The patch titled
Subject: mm/swap: convert lru_add to a folio_batch
has been added to the -mm mm-unstable branch. Its filename is
mm-swap-convert-lru_add-to-a-folio_batch.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-swap-convert-lru_add-to-a-folio_batch.patch
This patch will later appear in the mm-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
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 via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: "Matthew Wilcox (Oracle)" <willy@xxxxxxxxxxxxx>
Subject: mm/swap: convert lru_add to a folio_batch
Date: Fri, 17 Jun 2022 18:50:02 +0100
When adding folios to the LRU for the first time, the LRU flag will
already be clear, so skip the test-and-clear part of moving from one
LRU to another.
Removes 285 bytes from kernel text, mostly due to removing
__pagevec_lru_add().
Link: https://lkml.kernel.org/r/20220617175020.717127-5-willy@xxxxxxxxxxxxx
Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/swap.c | 75 ++++++++++++++++++----------------------------------
1 file changed, 26 insertions(+), 49 deletions(-)
--- a/mm/swap.c~mm-swap-convert-lru_add-to-a-folio_batch
+++ a/mm/swap.c
@@ -61,7 +61,7 @@ static DEFINE_PER_CPU(struct lru_rotate,
*/
struct lru_pvecs {
local_lock_t lock;
- struct pagevec lru_add;
+ struct folio_batch lru_add;
struct pagevec lru_deactivate_file;
struct pagevec lru_deactivate;
struct pagevec lru_lazyfree;
@@ -228,14 +228,13 @@ static bool pagevec_add_and_need_flush(s
typedef void (*move_fn_t)(struct lruvec *lruvec, struct folio *folio);
-static void __pagevec_lru_add_fn(struct folio *folio, struct lruvec *lruvec)
+static void lru_add_fn(struct lruvec *lruvec, struct folio *folio)
{
int was_unevictable = folio_test_clear_unevictable(folio);
long nr_pages = folio_nr_pages(folio);
VM_BUG_ON_FOLIO(folio_test_lru(folio), folio);
- folio_set_lru(folio);
/*
* Is an smp_mb__after_atomic() still required here, before
* folio_evictable() tests PageMlocked, to rule out the possibility
@@ -269,28 +268,6 @@ static void __pagevec_lru_add_fn(struct
trace_mm_lru_insertion(folio);
}
-/*
- * Add the passed pages to the LRU, then drop the caller's refcount
- * on them. Reinitialises the caller's pagevec.
- */
-static void __pagevec_lru_add(struct pagevec *pvec)
-{
- int i;
- struct lruvec *lruvec = NULL;
- unsigned long flags = 0;
-
- for (i = 0; i < pagevec_count(pvec); i++) {
- struct folio *folio = page_folio(pvec->pages[i]);
-
- lruvec = folio_lruvec_relock_irqsave(folio, lruvec, &flags);
- __pagevec_lru_add_fn(folio, lruvec);
- }
- if (lruvec)
- unlock_page_lruvec_irqrestore(lruvec, flags);
- release_pages(pvec->pages, pvec->nr);
- pagevec_reinit(pvec);
-}
-
static void folio_batch_move_lru(struct folio_batch *fbatch, move_fn_t move_fn)
{
int i;
@@ -301,7 +278,7 @@ static void folio_batch_move_lru(struct
struct folio *folio = fbatch->folios[i];
/* block memcg migration while the folio moves between lru */
- if (!folio_test_clear_lru(folio))
+ if (move_fn != lru_add_fn && !folio_test_clear_lru(folio))
continue;
lruvec = folio_lruvec_relock_irqsave(folio, lruvec, &flags);
@@ -473,26 +450,26 @@ static void folio_activate(struct folio
static void __lru_cache_activate_folio(struct folio *folio)
{
- struct pagevec *pvec;
+ struct folio_batch *fbatch;
int i;
local_lock(&lru_pvecs.lock);
- pvec = this_cpu_ptr(&lru_pvecs.lru_add);
+ fbatch = this_cpu_ptr(&lru_pvecs.lru_add);
/*
- * Search backwards on the optimistic assumption that the page being
- * activated has just been added to this pagevec. Note that only
- * the local pagevec is examined as a !PageLRU page could be in the
+ * Search backwards on the optimistic assumption that the folio being
+ * activated has just been added to this batch. Note that only
+ * the local batch is examined as a !LRU folio could be in the
* process of being released, reclaimed, migrated or on a remote
- * pagevec that is currently being drained. Furthermore, marking
- * a remote pagevec's page PageActive potentially hits a race where
- * a page is marked PageActive just after it is added to the inactive
+ * batch that is currently being drained. Furthermore, marking
+ * a remote batch's folio active potentially hits a race where
+ * a folio is marked active just after it is added to the inactive
* list causing accounting errors and BUG_ON checks to trigger.
*/
- for (i = pagevec_count(pvec) - 1; i >= 0; i--) {
- struct page *pagevec_page = pvec->pages[i];
+ for (i = folio_batch_count(fbatch) - 1; i >= 0; i--) {
+ struct folio *batch_folio = fbatch->folios[i];
- if (pagevec_page == &folio->page) {
+ if (batch_folio == folio) {
folio_set_active(folio);
break;
}
@@ -551,16 +528,16 @@ EXPORT_SYMBOL(folio_mark_accessed);
*/
void folio_add_lru(struct folio *folio)
{
- struct pagevec *pvec;
+ struct folio_batch *fbatch;
- VM_BUG_ON_FOLIO(folio_test_active(folio) && folio_test_unevictable(folio), folio);
+ VM_BUG_ON_FOLIO(folio_test_active(folio) &&
+ folio_test_unevictable(folio), folio);
VM_BUG_ON_FOLIO(folio_test_lru(folio), folio);
folio_get(folio);
local_lock(&lru_pvecs.lock);
- pvec = this_cpu_ptr(&lru_pvecs.lru_add);
- if (pagevec_add_and_need_flush(pvec, &folio->page))
- __pagevec_lru_add(pvec);
+ fbatch = this_cpu_ptr(&lru_pvecs.lru_add);
+ folio_batch_add_and_move(fbatch, folio, lru_add_fn);
local_unlock(&lru_pvecs.lock);
}
EXPORT_SYMBOL(folio_add_lru);
@@ -691,11 +668,11 @@ static void lru_lazyfree_fn(struct page
*/
void lru_add_drain_cpu(int cpu)
{
- struct folio_batch *fbatch;
- struct pagevec *pvec = &per_cpu(lru_pvecs.lru_add, cpu);
+ struct folio_batch *fbatch = &per_cpu(lru_pvecs.lru_add, cpu);
+ struct pagevec *pvec;
- if (pagevec_count(pvec))
- __pagevec_lru_add(pvec);
+ if (folio_batch_count(fbatch))
+ folio_batch_move_lru(fbatch, lru_add_fn);
fbatch = &per_cpu(lru_rotate.fbatch, cpu);
/* Disabling interrupts below acts as a compiler barrier. */
@@ -920,7 +897,7 @@ static inline void __lru_add_drain_all(b
for_each_online_cpu(cpu) {
struct work_struct *work = &per_cpu(lru_add_drain_work, cpu);
- if (pagevec_count(&per_cpu(lru_pvecs.lru_add, cpu)) ||
+ if (folio_batch_count(&per_cpu(lru_pvecs.lru_add, cpu)) ||
data_race(folio_batch_count(&per_cpu(lru_rotate.fbatch, cpu))) ||
pagevec_count(&per_cpu(lru_pvecs.lru_deactivate_file, cpu)) ||
pagevec_count(&per_cpu(lru_pvecs.lru_deactivate, cpu)) ||
@@ -1084,8 +1061,8 @@ EXPORT_SYMBOL(release_pages);
* OK from a correctness point of view but is inefficient - those pages may be
* cache-warm and we want to give them back to the page allocator ASAP.
*
- * So __pagevec_release() will drain those queues here. __pagevec_lru_add()
- * and __pagevec_lru_add_active() call release_pages() directly to avoid
+ * So __pagevec_release() will drain those queues here.
+ * folio_batch_move_lru() calls folios_put() directly to avoid
* mutual recursion.
*/
void __pagevec_release(struct pagevec *pvec)
_
Patches currently in -mm which might be from willy@xxxxxxxxxxxxx are
mm-add-vma-iterator.patch
mmap-use-the-vma-iterator-in-count_vma_pages_range.patch
proc-remove-vma-rbtree-use-from-nommu.patch
arm64-remove-mmap-linked-list-from-vdso.patch
parisc-remove-mmap-linked-list-from-cache-handling.patch
powerpc-remove-mmap-linked-list-walks.patch
s390-remove-vma-linked-list-walks.patch
x86-remove-vma-linked-list-walks.patch
xtensa-remove-vma-linked-list-walks.patch
cxl-remove-vma-linked-list-walk.patch
optee-remove-vma-linked-list-walk.patch
um-remove-vma-linked-list-walk.patch
coredump-remove-vma-linked-list-walk.patch
exec-use-vma-iterator-instead-of-linked-list.patch
fs-proc-task_mmu-stop-using-linked-list-and-highest_vm_end.patch
acct-use-vma-iterator-instead-of-linked-list.patch
perf-use-vma-iterator.patch
sched-use-maple-tree-iterator-to-walk-vmas.patch
fork-use-vma-iterator.patch
mm-khugepaged-stop-using-vma-linked-list.patch
mm-ksm-use-vma-iterators-instead-of-vma-linked-list.patch
mm-mlock-use-vma-iterator-and-maple-state-instead-of-vma-linked-list.patch
mm-pagewalk-use-vma_find-instead-of-vma-linked-list.patch
i915-use-the-vma-iterator.patch
nommu-remove-uses-of-vma-linked-list.patch
mm-vmscan-convert-reclaim_clean_pages_from_list-to-folios.patch
mm-vmscan-convert-isolate_lru_pages-to-use-a-folio.patch
mm-vmscan-convert-move_pages_to_lru-to-use-a-folio.patch
mm-vmscan-convert-shrink_active_list-to-use-a-folio.patch
mm-vmscan-convert-reclaim_pages-to-use-a-folio.patch
mm-add-folios_put.patch
mm-swap-add-folio_batch_move_lru.patch
mm-swap-make-__pagevec_lru_add-static.patch
mm-swap-convert-lru_add-to-a-folio_batch.patch
mm-swap-convert-lru_deactivate_file-to-a-folio_batch.patch
mm-swap-convert-lru_deactivate-to-a-folio_batch.patch
mm-swap-convert-lru_lazyfree-to-a-folio_batch.patch
mm-swap-convert-activate_page-to-a-folio_batch.patch
mm-swap-rename-lru_pvecs-to-cpu_fbatches.patch
mm-swap-pull-the-cpu-conditional-out-of-__lru_add_drain_all.patch
mm-swap-optimise-lru_add_drain_cpu.patch
mm-swap-convert-try_to_free_swap-to-use-a-folio.patch
mm-swap-convert-release_pages-to-use-a-folio-internally.patch
mm-swap-convert-put_pages_list-to-use-folios.patch
mm-swap-convert-__put_page-to-__folio_put.patch
mm-swap-convert-__put_single_page-to-__folio_put_small.patch
mm-swap-convert-__put_compound_page-to-__folio_put_large.patch
mm-swap-convert-__page_cache_release-to-use-a-folio.patch
mm-convert-destroy_compound_page-to-destroy_large_folio.patch
mm-convert-page_swap_flags-to-folio_swap_flags.patch
mm-swap-convert-delete_from_swap_cache-to-take-a-folio.patch
mm-swap-convert-__delete_from_swap_cache-to-a-folio.patch
