The patch titled
Subject: mm/swap: convert lru_deactivate_file to a folio_batch
has been added to the -mm mm-unstable branch. Its filename is
mm-swap-convert-lru_deactivate_file-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_deactivate_file-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_deactivate_file to a folio_batch
Date: Fri, 17 Jun 2022 18:50:03 +0100
Use a folio throughout lru_deactivate_file_fn(), removing many hidden
calls to compound_head(). Shrinks the kernel by 864 bytes of text.
Link: https://lkml.kernel.org/r/20220617175020.717127-6-willy@xxxxxxxxxxxxx
Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/swap.c | 82 ++++++++++++++++++++++++----------------------------
1 file changed, 39 insertions(+), 43 deletions(-)
--- a/mm/swap.c~mm-swap-convert-lru_deactivate_file-to-a-folio_batch
+++ a/mm/swap.c
@@ -62,7 +62,7 @@ static DEFINE_PER_CPU(struct lru_rotate,
struct lru_pvecs {
local_lock_t lock;
struct folio_batch lru_add;
- struct pagevec lru_deactivate_file;
+ struct folio_batch lru_deactivate_file;
struct pagevec lru_deactivate;
struct pagevec lru_lazyfree;
#ifdef CONFIG_SMP
@@ -562,56 +562,57 @@ void lru_cache_add_inactive_or_unevictab
}
/*
- * If the page can not be invalidated, it is moved to the
+ * If the folio cannot be invalidated, it is moved to the
* inactive list to speed up its reclaim. It is moved to the
* head of the list, rather than the tail, to give the flusher
* threads some time to write it out, as this is much more
* effective than the single-page writeout from reclaim.
*
- * If the page isn't page_mapped and dirty/writeback, the page
- * could reclaim asap using PG_reclaim.
+ * If the folio isn't mapped and dirty/writeback, the folio
+ * could be reclaimed asap using the reclaim flag.
*
- * 1. active, mapped page -> none
- * 2. active, dirty/writeback page -> inactive, head, PG_reclaim
- * 3. inactive, mapped page -> none
- * 4. inactive, dirty/writeback page -> inactive, head, PG_reclaim
+ * 1. active, mapped folio -> none
+ * 2. active, dirty/writeback folio -> inactive, head, reclaim
+ * 3. inactive, mapped folio -> none
+ * 4. inactive, dirty/writeback folio -> inactive, head, reclaim
* 5. inactive, clean -> inactive, tail
* 6. Others -> none
*
- * In 4, why it moves inactive's head, the VM expects the page would
- * be write it out by flusher threads as this is much more effective
+ * In 4, it moves to the head of the inactive list so the folio is
+ * written out by flusher threads as this is much more efficient
* than the single-page writeout from reclaim.
*/
-static void lru_deactivate_file_fn(struct page *page, struct lruvec *lruvec)
+static void lru_deactivate_file_fn(struct lruvec *lruvec, struct folio *folio)
{
- bool active = PageActive(page);
- int nr_pages = thp_nr_pages(page);
+ bool active = folio_test_active(folio);
+ long nr_pages = folio_nr_pages(folio);
- if (PageUnevictable(page))
+ if (folio_test_unevictable(folio))
return;
- /* Some processes are using the page */
- if (page_mapped(page))
+ /* Some processes are using the folio */
+ if (folio_mapped(folio))
return;
- del_page_from_lru_list(page, lruvec);
- ClearPageActive(page);
- ClearPageReferenced(page);
+ lruvec_del_folio(lruvec, folio);
+ folio_clear_active(folio);
+ folio_clear_referenced(folio);
- if (PageWriteback(page) || PageDirty(page)) {
+ if (folio_test_writeback(folio) || folio_test_dirty(folio)) {
/*
- * PG_reclaim could be raced with end_page_writeback
- * It can make readahead confusing. But race window
- * is _really_ small and it's non-critical problem.
+ * Setting the reclaim flag could race with
+ * folio_end_writeback() and confuse readahead. But the
+ * race window is _really_ small and it's not a critical
+ * problem.
*/
- add_page_to_lru_list(page, lruvec);
- SetPageReclaim(page);
+ lruvec_add_folio(lruvec, folio);
+ folio_set_reclaim(folio);
} else {
/*
- * The page's writeback ends up during pagevec
- * We move that page into tail of inactive.
+ * The folio's writeback ended while it was in the batch.
+ * We move that folio to the tail of the inactive list.
*/
- add_page_to_lru_list_tail(page, lruvec);
+ lruvec_add_folio_tail(lruvec, folio);
__count_vm_events(PGROTATED, nr_pages);
}
@@ -685,9 +686,9 @@ void lru_add_drain_cpu(int cpu)
local_unlock_irqrestore(&lru_rotate.lock, flags);
}
- pvec = &per_cpu(lru_pvecs.lru_deactivate_file, cpu);
- if (pagevec_count(pvec))
- pagevec_lru_move_fn(pvec, lru_deactivate_file_fn);
+ fbatch = &per_cpu(lru_pvecs.lru_deactivate_file, cpu);
+ if (folio_batch_count(fbatch))
+ folio_batch_move_lru(fbatch, lru_deactivate_file_fn);
pvec = &per_cpu(lru_pvecs.lru_deactivate, cpu);
if (pagevec_count(pvec))
@@ -701,32 +702,27 @@ void lru_add_drain_cpu(int cpu)
}
/**
- * deactivate_file_folio() - Forcefully deactivate a file folio.
+ * deactivate_file_folio() - Deactivate a file folio.
* @folio: Folio to deactivate.
*
* This function hints to the VM that @folio is a good reclaim candidate,
* for example if its invalidation fails due to the folio being dirty
* or under writeback.
*
- * Context: Caller holds a reference on the page.
+ * Context: Caller holds a reference on the folio.
*/
void deactivate_file_folio(struct folio *folio)
{
- struct pagevec *pvec;
+ struct folio_batch *fbatch;
- /*
- * In a workload with many unevictable pages such as mprotect,
- * unevictable folio deactivation for accelerating reclaim is pointless.
- */
+ /* Deactivating an unevictable folio will not accelerate reclaim */
if (folio_test_unevictable(folio))
return;
folio_get(folio);
local_lock(&lru_pvecs.lock);
- pvec = this_cpu_ptr(&lru_pvecs.lru_deactivate_file);
-
- if (pagevec_add_and_need_flush(pvec, &folio->page))
- pagevec_lru_move_fn(pvec, lru_deactivate_file_fn);
+ fbatch = this_cpu_ptr(&lru_pvecs.lru_deactivate_file);
+ folio_batch_add_and_move(fbatch, folio, lru_deactivate_file_fn);
local_unlock(&lru_pvecs.lock);
}
@@ -899,7 +895,7 @@ static inline void __lru_add_drain_all(b
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)) ||
+ folio_batch_count(&per_cpu(lru_pvecs.lru_deactivate_file, cpu)) ||
pagevec_count(&per_cpu(lru_pvecs.lru_deactivate, cpu)) ||
pagevec_count(&per_cpu(lru_pvecs.lru_lazyfree, cpu)) ||
need_activate_page_drain(cpu) ||
_
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
