linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: "Matthew Wilcox (Oracle)" <willy@infradead.org>
To: linux-mm@kvack.org
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>,
	linux-kernel@vger.kernel.org
Subject: [PATCH 49/75] mm/mlock: Turn munlock_vma_page() into munlock_vma_folio()
Date: Fri,  4 Feb 2022 19:58:26 +0000	[thread overview]
Message-ID: <20220204195852.1751729-50-willy@infradead.org> (raw)
In-Reply-To: <20220204195852.1751729-1-willy@infradead.org>

Add back munlock_vma_page() as a wrapper function.  Saves a few calls
to compound_head() and an assertion that the page is not a tail page.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 mm/folio-compat.c |  5 +++
 mm/internal.h     |  3 +-
 mm/mlock.c        | 86 +++++++++++++++++++++++------------------------
 3 files changed, 50 insertions(+), 44 deletions(-)

diff --git a/mm/folio-compat.c b/mm/folio-compat.c
index 90f03187a5e3..3804fd8c1f20 100644
--- a/mm/folio-compat.c
+++ b/mm/folio-compat.c
@@ -176,6 +176,11 @@ void mlock_vma_page(struct page *page)
 	mlock_vma_folio(page_folio(page));
 }
 
+unsigned long munlock_vma_page(struct page *page)
+{
+	return munlock_vma_folio(page_folio(page));
+}
+
 void page_mlock(struct page *page)
 {
 	folio_mlock(page_folio(page));
diff --git a/mm/internal.h b/mm/internal.h
index 18b024aa7e59..66645972cbd7 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -413,7 +413,8 @@ static inline void munlock_vma_pages_all(struct vm_area_struct *vma)
  */
 void mlock_vma_page(struct page *page);
 void mlock_vma_folio(struct folio *folio);
-extern unsigned int munlock_vma_page(struct page *page);
+unsigned long munlock_vma_page(struct page *page);
+unsigned long munlock_vma_folio(struct folio *folio);
 
 extern int mlock_future_check(struct mm_struct *mm, unsigned long flags,
 			      unsigned long len);
diff --git a/mm/mlock.c b/mm/mlock.c
index d998fd5c84bf..f188038ef48e 100644
--- a/mm/mlock.c
+++ b/mm/mlock.c
@@ -115,82 +115,81 @@ void mlock_vma_folio(struct folio *folio)
 /*
  * Finish munlock after successful page isolation
  *
- * Page must be locked. This is a wrapper for page_mlock()
- * and putback_lru_page() with munlock accounting.
+ * Folio must be locked. This is a wrapper for folio_mlock()
+ * and folio_putback_lru() with munlock accounting.
  */
-static void __munlock_isolated_page(struct page *page)
+static void __munlock_isolated_folio(struct folio *folio)
 {
 	/*
 	 * Optimization: if the page was mapped just once, that's our mapping
 	 * and we don't need to check all the other vmas.
 	 */
-	if (page_mapcount(page) > 1)
-		page_mlock(page);
+	/* XXX: should be folio_mapcount(), surely? */
+	if (page_mapcount(&folio->page) > 1)
+		folio_mlock(folio);
 
 	/* Did try_to_unlock() succeed or punt? */
-	if (!PageMlocked(page))
-		count_vm_events(UNEVICTABLE_PGMUNLOCKED, thp_nr_pages(page));
+	if (!folio_test_mlocked(folio))
+		count_vm_events(UNEVICTABLE_PGMUNLOCKED, folio_nr_pages(folio));
 
-	putback_lru_page(page);
+	folio_putback_lru(folio);
 }
 
 /*
- * Accounting for page isolation fail during munlock
+ * Accounting for folio isolation fail during munlock
  *
- * Performs accounting when page isolation fails in munlock. There is nothing
- * else to do because it means some other task has already removed the page
- * from the LRU. putback_lru_page() will take care of removing the page from
+ * Performs accounting when folio isolation fails in munlock. There is nothing
+ * else to do because it means some other task has already removed the folio
+ * from the LRU. folio_putback_lru() will take care of removing the folio from
  * the unevictable list, if necessary. vmscan [folio_referenced()] will move
- * the page back to the unevictable list if some other vma has it mlocked.
+ * the folio back to the unevictable list if some other vma has it mlocked.
  */
-static void __munlock_isolation_failed(struct page *page)
+static void __munlock_isolation_failed(struct folio *folio)
 {
-	int nr_pages = thp_nr_pages(page);
+	long nr_pages = folio_nr_pages(folio);
 
-	if (PageUnevictable(page))
+	if (folio_test_unevictable(folio))
 		__count_vm_events(UNEVICTABLE_PGSTRANDED, nr_pages);
 	else
 		__count_vm_events(UNEVICTABLE_PGMUNLOCKED, nr_pages);
 }
 
 /**
- * munlock_vma_page - munlock a vma page
- * @page: page to be unlocked, either a normal page or THP page head
+ * munlock_vma_folio() - munlock a vma folio.
+ * @folio: Folio to be unlocked.
  *
- * returns the size of the page as a page mask (0 for normal page,
- *         HPAGE_PMD_NR - 1 for THP head page)
- *
- * called from munlock()/munmap() path with page supposedly on the LRU.
- * When we munlock a page, because the vma where we found the page is being
+ * called from munlock()/munmap() path with folio supposedly on the LRU.
+ * When we munlock a folio, because the vma where we found the folio is being
  * munlock()ed or munmap()ed, we want to check whether other vmas hold the
- * page locked so that we can leave it on the unevictable lru list and not
- * bother vmscan with it.  However, to walk the page's rmap list in
- * page_mlock() we must isolate the page from the LRU.  If some other
- * task has removed the page from the LRU, we won't be able to do that.
- * So we clear the PageMlocked as we might not get another chance.  If we
- * can't isolate the page, we leave it for putback_lru_page() and vmscan
+ * folio locked so that we can leave it on the unevictable lru list and not
+ * bother vmscan with it.  However, to walk the folio's rmap list in
+ * folio_mlock() we must isolate the folio from the LRU.  If some other
+ * task has removed the folio from the LRU, we won't be able to do that.
+ * So we clear the folio mlocked flag as we might not get another chance.  If
+ * we can't isolate the folio, we leave it for folio_putback_lru() and vmscan
  * [folio_referenced()/try_to_unmap()] to deal with.
+ *
+ * Return: The size of the folio as a page mask (2^order - 1).
  */
-unsigned int munlock_vma_page(struct page *page)
+unsigned long munlock_vma_folio(struct folio *folio)
 {
-	int nr_pages;
+	long nr_pages;
 
-	/* For page_mlock() and to serialize with page migration */
-	BUG_ON(!PageLocked(page));
-	VM_BUG_ON_PAGE(PageTail(page), page);
+	/* For folio_mlock() and to serialize with page migration */
+	BUG_ON(!folio_test_locked(folio));
 
-	if (!TestClearPageMlocked(page)) {
-		/* Potentially, PTE-mapped THP: do not skip the rest PTEs */
+	if (!folio_test_clear_mlocked(folio)) {
+		/* Potentially, PTE-mapped folio: do not skip the other PTEs */
 		return 0;
 	}
 
-	nr_pages = thp_nr_pages(page);
-	mod_zone_page_state(page_zone(page), NR_MLOCK, -nr_pages);
+	nr_pages = folio_nr_pages(folio);
+	zone_stat_mod_folio(folio, NR_MLOCK, -nr_pages);
 
-	if (!isolate_lru_page(page))
-		__munlock_isolated_page(page);
+	if (!folio_isolate_lru(folio))
+		__munlock_isolated_folio(folio);
 	else
-		__munlock_isolation_failed(page);
+		__munlock_isolation_failed(folio);
 
 	return nr_pages - 1;
 }
@@ -289,7 +288,7 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
 				del_page_from_lru_list(page, lruvec);
 				continue;
 			} else
-				__munlock_isolation_failed(page);
+				__munlock_isolation_failed(folio);
 		} else {
 			delta_munlocked++;
 		}
@@ -318,6 +317,7 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
 		struct page *page = pvec->pages[i];
 
 		if (page) {
+			struct folio *folio = page_folio(page);
 			lock_page(page);
 			if (!__putback_lru_fast_prepare(page, &pvec_putback,
 					&pgrescued)) {
@@ -326,7 +326,7 @@ static void __munlock_pagevec(struct pagevec *pvec, struct zone *zone)
 				 * pin before unlock_page()
 				 */
 				get_page(page); /* for putback_lru_page() */
-				__munlock_isolated_page(page);
+				__munlock_isolated_folio(folio);
 				unlock_page(page);
 				put_page(page); /* from follow_page_mask() */
 			}
-- 
2.34.1



  parent reply	other threads:[~2022-02-04 20:22 UTC|newest]

Thread overview: 115+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-04 19:57 [PATCH 00/75] MM folio patches for 5.18 Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 01/75] mm/gup: Increment the page refcount before the pincount Matthew Wilcox (Oracle)
2022-02-04 21:13   ` John Hubbard
2022-02-04 21:28     ` Matthew Wilcox
2022-02-07  7:45   ` Christoph Hellwig
2022-02-04 19:57 ` [PATCH 02/75] mm/gup: Remove for_each_compound_range() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 03/75] mm/gup: Remove for_each_compound_head() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 04/75] mm/gup: Change the calling convention for compound_range_next() Matthew Wilcox (Oracle)
2022-02-07  7:45   ` Christoph Hellwig
2022-02-04 19:57 ` [PATCH 05/75] mm/gup: Optimise compound_range_next() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 06/75] mm/gup: Change the calling convention for compound_next() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 07/75] mm/gup: Fix some contiguous memmap assumptions Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 08/75] mm/gup: Remove an assumption of a contiguous memmap Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 09/75] mm/gup: Handle page split race more efficiently Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 10/75] mm/gup: Remove hpage_pincount_add() Matthew Wilcox (Oracle)
2022-02-04 21:29   ` John Hubbard
2022-02-07  7:46   ` Christoph Hellwig
2022-02-04 19:57 ` [PATCH 11/75] mm/gup: Remove hpage_pincount_sub() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 12/75] mm: Make compound_pincount always available Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 13/75] mm: Add folio_pincount_ptr() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 14/75] mm: Turn page_maybe_dma_pinned() into folio_maybe_dma_pinned() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 15/75] mm/gup: Add try_get_folio() and try_grab_folio() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 16/75] mm/gup: Convert try_grab_page() to use a folio Matthew Wilcox (Oracle)
2022-02-06  2:12   ` John Hubbard
2022-02-07  7:47   ` Christoph Hellwig
2022-02-04 19:57 ` [PATCH 17/75] mm: Remove page_cache_add_speculative() and page_cache_get_speculative() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 18/75] mm/gup: Add gup_put_folio() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 19/75] mm/hugetlb: Use try_grab_folio() instead of try_grab_compound_head() Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 20/75] mm/gup: Convert gup_pte_range() to use a folio Matthew Wilcox (Oracle)
2022-02-06 14:52   ` Mark Hemment
2022-02-11 20:20     ` Matthew Wilcox
2022-02-04 19:57 ` [PATCH 21/75] mm/gup: Convert gup_hugepte() " Matthew Wilcox (Oracle)
2022-02-04 19:57 ` [PATCH 22/75] mm/gup: Convert gup_huge_pmd() " Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 23/75] mm/gup: Convert gup_huge_pud() " Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 24/75] mm/gup: Convert gup_huge_pgd() " Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 25/75] mm/gup: Turn compound_next() into gup_folio_next() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 26/75] mm/gup: Turn compound_range_next() into gup_folio_range_next() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 27/75] mm: Turn isolate_lru_page() into folio_isolate_lru() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 28/75] mm/gup: Convert check_and_migrate_movable_pages() to use a folio Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 29/75] mm/workingset: Convert workingset_eviction() to take " Matthew Wilcox (Oracle)
2022-02-07  7:49   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 30/75] mm/memcg: Convert mem_cgroup_swapout() " Matthew Wilcox (Oracle)
2022-02-07  7:49   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 31/75] mm: Add lru_to_folio() Matthew Wilcox (Oracle)
2022-02-07  7:50   ` Christoph Hellwig
2022-02-11 20:24     ` Matthew Wilcox
2022-02-04 19:58 ` [PATCH 32/75] mm: Turn putback_lru_page() into folio_putback_lru() Matthew Wilcox (Oracle)
2022-02-07  7:50   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 33/75] mm/vmscan: Convert __remove_mapping() to take a folio Matthew Wilcox (Oracle)
2022-02-07  7:51   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 34/75] mm/vmscan: Turn page_check_dirty_writeback() into folio_check_dirty_writeback() Matthew Wilcox (Oracle)
2022-02-07  7:51   ` Christoph Hellwig
2022-02-12  1:49     ` Matthew Wilcox
2022-02-04 19:58 ` [PATCH 35/75] mm: Turn head_compound_mapcount() into folio_entire_mapcount() Matthew Wilcox (Oracle)
2022-02-07  7:52   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 36/75] mm: Add folio_mapcount() Matthew Wilcox (Oracle)
2022-02-07  7:53   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 37/75] mm: Add split_folio_to_list() Matthew Wilcox (Oracle)
2022-02-07  7:54   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 38/75] mm: Add folio_is_zone_device() and folio_is_device_private() Matthew Wilcox (Oracle)
2022-02-07  7:54   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 39/75] mm: Add folio_pgoff() Matthew Wilcox (Oracle)
2022-02-07  7:55   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 40/75] mm: Add pvmw_set_page() and pvmw_set_folio() Matthew Wilcox (Oracle)
2022-02-07  7:55   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 41/75] hexagon: Add pmd_pfn() Matthew Wilcox (Oracle)
2022-02-06 18:13   ` Mike Rapoport
2022-02-06 20:46     ` Matthew Wilcox
2022-02-06 21:33       ` Mike Rapoport
2022-02-06 22:05         ` Matthew Wilcox
2022-02-07 14:24           ` Mike Rapoport
2022-02-04 19:58 ` [PATCH 42/75] mm: Convert page_vma_mapped_walk to work on PFNs Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 43/75] mm/page_idle: Convert page_idle_clear_pte_refs() to use a folio Matthew Wilcox (Oracle)
2022-02-07  7:57   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 44/75] mm/rmap: Use a folio in page_mkclean_one() Matthew Wilcox (Oracle)
2022-02-07  7:57   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 45/75] mm/rmap: Turn page_referenced() into folio_referenced() Matthew Wilcox (Oracle)
2022-02-07  7:58   ` Christoph Hellwig
2022-02-04 19:58 ` [PATCH 46/75] mm/mlock: Turn clear_page_mlock() into folio_end_mlock() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 47/75] mm/mlock: Turn mlock_vma_page() into mlock_vma_folio() Matthew Wilcox (Oracle)
2022-02-07 10:46   ` Mike Rapoport
2022-02-04 19:58 ` [PATCH 48/75] mm/rmap: Turn page_mlock() into folio_mlock() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` Matthew Wilcox (Oracle) [this message]
2022-02-04 19:58 ` [PATCH 50/75] mm/huge_memory: Convert __split_huge_pmd() to take a folio Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 51/75] mm/rmap: Convert try_to_unmap() " Matthew Wilcox (Oracle)
2022-02-09 14:24   ` Mauricio Faria de Oliveira
2022-02-09 14:29     ` Matthew Wilcox
2022-02-04 19:58 ` [PATCH 52/75] mm/rmap: Convert try_to_migrate() to folios Matthew Wilcox (Oracle)
2022-02-09 15:27   ` Zi Yan
2022-02-04 19:58 ` [PATCH 53/75] mm/rmap: Convert make_device_exclusive_range() to use folios Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 54/75] mm/migrate: Convert remove_migration_ptes() to folios Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 55/75] mm/damon: Convert damon_pa_mkold() to use a folio Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 56/75] mm/damon: Convert damon_pa_young() " Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 57/75] mm/rmap: Turn page_lock_anon_vma_read() into folio_lock_anon_vma_read() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 58/75] mm: Turn page_anon_vma() into folio_anon_vma() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 59/75] mm/rmap: Convert rmap_walk() to take a folio Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 60/75] mm/rmap: Constify the rmap_walk_control argument Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 61/75] mm/vmscan: Free non-shmem folios without splitting them Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 62/75] mm/vmscan: Optimise shrink_page_list for non-PMD-sized folios Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 63/75] mm/vmscan: Account large folios correctly Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 64/75] mm/vmscan: Turn page_check_references() into folio_check_references() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 65/75] mm/vmscan: Convert pageout() to take a folio Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 66/75] mm: Turn can_split_huge_page() into can_split_folio() Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 67/75] mm/filemap: Allow large folios to be added to the page cache Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 68/75] mm: Fix READ_ONLY_THP warning Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 69/75] mm: Make large folios depend on THP Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 70/75] mm: Support arbitrary THP sizes Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 71/75] mm/readahead: Add large folio readahead Matthew Wilcox (Oracle)
2022-02-06 13:10   ` Mark Hemment
2022-02-04 19:58 ` [PATCH 72/75] mm/readahead: Align file mappings for non-DAX Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 73/75] mm/readahead: Switch to page_cache_ra_order Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 74/75] mm/filemap: Support VM_HUGEPAGE for file mappings Matthew Wilcox (Oracle)
2022-02-04 19:58 ` [PATCH 75/75] selftests/vm/transhuge-stress: Support file-backed PMD folios Matthew Wilcox (Oracle)
2022-02-13 22:31 ` [PATCH 00/75] MM folio patches for 5.18 John Hubbard
2022-02-14  4:33 ` Matthew Wilcox

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220204195852.1751729-50-willy@infradead.org \
    --to=willy@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox