Commit eca3654e authored by Darrick J. Wong's avatar Darrick J. Wong Committed by Dave Chinner
Browse files

vfs: enable remap callers that can handle short operations 



Plumb in a remap flag that enables the filesystem remap handler to
shorten remapping requests for callers that can handle it.  Now
copy_file_range can report partial success (in case we run up against
alignment problems, resource limits, etc.).

We also enable CAN_SHORTEN for fideduperange to maintain existing
userspace-visible behavior where xfs/btrfs shorten the dedupe range to
avoid stale post-eof data exposure.

Signed-off-by: default avatarDarrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: default avatarAmir Goldstein <amir73il@gmail.com>
Signed-off-by: default avatarDave Chinner <david@fromorbit.com>
parent df365836

Documentation/filesystems/vfs.txt

+3 −1
Original line number Diff line number Diff line
@@ -970,7 +970,9 @@ otherwise noted. 
	negative error code if errors occurred before any bytes were remapped.

	The remap_flags parameter accepts REMAP_FILE_* flags.  If

	REMAP_FILE_DEDUP is set then the implementation must only remap if the

	requested file ranges have identical contents.

	requested file ranges have identical contents.  If REMAP_CAN_SHORTEN is

	set, the caller is ok with the implementation shortening the request

	length to satisfy alignment or EOF requirements (or any other reason).



  fadvise: possibly called by the fadvise64() system call.

fs/read_write.c

+20 −8
Original line number Diff line number Diff line
@@ -1593,7 +1593,8 @@ ssize_t vfs_copy_file_range(struct file *file_in, loff_t pos_in, 


		cloned = file_in->f_op->remap_file_range(file_in, pos_in,

				file_out, pos_out,

				min_t(loff_t, MAX_RW_COUNT, len), 0);

				min_t(loff_t, MAX_RW_COUNT, len),

				REMAP_FILE_CAN_SHORTEN);

		if (cloned > 0) {

			ret = cloned;

			goto done;
@@ -1721,6 +1722,8 @@ static int remap_verify_area(struct file *file, loff_t pos, loff_t len, 
 * can't meaningfully compare post-EOF contents.

 *

 * For clone we only link a partial EOF block above the destination file's EOF.

 *

 * Shorten the request if possible.

 */

static int generic_remap_check_len(struct inode *inode_in,

				   struct inode *inode_out,
@@ -1729,18 +1732,26 @@ static int generic_remap_check_len(struct inode *inode_in, 
				   unsigned int remap_flags)

{

	u64 blkmask = i_blocksize(inode_in) - 1;

	loff_t new_len = *len;



	if ((*len & blkmask) == 0)

		return 0;



	if (remap_flags & REMAP_FILE_DEDUP)

		*len &= ~blkmask;

	else if (pos_out + *len < i_size_read(inode_out))

		return -EINVAL;

	if ((remap_flags & REMAP_FILE_DEDUP) ||

	    pos_out + *len < i_size_read(inode_out))

		new_len &= ~blkmask;



	if (new_len == *len)

		return 0;



	if (remap_flags & REMAP_FILE_CAN_SHORTEN) {

		*len = new_len;

		return 0;

	}



	return (remap_flags & REMAP_FILE_DEDUP) ? -EBADE : -EINVAL;

}



/*

 * Check that the two inodes are eligible for cloning, the ranges make

 * sense, and then flush all dirty data.  Caller must ensure that the
@@ -2014,7 +2025,8 @@ loff_t vfs_dedupe_file_range_one(struct file *src_file, loff_t src_pos, 
{

	loff_t ret;



	WARN_ON_ONCE(remap_flags & ~(REMAP_FILE_DEDUP));

	WARN_ON_ONCE(remap_flags & ~(REMAP_FILE_DEDUP |

				     REMAP_FILE_CAN_SHORTEN));



	ret = mnt_want_write_file(dst_file);

	if (ret)
@@ -2115,7 +2127,7 @@ int vfs_dedupe_file_range(struct file *file, struct file_dedupe_range *same) 


		deduped = vfs_dedupe_file_range_one(file, off, dst_file,

						    info->dest_offset, len,

						    0);

						    REMAP_FILE_CAN_SHORTEN);

		if (deduped == -EBADE)

			info->status = FILE_DEDUPE_RANGE_DIFFERS;

		else if (deduped < 0)

include/linux/fs.h

+3 −2
Original line number Diff line number Diff line
@@ -1727,8 +1727,10 @@ struct block_device_operations; 
 * See Documentation/filesystems/vfs.txt for more details about this call.

 *

 * REMAP_FILE_DEDUP: only remap if contents identical (i.e. deduplicate)

 * REMAP_FILE_CAN_SHORTEN: caller can handle a shortened request

 */

#define REMAP_FILE_DEDUP		(1 << 0)

#define REMAP_FILE_CAN_SHORTEN		(1 << 1)



/*

 * These flags signal that the caller is ok with altering various aspects of
@@ -1736,9 +1738,8 @@ struct block_device_operations; 
 * implementation; the vfs remap helper functions can take advantage of them.

 * Flags in this category exist to preserve the quirky behavior of the hoisted

 * btrfs clone/dedupe ioctls.

 * There are no flags yet, but subsequent commits will add some.

 */

#define REMAP_FILE_ADVISORY		(0)

#define REMAP_FILE_ADVISORY		(REMAP_FILE_CAN_SHORTEN)



struct iov_iter;

mm/filemap.c

+7 −4
Original line number Diff line number Diff line
@@ -3045,8 +3045,7 @@ int generic_remap_checks(struct file *file_in, loff_t pos_in, 
		bcount = ALIGN(size_in, bs) - pos_in;

	} else {

		if (!IS_ALIGNED(count, bs))

			return -EINVAL;



			count = ALIGN_DOWN(count, bs);

		bcount = count;

	}
@@ -3056,10 +3055,14 @@ int generic_remap_checks(struct file *file_in, loff_t pos_in, 
	    pos_out < pos_in + bcount)

		return -EINVAL;



	/* For now we don't support changing the length. */

	if (*req_count != count)

	/*

	 * We shortened the request but the caller can't deal with that, so

	 * bounce the request back to userspace.

	 */

	if (*req_count != count && !(remap_flags & REMAP_FILE_CAN_SHORTEN))

		return -EINVAL;



	*req_count = count;

	return 0;

}