The patch titled
Subject: mm/thp: narrow lru locking
has been added to the -mm tree. Its filename is
mm-thp-narrow-lru-locking.patch
This patch should soon appear at
https://ozlabs.org/~akpm/mmots/broken-out/mm-thp-narrow-lru-locking.patch
and later at
https://ozlabs.org/~akpm/mmotm/broken-out/mm-thp-narrow-lru-locking.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Alex Shi <alex.shi@xxxxxxxxxxxxxxxxx>
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.
[hughd@xxxxxxxxxx: rewrite commit log]
Link: https://lkml.kernel.org/r/1604566549-62481-5-git-send-email-alex.shi@xxxxxxxxxxxxxxxxx
Signed-off-by: Alex Shi <alex.shi@xxxxxxxxxxxxxxxxx>
Reviewed-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Acked-by: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Kirill A. Shutemov <kirill@xxxxxxxxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx>
Cc: Alexander Duyck <alexander.duyck@xxxxxxxxx>
Cc: Andrey Ryabinin <aryabinin@xxxxxxxxxxxxx>
Cc: "Chen, Rong A" <rong.a.chen@xxxxxxxxx>
Cc: Daniel Jordan <daniel.m.jordan@xxxxxxxxxx>
Cc: "Huang, Ying" <ying.huang@xxxxxxxxx>
Cc: Jann Horn <jannh@xxxxxxxxxx>
Cc: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Cc: Konstantin Khlebnikov <khlebnikov@xxxxxxxxxxxxxx>
Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxx>
Cc: Mika Penttilä <mika.penttila@xxxxxxxxxxxx>
Cc: Minchan Kim <minchan@xxxxxxxxxx>
Cc: Shakeel Butt <shakeelb@xxxxxxxxxx>
Cc: Tejun Heo <tj@xxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Vladimir Davydov <vdavydov.dev@xxxxxxxxx>
Cc: Vlastimil Babka <vbabka@xxxxxxx>
Cc: Wei Yang <richard.weiyang@xxxxxxxxx>
Cc: Yang Shi <yang.shi@xxxxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/huge_memory.c | 25 +++++++++++++------------
1 file changed, 13 insertions(+), 12 deletions(-)
--- a/mm/huge_memory.c~mm-thp-narrow-lru-locking
+++ a/mm/huge_memory.c
@@ -2432,7 +2432,7 @@ 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);
@@ -2442,8 +2442,6 @@ 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);
@@ -2455,6 +2453,11 @@ 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 */
@@ -2474,6 +2477,8 @@ 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);
@@ -2491,8 +2496,7 @@ 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);
@@ -2638,12 +2642,10 @@ 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);
@@ -2704,9 +2706,8 @@ 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));
@@ -2736,7 +2737,7 @@ 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) {
@@ -2750,7 +2751,7 @@ 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;
}
_
Patches currently in -mm which might be from alex.shi@xxxxxxxxxxxxxxxxx are
mm-filemap-add-static-for-function-__add_to_page_cache_locked.patch
fs-ntfs-remove-unused-varibles.patch
fs-ntfs-remove-unused-varible-attr_len.patch
mm-memcg-update-page-struct-member-in-comments.patch
mm-thp-move-lru_add_page_tail-func-to-huge_memoryc.patch
mm-thp-use-head-for-head-page-in-lru_add_page_tail.patch
mm-thp-simplify-lru_add_page_tail.patch
mm-thp-narrow-lru-locking.patch
mm-vmscan-remove-unnecessary-lruvec-adding.patch
mm-rmap-stop-store-reordering-issue-on-page-mapping.patch
mm-rmap-stop-store-reordering-issue-on-page-mapping-fix.patch
mm-memcg-add-debug-checking-in-lock_page_memcg.patch
mm-swapc-fold-vm-event-pgrotated-into-pagevec_move_tail_fn.patch
mm-lru-move-lock-into-lru_note_cost.patch
mm-vmscan-remove-lruvec-reget-in-move_pages_to_lru.patch
mm-mlock-remove-lru_lock-on-testclearpagemlocked.patch
mm-mlock-remove-__munlock_isolate_lru_page.patch
mm-lru-introduce-testclearpagelru.patch
mm-compaction-do-page-isolation-first-in-compaction.patch
mm-swapc-serialize-memcg-changes-in-pagevec_lru_move_fn.patch
mm-lru-replace-pgdat-lru_lock-with-lruvec-lock.patch
mm-lru-replace-pgdat-lru_lock-with-lruvec-lock-fix.patch
mm-lru-replace-pgdat-lru_lock-with-lruvec-lock-fix-2.patch
mm-lru-introduce-the-relock_page_lruvec-function-fix.patch
docs-vm-remove-unused-3-items-explanation-for-proc-vmstat.patch
mm-memcg-bail-early-from-swap-accounting-if-memcg-disabled.patch
mm-memcg-warning-on-memcg-after-readahead-page-charged.patch
