During the pagevec locking, a new page's lruvec is may same as
previous one. Thus we could save a re-locking, and only
change lock iff lruvec is new.
Function named relock_page_lruvec following Hugh Dickins patch.
The first version of this patch used rcu_read_lock to guard
lruvec assign and comparsion with locked_lruvev in relock_page_lruvec.
But Rong Chen <rong.a.chen@xxxxxxxxx> report a regression with
PROVE_LOCKING config. The rcu_read locking causes qspinlock waiting to
be locked for too long.
Since we had hold a spinlock, rcu_read locking isn't necessary.
[lkp@xxxxxxxxx: Fix RCU-related regression reported by LKP robot]
Signed-off-by: Alex Shi <alex.shi@xxxxxxxxxxxxxxxxx>
Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Vladimir Davydov <vdavydov.dev@xxxxxxxxx>
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: Roman Gushchin <guro@xxxxxx>
Cc: Shakeel Butt <shakeelb@xxxxxxxxxx>
Cc: Chris Down <chris@xxxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Andrey Ryabinin <aryabinin@xxxxxxxxxxxxx>
Cc: swkhack <swkhack@xxxxxxxxx>
Cc: "Potyra, Stefan" <Stefan.Potyra@xxxxxxxxxxxxxx>
Cc: Jason Gunthorpe <jgg@xxxxxxxx>
Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx>
Cc: Mauro Carvalho Chehab <mchehab+samsung@xxxxxxxxxx>
Cc: Peng Fan <peng.fan@xxxxxxx>
Cc: Nikolay Borisov <nborisov@xxxxxxxx>
Cc: Ira Weiny <ira.weiny@xxxxxxxxx>
Cc: Kirill Tkhai <ktkhai@xxxxxxxxxxxxx>
Cc: Yang Shi <yang.shi@xxxxxxxxxxxxxxxxx>
Cc: Yafang Shao <laoar.shao@xxxxxxxxx>
Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Cc: Konstantin Khlebnikov <khlebnikov@xxxxxxxxxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Tejun Heo <tj@xxxxxxxxxx>
Cc: linux-kernel@xxxxxxxxxxxxxxx
Cc: cgroups@xxxxxxxxxxxxxxx
Cc: linux-mm@xxxxxxxxx
---
include/linux/memcontrol.h | 44 ++++++++++++++++++++++++++++++++++++++++++++
mm/mlock.c | 16 +++++++++-------
mm/swap.c | 14 ++++++--------
mm/vmscan.c | 24 ++++++++++++------------
4 files changed, 71 insertions(+), 27 deletions(-)
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 0b32eadd0eda..eaec01fb627f 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -1290,6 +1290,50 @@ static inline void dec_lruvec_page_state(struct page *page,
mod_lruvec_page_state(page, idx, -1);
}
+/* Don't lock again iff page's lruvec locked */
+static inline struct lruvec *relock_page_lruvec_irq(struct page *page,
+ struct lruvec *locked_lruvec)
+{
+ struct pglist_data *pgdat = page_pgdat(page);
+ struct lruvec *lruvec;
+
+ if (!locked_lruvec)
+ goto lock;
+
+ lruvec = mem_cgroup_page_lruvec(page, pgdat);
+
+ if (locked_lruvec == lruvec)
+ return lruvec;
+
+ spin_unlock_irq(&locked_lruvec->lru_lock);
+
+lock:
+ lruvec = lock_page_lruvec_irq(page, pgdat);
+ return lruvec;
+}
+
+/* Don't lock again iff page's lruvec locked */
+static inline struct lruvec *relock_page_lruvec_irqsave(struct page *page,
+ struct lruvec *locked_lruvec)
+{
+ struct pglist_data *pgdat = page_pgdat(page);
+ struct lruvec *lruvec;
+
+ if (!locked_lruvec)
+ goto lock;
+
+ lruvec = mem_cgroup_page_lruvec(page, pgdat);
+
+ if (locked_lruvec == lruvec)
+ return lruvec;
+
+ spin_unlock_irqrestore(&locked_lruvec->lru_lock, locked_lruvec->irqflags);
+
+lock:
+ lruvec = lock_page_lruvec_irqsave(page, pgdat);
+ return lruvec;
+}
+
#ifdef CONFIG_CGROUP_WRITEBACK
struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
diff --git a/mm/mlock.c b/mm/mlock.c
index b509b80b8513..8b3a97b62c0a 100644
--- a/mm/mlock.c
+++ b/mm/mlock.c
@@ -290,6 +290,7 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
{
int i;
int nr = pagevec_count(pvec);
+ int delta_munlocked = -nr;
struct pagevec pvec_putback;
struct lruvec *lruvec = NULL;
int pgrescued = 0;
@@ -300,20 +301,19 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
for (i = 0; i < nr; i++) {
struct page *page = pvec->pages[i];
- lruvec = lock_page_lruvec_irq(page, page_pgdat(page));
+ lruvec = relock_page_lruvec_irq(page, lruvec);
if (TestClearPageMlocked(page)) {
/*
* We already have pin from follow_page_mask()
* so we can spare the get_page() here.
*/
- if (__munlock_isolate_lru_page(page, lruvec, false)) {
- __mod_zone_page_state(zone, NR_MLOCK, -1);
- spin_unlock_irq(&lruvec->lru_lock);
+ if (__munlock_isolate_lru_page(page, lruvec, false))
continue;
- } else
+ else
__munlock_isolation_failed(page);
- }
+ } else
+ delta_munlocked++;
/*
* We won't be munlocking this page in the next phase
@@ -323,8 +323,10 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
*/
pagevec_add(&pvec_putback, pvec->pages[i]);
pvec->pages[i] = NULL;
- spin_unlock_irq(&lruvec->lru_lock);
}
+ __mod_zone_page_state(zone, NR_MLOCK, delta_munlocked);
+ if (lruvec)
+ spin_unlock_irq(&lruvec->lru_lock);
/* Now we can release pins of pages that we are not munlocking */
pagevec_release(&pvec_putback);
diff --git a/mm/swap.c b/mm/swap.c
index 60f04cb2b49e..a26e19738e96 100644
--- a/mm/swap.c
+++ b/mm/swap.c
@@ -195,11 +195,12 @@ static void pagevec_lru_move_fn(struct pagevec *pvec,
for (i = 0; i < pagevec_count(pvec); i++) {
struct page *page = pvec->pages[i];
- lruvec = lock_page_lruvec_irqsave(page, page_pgdat(page));
+ lruvec = relock_page_lruvec_irqsave(page, lruvec);
(*move_fn)(page, lruvec, arg);
- spin_unlock_irqrestore(&lruvec->lru_lock, lruvec->irqflags);
}
+ if (lruvec)
+ spin_unlock_irqrestore(&lruvec->lru_lock, lruvec->irqflags);
release_pages(pvec->pages, pvec->nr);
pagevec_reinit(pvec);
@@ -821,15 +822,12 @@ void release_pages(struct page **pages, int nr)
}
if (PageLRU(page)) {
- struct lruvec *new_lruvec = mem_cgroup_page_lruvec(page, page_pgdat(page));
+ struct lruvec *pre_lruvec = lruvec;
- if (new_lruvec != lruvec) {
- if (lruvec)
- spin_unlock_irqrestore(&lruvec->lru_lock, lruvec->irqflags);
+ lruvec = relock_page_lruvec_irqsave(page, lruvec);
+ if (pre_lruvec != lruvec)
lock_batch = 0;
- lruvec = lock_page_lruvec_irqsave(page, page_pgdat(page));
- }
VM_BUG_ON_PAGE(!PageLRU(page), page);
__ClearPageLRU(page);
del_page_from_lru_list(page, lruvec, page_off_lru(page));
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 3cdf343e7a27..ba57c55a6a41 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1825,22 +1825,25 @@ static int too_many_isolated(struct pglist_data *pgdat, int file,
* Returns the number of pages moved to the given lruvec.
*/
-static unsigned noinline_for_stack move_pages_to_lru(struct lruvec *lruvec,
+static unsigned noinline_for_stack move_pages_to_lru(struct lruvec *llvec,
struct list_head *list)
{
int nr_pages, nr_moved = 0;
LIST_HEAD(pages_to_free);
struct page *page;
enum lru_list lru;
+ struct lruvec *lruvec = llvec;
while (!list_empty(list)) {
page = lru_to_page(list);
+ lruvec = relock_page_lruvec_irq(page, lruvec);
+
VM_BUG_ON_PAGE(PageLRU(page), page);
if (unlikely(!page_evictable(page))) {
list_del(&page->lru);
spin_unlock_irq(&lruvec->lru_lock);
+ lruvec = NULL;
putback_lru_page(page);
- spin_lock_irq(&lruvec->lru_lock);
continue;
}
@@ -1858,8 +1861,8 @@ static unsigned noinline_for_stack move_pages_to_lru(struct lruvec *lruvec,
if (unlikely(PageCompound(page))) {
spin_unlock_irq(&lruvec->lru_lock);
+ lruvec = NULL;
(*get_compound_page_dtor(page))(page);
- spin_lock_irq(&lruvec->lru_lock);
} else
list_add(&page->lru, &pages_to_free);
} else {
@@ -1867,6 +1870,11 @@ static unsigned noinline_for_stack move_pages_to_lru(struct lruvec *lruvec,
}
}
+ if (lruvec != llvec) {
+ if (lruvec)
+ spin_unlock_irq(&lruvec->lru_lock);
+ spin_lock_irq(&llvec->lru_lock);
+ }
/*
* To save our caller's stack, now use input list for pages to free.
*/
@@ -4289,18 +4297,10 @@ void check_move_unevictable_pages(struct pagevec *pvec)
for (i = 0; i < pvec->nr; i++) {
struct page *page = pvec->pages[i];
- struct pglist_data *pgdat = page_pgdat(page);
- struct lruvec *new_lruvec = mem_cgroup_page_lruvec(page, pgdat);
-
pgscanned++;
- if (lruvec != new_lruvec) {
- if (lruvec)
- spin_unlock_irq(&lruvec->lru_lock);
- lruvec = new_lruvec;
- spin_lock_irq(&lruvec->lru_lock);
- }
+ lruvec = relock_page_lruvec_irq(page, lruvec);
if (!PageLRU(page) || !PageUnevictable(page))
continue;
--
1.8.3.1
