From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Fri, 12 Jan 2024 04:00:22 +0000 (-0800)
Subject: Merge tag 'pull-rename' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
X-Git-Url: http://git.maquefel.me/?a=commitdiff_plain;h=bf4e7080aeed29354cb156a8eb5d221ab2b6a8cc;p=linux.git

Merge tag 'pull-rename' of git://git./linux/kernel/git/viro/vfs

Pull rename updates from Al Viro:
 "Fix directory locking scheme on rename

  This was broken in 6.5; we really can't lock two unrelated directories
  without holding ->s_vfs_rename_mutex first and in case of same-parent
  rename of a subdirectory 6.5 ends up doing just that"

* tag 'pull-rename' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  rename(): avoid a deadlock in the case of parents having no common ancestor
  kill lock_two_inodes()
  rename(): fix the locking of subdirectories
  f2fs: Avoid reading renamed directory if parent does not change
  ext4: don't access the source subdirectory content on same-directory rename
  ext2: Avoid reading renamed directory if parent does not change
  udf_rename(): only access the child content on cross-directory rename
  ocfs2: Avoid touching renamed directory if parent does not change
  reiserfs: Avoid touching renamed directory if parent does not change
---

bf4e7080aeed29354cb156a8eb5d221ab2b6a8cc
diff --cc Documentation/filesystems/porting.rst
index ced3a67613293,33cd56e2ca1a5..c549fb2fc3ba7
--- a/Documentation/filesystems/porting.rst
+++ b/Documentation/filesystems/porting.rst
@@@ -1064,12 -1064,27 +1064,39 @@@ generic_encode_ino32_fh() explicitly
  
  ---
  
+ **mandatory**
+ 
+ If ->rename() update of .. on cross-directory move needs an exclusion with
+ directory modifications, do *not* lock the subdirectory in question in your
+ ->rename() - it's done by the caller now [that item should've been added in
+ 28eceeda130f "fs: Lock moved directories"].
+ 
+ ---
+ 
+ **mandatory**
+ 
+ On same-directory ->rename() the (tautological) update of .. is not protected
+ by any locks; just don't do it if the old parent is the same as the new one.
+ We really can't lock two subdirectories in same-directory rename - not without
+ deadlocks.
+ 
+ ---
+ 
+ **mandatory**
+ 
+ lock_rename() and lock_rename_child() may fail in cross-directory case, if
+ their arguments do not have a common ancestor.  In that case ERR_PTR(-EXDEV)
+ is returned, with no locks taken.  In-tree users updated; out-of-tree ones
+ would need to do so.
++
++---
++
 +**recommended**
 +
 +Block device freezing and thawing have been moved to holder operations.
 +
 +Before this change, get_active_super() would only be able to find the
 +superblock of the main block device, i.e., the one stored in sb->s_bdev. Block
 +device freezing now works for any block device owned by a given superblock, not
 +just the main block device. The get_active_super() helper and bd_fsfreeze_sb
 +pointer are gone.
diff --cc fs/overlayfs/copy_up.c
index 696478f09cc1b,e44dc5f661610..b8e25ca51016d
--- a/fs/overlayfs/copy_up.c
+++ b/fs/overlayfs/copy_up.c
@@@ -779,13 -757,14 +779,15 @@@ static int ovl_copy_up_workdir(struct o
  	 * lock ordering with sb_writers, which shouldn't be held when calling
  	 * ovl_copy_up_data(), so lock workdir and destdir and make sure that
  	 * temp wasn't moved before copy up completion or cleanup.
 -	 * If temp was moved, abort without the cleanup.
  	 */
  	ovl_start_write(c->dentry);
- 	if (lock_rename(c->workdir, c->destdir) != NULL ||
- 	    temp->d_parent != c->workdir) {
+ 	trap = lock_rename(c->workdir, c->destdir);
+ 	if (trap || temp->d_parent != c->workdir) {
 +		/* temp or workdir moved underneath us? abort without cleanup */
 +		dput(temp);
  		err = -EIO;
+ 		if (IS_ERR(trap))
+ 			goto out;
  		goto unlock;
  	} else if (err) {
  		goto cleanup;