Linux-Fsdevel Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH] fs: Support THPs in vfs_dedupe_file_range
@ 2020-09-15 14:46 Matthew Wilcox (Oracle)
  2020-09-15 16:38 ` Darrick J. Wong
  0 siblings, 1 reply; 2+ messages in thread
From: Matthew Wilcox (Oracle) @ 2020-09-15 14:46 UTC (permalink / raw)
  To: Al Viro
  Cc: Matthew Wilcox (Oracle),
	linux-fsdevel, Darrick J . Wong, Bill O'Donnell

We may get tail pages returned from vfs_dedupe_get_page().  If we do,
we have to call page_mapping() instead of dereferencing page->mapping
directly.  We may also deadlock trying to lock the page twice if they're
subpages of the same THP, so compare the head pages instead.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/read_write.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/fs/read_write.c b/fs/read_write.c
index 5db58b8c78d0..c4d5eb47a21e 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1906,6 +1906,8 @@ static struct page *vfs_dedupe_get_page(struct inode *inode, loff_t offset)
  */
 static void vfs_lock_two_pages(struct page *page1, struct page *page2)
 {
+	page1 = thp_head(page1);
+	page2 = thp_head(page2);
 	/* Always lock in order of increasing index. */
 	if (page1->index > page2->index)
 		swap(page1, page2);
@@ -1918,6 +1920,8 @@ static void vfs_lock_two_pages(struct page *page1, struct page *page2)
 /* Unlock two pages, being careful not to unlock the same page twice. */
 static void vfs_unlock_two_pages(struct page *page1, struct page *page2)
 {
+	page1 = thp_head(page1);
+	page2 = thp_head(page2);
 	unlock_page(page1);
 	if (page1 != page2)
 		unlock_page(page2);
@@ -1972,8 +1976,8 @@ static int vfs_dedupe_file_range_compare(struct inode *src, loff_t srcoff,
 		 * someone is invalidating pages on us and we lose.
 		 */
 		if (!PageUptodate(src_page) || !PageUptodate(dest_page) ||
-		    src_page->mapping != src->i_mapping ||
-		    dest_page->mapping != dest->i_mapping) {
+		    page_mapping(src_page) != src->i_mapping ||
+		    page_mapping(dest_page) != dest->i_mapping) {
 			same = false;
 			goto unlock;
 		}
-- 
2.28.0


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] fs: Support THPs in vfs_dedupe_file_range
  2020-09-15 14:46 [PATCH] fs: Support THPs in vfs_dedupe_file_range Matthew Wilcox (Oracle)
@ 2020-09-15 16:38 ` Darrick J. Wong
  0 siblings, 0 replies; 2+ messages in thread
From: Darrick J. Wong @ 2020-09-15 16:38 UTC (permalink / raw)
  To: Matthew Wilcox (Oracle); +Cc: Al Viro, linux-fsdevel, Bill O'Donnell

On Tue, Sep 15, 2020 at 03:46:16PM +0100, Matthew Wilcox (Oracle) wrote:
> We may get tail pages returned from vfs_dedupe_get_page().  If we do,
> we have to call page_mapping() instead of dereferencing page->mapping
> directly.  We may also deadlock trying to lock the page twice if they're
> subpages of the same THP, so compare the head pages instead.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>

Seems fine to me...

Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>

--D

> ---
>  fs/read_write.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/read_write.c b/fs/read_write.c
> index 5db58b8c78d0..c4d5eb47a21e 100644
> --- a/fs/read_write.c
> +++ b/fs/read_write.c
> @@ -1906,6 +1906,8 @@ static struct page *vfs_dedupe_get_page(struct inode *inode, loff_t offset)
>   */
>  static void vfs_lock_two_pages(struct page *page1, struct page *page2)
>  {
> +	page1 = thp_head(page1);
> +	page2 = thp_head(page2);
>  	/* Always lock in order of increasing index. */
>  	if (page1->index > page2->index)
>  		swap(page1, page2);
> @@ -1918,6 +1920,8 @@ static void vfs_lock_two_pages(struct page *page1, struct page *page2)
>  /* Unlock two pages, being careful not to unlock the same page twice. */
>  static void vfs_unlock_two_pages(struct page *page1, struct page *page2)
>  {
> +	page1 = thp_head(page1);
> +	page2 = thp_head(page2);
>  	unlock_page(page1);
>  	if (page1 != page2)
>  		unlock_page(page2);
> @@ -1972,8 +1976,8 @@ static int vfs_dedupe_file_range_compare(struct inode *src, loff_t srcoff,
>  		 * someone is invalidating pages on us and we lose.
>  		 */
>  		if (!PageUptodate(src_page) || !PageUptodate(dest_page) ||
> -		    src_page->mapping != src->i_mapping ||
> -		    dest_page->mapping != dest->i_mapping) {
> +		    page_mapping(src_page) != src->i_mapping ||
> +		    page_mapping(dest_page) != dest->i_mapping) {
>  			same = false;
>  			goto unlock;
>  		}
> -- 
> 2.28.0
> 

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2020-09-15 22:14 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-09-15 14:46 [PATCH] fs: Support THPs in vfs_dedupe_file_range Matthew Wilcox (Oracle)
2020-09-15 16:38 ` Darrick J. Wong

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).