[10/15] vfs: make cloning to source file eof more explicit
diff mbox series

Message ID 153870034823.29072.4817036324117498634.stgit@magnolia
State New
Headers show
Series
  • fs: fixes for serious clone/dedupe problems
Related show

Commit Message

Darrick J. Wong Oct. 5, 2018, 12:45 a.m. UTC
From: Darrick J. Wong <darrick.wong@oracle.com>

Use the new CLONE_RANGE_EOF flag to explicitly declare that the caller
wants the clone implementation to set *len appropriately once the files
are locked.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fs/read_write.c |   14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

Comments

Amir Goldstein Oct. 5, 2018, 6:47 a.m. UTC | #1
On Fri, Oct 5, 2018 at 3:46 AM Darrick J. Wong <darrick.wong@oracle.com> wrote:
>
> From: Darrick J. Wong <darrick.wong@oracle.com>
>
> Use the new CLONE_RANGE_EOF flag to explicitly declare that the caller
> wants the clone implementation to set *len appropriately once the files
> are locked.
>
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> ---
>  fs/read_write.c |   14 +++++++++-----
>  1 file changed, 9 insertions(+), 5 deletions(-)
>
>
> diff --git a/fs/read_write.c b/fs/read_write.c
> index 7cfff497263b..4eaea52f70a8 100644
> --- a/fs/read_write.c
> +++ b/fs/read_write.c
> @@ -1747,11 +1747,12 @@ int vfs_clone_file_prep(struct file *file_in, loff_t pos_in,
>
>         isize = i_size_read(inode_in);
>
> -       /* Zero length dedupe exits immediately; reflink goes to EOF. */
> -       if (*len == 0) {
> -               if (is_dedupe || pos_in == isize)
> -                       return 0;
> -               if (pos_in > isize)
> +       /*
> +        * If the caller asked to go all the way to the end of the source file,
> +        * set *len now that we have the file locked.
> +        */
> +       if ((flags & CLONERANGE_EOF) && *len == 0) {
> +               if (pos_in >= isize)
>                         return -EINVAL;

Slight change of behavior with clone of 0 bytes at EOF.
Is that intentional?
Does not change behavior through the entire series?
Please drop a work about this in commit message.

Thanks,
Amir.

Patch
diff mbox series

diff --git a/fs/read_write.c b/fs/read_write.c
index 7cfff497263b..4eaea52f70a8 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1747,11 +1747,12 @@  int vfs_clone_file_prep(struct file *file_in, loff_t pos_in,
 
 	isize = i_size_read(inode_in);
 
-	/* Zero length dedupe exits immediately; reflink goes to EOF. */
-	if (*len == 0) {
-		if (is_dedupe || pos_in == isize)
-			return 0;
-		if (pos_in > isize)
+	/*
+	 * If the caller asked to go all the way to the end of the source file,
+	 * set *len now that we have the file locked.
+	 */
+	if ((flags & CLONERANGE_EOF) && *len == 0) {
+		if (pos_in >= isize)
 			return -EINVAL;
 		*len = isize - pos_in;
 	}
@@ -1836,6 +1837,9 @@  int vfs_clone_file_range(struct file *file_in, loff_t pos_in,
 	if (ret)
 		return ret;
 
+	if (len == 0)
+		flags |= CLONERANGE_EOF;
+
 	cloned = file_in->f_op->clone_file_range(file_in, pos_in,
 			file_out, pos_out, len, flags);
 	if (cloned < 0)