[04/19] mm/thp: narrow lru locking

Commit Message

Andrew Morton Dec. 15, 2020, 8:33 p.m. UTC

From: Alex Shi <[e-mail]>
Subject: mm/thp: narrow lru locking

lru_lock and page cache xa_lock have no obvious reason to be taken one way
round or the other: until now, lru_lock has been taken before page cache
xa_lock, when splitting a THP; but nothing else takes them together. 
Reverse that ordering: let's narrow the lru locking - but leave
local_irq_disable to block interrupts throughout, like before.

Hugh Dickins point: split_huge_page_to_list() was already silly, to be
using the _irqsave variant: it's just been taking sleeping locks, so would
already be broken if entered with interrupts enabled.  So we can save
passing flags argument down to __split_huge_page().

Why change the lock ordering here?  That was hard to decide.  One reason:
when this series reaches per-memcg lru locking, it relies on the THP's
memcg to be stable when taking the lru_lock: that is now done after the
THP's refcount has been frozen, which ensures page memcg cannot change.

Another reason: previously, lock_page_memcg()'s move_lock was presumed to
nest inside lru_lock; but now lru_lock must nest inside (page cache lock
inside) move_lock, so it becomes possible to use lock_page_memcg() to
stabilize page memcg before taking its lru_lock.  That is not the
mechanism used in this series, but it is an option we want to keep open.

[[e-mail]: rewrite commit log]
Link: https://lkml.kernel.org/r/[e-mail]
Signed-off-by: Alex Shi <[e-mail]>
Reviewed-by: Kirill A. Shutemov <[e-mail]>
Acked-by: Hugh Dickins <[e-mail]>
Cc: Kirill A. Shutemov <[e-mail]>
Cc: Andrea Arcangeli <[e-mail]>
Cc: Johannes Weiner <[e-mail]>
Cc: Matthew Wilcox <[e-mail]>
Cc: Alexander Duyck <[e-mail]>
Cc: Andrey Ryabinin <[e-mail]>
Cc: "Chen, Rong A" <[e-mail]>
Cc: Daniel Jordan <[e-mail]>
Cc: "Huang, Ying" <[e-mail]>
Cc: Jann Horn <[e-mail]>
Cc: Joonsoo Kim <[e-mail]>
Cc: Konstantin Khlebnikov <[e-mail]>
Cc: Mel Gorman <[e-mail]>
Cc: Michal Hocko <[e-mail]>
Cc: Michal Hocko <[e-mail]>
Cc: Mika Penttilä <[e-mail]>
Cc: Minchan Kim <[e-mail]>
Cc: Shakeel Butt <[e-mail]>
Cc: Tejun Heo <[e-mail]>
Cc: Thomas Gleixner <[e-mail]>
Cc: Vladimir Davydov <[e-mail]>
Cc: Vlastimil Babka <[e-mail]>
Cc: Wei Yang <[e-mail]>
Cc: Yang Shi <[e-mail]>
Signed-off-by: Andrew Morton <[e-mail]>
---

 mm/huge_memory.c |   25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)

Patch

--- a/mm/huge_memory.c~mm-thp-narrow-lru-locking
+++ a/mm/huge_memory.c
@@ -2446,7 +2446,7 @@  static void __split_huge_page_tail(struc
 }
 
 static void __split_huge_page(struct page *page, struct list_head *list,
-		pgoff_t end, unsigned long flags)
+		pgoff_t end)
 {
 	struct page *head = compound_head(page);
 	pg_data_t *pgdat = page_pgdat(head);
@@ -2456,8 +2456,6 @@  static void __split_huge_page(struct pag
 	unsigned int nr = thp_nr_pages(head);
 	int i;
 
-	lruvec = mem_cgroup_page_lruvec(head, pgdat);
-
 	/* complete memcg works before add pages to LRU */
 	mem_cgroup_split_huge_fixup(head);
 
@@ -2469,6 +2467,11 @@  static void __split_huge_page(struct pag
 		xa_lock(&swap_cache->i_pages);
 	}
 
+	/* prevent PageLRU to go away from under us, and freeze lru stats */
+	spin_lock(&pgdat->lru_lock);
+
+	lruvec = mem_cgroup_page_lruvec(head, pgdat);
+
 	for (i = nr - 1; i >= 1; i--) {
 		__split_huge_page_tail(head, i, lruvec, list);
 		/* Some pages can be beyond i_size: drop them from page cache */
@@ -2488,6 +2491,8 @@  static void __split_huge_page(struct pag
 	}
 
 	ClearPageCompound(head);
+	spin_unlock(&pgdat->lru_lock);
+	/* Caller disabled irqs, so they are still disabled here */
 
 	split_page_owner(head, nr);
 
@@ -2505,8 +2510,7 @@  static void __split_huge_page(struct pag
 		page_ref_add(head, 2);
 		xa_unlock(&head->mapping->i_pages);
 	}
-
-	spin_unlock_irqrestore(&pgdat->lru_lock, flags);
+	local_irq_enable();
 
 	remap_page(head, nr);
 
@@ -2652,12 +2656,10 @@  bool can_split_huge_page(struct page *pa
 int split_huge_page_to_list(struct page *page, struct list_head *list)
 {
 	struct page *head = compound_head(page);
-	struct pglist_data *pgdata = NODE_DATA(page_to_nid(head));
 	struct deferred_split *ds_queue = get_deferred_split_queue(head);
 	struct anon_vma *anon_vma = NULL;
 	struct address_space *mapping = NULL;
 	int count, mapcount, extra_pins, ret;
-	unsigned long flags;
 	pgoff_t end;
 
 	VM_BUG_ON_PAGE(is_huge_zero_page(head), head);
@@ -2718,9 +2720,8 @@  int split_huge_page_to_list(struct page
 	unmap_page(head);
 	VM_BUG_ON_PAGE(compound_mapcount(head), head);
 
-	/* prevent PageLRU to go away from under us, and freeze lru stats */
-	spin_lock_irqsave(&pgdata->lru_lock, flags);
-
+	/* block interrupt reentry in xa_lock and spinlock */
+	local_irq_disable();
 	if (mapping) {
 		XA_STATE(xas, &mapping->i_pages, page_index(head));
 
@@ -2750,7 +2751,7 @@  int split_huge_page_to_list(struct page
 				__dec_lruvec_page_state(head, NR_FILE_THPS);
 		}
 
-		__split_huge_page(page, list, end, flags);
+		__split_huge_page(page, list, end);
 		ret = 0;
 	} else {
 		if (IS_ENABLED(CONFIG_DEBUG_VM) && mapcount) {
@@ -2764,7 +2765,7 @@  int split_huge_page_to_list(struct page
 		spin_unlock(&ds_queue->split_queue_lock);
 fail:		if (mapping)
 			xa_unlock(&mapping->i_pages);
-		spin_unlock_irqrestore(&pgdata->lru_lock, flags);
+		local_irq_enable();
 		remap_page(head, thp_nr_pages(head));
 		ret = -EBUSY;
 	}