linux-kernel - Re: [PATCH 3/4] mm/thp: allow drop THP from page cache

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 17 Oct 2019 09:12:47 -0700
From:   Matthew Wilcox <willy@...radead.org>
To:     Song Liu <songliubraving@...com>
Cc:     linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        akpm@...ux-foundation.org, matthew.wilcox@...cle.com,
        kernel-team@...com, william.kucharski@...cle.com,
        "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>
Subject: Re: [PATCH 3/4] mm/thp: allow drop THP from page cache

On Wed, Oct 16, 2019 at 12:37:30AM -0700, Song Liu wrote:
> From: "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>
> 
> 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() and __remove_mapping().
> 
> Fixes: 99cb0dbd47a1 ("mm,thp: add read-only THP support for (non-shmem) FS")
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@...ux.intel.com>
> Tested-by: Song Liu <songliubraving@...com>
> Signed-off-by: Song Liu <songliubraving@...com>
> ---
>  mm/truncate.c | 12 ++++++++++++
>  mm/vmscan.c   |  3 ++-
>  2 files changed, 14 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index c6659bb758a4..1d80a188ad4a 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -932,7 +932,8 @@ static int __remove_mapping(struct address_space *mapping, struct page *page,
>  	 * Note that if SetPageDirty is always performed via set_page_dirty,
>  	 * and thus under the i_pages lock, then this ordering is not required.
>  	 */
> -	if (unlikely(PageTransHuge(page)) && PageSwapCache(page))
> +	if (unlikely(PageTransHuge(page)) &&
> +			(PageSwapCache(page) || !PageSwapBacked(page)))
>  		refcount = 1 + HPAGE_PMD_NR;
>  	else
>  		refcount = 2;

Kirill suggests that this patch would be better (for this part of the patch;
the part in truncate.c should remain as it is)

commit ddcee327f96d57cb9d5310486d21e43892b7a368
Author: William Kucharski <william.kucharski@...cle.com>
Date:   Fri Sep 20 16:14:51 2019 -0400

    mm: Support removing arbitrary sized pages from mapping
    
    __remove_mapping() assumes that pages can only be either base pages
    or HPAGE_PMD_SIZE.  Further, it assumes that large pages are
    swap-backed.  Support all kinds of pages by unconditionally asking how
    many pages this page references.
    
    Signed-off-by: William Kucharski <william.kucharski@...cle.com>
    Signed-off-by: Matthew Wilcox (Oracle) <willy@...radead.org>

diff --git a/mm/vmscan.c b/mm/vmscan.c
index c6659bb758a4..f870da1f4bb7 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -932,10 +932,7 @@ static int __remove_mapping(struct address_space *mapping, struct page *page,
 	 * Note that if SetPageDirty is always performed via set_page_dirty,
 	 * and thus under the i_pages lock, then this ordering is not required.
 	 */
-	if (unlikely(PageTransHuge(page)) && PageSwapCache(page))
-		refcount = 1 + HPAGE_PMD_NR;
-	else
-		refcount = 2;
+	refcount = 1 + compound_nr(page);
 	if (!page_ref_freeze(page, refcount))
 		goto cannot_free;
 	/* note: atomic_cmpxchg in page_ref_freeze provides the smp_rmb */