mm, THP, swap: support splitting THP for THP swap out

After adding swapping out support for THP (Transparent Huge Page), it is
possible that a THP in swap cache (partly swapped out) need to be split.
To split such a THP, the swap cluster backing the THP need to be split
too, that is, the CLUSTER_FLAG_HUGE flag need to be cleared for the swap
cluster.  The patch implemented this.

And because the THP swap writing needs the THP keeps as huge page during
writing.  The PageWriteback flag is checked before splitting.

Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: "Huang, Ying" <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Minchan Kim <[email protected]>
Cc: Hugh Dickins <[email protected]>
Cc: Shaohua Li <[email protected]>
Cc: Rik van Riel <[email protected]>
Cc: Andrea Arcangeli <[email protected]>
Cc: "Kirill A . Shutemov" <[email protected]>
Cc: Dan Williams <[email protected]>
Cc: Jens Axboe <[email protected]>
Cc: Michal Hocko <[email protected]>
Cc: Ross Zwisler <[email protected]> [for brd.c, zram_drv.c, pmem.c]
Cc: Vishal L Verma <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>

3 files changed, 33 insertions, 1 deletions

diff --git a/include/linux/swap.h b/include/linux/swap.h
index 7176ba780e83..461cf107ad52 100644
--- a/include/linux/swap.h
+++ b/include/linux/swap.h
@@ -527,6 +527,15 @@ static inline swp_entry_t get_swap_page(struct page *page)
 
 #endif /* CONFIG_SWAP */
 
+#ifdef CONFIG_THP_SWAP
+extern int split_swap_cluster(swp_entry_t entry);
+#else
+static inline int split_swap_cluster(swp_entry_t entry)
+{
+	return 0;
+}
+#endif
+
 #ifdef CONFIG_MEMCG
 static inline int mem_cgroup_swappiness(struct mem_cgroup *memcg)
 {
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 02dfe635c9fe..772048c233d1 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -2481,6 +2481,9 @@ int split_huge_page_to_list(struct page *page, struct list_head *list)
 	VM_BUG_ON_PAGE(!PageLocked(page), page);
 	VM_BUG_ON_PAGE(!PageCompound(page), page);
 
+	if (PageWriteback(page))
+		return -EBUSY;
+
 	if (PageAnon(head)) {
 		/*
 		 * The caller does not necessarily hold an mmap_sem that would
@@ -2558,7 +2561,12 @@ int split_huge_page_to_list(struct page *page, struct list_head *list)
 			__dec_node_page_state(page, NR_SHMEM_THPS);
 		spin_unlock(&pgdata->split_queue_lock);
 		__split_huge_page(page, list, flags);
-		ret = 0;
+		if (PageSwapCache(head)) {
+			swp_entry_t entry = { .val = page_private(head) };
+
+			ret = split_swap_cluster(entry);
+		} else
+			ret = 0;
 	} else {
 		if (IS_ENABLED(CONFIG_DEBUG_VM) && mapcount) {
 			pr_alert("total_mapcount: %u, page_count(): %u\n",
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 267b1fe41844..42eff9e4e972 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -1216,6 +1216,21 @@ static void swapcache_free_cluster(swp_entry_t entry)
 		}
 	}
 }
+
+int split_swap_cluster(swp_entry_t entry)
+{
+	struct swap_info_struct *si;
+	struct swap_cluster_info *ci;
+	unsigned long offset = swp_offset(entry);
+
+	si = _swap_info_get(entry);
+	if (!si)
+		return -EBUSY;
+	ci = lock_cluster(si, offset);
+	cluster_clear_huge(ci);
+	unlock_cluster(ci);
+	return 0;
+}
 #else
 static inline void swapcache_free_cluster(swp_entry_t entry)
 {