diff mbox series

[24/26] mm/thp: allow dropping THP from page cache

Message ID	20191019032036.xKRTLdCqf%akpm@linux-foundation.org (mailing list archive)
State	New, archived
Headers	show Return-Path: <SRS0=Badx=YM=kvack.org=owner-linux-mm@kernel.org> DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 8DBD022466 Date: Fri, 18 Oct 2019 20:20:36 -0700 From: akpm@linux-foundation.org To: akpm@linux-foundation.org, kirill.shutemov@linux.intel.com, linux-mm@kvack.org, mm-commits@vger.kernel.org, oleg@redhat.com, songliubraving@fb.com, srikar@linux.vnet.ibm.com, torvalds@linux-foundation.org, william.kucharski@oracle.com, willy@infradead.org, yang.shi@linux.alibaba.com Subject: [patch 24/26] mm/thp: allow dropping THP from page cache Message-ID: <20191019032036.xKRTLdCqf%akpm@linux-foundation.org> User-Agent: s-nail v14.8.16 Sender: owner-linux-mm@kvack.org Precedence: bulk
Series	[01/26] drivers/base/memory.c: don't access uninitialized memmaps in soft_offline_page_store() \| expand [01/26] drivers/base/memory.c: don't access uninitialized memmaps in soft_offline_page_store() [02/26] fs/proc/page.c: don't access uninitialized memmaps in fs/proc/page.c [03/26] mm/memory-failure.c: don't access uninitialized memmaps in memory_failure() [04/26] scripts/gdb: fix lx-dmesg when CONFIG_PRINTK_CALLER is set [05/26] mm/page_owner: don't access uninitialized memmaps when reading /proc/pagetypeinfo [06/26] mm/memory_hotplug: don't access uninitialized memmaps in shrink_pgdat_span() [07/26] mm/memunmap: don't access uninitialized memmap in memunmap_pages() [08/26] mm: memcg/slab: fix panic in __free_slab() caused by premature memcg pointer release [09/26] ocfs2: fix error handling in ocfs2_setattr() [10/26] mm/gup_benchmark: add a missing "w" to getopt string [11/26] mm/gup: fix a misnamed "write" argument, and a related bug [12/26] mm: memcg: get number of pages on the LRU list in memcgroup base on lru_zone_size [13/26] mm: memblock: do not enforce current limit for memblock_phys* family [14/26] hugetlbfs: don't access uninitialized memmaps in pfn_range_valid_gigantic() [15/26] ocfs2: fix panic due to ocfs2_wq is null [16/26] mm/memcontrol: update lruvec counters in mem_cgroup_move_account [17/26] zram: fix race between backing_dev_show and backing_dev_store [18/26] mm: include <linux/huge_mm.h> for is_vma_temporary_stack [19/26] mm/filemap.c: include <linux/ramfs.h> for generic_file_vm_ops definition [20/26] mm/init-mm.c: include <linux/mman.h> for vm_committed_as_batch [21/26] proc/meminfo: fix output alignment [22/26] mm/thp: fix node page state in split_huge_page_to_list() [23/26] mm/vmscan.c: support removing arbitrary sized pages from mapping [24/26] mm/thp: allow dropping THP from page cache [25/26] kernel/events/uprobes.c: only do FOLL_SPLIT_PMD for uprobe register [26/26] scripts/gdb: fix debugging modules on s390

Commit Message

Andrew Morton Oct. 19, 2019, 3:20 a.m. UTC

From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Subject: mm/thp: allow dropping THP from page cache

Once a THP is added to the page cache, it cannot be dropped via
/proc/sys/vm/drop_caches.  Fix this issue with proper handling in
invalidate_mapping_pages().

Link: http://lkml.kernel.org/r/20191017164223.2762148-5-songliubraving@fb.com
Fixes: 99cb0dbd47a1 ("mm,thp: add read-only THP support for (non-shmem) FS")
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
Tested-by: Song Liu <songliubraving@fb.com>
Acked-by: Yang Shi <yang.shi@linux.alibaba.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: William Kucharski <william.kucharski@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 mm/truncate.c |   12 ++++++++++++
 1 file changed, 12 insertions(+)

diff mbox series

Patch

--- a/mm/truncate.c~mm-thp-allow-drop-thp-from-page-cache
+++ a/mm/truncate.c
@@ -592,6 +592,16 @@  unsigned long invalidate_mapping_pages(s
 					unlock_page(page);
 					continue;
 				}
+
+				/* Take a pin outside pagevec */
+				get_page(page);
+
+				/*
+				 * Drop extra pins before trying to invalidate
+				 * the huge page.
+				 */
+				pagevec_remove_exceptionals(&pvec);
+				pagevec_release(&pvec);
 			}
 
 			ret = invalidate_inode_page(page);
@@ -602,6 +612,8 @@  unsigned long invalidate_mapping_pages(s
 			 */
 			if (!ret)
 				deactivate_file_page(page);
+			if (PageTransHuge(page))
+				put_page(page);
 			count += ret;
 		}
 		pagevec_remove_exceptionals(&pvec);