[RFC,01/35] vfs: clean up dedup

Commit Message

Miklos Szeredi April 12, 2018, 3:07 p.m. UTC

Extract vfs_dedupe_file_range_one() helper to deal with a single dedup
request.

Export this helper to modules, so it can be used by overlayfs to stack the
dedupe method.

Signed-off-by: Miklos Szeredi <[e-mail]>
---
 fs/read_write.c    | 90 ++++++++++++++++++++++++++++++------------------------
 include/linux/fs.h |  3 ++
 2 files changed, 53 insertions(+), 40 deletions(-)

Comments

Matthew Wilcox (Oracle) April 12, 2018, 4:25 p.m. UTC | #1

On Thu, Apr 12, 2018 at 05:07:52PM +0200, Miklos Szeredi wrote:
> +ssize_t vfs_dedupe_file_range_one(struct file *src_file, u64 src_pos, u64 len,
> +			      struct file *dst_file, u64 dst_pos)

Why u64 instead of loff_t?

Miklos Szeredi April 12, 2018, 5:24 p.m. UTC | #2

On Thu, Apr 12, 2018 at 6:25 PM, Matthew Wilcox <[e-mail]> wrote:
> On Thu, Apr 12, 2018 at 05:07:52PM +0200, Miklos Szeredi wrote:
>> +ssize_t vfs_dedupe_file_range_one(struct file *src_file, u64 src_pos, u64 len,
>> +                           struct file *dst_file, u64 dst_pos)
>
> Why u64 instead of loff_t?

Peculiarity of f_op->dedupe_file_range().

I think a cleanup of copyfile/clone/dedupe interfaces would be good.
Perhaps merge them into one?  Or just merge copyfile/clone?  Even if
not merging it should make sense to make argument types and argument
order the same.

Thanks,
Miklos

Patch

diff --git a/fs/read_write.c b/fs/read_write.c
index f8547b82dfb3..82aa0d32d0cd 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -1937,6 +1937,44 @@  int vfs_dedupe_file_range_compare(struct inode *src, loff_t srcoff,
 }
 EXPORT_SYMBOL(vfs_dedupe_file_range_compare);
 
+ssize_t vfs_dedupe_file_range_one(struct file *src_file, u64 src_pos, u64 len,
+			      struct file *dst_file, u64 dst_pos)
+{
+	ssize_t ret;
+
+	ret = mnt_want_write_file(dst_file);
+	if (ret)
+		return ret;
+
+	ret = clone_verify_area(dst_file, dst_pos, len, true);
+	if (ret < 0)
+		goto out_drop_write;
+
+	ret = -EINVAL;
+	if (!(capable(CAP_SYS_ADMIN) || (dst_file->f_mode & FMODE_WRITE)))
+		goto out_drop_write;
+
+	ret = -EXDEV;
+	if (src_file->f_path.mnt != dst_file->f_path.mnt)
+		goto out_drop_write;
+
+	ret = -EISDIR;
+	if (S_ISDIR(file_inode(dst_file)->i_mode))
+		goto out_drop_write;
+
+	ret = -EINVAL;
+	if (!dst_file->f_op->dedupe_file_range)
+		goto out_drop_write;
+
+	ret = dst_file->f_op->dedupe_file_range(src_file, src_pos, len,
+						dst_file, dst_pos);
+out_drop_write:
+	mnt_drop_write_file(dst_file);
+
+	return ret;
+}
+EXPORT_SYMBOL(vfs_dedupe_file_range_one);
+
 int vfs_dedupe_file_range(struct file *file, struct file_dedupe_range *same)
 {
 	struct file_dedupe_range_info *info;
@@ -1945,10 +1983,7 @@  int vfs_dedupe_file_range(struct file *file, struct file_dedupe_range *same)
 	u64 len;
 	int i;
 	int ret;
-	bool is_admin = capable(CAP_SYS_ADMIN);
 	u16 count = same->dest_count;
-	struct file *dst_file;
-	loff_t dst_off;
 	ssize_t deduped;
 
 	if (!(file->f_mode & FMODE_READ))
@@ -1983,54 +2018,29 @@  int vfs_dedupe_file_range(struct file *file, struct file_dedupe_range *same)
 	}
 
 	for (i = 0, info = same->info; i < count; i++, info++) {
-		struct inode *dst;
 		struct fd dst_fd = fdget(info->dest_fd);
+		struct file *dst_file = dst_fd.file;
 
-		dst_file = dst_fd.file;
 		if (!dst_file) {
 			info->status = -EBADF;
 			goto next_loop;
 		}
-		dst = file_inode(dst_file);
-
-		ret = mnt_want_write_file(dst_file);
-		if (ret) {
-			info->status = ret;
-			goto next_loop;
-		}
-
-		dst_off = info->dest_offset;
-		ret = clone_verify_area(dst_file, dst_off, len, true);
-		if (ret < 0) {
-			info->status = ret;
-			goto next_file;
-		}
-		ret = 0;
 
 		if (info->reserved) {
 			info->status = -EINVAL;
-		} else if (!(is_admin || (dst_file->f_mode & FMODE_WRITE))) {
-			info->status = -EINVAL;
-		} else if (file->f_path.mnt != dst_file->f_path.mnt) {
-			info->status = -EXDEV;
-		} else if (S_ISDIR(dst->i_mode)) {
-			info->status = -EISDIR;
-		} else if (dst_file->f_op->dedupe_file_range == NULL) {
-			info->status = -EINVAL;
-		} else {
-			deduped = dst_file->f_op->dedupe_file_range(file, off,
-							len, dst_file,
-							info->dest_offset);
-			if (deduped == -EBADE)
-				info->status = FILE_DEDUPE_RANGE_DIFFERS;
-			else if (deduped < 0)
-				info->status = deduped;
-			else
-				info->bytes_deduped += deduped;
+			goto next_loop;
 		}
 
-next_file:
-		mnt_drop_write_file(dst_file);
+		deduped = vfs_dedupe_file_range_one(file, off, len,
+						    dst_file,
+						    info->dest_offset);
+		if (deduped == -EBADE)
+			info->status = FILE_DEDUPE_RANGE_DIFFERS;
+		else if (deduped < 0)
+			info->status = deduped;
+		else
+			info->bytes_deduped += deduped;
+
 next_loop:
 		fdput(dst_fd);
 
diff --git a/include/linux/fs.h b/include/linux/fs.h
index c6baf767619e..f0f87f2beb79 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1809,6 +1809,9 @@  extern int vfs_dedupe_file_range_compare(struct inode *src, loff_t srcoff,
 					 loff_t len, bool *is_same);
 extern int vfs_dedupe_file_range(struct file *file,
 				 struct file_dedupe_range *same);
+extern ssize_t vfs_dedupe_file_range_one(struct file *src_file, u64 src_pos,
+					 u64 len, struct file *dst_file,
+					 u64 dst_pos);
 
 struct super_operations {
    	struct inode *(*alloc_inode)(struct super_block *sb);