From e789ca0cc1d51296832b8424fa4008ce6e9d1703 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:37 +0100
Subject: ext4: combine ext4_handle_error() and save_error_info()

save_error_info() is always called together with ext4_handle_error().
Combine them into a single call and move unconditional bits out of
save_error_info() into ext4_handle_error().

Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-2-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/super.c | 34 +++++++++++++++++-----------------
 1 file changed, 17 insertions(+), 17 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 4bbfb05aae58..cdf2a377d884 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -592,9 +592,6 @@ static void __save_error_info(struct super_block *sb, int error,
 {
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 
-	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
-	if (bdev_read_only(sb->s_bdev))
-		return;
 	/* We default to EFSCORRUPTED error... */
 	if (error == 0)
 		error = EFSCORRUPTED;
@@ -647,13 +644,19 @@ static void save_error_info(struct super_block *sb, int error,
  * used to deal with unrecoverable failures such as journal IO errors or ENOMEM
  * at a critical moment in log management.
  */
-static void ext4_handle_error(struct super_block *sb, bool force_ro)
+static void ext4_handle_error(struct super_block *sb, bool force_ro, int error,
+			      __u32 ino, __u64 block,
+			      const char *func, unsigned int line)
 {
 	journal_t *journal = EXT4_SB(sb)->s_journal;
 
+	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
 	if (test_opt(sb, WARN_ON_ERROR))
 		WARN_ON_ONCE(1);
 
+	if (!bdev_read_only(sb->s_bdev))
+		save_error_info(sb, error, ino, block, func, line);
+
 	if (sb_rdonly(sb) || (!force_ro && test_opt(sb, ERRORS_CONT)))
 		return;
 
@@ -710,8 +713,7 @@ void __ext4_error(struct super_block *sb, const char *function,
 		       sb->s_id, function, line, current->comm, &vaf);
 		va_end(args);
 	}
-	save_error_info(sb, error, 0, block, function, line);
-	ext4_handle_error(sb, force_ro);
+	ext4_handle_error(sb, force_ro, error, 0, block, function, line);
 }
 
 void __ext4_error_inode(struct inode *inode, const char *function,
@@ -741,9 +743,8 @@ void __ext4_error_inode(struct inode *inode, const char *function,
 			       current->comm, &vaf);
 		va_end(args);
 	}
-	save_error_info(inode->i_sb, error, inode->i_ino, block,
-			function, line);
-	ext4_handle_error(inode->i_sb, false);
+	ext4_handle_error(inode->i_sb, false, error, inode->i_ino, block,
+			  function, line);
 }
 
 void __ext4_error_file(struct file *file, const char *function,
@@ -780,9 +781,8 @@ void __ext4_error_file(struct file *file, const char *function,
 			       current->comm, path, &vaf);
 		va_end(args);
 	}
-	save_error_info(inode->i_sb, EFSCORRUPTED, inode->i_ino, block,
-			function, line);
-	ext4_handle_error(inode->i_sb, false);
+	ext4_handle_error(inode->i_sb, false, EFSCORRUPTED, inode->i_ino, block,
+			  function, line);
 }
 
 const char *ext4_decode_error(struct super_block *sb, int errno,
@@ -849,8 +849,7 @@ void __ext4_std_error(struct super_block *sb, const char *function,
 		       sb->s_id, function, line, errstr);
 	}
 
-	save_error_info(sb, -errno, 0, 0, function, line);
-	ext4_handle_error(sb, false);
+	ext4_handle_error(sb, false, -errno, 0, 0, function, line);
 }
 
 void __ext4_msg(struct super_block *sb,
@@ -944,13 +943,14 @@ __acquires(bitlock)
 	if (test_opt(sb, ERRORS_CONT)) {
 		if (test_opt(sb, WARN_ON_ERROR))
 			WARN_ON_ONCE(1);
+		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
 		__save_error_info(sb, EFSCORRUPTED, ino, block, function, line);
-		schedule_work(&EXT4_SB(sb)->s_error_work);
+		if (!bdev_read_only(sb->s_bdev))
+			schedule_work(&EXT4_SB(sb)->s_error_work);
 		return;
 	}
 	ext4_unlock_group(sb, grp);
-	save_error_info(sb, EFSCORRUPTED, ino, block, function, line);
-	ext4_handle_error(sb, false);
+	ext4_handle_error(sb, false, EFSCORRUPTED, ino, block, function, line);
 	/*
 	 * We only get here in the ERRORS_RO case; relocking the group
 	 * may be dangerous, but nothing bad will happen since the
-- 
cgit v1.2.3-58-ga151


From 4392fbc4bab57db3760f0fb61258cb7089b37665 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:38 +0100
Subject: ext4: drop sync argument of ext4_commit_super()

Everybody passes 1 as sync argument of ext4_commit_super(). Just drop
it.

Reviewed-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-3-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/super.c | 47 ++++++++++++++++++++++-------------------------
 1 file changed, 22 insertions(+), 25 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index cdf2a377d884..8bf31003416a 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -65,7 +65,7 @@ static struct ratelimit_state ext4_mount_msg_ratelimit;
 static int ext4_load_journal(struct super_block *, struct ext4_super_block *,
 			     unsigned long journal_devnum);
 static int ext4_show_options(struct seq_file *seq, struct dentry *root);
-static int ext4_commit_super(struct super_block *sb, int sync);
+static int ext4_commit_super(struct super_block *sb);
 static int ext4_mark_recovery_complete(struct super_block *sb,
 					struct ext4_super_block *es);
 static int ext4_clear_journal_err(struct super_block *sb,
@@ -621,7 +621,7 @@ static void save_error_info(struct super_block *sb, int error,
 {
 	__save_error_info(sb, error, ino, block, func, line);
 	if (!bdev_read_only(sb->s_bdev))
-		ext4_commit_super(sb, 1);
+		ext4_commit_super(sb);
 }
 
 /* Deal with the reporting of failure conditions on a filesystem such as
@@ -686,7 +686,7 @@ static void flush_stashed_error_work(struct work_struct *work)
 	struct ext4_sb_info *sbi = container_of(work, struct ext4_sb_info,
 						s_error_work);
 
-	ext4_commit_super(sbi->s_sb, 1);
+	ext4_commit_super(sbi->s_sb);
 }
 
 #define ext4_error_ratelimit(sb)					\
@@ -1152,7 +1152,7 @@ static void ext4_put_super(struct super_block *sb)
 		es->s_state = cpu_to_le16(sbi->s_mount_state);
 	}
 	if (!sb_rdonly(sb))
-		ext4_commit_super(sb, 1);
+		ext4_commit_super(sb);
 
 	rcu_read_lock();
 	group_desc = rcu_dereference(sbi->s_group_desc);
@@ -2642,7 +2642,7 @@ static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es,
 	if (sbi->s_journal)
 		ext4_set_feature_journal_needs_recovery(sb);
 
-	err = ext4_commit_super(sb, 1);
+	err = ext4_commit_super(sb);
 done:
 	if (test_opt(sb, DEBUG))
 		printk(KERN_INFO "[EXT4 FS bs=%lu, gc=%u, "
@@ -4869,7 +4869,7 @@ no_journal:
 	if (DUMMY_ENCRYPTION_ENABLED(sbi) && !sb_rdonly(sb) &&
 	    !ext4_has_feature_encrypt(sb)) {
 		ext4_set_feature_encrypt(sb);
-		ext4_commit_super(sb, 1);
+		ext4_commit_super(sb);
 	}
 
 	/*
@@ -5424,7 +5424,7 @@ static int ext4_load_journal(struct super_block *sb,
 		es->s_journal_dev = cpu_to_le32(journal_devnum);
 
 		/* Make sure we flush the recovery flag to disk. */
-		ext4_commit_super(sb, 1);
+		ext4_commit_super(sb);
 	}
 
 	return 0;
@@ -5434,7 +5434,7 @@ err_out:
 	return err;
 }
 
-static int ext4_commit_super(struct super_block *sb, int sync)
+static int ext4_commit_super(struct super_block *sb)
 {
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
@@ -5515,8 +5515,7 @@ static int ext4_commit_super(struct super_block *sb, int sync)
 
 	BUFFER_TRACE(sbh, "marking dirty");
 	ext4_superblock_csum_set(sb);
-	if (sync)
-		lock_buffer(sbh);
+	lock_buffer(sbh);
 	if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) {
 		/*
 		 * Oh, dear.  A previous attempt to write the
@@ -5532,16 +5531,14 @@ static int ext4_commit_super(struct super_block *sb, int sync)
 		set_buffer_uptodate(sbh);
 	}
 	mark_buffer_dirty(sbh);
-	if (sync) {
-		unlock_buffer(sbh);
-		error = __sync_dirty_buffer(sbh,
-			REQ_SYNC | (test_opt(sb, BARRIER) ? REQ_FUA : 0));
-		if (buffer_write_io_error(sbh)) {
-			ext4_msg(sb, KERN_ERR, "I/O error while writing "
-			       "superblock");
-			clear_buffer_write_io_error(sbh);
-			set_buffer_uptodate(sbh);
-		}
+	unlock_buffer(sbh);
+	error = __sync_dirty_buffer(sbh,
+		REQ_SYNC | (test_opt(sb, BARRIER) ? REQ_FUA : 0));
+	if (buffer_write_io_error(sbh)) {
+		ext4_msg(sb, KERN_ERR, "I/O error while writing "
+		       "superblock");
+		clear_buffer_write_io_error(sbh);
+		set_buffer_uptodate(sbh);
 	}
 	return error;
 }
@@ -5572,7 +5569,7 @@ static int ext4_mark_recovery_complete(struct super_block *sb,
 
 	if (ext4_has_feature_journal_needs_recovery(sb) && sb_rdonly(sb)) {
 		ext4_clear_feature_journal_needs_recovery(sb);
-		ext4_commit_super(sb, 1);
+		ext4_commit_super(sb);
 	}
 out:
 	jbd2_journal_unlock_updates(journal);
@@ -5614,7 +5611,7 @@ static int ext4_clear_journal_err(struct super_block *sb,
 
 		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
 		es->s_state |= cpu_to_le16(EXT4_ERROR_FS);
-		ext4_commit_super(sb, 1);
+		ext4_commit_super(sb);
 
 		jbd2_journal_clear_err(journal);
 		jbd2_journal_update_sb_errno(journal);
@@ -5716,7 +5713,7 @@ static int ext4_freeze(struct super_block *sb)
 		ext4_clear_feature_journal_needs_recovery(sb);
 	}
 
-	error = ext4_commit_super(sb, 1);
+	error = ext4_commit_super(sb);
 out:
 	if (journal)
 		/* we rely on upper layer to stop further updates */
@@ -5738,7 +5735,7 @@ static int ext4_unfreeze(struct super_block *sb)
 		ext4_set_feature_journal_needs_recovery(sb);
 	}
 
-	ext4_commit_super(sb, 1);
+	ext4_commit_super(sb);
 	return 0;
 }
 
@@ -5998,7 +5995,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data)
 	}
 
 	if (sbi->s_journal == NULL && !(old_sb_flags & SB_RDONLY)) {
-		err = ext4_commit_super(sb, 1);
+		err = ext4_commit_super(sb);
 		if (err)
 			goto restore_opts;
 	}
-- 
cgit v1.2.3-58-ga151


From 05c2c00f3769abb9e323fcaca70d2de0b48af7ba Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:39 +0100
Subject: ext4: protect superblock modifications with a buffer lock

Protect all superblock modifications (including checksum computation)
with a superblock buffer lock. That way we are sure computed checksum
matches current superblock contents (a mismatch could cause checksum
failures in nojournal mode or if an unjournalled superblock update races
with a journalled one). Also we avoid modifying superblock contents
while it is being written out (which can cause DIF/DIX failures if we
are running in nojournal mode).

Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-4-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/ext4_jbd2.c |  1 -
 fs/ext4/file.c      |  3 +++
 fs/ext4/inode.c     |  3 +++
 fs/ext4/namei.c     |  6 ++++++
 fs/ext4/resize.c    | 12 ++++++++++++
 fs/ext4/super.c     |  2 +-
 fs/ext4/xattr.c     |  3 +++
 7 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index 1a0a827a7f34..c7e410c4ab7d 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -379,7 +379,6 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line,
 	struct buffer_head *bh = EXT4_SB(sb)->s_sbh;
 	int err = 0;
 
-	ext4_superblock_csum_set(sb);
 	if (ext4_handle_valid(handle)) {
 		err = jbd2_journal_dirty_metadata(handle, bh);
 		if (err)
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 3ed8c048fb12..26907d5835d0 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -809,8 +809,11 @@ static int ext4_sample_last_mounted(struct super_block *sb,
 	err = ext4_journal_get_write_access(handle, sbi->s_sbh);
 	if (err)
 		goto out_journal;
+	lock_buffer(sbi->s_sbh);
 	strlcpy(sbi->s_es->s_last_mounted, cp,
 		sizeof(sbi->s_es->s_last_mounted));
+	ext4_superblock_csum_set(sb);
+	unlock_buffer(sbi->s_sbh);
 	ext4_handle_dirty_super(handle, sb);
 out_journal:
 	ext4_journal_stop(handle);
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 27946882d4ce..27de6f0a33c8 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5150,7 +5150,10 @@ static int ext4_do_update_inode(handle_t *handle,
 		err = ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh);
 		if (err)
 			goto out_brelse;
+		lock_buffer(EXT4_SB(sb)->s_sbh);
 		ext4_set_feature_large_file(sb);
+		ext4_superblock_csum_set(sb);
+		unlock_buffer(EXT4_SB(sb)->s_sbh);
 		ext4_handle_sync(handle);
 		err = ext4_handle_dirty_super(handle, sb);
 	}
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 7a890ff214f1..40970a509dcc 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -2978,7 +2978,10 @@ int ext4_orphan_add(handle_t *handle, struct inode *inode)
 	    (le32_to_cpu(sbi->s_es->s_inodes_count))) {
 		/* Insert this inode at the head of the on-disk orphan list */
 		NEXT_ORPHAN(inode) = le32_to_cpu(sbi->s_es->s_last_orphan);
+		lock_buffer(sbi->s_sbh);
 		sbi->s_es->s_last_orphan = cpu_to_le32(inode->i_ino);
+		ext4_superblock_csum_set(sb);
+		unlock_buffer(sbi->s_sbh);
 		dirty = true;
 	}
 	list_add(&EXT4_I(inode)->i_orphan, &sbi->s_orphan);
@@ -3061,7 +3064,10 @@ int ext4_orphan_del(handle_t *handle, struct inode *inode)
 			mutex_unlock(&sbi->s_orphan_lock);
 			goto out_brelse;
 		}
+		lock_buffer(sbi->s_sbh);
 		sbi->s_es->s_last_orphan = cpu_to_le32(ino_next);
+		ext4_superblock_csum_set(inode->i_sb);
+		unlock_buffer(sbi->s_sbh);
 		mutex_unlock(&sbi->s_orphan_lock);
 		err = ext4_handle_dirty_super(handle, inode->i_sb);
 	} else {
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 928700d57eb6..6155f2b9538c 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -899,7 +899,10 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	EXT4_SB(sb)->s_gdb_count++;
 	ext4_kvfree_array_rcu(o_group_desc);
 
+	lock_buffer(EXT4_SB(sb)->s_sbh);
 	le16_add_cpu(&es->s_reserved_gdt_blocks, -1);
+	ext4_superblock_csum_set(sb);
+	unlock_buffer(EXT4_SB(sb)->s_sbh);
 	err = ext4_handle_dirty_super(handle, sb);
 	if (err)
 		ext4_std_error(sb, err);
@@ -1384,6 +1387,7 @@ static void ext4_update_super(struct super_block *sb,
 	reserved_blocks *= blocks_count;
 	do_div(reserved_blocks, 100);
 
+	lock_buffer(sbi->s_sbh);
 	ext4_blocks_count_set(es, ext4_blocks_count(es) + blocks_count);
 	ext4_free_blocks_count_set(es, ext4_free_blocks_count(es) + free_blocks);
 	le32_add_cpu(&es->s_inodes_count, EXT4_INODES_PER_GROUP(sb) *
@@ -1421,6 +1425,8 @@ static void ext4_update_super(struct super_block *sb,
 	 * active. */
 	ext4_r_blocks_count_set(es, ext4_r_blocks_count(es) +
 				reserved_blocks);
+	ext4_superblock_csum_set(sb);
+	unlock_buffer(sbi->s_sbh);
 
 	/* Update the free space counts */
 	percpu_counter_add(&sbi->s_freeclusters_counter,
@@ -1717,8 +1723,11 @@ static int ext4_group_extend_no_check(struct super_block *sb,
 		goto errout;
 	}
 
+	lock_buffer(EXT4_SB(sb)->s_sbh);
 	ext4_blocks_count_set(es, o_blocks_count + add);
 	ext4_free_blocks_count_set(es, ext4_free_blocks_count(es) + add);
+	ext4_superblock_csum_set(sb);
+	unlock_buffer(EXT4_SB(sb)->s_sbh);
 	ext4_debug("freeing blocks %llu through %llu\n", o_blocks_count,
 		   o_blocks_count + add);
 	/* We add the blocks to the bitmap and set the group need init bit */
@@ -1874,10 +1883,13 @@ static int ext4_convert_meta_bg(struct super_block *sb, struct inode *inode)
 	if (err)
 		goto errout;
 
+	lock_buffer(sbi->s_sbh);
 	ext4_clear_feature_resize_inode(sb);
 	ext4_set_feature_meta_bg(sb);
 	sbi->s_es->s_first_meta_bg =
 		cpu_to_le32(num_desc_blocks(sb, sbi->s_groups_count));
+	ext4_superblock_csum_set(sb);
+	unlock_buffer(sbi->s_sbh);
 
 	err = ext4_handle_dirty_super(handle, sb);
 	if (err) {
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 8bf31003416a..2d653ad14200 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -5444,6 +5444,7 @@ static int ext4_commit_super(struct super_block *sb)
 	if (!sbh || block_device_ejected(sb))
 		return error;
 
+	lock_buffer(sbh);
 	/*
 	 * If the file system is mounted read-only, don't update the
 	 * superblock write time.  This avoids updating the superblock
@@ -5515,7 +5516,6 @@ static int ext4_commit_super(struct super_block *sb)
 
 	BUFFER_TRACE(sbh, "marking dirty");
 	ext4_superblock_csum_set(sb);
-	lock_buffer(sbh);
 	if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) {
 		/*
 		 * Oh, dear.  A previous attempt to write the
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 4e3b1f8c2e81..1db7ca778d69 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -792,7 +792,10 @@ static void ext4_xattr_update_super_block(handle_t *handle,
 
 	BUFFER_TRACE(EXT4_SB(sb)->s_sbh, "get_write_access");
 	if (ext4_journal_get_write_access(handle, EXT4_SB(sb)->s_sbh) == 0) {
+		lock_buffer(EXT4_SB(sb)->s_sbh);
 		ext4_set_feature_xattr(sb);
+		ext4_superblock_csum_set(sb);
+		unlock_buffer(EXT4_SB(sb)->s_sbh);
 		ext4_handle_dirty_super(handle, sb);
 	}
 }
-- 
cgit v1.2.3-58-ga151


From 2d01ddc86606564fb08c56e3bc93a0693895f710 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:40 +0100
Subject: ext4: save error info to sb through journal if available

If journalling is still working at the moment we get to writing error
information to the superblock we cannot write directly to the superblock
as such write could race with journalled update of the superblock and
cause journal checksum failures, writing inconsistent information to the
journal or other problems. We cannot journal the superblock directly
from the error handling functions as we are running in uncertain context
and could deadlock so just punt journalled superblock update to a
workqueue.

Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-5-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/super.c | 101 +++++++++++++++++++++++++++++++++++++++++---------------
 1 file changed, 75 insertions(+), 26 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 2d653ad14200..9a256c558e04 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -65,6 +65,7 @@ static struct ratelimit_state ext4_mount_msg_ratelimit;
 static int ext4_load_journal(struct super_block *, struct ext4_super_block *,
 			     unsigned long journal_devnum);
 static int ext4_show_options(struct seq_file *seq, struct dentry *root);
+static void ext4_update_super(struct super_block *sb);
 static int ext4_commit_super(struct super_block *sb);
 static int ext4_mark_recovery_complete(struct super_block *sb,
 					struct ext4_super_block *es);
@@ -586,9 +587,9 @@ static int ext4_errno_to_code(int errno)
 	return EXT4_ERR_UNKNOWN;
 }
 
-static void __save_error_info(struct super_block *sb, int error,
-			      __u32 ino, __u64 block,
-			      const char *func, unsigned int line)
+static void save_error_info(struct super_block *sb, int error,
+			    __u32 ino, __u64 block,
+			    const char *func, unsigned int line)
 {
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 
@@ -615,15 +616,6 @@ static void __save_error_info(struct super_block *sb, int error,
 	spin_unlock(&sbi->s_error_lock);
 }
 
-static void save_error_info(struct super_block *sb, int error,
-			    __u32 ino, __u64 block,
-			    const char *func, unsigned int line)
-{
-	__save_error_info(sb, error, ino, block, func, line);
-	if (!bdev_read_only(sb->s_bdev))
-		ext4_commit_super(sb);
-}
-
 /* Deal with the reporting of failure conditions on a filesystem such as
  * inconsistencies detected or read IO failures.
  *
@@ -649,20 +641,35 @@ static void ext4_handle_error(struct super_block *sb, bool force_ro, int error,
 			      const char *func, unsigned int line)
 {
 	journal_t *journal = EXT4_SB(sb)->s_journal;
+	bool continue_fs = !force_ro && test_opt(sb, ERRORS_CONT);
 
 	EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
 	if (test_opt(sb, WARN_ON_ERROR))
 		WARN_ON_ONCE(1);
 
-	if (!bdev_read_only(sb->s_bdev))
+	if (!continue_fs && !sb_rdonly(sb)) {
+		ext4_set_mount_flag(sb, EXT4_MF_FS_ABORTED);
+		if (journal)
+			jbd2_journal_abort(journal, -EIO);
+	}
+
+	if (!bdev_read_only(sb->s_bdev)) {
 		save_error_info(sb, error, ino, block, func, line);
+		/*
+		 * In case the fs should keep running, we need to writeout
+		 * superblock through the journal. Due to lock ordering
+		 * constraints, it may not be safe to do it right here so we
+		 * defer superblock flushing to a workqueue.
+		 */
+		if (continue_fs)
+			schedule_work(&EXT4_SB(sb)->s_error_work);
+		else
+			ext4_commit_super(sb);
+	}
 
-	if (sb_rdonly(sb) || (!force_ro && test_opt(sb, ERRORS_CONT)))
+	if (sb_rdonly(sb) || continue_fs)
 		return;
 
-	ext4_set_mount_flag(sb, EXT4_MF_FS_ABORTED);
-	if (journal)
-		jbd2_journal_abort(journal, -EIO);
 	/*
 	 * We force ERRORS_RO behavior when system is rebooting. Otherwise we
 	 * could panic during 'reboot -f' as the underlying device got already
@@ -685,7 +692,38 @@ static void flush_stashed_error_work(struct work_struct *work)
 {
 	struct ext4_sb_info *sbi = container_of(work, struct ext4_sb_info,
 						s_error_work);
+	journal_t *journal = sbi->s_journal;
+	handle_t *handle;
 
+	/*
+	 * If the journal is still running, we have to write out superblock
+	 * through the journal to avoid collisions of other journalled sb
+	 * updates.
+	 *
+	 * We use directly jbd2 functions here to avoid recursing back into
+	 * ext4 error handling code during handling of previous errors.
+	 */
+	if (!sb_rdonly(sbi->s_sb) && journal) {
+		handle = jbd2_journal_start(journal, 1);
+		if (IS_ERR(handle))
+			goto write_directly;
+		if (jbd2_journal_get_write_access(handle, sbi->s_sbh)) {
+			jbd2_journal_stop(handle);
+			goto write_directly;
+		}
+		ext4_update_super(sbi->s_sb);
+		if (jbd2_journal_dirty_metadata(handle, sbi->s_sbh)) {
+			jbd2_journal_stop(handle);
+			goto write_directly;
+		}
+		jbd2_journal_stop(handle);
+		return;
+	}
+write_directly:
+	/*
+	 * Write through journal failed. Write sb directly to get error info
+	 * out and hope for the best.
+	 */
 	ext4_commit_super(sbi->s_sb);
 }
 
@@ -944,9 +982,11 @@ __acquires(bitlock)
 		if (test_opt(sb, WARN_ON_ERROR))
 			WARN_ON_ONCE(1);
 		EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS;
-		__save_error_info(sb, EFSCORRUPTED, ino, block, function, line);
-		if (!bdev_read_only(sb->s_bdev))
+		if (!bdev_read_only(sb->s_bdev)) {
+			save_error_info(sb, EFSCORRUPTED, ino, block, function,
+					line);
 			schedule_work(&EXT4_SB(sb)->s_error_work);
+		}
 		return;
 	}
 	ext4_unlock_group(sb, grp);
@@ -5434,15 +5474,12 @@ err_out:
 	return err;
 }
 
-static int ext4_commit_super(struct super_block *sb)
+/* Copy state of EXT4_SB(sb) into buffer for on-disk superblock */
+static void ext4_update_super(struct super_block *sb)
 {
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
 	struct buffer_head *sbh = EXT4_SB(sb)->s_sbh;
-	int error = 0;
-
-	if (!sbh || block_device_ejected(sb))
-		return error;
 
 	lock_buffer(sbh);
 	/*
@@ -5514,8 +5551,20 @@ static int ext4_commit_super(struct super_block *sb)
 	}
 	spin_unlock(&sbi->s_error_lock);
 
-	BUFFER_TRACE(sbh, "marking dirty");
 	ext4_superblock_csum_set(sb);
+	unlock_buffer(sbh);
+}
+
+static int ext4_commit_super(struct super_block *sb)
+{
+	struct buffer_head *sbh = EXT4_SB(sb)->s_sbh;
+	int error = 0;
+
+	if (!sbh || block_device_ejected(sb))
+		return error;
+
+	ext4_update_super(sb);
+
 	if (buffer_write_io_error(sbh) || !buffer_uptodate(sbh)) {
 		/*
 		 * Oh, dear.  A previous attempt to write the
@@ -5530,8 +5579,8 @@ static int ext4_commit_super(struct super_block *sb)
 		clear_buffer_write_io_error(sbh);
 		set_buffer_uptodate(sbh);
 	}
+	BUFFER_TRACE(sbh, "marking dirty");
 	mark_buffer_dirty(sbh);
-	unlock_buffer(sbh);
 	error = __sync_dirty_buffer(sbh,
 		REQ_SYNC | (test_opt(sb, BARRIER) ? REQ_FUA : 0));
 	if (buffer_write_io_error(sbh)) {
-- 
cgit v1.2.3-58-ga151


From e92ad03fa53498f12b3f5ecb8822adc3bf815b28 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:41 +0100
Subject: ext4: use sbi instead of EXT4_SB(sb) in ext4_update_super()

No behavioral change.

Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-6-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/super.c | 21 ++++++++++-----------
 1 file changed, 10 insertions(+), 11 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 9a256c558e04..0f0db49031dc 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -5478,8 +5478,8 @@ err_out:
 static void ext4_update_super(struct super_block *sb)
 {
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
-	struct ext4_super_block *es = EXT4_SB(sb)->s_es;
-	struct buffer_head *sbh = EXT4_SB(sb)->s_sbh;
+	struct ext4_super_block *es = sbi->s_es;
+	struct buffer_head *sbh = sbi->s_sbh;
 
 	lock_buffer(sbh);
 	/*
@@ -5496,21 +5496,20 @@ static void ext4_update_super(struct super_block *sb)
 		ext4_update_tstamp(es, s_wtime);
 	if (sb->s_bdev->bd_part)
 		es->s_kbytes_written =
-			cpu_to_le64(EXT4_SB(sb)->s_kbytes_written +
+			cpu_to_le64(sbi->s_kbytes_written +
 			    ((part_stat_read(sb->s_bdev->bd_part,
 					     sectors[STAT_WRITE]) -
-			      EXT4_SB(sb)->s_sectors_written_start) >> 1));
+			      sbi->s_sectors_written_start) >> 1));
 	else
-		es->s_kbytes_written =
-			cpu_to_le64(EXT4_SB(sb)->s_kbytes_written);
-	if (percpu_counter_initialized(&EXT4_SB(sb)->s_freeclusters_counter))
+		es->s_kbytes_written = cpu_to_le64(sbi->s_kbytes_written);
+	if (percpu_counter_initialized(&sbi->s_freeclusters_counter))
 		ext4_free_blocks_count_set(es,
-			EXT4_C2B(EXT4_SB(sb), percpu_counter_sum_positive(
-				&EXT4_SB(sb)->s_freeclusters_counter)));
-	if (percpu_counter_initialized(&EXT4_SB(sb)->s_freeinodes_counter))
+			EXT4_C2B(sbi, percpu_counter_sum_positive(
+				&sbi->s_freeclusters_counter)));
+	if (percpu_counter_initialized(&sbi->s_freeinodes_counter))
 		es->s_free_inodes_count =
 			cpu_to_le32(percpu_counter_sum_positive(
-				&EXT4_SB(sb)->s_freeinodes_counter));
+				&sbi->s_freeinodes_counter));
 	/* Copy error information to the on-disk superblock */
 	spin_lock(&sbi->s_error_lock);
 	if (sbi->s_add_error_count > 0) {
-- 
cgit v1.2.3-58-ga151


From dfd56c2c0c0dbb11be939b804ddc8d5395ab3432 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:43 +0100
Subject: ext4: fix superblock checksum failure when setting password salt

When setting password salt in the superblock, we forget to recompute the
superblock checksum so it will not match until the next superblock
modification which recomputes the checksum. Fix it.

CC: Michael Halcrow <mhalcrow@google.com>
Reported-by: Andreas Dilger <adilger@dilger.ca>
Fixes: 9bd8212f981e ("ext4 crypto: add encryption policy and password salt support")
Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-8-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/ioctl.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index f0381876a7e5..106bf149e8ca 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -1157,7 +1157,10 @@ resizefs_out:
 			err = ext4_journal_get_write_access(handle, sbi->s_sbh);
 			if (err)
 				goto pwsalt_err_journal;
+			lock_buffer(sbi->s_sbh);
 			generate_random_uuid(sbi->s_es->s_encrypt_pw_salt);
+			ext4_superblock_csum_set(sb);
+			unlock_buffer(sbi->s_sbh);
 			err = ext4_handle_dirty_metadata(handle, NULL,
 							 sbi->s_sbh);
 		pwsalt_err_journal:
-- 
cgit v1.2.3-58-ga151


From a3f5cf14ff917d46a4d491cf86210fd639d1ff38 Mon Sep 17 00:00:00 2001
From: Jan Kara <jack@suse.cz>
Date: Wed, 16 Dec 2020 11:18:44 +0100
Subject: ext4: drop ext4_handle_dirty_super()

The wrapper is now useless since it does what
ext4_handle_dirty_metadata() does. Just remove it.

Signed-off-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201216101844.22917-9-jack@suse.cz
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/ext4_jbd2.c | 16 ----------------
 fs/ext4/ext4_jbd2.h |  5 -----
 fs/ext4/file.c      |  2 +-
 fs/ext4/inode.c     |  3 ++-
 fs/ext4/namei.c     |  4 ++--
 fs/ext4/resize.c    |  8 ++++----
 fs/ext4/xattr.c     |  2 +-
 7 files changed, 10 insertions(+), 30 deletions(-)

diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c
index c7e410c4ab7d..be799040a415 100644
--- a/fs/ext4/ext4_jbd2.c
+++ b/fs/ext4/ext4_jbd2.c
@@ -372,19 +372,3 @@ int __ext4_handle_dirty_metadata(const char *where, unsigned int line,
 	}
 	return err;
 }
-
-int __ext4_handle_dirty_super(const char *where, unsigned int line,
-			      handle_t *handle, struct super_block *sb)
-{
-	struct buffer_head *bh = EXT4_SB(sb)->s_sbh;
-	int err = 0;
-
-	if (ext4_handle_valid(handle)) {
-		err = jbd2_journal_dirty_metadata(handle, bh);
-		if (err)
-			ext4_journal_abort_handle(where, line, __func__,
-						  bh, handle, err);
-	} else
-		mark_buffer_dirty(bh);
-	return err;
-}
diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h
index a124c68b0c75..0d2fa423b7ad 100644
--- a/fs/ext4/ext4_jbd2.h
+++ b/fs/ext4/ext4_jbd2.h
@@ -244,9 +244,6 @@ int __ext4_handle_dirty_metadata(const char *where, unsigned int line,
 				 handle_t *handle, struct inode *inode,
 				 struct buffer_head *bh);
 
-int __ext4_handle_dirty_super(const char *where, unsigned int line,
-			      handle_t *handle, struct super_block *sb);
-
 #define ext4_journal_get_write_access(handle, bh) \
 	__ext4_journal_get_write_access(__func__, __LINE__, (handle), (bh))
 #define ext4_forget(handle, is_metadata, inode, bh, block_nr) \
@@ -257,8 +254,6 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line,
 #define ext4_handle_dirty_metadata(handle, inode, bh) \
 	__ext4_handle_dirty_metadata(__func__, __LINE__, (handle), (inode), \
 				     (bh))
-#define ext4_handle_dirty_super(handle, sb) \
-	__ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb))
 
 handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line,
 				  int type, int blocks, int rsv_blocks,
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 26907d5835d0..1cd3d26e3217 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -814,7 +814,7 @@ static int ext4_sample_last_mounted(struct super_block *sb,
 		sizeof(sbi->s_es->s_last_mounted));
 	ext4_superblock_csum_set(sb);
 	unlock_buffer(sbi->s_sbh);
-	ext4_handle_dirty_super(handle, sb);
+	ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
 out_journal:
 	ext4_journal_stop(handle);
 out:
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 27de6f0a33c8..c173c8405856 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5155,7 +5155,8 @@ static int ext4_do_update_inode(handle_t *handle,
 		ext4_superblock_csum_set(sb);
 		unlock_buffer(EXT4_SB(sb)->s_sbh);
 		ext4_handle_sync(handle);
-		err = ext4_handle_dirty_super(handle, sb);
+		err = ext4_handle_dirty_metadata(handle, NULL,
+						 EXT4_SB(sb)->s_sbh);
 	}
 	ext4_update_inode_fsync_trans(handle, inode, need_datasync);
 out_brelse:
diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 40970a509dcc..a3b28ef2455a 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -2988,7 +2988,7 @@ int ext4_orphan_add(handle_t *handle, struct inode *inode)
 	mutex_unlock(&sbi->s_orphan_lock);
 
 	if (dirty) {
-		err = ext4_handle_dirty_super(handle, sb);
+		err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
 		rc = ext4_mark_iloc_dirty(handle, inode, &iloc);
 		if (!err)
 			err = rc;
@@ -3069,7 +3069,7 @@ int ext4_orphan_del(handle_t *handle, struct inode *inode)
 		ext4_superblock_csum_set(inode->i_sb);
 		unlock_buffer(sbi->s_sbh);
 		mutex_unlock(&sbi->s_orphan_lock);
-		err = ext4_handle_dirty_super(handle, inode->i_sb);
+		err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
 	} else {
 		struct ext4_iloc iloc2;
 		struct inode *i_prev =
diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c
index 6155f2b9538c..bd0d185654f3 100644
--- a/fs/ext4/resize.c
+++ b/fs/ext4/resize.c
@@ -903,7 +903,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode,
 	le16_add_cpu(&es->s_reserved_gdt_blocks, -1);
 	ext4_superblock_csum_set(sb);
 	unlock_buffer(EXT4_SB(sb)->s_sbh);
-	err = ext4_handle_dirty_super(handle, sb);
+	err = ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
 	if (err)
 		ext4_std_error(sb, err);
 	return err;
@@ -1521,7 +1521,7 @@ static int ext4_flex_group_add(struct super_block *sb,
 
 	ext4_update_super(sb, flex_gd);
 
-	err = ext4_handle_dirty_super(handle, sb);
+	err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
 
 exit_journal:
 	err2 = ext4_journal_stop(handle);
@@ -1734,7 +1734,7 @@ static int ext4_group_extend_no_check(struct super_block *sb,
 	err = ext4_group_add_blocks(handle, sb, o_blocks_count, add);
 	if (err)
 		goto errout;
-	ext4_handle_dirty_super(handle, sb);
+	ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
 	ext4_debug("freed blocks %llu through %llu\n", o_blocks_count,
 		   o_blocks_count + add);
 errout:
@@ -1891,7 +1891,7 @@ static int ext4_convert_meta_bg(struct super_block *sb, struct inode *inode)
 	ext4_superblock_csum_set(sb);
 	unlock_buffer(sbi->s_sbh);
 
-	err = ext4_handle_dirty_super(handle, sb);
+	err = ext4_handle_dirty_metadata(handle, NULL, sbi->s_sbh);
 	if (err) {
 		ext4_std_error(sb, err);
 		goto errout;
diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index 1db7ca778d69..372208500f4e 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -796,7 +796,7 @@ static void ext4_xattr_update_super_block(handle_t *handle,
 		ext4_set_feature_xattr(sb);
 		ext4_superblock_csum_set(sb);
 		unlock_buffer(EXT4_SB(sb)->s_sbh);
-		ext4_handle_dirty_super(handle, sb);
+		ext4_handle_dirty_metadata(handle, NULL, EXT4_SB(sb)->s_sbh);
 	}
 }
 
-- 
cgit v1.2.3-58-ga151


From 5a3b590d4b2db187faa6f06adc9a53d6199fb1f9 Mon Sep 17 00:00:00 2001
From: Theodore Ts'o <tytso@mit.edu>
Date: Thu, 17 Dec 2020 13:24:15 -0500
Subject: ext4: don't leak old mountpoint samples

When the first file is opened, ext4 samples the mountpoint of the
filesystem in 64 bytes of the super block.  It does so using
strlcpy(), this means that the remaining bytes in the super block
string buffer are untouched.  If the mount point before had a longer
path than the current one, it can be reconstructed.

Consider the case where the fs was mounted to "/media/johnjdeveloper"
and later to "/".  The super block buffer then contains
"/\x00edia/johnjdeveloper".

This case was seen in the wild and caused confusion how the name
of a developer ands up on the super block of a filesystem used
in production...

Fix this by using strncpy() instead of strlcpy().  The superblock
field is defined to be a fixed-size char array, and it is already
marked using __nonstring in fs/ext4/ext4.h.  The consumer of the field
in e2fsprogs already assumes that in the case of a 64+ byte mount
path, that s_last_mounted will not be NUL terminated.

Link: https://lore.kernel.org/r/X9ujIOJG/HqMr88R@mit.edu
Reported-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
---
 fs/ext4/file.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 1cd3d26e3217..349b27f0dda0 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -810,7 +810,7 @@ static int ext4_sample_last_mounted(struct super_block *sb,
 	if (err)
 		goto out_journal;
 	lock_buffer(sbi->s_sbh);
-	strlcpy(sbi->s_es->s_last_mounted, cp,
+	strncpy(sbi->s_es->s_last_mounted, cp,
 		sizeof(sbi->s_es->s_last_mounted));
 	ext4_superblock_csum_set(sb);
 	unlock_buffer(sbi->s_sbh);
-- 
cgit v1.2.3-58-ga151


From 23dd561ad9eae02b4d51bb502fe4e1a0666e9567 Mon Sep 17 00:00:00 2001
From: Yi Li <yili@winhong.com>
Date: Wed, 30 Dec 2020 11:38:27 +0800
Subject: ext4: use IS_ERR instead of IS_ERR_OR_NULL and set inode null when
 IS_ERR

1: ext4_iget/ext4_find_extent never returns NULL, use IS_ERR
instead of IS_ERR_OR_NULL to fix this.

2: ext4_fc_replay_inode should set the inode to NULL when IS_ERR.
and go to call iput properly.

Fixes: 8016e29f4362 ("ext4: fast commit recovery path")
Signed-off-by: Yi Li <yili@winhong.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Link: https://lore.kernel.org/r/20201230033827.3996064-1-yili@winhong.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
---
 fs/ext4/fast_commit.c | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/fs/ext4/fast_commit.c b/fs/ext4/fast_commit.c
index 4fcc21c25e79..6b5489273c85 100644
--- a/fs/ext4/fast_commit.c
+++ b/fs/ext4/fast_commit.c
@@ -1318,14 +1318,14 @@ static int ext4_fc_replay_unlink(struct super_block *sb, struct ext4_fc_tl *tl)
 	entry.len = darg.dname_len;
 	inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL);
 
-	if (IS_ERR_OR_NULL(inode)) {
+	if (IS_ERR(inode)) {
 		jbd_debug(1, "Inode %d not found", darg.ino);
 		return 0;
 	}
 
 	old_parent = ext4_iget(sb, darg.parent_ino,
 				EXT4_IGET_NORMAL);
-	if (IS_ERR_OR_NULL(old_parent)) {
+	if (IS_ERR(old_parent)) {
 		jbd_debug(1, "Dir with inode  %d not found", darg.parent_ino);
 		iput(inode);
 		return 0;
@@ -1410,7 +1410,7 @@ static int ext4_fc_replay_link(struct super_block *sb, struct ext4_fc_tl *tl)
 			darg.parent_ino, darg.dname_len);
 
 	inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL);
-	if (IS_ERR_OR_NULL(inode)) {
+	if (IS_ERR(inode)) {
 		jbd_debug(1, "Inode not found.");
 		return 0;
 	}
@@ -1466,10 +1466,11 @@ static int ext4_fc_replay_inode(struct super_block *sb, struct ext4_fc_tl *tl)
 	trace_ext4_fc_replay(sb, tag, ino, 0, 0);
 
 	inode = ext4_iget(sb, ino, EXT4_IGET_NORMAL);
-	if (!IS_ERR_OR_NULL(inode)) {
+	if (!IS_ERR(inode)) {
 		ext4_ext_clear_bb(inode);
 		iput(inode);
 	}
+	inode = NULL;
 
 	ext4_fc_record_modified_inode(sb, ino);
 
@@ -1512,7 +1513,7 @@ static int ext4_fc_replay_inode(struct super_block *sb, struct ext4_fc_tl *tl)
 
 	/* Given that we just wrote the inode on disk, this SHOULD succeed. */
 	inode = ext4_iget(sb, ino, EXT4_IGET_NORMAL);
-	if (IS_ERR_OR_NULL(inode)) {
+	if (IS_ERR(inode)) {
 		jbd_debug(1, "Inode not found.");
 		return -EFSCORRUPTED;
 	}
@@ -1564,7 +1565,7 @@ static int ext4_fc_replay_create(struct super_block *sb, struct ext4_fc_tl *tl)
 		goto out;
 
 	inode = ext4_iget(sb, darg.ino, EXT4_IGET_NORMAL);
-	if (IS_ERR_OR_NULL(inode)) {
+	if (IS_ERR(inode)) {
 		jbd_debug(1, "inode %d not found.", darg.ino);
 		inode = NULL;
 		ret = -EINVAL;
@@ -1577,7 +1578,7 @@ static int ext4_fc_replay_create(struct super_block *sb, struct ext4_fc_tl *tl)
 		 * dot and dot dot dirents are setup properly.
 		 */
 		dir = ext4_iget(sb, darg.parent_ino, EXT4_IGET_NORMAL);
-		if (IS_ERR_OR_NULL(dir)) {
+		if (IS_ERR(dir)) {
 			jbd_debug(1, "Dir %d not found.", darg.ino);
 			goto out;
 		}
@@ -1653,7 +1654,7 @@ static int ext4_fc_replay_add_range(struct super_block *sb,
 
 	inode = ext4_iget(sb, le32_to_cpu(fc_add_ex->fc_ino),
 				EXT4_IGET_NORMAL);
-	if (IS_ERR_OR_NULL(inode)) {
+	if (IS_ERR(inode)) {
 		jbd_debug(1, "Inode not found.");
 		return 0;
 	}
@@ -1777,7 +1778,7 @@ ext4_fc_replay_del_range(struct super_block *sb, struct ext4_fc_tl *tl)
 		le32_to_cpu(lrange->fc_ino), cur, remaining);
 
 	inode = ext4_iget(sb, le32_to_cpu(lrange->fc_ino), EXT4_IGET_NORMAL);
-	if (IS_ERR_OR_NULL(inode)) {
+	if (IS_ERR(inode)) {
 		jbd_debug(1, "Inode %d not found", le32_to_cpu(lrange->fc_ino));
 		return 0;
 	}
@@ -1832,7 +1833,7 @@ static void ext4_fc_set_bitmaps_and_counters(struct super_block *sb)
 	for (i = 0; i < state->fc_modified_inodes_used; i++) {
 		inode = ext4_iget(sb, state->fc_modified_inodes[i],
 			EXT4_IGET_NORMAL);
-		if (IS_ERR_OR_NULL(inode)) {
+		if (IS_ERR(inode)) {
 			jbd_debug(1, "Inode %d not found.",
 				state->fc_modified_inodes[i]);
 			continue;
@@ -1849,7 +1850,7 @@ static void ext4_fc_set_bitmaps_and_counters(struct super_block *sb)
 
 			if (ret > 0) {
 				path = ext4_find_extent(inode, map.m_lblk, NULL, 0);
-				if (!IS_ERR_OR_NULL(path)) {
+				if (!IS_ERR(path)) {
 					for (j = 0; j < path->p_depth; j++)
 						ext4_mb_mark_bb(inode->i_sb,
 							path[j].p_block, 1, 1);
-- 
cgit v1.2.3-58-ga151


From 31e203e09f036f48e7c567c2d32df0196bbd303f Mon Sep 17 00:00:00 2001
From: Daejun Park <daejun7.park@samsung.com>
Date: Wed, 30 Dec 2020 18:48:51 +0900
Subject: ext4: fix wrong list_splice in ext4_fc_cleanup

After full/fast commit, entries in staging queue are promoted to main
queue. In ext4_fs_cleanup function, it splice to staging queue to
staging queue.

Fixes: aa75f4d3daaeb ("ext4: main fast-commit commit path")
Signed-off-by: Daejun Park <daejun7.park@samsung.com>
Reviewed-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
Link: https://lore.kernel.org/r/20201230094851epcms2p6eeead8cc984379b37b2efd21af90fd1a@epcms2p6
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
---
 fs/ext4/fast_commit.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ext4/fast_commit.c b/fs/ext4/fast_commit.c
index 6b5489273c85..ec5316fbcd1d 100644
--- a/fs/ext4/fast_commit.c
+++ b/fs/ext4/fast_commit.c
@@ -1268,7 +1268,7 @@ static void ext4_fc_cleanup(journal_t *journal, int full)
 	list_splice_init(&sbi->s_fc_dentry_q[FC_Q_STAGING],
 				&sbi->s_fc_dentry_q[FC_Q_MAIN]);
 	list_splice_init(&sbi->s_fc_q[FC_Q_STAGING],
-				&sbi->s_fc_q[FC_Q_STAGING]);
+				&sbi->s_fc_q[FC_Q_MAIN]);
 
 	ext4_clear_mount_flag(sb, EXT4_MF_FC_COMMITTING);
 	ext4_clear_mount_flag(sb, EXT4_MF_FC_INELIGIBLE);
-- 
cgit v1.2.3-58-ga151


From 6b4b8e6b4ad8553660421d6360678b3811d5deb9 Mon Sep 17 00:00:00 2001
From: yangerkun <yangerkun@huawei.com>
Date: Tue, 5 Jan 2021 14:28:57 +0800
Subject: ext4: fix bug for rename with RENAME_WHITEOUT

We got a "deleted inode referenced" warning cross our fsstress test. The
bug can be reproduced easily with following steps:

  cd /dev/shm
  mkdir test/
  fallocate -l 128M img
  mkfs.ext4 -b 1024 img
  mount img test/
  dd if=/dev/zero of=test/foo bs=1M count=128
  mkdir test/dir/ && cd test/dir/
  for ((i=0;i<1000;i++)); do touch file$i; done # consume all block
  cd ~ && renameat2(AT_FDCWD, /dev/shm/test/dir/file1, AT_FDCWD,
    /dev/shm/test/dir/dst_file, RENAME_WHITEOUT) # ext4_add_entry in
    ext4_rename will return ENOSPC!!
  cd /dev/shm/ && umount test/ && mount img test/ && ls -li test/dir/file1
  We will get the output:
  "ls: cannot access 'test/dir/file1': Structure needs cleaning"
  and the dmesg show:
  "EXT4-fs error (device loop0): ext4_lookup:1626: inode #2049: comm ls:
  deleted inode referenced: 139"

ext4_rename will create a special inode for whiteout and use this 'ino'
to replace the source file's dir entry 'ino'. Once error happens
latter(the error above was the ENOSPC return from ext4_add_entry in
ext4_rename since all space has been consumed), the cleanup do drop the
nlink for whiteout, but forget to restore 'ino' with source file. This
will trigger the bug describle as above.

Signed-off-by: yangerkun <yangerkun@huawei.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: stable@vger.kernel.org
Fixes: cd808deced43 ("ext4: support RENAME_WHITEOUT")
Link: https://lore.kernel.org/r/20210105062857.3566-1-yangerkun@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/namei.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index a3b28ef2455a..fa625a247e9a 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -3601,9 +3601,6 @@ static int ext4_setent(handle_t *handle, struct ext4_renament *ent,
 			return retval2;
 		}
 	}
-	brelse(ent->bh);
-	ent->bh = NULL;
-
 	return retval;
 }
 
@@ -3802,6 +3799,7 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 		}
 	}
 
+	old_file_type = old.de->file_type;
 	if (IS_DIRSYNC(old.dir) || IS_DIRSYNC(new.dir))
 		ext4_handle_sync(handle);
 
@@ -3829,7 +3827,6 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 	force_reread = (new.dir->i_ino == old.dir->i_ino &&
 			ext4_test_inode_flag(new.dir, EXT4_INODE_INLINE_DATA));
 
-	old_file_type = old.de->file_type;
 	if (whiteout) {
 		/*
 		 * Do this before adding a new entry, so the old entry is sure
@@ -3927,15 +3924,19 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry,
 	retval = 0;
 
 end_rename:
-	brelse(old.dir_bh);
-	brelse(old.bh);
-	brelse(new.bh);
 	if (whiteout) {
-		if (retval)
+		if (retval) {
+			ext4_setent(handle, &old,
+				old.inode->i_ino, old_file_type);
 			drop_nlink(whiteout);
+		}
 		unlock_new_inode(whiteout);
 		iput(whiteout);
+
 	}
+	brelse(old.dir_bh);
+	brelse(old.bh);
+	brelse(new.bh);
 	if (handle)
 		ext4_journal_stop(handle);
 	return retval;
-- 
cgit v1.2.3-58-ga151


From e9f53353e166a67dfe4f8295100f8ac39d6cf10b Mon Sep 17 00:00:00 2001
From: Daejun Park <daejun7.park@samsung.com>
Date: Wed, 6 Jan 2021 10:32:42 +0900
Subject: ext4: remove expensive flush on fast commit

In the fast commit, it adds REQ_FUA and REQ_PREFLUSH on each fast
commit block when barrier is enabled.  However, in recovery phase,
ext4 compares CRC value in the tail.  So it is sufficient to add
REQ_FUA and REQ_PREFLUSH on the block that has tail.

Signed-off-by: Daejun Park <daejun7.park@samsung.com>
Reviewed-by: Harshad Shirwadkar <harshadshirwadkar@gmail.com>
Link: https://lore.kernel.org/r/20210106013242epcms2p5b6b4ed8ca86f29456fdf56aa580e74b4@epcms2p5
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/fast_commit.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/fs/ext4/fast_commit.c b/fs/ext4/fast_commit.c
index ec5316fbcd1d..0a14a7c87bf8 100644
--- a/fs/ext4/fast_commit.c
+++ b/fs/ext4/fast_commit.c
@@ -604,13 +604,13 @@ void ext4_fc_track_range(handle_t *handle, struct inode *inode, ext4_lblk_t star
 	trace_ext4_fc_track_range(inode, start, end, ret);
 }
 
-static void ext4_fc_submit_bh(struct super_block *sb)
+static void ext4_fc_submit_bh(struct super_block *sb, bool is_tail)
 {
 	int write_flags = REQ_SYNC;
 	struct buffer_head *bh = EXT4_SB(sb)->s_fc_bh;
 
-	/* TODO: REQ_FUA | REQ_PREFLUSH is unnecessarily expensive. */
-	if (test_opt(sb, BARRIER))
+	/* Add REQ_FUA | REQ_PREFLUSH only its tail */
+	if (test_opt(sb, BARRIER) && is_tail)
 		write_flags |= REQ_FUA | REQ_PREFLUSH;
 	lock_buffer(bh);
 	set_buffer_dirty(bh);
@@ -684,7 +684,7 @@ static u8 *ext4_fc_reserve_space(struct super_block *sb, int len, u32 *crc)
 		*crc = ext4_chksum(sbi, *crc, tl, sizeof(*tl));
 	if (pad_len > 0)
 		ext4_fc_memzero(sb, tl + 1, pad_len, crc);
-	ext4_fc_submit_bh(sb);
+	ext4_fc_submit_bh(sb, false);
 
 	ret = jbd2_fc_get_buf(EXT4_SB(sb)->s_journal, &bh);
 	if (ret)
@@ -741,7 +741,7 @@ static int ext4_fc_write_tail(struct super_block *sb, u32 crc)
 	tail.fc_crc = cpu_to_le32(crc);
 	ext4_fc_memcpy(sb, dst, &tail.fc_crc, sizeof(tail.fc_crc), NULL);
 
-	ext4_fc_submit_bh(sb);
+	ext4_fc_submit_bh(sb, true);
 
 	return 0;
 }
-- 
cgit v1.2.3-58-ga151