Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To:     linux-kernel@vger.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        stable@vger.kernel.org, stable@kernel.org,
        Harshad Shirwadkar <harshadshirwadkar@gmail.com>,
        Theodore Tso <tytso@mit.edu>
Subject: [PATCH 5.10 150/157] ext4: fix rename whiteout with fast commit
Date:   Mon, 22 Mar 2021 13:28:27 +0100
Message-Id: <20210322121938.499373097@linuxfoundation.org>
In-Reply-To: <20210322121933.746237845@linuxfoundation.org>
References: <20210322121933.746237845@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Harshad Shirwadkar <harshadshirwadkar@gmail.com>

commit 8210bb29c1b66200cff7b25febcf6e39baf49fbf upstream.

This patch adds rename whiteout support in fast commits. Note that the
whiteout object that gets created is actually char device. Which
imples, the function ext4_inode_journal_mode(struct inode *inode)
would return "JOURNAL_DATA" for this inode. This has a consequence in
fast commit code that it will make creation of the whiteout object a
fast-commit ineligible behavior and thus will fall back to full
commits. With this patch, this can be observed by running fast commits
with rename whiteout and seeing the stats generated by ext4_fc_stats
tracepoint as follows:

ext4_fc_stats: dev 254:32 fc ineligible reasons:
XATTR:0, CROSS_RENAME:0, JOURNAL_FLAG_CHANGE:0, NO_MEM:0, SWAP_BOOT:0,
RESIZE:0, RENAME_DIR:0, FALLOC_RANGE:0, INODE_JOURNAL_DATA:16;
num_commits:6, ineligible: 6, numblks: 3

So in short, this patch guarantees that in case of rename whiteout, we
fall back to full commits.

Amir mentioned that instead of creating a new whiteout object for
every rename, we can create a static whiteout object with irrelevant
nlink. That will make fast commits to not fall back to full
commit. But until this happens, this patch will ensure correctness by
falling back to full commits.

Fixes: 8016e29f4362 ("ext4: fast commit recovery path")
Cc: stable@kernel.org
Signed-off-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
Link: https://lore.kernel.org/r/20210316221921.1124955-1-harshadshirwadkar@gmail.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/ext4/ext4.h        |    2 ++
 fs/ext4/fast_commit.c |    9 +++++++--
 fs/ext4/namei.c       |    3 +++
 3 files changed, 12 insertions(+), 2 deletions(-)

--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -2762,6 +2762,8 @@ void __ext4_fc_track_link(handle_t *hand
 	struct dentry *dentry);
 void ext4_fc_track_unlink(handle_t *handle, struct dentry *dentry);
 void ext4_fc_track_link(handle_t *handle, struct dentry *dentry);
+void __ext4_fc_track_create(handle_t *handle, struct inode *inode,
+			    struct dentry *dentry);
 void ext4_fc_track_create(handle_t *handle, struct dentry *dentry);
 void ext4_fc_track_inode(handle_t *handle, struct inode *inode);
 void ext4_fc_mark_ineligible(struct super_block *sb, int reason);
--- a/fs/ext4/fast_commit.c
+++ b/fs/ext4/fast_commit.c
@@ -452,10 +452,10 @@ void ext4_fc_track_link(handle_t *handle
 	__ext4_fc_track_link(handle, d_inode(dentry), dentry);
 }
 
-void ext4_fc_track_create(handle_t *handle, struct dentry *dentry)
+void __ext4_fc_track_create(handle_t *handle, struct inode *inode,
+			  struct dentry *dentry)
 {
 	struct __track_dentry_update_args args;
-	struct inode *inode = d_inode(dentry);
 	int ret;
 
 	args.dentry = dentry;
@@ -466,6 +466,11 @@ void ext4_fc_track_create(handle_t *hand
 	trace_ext4_fc_track_create(inode, dentry, ret);
 }
 
+void ext4_fc_track_create(handle_t *handle, struct dentry *dentry)
+{
+	__ext4_fc_track_create(handle, d_inode(dentry), dentry);
+}
+
 /* __track_fn for inode tracking */
 static int __track_inode(struct inode *inode, void *arg, bool update)
 {
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -3864,6 +3864,7 @@ static int ext4_rename(struct inode *old
 		retval = ext4_mark_inode_dirty(handle, whiteout);
 		if (unlikely(retval))
 			goto end_rename;
+
 	}
 	if (!new.bh) {
 		retval = ext4_add_entry(handle, new.dentry, old.inode);
@@ -3937,6 +3938,8 @@ static int ext4_rename(struct inode *old
 			ext4_fc_track_unlink(handle, new.dentry);
 		__ext4_fc_track_link(handle, old.inode, new.dentry);
 		__ext4_fc_track_unlink(handle, old.inode, old.dentry);
+		if (whiteout)
+			__ext4_fc_track_create(handle, whiteout, old.dentry);
 	}
 
 	if (new.inode) {