,7 +1732,9 @@ repeat:
 	 * dissolve free hugepages in the memory block before doing offlining
 	 * actually in order to make hugetlbfs's object counting consistent.
 	 */
-	dissolve_free_huge_pages(start_pfn, end_pfn);
+	ret = dissolve_free_huge_pages(start_pfn, end_pfn);
+	if (ret)
+		goto failed_removal;
 	/* check again */
 	offlined_pages = check_pages_isolated(start_pfn, end_pfn);
 	if (offlined_pages < 0) {